- video-ports: 24 bits value which defines how the video controller
output is wired to the TDA998x input - default: <0x230145>
+ - audio-ports: array of 8-bit values, 2 values per one DAI[1].
+ The first value defines the DAI type: TDA998x_SPDIF or TDA998x_I2S[2].
+ The second value defines the tda998x AP_ENA reg content when the DAI
+ in question is used. The implementation allows one or two DAIs. If two
+ DAIs are defined, they must be of different type.
+
+[1] Documentation/sound/alsa/soc/DAI.txt
+[2] include/dt-bindings/display/tda998x.h
+
Example:
+#include <dt-bindings/display/tda998x.h>
+
tda998x: hdmi-encoder {
compatible = "nxp,tda998x";
reg = <0x70>;
interrupts = <27 2>; /* falling edge */
pinctrl-0 = <&pmx_camera>;
pinctrl-names = "default";
+ video-ports = <0x230145>;
+
+ #sound-dai-cells = <2>;
+ /* DAI-format AP_ENA reg value */
+ audio-ports = < TDA998x_SPDIF 0x04
+ TDA998x_I2S 0x03>;
+
};
Required properties:
- compatible: value should be one of the following
- "rockchip,rk3288-vop";
"rockchip,rk3036-vop";
+ "rockchip,rk3288-vop";
+ "rockchip,rk3399-vop-big";
+ "rockchip,rk3399-vop-lit";
- interrupts: should contain a list of all VOP IP block interrupts in the
order: VSYNC, LCD_SYSTEM. The interrupt specifier
The TCON acts as a timing controller for RGB, LVDS and TV interfaces.
Required properties:
- - compatible: value should be "allwinner,sun5i-a13-tcon".
+ - compatible: value must be either:
+ * allwinner,sun5i-a13-tcon
+ * allwinner,sun8i-a33-tcon
- reg: base address and size of memory-mapped region
- interrupts: interrupt associated to this IP
- clocks: phandles to the clocks feeding the TCON. Three are needed:
- 'ahb': the interface clocks
- 'tcon-ch0': The clock driving the TCON channel 0
- - 'tcon-ch1': The clock driving the TCON channel 1
- resets: phandles to the reset controllers driving the encoder
- "lcd": the reset line for the TCON channel 0
second the block connected to the TCON channel 1 (usually the TV
encoder)
+On the A13, there is one more clock required:
+ - 'tcon-ch1': The clock driving the TCON channel 1
+
+DRC
+---
+
+The DRC (Dynamic Range Controller), found in the latest Allwinner SoCs
+(A31, A23, A33), allows to dynamically adjust pixel
+brightness/contrast based on histogram measurements for LCD content
+adaptive backlight control.
+
+
+Required properties:
+ - compatible: value must be one of:
+ * allwinner,sun8i-a33-drc
+ - reg: base address and size of the memory-mapped region.
+ - interrupts: interrupt associated to this IP
+ - clocks: phandles to the clocks feeding the DRC
+ * ahb: the DRC interface clock
+ * mod: the DRC module clock
+ * ram: the DRC DRAM clock
+ - clock-names: the clock names mentioned above
+ - resets: phandles to the reset line driving the DRC
+
+- ports: A ports node with endpoint definitions as defined in
+ Documentation/devicetree/bindings/media/video-interfaces.txt. The
+ first port should be the input endpoints, the second one the outputs
Display Engine Backend
----------------------
Required properties:
- compatible: value must be one of:
* allwinner,sun5i-a13-display-backend
+ * allwinner,sun8i-a33-display-backend
- reg: base address and size of the memory-mapped region.
- clocks: phandles to the clocks feeding the frontend and backend
* ahb: the backend interface clock
Documentation/devicetree/bindings/media/video-interfaces.txt. The
first port should be the input endpoints, the second one the output
+On the A33, some additional properties are required:
+ - reg needs to have an additional region corresponding to the SAT
+ - reg-names need to be set, with "be" and "sat"
+ - clocks and clock-names need to have a phandle to the SAT bus
+ clocks, whose name will be "sat"
+ - resets and reset-names need to have a phandle to the SAT bus
+ resets, whose name will be "sat"
+
Display Engine Frontend
-----------------------
Required properties:
- compatible: value must be one of:
* allwinner,sun5i-a13-display-frontend
+ * allwinner,sun8i-a33-display-frontend
- reg: base address and size of the memory-mapped region.
- interrupts: interrupt associated to this IP
- clocks: phandles to the clocks feeding the frontend and backend
Required properties:
- compatible: value must be one of:
* allwinner,sun5i-a13-display-engine
+ * allwinner,sun8i-a33-display-engine
- allwinner,pipelines: list of phandle to the display engine
frontends available.
the lcd controller.
- max-pixelclock: The maximum pixel clock that can be supported
by the lcd controller in KHz.
+ - blue-and-red-wiring: Recognized values "default", "straight" or
+ "crossed". This property deals with the LCDC revision 2 (found on
+ AM335x) color errata [1].
+ - "straight" indicates normal wiring that supports RGB565,
+ BGR888, and XBGR8888 color formats.
+ - "crossed" indicates wiring that has blue and red wires
+ crossed. This setup supports BGR565, RGB888 and XRGB8888
+ formats.
+ - If the property is not present or its value is not recognized
+ the legacy mode is assumed. This configuration supports RGB565,
+ RGB888 and XRGB8888 formats. However, depending on wiring, the red
+ and blue colors are swapped in either 16 or 24-bit color modes.
Optional nodes:
Documentation/devicetree/bindings/display/tilcdc/tfp410.txt for connecting
tfp410 DVI encoder or lcd panel to lcdc
+[1] There is an errata about AM335x color wiring. For 16-bit color mode
+ the wires work as they should (LCD_DATA[0:4] is for Blue[3:7]),
+ but for 24 bit color modes the wiring of blue and red components is
+ crossed and LCD_DATA[0:4] is for Red[3:7] and LCD_DATA[11:15] is
+ for Blue[3-7]. For more details see section 3.1.1 in AM335x
+ Silicon Errata:
+ http://www.ti.com/general/docs/lit/getliterature.tsp?baseLiteratureNumber=sprz360
+
Example:
fb: fb@4830e000 {
interrupts = <36>;
ti,hwmods = "lcdc";
+ blue-and-red-wiring = "crossed";
+
port {
lcdc_0: endpoint@0 {
remote-endpoint = <&hdmi_0>;
- struct drm_mode_config_funcs \*funcs;
Mode setting functions.
+Modeset Base Object Abstraction
+===============================
+
+.. kernel-doc:: include/drm/drm_mode_object.h
+ :internal:
+
+.. kernel-doc:: drivers/gpu/drm/drm_mode_object.c
+ :export:
+
KMS Data Structures
===================
.. kernel-doc:: drivers/gpu/drm/drm_connector.c
:export:
+Encoder Abstraction
+===================
+
+.. kernel-doc:: drivers/gpu/drm/drm_encoder.c
+ :doc: overview
+
+Encoder Functions Reference
+---------------------------
+
+.. kernel-doc:: include/drm/drm_encoder.h
+ :internal:
+
+.. kernel-doc:: drivers/gpu/drm/drm_encoder.c
+ :export:
+
KMS Initialization and Cleanup
==============================
primary planes may make use of the helper functions described in ? to
create and register a primary plane with standard capabilities.
-Encoders (:c:type:`struct drm_encoder <drm_encoder>`)
------------------------------------------------------
-
-An encoder takes pixel data from a CRTC and converts it to a format
-suitable for any attached connectors. On some devices, it may be
-possible to have a CRTC send data to more than one encoder. In that
-case, both encoders would receive data from the same scanout buffer,
-resulting in a "cloned" display configuration across the connectors
-attached to each encoder.
-
-Encoder Initialization
-~~~~~~~~~~~~~~~~~~~~~~
-
-As for CRTCs, a KMS driver must create, initialize and register at least
-one :c:type:`struct drm_encoder <drm_encoder>` instance. The
-instance is allocated and zeroed by the driver, possibly as part of a
-larger structure.
-
-Drivers must initialize the :c:type:`struct drm_encoder
-<drm_encoder>` possible_crtcs and possible_clones fields before
-registering the encoder. Both fields are bitmasks of respectively the
-CRTCs that the encoder can be connected to, and sibling encoders
-candidate for cloning.
-
-After being initialized, the encoder must be registered with a call to
-:c:func:`drm_encoder_init()`. The function takes a pointer to the
-encoder functions and an encoder type. Supported types are
-
-- DRM_MODE_ENCODER_DAC for VGA and analog on DVI-I/DVI-A
-- DRM_MODE_ENCODER_TMDS for DVI, HDMI and (embedded) DisplayPort
-- DRM_MODE_ENCODER_LVDS for display panels
-- DRM_MODE_ENCODER_TVDAC for TV output (Composite, S-Video,
- Component, SCART)
-- DRM_MODE_ENCODER_VIRTUAL for virtual machine displays
-
-Encoders must be attached to a CRTC to be used. DRM drivers leave
-encoders unattached at initialization time. Applications (or the fbdev
-compatibility layer when implemented) are responsible for attaching the
-encoders they want to use to a CRTC.
-
Cleanup
-------
KMS Properties
==============
-Drivers may need to expose additional parameters to applications than
-those described in the previous sections. KMS supports attaching
-properties to CRTCs, connectors and planes and offers a userspace API to
-list, get and set the property values.
-
-Properties are identified by a name that uniquely defines the property
-purpose, and store an associated value. For all property types except
-blob properties the value is a 64-bit unsigned integer.
-
-KMS differentiates between properties and property instances. Drivers
-first create properties and then create and associate individual
-instances of those properties to objects. A property can be instantiated
-multiple times and associated with different objects. Values are stored
-in property instances, and all other property information are stored in
-the property and shared between all instances of the property.
-
-Every property is created with a type that influences how the KMS core
-handles the property. Supported property types are
-
-DRM_MODE_PROP_RANGE
- Range properties report their minimum and maximum admissible values.
- The KMS core verifies that values set by application fit in that
- range.
-
-DRM_MODE_PROP_ENUM
- Enumerated properties take a numerical value that ranges from 0 to
- the number of enumerated values defined by the property minus one,
- and associate a free-formed string name to each value. Applications
- can retrieve the list of defined value-name pairs and use the
- numerical value to get and set property instance values.
-
-DRM_MODE_PROP_BITMASK
- Bitmask properties are enumeration properties that additionally
- restrict all enumerated values to the 0..63 range. Bitmask property
- instance values combine one or more of the enumerated bits defined
- by the property.
-
-DRM_MODE_PROP_BLOB
- Blob properties store a binary blob without any format restriction.
- The binary blobs are created as KMS standalone objects, and blob
- property instance values store the ID of their associated blob
- object.
-
- Blob properties are only used for the connector EDID property and
- cannot be created by drivers.
-
-To create a property drivers call one of the following functions
-depending on the property type. All property creation functions take
-property flags and name, as well as type-specific arguments.
-
-- struct drm_property \*drm_property_create_range(struct
- drm_device \*dev, int flags, const char \*name, uint64_t min,
- uint64_t max);
- Create a range property with the given minimum and maximum values.
-
-- struct drm_property \*drm_property_create_enum(struct drm_device
- \*dev, int flags, const char \*name, const struct
- drm_prop_enum_list \*props, int num_values);
- Create an enumerated property. The ``props`` argument points to an
- array of ``num_values`` value-name pairs.
-
-- struct drm_property \*drm_property_create_bitmask(struct
- drm_device \*dev, int flags, const char \*name, const struct
- drm_prop_enum_list \*props, int num_values);
- Create a bitmask property. The ``props`` argument points to an array
- of ``num_values`` value-name pairs.
-
-Properties can additionally be created as immutable, in which case they
-will be read-only for applications but can be modified by the driver. To
-create an immutable property drivers must set the
-DRM_MODE_PROP_IMMUTABLE flag at property creation time.
-
-When no array of value-name pairs is readily available at property
-creation time for enumerated or range properties, drivers can create the
-property using the :c:func:`drm_property_create()` function and
-manually add enumeration value-name pairs by calling the
-:c:func:`drm_property_add_enum()` function. Care must be taken to
-properly specify the property type through the ``flags`` argument.
-
-After creating properties drivers can attach property instances to CRTC,
-connector and plane objects by calling the
-:c:func:`drm_object_attach_property()`. The function takes a
-pointer to the target object, a pointer to the previously created
-property and an initial instance value.
+Property Types and Blob Property Support
+----------------------------------------
+
+.. kernel-doc:: drivers/gpu/drm/drm_property.c
+ :doc: overview
+
+.. kernel-doc:: include/drm/drm_property.h
+ :internal:
+
+.. kernel-doc:: drivers/gpu/drm/drm_property.c
+ :export:
Blending and Z-Position properties
----------------------------------
visible to user-space and accessible beyond open-file boundaries, they
cannot support render nodes.
+Validating changes with IGT
+===========================
+
+There's a collection of tests that aims to cover the whole functionality of
+DRM drivers and that can be used to check that changes to DRM drivers or the
+core don't regress existing functionality. This test suite is called IGT and
+its code can be found in https://cgit.freedesktop.org/drm/igt-gpu-tools/.
+
+To build IGT, start by installing its build dependencies. In Debian-based
+systems::
+
+ # apt-get build-dep intel-gpu-tools
+
+And in Fedora-based systems::
+
+ # dnf builddep intel-gpu-tools
+
+Then clone the repository::
+
+ $ git clone git://anongit.freedesktop.org/drm/igt-gpu-tools
+
+Configure the build system and start the build::
+
+ $ cd igt-gpu-tools && ./autogen.sh && make -j6
+
+Download the piglit dependency::
+
+ $ ./scripts/run-tests.sh -d
+
+And run the tests::
+
+ $ ./scripts/run-tests.sh -t kms -t core -s
+
+run-tests.sh is a wrapper around piglit that will execute the tests matching
+the -t options. A report in HTML format will be available in
+./results/html/index.html. Results can be compared with piglit.
+
VBlank event handling
=====================
#include "am33xx.dtsi"
#include "am335x-bone-common.dtsi"
+#include <dt-bindings/display/tda998x.h>
/ {
model = "TI AM335x BeagleBone Black";
AM33XX_IOPAD(0x9b0, PIN_OUTPUT_PULLDOWN | MUX_MODE3) /* xdma_event_intr0 */
>;
};
+
+ mcasp0_pins: mcasp0_pins {
+ pinctrl-single,pins = <
+ AM33XX_IOPAD(0x9ac, PIN_INPUT_PULLUP | MUX_MODE0) /* mcasp0_ahcklx.mcasp0_ahclkx */
+ AM33XX_IOPAD(0x99c, PIN_OUTPUT_PULLDOWN | MUX_MODE2) /* mcasp0_ahclkr.mcasp0_axr2*/
+ AM33XX_IOPAD(0x994, PIN_OUTPUT_PULLUP | MUX_MODE0) /* mcasp0_fsx.mcasp0_fsx */
+ AM33XX_IOPAD(0x990, PIN_OUTPUT_PULLDOWN | MUX_MODE0) /* mcasp0_aclkx.mcasp0_aclkx */
+ AM33XX_IOPAD(0x86c, PIN_OUTPUT_PULLDOWN | MUX_MODE7) /* gpmc_a11.GPIO1_27 */
+ >;
+ };
};
&lcdc {
};
&i2c0 {
- tda19988 {
+ tda19988: tda19988 {
compatible = "nxp,tda998x";
reg = <0x70>;
+
pinctrl-names = "default", "off";
pinctrl-0 = <&nxp_hdmi_bonelt_pins>;
pinctrl-1 = <&nxp_hdmi_bonelt_off_pins>;
- port {
- hdmi_0: endpoint@0 {
- remote-endpoint = <&lcdc_0>;
+ #sound-dai-cells = <0>;
+ audio-ports = < TDA998x_I2S 0x03>;
+
+ ports {
+ port@0 {
+ hdmi_0: endpoint@0 {
+ remote-endpoint = <&lcdc_0>;
+ };
};
};
};
&rtc {
system-power-controller;
};
+
+&mcasp0 {
+ #sound-dai-cells = <0>;
+ pinctrl-names = "default";
+ pinctrl-0 = <&mcasp0_pins>;
+ status = "okay";
+ op-mode = <0>; /* MCASP_IIS_MODE */
+ tdm-slots = <2>;
+ serial-dir = < /* 0: INACTIVE, 1: TX, 2: RX */
+ 0 0 1 0
+ >;
+ tx-num-evt = <32>;
+ rx-num-evt = <32>;
+};
+
+/ {
+ clk_mcasp0_fixed: clk_mcasp0_fixed {
+ #clock-cells = <0>;
+ compatible = "fixed-clock";
+ clock-frequency = <24576000>;
+ };
+
+ clk_mcasp0: clk_mcasp0 {
+ #clock-cells = <0>;
+ compatible = "gpio-gate-clock";
+ clocks = <&clk_mcasp0_fixed>;
+ enable-gpios = <&gpio1 27 0>; /* BeagleBone Black Clk enable on GPIO1_27 */
+ };
+
+ sound {
+ compatible = "simple-audio-card";
+ simple-audio-card,name = "TI BeagleBone Black";
+ simple-audio-card,format = "i2s";
+ simple-audio-card,bitclock-master = <&dailink0_master>;
+ simple-audio-card,frame-master = <&dailink0_master>;
+
+ dailink0_master: simple-audio-card,cpu {
+ sound-dai = <&mcasp0>;
+ clocks = <&clk_mcasp0>;
+ };
+
+ simple-audio-card,codec {
+ sound-dai = <&tda19988>;
+ };
+ };
+};
static phys_addr_t __init i865_stolen_base(int num, int slot, int func,
size_t stolen_size)
{
- u16 toud;
+ u16 toud = 0;
- /*
- * FIXME is the graphics stolen memory region
- * always at TOUD? Ie. is it always the last
- * one to be allocated by the BIOS?
- */
toud = read_pci_config_16(0, 0, 0, I865_TOUD);
- return (phys_addr_t)toud << 16;
+ return (phys_addr_t)(toud << 16) + i845_tseg_size();
}
static phys_addr_t __init gen3_stolen_base(int num, int slot, int func,
INTEL_I915GM_IDS(&gen3_early_ops),
INTEL_I945G_IDS(&gen3_early_ops),
INTEL_I945GM_IDS(&gen3_early_ops),
- INTEL_VLV_M_IDS(&gen6_early_ops),
- INTEL_VLV_D_IDS(&gen6_early_ops),
+ INTEL_VLV_IDS(&gen6_early_ops),
INTEL_PINEVIEW_IDS(&gen3_early_ops),
INTEL_I965G_IDS(&gen3_early_ops),
INTEL_G33_IDS(&gen3_early_ops),
INTEL_SNB_M_IDS(&gen6_early_ops),
INTEL_IVB_M_IDS(&gen6_early_ops),
INTEL_IVB_D_IDS(&gen6_early_ops),
- INTEL_HSW_D_IDS(&gen6_early_ops),
- INTEL_HSW_M_IDS(&gen6_early_ops),
- INTEL_BDW_M_IDS(&gen8_early_ops),
- INTEL_BDW_D_IDS(&gen8_early_ops),
+ INTEL_HSW_IDS(&gen6_early_ops),
+ INTEL_BDW_IDS(&gen8_early_ops),
INTEL_CHV_IDS(&chv_early_ops),
INTEL_SKL_IDS(&gen9_early_ops),
INTEL_BXT_IDS(&gen9_early_ops),
unsigned int flags)
{
intel_private.driver->write_entry(addr, pg, flags);
+ if (intel_private.driver->chipset_flush)
+ intel_private.driver->chipset_flush();
}
EXPORT_SYMBOL(intel_gtt_insert_page);
poll_wait(file, &sync_file->wq, wait);
- if (!test_and_set_bit(POLL_ENABLED, &sync_file->fence->flags)) {
+ if (!poll_does_not_wait(wait) &&
+ !test_and_set_bit(POLL_ENABLED, &sync_file->fence->flags)) {
if (fence_add_callback(sync_file->fence, &sync_file->cb,
fence_check_cb_func) < 0)
wake_up_all(&sync_file->wq);
If M is selected, the module will be called amdgpu.
source "drivers/gpu/drm/amd/amdgpu/Kconfig"
-source "drivers/gpu/drm/amd/powerplay/Kconfig"
-
-source "drivers/gpu/drm/amd/acp/Kconfig"
source "drivers/gpu/drm/nouveau/Kconfig"
menuconfig DRM_LEGACY
bool "Enable legacy drivers (DANGEROUS)"
depends on DRM
- depends on BROKEN
help
Enable legacy DRI1 drivers. Those drivers expose unsafe and dangerous
APIs to user-space, which can be used to circumvent access
drm_trace_points.o drm_global.o drm_prime.o \
drm_rect.o drm_vma_manager.o drm_flip_work.o \
drm_modeset_lock.o drm_atomic.o drm_bridge.o \
- drm_framebuffer.o drm_connector.o drm_blend.o
+ drm_framebuffer.o drm_connector.o drm_blend.o \
+ drm_encoder.o drm_mode_object.o drm_property.o
drm-$(CONFIG_COMPAT) += drm_ioc32.o
drm-$(CONFIG_DRM_GEM_CMA_HELPER) += drm_gem_cma_helper.o
obj-$(CONFIG_DRM_AMDGPU)+= amd/amdgpu/
obj-$(CONFIG_DRM_MGA) += mga/
obj-$(CONFIG_DRM_I810) += i810/
-obj-$(CONFIG_DRM_I915) += i915/
+obj-$(CONFIG_DRM_I915) += i915/
obj-$(CONFIG_DRM_MGAG200) += mgag200/
obj-$(CONFIG_DRM_VC4) += vc4/
obj-$(CONFIG_DRM_CIRRUS_QEMU) += cirrus/
+config DRM_AMDGPU_SI
+ bool "Enable amdgpu support for SI parts"
+ depends on DRM_AMDGPU
+ help
+ Choose this option if you want to enable experimental support
+ for SI asics.
+
config DRM_AMDGPU_CIK
bool "Enable amdgpu support for CIK parts"
depends on DRM_AMDGPU
Selecting this option creates a debugfs file to inspect the mapped
pages. Uses more memory for housekeeping, enable only for debugging.
+source "drivers/gpu/drm/amd/powerplay/Kconfig"
+source "drivers/gpu/drm/amd/acp/Kconfig"
ci_smc.o ci_dpm.o dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o \
amdgpu_amdkfd_gfx_v7.o
+amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o si_dpm.o si_smc.o
+
amdgpu-y += \
vi.o
# add DCE block
amdgpu-y += \
dce_v10_0.o \
- dce_v11_0.o
+ dce_v11_0.o \
+ dce_virtual.o
# add GFX block
amdgpu-y += \
#define ENCODER_OBJECT_ID_INTERNAL_VCE 0x24
#define ENCODER_OBJECT_ID_INTERNAL_UNIPHY3 0x25
#define ENCODER_OBJECT_ID_INTERNAL_AMCLK 0x27
+#define ENCODER_OBJECT_ID_VIRTUAL 0x28
#define ENCODER_OBJECT_ID_GENERAL_EXTERNAL_DVO 0xFF
#define CONNECTOR_OBJECT_ID_eDP 0x14
#define CONNECTOR_OBJECT_ID_MXM 0x15
#define CONNECTOR_OBJECT_ID_LVDS_eDP 0x16
+#define CONNECTOR_OBJECT_ID_VIRTUAL 0x17
/* deleted */
#define GRAPH_OBJECT_ENUM_ID5 0x05
#define GRAPH_OBJECT_ENUM_ID6 0x06
#define GRAPH_OBJECT_ENUM_ID7 0x07
+#define GRAPH_OBJECT_ENUM_VIRTUAL 0x08
/****************************************************/
/* Graphics Object ID Bit definition */
GRAPH_OBJECT_ENUM_ID1 << ENUM_ID_SHIFT |\
ENCODER_OBJECT_ID_HDMI_ANX9805 << OBJECT_ID_SHIFT)
+#define ENCODER_VIRTUAL_ENUM_VIRTUAL ( GRAPH_OBJECT_TYPE_ENCODER << OBJECT_TYPE_SHIFT |\
+ GRAPH_OBJECT_ENUM_VIRTUAL << ENUM_ID_SHIFT |\
+ ENCODER_OBJECT_ID_VIRTUAL << OBJECT_ID_SHIFT)
+
/****************************************************/
/* Connector Object ID definition - Shared with BIOS */
/****************************************************/
#include "amdgpu_ih.h"
#include "amdgpu_irq.h"
#include "amdgpu_ucode.h"
+#include "amdgpu_ttm.h"
#include "amdgpu_gds.h"
#include "amd_powerplay.h"
#include "amdgpu_acp.h"
extern int amdgpu_modeset;
extern int amdgpu_vram_limit;
extern int amdgpu_gart_size;
+extern int amdgpu_moverate;
extern int amdgpu_benchmarking;
extern int amdgpu_testing;
extern int amdgpu_audio;
extern unsigned amdgpu_cg_mask;
extern unsigned amdgpu_pg_mask;
extern char *amdgpu_disable_cu;
+extern int amdgpu_sclk_deep_sleep_en;
+extern char *amdgpu_virtual_display;
+extern unsigned amdgpu_pp_feature_mask;
#define AMDGPU_WAIT_IDLE_TIMEOUT_IN_MS 3000
#define AMDGPU_MAX_USEC_TIMEOUT 100000 /* 100 ms */
#define AMDGPU_MAX_RINGS 16
#define AMDGPU_MAX_GFX_RINGS 1
#define AMDGPU_MAX_COMPUTE_RINGS 8
-#define AMDGPU_MAX_VCE_RINGS 2
+#define AMDGPU_MAX_VCE_RINGS 3
/* max number of IP instances */
#define AMDGPU_MAX_SDMA_INSTANCES 2
uint64_t pe, uint64_t src,
unsigned count);
/* write pte one entry at a time with addr mapping */
- void (*write_pte)(struct amdgpu_ib *ib,
- const dma_addr_t *pages_addr, uint64_t pe,
- uint64_t addr, unsigned count,
- uint32_t incr, uint32_t flags);
+ void (*write_pte)(struct amdgpu_ib *ib, uint64_t pe,
+ uint64_t value, unsigned count,
+ uint32_t incr);
/* for linear pte/pde updates without addr mapping */
void (*set_pte_pde)(struct amdgpu_ib *ib,
uint64_t pe,
/* note usage for clock and power gating */
void (*begin_use)(struct amdgpu_ring *ring);
void (*end_use)(struct amdgpu_ring *ring);
+ void (*emit_switch_buffer) (struct amdgpu_ring *ring);
+ void (*emit_cntxcntl) (struct amdgpu_ring *ring, uint32_t flags);
+ unsigned (*get_emit_ib_size) (struct amdgpu_ring *ring);
+ unsigned (*get_dma_frame_size) (struct amdgpu_ring *ring);
};
/*
unsigned amdgpu_fence_count_emitted(struct amdgpu_ring *ring);
/*
- * TTM.
+ * BO.
*/
-#define AMDGPU_TTM_LRU_SIZE 20
-
-struct amdgpu_mman_lru {
- struct list_head *lru[TTM_NUM_MEM_TYPES];
- struct list_head *swap_lru;
-};
-
-struct amdgpu_mman {
- struct ttm_bo_global_ref bo_global_ref;
- struct drm_global_reference mem_global_ref;
- struct ttm_bo_device bdev;
- bool mem_global_referenced;
- bool initialized;
-
-#if defined(CONFIG_DEBUG_FS)
- struct dentry *vram;
- struct dentry *gtt;
-#endif
-
- /* buffer handling */
- const struct amdgpu_buffer_funcs *buffer_funcs;
- struct amdgpu_ring *buffer_funcs_ring;
- /* Scheduler entity for buffer moves */
- struct amd_sched_entity entity;
-
- /* custom LRU management */
- struct amdgpu_mman_lru log2_size[AMDGPU_TTM_LRU_SIZE];
-};
-
-int amdgpu_copy_buffer(struct amdgpu_ring *ring,
- uint64_t src_offset,
- uint64_t dst_offset,
- uint32_t byte_count,
- struct reservation_object *resv,
- struct fence **fence);
-int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma);
-
struct amdgpu_bo_list_entry {
struct amdgpu_bo *robj;
struct ttm_validate_buffer tv;
struct amdgpu_device *adev;
struct drm_gem_object gem_base;
struct amdgpu_bo *parent;
+ struct amdgpu_bo *shadow;
struct ttm_bo_kmap_obj dma_buf_vmap;
struct amdgpu_mn *mn;
struct list_head mn_list;
+ struct list_head shadow_list;
};
#define gem_to_amdgpu_bo(gobj) container_of((gobj), struct amdgpu_bo, gem_base)
int amdgpu_gart_bind(struct amdgpu_device *adev, unsigned offset,
int pages, struct page **pagelist,
dma_addr_t *dma_addr, uint32_t flags);
+int amdgpu_ttm_recover_gart(struct amdgpu_device *adev);
/*
* GPU MC structures, functions & helpers
uint32_t fw_version;
struct amdgpu_irq_src vm_fault;
uint32_t vram_type;
+ uint32_t srbm_soft_reset;
+ struct amdgpu_mode_mc_save save;
};
/*
*/
struct amdgpu_flip_work {
- struct work_struct flip_work;
+ struct delayed_work flip_work;
struct work_struct unpin_work;
struct amdgpu_device *adev;
int crtc_id;
+ u32 target_vblank;
uint64_t base;
struct drm_pending_vblank_event *event;
struct amdgpu_bo *old_rbo;
/* maximum number of VMIDs */
#define AMDGPU_NUM_VM 16
+/* Maximum number of PTEs the hardware can write with one command */
+#define AMDGPU_VM_MAX_UPDATE_SIZE 0x3FFFF
+
/* number of entries in page table */
#define AMDGPU_VM_PTE_COUNT (1 << amdgpu_vm_block_size)
/* PTBs (Page Table Blocks) need to be aligned to 32K */
#define AMDGPU_VM_PTB_ALIGN_SIZE 32768
-#define AMDGPU_VM_PTB_ALIGN_MASK (AMDGPU_VM_PTB_ALIGN_SIZE - 1)
-#define AMDGPU_VM_PTB_ALIGN(a) (((a) + AMDGPU_VM_PTB_ALIGN_MASK) & ~AMDGPU_VM_PTB_ALIGN_MASK)
+
+/* LOG2 number of continuous pages for the fragment field */
+#define AMDGPU_LOG2_PAGES_PER_FRAG 4
#define AMDGPU_PTE_VALID (1 << 0)
#define AMDGPU_PTE_SYSTEM (1 << 1)
#define AMDGPU_PTE_READABLE (1 << 5)
#define AMDGPU_PTE_WRITEABLE (1 << 6)
-/* PTE (Page Table Entry) fragment field for different page sizes */
-#define AMDGPU_PTE_FRAG_4KB (0 << 7)
-#define AMDGPU_PTE_FRAG_64KB (4 << 7)
-#define AMDGPU_LOG2_PAGES_PER_FRAG 4
+#define AMDGPU_PTE_FRAG(x) ((x & 0x1f) << 7)
/* How to programm VM fault handling */
#define AMDGPU_VM_FAULT_STOP_NEVER 0
struct amdgpu_vm_pt {
struct amdgpu_bo_list_entry entry;
uint64_t addr;
+ uint64_t shadow_addr;
};
struct amdgpu_vm {
struct amdgpu_job *job);
int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job);
void amdgpu_vm_reset_id(struct amdgpu_device *adev, unsigned vm_id);
-uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr);
int amdgpu_vm_update_page_directory(struct amdgpu_device *adev,
struct amdgpu_vm *vm);
int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
struct amdgpu_sync *sync);
int amdgpu_vm_bo_update(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
- struct ttm_mem_reg *mem);
+ bool clear);
void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
struct amdgpu_bo *bo);
struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm,
spinlock_t ring_lock;
struct fence **fences;
struct amdgpu_ctx_ring rings[AMDGPU_MAX_RINGS];
+ bool preamble_presented;
};
struct amdgpu_ctx_mgr {
unsigned ce_ram_size;
struct amdgpu_cu_info cu_info;
const struct amdgpu_gfx_funcs *funcs;
+
+ /* reset mask */
+ uint32_t grbm_soft_reset;
+ uint32_t srbm_soft_reset;
};
int amdgpu_ib_get(struct amdgpu_device *adev, struct amdgpu_vm *vm,
struct fence *fence;
uint64_t bytes_moved_threshold;
uint64_t bytes_moved;
+ struct amdgpu_bo_list_entry *evictable;
/* user fence */
struct amdgpu_bo_list_entry uf_entry;
};
+#define AMDGPU_PREAMBLE_IB_PRESENT (1 << 0) /* bit set means command submit involves a preamble IB */
+#define AMDGPU_PREAMBLE_IB_PRESENT_FIRST (1 << 1) /* bit set means preamble IB is first presented in belonging context */
+#define AMDGPU_HAVE_CTX_SWITCH (1 << 2) /* bit set means context switch occured */
+
struct amdgpu_job {
struct amd_sched_job base;
struct amdgpu_device *adev;
struct amdgpu_sync sync;
struct amdgpu_ib *ibs;
struct fence *fence; /* the hw fence */
+ uint32_t preamble_status;
uint32_t num_ibs;
void *owner;
- uint64_t ctx;
+ uint64_t fence_ctx; /* the fence_context this job uses */
bool vm_needs_flush;
unsigned vm_id;
uint64_t vm_pd_addr;
bool address_64_bit;
bool use_ctx_buf;
struct amd_sched_entity entity;
+ uint32_t srbm_soft_reset;
};
/*
struct amdgpu_irq_src irq;
unsigned harvest_config;
struct amd_sched_entity entity;
+ uint32_t srbm_soft_reset;
+ unsigned num_rings;
};
/*
struct amdgpu_sdma {
struct amdgpu_sdma_instance instance[AMDGPU_MAX_SDMA_INSTANCES];
+#ifdef CONFIG_DRM_AMDGPU_SI
+ //SI DMA has a difference trap irq number for the second engine
+ struct amdgpu_irq_src trap_irq_1;
+#endif
struct amdgpu_irq_src trap_irq;
struct amdgpu_irq_src illegal_inst_irq;
int num_instances;
+ uint32_t srbm_soft_reset;
};
/*
int (*set_vce_clocks)(struct amdgpu_device *adev, u32 evclk, u32 ecclk);
/* query virtual capabilities */
u32 (*get_virtual_caps)(struct amdgpu_device *adev);
+ /* static power management */
+ int (*get_pcie_lanes)(struct amdgpu_device *adev);
+ void (*set_pcie_lanes)(struct amdgpu_device *adev, int lanes);
};
/*
bool valid;
bool sw;
bool hw;
+ bool hang;
};
struct amdgpu_device {
spinlock_t pcie_idx_lock;
amdgpu_rreg_t pcie_rreg;
amdgpu_wreg_t pcie_wreg;
+ amdgpu_rreg_t pciep_rreg;
+ amdgpu_wreg_t pciep_wreg;
/* protects concurrent UVD register access */
spinlock_t uvd_ctx_idx_lock;
amdgpu_rreg_t uvd_ctx_rreg;
atomic64_t num_evictions;
atomic_t gpu_reset_counter;
+ /* data for buffer migration throttling */
+ struct {
+ spinlock_t lock;
+ s64 last_update_us;
+ s64 accum_us; /* accumulated microseconds */
+ u32 log2_max_MBps;
+ } mm_stats;
+
/* display */
+ bool enable_virtual_display;
struct amdgpu_mode_info mode_info;
struct work_struct hotplug_work;
struct amdgpu_irq_src crtc_irq;
struct kfd_dev *kfd;
struct amdgpu_virtualization virtualization;
+
+ /* link all shadow bo */
+ struct list_head shadow_list;
+ struct mutex shadow_list_lock;
+ /* link all gtt */
+ spinlock_t gtt_list_lock;
+ struct list_head gtt_list;
+
};
bool amdgpu_device_is_px(struct drm_device *dev);
#define REG_GET(FIELD, v) (((v) << FIELD##_SHIFT) & FIELD##_MASK)
#define RREG32_PCIE(reg) adev->pcie_rreg(adev, (reg))
#define WREG32_PCIE(reg, v) adev->pcie_wreg(adev, (reg), (v))
+#define RREG32_PCIE_PORT(reg) adev->pciep_rreg(adev, (reg))
+#define WREG32_PCIE_PORT(reg, v) adev->pciep_wreg(adev, (reg), (v))
#define RREG32_SMC(reg) adev->smc_rreg(adev, (reg))
#define WREG32_SMC(reg, v) adev->smc_wreg(adev, (reg), (v))
#define RREG32_UVD_CTX(reg) adev->uvd_ctx_rreg(adev, (reg))
#define REG_GET_FIELD(value, reg, field) \
(((value) & REG_FIELD_MASK(reg, field)) >> REG_FIELD_SHIFT(reg, field))
+#define WREG32_FIELD(reg, field, val) \
+ WREG32(mm##reg, (RREG32(mm##reg) & ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field))
+
/*
* BIOS helpers.
*/
#define amdgpu_asic_set_uvd_clocks(adev, v, d) (adev)->asic_funcs->set_uvd_clocks((adev), (v), (d))
#define amdgpu_asic_set_vce_clocks(adev, ev, ec) (adev)->asic_funcs->set_vce_clocks((adev), (ev), (ec))
#define amdgpu_asic_get_virtual_caps(adev) ((adev)->asic_funcs->get_virtual_caps((adev)))
+#define amdgpu_get_pcie_lanes(adev) (adev)->asic_funcs->get_pcie_lanes((adev))
+#define amdgpu_set_pcie_lanes(adev, l) (adev)->asic_funcs->set_pcie_lanes((adev), (l))
+#define amdgpu_asic_get_gpu_clock_counter(adev) (adev)->asic_funcs->get_gpu_clock_counter((adev))
#define amdgpu_asic_read_disabled_bios(adev) (adev)->asic_funcs->read_disabled_bios((adev))
#define amdgpu_asic_read_bios_from_rom(adev, b, l) (adev)->asic_funcs->read_bios_from_rom((adev), (b), (l))
#define amdgpu_asic_read_register(adev, se, sh, offset, v)((adev)->asic_funcs->read_register((adev), (se), (sh), (offset), (v)))
#define amdgpu_gart_flush_gpu_tlb(adev, vmid) (adev)->gart.gart_funcs->flush_gpu_tlb((adev), (vmid))
#define amdgpu_gart_set_pte_pde(adev, pt, idx, addr, flags) (adev)->gart.gart_funcs->set_pte_pde((adev), (pt), (idx), (addr), (flags))
#define amdgpu_vm_copy_pte(adev, ib, pe, src, count) ((adev)->vm_manager.vm_pte_funcs->copy_pte((ib), (pe), (src), (count)))
-#define amdgpu_vm_write_pte(adev, ib, pa, pe, addr, count, incr, flags) ((adev)->vm_manager.vm_pte_funcs->write_pte((ib), (pa), (pe), (addr), (count), (incr), (flags)))
+#define amdgpu_vm_write_pte(adev, ib, pe, value, count, incr) ((adev)->vm_manager.vm_pte_funcs->write_pte((ib), (pe), (value), (count), (incr)))
#define amdgpu_vm_set_pte_pde(adev, ib, pe, addr, count, incr, flags) ((adev)->vm_manager.vm_pte_funcs->set_pte_pde((ib), (pe), (addr), (count), (incr), (flags)))
#define amdgpu_ring_parse_cs(r, p, ib) ((r)->funcs->parse_cs((p), (ib)))
#define amdgpu_ring_test_ring(r) (r)->funcs->test_ring((r))
#define amdgpu_ring_emit_gds_switch(r, v, db, ds, wb, ws, ab, as) (r)->funcs->emit_gds_switch((r), (v), (db), (ds), (wb), (ws), (ab), (as))
#define amdgpu_ring_emit_hdp_flush(r) (r)->funcs->emit_hdp_flush((r))
#define amdgpu_ring_emit_hdp_invalidate(r) (r)->funcs->emit_hdp_invalidate((r))
+#define amdgpu_ring_emit_switch_buffer(r) (r)->funcs->emit_switch_buffer((r))
+#define amdgpu_ring_emit_cntxcntl(r, d) (r)->funcs->emit_cntxcntl((r), (d))
#define amdgpu_ring_pad_ib(r, ib) ((r)->funcs->pad_ib((r), (ib)))
#define amdgpu_ring_init_cond_exec(r) (r)->funcs->init_cond_exec((r))
#define amdgpu_ring_patch_cond_exec(r,o) (r)->funcs->patch_cond_exec((r),(o))
+#define amdgpu_ring_get_emit_ib_size(r) (r)->funcs->get_emit_ib_size((r))
+#define amdgpu_ring_get_dma_frame_size(r) (r)->funcs->get_dma_frame_size((r))
#define amdgpu_ih_get_wptr(adev) (adev)->irq.ih_funcs->get_wptr((adev))
#define amdgpu_ih_decode_iv(adev, iv) (adev)->irq.ih_funcs->decode_iv((adev), (iv))
#define amdgpu_ih_set_rptr(adev) (adev)->irq.ih_funcs->set_rptr((adev))
/* Common functions */
int amdgpu_gpu_reset(struct amdgpu_device *adev);
+bool amdgpu_need_backup(struct amdgpu_device *adev);
void amdgpu_pci_config_reset(struct amdgpu_device *adev);
bool amdgpu_card_posted(struct amdgpu_device *adev);
void amdgpu_update_display_priority(struct amdgpu_device *adev);
void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64 base);
void amdgpu_gtt_location(struct amdgpu_device *adev, struct amdgpu_mc *mc);
void amdgpu_ttm_set_active_vram_size(struct amdgpu_device *adev, u64 size);
+u64 amdgpu_ttm_get_gtt_mem_size(struct amdgpu_device *adev);
+int amdgpu_ttm_global_init(struct amdgpu_device *adev);
+int amdgpu_ttm_init(struct amdgpu_device *adev);
+void amdgpu_ttm_fini(struct amdgpu_device *adev);
void amdgpu_program_register_sequence(struct amdgpu_device *adev,
const u32 *registers,
const u32 array_size);
struct drm_file *file_priv);
void amdgpu_driver_preclose_kms(struct drm_device *dev,
struct drm_file *file_priv);
-int amdgpu_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon);
-int amdgpu_resume_kms(struct drm_device *dev, bool resume, bool fbcon);
+int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon);
+int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon);
u32 amdgpu_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe);
int amdgpu_enable_vblank_kms(struct drm_device *dev, unsigned int pipe);
void amdgpu_disable_vblank_kms(struct drm_device *dev, unsigned int pipe);
struct amdgpu_bo_va_mapping *
amdgpu_cs_find_mapping(struct amdgpu_cs_parser *parser,
uint64_t addr, struct amdgpu_bo **bo);
+int amdgpu_cs_sysvm_access_required(struct amdgpu_cs_parser *parser);
#include "amdgpu_object.h"
#endif
DRM_MODE_CONNECTOR_Unknown
};
+bool amdgpu_atombios_has_dce_engine_info(struct amdgpu_device *adev)
+{
+ struct amdgpu_mode_info *mode_info = &adev->mode_info;
+ struct atom_context *ctx = mode_info->atom_context;
+ int index = GetIndexIntoMasterTable(DATA, Object_Header);
+ u16 size, data_offset;
+ u8 frev, crev;
+ ATOM_DISPLAY_OBJECT_PATH_TABLE *path_obj;
+ ATOM_OBJECT_HEADER *obj_header;
+
+ if (!amdgpu_atom_parse_data_header(ctx, index, &size, &frev, &crev, &data_offset))
+ return false;
+
+ if (crev < 2)
+ return false;
+
+ obj_header = (ATOM_OBJECT_HEADER *) (ctx->bios + data_offset);
+ path_obj = (ATOM_DISPLAY_OBJECT_PATH_TABLE *)
+ (ctx->bios + data_offset +
+ le16_to_cpu(obj_header->usDisplayPathTableOffset));
+
+ if (path_obj->ucNumOfDispPath)
+ return true;
+ else
+ return false;
+}
+
bool amdgpu_atombios_get_connector_info_from_object_table(struct amdgpu_device *adev)
{
struct amdgpu_mode_info *mode_info = &adev->mode_info;
return -EINVAL;
switch (crev) {
+ case 2:
+ case 3:
+ case 5:
+ /* r6xx, r7xx, evergreen, ni, si.
+ * TODO: add support for asic_type <= CHIP_RV770*/
+ if (clock_type == COMPUTE_ENGINE_PLL_PARAM) {
+ args.v3.ulClockParams = cpu_to_le32((clock_type << 24) | clock);
+
+ amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
+
+ dividers->post_div = args.v3.ucPostDiv;
+ dividers->enable_post_div = (args.v3.ucCntlFlag &
+ ATOM_PLL_CNTL_FLAG_PLL_POST_DIV_EN) ? true : false;
+ dividers->enable_dithen = (args.v3.ucCntlFlag &
+ ATOM_PLL_CNTL_FLAG_FRACTION_DISABLE) ? false : true;
+ dividers->whole_fb_div = le16_to_cpu(args.v3.ulFbDiv.usFbDiv);
+ dividers->frac_fb_div = le16_to_cpu(args.v3.ulFbDiv.usFbDivFrac);
+ dividers->ref_div = args.v3.ucRefDiv;
+ dividers->vco_mode = (args.v3.ucCntlFlag &
+ ATOM_PLL_CNTL_FLAG_MPLL_VCO_MODE) ? 1 : 0;
+ } else {
+ /* for SI we use ComputeMemoryClockParam for memory plls */
+ if (adev->asic_type >= CHIP_TAHITI)
+ return -EINVAL;
+ args.v5.ulClockParams = cpu_to_le32((clock_type << 24) | clock);
+ if (strobe_mode)
+ args.v5.ucInputFlag = ATOM_PLL_INPUT_FLAG_PLL_STROBE_MODE_EN;
+
+ amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
+
+ dividers->post_div = args.v5.ucPostDiv;
+ dividers->enable_post_div = (args.v5.ucCntlFlag &
+ ATOM_PLL_CNTL_FLAG_PLL_POST_DIV_EN) ? true : false;
+ dividers->enable_dithen = (args.v5.ucCntlFlag &
+ ATOM_PLL_CNTL_FLAG_FRACTION_DISABLE) ? false : true;
+ dividers->whole_fb_div = le16_to_cpu(args.v5.ulFbDiv.usFbDiv);
+ dividers->frac_fb_div = le16_to_cpu(args.v5.ulFbDiv.usFbDivFrac);
+ dividers->ref_div = args.v5.ucRefDiv;
+ dividers->vco_mode = (args.v5.ucCntlFlag &
+ ATOM_PLL_CNTL_FLAG_MPLL_VCO_MODE) ? 1 : 0;
+ }
+ break;
case 4:
/* fusion */
args.v4.ulClock = cpu_to_le32(clock); /* 10 khz */
amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
}
+void amdgpu_atombios_get_default_voltages(struct amdgpu_device *adev,
+ u16 *vddc, u16 *vddci, u16 *mvdd)
+{
+ struct amdgpu_mode_info *mode_info = &adev->mode_info;
+ int index = GetIndexIntoMasterTable(DATA, FirmwareInfo);
+ u8 frev, crev;
+ u16 data_offset;
+ union firmware_info *firmware_info;
+
+ *vddc = 0;
+ *vddci = 0;
+ *mvdd = 0;
+
+ if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL,
+ &frev, &crev, &data_offset)) {
+ firmware_info =
+ (union firmware_info *)(mode_info->atom_context->bios +
+ data_offset);
+ *vddc = le16_to_cpu(firmware_info->info_14.usBootUpVDDCVoltage);
+ if ((frev == 2) && (crev >= 2)) {
+ *vddci = le16_to_cpu(firmware_info->info_22.usBootUpVDDCIVoltage);
+ *mvdd = le16_to_cpu(firmware_info->info_22.usBootUpMVDDCVoltage);
+ }
+ }
+}
+
union set_voltage {
struct _SET_VOLTAGE_PS_ALLOCATION alloc;
struct _SET_VOLTAGE_PARAMETERS v1;
struct _SET_VOLTAGE_PARAMETERS_V1_3 v3;
};
+int amdgpu_atombios_get_max_vddc(struct amdgpu_device *adev, u8 voltage_type,
+ u16 voltage_id, u16 *voltage)
+{
+ union set_voltage args;
+ int index = GetIndexIntoMasterTable(COMMAND, SetVoltage);
+ u8 frev, crev;
+
+ if (!amdgpu_atom_parse_cmd_header(adev->mode_info.atom_context, index, &frev, &crev))
+ return -EINVAL;
+
+ switch (crev) {
+ case 1:
+ return -EINVAL;
+ case 2:
+ args.v2.ucVoltageType = SET_VOLTAGE_GET_MAX_VOLTAGE;
+ args.v2.ucVoltageMode = 0;
+ args.v2.usVoltageLevel = 0;
+
+ amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
+
+ *voltage = le16_to_cpu(args.v2.usVoltageLevel);
+ break;
+ case 3:
+ args.v3.ucVoltageType = voltage_type;
+ args.v3.ucVoltageMode = ATOM_GET_VOLTAGE_LEVEL;
+ args.v3.usVoltageLevel = cpu_to_le16(voltage_id);
+
+ amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
+
+ *voltage = le16_to_cpu(args.v3.usVoltageLevel);
+ break;
+ default:
+ DRM_ERROR("Unknown table version %d, %d\n", frev, crev);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+int amdgpu_atombios_get_leakage_vddc_based_on_leakage_idx(struct amdgpu_device *adev,
+ u16 *voltage,
+ u16 leakage_idx)
+{
+ return amdgpu_atombios_get_max_vddc(adev, VOLTAGE_TYPE_VDDC, leakage_idx, voltage);
+}
+
void amdgpu_atombios_set_voltage(struct amdgpu_device *adev,
u16 voltage_level,
u8 voltage_type)
return NULL;
}
+int amdgpu_atombios_get_svi2_info(struct amdgpu_device *adev,
+ u8 voltage_type,
+ u8 *svd_gpio_id, u8 *svc_gpio_id)
+{
+ int index = GetIndexIntoMasterTable(DATA, VoltageObjectInfo);
+ u8 frev, crev;
+ u16 data_offset, size;
+ union voltage_object_info *voltage_info;
+ union voltage_object *voltage_object = NULL;
+
+ if (amdgpu_atom_parse_data_header(adev->mode_info.atom_context, index, &size,
+ &frev, &crev, &data_offset)) {
+ voltage_info = (union voltage_object_info *)
+ (adev->mode_info.atom_context->bios + data_offset);
+
+ switch (frev) {
+ case 3:
+ switch (crev) {
+ case 1:
+ voltage_object = (union voltage_object *)
+ amdgpu_atombios_lookup_voltage_object_v3(&voltage_info->v3,
+ voltage_type,
+ VOLTAGE_OBJ_SVID2);
+ if (voltage_object) {
+ *svd_gpio_id = voltage_object->v3.asSVID2Obj.ucSVDGpioId;
+ *svc_gpio_id = voltage_object->v3.asSVID2Obj.ucSVCGpioId;
+ } else {
+ return -EINVAL;
+ }
+ break;
+ default:
+ DRM_ERROR("unknown voltage object table\n");
+ return -EINVAL;
+ }
+ break;
+ default:
+ DRM_ERROR("unknown voltage object table\n");
+ return -EINVAL;
+ }
+
+ }
+ return 0;
+}
+
bool
amdgpu_atombios_is_voltage_gpio(struct amdgpu_device *adev,
u8 voltage_type, u8 voltage_mode)
uint8_t id);
void amdgpu_atombios_i2c_init(struct amdgpu_device *adev);
+bool amdgpu_atombios_has_dce_engine_info(struct amdgpu_device *adev);
+
bool amdgpu_atombios_get_connector_info_from_object_table(struct amdgpu_device *adev);
int amdgpu_atombios_get_clock_info(struct amdgpu_device *adev);
void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev);
void amdgpu_atombios_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le);
-
+int amdgpu_atombios_get_max_vddc(struct amdgpu_device *adev, u8 voltage_type,
+ u16 voltage_id, u16 *voltage);
+int amdgpu_atombios_get_leakage_vddc_based_on_leakage_idx(struct amdgpu_device *adev,
+ u16 *voltage,
+ u16 leakage_idx);
+void amdgpu_atombios_get_default_voltages(struct amdgpu_device *adev,
+ u16 *vddc, u16 *vddci, u16 *mvdd);
+int amdgpu_atombios_get_clock_dividers(struct amdgpu_device *adev,
+ u8 clock_type,
+ u32 clock,
+ bool strobe_mode,
+ struct atom_clock_dividers *dividers);
+int amdgpu_atombios_get_svi2_info(struct amdgpu_device *adev,
+ u8 voltage_type,
+ u8 *svd_gpio_id, u8 *svc_gpio_id);
#endif
start_jiffies = jiffies;
for (i = 0; i < n; i++) {
struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring;
- r = amdgpu_copy_buffer(ring, saddr, daddr, size, NULL, &fence);
+ r = amdgpu_copy_buffer(ring, saddr, daddr, size, NULL, &fence,
+ false);
if (r)
goto exit_do_move;
r = fence_wait(fence, false);
return amdgpu_irq_put(adev, adev->irq.sources[src_id], type);
}
-int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device,
+static int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device,
enum amd_ip_block_type block_type,
enum amd_clockgating_state state)
{
return r;
}
-int amdgpu_cgs_set_powergating_state(struct cgs_device *cgs_device,
+static int amdgpu_cgs_set_powergating_state(struct cgs_device *cgs_device,
enum amd_ip_block_type block_type,
enum amd_powergating_state state)
{
case CGS_SYSTEM_INFO_GFX_SE_INFO:
sys_info->value = adev->gfx.config.max_shader_engines;
break;
+ case CGS_SYSTEM_INFO_PCIE_SUB_SYS_ID:
+ sys_info->value = adev->pdev->subsystem_device;
+ break;
+ case CGS_SYSTEM_INFO_PCIE_SUB_SYS_VENDOR_ID:
+ sys_info->value = adev->pdev->subsystem_vendor;
+ break;
default:
return -ENODEV;
}
.force = amdgpu_connector_dvi_force,
};
+static struct drm_encoder *
+amdgpu_connector_virtual_encoder(struct drm_connector *connector)
+{
+ int enc_id = connector->encoder_ids[0];
+ struct drm_encoder *encoder;
+ int i;
+ for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
+ if (connector->encoder_ids[i] == 0)
+ break;
+
+ encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
+ if (!encoder)
+ continue;
+
+ if (encoder->encoder_type == DRM_MODE_ENCODER_VIRTUAL)
+ return encoder;
+ }
+
+ /* pick the first one */
+ if (enc_id)
+ return drm_encoder_find(connector->dev, enc_id);
+ return NULL;
+}
+
+static int amdgpu_connector_virtual_get_modes(struct drm_connector *connector)
+{
+ struct drm_encoder *encoder = amdgpu_connector_best_single_encoder(connector);
+
+ if (encoder) {
+ amdgpu_connector_add_common_modes(encoder, connector);
+ }
+
+ return 0;
+}
+
+static int amdgpu_connector_virtual_mode_valid(struct drm_connector *connector,
+ struct drm_display_mode *mode)
+{
+ return MODE_OK;
+}
+
+int amdgpu_connector_virtual_dpms(struct drm_connector *connector, int mode)
+{
+ return 0;
+}
+
+static enum drm_connector_status
+
+amdgpu_connector_virtual_detect(struct drm_connector *connector, bool force)
+{
+ return connector_status_connected;
+}
+
+int amdgpu_connector_virtual_set_property(struct drm_connector *connector,
+ struct drm_property *property,
+ uint64_t val)
+{
+ return 0;
+}
+
+static void amdgpu_connector_virtual_force(struct drm_connector *connector)
+{
+ return;
+}
+
+static const struct drm_connector_helper_funcs amdgpu_connector_virtual_helper_funcs = {
+ .get_modes = amdgpu_connector_virtual_get_modes,
+ .mode_valid = amdgpu_connector_virtual_mode_valid,
+ .best_encoder = amdgpu_connector_virtual_encoder,
+};
+
+static const struct drm_connector_funcs amdgpu_connector_virtual_funcs = {
+ .dpms = amdgpu_connector_virtual_dpms,
+ .detect = amdgpu_connector_virtual_detect,
+ .fill_modes = drm_helper_probe_single_connector_modes,
+ .set_property = amdgpu_connector_virtual_set_property,
+ .destroy = amdgpu_connector_destroy,
+ .force = amdgpu_connector_virtual_force,
+};
+
void
amdgpu_connector_add(struct amdgpu_device *adev,
uint32_t connector_id,
connector->interlace_allowed = false;
connector->doublescan_allowed = false;
break;
+ case DRM_MODE_CONNECTOR_VIRTUAL:
+ amdgpu_dig_connector = kzalloc(sizeof(struct amdgpu_connector_atom_dig), GFP_KERNEL);
+ if (!amdgpu_dig_connector)
+ goto failed;
+ amdgpu_connector->con_priv = amdgpu_dig_connector;
+ drm_connector_init(dev, &amdgpu_connector->base, &amdgpu_connector_virtual_funcs, connector_type);
+ drm_connector_helper_add(&amdgpu_connector->base, &amdgpu_connector_virtual_helper_funcs);
+ subpixel_order = SubPixelHorizontalRGB;
+ connector->interlace_allowed = false;
+ connector->doublescan_allowed = false;
+ break;
}
}
uint32_t *offset)
{
struct drm_gem_object *gobj;
+ unsigned long size;
gobj = drm_gem_object_lookup(p->filp, data->handle);
if (gobj == NULL)
p->uf_entry.tv.bo = &p->uf_entry.robj->tbo;
p->uf_entry.tv.shared = true;
p->uf_entry.user_pages = NULL;
+
+ size = amdgpu_bo_size(p->uf_entry.robj);
+ if (size != PAGE_SIZE || (data->offset + 8) > size)
+ return -EINVAL;
+
*offset = data->offset;
drm_gem_object_unreference_unlocked(gobj);
return ret;
}
-/* Returns how many bytes TTM can move per IB.
+/* Convert microseconds to bytes. */
+static u64 us_to_bytes(struct amdgpu_device *adev, s64 us)
+{
+ if (us <= 0 || !adev->mm_stats.log2_max_MBps)
+ return 0;
+
+ /* Since accum_us is incremented by a million per second, just
+ * multiply it by the number of MB/s to get the number of bytes.
+ */
+ return us << adev->mm_stats.log2_max_MBps;
+}
+
+static s64 bytes_to_us(struct amdgpu_device *adev, u64 bytes)
+{
+ if (!adev->mm_stats.log2_max_MBps)
+ return 0;
+
+ return bytes >> adev->mm_stats.log2_max_MBps;
+}
+
+/* Returns how many bytes TTM can move right now. If no bytes can be moved,
+ * it returns 0. If it returns non-zero, it's OK to move at least one buffer,
+ * which means it can go over the threshold once. If that happens, the driver
+ * will be in debt and no other buffer migrations can be done until that debt
+ * is repaid.
+ *
+ * This approach allows moving a buffer of any size (it's important to allow
+ * that).
+ *
+ * The currency is simply time in microseconds and it increases as the clock
+ * ticks. The accumulated microseconds (us) are converted to bytes and
+ * returned.
*/
static u64 amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev)
{
- u64 real_vram_size = adev->mc.real_vram_size;
- u64 vram_usage = atomic64_read(&adev->vram_usage);
+ s64 time_us, increment_us;
+ u64 max_bytes;
+ u64 free_vram, total_vram, used_vram;
- /* This function is based on the current VRAM usage.
+ /* Allow a maximum of 200 accumulated ms. This is basically per-IB
+ * throttling.
*
- * - If all of VRAM is free, allow relocating the number of bytes that
- * is equal to 1/4 of the size of VRAM for this IB.
+ * It means that in order to get full max MBps, at least 5 IBs per
+ * second must be submitted and not more than 200ms apart from each
+ * other.
+ */
+ const s64 us_upper_bound = 200000;
- * - If more than one half of VRAM is occupied, only allow relocating
- * 1 MB of data for this IB.
- *
- * - From 0 to one half of used VRAM, the threshold decreases
- * linearly.
- * __________________
- * 1/4 of -|\ |
- * VRAM | \ |
- * | \ |
- * | \ |
- * | \ |
- * | \ |
- * | \ |
- * | \________|1 MB
- * |----------------|
- * VRAM 0 % 100 %
- * used used
- *
- * Note: It's a threshold, not a limit. The threshold must be crossed
- * for buffer relocations to stop, so any buffer of an arbitrary size
- * can be moved as long as the threshold isn't crossed before
- * the relocation takes place. We don't want to disable buffer
- * relocations completely.
+ if (!adev->mm_stats.log2_max_MBps)
+ return 0;
+
+ total_vram = adev->mc.real_vram_size - adev->vram_pin_size;
+ used_vram = atomic64_read(&adev->vram_usage);
+ free_vram = used_vram >= total_vram ? 0 : total_vram - used_vram;
+
+ spin_lock(&adev->mm_stats.lock);
+
+ /* Increase the amount of accumulated us. */
+ time_us = ktime_to_us(ktime_get());
+ increment_us = time_us - adev->mm_stats.last_update_us;
+ adev->mm_stats.last_update_us = time_us;
+ adev->mm_stats.accum_us = min(adev->mm_stats.accum_us + increment_us,
+ us_upper_bound);
+
+ /* This prevents the short period of low performance when the VRAM
+ * usage is low and the driver is in debt or doesn't have enough
+ * accumulated us to fill VRAM quickly.
*
- * The idea is that buffers should be placed in VRAM at creation time
- * and TTM should only do a minimum number of relocations during
- * command submission. In practice, you need to submit at least
- * a dozen IBs to move all buffers to VRAM if they are in GTT.
+ * The situation can occur in these cases:
+ * - a lot of VRAM is freed by userspace
+ * - the presence of a big buffer causes a lot of evictions
+ * (solution: split buffers into smaller ones)
*
- * Also, things can get pretty crazy under memory pressure and actual
- * VRAM usage can change a lot, so playing safe even at 50% does
- * consistently increase performance.
+ * If 128 MB or 1/8th of VRAM is free, start filling it now by setting
+ * accum_us to a positive number.
+ */
+ if (free_vram >= 128 * 1024 * 1024 || free_vram >= total_vram / 8) {
+ s64 min_us;
+
+ /* Be more aggresive on dGPUs. Try to fill a portion of free
+ * VRAM now.
+ */
+ if (!(adev->flags & AMD_IS_APU))
+ min_us = bytes_to_us(adev, free_vram / 4);
+ else
+ min_us = 0; /* Reset accum_us on APUs. */
+
+ adev->mm_stats.accum_us = max(min_us, adev->mm_stats.accum_us);
+ }
+
+ /* This returns 0 if the driver is in debt to disallow (optional)
+ * buffer moves.
+ */
+ max_bytes = us_to_bytes(adev, adev->mm_stats.accum_us);
+
+ spin_unlock(&adev->mm_stats.lock);
+ return max_bytes;
+}
+
+/* Report how many bytes have really been moved for the last command
+ * submission. This can result in a debt that can stop buffer migrations
+ * temporarily.
+ */
+static void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev,
+ u64 num_bytes)
+{
+ spin_lock(&adev->mm_stats.lock);
+ adev->mm_stats.accum_us -= bytes_to_us(adev, num_bytes);
+ spin_unlock(&adev->mm_stats.lock);
+}
+
+static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
+ struct amdgpu_bo *bo)
+{
+ u64 initial_bytes_moved;
+ uint32_t domain;
+ int r;
+
+ if (bo->pin_count)
+ return 0;
+
+ /* Don't move this buffer if we have depleted our allowance
+ * to move it. Don't move anything if the threshold is zero.
*/
+ if (p->bytes_moved < p->bytes_moved_threshold)
+ domain = bo->prefered_domains;
+ else
+ domain = bo->allowed_domains;
+
+retry:
+ amdgpu_ttm_placement_from_domain(bo, domain);
+ initial_bytes_moved = atomic64_read(&bo->adev->num_bytes_moved);
+ r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+ p->bytes_moved += atomic64_read(&bo->adev->num_bytes_moved) -
+ initial_bytes_moved;
+
+ if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) {
+ domain = bo->allowed_domains;
+ goto retry;
+ }
- u64 half_vram = real_vram_size >> 1;
- u64 half_free_vram = vram_usage >= half_vram ? 0 : half_vram - vram_usage;
- u64 bytes_moved_threshold = half_free_vram >> 1;
- return max(bytes_moved_threshold, 1024*1024ull);
+ return r;
}
-int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
+/* Last resort, try to evict something from the current working set */
+static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
+ struct amdgpu_bo_list_entry *lobj)
+{
+ uint32_t domain = lobj->robj->allowed_domains;
+ int r;
+
+ if (!p->evictable)
+ return false;
+
+ for (;&p->evictable->tv.head != &p->validated;
+ p->evictable = list_prev_entry(p->evictable, tv.head)) {
+
+ struct amdgpu_bo_list_entry *candidate = p->evictable;
+ struct amdgpu_bo *bo = candidate->robj;
+ u64 initial_bytes_moved;
+ uint32_t other;
+
+ /* If we reached our current BO we can forget it */
+ if (candidate == lobj)
+ break;
+
+ other = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
+
+ /* Check if this BO is in one of the domains we need space for */
+ if (!(other & domain))
+ continue;
+
+ /* Check if we can move this BO somewhere else */
+ other = bo->allowed_domains & ~domain;
+ if (!other)
+ continue;
+
+ /* Good we can try to move this BO somewhere else */
+ amdgpu_ttm_placement_from_domain(bo, other);
+ initial_bytes_moved = atomic64_read(&bo->adev->num_bytes_moved);
+ r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
+ p->bytes_moved += atomic64_read(&bo->adev->num_bytes_moved) -
+ initial_bytes_moved;
+
+ if (unlikely(r))
+ break;
+
+ p->evictable = list_prev_entry(p->evictable, tv.head);
+ list_move(&candidate->tv.head, &p->validated);
+
+ return true;
+ }
+
+ return false;
+}
+
+static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
struct list_head *validated)
{
struct amdgpu_bo_list_entry *lobj;
- u64 initial_bytes_moved;
int r;
list_for_each_entry(lobj, validated, tv.head) {
struct amdgpu_bo *bo = lobj->robj;
bool binding_userptr = false;
struct mm_struct *usermm;
- uint32_t domain;
usermm = amdgpu_ttm_tt_get_usermm(bo->tbo.ttm);
if (usermm && usermm != current->mm)
binding_userptr = true;
}
- if (bo->pin_count)
- continue;
-
- /* Avoid moving this one if we have moved too many buffers
- * for this IB already.
- *
- * Note that this allows moving at least one buffer of
- * any size, because it doesn't take the current "bo"
- * into account. We don't want to disallow buffer moves
- * completely.
- */
- if (p->bytes_moved <= p->bytes_moved_threshold)
- domain = bo->prefered_domains;
- else
- domain = bo->allowed_domains;
-
- retry:
- amdgpu_ttm_placement_from_domain(bo, domain);
- initial_bytes_moved = atomic64_read(&bo->adev->num_bytes_moved);
- r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
- p->bytes_moved += atomic64_read(&bo->adev->num_bytes_moved) -
- initial_bytes_moved;
+ if (p->evictable == lobj)
+ p->evictable = NULL;
- if (unlikely(r)) {
- if (r != -ERESTARTSYS && domain != bo->allowed_domains) {
- domain = bo->allowed_domains;
- goto retry;
- }
+ do {
+ r = amdgpu_cs_bo_validate(p, bo);
+ } while (r == -ENOMEM && amdgpu_cs_try_evict(p, lobj));
+ if (r)
return r;
+
+ if (bo->shadow) {
+ r = amdgpu_cs_bo_validate(p, bo);
+ if (r)
+ return r;
}
if (binding_userptr) {
r = ttm_eu_reserve_buffers(&p->ticket, &p->validated, true,
&duplicates);
- if (unlikely(r != 0))
+ if (unlikely(r != 0)) {
+ DRM_ERROR("ttm_eu_reserve_buffers failed.\n");
goto error_free_pages;
+ }
/* Without a BO list we don't have userptr BOs */
if (!p->bo_list)
/* Unreserve everything again. */
ttm_eu_backoff_reservation(&p->ticket, &p->validated);
- /* We tried to often, just abort */
+ /* We tried too many times, just abort */
if (!--tries) {
r = -EDEADLK;
+ DRM_ERROR("deadlock in %s\n", __func__);
goto error_free_pages;
}
sizeof(struct page*));
if (!e->user_pages) {
r = -ENOMEM;
+ DRM_ERROR("calloc failure in %s\n", __func__);
goto error_free_pages;
}
r = amdgpu_ttm_tt_get_user_pages(ttm, e->user_pages);
if (r) {
+ DRM_ERROR("amdgpu_ttm_tt_get_user_pages failed.\n");
drm_free_large(e->user_pages);
e->user_pages = NULL;
goto error_free_pages;
p->bytes_moved_threshold = amdgpu_cs_get_threshold_for_moves(p->adev);
p->bytes_moved = 0;
+ p->evictable = list_last_entry(&p->validated,
+ struct amdgpu_bo_list_entry,
+ tv.head);
r = amdgpu_cs_list_validate(p, &duplicates);
- if (r)
+ if (r) {
+ DRM_ERROR("amdgpu_cs_list_validate(duplicates) failed.\n");
goto error_validate;
+ }
r = amdgpu_cs_list_validate(p, &p->validated);
- if (r)
+ if (r) {
+ DRM_ERROR("amdgpu_cs_list_validate(validated) failed.\n");
goto error_validate;
+ }
+
+ amdgpu_cs_report_moved_bytes(p->adev, p->bytes_moved);
fpriv->vm.last_eviction_counter =
atomic64_read(&p->adev->num_evictions);
}
}
- if (p->uf_entry.robj)
- p->job->uf_addr += amdgpu_bo_gpu_offset(p->uf_entry.robj);
+ if (!r && p->uf_entry.robj) {
+ struct amdgpu_bo *uf = p->uf_entry.robj;
+
+ r = amdgpu_ttm_bind(uf->tbo.ttm, &uf->tbo.mem);
+ p->job->uf_addr += amdgpu_bo_gpu_offset(uf);
+ }
error_validate:
if (r) {
if (bo_va == NULL)
continue;
- r = amdgpu_vm_bo_update(adev, bo_va, &bo->tbo.mem);
+ r = amdgpu_vm_bo_update(adev, bo_va, false);
if (r)
return r;
if (r)
return r;
+ if (ib->flags & AMDGPU_IB_FLAG_PREAMBLE) {
+ parser->job->preamble_status |= AMDGPU_PREAMBLE_IB_PRESENT;
+ if (!parser->ctx->preamble_presented) {
+ parser->job->preamble_status |= AMDGPU_PREAMBLE_IB_PRESENT_FIRST;
+ parser->ctx->preamble_presented = true;
+ }
+ }
+
if (parser->job->ring && parser->job->ring != ring)
return -EINVAL;
}
job->owner = p->filp;
- job->ctx = entity->fence_context;
+ job->fence_ctx = entity->fence_context;
p->fence = fence_get(&job->base.s_fence->finished);
cs->out.handle = amdgpu_ctx_add_fence(p->ctx, ring, p->fence);
job->uf_sequence = cs->out.handle;
return NULL;
}
+
+/**
+ * amdgpu_cs_sysvm_access_required - make BOs accessible by the system VM
+ *
+ * @parser: command submission parser context
+ *
+ * Helper for UVD/VCE VM emulation, make sure BOs are accessible by the system VM.
+ */
+int amdgpu_cs_sysvm_access_required(struct amdgpu_cs_parser *parser)
+{
+ unsigned i;
+ int r;
+
+ if (!parser->bo_list)
+ return 0;
+
+ for (i = 0; i < parser->bo_list->num_entries; i++) {
+ struct amdgpu_bo *bo = parser->bo_list->array[i].robj;
+
+ r = amdgpu_ttm_bind(bo->tbo.ttm, &bo->tbo.mem);
+ if (unlikely(r))
+ return r;
+ }
+
+ return 0;
+}
#include "atom.h"
#include "amdgpu_atombios.h"
#include "amd_pcie.h"
+#ifdef CONFIG_DRM_AMDGPU_SI
+#include "si.h"
+#endif
#ifdef CONFIG_DRM_AMDGPU_CIK
#include "cik.h"
#endif
#include "vi.h"
#include "bif/bif_4_1_d.h"
+#include <linux/pci.h>
static int amdgpu_debugfs_regs_init(struct amdgpu_device *adev);
static void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev);
static const char *amdgpu_asic_name[] = {
+ "TAHITI",
+ "PITCAIRN",
+ "VERDE",
+ "OLAND",
+ "HAINAN",
"BONAIRE",
"KAVERI",
"KABINI",
/* don't suspend or resume card normally */
dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
- amdgpu_resume_kms(dev, true, true);
+ amdgpu_device_resume(dev, true, true);
dev->pdev->d3_delay = d3_delay;
printk(KERN_INFO "amdgpu: switched off\n");
drm_kms_helper_poll_disable(dev);
dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
- amdgpu_suspend_kms(dev, true, true);
+ amdgpu_device_suspend(dev, true, true);
dev->switch_power_state = DRM_SWITCH_POWER_OFF;
}
}
return 1;
}
+static void amdgpu_whether_enable_virtual_display(struct amdgpu_device *adev)
+{
+ adev->enable_virtual_display = false;
+
+ if (amdgpu_virtual_display) {
+ struct drm_device *ddev = adev->ddev;
+ const char *pci_address_name = pci_name(ddev->pdev);
+ char *pciaddstr, *pciaddstr_tmp, *pciaddname;
+
+ pciaddstr = kstrdup(amdgpu_virtual_display, GFP_KERNEL);
+ pciaddstr_tmp = pciaddstr;
+ while ((pciaddname = strsep(&pciaddstr_tmp, ";"))) {
+ if (!strcmp(pci_address_name, pciaddname)) {
+ adev->enable_virtual_display = true;
+ break;
+ }
+ }
+
+ DRM_INFO("virtual display string:%s, %s:virtual_display:%d\n",
+ amdgpu_virtual_display, pci_address_name,
+ adev->enable_virtual_display);
+
+ kfree(pciaddstr);
+ }
+}
+
static int amdgpu_early_init(struct amdgpu_device *adev)
{
int i, r;
+ amdgpu_whether_enable_virtual_display(adev);
+
switch (adev->asic_type) {
case CHIP_TOPAZ:
case CHIP_TONGA:
if (r)
return r;
break;
+#ifdef CONFIG_DRM_AMDGPU_SI
+ case CHIP_VERDE:
+ case CHIP_TAHITI:
+ case CHIP_PITCAIRN:
+ case CHIP_OLAND:
+ case CHIP_HAINAN:
+ adev->family = AMDGPU_FAMILY_SI;
+ r = si_set_ip_blocks(adev);
+ if (r)
+ return r;
+ break;
+#endif
#ifdef CONFIG_DRM_AMDGPU_CIK
case CHIP_BONAIRE:
case CHIP_HAWAII:
for (i = 0; i < adev->num_ip_blocks; i++) {
if (!adev->ip_block_status[i].valid)
continue;
+ if (adev->ip_blocks[i].type == AMD_IP_BLOCK_TYPE_UVD ||
+ adev->ip_blocks[i].type == AMD_IP_BLOCK_TYPE_VCE)
+ continue;
/* enable clockgating to save power */
r = adev->ip_blocks[i].funcs->set_clockgating_state((void *)adev,
AMD_CG_STATE_GATE);
{
int r, i;
bool runtime = false;
+ u32 max_MBps;
adev->shutdown = false;
adev->dev = &pdev->dev;
adev->smc_wreg = &amdgpu_invalid_wreg;
adev->pcie_rreg = &amdgpu_invalid_rreg;
adev->pcie_wreg = &amdgpu_invalid_wreg;
+ adev->pciep_rreg = &amdgpu_invalid_rreg;
+ adev->pciep_wreg = &amdgpu_invalid_wreg;
adev->uvd_ctx_rreg = &amdgpu_invalid_rreg;
adev->uvd_ctx_wreg = &amdgpu_invalid_wreg;
adev->didt_rreg = &amdgpu_invalid_rreg;
spin_lock_init(&adev->didt_idx_lock);
spin_lock_init(&adev->gc_cac_idx_lock);
spin_lock_init(&adev->audio_endpt_idx_lock);
+ spin_lock_init(&adev->mm_stats.lock);
+
+ INIT_LIST_HEAD(&adev->shadow_list);
+ mutex_init(&adev->shadow_list_lock);
+
+ INIT_LIST_HEAD(&adev->gtt_list);
+ spin_lock_init(&adev->gtt_list_lock);
+
+ if (adev->asic_type >= CHIP_BONAIRE) {
+ adev->rmmio_base = pci_resource_start(adev->pdev, 5);
+ adev->rmmio_size = pci_resource_len(adev->pdev, 5);
+ } else {
+ adev->rmmio_base = pci_resource_start(adev->pdev, 2);
+ adev->rmmio_size = pci_resource_len(adev->pdev, 2);
+ }
- adev->rmmio_base = pci_resource_start(adev->pdev, 5);
- adev->rmmio_size = pci_resource_len(adev->pdev, 5);
adev->rmmio = ioremap(adev->rmmio_base, adev->rmmio_size);
if (adev->rmmio == NULL) {
return -ENOMEM;
DRM_INFO("register mmio base: 0x%08X\n", (uint32_t)adev->rmmio_base);
DRM_INFO("register mmio size: %u\n", (unsigned)adev->rmmio_size);
- /* doorbell bar mapping */
- amdgpu_doorbell_init(adev);
+ if (adev->asic_type >= CHIP_BONAIRE)
+ /* doorbell bar mapping */
+ amdgpu_doorbell_init(adev);
/* io port mapping */
for (i = 0; i < DEVICE_COUNT_RESOURCE; i++) {
adev->accel_working = true;
+ /* Initialize the buffer migration limit. */
+ if (amdgpu_moverate >= 0)
+ max_MBps = amdgpu_moverate;
+ else
+ max_MBps = 8; /* Allow 8 MB/s. */
+ /* Get a log2 for easy divisions. */
+ adev->mm_stats.log2_max_MBps = ilog2(max(1u, max_MBps));
+
amdgpu_fbdev_init(adev);
r = amdgpu_ib_pool_init(adev);
adev->rio_mem = NULL;
iounmap(adev->rmmio);
adev->rmmio = NULL;
- amdgpu_doorbell_fini(adev);
+ if (adev->asic_type >= CHIP_BONAIRE)
+ amdgpu_doorbell_fini(adev);
amdgpu_debugfs_regs_cleanup(adev);
amdgpu_debugfs_remove_files(adev);
}
* Suspend & resume.
*/
/**
- * amdgpu_suspend_kms - initiate device suspend
+ * amdgpu_device_suspend - initiate device suspend
*
* @pdev: drm dev pointer
* @state: suspend state
* Returns 0 for success or an error on failure.
* Called at driver suspend.
*/
-int amdgpu_suspend_kms(struct drm_device *dev, bool suspend, bool fbcon)
+int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
{
struct amdgpu_device *adev;
struct drm_crtc *crtc;
adev = dev->dev_private;
- if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF ||
+ dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF)
return 0;
drm_kms_helper_poll_disable(dev);
/* Shut down the device */
pci_disable_device(dev->pdev);
pci_set_power_state(dev->pdev, PCI_D3hot);
+ } else {
+ r = amdgpu_asic_reset(adev);
+ if (r)
+ DRM_ERROR("amdgpu asic reset failed\n");
}
if (fbcon) {
}
/**
- * amdgpu_resume_kms - initiate device resume
+ * amdgpu_device_resume - initiate device resume
*
* @pdev: drm dev pointer
*
* Returns 0 for success or an error on failure.
* Called at driver resume.
*/
-int amdgpu_resume_kms(struct drm_device *dev, bool resume, bool fbcon)
+int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon)
{
struct drm_connector *connector;
struct amdgpu_device *adev = dev->dev_private;
struct drm_crtc *crtc;
int r;
- if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF ||
+ dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF)
return 0;
- if (fbcon) {
+ if (fbcon)
console_lock();
- }
+
if (resume) {
pci_set_power_state(dev->pdev, PCI_D0);
pci_restore_state(dev->pdev);
- if (pci_enable_device(dev->pdev)) {
+ r = pci_enable_device(dev->pdev);
+ if (r) {
if (fbcon)
console_unlock();
- return -1;
+ return r;
}
}
/* post card */
- if (!amdgpu_card_posted(adev))
- amdgpu_atom_asic_init(adev->mode_info.atom_context);
+ if (!amdgpu_card_posted(adev) || !resume) {
+ r = amdgpu_atom_asic_init(adev->mode_info.atom_context);
+ if (r)
+ DRM_ERROR("amdgpu asic init failed\n");
+ }
r = amdgpu_resume(adev);
if (r)
return 0;
}
+static bool amdgpu_check_soft_reset(struct amdgpu_device *adev)
+{
+ int i;
+ bool asic_hang = false;
+
+ for (i = 0; i < adev->num_ip_blocks; i++) {
+ if (!adev->ip_block_status[i].valid)
+ continue;
+ if (adev->ip_blocks[i].funcs->check_soft_reset)
+ adev->ip_blocks[i].funcs->check_soft_reset(adev);
+ if (adev->ip_block_status[i].hang) {
+ DRM_INFO("IP block:%d is hang!\n", i);
+ asic_hang = true;
+ }
+ }
+ return asic_hang;
+}
+
+int amdgpu_pre_soft_reset(struct amdgpu_device *adev)
+{
+ int i, r = 0;
+
+ for (i = 0; i < adev->num_ip_blocks; i++) {
+ if (!adev->ip_block_status[i].valid)
+ continue;
+ if (adev->ip_block_status[i].hang &&
+ adev->ip_blocks[i].funcs->pre_soft_reset) {
+ r = adev->ip_blocks[i].funcs->pre_soft_reset(adev);
+ if (r)
+ return r;
+ }
+ }
+
+ return 0;
+}
+
+static bool amdgpu_need_full_reset(struct amdgpu_device *adev)
+{
+ if (adev->ip_block_status[AMD_IP_BLOCK_TYPE_GMC].hang ||
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_SMC].hang ||
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_ACP].hang ||
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_DCE].hang) {
+ DRM_INFO("Some block need full reset!\n");
+ return true;
+ }
+ return false;
+}
+
+static int amdgpu_soft_reset(struct amdgpu_device *adev)
+{
+ int i, r = 0;
+
+ for (i = 0; i < adev->num_ip_blocks; i++) {
+ if (!adev->ip_block_status[i].valid)
+ continue;
+ if (adev->ip_block_status[i].hang &&
+ adev->ip_blocks[i].funcs->soft_reset) {
+ r = adev->ip_blocks[i].funcs->soft_reset(adev);
+ if (r)
+ return r;
+ }
+ }
+
+ return 0;
+}
+
+static int amdgpu_post_soft_reset(struct amdgpu_device *adev)
+{
+ int i, r = 0;
+
+ for (i = 0; i < adev->num_ip_blocks; i++) {
+ if (!adev->ip_block_status[i].valid)
+ continue;
+ if (adev->ip_block_status[i].hang &&
+ adev->ip_blocks[i].funcs->post_soft_reset)
+ r = adev->ip_blocks[i].funcs->post_soft_reset(adev);
+ if (r)
+ return r;
+ }
+
+ return 0;
+}
+
+bool amdgpu_need_backup(struct amdgpu_device *adev)
+{
+ if (adev->flags & AMD_IS_APU)
+ return false;
+
+ return amdgpu_lockup_timeout > 0 ? true : false;
+}
+
+static int amdgpu_recover_vram_from_shadow(struct amdgpu_device *adev,
+ struct amdgpu_ring *ring,
+ struct amdgpu_bo *bo,
+ struct fence **fence)
+{
+ uint32_t domain;
+ int r;
+
+ if (!bo->shadow)
+ return 0;
+
+ r = amdgpu_bo_reserve(bo, false);
+ if (r)
+ return r;
+ domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
+ /* if bo has been evicted, then no need to recover */
+ if (domain == AMDGPU_GEM_DOMAIN_VRAM) {
+ r = amdgpu_bo_restore_from_shadow(adev, ring, bo,
+ NULL, fence, true);
+ if (r) {
+ DRM_ERROR("recover page table failed!\n");
+ goto err;
+ }
+ }
+err:
+ amdgpu_bo_unreserve(bo);
+ return r;
+}
+
/**
* amdgpu_gpu_reset - reset the asic
*
{
int i, r;
int resched;
+ bool need_full_reset;
+
+ if (!amdgpu_check_soft_reset(adev)) {
+ DRM_INFO("No hardware hang detected. Did some blocks stall?\n");
+ return 0;
+ }
atomic_inc(&adev->gpu_reset_counter);
/* after all hw jobs are reset, hw fence is meaningless, so force_completion */
amdgpu_fence_driver_force_completion(adev);
- /* save scratch */
- amdgpu_atombios_scratch_regs_save(adev);
- r = amdgpu_suspend(adev);
+ need_full_reset = amdgpu_need_full_reset(adev);
-retry:
- /* Disable fb access */
- if (adev->mode_info.num_crtc) {
- struct amdgpu_mode_mc_save save;
- amdgpu_display_stop_mc_access(adev, &save);
- amdgpu_wait_for_idle(adev, AMD_IP_BLOCK_TYPE_GMC);
+ if (!need_full_reset) {
+ amdgpu_pre_soft_reset(adev);
+ r = amdgpu_soft_reset(adev);
+ amdgpu_post_soft_reset(adev);
+ if (r || amdgpu_check_soft_reset(adev)) {
+ DRM_INFO("soft reset failed, will fallback to full reset!\n");
+ need_full_reset = true;
+ }
}
- r = amdgpu_asic_reset(adev);
- /* post card */
- amdgpu_atom_asic_init(adev->mode_info.atom_context);
+ if (need_full_reset) {
+ /* save scratch */
+ amdgpu_atombios_scratch_regs_save(adev);
+ r = amdgpu_suspend(adev);
- if (!r) {
- dev_info(adev->dev, "GPU reset succeeded, trying to resume\n");
- r = amdgpu_resume(adev);
+retry:
+ /* Disable fb access */
+ if (adev->mode_info.num_crtc) {
+ struct amdgpu_mode_mc_save save;
+ amdgpu_display_stop_mc_access(adev, &save);
+ amdgpu_wait_for_idle(adev, AMD_IP_BLOCK_TYPE_GMC);
+ }
+
+ r = amdgpu_asic_reset(adev);
+ /* post card */
+ amdgpu_atom_asic_init(adev->mode_info.atom_context);
+
+ if (!r) {
+ dev_info(adev->dev, "GPU reset succeeded, trying to resume\n");
+ r = amdgpu_resume(adev);
+ }
+ /* restore scratch */
+ amdgpu_atombios_scratch_regs_restore(adev);
}
- /* restore scratch */
- amdgpu_atombios_scratch_regs_restore(adev);
if (!r) {
+ amdgpu_irq_gpu_reset_resume_helper(adev);
+ if (need_full_reset && amdgpu_need_backup(adev)) {
+ r = amdgpu_ttm_recover_gart(adev);
+ if (r)
+ DRM_ERROR("gart recovery failed!!!\n");
+ }
r = amdgpu_ib_ring_tests(adev);
if (r) {
dev_err(adev->dev, "ib ring test failed (%d).\n", r);
r = amdgpu_suspend(adev);
+ need_full_reset = true;
goto retry;
}
+ /**
+ * recovery vm page tables, since we cannot depend on VRAM is
+ * consistent after gpu full reset.
+ */
+ if (need_full_reset && amdgpu_need_backup(adev)) {
+ struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring;
+ struct amdgpu_bo *bo, *tmp;
+ struct fence *fence = NULL, *next = NULL;
+
+ DRM_INFO("recover vram bo from shadow\n");
+ mutex_lock(&adev->shadow_list_lock);
+ list_for_each_entry_safe(bo, tmp, &adev->shadow_list, shadow_list) {
+ amdgpu_recover_vram_from_shadow(adev, ring, bo, &next);
+ if (fence) {
+ r = fence_wait(fence, false);
+ if (r) {
+ WARN(r, "recovery from shadow isn't comleted\n");
+ break;
+ }
+ }
+ fence_put(fence);
+ fence = next;
+ }
+ mutex_unlock(&adev->shadow_list_lock);
+ if (fence) {
+ r = fence_wait(fence, false);
+ if (r)
+ WARN(r, "recovery from shadow isn't comleted\n");
+ }
+ fence_put(fence);
+ }
for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
struct amdgpu_ring *ring = adev->rings[i];
if (!ring)
continue;
+
amd_sched_job_recovery(&ring->sched);
kthread_unpark(ring->sched.thread);
}
/* bad news, how to tell it to userspace ? */
dev_info(adev->dev, "GPU reset failed\n");
}
- amdgpu_irq_gpu_reset_resume_helper(adev);
return r;
}
struct amdgpu_device *adev = f->f_inode->i_private;
ssize_t result = 0;
int r;
- bool use_bank;
+ bool pm_pg_lock, use_bank;
unsigned instance_bank, sh_bank, se_bank;
if (size & 0x3 || *pos & 0x3)
return -EINVAL;
+ /* are we reading registers for which a PG lock is necessary? */
+ pm_pg_lock = (*pos >> 23) & 1;
+
if (*pos & (1ULL << 62)) {
se_bank = (*pos >> 24) & 0x3FF;
sh_bank = (*pos >> 34) & 0x3FF;
instance_bank = (*pos >> 44) & 0x3FF;
use_bank = 1;
- *pos &= 0xFFFFFF;
} else {
use_bank = 0;
}
+ *pos &= 0x3FFFF;
+
if (use_bank) {
if (sh_bank >= adev->gfx.config.max_sh_per_se ||
se_bank >= adev->gfx.config.max_shader_engines)
sh_bank, instance_bank);
}
+ if (pm_pg_lock)
+ mutex_lock(&adev->pm.mutex);
+
while (size) {
uint32_t value;
mutex_unlock(&adev->grbm_idx_mutex);
}
+ if (pm_pg_lock)
+ mutex_unlock(&adev->pm.mutex);
+
return result;
}
while (size) {
uint32_t value;
- value = RREG32_SMC(*pos >> 2);
+ value = RREG32_SMC(*pos);
r = put_user(value, (uint32_t *)buf);
if (r)
return r;
if (r)
return r;
- WREG32_SMC(*pos >> 2, value);
+ WREG32_SMC(*pos, value);
result += 4;
buf += 4;
return -ENOMEM;
/* version, increment each time something is added */
- config[no_regs++] = 0;
+ config[no_regs++] = 2;
config[no_regs++] = adev->gfx.config.max_shader_engines;
config[no_regs++] = adev->gfx.config.max_tile_pipes;
config[no_regs++] = adev->gfx.config.max_cu_per_sh;
config[no_regs++] = adev->gfx.config.gb_addr_config;
config[no_regs++] = adev->gfx.config.num_rbs;
+ /* rev==1 */
+ config[no_regs++] = adev->rev_id;
+ config[no_regs++] = adev->pg_flags;
+ config[no_regs++] = adev->cg_flags;
+
+ /* rev==2 */
+ config[no_regs++] = adev->family;
+ config[no_regs++] = adev->external_rev_id;
+
while (size && (*pos < no_regs * 4)) {
uint32_t value;
container_of(cb, struct amdgpu_flip_work, cb);
fence_put(f);
- schedule_work(&work->flip_work);
+ schedule_work(&work->flip_work.work);
}
static bool amdgpu_flip_handle_fence(struct amdgpu_flip_work *work,
static void amdgpu_flip_work_func(struct work_struct *__work)
{
+ struct delayed_work *delayed_work =
+ container_of(__work, struct delayed_work, work);
struct amdgpu_flip_work *work =
- container_of(__work, struct amdgpu_flip_work, flip_work);
+ container_of(delayed_work, struct amdgpu_flip_work, flip_work);
struct amdgpu_device *adev = work->adev;
struct amdgpu_crtc *amdgpuCrtc = adev->mode_info.crtcs[work->crtc_id];
struct drm_crtc *crtc = &amdgpuCrtc->base;
unsigned long flags;
- unsigned i, repcnt = 4;
- int vpos, hpos, stat, min_udelay = 0;
- struct drm_vblank_crtc *vblank = &crtc->dev->vblank[work->crtc_id];
+ unsigned i;
+ int vpos, hpos;
if (amdgpu_flip_handle_fence(work, &work->excl))
return;
if (amdgpu_flip_handle_fence(work, &work->shared[i]))
return;
- /* We borrow the event spin lock for protecting flip_status */
- spin_lock_irqsave(&crtc->dev->event_lock, flags);
-
- /* If this happens to execute within the "virtually extended" vblank
- * interval before the start of the real vblank interval then it needs
- * to delay programming the mmio flip until the real vblank is entered.
- * This prevents completing a flip too early due to the way we fudge
- * our vblank counter and vblank timestamps in order to work around the
- * problem that the hw fires vblank interrupts before actual start of
- * vblank (when line buffer refilling is done for a frame). It
- * complements the fudging logic in amdgpu_get_crtc_scanoutpos() for
- * timestamping and amdgpu_get_vblank_counter_kms() for vblank counts.
- *
- * In practice this won't execute very often unless on very fast
- * machines because the time window for this to happen is very small.
+ /* Wait until we're out of the vertical blank period before the one
+ * targeted by the flip
*/
- while (amdgpuCrtc->enabled && --repcnt) {
- /* GET_DISTANCE_TO_VBLANKSTART returns distance to real vblank
- * start in hpos, and to the "fudged earlier" vblank start in
- * vpos.
- */
- stat = amdgpu_get_crtc_scanoutpos(adev->ddev, work->crtc_id,
- GET_DISTANCE_TO_VBLANKSTART,
- &vpos, &hpos, NULL, NULL,
- &crtc->hwmode);
-
- if ((stat & (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE)) !=
- (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE) ||
- !(vpos >= 0 && hpos <= 0))
- break;
-
- /* Sleep at least until estimated real start of hw vblank */
- min_udelay = (-hpos + 1) * max(vblank->linedur_ns / 1000, 5);
- if (min_udelay > vblank->framedur_ns / 2000) {
- /* Don't wait ridiculously long - something is wrong */
- repcnt = 0;
- break;
- }
- spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
- usleep_range(min_udelay, 2 * min_udelay);
- spin_lock_irqsave(&crtc->dev->event_lock, flags);
+ if (amdgpuCrtc->enabled &&
+ (amdgpu_get_crtc_scanoutpos(adev->ddev, work->crtc_id, 0,
+ &vpos, &hpos, NULL, NULL,
+ &crtc->hwmode)
+ & (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK)) ==
+ (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK) &&
+ (int)(work->target_vblank -
+ amdgpu_get_vblank_counter_kms(adev->ddev, amdgpuCrtc->crtc_id)) > 0) {
+ schedule_delayed_work(&work->flip_work, usecs_to_jiffies(1000));
+ return;
}
- if (!repcnt)
- DRM_DEBUG_DRIVER("Delay problem on crtc %d: min_udelay %d, "
- "framedur %d, linedur %d, stat %d, vpos %d, "
- "hpos %d\n", work->crtc_id, min_udelay,
- vblank->framedur_ns / 1000,
- vblank->linedur_ns / 1000, stat, vpos, hpos);
+ /* We borrow the event spin lock for protecting flip_status */
+ spin_lock_irqsave(&crtc->dev->event_lock, flags);
/* Do the flip (mmio) */
adev->mode_info.funcs->page_flip(adev, work->crtc_id, work->base, work->async);
kfree(work);
}
-int amdgpu_crtc_page_flip(struct drm_crtc *crtc,
- struct drm_framebuffer *fb,
- struct drm_pending_vblank_event *event,
- uint32_t page_flip_flags)
+int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc,
+ struct drm_framebuffer *fb,
+ struct drm_pending_vblank_event *event,
+ uint32_t page_flip_flags, uint32_t target)
{
struct drm_device *dev = crtc->dev;
struct amdgpu_device *adev = dev->dev_private;
if (work == NULL)
return -ENOMEM;
- INIT_WORK(&work->flip_work, amdgpu_flip_work_func);
+ INIT_DELAYED_WORK(&work->flip_work, amdgpu_flip_work_func);
INIT_WORK(&work->unpin_work, amdgpu_unpin_work_func);
work->event = event;
amdgpu_bo_unreserve(new_rbo);
work->base = base;
-
- r = drm_crtc_vblank_get(crtc);
- if (r) {
- DRM_ERROR("failed to get vblank before flip\n");
- goto pflip_cleanup;
- }
+ work->target_vblank = target - drm_crtc_vblank_count(crtc) +
+ amdgpu_get_vblank_counter_kms(dev, work->crtc_id);
/* we borrow the event spin lock for protecting flip_wrok */
spin_lock_irqsave(&crtc->dev->event_lock, flags);
DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
r = -EBUSY;
- goto vblank_cleanup;
+ goto pflip_cleanup;
}
amdgpu_crtc->pflip_status = AMDGPU_FLIP_PENDING;
/* update crtc fb */
crtc->primary->fb = fb;
spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
- amdgpu_flip_work_func(&work->flip_work);
+ amdgpu_flip_work_func(&work->flip_work.work);
return 0;
-vblank_cleanup:
- drm_crtc_vblank_put(crtc);
-
pflip_cleanup:
if (unlikely(amdgpu_bo_reserve(new_rbo, false) != 0)) {
DRM_ERROR("failed to reserve new rbo in error path\n");
return ret;
}
-static const char *encoder_names[38] = {
+static const char *encoder_names[41] = {
"NONE",
"INTERNAL_LVDS",
"INTERNAL_TMDS1",
"TRAVIS",
"INTERNAL_VCE",
"INTERNAL_UNIPHY3",
+ "HDMI_ANX9805",
+ "INTERNAL_AMCLK",
+ "VIRTUAL",
};
static const char *hpd_names[6] = {
* - 3.2.0 - GFX8: Uses EOP_TC_WB_ACTION_EN, so UMDs don't have to do the same
* at the end of IBs.
* - 3.3.0 - Add VM support for UVD on supported hardware.
+ * - 3.4.0 - Add AMDGPU_INFO_NUM_EVICTIONS.
+ * - 3.5.0 - Add support for new UVD_NO_OP register.
+ * - 3.6.0 - kmd involves use CONTEXT_CONTROL in ring buffer.
*/
#define KMS_DRIVER_MAJOR 3
-#define KMS_DRIVER_MINOR 3
+#define KMS_DRIVER_MINOR 6
#define KMS_DRIVER_PATCHLEVEL 0
int amdgpu_vram_limit = 0;
int amdgpu_gart_size = -1; /* auto */
+int amdgpu_moverate = -1; /* auto */
int amdgpu_benchmarking = 0;
int amdgpu_testing = 0;
int amdgpu_audio = -1;
int amdgpu_sched_hw_submission = 2;
int amdgpu_powerplay = -1;
int amdgpu_powercontainment = 1;
+int amdgpu_sclk_deep_sleep_en = 1;
unsigned amdgpu_pcie_gen_cap = 0;
unsigned amdgpu_pcie_lane_cap = 0;
unsigned amdgpu_cg_mask = 0xffffffff;
unsigned amdgpu_pg_mask = 0xffffffff;
char *amdgpu_disable_cu = NULL;
+char *amdgpu_virtual_display = NULL;
+unsigned amdgpu_pp_feature_mask = 0xffffffff;
MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes");
module_param_named(vramlimit, amdgpu_vram_limit, int, 0600);
MODULE_PARM_DESC(gartsize, "Size of PCIE/IGP gart to setup in megabytes (32, 64, etc., -1 = auto)");
module_param_named(gartsize, amdgpu_gart_size, int, 0600);
+MODULE_PARM_DESC(moverate, "Maximum buffer migration rate in MB/s. (32, 64, etc., -1=auto, 0=1=disabled)");
+module_param_named(moverate, amdgpu_moverate, int, 0600);
+
MODULE_PARM_DESC(benchmark, "Run benchmark");
module_param_named(benchmark, amdgpu_benchmarking, int, 0444);
MODULE_PARM_DESC(powercontainment, "Power Containment (1 = enable (default), 0 = disable)");
module_param_named(powercontainment, amdgpu_powercontainment, int, 0444);
+
+MODULE_PARM_DESC(ppfeaturemask, "all power features enabled (default))");
+module_param_named(ppfeaturemask, amdgpu_pp_feature_mask, int, 0444);
#endif
+MODULE_PARM_DESC(sclkdeepsleep, "SCLK Deep Sleep (1 = enable (default), 0 = disable)");
+module_param_named(sclkdeepsleep, amdgpu_sclk_deep_sleep_en, int, 0444);
+
MODULE_PARM_DESC(pcie_gen_cap, "PCIE Gen Caps (0: autodetect (default))");
module_param_named(pcie_gen_cap, amdgpu_pcie_gen_cap, uint, 0444);
MODULE_PARM_DESC(disable_cu, "Disable CUs (se.sh.cu,...)");
module_param_named(disable_cu, amdgpu_disable_cu, charp, 0444);
+MODULE_PARM_DESC(virtual_display, "Enable virtual display feature (the virtual_display will be set like xxxx:xx:xx.x;xxxx:xx:xx.x)");
+module_param_named(virtual_display, amdgpu_virtual_display, charp, 0444);
+
static const struct pci_device_id pciidlist[] = {
+#ifdef CONFIG_DRM_AMDGPU_SI
+ {0x1002, 0x6780, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x6784, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x6788, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x678A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x6790, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x6791, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x6792, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x6798, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x6799, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x679A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x679B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x679E, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x679F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
+ {0x1002, 0x6800, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN|AMD_IS_MOBILITY},
+ {0x1002, 0x6801, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN|AMD_IS_MOBILITY},
+ {0x1002, 0x6802, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN|AMD_IS_MOBILITY},
+ {0x1002, 0x6806, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6808, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6809, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6810, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6811, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6816, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6817, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6818, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6819, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_PITCAIRN},
+ {0x1002, 0x6600, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6601, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6602, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6603, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6604, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6605, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6606, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6607, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6608, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND},
+ {0x1002, 0x6610, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND},
+ {0x1002, 0x6611, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND},
+ {0x1002, 0x6613, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND},
+ {0x1002, 0x6617, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6620, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6621, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6623, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND|AMD_IS_MOBILITY},
+ {0x1002, 0x6631, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_OLAND},
+ {0x1002, 0x6820, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6821, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6822, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6823, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6824, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6825, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6826, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6827, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6828, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x6829, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x682A, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x682B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x682C, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x682D, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x682F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6830, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6831, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE|AMD_IS_MOBILITY},
+ {0x1002, 0x6835, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x6837, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x6838, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x6839, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x683B, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x683D, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x683F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VERDE},
+ {0x1002, 0x6660, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY},
+ {0x1002, 0x6663, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY},
+ {0x1002, 0x6664, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY},
+ {0x1002, 0x6665, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY},
+ {0x1002, 0x6667, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY},
+ {0x1002, 0x666F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_HAINAN|AMD_IS_MOBILITY},
+#endif
#ifdef CONFIG_DRM_AMDGPU_CIK
/* Kaveri */
{0x1002, 0x1304, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_KAVERI|AMD_IS_MOBILITY|AMD_IS_APU},
drm_put_dev(dev);
}
+static void
+amdgpu_pci_shutdown(struct pci_dev *pdev)
+{
+ struct drm_device *dev = pci_get_drvdata(pdev);
+ struct amdgpu_device *adev = dev->dev_private;
+
+ /* if we are running in a VM, make sure the device
+ * torn down properly on reboot/shutdown
+ */
+ if (adev->virtualization.is_virtual)
+ amdgpu_pci_remove(pdev);
+}
+
static int amdgpu_pmops_suspend(struct device *dev)
{
struct pci_dev *pdev = to_pci_dev(dev);
+
struct drm_device *drm_dev = pci_get_drvdata(pdev);
- return amdgpu_suspend_kms(drm_dev, true, true);
+ return amdgpu_device_suspend(drm_dev, true, true);
}
static int amdgpu_pmops_resume(struct device *dev)
{
struct pci_dev *pdev = to_pci_dev(dev);
struct drm_device *drm_dev = pci_get_drvdata(pdev);
- return amdgpu_resume_kms(drm_dev, true, true);
+
+ /* GPU comes up enabled by the bios on resume */
+ if (amdgpu_device_is_px(drm_dev)) {
+ pm_runtime_disable(dev);
+ pm_runtime_set_active(dev);
+ pm_runtime_enable(dev);
+ }
+
+ return amdgpu_device_resume(drm_dev, true, true);
}
static int amdgpu_pmops_freeze(struct device *dev)
{
struct pci_dev *pdev = to_pci_dev(dev);
+
struct drm_device *drm_dev = pci_get_drvdata(pdev);
- return amdgpu_suspend_kms(drm_dev, false, true);
+ return amdgpu_device_suspend(drm_dev, false, true);
}
static int amdgpu_pmops_thaw(struct device *dev)
{
struct pci_dev *pdev = to_pci_dev(dev);
+
+ struct drm_device *drm_dev = pci_get_drvdata(pdev);
+ return amdgpu_device_resume(drm_dev, false, true);
+}
+
+static int amdgpu_pmops_poweroff(struct device *dev)
+{
+ struct pci_dev *pdev = to_pci_dev(dev);
+
+ struct drm_device *drm_dev = pci_get_drvdata(pdev);
+ return amdgpu_device_suspend(drm_dev, true, true);
+}
+
+static int amdgpu_pmops_restore(struct device *dev)
+{
+ struct pci_dev *pdev = to_pci_dev(dev);
+
struct drm_device *drm_dev = pci_get_drvdata(pdev);
- return amdgpu_resume_kms(drm_dev, false, true);
+ return amdgpu_device_resume(drm_dev, false, true);
}
static int amdgpu_pmops_runtime_suspend(struct device *dev)
drm_kms_helper_poll_disable(drm_dev);
vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_OFF);
- ret = amdgpu_suspend_kms(drm_dev, false, false);
+ ret = amdgpu_device_suspend(drm_dev, false, false);
pci_save_state(pdev);
pci_disable_device(pdev);
pci_ignore_hotplug(pdev);
return ret;
pci_set_master(pdev);
- ret = amdgpu_resume_kms(drm_dev, false, false);
+ ret = amdgpu_device_resume(drm_dev, false, false);
drm_kms_helper_poll_enable(drm_dev);
vga_switcheroo_set_dynamic_switch(pdev, VGA_SWITCHEROO_ON);
drm_dev->switch_power_state = DRM_SWITCH_POWER_ON;
.resume = amdgpu_pmops_resume,
.freeze = amdgpu_pmops_freeze,
.thaw = amdgpu_pmops_thaw,
- .poweroff = amdgpu_pmops_freeze,
- .restore = amdgpu_pmops_resume,
+ .poweroff = amdgpu_pmops_poweroff,
+ .restore = amdgpu_pmops_restore,
.runtime_suspend = amdgpu_pmops_runtime_suspend,
.runtime_resume = amdgpu_pmops_runtime_resume,
.runtime_idle = amdgpu_pmops_runtime_idle,
.id_table = pciidlist,
.probe = amdgpu_pci_probe,
.remove = amdgpu_pci_remove,
+ .shutdown = amdgpu_pci_shutdown,
.driver.pm = &amdgpu_pm_ops,
};
*/
#include <linux/module.h>
#include <linux/slab.h>
+#include <linux/pm_runtime.h>
#include <drm/drmP.h>
#include <drm/drm_crtc.h>
struct amdgpu_device *adev;
};
+static int
+amdgpufb_open(struct fb_info *info, int user)
+{
+ struct amdgpu_fbdev *rfbdev = info->par;
+ struct amdgpu_device *adev = rfbdev->adev;
+ int ret = pm_runtime_get_sync(adev->ddev->dev);
+ if (ret < 0 && ret != -EACCES) {
+ pm_runtime_mark_last_busy(adev->ddev->dev);
+ pm_runtime_put_autosuspend(adev->ddev->dev);
+ return ret;
+ }
+ return 0;
+}
+
+static int
+amdgpufb_release(struct fb_info *info, int user)
+{
+ struct amdgpu_fbdev *rfbdev = info->par;
+ struct amdgpu_device *adev = rfbdev->adev;
+
+ pm_runtime_mark_last_busy(adev->ddev->dev);
+ pm_runtime_put_autosuspend(adev->ddev->dev);
+ return 0;
+}
+
static struct fb_ops amdgpufb_ops = {
.owner = THIS_MODULE,
+ .fb_open = amdgpufb_open,
+ .fb_release = amdgpufb_release,
.fb_check_var = drm_fb_helper_check_var,
.fb_set_par = drm_fb_helper_set_par,
.fb_fillrect = drm_fb_helper_cfb_fillrect,
#define AMDGPU_GWS_SHIFT PAGE_SHIFT
#define AMDGPU_OA_SHIFT PAGE_SHIFT
-#define AMDGPU_PL_GDS TTM_PL_PRIV0
-#define AMDGPU_PL_GWS TTM_PL_PRIV1
-#define AMDGPU_PL_OA TTM_PL_PRIV2
-
-#define AMDGPU_PL_FLAG_GDS TTM_PL_FLAG_PRIV0
-#define AMDGPU_PL_FLAG_GWS TTM_PL_FLAG_PRIV1
-#define AMDGPU_PL_FLAG_OA TTM_PL_FLAG_PRIV2
-
struct amdgpu_ring;
struct amdgpu_bo;
"AMDGPU i2c hw bus %s", name);
i2c->adapter.algo = &amdgpu_atombios_i2c_algo;
ret = i2c_add_adapter(&i2c->adapter);
- if (ret) {
- DRM_ERROR("Failed to register hw i2c %s\n", name);
+ if (ret)
goto out_free;
- }
} else {
/* set the amdgpu bit adapter */
snprintf(i2c->adapter.name, sizeof(i2c->adapter.name),
bool skip_preamble, need_ctx_switch;
unsigned patch_offset = ~0;
struct amdgpu_vm *vm;
- uint64_t ctx;
+ uint64_t fence_ctx;
+ uint32_t status = 0, alloc_size;
unsigned i;
int r = 0;
/* ring tests don't use a job */
if (job) {
vm = job->vm;
- ctx = job->ctx;
+ fence_ctx = job->fence_ctx;
} else {
vm = NULL;
- ctx = 0;
+ fence_ctx = 0;
}
if (!ring->ready) {
- dev_err(adev->dev, "couldn't schedule ib\n");
+ dev_err(adev->dev, "couldn't schedule ib on ring <%s>\n", ring->name);
return -EINVAL;
}
return -EINVAL;
}
- r = amdgpu_ring_alloc(ring, 256 * num_ibs);
+ alloc_size = amdgpu_ring_get_dma_frame_size(ring) +
+ num_ibs * amdgpu_ring_get_emit_ib_size(ring);
+
+ r = amdgpu_ring_alloc(ring, alloc_size);
if (r) {
dev_err(adev->dev, "scheduling IB failed (%d).\n", r);
return r;
/* always set cond_exec_polling to CONTINUE */
*ring->cond_exe_cpu_addr = 1;
- skip_preamble = ring->current_ctx == ctx;
- need_ctx_switch = ring->current_ctx != ctx;
+ skip_preamble = ring->current_ctx == fence_ctx;
+ need_ctx_switch = ring->current_ctx != fence_ctx;
+ if (job && ring->funcs->emit_cntxcntl) {
+ if (need_ctx_switch)
+ status |= AMDGPU_HAVE_CTX_SWITCH;
+ status |= job->preamble_status;
+ amdgpu_ring_emit_cntxcntl(ring, status);
+ }
+
for (i = 0; i < num_ibs; ++i) {
ib = &ibs[i];
/* drop preamble IBs if we don't have a context switch */
- if ((ib->flags & AMDGPU_IB_FLAG_PREAMBLE) && skip_preamble)
+ if ((ib->flags & AMDGPU_IB_FLAG_PREAMBLE) &&
+ skip_preamble &&
+ !(status & AMDGPU_PREAMBLE_IB_PRESENT_FIRST))
continue;
amdgpu_ring_emit_ib(ring, ib, job ? job->vm_id : 0,
if (patch_offset != ~0 && ring->funcs->patch_cond_exec)
amdgpu_ring_patch_cond_exec(ring, patch_offset);
- ring->current_ctx = ctx;
+ ring->current_ctx = fence_ctx;
+ if (ring->funcs->emit_switch_buffer)
+ amdgpu_ring_emit_switch_buffer(ring);
amdgpu_ring_commit(ring);
return 0;
}
/* Allocate ring buffer */
if (adev->irq.ih.ring_obj == NULL) {
- r = amdgpu_bo_create(adev, adev->irq.ih.ring_size,
- PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_GTT, 0,
- NULL, NULL, &adev->irq.ih.ring_obj);
+ r = amdgpu_bo_create_kernel(adev, adev->irq.ih.ring_size,
+ PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT,
+ &adev->irq.ih.ring_obj,
+ &adev->irq.ih.gpu_addr,
+ (void **)&adev->irq.ih.ring);
if (r) {
DRM_ERROR("amdgpu: failed to create ih ring buffer (%d).\n", r);
return r;
}
- r = amdgpu_bo_reserve(adev->irq.ih.ring_obj, false);
- if (unlikely(r != 0))
- return r;
- r = amdgpu_bo_pin(adev->irq.ih.ring_obj,
- AMDGPU_GEM_DOMAIN_GTT,
- &adev->irq.ih.gpu_addr);
- if (r) {
- amdgpu_bo_unreserve(adev->irq.ih.ring_obj);
- DRM_ERROR("amdgpu: failed to pin ih ring buffer (%d).\n", r);
- return r;
- }
- r = amdgpu_bo_kmap(adev->irq.ih.ring_obj,
- (void **)&adev->irq.ih.ring);
- amdgpu_bo_unreserve(adev->irq.ih.ring_obj);
- if (r) {
- DRM_ERROR("amdgpu: failed to map ih ring buffer (%d).\n", r);
- return r;
- }
}
return 0;
}
*/
void amdgpu_ih_ring_fini(struct amdgpu_device *adev)
{
- int r;
-
if (adev->irq.ih.use_bus_addr) {
if (adev->irq.ih.ring) {
/* add 8 bytes for the rptr/wptr shadows and
adev->irq.ih.ring = NULL;
}
} else {
- if (adev->irq.ih.ring_obj) {
- r = amdgpu_bo_reserve(adev->irq.ih.ring_obj, false);
- if (likely(r == 0)) {
- amdgpu_bo_kunmap(adev->irq.ih.ring_obj);
- amdgpu_bo_unpin(adev->irq.ih.ring_obj);
- amdgpu_bo_unreserve(adev->irq.ih.ring_obj);
- }
- amdgpu_bo_unref(&adev->irq.ih.ring_obj);
- adev->irq.ih.ring = NULL;
- adev->irq.ih.ring_obj = NULL;
- }
+ amdgpu_bo_free_kernel(&adev->irq.ih.ring_obj,
+ &adev->irq.ih.gpu_addr,
+ (void **)&adev->irq.ih.ring);
amdgpu_wb_free(adev, adev->irq.ih.wptr_offs);
amdgpu_wb_free(adev, adev->irq.ih.rptr_offs);
}
/* gen irq stuff */
struct irq_domain *domain; /* GPU irq controller domain */
unsigned virq[AMDGPU_MAX_IRQ_SRC_ID];
+ uint32_t srbm_soft_reset;
};
void amdgpu_irq_preinstall(struct drm_device *dev);
amdgpu_ib_free(job->adev, &job->ibs[i], f);
}
-void amdgpu_job_free_cb(struct amd_sched_job *s_job)
+static void amdgpu_job_free_cb(struct amd_sched_job *s_job)
{
struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base);
return r;
job->owner = owner;
- job->ctx = entity->fence_context;
+ job->fence_ctx = entity->fence_context;
*f = fence_get(&job->base.s_fence->finished);
amdgpu_job_free_resources(job);
amd_sched_entity_push_job(&job->base);
type = AMD_IP_BLOCK_TYPE_UVD;
ring_mask = adev->uvd.ring.ready ? 1 : 0;
ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
- ib_size_alignment = 8;
+ ib_size_alignment = 16;
break;
case AMDGPU_HW_IP_VCE:
type = AMD_IP_BLOCK_TYPE_VCE;
- for (i = 0; i < AMDGPU_MAX_VCE_RINGS; i++)
+ for (i = 0; i < adev->vce.num_rings; i++)
ring_mask |= ((adev->vce.ring[i].ready ? 1 : 0) << i);
ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
- ib_size_alignment = 8;
+ ib_size_alignment = 1;
break;
default:
return -EINVAL;
case AMDGPU_INFO_NUM_BYTES_MOVED:
ui64 = atomic64_read(&adev->num_bytes_moved);
return copy_to_user(out, &ui64, min(size, 8u)) ? -EFAULT : 0;
+ case AMDGPU_INFO_NUM_EVICTIONS:
+ ui64 = atomic64_read(&adev->num_evictions);
+ return copy_to_user(out, &ui64, min(size, 8u)) ? -EFAULT : 0;
case AMDGPU_INFO_VRAM_USAGE:
ui64 = atomic64_read(&adev->vram_usage);
return copy_to_user(out, &ui64, min(size, 8u)) ? -EFAULT : 0;
return r;
fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL);
- if (unlikely(!fpriv))
- return -ENOMEM;
+ if (unlikely(!fpriv)) {
+ r = -ENOMEM;
+ goto out_suspend;
+ }
r = amdgpu_vm_init(adev, &fpriv->vm);
- if (r)
- goto error_free;
+ if (r) {
+ kfree(fpriv);
+ goto out_suspend;
+ }
mutex_init(&fpriv->bo_list_lock);
idr_init(&fpriv->bo_list_handles);
file_priv->driver_priv = fpriv;
+out_suspend:
pm_runtime_mark_last_busy(dev->dev);
pm_runtime_put_autosuspend(dev->dev);
- return 0;
-
-error_free:
- kfree(fpriv);
return r;
}
kfree(fpriv);
file_priv->driver_priv = NULL;
+
+ pm_runtime_mark_last_busy(dev->dev);
+ pm_runtime_put_autosuspend(dev->dev);
}
/**
void amdgpu_driver_preclose_kms(struct drm_device *dev,
struct drm_file *file_priv)
{
+ pm_runtime_get_sync(dev->dev);
}
/*
#include <drm/drm_plane_helper.h>
#include <linux/i2c.h>
#include <linux/i2c-algo-bit.h>
+#include <linux/hrtimer.h>
+#include "amdgpu_irq.h"
struct amdgpu_bo;
struct amdgpu_device;
int num_dig; /* number of dig blocks */
int disp_priority;
const struct amdgpu_display_funcs *funcs;
+ struct hrtimer vblank_timer;
+ enum amdgpu_interrupt_state vsync_timer_enabled;
};
#define AMDGPU_MAX_BL_LEVEL 0xFF
void amdgpu_print_display_setup(struct drm_device *dev);
int amdgpu_modeset_create_props(struct amdgpu_device *adev);
int amdgpu_crtc_set_config(struct drm_mode_set *set);
-int amdgpu_crtc_page_flip(struct drm_crtc *crtc,
- struct drm_framebuffer *fb,
- struct drm_pending_vblank_event *event,
- uint32_t page_flip_flags);
+int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc,
+ struct drm_framebuffer *fb,
+ struct drm_pending_vblank_event *event,
+ uint32_t page_flip_flags, uint32_t target);
extern const struct drm_mode_config_funcs amdgpu_mode_funcs;
#endif
#include "amdgpu_trace.h"
-int amdgpu_ttm_init(struct amdgpu_device *adev);
-void amdgpu_ttm_fini(struct amdgpu_device *adev);
static u64 amdgpu_get_vis_part_size(struct amdgpu_device *adev,
struct ttm_mem_reg *mem)
{
- u64 ret = 0;
- if (mem->start << PAGE_SHIFT < adev->mc.visible_vram_size) {
- ret = (u64)((mem->start << PAGE_SHIFT) + mem->size) >
- adev->mc.visible_vram_size ?
- adev->mc.visible_vram_size - (mem->start << PAGE_SHIFT) :
- mem->size;
- }
- return ret;
+ if (mem->start << PAGE_SHIFT >= adev->mc.visible_vram_size)
+ return 0;
+
+ return ((mem->start << PAGE_SHIFT) + mem->size) >
+ adev->mc.visible_vram_size ?
+ adev->mc.visible_vram_size - (mem->start << PAGE_SHIFT) :
+ mem->size;
}
static void amdgpu_update_memory_usage(struct amdgpu_device *adev,
drm_gem_object_release(&bo->gem_base);
amdgpu_bo_unref(&bo->parent);
+ if (!list_empty(&bo->shadow_list)) {
+ mutex_lock(&bo->adev->shadow_list_lock);
+ list_del_init(&bo->shadow_list);
+ mutex_unlock(&bo->adev->shadow_list_lock);
+ }
kfree(bo->metadata);
kfree(bo);
}
static void amdgpu_ttm_placement_init(struct amdgpu_device *adev,
struct ttm_placement *placement,
- struct ttm_place *placements,
+ struct ttm_place *places,
u32 domain, u64 flags)
{
- u32 c = 0, i;
-
- placement->placement = placements;
- placement->busy_placement = placements;
+ u32 c = 0;
if (domain & AMDGPU_GEM_DOMAIN_VRAM) {
+ unsigned visible_pfn = adev->mc.visible_vram_size >> PAGE_SHIFT;
+
if (flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS &&
- adev->mc.visible_vram_size < adev->mc.real_vram_size) {
- placements[c].fpfn =
- adev->mc.visible_vram_size >> PAGE_SHIFT;
- placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
- TTM_PL_FLAG_VRAM | TTM_PL_FLAG_TOPDOWN;
+ !(flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) &&
+ adev->mc.visible_vram_size < adev->mc.real_vram_size) {
+ places[c].fpfn = visible_pfn;
+ places[c].lpfn = 0;
+ places[c].flags = TTM_PL_FLAG_WC |
+ TTM_PL_FLAG_UNCACHED | TTM_PL_FLAG_VRAM |
+ TTM_PL_FLAG_TOPDOWN;
+ c++;
}
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
+
+ places[c].fpfn = 0;
+ places[c].lpfn = 0;
+ places[c].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_UNCACHED |
TTM_PL_FLAG_VRAM;
- if (!(flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED))
- placements[c - 1].flags |= TTM_PL_FLAG_TOPDOWN;
+ if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)
+ places[c].lpfn = visible_pfn;
+ else
+ places[c].flags |= TTM_PL_FLAG_TOPDOWN;
+ c++;
}
if (domain & AMDGPU_GEM_DOMAIN_GTT) {
- if (flags & AMDGPU_GEM_CREATE_CPU_GTT_USWC) {
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_TT |
+ places[c].fpfn = 0;
+ places[c].lpfn = 0;
+ places[c].flags = TTM_PL_FLAG_TT;
+ if (flags & AMDGPU_GEM_CREATE_CPU_GTT_USWC)
+ places[c].flags |= TTM_PL_FLAG_WC |
TTM_PL_FLAG_UNCACHED;
- } else {
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_TT;
- }
+ else
+ places[c].flags |= TTM_PL_FLAG_CACHED;
+ c++;
}
if (domain & AMDGPU_GEM_DOMAIN_CPU) {
- if (flags & AMDGPU_GEM_CREATE_CPU_GTT_USWC) {
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_FLAG_WC | TTM_PL_FLAG_SYSTEM |
+ places[c].fpfn = 0;
+ places[c].lpfn = 0;
+ places[c].flags = TTM_PL_FLAG_SYSTEM;
+ if (flags & AMDGPU_GEM_CREATE_CPU_GTT_USWC)
+ places[c].flags |= TTM_PL_FLAG_WC |
TTM_PL_FLAG_UNCACHED;
- } else {
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_SYSTEM;
- }
+ else
+ places[c].flags |= TTM_PL_FLAG_CACHED;
+ c++;
}
if (domain & AMDGPU_GEM_DOMAIN_GDS) {
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_FLAG_UNCACHED |
- AMDGPU_PL_FLAG_GDS;
+ places[c].fpfn = 0;
+ places[c].lpfn = 0;
+ places[c].flags = TTM_PL_FLAG_UNCACHED | AMDGPU_PL_FLAG_GDS;
+ c++;
}
+
if (domain & AMDGPU_GEM_DOMAIN_GWS) {
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_FLAG_UNCACHED |
- AMDGPU_PL_FLAG_GWS;
+ places[c].fpfn = 0;
+ places[c].lpfn = 0;
+ places[c].flags = TTM_PL_FLAG_UNCACHED | AMDGPU_PL_FLAG_GWS;
+ c++;
}
+
if (domain & AMDGPU_GEM_DOMAIN_OA) {
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_FLAG_UNCACHED |
- AMDGPU_PL_FLAG_OA;
+ places[c].fpfn = 0;
+ places[c].lpfn = 0;
+ places[c].flags = TTM_PL_FLAG_UNCACHED | AMDGPU_PL_FLAG_OA;
+ c++;
}
if (!c) {
- placements[c].fpfn = 0;
- placements[c++].flags = TTM_PL_MASK_CACHING |
- TTM_PL_FLAG_SYSTEM;
+ places[c].fpfn = 0;
+ places[c].lpfn = 0;
+ places[c].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM;
+ c++;
}
+
placement->num_placement = c;
- placement->num_busy_placement = c;
+ placement->placement = places;
- for (i = 0; i < c; i++) {
- if ((flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) &&
- (placements[i].flags & TTM_PL_FLAG_VRAM) &&
- !placements[i].fpfn)
- placements[i].lpfn =
- adev->mc.visible_vram_size >> PAGE_SHIFT;
- else
- placements[i].lpfn = 0;
- }
+ placement->num_busy_placement = c;
+ placement->busy_placement = places;
}
void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *rbo, u32 domain)
bo->placement.busy_placement = bo->placements;
}
+/**
+ * amdgpu_bo_create_kernel - create BO for kernel use
+ *
+ * @adev: amdgpu device object
+ * @size: size for the new BO
+ * @align: alignment for the new BO
+ * @domain: where to place it
+ * @bo_ptr: resulting BO
+ * @gpu_addr: GPU addr of the pinned BO
+ * @cpu_addr: optional CPU address mapping
+ *
+ * Allocates and pins a BO for kernel internal use.
+ *
+ * Returns 0 on success, negative error code otherwise.
+ */
+int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
+ unsigned long size, int align,
+ u32 domain, struct amdgpu_bo **bo_ptr,
+ u64 *gpu_addr, void **cpu_addr)
+{
+ int r;
+
+ r = amdgpu_bo_create(adev, size, align, true, domain,
+ AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
+ NULL, NULL, bo_ptr);
+ if (r) {
+ dev_err(adev->dev, "(%d) failed to allocate kernel bo\n", r);
+ return r;
+ }
+
+ r = amdgpu_bo_reserve(*bo_ptr, false);
+ if (r) {
+ dev_err(adev->dev, "(%d) failed to reserve kernel bo\n", r);
+ goto error_free;
+ }
+
+ r = amdgpu_bo_pin(*bo_ptr, domain, gpu_addr);
+ if (r) {
+ dev_err(adev->dev, "(%d) kernel bo pin failed\n", r);
+ goto error_unreserve;
+ }
+
+ if (cpu_addr) {
+ r = amdgpu_bo_kmap(*bo_ptr, cpu_addr);
+ if (r) {
+ dev_err(adev->dev, "(%d) kernel bo map failed\n", r);
+ goto error_unreserve;
+ }
+ }
+
+ amdgpu_bo_unreserve(*bo_ptr);
+
+ return 0;
+
+error_unreserve:
+ amdgpu_bo_unreserve(*bo_ptr);
+
+error_free:
+ amdgpu_bo_unref(bo_ptr);
+
+ return r;
+}
+
+/**
+ * amdgpu_bo_free_kernel - free BO for kernel use
+ *
+ * @bo: amdgpu BO to free
+ *
+ * unmaps and unpin a BO for kernel internal use.
+ */
+void amdgpu_bo_free_kernel(struct amdgpu_bo **bo, u64 *gpu_addr,
+ void **cpu_addr)
+{
+ if (*bo == NULL)
+ return;
+
+ if (likely(amdgpu_bo_reserve(*bo, false) == 0)) {
+ if (cpu_addr)
+ amdgpu_bo_kunmap(*bo);
+
+ amdgpu_bo_unpin(*bo);
+ amdgpu_bo_unreserve(*bo);
+ }
+ amdgpu_bo_unref(bo);
+
+ if (gpu_addr)
+ *gpu_addr = 0;
+
+ if (cpu_addr)
+ *cpu_addr = NULL;
+}
+
int amdgpu_bo_create_restricted(struct amdgpu_device *adev,
unsigned long size, int byte_align,
bool kernel, u32 domain, u64 flags,
}
bo->adev = adev;
INIT_LIST_HEAD(&bo->list);
+ INIT_LIST_HEAD(&bo->shadow_list);
INIT_LIST_HEAD(&bo->va);
bo->prefered_domains = domain & (AMDGPU_GEM_DOMAIN_VRAM |
AMDGPU_GEM_DOMAIN_GTT |
if (unlikely(r != 0)) {
return r;
}
+
+ if (flags & AMDGPU_GEM_CREATE_VRAM_CLEARED &&
+ bo->tbo.mem.placement & TTM_PL_FLAG_VRAM) {
+ struct fence *fence;
+
+ if (adev->mman.buffer_funcs_ring == NULL ||
+ !adev->mman.buffer_funcs_ring->ready) {
+ r = -EBUSY;
+ goto fail_free;
+ }
+
+ r = amdgpu_bo_reserve(bo, false);
+ if (unlikely(r != 0))
+ goto fail_free;
+
+ amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM);
+ r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
+ if (unlikely(r != 0))
+ goto fail_unreserve;
+
+ amdgpu_fill_buffer(bo, 0, bo->tbo.resv, &fence);
+ amdgpu_bo_fence(bo, fence, false);
+ amdgpu_bo_unreserve(bo);
+ fence_put(bo->tbo.moving);
+ bo->tbo.moving = fence_get(fence);
+ fence_put(fence);
+ }
*bo_ptr = bo;
trace_amdgpu_bo_create(bo);
return 0;
+
+fail_unreserve:
+ amdgpu_bo_unreserve(bo);
+fail_free:
+ amdgpu_bo_unref(&bo);
+ return r;
+}
+
+static int amdgpu_bo_create_shadow(struct amdgpu_device *adev,
+ unsigned long size, int byte_align,
+ struct amdgpu_bo *bo)
+{
+ struct ttm_placement placement = {0};
+ struct ttm_place placements[AMDGPU_GEM_DOMAIN_MAX + 1];
+ int r;
+
+ if (bo->shadow)
+ return 0;
+
+ bo->flags |= AMDGPU_GEM_CREATE_SHADOW;
+ memset(&placements, 0,
+ (AMDGPU_GEM_DOMAIN_MAX + 1) * sizeof(struct ttm_place));
+
+ amdgpu_ttm_placement_init(adev, &placement,
+ placements, AMDGPU_GEM_DOMAIN_GTT,
+ AMDGPU_GEM_CREATE_CPU_GTT_USWC);
+
+ r = amdgpu_bo_create_restricted(adev, size, byte_align, true,
+ AMDGPU_GEM_DOMAIN_GTT,
+ AMDGPU_GEM_CREATE_CPU_GTT_USWC,
+ NULL, &placement,
+ bo->tbo.resv,
+ &bo->shadow);
+ if (!r) {
+ bo->shadow->parent = amdgpu_bo_ref(bo);
+ mutex_lock(&adev->shadow_list_lock);
+ list_add_tail(&bo->shadow_list, &adev->shadow_list);
+ mutex_unlock(&adev->shadow_list_lock);
+ }
+
+ return r;
}
int amdgpu_bo_create(struct amdgpu_device *adev,
{
struct ttm_placement placement = {0};
struct ttm_place placements[AMDGPU_GEM_DOMAIN_MAX + 1];
+ int r;
memset(&placements, 0,
(AMDGPU_GEM_DOMAIN_MAX + 1) * sizeof(struct ttm_place));
amdgpu_ttm_placement_init(adev, &placement,
placements, domain, flags);
- return amdgpu_bo_create_restricted(adev, size, byte_align, kernel,
- domain, flags, sg, &placement,
- resv, bo_ptr);
+ r = amdgpu_bo_create_restricted(adev, size, byte_align, kernel,
+ domain, flags, sg, &placement,
+ resv, bo_ptr);
+ if (r)
+ return r;
+
+ if (amdgpu_need_backup(adev) && (flags & AMDGPU_GEM_CREATE_SHADOW)) {
+ r = amdgpu_bo_create_shadow(adev, size, byte_align, (*bo_ptr));
+ if (r)
+ amdgpu_bo_unref(bo_ptr);
+ }
+
+ return r;
+}
+
+int amdgpu_bo_backup_to_shadow(struct amdgpu_device *adev,
+ struct amdgpu_ring *ring,
+ struct amdgpu_bo *bo,
+ struct reservation_object *resv,
+ struct fence **fence,
+ bool direct)
+
+{
+ struct amdgpu_bo *shadow = bo->shadow;
+ uint64_t bo_addr, shadow_addr;
+ int r;
+
+ if (!shadow)
+ return -EINVAL;
+
+ bo_addr = amdgpu_bo_gpu_offset(bo);
+ shadow_addr = amdgpu_bo_gpu_offset(bo->shadow);
+
+ r = reservation_object_reserve_shared(bo->tbo.resv);
+ if (r)
+ goto err;
+
+ r = amdgpu_copy_buffer(ring, bo_addr, shadow_addr,
+ amdgpu_bo_size(bo), resv, fence,
+ direct);
+ if (!r)
+ amdgpu_bo_fence(bo, *fence, true);
+
+err:
+ return r;
+}
+
+int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev,
+ struct amdgpu_ring *ring,
+ struct amdgpu_bo *bo,
+ struct reservation_object *resv,
+ struct fence **fence,
+ bool direct)
+
+{
+ struct amdgpu_bo *shadow = bo->shadow;
+ uint64_t bo_addr, shadow_addr;
+ int r;
+
+ if (!shadow)
+ return -EINVAL;
+
+ bo_addr = amdgpu_bo_gpu_offset(bo);
+ shadow_addr = amdgpu_bo_gpu_offset(bo->shadow);
+
+ r = reservation_object_reserve_shared(bo->tbo.resv);
+ if (r)
+ goto err;
+
+ r = amdgpu_copy_buffer(ring, shadow_addr, bo_addr,
+ amdgpu_bo_size(bo), resv, fence,
+ direct);
+ if (!r)
+ amdgpu_bo_fence(bo, *fence, true);
+
+err:
+ return r;
}
int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr)
return -EINVAL;
if (bo->pin_count) {
+ uint32_t mem_type = bo->tbo.mem.mem_type;
+
+ if (domain != amdgpu_mem_type_to_domain(mem_type))
+ return -EINVAL;
+
bo->pin_count++;
if (gpu_addr)
*gpu_addr = amdgpu_bo_gpu_offset(bo);
if (max_offset != 0) {
- u64 domain_start;
- if (domain == AMDGPU_GEM_DOMAIN_VRAM)
- domain_start = bo->adev->mc.vram_start;
- else
- domain_start = bo->adev->mc.gtt_start;
+ u64 domain_start = bo->tbo.bdev->man[mem_type].gpu_offset;
WARN_ON_ONCE(max_offset <
(amdgpu_bo_gpu_offset(bo) - domain_start));
}
/* force to pin into visible video ram */
if ((bo->placements[i].flags & TTM_PL_FLAG_VRAM) &&
!(bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS) &&
- (!max_offset || max_offset > bo->adev->mc.visible_vram_size)) {
+ (!max_offset || max_offset >
+ bo->adev->mc.visible_vram_size)) {
if (WARN_ON_ONCE(min_offset >
bo->adev->mc.visible_vram_size))
return -EINVAL;
}
r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
- if (likely(r == 0)) {
- bo->pin_count = 1;
- if (gpu_addr != NULL)
- *gpu_addr = amdgpu_bo_gpu_offset(bo);
- if (domain == AMDGPU_GEM_DOMAIN_VRAM) {
- bo->adev->vram_pin_size += amdgpu_bo_size(bo);
- if (bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS)
- bo->adev->invisible_pin_size += amdgpu_bo_size(bo);
- } else
- bo->adev->gart_pin_size += amdgpu_bo_size(bo);
- } else {
+ if (unlikely(r)) {
dev_err(bo->adev->dev, "%p pin failed\n", bo);
+ goto error;
}
+ r = amdgpu_ttm_bind(bo->tbo.ttm, &bo->tbo.mem);
+ if (unlikely(r)) {
+ dev_err(bo->adev->dev, "%p bind failed\n", bo);
+ goto error;
+ }
+
+ bo->pin_count = 1;
+ if (gpu_addr != NULL)
+ *gpu_addr = amdgpu_bo_gpu_offset(bo);
+ if (domain == AMDGPU_GEM_DOMAIN_VRAM) {
+ bo->adev->vram_pin_size += amdgpu_bo_size(bo);
+ if (bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS)
+ bo->adev->invisible_pin_size += amdgpu_bo_size(bo);
+ } else if (domain == AMDGPU_GEM_DOMAIN_GTT) {
+ bo->adev->gart_pin_size += amdgpu_bo_size(bo);
+ }
+
+error:
return r;
}
bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
}
r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
- if (likely(r == 0)) {
- if (bo->tbo.mem.mem_type == TTM_PL_VRAM) {
- bo->adev->vram_pin_size -= amdgpu_bo_size(bo);
- if (bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS)
- bo->adev->invisible_pin_size -= amdgpu_bo_size(bo);
- } else
- bo->adev->gart_pin_size -= amdgpu_bo_size(bo);
- } else {
+ if (unlikely(r)) {
dev_err(bo->adev->dev, "%p validate failed for unpin\n", bo);
+ goto error;
+ }
+
+ if (bo->tbo.mem.mem_type == TTM_PL_VRAM) {
+ bo->adev->vram_pin_size -= amdgpu_bo_size(bo);
+ if (bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS)
+ bo->adev->invisible_pin_size -= amdgpu_bo_size(bo);
+ } else if (bo->tbo.mem.mem_type == TTM_PL_TT) {
+ bo->adev->gart_pin_size -= amdgpu_bo_size(bo);
}
+
+error:
return r;
}
for (i = 0; i < abo->placement.num_placement; i++) {
/* Force into visible VRAM */
if ((abo->placements[i].flags & TTM_PL_FLAG_VRAM) &&
- (!abo->placements[i].lpfn || abo->placements[i].lpfn > lpfn))
+ (!abo->placements[i].lpfn ||
+ abo->placements[i].lpfn > lpfn))
abo->placements[i].lpfn = lpfn;
}
r = ttm_bo_validate(bo, &abo->placement, false, false);
else
reservation_object_add_excl_fence(resv, fence);
}
+
+/**
+ * amdgpu_bo_gpu_offset - return GPU offset of bo
+ * @bo: amdgpu object for which we query the offset
+ *
+ * Returns current GPU offset of the object.
+ *
+ * Note: object should either be pinned or reserved when calling this
+ * function, it might be useful to add check for this for debugging.
+ */
+u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo)
+{
+ WARN_ON_ONCE(bo->tbo.mem.mem_type == TTM_PL_SYSTEM);
+ WARN_ON_ONCE(bo->tbo.mem.mem_type == TTM_PL_TT &&
+ !amdgpu_ttm_is_bound(bo->tbo.ttm));
+ WARN_ON_ONCE(!ww_mutex_is_locked(&bo->tbo.resv->lock) &&
+ !bo->pin_count);
+ WARN_ON_ONCE(bo->tbo.mem.start == AMDGPU_BO_INVALID_OFFSET);
+
+ return bo->tbo.offset;
+}
#include <drm/amdgpu_drm.h>
#include "amdgpu.h"
+#define AMDGPU_BO_INVALID_OFFSET LONG_MAX
+
/**
* amdgpu_mem_type_to_domain - return domain corresponding to mem_type
* @mem_type: ttm memory type
ttm_bo_unreserve(&bo->tbo);
}
-/**
- * amdgpu_bo_gpu_offset - return GPU offset of bo
- * @bo: amdgpu object for which we query the offset
- *
- * Returns current GPU offset of the object.
- *
- * Note: object should either be pinned or reserved when calling this
- * function, it might be useful to add check for this for debugging.
- */
-static inline u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo)
-{
- WARN_ON_ONCE(bo->tbo.mem.mem_type == TTM_PL_SYSTEM);
- return bo->tbo.offset;
-}
-
static inline unsigned long amdgpu_bo_size(struct amdgpu_bo *bo)
{
return bo->tbo.num_pages << PAGE_SHIFT;
struct ttm_placement *placement,
struct reservation_object *resv,
struct amdgpu_bo **bo_ptr);
+int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
+ unsigned long size, int align,
+ u32 domain, struct amdgpu_bo **bo_ptr,
+ u64 *gpu_addr, void **cpu_addr);
+void amdgpu_bo_free_kernel(struct amdgpu_bo **bo, u64 *gpu_addr,
+ void **cpu_addr);
int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr);
void amdgpu_bo_kunmap(struct amdgpu_bo *bo);
struct amdgpu_bo *amdgpu_bo_ref(struct amdgpu_bo *bo);
int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo);
void amdgpu_bo_fence(struct amdgpu_bo *bo, struct fence *fence,
bool shared);
+u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo);
+int amdgpu_bo_backup_to_shadow(struct amdgpu_device *adev,
+ struct amdgpu_ring *ring,
+ struct amdgpu_bo *bo,
+ struct reservation_object *resv,
+ struct fence **fence, bool direct);
+int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev,
+ struct amdgpu_ring *ring,
+ struct amdgpu_bo *bo,
+ struct reservation_object *resv,
+ struct fence **fence,
+ bool direct);
+
/*
* sub allocation
#include "amdgpu.h"
#include "atom.h"
#include "atombios_encoders.h"
+#include "amdgpu_pll.h"
#include <asm/div64.h>
#include <linux/gcd.h>
void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable)
{
- if (adev->pp_enabled)
+ if (adev->pp_enabled || adev->pm.funcs->powergate_uvd) {
+ /* enable/disable UVD */
+ mutex_lock(&adev->pm.mutex);
amdgpu_dpm_powergate_uvd(adev, !enable);
- else {
- if (adev->pm.funcs->powergate_uvd) {
+ mutex_unlock(&adev->pm.mutex);
+ } else {
+ if (enable) {
mutex_lock(&adev->pm.mutex);
- /* enable/disable UVD */
- amdgpu_dpm_powergate_uvd(adev, !enable);
+ adev->pm.dpm.uvd_active = true;
+ adev->pm.dpm.state = POWER_STATE_TYPE_INTERNAL_UVD;
mutex_unlock(&adev->pm.mutex);
} else {
- if (enable) {
- mutex_lock(&adev->pm.mutex);
- adev->pm.dpm.uvd_active = true;
- adev->pm.dpm.state = POWER_STATE_TYPE_INTERNAL_UVD;
- mutex_unlock(&adev->pm.mutex);
- } else {
- mutex_lock(&adev->pm.mutex);
- adev->pm.dpm.uvd_active = false;
- mutex_unlock(&adev->pm.mutex);
- }
- amdgpu_pm_compute_clocks(adev);
+ mutex_lock(&adev->pm.mutex);
+ adev->pm.dpm.uvd_active = false;
+ mutex_unlock(&adev->pm.mutex);
}
-
+ amdgpu_pm_compute_clocks(adev);
}
}
void amdgpu_dpm_enable_vce(struct amdgpu_device *adev, bool enable)
{
- if (adev->pp_enabled)
+ if (adev->pp_enabled || adev->pm.funcs->powergate_vce) {
+ /* enable/disable VCE */
+ mutex_lock(&adev->pm.mutex);
amdgpu_dpm_powergate_vce(adev, !enable);
- else {
- if (adev->pm.funcs->powergate_vce) {
+ mutex_unlock(&adev->pm.mutex);
+ } else {
+ if (enable) {
mutex_lock(&adev->pm.mutex);
- amdgpu_dpm_powergate_vce(adev, !enable);
+ adev->pm.dpm.vce_active = true;
+ /* XXX select vce level based on ring/task */
+ adev->pm.dpm.vce_level = AMDGPU_VCE_LEVEL_AC_ALL;
mutex_unlock(&adev->pm.mutex);
} else {
- if (enable) {
- mutex_lock(&adev->pm.mutex);
- adev->pm.dpm.vce_active = true;
- /* XXX select vce level based on ring/task */
- adev->pm.dpm.vce_level = AMDGPU_VCE_LEVEL_AC_ALL;
- mutex_unlock(&adev->pm.mutex);
- } else {
- mutex_lock(&adev->pm.mutex);
- adev->pm.dpm.vce_active = false;
- mutex_unlock(&adev->pm.mutex);
- }
- amdgpu_pm_compute_clocks(adev);
+ mutex_lock(&adev->pm.mutex);
+ adev->pm.dpm.vce_active = false;
+ mutex_unlock(&adev->pm.mutex);
}
+ amdgpu_pm_compute_clocks(adev);
}
}
#include "amdgpu_pm.h"
#include <drm/amdgpu_drm.h>
#include "amdgpu_powerplay.h"
+#include "si_dpm.h"
#include "cik_dpm.h"
#include "vi_dpm.h"
pp_init->chip_family = adev->family;
pp_init->chip_id = adev->asic_type;
pp_init->device = amdgpu_cgs_create_device(adev);
- pp_init->powercontainment_enabled = amdgpu_powercontainment;
-
ret = amd_powerplay_init(pp_init, amd_pp);
kfree(pp_init);
#endif
amd_pp->pp_handle = (void *)adev;
switch (adev->asic_type) {
+#ifdef CONFIG_DRM_AMDGPU_SI
+ case CHIP_TAHITI:
+ case CHIP_PITCAIRN:
+ case CHIP_VERDE:
+ case CHIP_OLAND:
+ case CHIP_HAINAN:
+ amd_pp->ip_funcs = &si_dpm_ip_funcs;
+ break;
+#endif
#ifdef CONFIG_DRM_AMDGPU_CIK
case CHIP_BONAIRE:
case CHIP_HAWAII:
break;
case CHIP_TONGA:
case CHIP_FIJI:
- adev->pp_enabled = (amdgpu_powerplay == 0) ? false : true;
- break;
+ case CHIP_TOPAZ:
case CHIP_CARRIZO:
case CHIP_STONEY:
- adev->pp_enabled = (amdgpu_powerplay > 0) ? true : false;
+ adev->pp_enabled = (amdgpu_powerplay == 0) ? false : true;
break;
/* These chips don't have powerplay implemenations */
case CHIP_BONAIRE:
case CHIP_KABINI:
case CHIP_MULLINS:
case CHIP_KAVERI:
- case CHIP_TOPAZ:
default:
adev->pp_enabled = false;
break;
/* Allocate ring buffer */
if (ring->ring_obj == NULL) {
- r = amdgpu_bo_create(adev, ring->ring_size, PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_GTT, 0,
- NULL, NULL, &ring->ring_obj);
+ r = amdgpu_bo_create_kernel(adev, ring->ring_size, PAGE_SIZE,
+ AMDGPU_GEM_DOMAIN_GTT,
+ &ring->ring_obj,
+ &ring->gpu_addr,
+ (void **)&ring->ring);
if (r) {
dev_err(adev->dev, "(%d) ring create failed\n", r);
return r;
}
- r = amdgpu_bo_reserve(ring->ring_obj, false);
- if (unlikely(r != 0))
- return r;
- r = amdgpu_bo_pin(ring->ring_obj, AMDGPU_GEM_DOMAIN_GTT,
- &ring->gpu_addr);
- if (r) {
- amdgpu_bo_unreserve(ring->ring_obj);
- dev_err(adev->dev, "(%d) ring pin failed\n", r);
- return r;
- }
- r = amdgpu_bo_kmap(ring->ring_obj,
- (void **)&ring->ring);
-
memset((void *)ring->ring, 0, ring->ring_size);
-
- amdgpu_bo_unreserve(ring->ring_obj);
- if (r) {
- dev_err(adev->dev, "(%d) ring map failed\n", r);
- return r;
- }
}
ring->ptr_mask = (ring->ring_size / 4) - 1;
ring->max_dw = max_dw;
*/
void amdgpu_ring_fini(struct amdgpu_ring *ring)
{
- int r;
- struct amdgpu_bo *ring_obj;
-
- ring_obj = ring->ring_obj;
ring->ready = false;
- ring->ring = NULL;
- ring->ring_obj = NULL;
amdgpu_wb_free(ring->adev, ring->cond_exe_offs);
amdgpu_wb_free(ring->adev, ring->fence_offs);
amdgpu_wb_free(ring->adev, ring->rptr_offs);
amdgpu_wb_free(ring->adev, ring->wptr_offs);
- if (ring_obj) {
- r = amdgpu_bo_reserve(ring_obj, false);
- if (likely(r == 0)) {
- amdgpu_bo_kunmap(ring_obj);
- amdgpu_bo_unpin(ring_obj);
- amdgpu_bo_unreserve(ring_obj);
- }
- amdgpu_bo_unref(&ring_obj);
- }
+ amdgpu_bo_free_kernel(&ring->ring_obj,
+ &ring->gpu_addr,
+ (void **)&ring->ring);
+
amdgpu_debugfs_ring_fini(ring);
}
amdgpu_bo_kunmap(gtt_obj[i]);
r = amdgpu_copy_buffer(ring, gtt_addr, vram_addr,
- size, NULL, &fence);
+ size, NULL, &fence, false);
if (r) {
DRM_ERROR("Failed GTT->VRAM copy %d\n", i);
amdgpu_bo_kunmap(vram_obj);
r = amdgpu_copy_buffer(ring, vram_addr, gtt_addr,
- size, NULL, &fence);
+ size, NULL, &fence, false);
if (r) {
DRM_ERROR("Failed VRAM->GTT copy %d\n", i);
#include <ttm/ttm_placement.h>
#include <ttm/ttm_module.h>
#include <ttm/ttm_page_alloc.h>
+#include <ttm/ttm_memory.h>
#include <drm/drmP.h>
#include <drm/amdgpu_drm.h>
#include <linux/seq_file.h>
ttm_mem_global_release(ref->object);
}
-static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
+int amdgpu_ttm_global_init(struct amdgpu_device *adev)
{
struct drm_global_reference *global_ref;
struct amdgpu_ring *ring;
global_ref->init = &amdgpu_ttm_mem_global_init;
global_ref->release = &amdgpu_ttm_mem_global_release;
r = drm_global_item_ref(global_ref);
- if (r != 0) {
+ if (r) {
DRM_ERROR("Failed setting up TTM memory accounting "
"subsystem.\n");
- return r;
+ goto error_mem;
}
adev->mman.bo_global_ref.mem_glob =
global_ref->init = &ttm_bo_global_init;
global_ref->release = &ttm_bo_global_release;
r = drm_global_item_ref(global_ref);
- if (r != 0) {
+ if (r) {
DRM_ERROR("Failed setting up TTM BO subsystem.\n");
- drm_global_item_unref(&adev->mman.mem_global_ref);
- return r;
+ goto error_bo;
}
ring = adev->mman.buffer_funcs_ring;
rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_KERNEL];
r = amd_sched_entity_init(&ring->sched, &adev->mman.entity,
rq, amdgpu_sched_jobs);
- if (r != 0) {
+ if (r) {
DRM_ERROR("Failed setting up TTM BO move run queue.\n");
- drm_global_item_unref(&adev->mman.mem_global_ref);
- drm_global_item_unref(&adev->mman.bo_global_ref.ref);
- return r;
+ goto error_entity;
}
adev->mman.mem_global_referenced = true;
return 0;
+
+error_entity:
+ drm_global_item_unref(&adev->mman.bo_global_ref.ref);
+error_bo:
+ drm_global_item_unref(&adev->mman.mem_global_ref);
+error_mem:
+ return r;
}
static void amdgpu_ttm_global_fini(struct amdgpu_device *adev)
.lpfn = 0,
.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
};
+ unsigned i;
if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) {
placement->placement = &placements;
rbo = container_of(bo, struct amdgpu_bo, tbo);
switch (bo->mem.mem_type) {
case TTM_PL_VRAM:
- if (rbo->adev->mman.buffer_funcs_ring->ready == false)
+ if (rbo->adev->mman.buffer_funcs_ring->ready == false) {
amdgpu_ttm_placement_from_domain(rbo, AMDGPU_GEM_DOMAIN_CPU);
- else
+ } else {
amdgpu_ttm_placement_from_domain(rbo, AMDGPU_GEM_DOMAIN_GTT);
+ for (i = 0; i < rbo->placement.num_placement; ++i) {
+ if (!(rbo->placements[i].flags &
+ TTM_PL_FLAG_TT))
+ continue;
+
+ if (rbo->placements[i].lpfn)
+ continue;
+
+ /* set an upper limit to force directly
+ * allocating address space for the BO.
+ */
+ rbo->placements[i].lpfn =
+ rbo->adev->mc.gtt_size >> PAGE_SHIFT;
+ }
+ }
break;
case TTM_PL_TT:
default:
new_start = new_mem->start << PAGE_SHIFT;
switch (old_mem->mem_type) {
- case TTM_PL_VRAM:
- old_start += adev->mc.vram_start;
- break;
case TTM_PL_TT:
- old_start += adev->mc.gtt_start;
+ r = amdgpu_ttm_bind(bo->ttm, old_mem);
+ if (r)
+ return r;
+
+ case TTM_PL_VRAM:
+ old_start += bo->bdev->man[old_mem->mem_type].gpu_offset;
break;
default:
DRM_ERROR("Unknown placement %d\n", old_mem->mem_type);
return -EINVAL;
}
switch (new_mem->mem_type) {
- case TTM_PL_VRAM:
- new_start += adev->mc.vram_start;
- break;
case TTM_PL_TT:
- new_start += adev->mc.gtt_start;
+ r = amdgpu_ttm_bind(bo->ttm, new_mem);
+ if (r)
+ return r;
+
+ case TTM_PL_VRAM:
+ new_start += bo->bdev->man[new_mem->mem_type].gpu_offset;
break;
default:
DRM_ERROR("Unknown placement %d\n", old_mem->mem_type);
r = amdgpu_copy_buffer(ring, old_start, new_start,
new_mem->num_pages * PAGE_SIZE, /* bytes */
- bo->resv, &fence);
+ bo->resv, &fence, false);
if (r)
return r;
placement.num_busy_placement = 1;
placement.busy_placement = &placements;
placements.fpfn = 0;
- placements.lpfn = 0;
+ placements.lpfn = adev->mc.gtt_size >> PAGE_SHIFT;
placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
interruptible, no_wait_gpu);
if (unlikely(r)) {
goto out_cleanup;
}
- r = ttm_bo_move_ttm(bo, true, interruptible, no_wait_gpu, new_mem);
+ r = ttm_bo_move_ttm(bo, interruptible, no_wait_gpu, new_mem);
out_cleanup:
ttm_bo_mem_put(bo, &tmp_mem);
return r;
placement.num_busy_placement = 1;
placement.busy_placement = &placements;
placements.fpfn = 0;
- placements.lpfn = 0;
+ placements.lpfn = adev->mc.gtt_size >> PAGE_SHIFT;
placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT;
r = ttm_bo_mem_space(bo, &placement, &tmp_mem,
interruptible, no_wait_gpu);
if (unlikely(r)) {
return r;
}
- r = ttm_bo_move_ttm(bo, true, interruptible, no_wait_gpu, &tmp_mem);
+ r = ttm_bo_move_ttm(bo, interruptible, no_wait_gpu, &tmp_mem);
if (unlikely(r)) {
goto out_cleanup;
}
if (r) {
memcpy:
- r = ttm_bo_move_memcpy(bo, evict, interruptible,
- no_wait_gpu, new_mem);
+ r = ttm_bo_move_memcpy(bo, interruptible, no_wait_gpu, new_mem);
if (r) {
return r;
}
spinlock_t guptasklock;
struct list_head guptasks;
atomic_t mmu_invalidations;
+ struct list_head list;
};
int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
struct ttm_mem_reg *bo_mem)
{
struct amdgpu_ttm_tt *gtt = (void*)ttm;
- uint32_t flags = amdgpu_ttm_tt_pte_flags(gtt->adev, ttm, bo_mem);
int r;
if (gtt->userptr) {
return r;
}
}
- gtt->offset = (unsigned long)(bo_mem->start << PAGE_SHIFT);
+ gtt->offset = (u64)bo_mem->start << PAGE_SHIFT;
if (!ttm->num_pages) {
WARN(1, "nothing to bind %lu pages for mreg %p back %p!\n",
ttm->num_pages, bo_mem, ttm);
bo_mem->mem_type == AMDGPU_PL_OA)
return -EINVAL;
+ return 0;
+}
+
+bool amdgpu_ttm_is_bound(struct ttm_tt *ttm)
+{
+ struct amdgpu_ttm_tt *gtt = (void *)ttm;
+
+ return gtt && !list_empty(>t->list);
+}
+
+int amdgpu_ttm_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem)
+{
+ struct amdgpu_ttm_tt *gtt = (void *)ttm;
+ uint32_t flags;
+ int r;
+
+ if (!ttm || amdgpu_ttm_is_bound(ttm))
+ return 0;
+
+ flags = amdgpu_ttm_tt_pte_flags(gtt->adev, ttm, bo_mem);
r = amdgpu_gart_bind(gtt->adev, gtt->offset, ttm->num_pages,
ttm->pages, gtt->ttm.dma_address, flags);
if (r) {
- DRM_ERROR("failed to bind %lu pages at 0x%08X\n",
- ttm->num_pages, (unsigned)gtt->offset);
+ DRM_ERROR("failed to bind %lu pages at 0x%08llX\n",
+ ttm->num_pages, gtt->offset);
return r;
}
+ spin_lock(>t->adev->gtt_list_lock);
+ list_add_tail(>t->list, >t->adev->gtt_list);
+ spin_unlock(>t->adev->gtt_list_lock);
+ return 0;
+}
+
+int amdgpu_ttm_recover_gart(struct amdgpu_device *adev)
+{
+ struct amdgpu_ttm_tt *gtt, *tmp;
+ struct ttm_mem_reg bo_mem;
+ uint32_t flags;
+ int r;
+
+ bo_mem.mem_type = TTM_PL_TT;
+ spin_lock(&adev->gtt_list_lock);
+ list_for_each_entry_safe(gtt, tmp, &adev->gtt_list, list) {
+ flags = amdgpu_ttm_tt_pte_flags(gtt->adev, >t->ttm.ttm, &bo_mem);
+ r = amdgpu_gart_bind(adev, gtt->offset, gtt->ttm.ttm.num_pages,
+ gtt->ttm.ttm.pages, gtt->ttm.dma_address,
+ flags);
+ if (r) {
+ spin_unlock(&adev->gtt_list_lock);
+ DRM_ERROR("failed to bind %lu pages at 0x%08llX\n",
+ gtt->ttm.ttm.num_pages, gtt->offset);
+ return r;
+ }
+ }
+ spin_unlock(&adev->gtt_list_lock);
return 0;
}
{
struct amdgpu_ttm_tt *gtt = (void *)ttm;
+ if (!amdgpu_ttm_is_bound(ttm))
+ return 0;
+
/* unbind shouldn't be done for GDS/GWS/OA in ttm_bo_clean_mm */
if (gtt->adev->gart.ready)
amdgpu_gart_unbind(gtt->adev, gtt->offset, ttm->num_pages);
if (gtt->userptr)
amdgpu_ttm_tt_unpin_userptr(ttm);
+ spin_lock(>t->adev->gtt_list_lock);
+ list_del_init(>t->list);
+ spin_unlock(>t->adev->gtt_list_lock);
+
return 0;
}
kfree(gtt);
return NULL;
}
+ INIT_LIST_HEAD(>t->list);
return >t->ttm.ttm;
}
struct list_head *res = lru->lru[tbo->mem.mem_type];
lru->lru[tbo->mem.mem_type] = &tbo->lru;
+ while ((++lru)->lru[tbo->mem.mem_type] == res)
+ lru->lru[tbo->mem.mem_type] = &tbo->lru;
return res;
}
struct list_head *res = lru->swap_lru;
lru->swap_lru = &tbo->swap;
+ while ((++lru)->swap_lru == res)
+ lru->swap_lru = &tbo->swap;
return res;
}
unsigned i, j;
int r;
- r = amdgpu_ttm_global_init(adev);
- if (r) {
- return r;
- }
/* No others user of address space so set it to 0 */
r = ttm_bo_device_init(&adev->mman.bdev,
adev->mman.bo_global_ref.ref.object,
lru->swap_lru = &adev->mman.bdev.glob->swap_lru;
}
+ for (j = 0; j < TTM_NUM_MEM_TYPES; ++j)
+ adev->mman.guard.lru[j] = NULL;
+ adev->mman.guard.swap_lru = NULL;
+
adev->mman.initialized = true;
r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_VRAM,
adev->mc.real_vram_size >> PAGE_SHIFT);
uint64_t dst_offset,
uint32_t byte_count,
struct reservation_object *resv,
- struct fence **fence)
+ struct fence **fence, bool direct_submit)
{
struct amdgpu_device *adev = ring->adev;
struct amdgpu_job *job;
byte_count -= cur_size_in_bytes;
}
+ amdgpu_ring_pad_ib(ring, &job->ibs[0]);
+ WARN_ON(job->ibs[0].length_dw > num_dw);
+ if (direct_submit) {
+ r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs,
+ NULL, NULL, fence);
+ job->fence = fence_get(*fence);
+ if (r)
+ DRM_ERROR("Error scheduling IBs (%d)\n", r);
+ amdgpu_job_free(job);
+ } else {
+ r = amdgpu_job_submit(job, ring, &adev->mman.entity,
+ AMDGPU_FENCE_OWNER_UNDEFINED, fence);
+ if (r)
+ goto error_free;
+ }
+
+ return r;
+
+error_free:
+ amdgpu_job_free(job);
+ return r;
+}
+
+int amdgpu_fill_buffer(struct amdgpu_bo *bo,
+ uint32_t src_data,
+ struct reservation_object *resv,
+ struct fence **fence)
+{
+ struct amdgpu_device *adev = bo->adev;
+ struct amdgpu_job *job;
+ struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring;
+
+ uint32_t max_bytes, byte_count;
+ uint64_t dst_offset;
+ unsigned int num_loops, num_dw;
+ unsigned int i;
+ int r;
+
+ byte_count = bo->tbo.num_pages << PAGE_SHIFT;
+ max_bytes = adev->mman.buffer_funcs->fill_max_bytes;
+ num_loops = DIV_ROUND_UP(byte_count, max_bytes);
+ num_dw = num_loops * adev->mman.buffer_funcs->fill_num_dw;
+
+ /* for IB padding */
+ while (num_dw & 0x7)
+ num_dw++;
+
+ r = amdgpu_job_alloc_with_ib(adev, num_dw * 4, &job);
+ if (r)
+ return r;
+
+ if (resv) {
+ r = amdgpu_sync_resv(adev, &job->sync, resv,
+ AMDGPU_FENCE_OWNER_UNDEFINED);
+ if (r) {
+ DRM_ERROR("sync failed (%d).\n", r);
+ goto error_free;
+ }
+ }
+
+ dst_offset = bo->tbo.mem.start << PAGE_SHIFT;
+ for (i = 0; i < num_loops; i++) {
+ uint32_t cur_size_in_bytes = min(byte_count, max_bytes);
+
+ amdgpu_emit_fill_buffer(adev, &job->ibs[0], src_data,
+ dst_offset, cur_size_in_bytes);
+
+ dst_offset += cur_size_in_bytes;
+ byte_count -= cur_size_in_bytes;
+ }
+
amdgpu_ring_pad_ib(ring, &job->ibs[0]);
WARN_ON(job->ibs[0].length_dw > num_dw);
r = amdgpu_job_submit(job, ring, &adev->mman.entity,
- AMDGPU_FENCE_OWNER_UNDEFINED, fence);
+ AMDGPU_FENCE_OWNER_UNDEFINED, fence);
if (r)
goto error_free;
#endif
}
+
+u64 amdgpu_ttm_get_gtt_mem_size(struct amdgpu_device *adev)
+{
+ return ttm_get_kernel_zone_memory_size(adev->mman.mem_global_ref.object);
+}
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __AMDGPU_TTM_H__
+#define __AMDGPU_TTM_H__
+
+#include "gpu_scheduler.h"
+
+#define AMDGPU_PL_GDS (TTM_PL_PRIV + 0)
+#define AMDGPU_PL_GWS (TTM_PL_PRIV + 1)
+#define AMDGPU_PL_OA (TTM_PL_PRIV + 2)
+
+#define AMDGPU_PL_FLAG_GDS (TTM_PL_FLAG_PRIV << 0)
+#define AMDGPU_PL_FLAG_GWS (TTM_PL_FLAG_PRIV << 1)
+#define AMDGPU_PL_FLAG_OA (TTM_PL_FLAG_PRIV << 2)
+
+#define AMDGPU_TTM_LRU_SIZE 20
+
+struct amdgpu_mman_lru {
+ struct list_head *lru[TTM_NUM_MEM_TYPES];
+ struct list_head *swap_lru;
+};
+
+struct amdgpu_mman {
+ struct ttm_bo_global_ref bo_global_ref;
+ struct drm_global_reference mem_global_ref;
+ struct ttm_bo_device bdev;
+ bool mem_global_referenced;
+ bool initialized;
+
+#if defined(CONFIG_DEBUG_FS)
+ struct dentry *vram;
+ struct dentry *gtt;
+#endif
+
+ /* buffer handling */
+ const struct amdgpu_buffer_funcs *buffer_funcs;
+ struct amdgpu_ring *buffer_funcs_ring;
+ /* Scheduler entity for buffer moves */
+ struct amd_sched_entity entity;
+
+ /* custom LRU management */
+ struct amdgpu_mman_lru log2_size[AMDGPU_TTM_LRU_SIZE];
+ /* guard for log2_size array, don't add anything in between */
+ struct amdgpu_mman_lru guard;
+};
+
+int amdgpu_copy_buffer(struct amdgpu_ring *ring,
+ uint64_t src_offset,
+ uint64_t dst_offset,
+ uint32_t byte_count,
+ struct reservation_object *resv,
+ struct fence **fence, bool direct_submit);
+int amdgpu_fill_buffer(struct amdgpu_bo *bo,
+ uint32_t src_data,
+ struct reservation_object *resv,
+ struct fence **fence);
+
+int amdgpu_mmap(struct file *filp, struct vm_area_struct *vma);
+bool amdgpu_ttm_is_bound(struct ttm_tt *ttm);
+int amdgpu_ttm_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem);
+
+#endif
const struct common_firmware_header *header = NULL;
err = amdgpu_bo_create(adev, adev->firmware.fw_size, PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_GTT, 0, NULL, NULL, bo);
+ AMDGPU_GEM_DOMAIN_GTT, 0, NULL, NULL, bo);
if (err) {
dev_err(adev->dev, "(%d) Firmware buffer allocate failed\n", err);
- err = -ENOMEM;
goto failed;
}
err = amdgpu_bo_reserve(*bo, false);
if (err) {
- amdgpu_bo_unref(bo);
dev_err(adev->dev, "(%d) Firmware buffer reserve failed\n", err);
- goto failed;
+ goto failed_reserve;
}
err = amdgpu_bo_pin(*bo, AMDGPU_GEM_DOMAIN_GTT, &fw_mc_addr);
if (err) {
- amdgpu_bo_unreserve(*bo);
- amdgpu_bo_unref(bo);
dev_err(adev->dev, "(%d) Firmware buffer pin failed\n", err);
- goto failed;
+ goto failed_pin;
}
err = amdgpu_bo_kmap(*bo, &fw_buf_ptr);
if (err) {
dev_err(adev->dev, "(%d) Firmware buffer kmap failed\n", err);
- amdgpu_bo_unpin(*bo);
- amdgpu_bo_unreserve(*bo);
- amdgpu_bo_unref(bo);
- goto failed;
+ goto failed_kmap;
}
amdgpu_bo_unreserve(*bo);
fw_offset += ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE);
}
}
+ return 0;
+failed_kmap:
+ amdgpu_bo_unpin(*bo);
+failed_pin:
+ amdgpu_bo_unreserve(*bo);
+failed_reserve:
+ amdgpu_bo_unref(bo);
failed:
- if (err)
- adev->firmware.smu_load = false;
+ adev->firmware.smu_load = false;
return err;
}
bo_size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8)
+ AMDGPU_UVD_STACK_SIZE + AMDGPU_UVD_HEAP_SIZE
+ AMDGPU_UVD_SESSION_SIZE * adev->uvd.max_handles;
- r = amdgpu_bo_create(adev, bo_size, PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_VRAM,
- AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
- NULL, NULL, &adev->uvd.vcpu_bo);
+ r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE,
+ AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.vcpu_bo,
+ &adev->uvd.gpu_addr, &adev->uvd.cpu_addr);
if (r) {
dev_err(adev->dev, "(%d) failed to allocate UVD bo\n", r);
return r;
}
- r = amdgpu_bo_reserve(adev->uvd.vcpu_bo, false);
- if (r) {
- amdgpu_bo_unref(&adev->uvd.vcpu_bo);
- dev_err(adev->dev, "(%d) failed to reserve UVD bo\n", r);
- return r;
- }
-
- r = amdgpu_bo_pin(adev->uvd.vcpu_bo, AMDGPU_GEM_DOMAIN_VRAM,
- &adev->uvd.gpu_addr);
- if (r) {
- amdgpu_bo_unreserve(adev->uvd.vcpu_bo);
- amdgpu_bo_unref(&adev->uvd.vcpu_bo);
- dev_err(adev->dev, "(%d) UVD bo pin failed\n", r);
- return r;
- }
-
- r = amdgpu_bo_kmap(adev->uvd.vcpu_bo, &adev->uvd.cpu_addr);
- if (r) {
- dev_err(adev->dev, "(%d) UVD map failed\n", r);
- return r;
- }
-
- amdgpu_bo_unreserve(adev->uvd.vcpu_bo);
-
ring = &adev->uvd.ring;
rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_NORMAL];
r = amd_sched_entity_init(&ring->sched, &adev->uvd.entity,
int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
{
- int r;
-
kfree(adev->uvd.saved_bo);
amd_sched_entity_fini(&adev->uvd.ring.sched, &adev->uvd.entity);
- if (adev->uvd.vcpu_bo) {
- r = amdgpu_bo_reserve(adev->uvd.vcpu_bo, false);
- if (!r) {
- amdgpu_bo_kunmap(adev->uvd.vcpu_bo);
- amdgpu_bo_unpin(adev->uvd.vcpu_bo);
- amdgpu_bo_unreserve(adev->uvd.vcpu_bo);
- }
-
- amdgpu_bo_unref(&adev->uvd.vcpu_bo);
- }
+ amdgpu_bo_free_kernel(&adev->uvd.vcpu_bo,
+ &adev->uvd.gpu_addr,
+ (void **)&adev->uvd.cpu_addr);
amdgpu_ring_fini(&adev->uvd.ring);
if (!adev->uvd.saved_bo)
return -ENOMEM;
- memcpy(adev->uvd.saved_bo, ptr, size);
+ memcpy_fromio(adev->uvd.saved_bo, ptr, size);
return 0;
}
ptr = adev->uvd.cpu_addr;
if (adev->uvd.saved_bo != NULL) {
- memcpy(ptr, adev->uvd.saved_bo, size);
+ memcpy_toio(ptr, adev->uvd.saved_bo, size);
kfree(adev->uvd.saved_bo);
adev->uvd.saved_bo = NULL;
} else {
hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
- memcpy(adev->uvd.cpu_addr, (adev->uvd.fw->data) + offset,
- (adev->uvd.fw->size) - offset);
+ memcpy_toio(adev->uvd.cpu_addr, adev->uvd.fw->data + offset,
+ le32_to_cpu(hdr->ucode_size_bytes));
size -= le32_to_cpu(hdr->ucode_size_bytes);
ptr += le32_to_cpu(hdr->ucode_size_bytes);
- memset(ptr, 0, size);
+ memset_io(ptr, 0, size);
}
return 0;
return r;
break;
case mmUVD_ENGINE_CNTL:
+ case mmUVD_NO_OP:
break;
default:
DRM_ERROR("Invalid reg 0x%X!\n", reg);
return -EINVAL;
}
+ r = amdgpu_cs_sysvm_access_required(parser);
+ if (r)
+ return r;
+
ctx.parser = parser;
ctx.buf_sizes = buf_sizes;
ctx.ib_idx = ib_idx;
ib->ptr[3] = addr >> 32;
ib->ptr[4] = PACKET0(mmUVD_GPCOM_VCPU_CMD, 0);
ib->ptr[5] = 0;
- for (i = 6; i < 16; ++i)
- ib->ptr[i] = PACKET2(0);
+ for (i = 6; i < 16; i += 2) {
+ ib->ptr[i] = PACKET0(mmUVD_NO_OP, 0);
+ ib->ptr[i+1] = 0;
+ }
ib->length_dw = 16;
if (direct) {
{
struct amdgpu_device *adev =
container_of(work, struct amdgpu_device, uvd.idle_work.work);
- unsigned i, fences, handles = 0;
-
- fences = amdgpu_fence_count_emitted(&adev->uvd.ring);
-
- for (i = 0; i < adev->uvd.max_handles; ++i)
- if (atomic_read(&adev->uvd.handles[i]))
- ++handles;
+ unsigned fences = amdgpu_fence_count_emitted(&adev->uvd.ring);
- if (fences == 0 && handles == 0) {
+ if (fences == 0) {
if (adev->pm.dpm_enabled) {
amdgpu_dpm_enable_uvd(adev, false);
} else {
hdr = (const struct common_firmware_header *)adev->vce.fw->data;
offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
- memcpy(cpu_addr, (adev->vce.fw->data) + offset,
- (adev->vce.fw->size) - offset);
+ memcpy_toio(cpu_addr, adev->vce.fw->data + offset,
+ adev->vce.fw->size - offset);
amdgpu_bo_kunmap(adev->vce.vcpu_bo);
uint32_t allocated = 0;
uint32_t tmp, handle = 0;
uint32_t *size = &tmp;
- int i, r = 0, idx = 0;
+ int i, r, idx = 0;
+
+ r = amdgpu_cs_sysvm_access_required(p);
+ if (r)
+ return r;
while (idx < ib->length_dw) {
uint32_t len = amdgpu_get_ib_value(p, ib_idx, idx);
amdgpu_ring_write(ring, VCE_CMD_END);
}
+unsigned amdgpu_vce_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 4; /* amdgpu_vce_ring_emit_ib */
+}
+
+unsigned amdgpu_vce_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 6; /* amdgpu_vce_ring_emit_fence x1 no user fence */
+}
+
/**
* amdgpu_vce_ring_test_ring - test if VCE ring is working
*
struct fence *fence = NULL;
long r;
- /* skip vce ring1 ib test for now, since it's not reliable */
- if (ring == &ring->adev->vce.ring[1])
+ /* skip vce ring1/2 ib test for now, since it's not reliable */
+ if (ring != &ring->adev->vce.ring[0])
return 0;
r = amdgpu_vce_get_create_msg(ring, 1, NULL);
int amdgpu_vce_ring_test_ib(struct amdgpu_ring *ring, long timeout);
void amdgpu_vce_ring_begin_use(struct amdgpu_ring *ring);
void amdgpu_vce_ring_end_use(struct amdgpu_ring *ring);
+unsigned amdgpu_vce_ring_get_emit_ib_size(struct amdgpu_ring *ring);
+unsigned amdgpu_vce_ring_get_dma_frame_size(struct amdgpu_ring *ring);
#endif
* SI supports 16.
*/
-/* Special value that no flush is necessary */
-#define AMDGPU_VM_NO_FLUSH (~0ll)
-
/* Local structure. Encapsulate some VM table update parameters to reduce
* the number of function parameters
*/
-struct amdgpu_vm_update_params {
+struct amdgpu_pte_update_params {
+ /* amdgpu device we do this update for */
+ struct amdgpu_device *adev;
/* address where to copy page table entries from */
uint64_t src;
- /* DMA addresses to use for mapping */
- dma_addr_t *pages_addr;
/* indirect buffer to fill with commands */
struct amdgpu_ib *ib;
+ /* Function which actually does the update */
+ void (*func)(struct amdgpu_pte_update_params *params, uint64_t pe,
+ uint64_t addr, unsigned count, uint32_t incr,
+ uint32_t flags);
+ /* indicate update pt or its shadow */
+ bool shadow;
};
/**
}
/**
- * amdgpu_vm_update_pages - helper to call the right asic function
+ * amdgpu_vm_do_set_ptes - helper to call the right asic function
*
- * @adev: amdgpu_device pointer
- * @vm_update_params: see amdgpu_vm_update_params definition
+ * @params: see amdgpu_pte_update_params definition
* @pe: addr of the page entry
* @addr: dst addr to write into pe
* @count: number of page entries to update
* Traces the parameters and calls the right asic functions
* to setup the page table using the DMA.
*/
-static void amdgpu_vm_update_pages(struct amdgpu_device *adev,
- struct amdgpu_vm_update_params
- *vm_update_params,
- uint64_t pe, uint64_t addr,
- unsigned count, uint32_t incr,
- uint32_t flags)
+static void amdgpu_vm_do_set_ptes(struct amdgpu_pte_update_params *params,
+ uint64_t pe, uint64_t addr,
+ unsigned count, uint32_t incr,
+ uint32_t flags)
{
trace_amdgpu_vm_set_page(pe, addr, count, incr, flags);
- if (vm_update_params->src) {
- amdgpu_vm_copy_pte(adev, vm_update_params->ib,
- pe, (vm_update_params->src + (addr >> 12) * 8), count);
-
- } else if (vm_update_params->pages_addr) {
- amdgpu_vm_write_pte(adev, vm_update_params->ib,
- vm_update_params->pages_addr,
- pe, addr, count, incr, flags);
-
- } else if (count < 3) {
- amdgpu_vm_write_pte(adev, vm_update_params->ib, NULL, pe, addr,
- count, incr, flags);
+ if (count < 3) {
+ amdgpu_vm_write_pte(params->adev, params->ib, pe,
+ addr | flags, count, incr);
} else {
- amdgpu_vm_set_pte_pde(adev, vm_update_params->ib, pe, addr,
+ amdgpu_vm_set_pte_pde(params->adev, params->ib, pe, addr,
count, incr, flags);
}
}
+/**
+ * amdgpu_vm_do_copy_ptes - copy the PTEs from the GART
+ *
+ * @params: see amdgpu_pte_update_params definition
+ * @pe: addr of the page entry
+ * @addr: dst addr to write into pe
+ * @count: number of page entries to update
+ * @incr: increase next addr by incr bytes
+ * @flags: hw access flags
+ *
+ * Traces the parameters and calls the DMA function to copy the PTEs.
+ */
+static void amdgpu_vm_do_copy_ptes(struct amdgpu_pte_update_params *params,
+ uint64_t pe, uint64_t addr,
+ unsigned count, uint32_t incr,
+ uint32_t flags)
+{
+ trace_amdgpu_vm_set_page(pe, addr, count, incr, flags);
+
+ amdgpu_vm_copy_pte(params->adev, params->ib, pe,
+ (params->src + (addr >> 12) * 8), count);
+}
+
/**
* amdgpu_vm_clear_bo - initially clear the page dir/table
*
struct amdgpu_ring *ring;
struct fence *fence = NULL;
struct amdgpu_job *job;
- struct amdgpu_vm_update_params vm_update_params;
+ struct amdgpu_pte_update_params params;
unsigned entries;
uint64_t addr;
int r;
- memset(&vm_update_params, 0, sizeof(vm_update_params));
ring = container_of(vm->entity.sched, struct amdgpu_ring, sched);
r = reservation_object_reserve_shared(bo->tbo.resv);
if (r)
goto error;
- vm_update_params.ib = &job->ibs[0];
- amdgpu_vm_update_pages(adev, &vm_update_params, addr, 0, entries,
- 0, 0);
+ memset(¶ms, 0, sizeof(params));
+ params.adev = adev;
+ params.ib = &job->ibs[0];
+ amdgpu_vm_do_set_ptes(¶ms, addr, 0, entries, 0, 0);
amdgpu_ring_pad_ib(ring, &job->ibs[0]);
WARN_ON(job->ibs[0].length_dw > 64);
* Look up the physical address of the page that the pte resolves
* to and return the pointer for the page table entry.
*/
-uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr)
+static uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr)
{
uint64_t result;
- if (pages_addr) {
- /* page table offset */
- result = pages_addr[addr >> PAGE_SHIFT];
-
- /* in case cpu page size != gpu page size*/
- result |= addr & (~PAGE_MASK);
+ /* page table offset */
+ result = pages_addr[addr >> PAGE_SHIFT];
- } else {
- /* No mapping required */
- result = addr;
- }
+ /* in case cpu page size != gpu page size*/
+ result |= addr & (~PAGE_MASK);
result &= 0xFFFFFFFFFFFFF000ULL;
return result;
}
-/**
- * amdgpu_vm_update_pdes - make sure that page directory is valid
- *
- * @adev: amdgpu_device pointer
- * @vm: requested vm
- * @start: start of GPU address range
- * @end: end of GPU address range
- *
- * Allocates new page tables if necessary
- * and updates the page directory.
- * Returns 0 for success, error for failure.
- */
-int amdgpu_vm_update_page_directory(struct amdgpu_device *adev,
- struct amdgpu_vm *vm)
+static int amdgpu_vm_update_pd_or_shadow(struct amdgpu_device *adev,
+ struct amdgpu_vm *vm,
+ bool shadow)
{
struct amdgpu_ring *ring;
- struct amdgpu_bo *pd = vm->page_directory;
- uint64_t pd_addr = amdgpu_bo_gpu_offset(pd);
+ struct amdgpu_bo *pd = shadow ? vm->page_directory->shadow :
+ vm->page_directory;
+ uint64_t pd_addr;
uint32_t incr = AMDGPU_VM_PTE_COUNT * 8;
uint64_t last_pde = ~0, last_pt = ~0;
unsigned count = 0, pt_idx, ndw;
struct amdgpu_job *job;
- struct amdgpu_vm_update_params vm_update_params;
+ struct amdgpu_pte_update_params params;
struct fence *fence = NULL;
int r;
- memset(&vm_update_params, 0, sizeof(vm_update_params));
+ if (!pd)
+ return 0;
+ pd_addr = amdgpu_bo_gpu_offset(pd);
ring = container_of(vm->entity.sched, struct amdgpu_ring, sched);
/* padding, etc. */
if (r)
return r;
- vm_update_params.ib = &job->ibs[0];
+ memset(¶ms, 0, sizeof(params));
+ params.adev = adev;
+ params.ib = &job->ibs[0];
/* walk over the address space and update the page directory */
for (pt_idx = 0; pt_idx <= vm->max_pde_used; ++pt_idx) {
continue;
pt = amdgpu_bo_gpu_offset(bo);
- if (vm->page_tables[pt_idx].addr == pt)
- continue;
- vm->page_tables[pt_idx].addr = pt;
+ if (!shadow) {
+ if (vm->page_tables[pt_idx].addr == pt)
+ continue;
+ vm->page_tables[pt_idx].addr = pt;
+ } else {
+ if (vm->page_tables[pt_idx].shadow_addr == pt)
+ continue;
+ vm->page_tables[pt_idx].shadow_addr = pt;
+ }
pde = pd_addr + pt_idx * 8;
if (((last_pde + 8 * count) != pde) ||
- ((last_pt + incr * count) != pt)) {
+ ((last_pt + incr * count) != pt) ||
+ (count == AMDGPU_VM_MAX_UPDATE_SIZE)) {
if (count) {
- amdgpu_vm_update_pages(adev, &vm_update_params,
- last_pde, last_pt,
- count, incr,
- AMDGPU_PTE_VALID);
+ amdgpu_vm_do_set_ptes(¶ms, last_pde,
+ last_pt, count, incr,
+ AMDGPU_PTE_VALID);
}
count = 1;
}
if (count)
- amdgpu_vm_update_pages(adev, &vm_update_params,
- last_pde, last_pt,
- count, incr, AMDGPU_PTE_VALID);
+ amdgpu_vm_do_set_ptes(¶ms, last_pde, last_pt,
+ count, incr, AMDGPU_PTE_VALID);
- if (vm_update_params.ib->length_dw != 0) {
- amdgpu_ring_pad_ib(ring, vm_update_params.ib);
+ if (params.ib->length_dw != 0) {
+ amdgpu_ring_pad_ib(ring, params.ib);
amdgpu_sync_resv(adev, &job->sync, pd->tbo.resv,
AMDGPU_FENCE_OWNER_VM);
- WARN_ON(vm_update_params.ib->length_dw > ndw);
+ WARN_ON(params.ib->length_dw > ndw);
r = amdgpu_job_submit(job, ring, &vm->entity,
AMDGPU_FENCE_OWNER_VM, &fence);
if (r)
return r;
}
-/**
- * amdgpu_vm_frag_ptes - add fragment information to PTEs
+/*
+ * amdgpu_vm_update_pdes - make sure that page directory is valid
*
* @adev: amdgpu_device pointer
- * @vm_update_params: see amdgpu_vm_update_params definition
- * @pe_start: first PTE to handle
- * @pe_end: last PTE to handle
- * @addr: addr those PTEs should point to
- * @flags: hw mapping flags
+ * @vm: requested vm
+ * @start: start of GPU address range
+ * @end: end of GPU address range
+ *
+ * Allocates new page tables if necessary
+ * and updates the page directory.
+ * Returns 0 for success, error for failure.
*/
-static void amdgpu_vm_frag_ptes(struct amdgpu_device *adev,
- struct amdgpu_vm_update_params
- *vm_update_params,
- uint64_t pe_start, uint64_t pe_end,
- uint64_t addr, uint32_t flags)
+int amdgpu_vm_update_page_directory(struct amdgpu_device *adev,
+ struct amdgpu_vm *vm)
{
- /**
- * The MC L1 TLB supports variable sized pages, based on a fragment
- * field in the PTE. When this field is set to a non-zero value, page
- * granularity is increased from 4KB to (1 << (12 + frag)). The PTE
- * flags are considered valid for all PTEs within the fragment range
- * and corresponding mappings are assumed to be physically contiguous.
- *
- * The L1 TLB can store a single PTE for the whole fragment,
- * significantly increasing the space available for translation
- * caching. This leads to large improvements in throughput when the
- * TLB is under pressure.
- *
- * The L2 TLB distributes small and large fragments into two
- * asymmetric partitions. The large fragment cache is significantly
- * larger. Thus, we try to use large fragments wherever possible.
- * Userspace can support this by aligning virtual base address and
- * allocation size to the fragment size.
- */
-
- /* SI and newer are optimized for 64KB */
- uint64_t frag_flags = AMDGPU_PTE_FRAG_64KB;
- uint64_t frag_align = 0x80;
-
- uint64_t frag_start = ALIGN(pe_start, frag_align);
- uint64_t frag_end = pe_end & ~(frag_align - 1);
-
- unsigned count;
-
- /* Abort early if there isn't anything to do */
- if (pe_start == pe_end)
- return;
-
- /* system pages are non continuously */
- if (vm_update_params->src || vm_update_params->pages_addr ||
- !(flags & AMDGPU_PTE_VALID) || (frag_start >= frag_end)) {
-
- count = (pe_end - pe_start) / 8;
- amdgpu_vm_update_pages(adev, vm_update_params, pe_start,
- addr, count, AMDGPU_GPU_PAGE_SIZE,
- flags);
- return;
- }
-
- /* handle the 4K area at the beginning */
- if (pe_start != frag_start) {
- count = (frag_start - pe_start) / 8;
- amdgpu_vm_update_pages(adev, vm_update_params, pe_start, addr,
- count, AMDGPU_GPU_PAGE_SIZE, flags);
- addr += AMDGPU_GPU_PAGE_SIZE * count;
- }
-
- /* handle the area in the middle */
- count = (frag_end - frag_start) / 8;
- amdgpu_vm_update_pages(adev, vm_update_params, frag_start, addr, count,
- AMDGPU_GPU_PAGE_SIZE, flags | frag_flags);
+ int r;
- /* handle the 4K area at the end */
- if (frag_end != pe_end) {
- addr += AMDGPU_GPU_PAGE_SIZE * count;
- count = (pe_end - frag_end) / 8;
- amdgpu_vm_update_pages(adev, vm_update_params, frag_end, addr,
- count, AMDGPU_GPU_PAGE_SIZE, flags);
- }
+ r = amdgpu_vm_update_pd_or_shadow(adev, vm, true);
+ if (r)
+ return r;
+ return amdgpu_vm_update_pd_or_shadow(adev, vm, false);
}
/**
* amdgpu_vm_update_ptes - make sure that page tables are valid
*
- * @adev: amdgpu_device pointer
- * @vm_update_params: see amdgpu_vm_update_params definition
+ * @params: see amdgpu_pte_update_params definition
* @vm: requested vm
* @start: start of GPU address range
* @end: end of GPU address range
*
* Update the page tables in the range @start - @end.
*/
-static void amdgpu_vm_update_ptes(struct amdgpu_device *adev,
- struct amdgpu_vm_update_params
- *vm_update_params,
+static void amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params,
struct amdgpu_vm *vm,
uint64_t start, uint64_t end,
uint64_t dst, uint32_t flags)
{
const uint64_t mask = AMDGPU_VM_PTE_COUNT - 1;
- uint64_t cur_pe_start, cur_pe_end, cur_dst;
+ uint64_t cur_pe_start, cur_nptes, cur_dst;
uint64_t addr; /* next GPU address to be updated */
uint64_t pt_idx;
struct amdgpu_bo *pt;
addr = start;
pt_idx = addr >> amdgpu_vm_block_size;
pt = vm->page_tables[pt_idx].entry.robj;
-
+ if (params->shadow) {
+ if (!pt->shadow)
+ return;
+ pt = vm->page_tables[pt_idx].entry.robj->shadow;
+ }
if ((addr & ~mask) == (end & ~mask))
nptes = end - addr;
else
cur_pe_start = amdgpu_bo_gpu_offset(pt);
cur_pe_start += (addr & mask) * 8;
- cur_pe_end = cur_pe_start + 8 * nptes;
+ cur_nptes = nptes;
cur_dst = dst;
/* for next ptb*/
while (addr < end) {
pt_idx = addr >> amdgpu_vm_block_size;
pt = vm->page_tables[pt_idx].entry.robj;
+ if (params->shadow) {
+ if (!pt->shadow)
+ return;
+ pt = vm->page_tables[pt_idx].entry.robj->shadow;
+ }
if ((addr & ~mask) == (end & ~mask))
nptes = end - addr;
next_pe_start = amdgpu_bo_gpu_offset(pt);
next_pe_start += (addr & mask) * 8;
- if (cur_pe_end == next_pe_start) {
+ if ((cur_pe_start + 8 * cur_nptes) == next_pe_start &&
+ ((cur_nptes + nptes) <= AMDGPU_VM_MAX_UPDATE_SIZE)) {
/* The next ptb is consecutive to current ptb.
- * Don't call amdgpu_vm_frag_ptes now.
+ * Don't call the update function now.
* Will update two ptbs together in future.
*/
- cur_pe_end += 8 * nptes;
+ cur_nptes += nptes;
} else {
- amdgpu_vm_frag_ptes(adev, vm_update_params,
- cur_pe_start, cur_pe_end,
- cur_dst, flags);
+ params->func(params, cur_pe_start, cur_dst, cur_nptes,
+ AMDGPU_GPU_PAGE_SIZE, flags);
cur_pe_start = next_pe_start;
- cur_pe_end = next_pe_start + 8 * nptes;
+ cur_nptes = nptes;
cur_dst = dst;
}
dst += nptes * AMDGPU_GPU_PAGE_SIZE;
}
- amdgpu_vm_frag_ptes(adev, vm_update_params, cur_pe_start,
- cur_pe_end, cur_dst, flags);
+ params->func(params, cur_pe_start, cur_dst, cur_nptes,
+ AMDGPU_GPU_PAGE_SIZE, flags);
+}
+
+/*
+ * amdgpu_vm_frag_ptes - add fragment information to PTEs
+ *
+ * @params: see amdgpu_pte_update_params definition
+ * @vm: requested vm
+ * @start: first PTE to handle
+ * @end: last PTE to handle
+ * @dst: addr those PTEs should point to
+ * @flags: hw mapping flags
+ */
+static void amdgpu_vm_frag_ptes(struct amdgpu_pte_update_params *params,
+ struct amdgpu_vm *vm,
+ uint64_t start, uint64_t end,
+ uint64_t dst, uint32_t flags)
+{
+ /**
+ * The MC L1 TLB supports variable sized pages, based on a fragment
+ * field in the PTE. When this field is set to a non-zero value, page
+ * granularity is increased from 4KB to (1 << (12 + frag)). The PTE
+ * flags are considered valid for all PTEs within the fragment range
+ * and corresponding mappings are assumed to be physically contiguous.
+ *
+ * The L1 TLB can store a single PTE for the whole fragment,
+ * significantly increasing the space available for translation
+ * caching. This leads to large improvements in throughput when the
+ * TLB is under pressure.
+ *
+ * The L2 TLB distributes small and large fragments into two
+ * asymmetric partitions. The large fragment cache is significantly
+ * larger. Thus, we try to use large fragments wherever possible.
+ * Userspace can support this by aligning virtual base address and
+ * allocation size to the fragment size.
+ */
+
+ const uint64_t frag_align = 1 << AMDGPU_LOG2_PAGES_PER_FRAG;
+
+ uint64_t frag_start = ALIGN(start, frag_align);
+ uint64_t frag_end = end & ~(frag_align - 1);
+
+ uint32_t frag;
+
+ /* system pages are non continuously */
+ if (params->src || !(flags & AMDGPU_PTE_VALID) ||
+ (frag_start >= frag_end)) {
+
+ amdgpu_vm_update_ptes(params, vm, start, end, dst, flags);
+ return;
+ }
+
+ /* use more than 64KB fragment size if possible */
+ frag = lower_32_bits(frag_start | frag_end);
+ frag = likely(frag) ? __ffs(frag) : 31;
+
+ /* handle the 4K area at the beginning */
+ if (start != frag_start) {
+ amdgpu_vm_update_ptes(params, vm, start, frag_start,
+ dst, flags);
+ dst += (frag_start - start) * AMDGPU_GPU_PAGE_SIZE;
+ }
+
+ /* handle the area in the middle */
+ amdgpu_vm_update_ptes(params, vm, frag_start, frag_end, dst,
+ flags | AMDGPU_PTE_FRAG(frag));
+
+ /* handle the 4K area at the end */
+ if (frag_end != end) {
+ dst += (frag_end - frag_start) * AMDGPU_GPU_PAGE_SIZE;
+ amdgpu_vm_update_ptes(params, vm, frag_end, end, dst, flags);
+ }
}
/**
void *owner = AMDGPU_FENCE_OWNER_VM;
unsigned nptes, ncmds, ndw;
struct amdgpu_job *job;
- struct amdgpu_vm_update_params vm_update_params;
+ struct amdgpu_pte_update_params params;
struct fence *f = NULL;
int r;
+ memset(¶ms, 0, sizeof(params));
+ params.adev = adev;
+ params.src = src;
+
ring = container_of(vm->entity.sched, struct amdgpu_ring, sched);
- memset(&vm_update_params, 0, sizeof(vm_update_params));
- vm_update_params.src = src;
- vm_update_params.pages_addr = pages_addr;
+
+ memset(¶ms, 0, sizeof(params));
+ params.adev = adev;
+ params.src = src;
/* sync to everything on unmapping */
if (!(flags & AMDGPU_PTE_VALID))
/* padding, etc. */
ndw = 64;
- if (vm_update_params.src) {
+ if (src) {
/* only copy commands needed */
ndw += ncmds * 7;
- } else if (vm_update_params.pages_addr) {
- /* header for write data commands */
- ndw += ncmds * 4;
+ params.func = amdgpu_vm_do_copy_ptes;
+
+ } else if (pages_addr) {
+ /* copy commands needed */
+ ndw += ncmds * 7;
- /* body of write data command */
+ /* and also PTEs */
ndw += nptes * 2;
+ params.func = amdgpu_vm_do_copy_ptes;
+
} else {
/* set page commands needed */
ndw += ncmds * 10;
/* two extra commands for begin/end of fragment */
ndw += 2 * 10;
+
+ params.func = amdgpu_vm_do_set_ptes;
}
r = amdgpu_job_alloc_with_ib(adev, ndw * 4, &job);
if (r)
return r;
- vm_update_params.ib = &job->ibs[0];
+ params.ib = &job->ibs[0];
+
+ if (!src && pages_addr) {
+ uint64_t *pte;
+ unsigned i;
+
+ /* Put the PTEs at the end of the IB. */
+ i = ndw - nptes * 2;
+ pte= (uint64_t *)&(job->ibs->ptr[i]);
+ params.src = job->ibs->gpu_addr + i * 4;
+
+ for (i = 0; i < nptes; ++i) {
+ pte[i] = amdgpu_vm_map_gart(pages_addr, addr + i *
+ AMDGPU_GPU_PAGE_SIZE);
+ pte[i] |= flags;
+ }
+ }
r = amdgpu_sync_fence(adev, &job->sync, exclusive);
if (r)
if (r)
goto error_free;
- amdgpu_vm_update_ptes(adev, &vm_update_params, vm, start,
- last + 1, addr, flags);
+ params.shadow = true;
+ amdgpu_vm_frag_ptes(¶ms, vm, start, last + 1, addr, flags);
+ params.shadow = false;
+ amdgpu_vm_frag_ptes(¶ms, vm, start, last + 1, addr, flags);
- amdgpu_ring_pad_ib(ring, vm_update_params.ib);
- WARN_ON(vm_update_params.ib->length_dw > ndw);
+ amdgpu_ring_pad_ib(ring, params.ib);
+ WARN_ON(params.ib->length_dw > ndw);
r = amdgpu_job_submit(job, ring, &vm->entity,
AMDGPU_FENCE_OWNER_VM, &f);
if (r)
*
* @adev: amdgpu_device pointer
* @bo_va: requested BO and VM object
- * @mem: ttm mem
+ * @clear: if true clear the entries
*
* Fill in the page table entries for @bo_va.
* Returns 0 for success, -EINVAL for failure.
- *
- * Object have to be reserved and mutex must be locked!
*/
int amdgpu_vm_bo_update(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va,
- struct ttm_mem_reg *mem)
+ bool clear)
{
struct amdgpu_vm *vm = bo_va->vm;
struct amdgpu_bo_va_mapping *mapping;
dma_addr_t *pages_addr = NULL;
uint32_t gtt_flags, flags;
+ struct ttm_mem_reg *mem;
struct fence *exclusive;
uint64_t addr;
int r;
- if (mem) {
+ if (clear) {
+ mem = NULL;
+ addr = 0;
+ exclusive = NULL;
+ } else {
struct ttm_dma_tt *ttm;
+ mem = &bo_va->bo->tbo.mem;
addr = (u64)mem->start << PAGE_SHIFT;
switch (mem->mem_type) {
case TTM_PL_TT:
}
exclusive = reservation_object_get_excl(bo_va->bo->tbo.resv);
- } else {
- addr = 0;
- exclusive = NULL;
}
flags = amdgpu_ttm_tt_pte_flags(adev, bo_va->bo->tbo.ttm, mem);
- gtt_flags = (adev == bo_va->bo->adev) ? flags : 0;
+ gtt_flags = (amdgpu_ttm_is_bound(bo_va->bo->tbo.ttm) &&
+ adev == bo_va->bo->adev) ? flags : 0;
spin_lock(&vm->status_lock);
if (!list_empty(&bo_va->vm_status))
spin_lock(&vm->status_lock);
list_splice_init(&bo_va->invalids, &bo_va->valids);
list_del_init(&bo_va->vm_status);
- if (!mem)
+ if (clear)
list_add(&bo_va->vm_status, &vm->cleared);
spin_unlock(&vm->status_lock);
struct amdgpu_bo_va, vm_status);
spin_unlock(&vm->status_lock);
- r = amdgpu_vm_bo_update(adev, bo_va, NULL);
+ r = amdgpu_vm_bo_update(adev, bo_va, true);
if (r)
return r;
r = amdgpu_bo_create(adev, AMDGPU_VM_PTE_COUNT * 8,
AMDGPU_GPU_PAGE_SIZE, true,
AMDGPU_GEM_DOMAIN_VRAM,
- AMDGPU_GEM_CREATE_NO_CPU_ACCESS,
+ AMDGPU_GEM_CREATE_NO_CPU_ACCESS |
+ AMDGPU_GEM_CREATE_SHADOW,
NULL, resv, &pt);
if (r)
goto error_free;
r = amdgpu_bo_create(adev, pd_size, align, true,
AMDGPU_GEM_DOMAIN_VRAM,
- AMDGPU_GEM_CREATE_NO_CPU_ACCESS,
+ AMDGPU_GEM_CREATE_NO_CPU_ACCESS |
+ AMDGPU_GEM_CREATE_SHADOW,
NULL, NULL, &vm->page_directory);
if (r)
goto error_free_sched_entity;
kfree(mapping);
}
- for (i = 0; i < amdgpu_vm_num_pdes(adev); i++)
+ for (i = 0; i < amdgpu_vm_num_pdes(adev); i++) {
+ if (vm->page_tables[i].entry.robj &&
+ vm->page_tables[i].entry.robj->shadow)
+ amdgpu_bo_unref(&vm->page_tables[i].entry.robj->shadow);
amdgpu_bo_unref(&vm->page_tables[i].entry.robj);
+ }
drm_free_large(vm->page_tables);
+ if (vm->page_directory->shadow)
+ amdgpu_bo_unref(&vm->page_directory->shadow);
amdgpu_bo_unref(&vm->page_directory);
fence_put(vm->page_directory_fence);
}
* SetPixelClock provides the dividers
*/
args.v6.ulDispEngClkFreq = cpu_to_le32(dispclk);
- args.v6.ucPpll = ATOM_EXT_PLL1;
+ if (adev->asic_type == CHIP_TAHITI ||
+ adev->asic_type == CHIP_PITCAIRN ||
+ adev->asic_type == CHIP_VERDE ||
+ adev->asic_type == CHIP_OLAND)
+ args.v6.ucPpll = ATOM_PPLL0;
+ else
+ args.v6.ucPpll = ATOM_EXT_PLL1;
break;
default:
DRM_ERROR("Unknown table version %d %d\n", frev, crev);
/* timeout */
if (args.v2.ucReplyStatus == 1) {
- DRM_DEBUG_KMS("dp_aux_ch timeout\n");
r = -ETIMEDOUT;
goto done;
}
{
struct amdgpu_connector_atom_dig *dig_connector = amdgpu_connector->con_priv;
u8 msg[DP_DPCD_SIZE];
- int ret, i;
+ int ret;
- for (i = 0; i < 7; i++) {
- ret = drm_dp_dpcd_read(&amdgpu_connector->ddc_bus->aux, DP_DPCD_REV, msg,
- DP_DPCD_SIZE);
- if (ret == DP_DPCD_SIZE) {
- memcpy(dig_connector->dpcd, msg, DP_DPCD_SIZE);
+ ret = drm_dp_dpcd_read(&amdgpu_connector->ddc_bus->aux, DP_DPCD_REV,
+ msg, DP_DPCD_SIZE);
+ if (ret == DP_DPCD_SIZE) {
+ memcpy(dig_connector->dpcd, msg, DP_DPCD_SIZE);
- DRM_DEBUG_KMS("DPCD: %*ph\n", (int)sizeof(dig_connector->dpcd),
- dig_connector->dpcd);
+ DRM_DEBUG_KMS("DPCD: %*ph\n", (int)sizeof(dig_connector->dpcd),
+ dig_connector->dpcd);
- amdgpu_atombios_dp_probe_oui(amdgpu_connector);
+ amdgpu_atombios_dp_probe_oui(amdgpu_connector);
- return 0;
- }
+ return 0;
}
+
dig_connector->dpcd[0] = 0;
return -EINVAL;
}
#include "amdgpu.h"
#include "atom.h"
#include "amdgpu_atombios.h"
+#include "atombios_i2c.h"
#define TARGET_HW_I2C_CLOCK 50
amdgpu_irq_put(adev, &adev->pm.dpm.thermal.irq,
AMDGPU_THERMAL_IRQ_HIGH_TO_LOW);
- ci_dpm_powergate_uvd(adev, false);
+ ci_dpm_powergate_uvd(adev, true);
if (!amdgpu_ci_is_smc_running(adev))
return;
pi->pcie_dpm_key_disabled = 0;
pi->thermal_sclk_dpm_enabled = 0;
- pi->caps_sclk_ds = true;
+ if (amdgpu_sclk_deep_sleep_en)
+ pi->caps_sclk_ds = true;
+ else
+ pi->caps_sclk_ds = false;
pi->mclk_strobe_mode_threshold = 40000;
pi->mclk_stutter_mode_threshold = 40000;
pi->caps_dynamic_ac_timing = true;
- pi->uvd_power_gated = false;
+ pi->uvd_power_gated = true;
/* make sure dc limits are valid */
if ((adev->pm.dpm.dyn_state.max_clock_voltage_on_dc.sclk == 0) ||
if (ret)
return ret;
- ci_dpm_powergate_uvd(adev, true);
-
return 0;
}
#include "amdgpu_amdkfd.h"
#include "amdgpu_powerplay.h"
+#include "dce_virtual.h"
/*
* Indirect registers accessor
},
};
+static const struct amdgpu_ip_block_version bonaire_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 8,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 7,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &gfx_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_sdma_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 4,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &uvd_v4_2_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vce_v2_0_ip_funcs,
+ },
+};
+
static const struct amdgpu_ip_block_version hawaii_ip_blocks[] =
{
/* ORDER MATTERS! */
},
};
+static const struct amdgpu_ip_block_version hawaii_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 8,
+ .minor = 5,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 7,
+ .minor = 3,
+ .rev = 0,
+ .funcs = &gfx_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_sdma_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 4,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &uvd_v4_2_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vce_v2_0_ip_funcs,
+ },
+};
+
static const struct amdgpu_ip_block_version kabini_ip_blocks[] =
{
/* ORDER MATTERS! */
},
};
+static const struct amdgpu_ip_block_version kabini_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 8,
+ .minor = 3,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 7,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &gfx_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_sdma_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 4,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &uvd_v4_2_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vce_v2_0_ip_funcs,
+ },
+};
+
static const struct amdgpu_ip_block_version mullins_ip_blocks[] =
{
/* ORDER MATTERS! */
},
};
+static const struct amdgpu_ip_block_version mullins_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 8,
+ .minor = 3,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 7,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &gfx_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_sdma_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 4,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &uvd_v4_2_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vce_v2_0_ip_funcs,
+ },
+};
+
static const struct amdgpu_ip_block_version kaveri_ip_blocks[] =
{
/* ORDER MATTERS! */
},
};
+static const struct amdgpu_ip_block_version kaveri_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 8,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 7,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &gfx_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cik_sdma_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 4,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &uvd_v4_2_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vce_v2_0_ip_funcs,
+ },
+};
+
int cik_set_ip_blocks(struct amdgpu_device *adev)
{
- switch (adev->asic_type) {
- case CHIP_BONAIRE:
- adev->ip_blocks = bonaire_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(bonaire_ip_blocks);
- break;
- case CHIP_HAWAII:
- adev->ip_blocks = hawaii_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(hawaii_ip_blocks);
- break;
- case CHIP_KAVERI:
- adev->ip_blocks = kaveri_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(kaveri_ip_blocks);
- break;
- case CHIP_KABINI:
- adev->ip_blocks = kabini_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(kabini_ip_blocks);
- break;
- case CHIP_MULLINS:
- adev->ip_blocks = mullins_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(mullins_ip_blocks);
- break;
- default:
- /* FIXME: not supported yet */
- return -EINVAL;
+ if (adev->enable_virtual_display) {
+ switch (adev->asic_type) {
+ case CHIP_BONAIRE:
+ adev->ip_blocks = bonaire_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(bonaire_ip_blocks_vd);
+ break;
+ case CHIP_HAWAII:
+ adev->ip_blocks = hawaii_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(hawaii_ip_blocks_vd);
+ break;
+ case CHIP_KAVERI:
+ adev->ip_blocks = kaveri_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(kaveri_ip_blocks_vd);
+ break;
+ case CHIP_KABINI:
+ adev->ip_blocks = kabini_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(kabini_ip_blocks_vd);
+ break;
+ case CHIP_MULLINS:
+ adev->ip_blocks = mullins_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(mullins_ip_blocks_vd);
+ break;
+ default:
+ /* FIXME: not supported yet */
+ return -EINVAL;
+ }
+ } else {
+ switch (adev->asic_type) {
+ case CHIP_BONAIRE:
+ adev->ip_blocks = bonaire_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(bonaire_ip_blocks);
+ break;
+ case CHIP_HAWAII:
+ adev->ip_blocks = hawaii_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(hawaii_ip_blocks);
+ break;
+ case CHIP_KAVERI:
+ adev->ip_blocks = kaveri_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(kaveri_ip_blocks);
+ break;
+ case CHIP_KABINI:
+ adev->ip_blocks = kabini_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(kabini_ip_blocks);
+ break;
+ case CHIP_MULLINS:
+ adev->ip_blocks = mullins_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(mullins_ip_blocks);
+ break;
+ default:
+ /* FIXME: not supported yet */
+ return -EINVAL;
+ }
}
return 0;
uint64_t pe, uint64_t src,
unsigned count)
{
- while (count) {
- unsigned bytes = count * 8;
- if (bytes > 0x1FFFF8)
- bytes = 0x1FFFF8;
-
- ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_COPY,
- SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
- ib->ptr[ib->length_dw++] = bytes;
- ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */
- ib->ptr[ib->length_dw++] = lower_32_bits(src);
- ib->ptr[ib->length_dw++] = upper_32_bits(src);
- ib->ptr[ib->length_dw++] = lower_32_bits(pe);
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
-
- pe += bytes;
- src += bytes;
- count -= bytes / 8;
- }
+ unsigned bytes = count * 8;
+
+ ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_COPY,
+ SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
+ ib->ptr[ib->length_dw++] = bytes;
+ ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */
+ ib->ptr[ib->length_dw++] = lower_32_bits(src);
+ ib->ptr[ib->length_dw++] = upper_32_bits(src);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
}
/**
*
* @ib: indirect buffer to fill with commands
* @pe: addr of the page entry
- * @addr: dst addr to write into pe
+ * @value: dst addr to write into pe
* @count: number of page entries to update
* @incr: increase next addr by incr bytes
- * @flags: access flags
*
* Update PTEs by writing them manually using sDMA (CIK).
*/
-static void cik_sdma_vm_write_pte(struct amdgpu_ib *ib,
- const dma_addr_t *pages_addr, uint64_t pe,
- uint64_t addr, unsigned count,
- uint32_t incr, uint32_t flags)
+static void cik_sdma_vm_write_pte(struct amdgpu_ib *ib, uint64_t pe,
+ uint64_t value, unsigned count,
+ uint32_t incr)
{
- uint64_t value;
- unsigned ndw;
-
- while (count) {
- ndw = count * 2;
- if (ndw > 0xFFFFE)
- ndw = 0xFFFFE;
-
- /* for non-physically contiguous pages (system) */
- ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_WRITE,
- SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
- ib->ptr[ib->length_dw++] = pe;
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
- ib->ptr[ib->length_dw++] = ndw;
- for (; ndw > 0; ndw -= 2, --count, pe += 8) {
- value = amdgpu_vm_map_gart(pages_addr, addr);
- addr += incr;
- value |= flags;
- ib->ptr[ib->length_dw++] = value;
- ib->ptr[ib->length_dw++] = upper_32_bits(value);
- }
+ unsigned ndw = count * 2;
+
+ ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_WRITE,
+ SDMA_WRITE_SUB_OPCODE_LINEAR, 0);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
+ ib->ptr[ib->length_dw++] = ndw;
+ for (; ndw > 0; ndw -= 2) {
+ ib->ptr[ib->length_dw++] = lower_32_bits(value);
+ ib->ptr[ib->length_dw++] = upper_32_bits(value);
+ value += incr;
}
}
*
* Update the page tables using sDMA (CIK).
*/
-static void cik_sdma_vm_set_pte_pde(struct amdgpu_ib *ib,
- uint64_t pe,
+static void cik_sdma_vm_set_pte_pde(struct amdgpu_ib *ib, uint64_t pe,
uint64_t addr, unsigned count,
uint32_t incr, uint32_t flags)
{
- uint64_t value;
- unsigned ndw;
-
- while (count) {
- ndw = count;
- if (ndw > 0x7FFFF)
- ndw = 0x7FFFF;
-
- if (flags & AMDGPU_PTE_VALID)
- value = addr;
- else
- value = 0;
-
- /* for physically contiguous pages (vram) */
- ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_GENERATE_PTE_PDE, 0, 0);
- ib->ptr[ib->length_dw++] = pe; /* dst addr */
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
- ib->ptr[ib->length_dw++] = flags; /* mask */
- ib->ptr[ib->length_dw++] = 0;
- ib->ptr[ib->length_dw++] = value; /* value */
- ib->ptr[ib->length_dw++] = upper_32_bits(value);
- ib->ptr[ib->length_dw++] = incr; /* increment size */
- ib->ptr[ib->length_dw++] = 0;
- ib->ptr[ib->length_dw++] = ndw; /* number of entries */
-
- pe += ndw * 8;
- addr += ndw * incr;
- count -= ndw;
- }
+ /* for physically contiguous pages (vram) */
+ ib->ptr[ib->length_dw++] = SDMA_PACKET(SDMA_OPCODE_GENERATE_PTE_PDE, 0, 0);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe); /* dst addr */
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
+ ib->ptr[ib->length_dw++] = flags; /* mask */
+ ib->ptr[ib->length_dw++] = 0;
+ ib->ptr[ib->length_dw++] = lower_32_bits(addr); /* value */
+ ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+ ib->ptr[ib->length_dw++] = incr; /* increment size */
+ ib->ptr[ib->length_dw++] = 0;
+ ib->ptr[ib->length_dw++] = count; /* number of entries */
}
/**
amdgpu_ring_write(ring, (0xfff << 16) | 10); /* retry count, poll interval */
}
+static unsigned cik_sdma_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 7 + 4; /* cik_sdma_ring_emit_ib */
+}
+
+static unsigned cik_sdma_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 6 + /* cik_sdma_ring_emit_hdp_flush */
+ 3 + /* cik_sdma_ring_emit_hdp_invalidate */
+ 6 + /* cik_sdma_ring_emit_pipeline_sync */
+ 12 + /* cik_sdma_ring_emit_vm_flush */
+ 9 + 9 + 9; /* cik_sdma_ring_emit_fence x3 for user fence, vm fence */
+}
+
static void cik_enable_sdma_mgcg(struct amdgpu_device *adev,
bool enable)
{
.test_ib = cik_sdma_ring_test_ib,
.insert_nop = cik_sdma_ring_insert_nop,
.pad_ib = cik_sdma_ring_pad_ib,
+ .get_emit_ib_size = cik_sdma_ring_get_emit_ib_size,
+ .get_dma_frame_size = cik_sdma_ring_get_dma_frame_size,
};
static void cik_sdma_set_ring_funcs(struct amdgpu_device *adev)
static void cz_dpm_powergate_uvd(struct amdgpu_device *adev, bool gate);
static void cz_dpm_powergate_vce(struct amdgpu_device *adev, bool gate);
+static void cz_dpm_fini(struct amdgpu_device *adev);
static struct cz_ps *cz_get_ps(struct amdgpu_ps *rps)
{
ps = kzalloc(sizeof(struct cz_ps), GFP_KERNEL);
if (ps == NULL) {
+ for (j = 0; j < i; j++)
+ kfree(adev->pm.dpm.ps[j].ps_priv);
kfree(adev->pm.dpm.ps);
return -ENOMEM;
}
ret = amdgpu_get_platform_caps(adev);
if (ret)
- return ret;
+ goto err;
ret = amdgpu_parse_extended_power_table(adev);
if (ret)
- return ret;
+ goto err;
pi->sram_end = SMC_RAM_END;
pi->caps_td_ramping = true;
pi->caps_tcp_ramping = true;
}
- pi->caps_sclk_ds = true;
+ if (amdgpu_sclk_deep_sleep_en)
+ pi->caps_sclk_ds = true;
+ else
+ pi->caps_sclk_ds = false;
+
pi->voting_clients = 0x00c00033;
pi->auto_thermal_throttling_enabled = true;
pi->bapm_enabled = false;
ret = cz_parse_sys_info_table(adev);
if (ret)
- return ret;
+ goto err;
cz_patch_voltage_values(adev);
cz_construct_boot_state(adev);
ret = cz_parse_power_table(adev);
if (ret)
- return ret;
+ goto err;
ret = cz_process_firmware_header(adev);
if (ret)
- return ret;
+ goto err;
pi->dpm_enabled = true;
pi->uvd_dynamic_pg = false;
return 0;
+err:
+ cz_dpm_fini(adev);
+ return ret;
}
static void cz_dpm_fini(struct amdgpu_device *adev)
struct cz_power_info *pi = cz_get_pi(adev);
pi->active_process_mask = 0;
-
}
static int cz_dpm_download_pptable_from_smu(struct amdgpu_device *adev,
void **table)
{
- int ret = 0;
-
- ret = cz_smu_download_pptable(adev, table);
-
- return ret;
+ return cz_smu_download_pptable(adev, table);
}
static int cz_dpm_upload_pptable_to_smu(struct amdgpu_device *adev)
pi->sclk_dpm.hard_min_clk = 0;
cz_send_msg_to_smc(adev, PPSMC_MSG_GetMaxSclkLevel);
level = cz_get_argument(adev);
- if (level < table->count)
+ if (level < table->count) {
clock = table->entries[level].clk;
- else {
+ } else {
DRM_ERROR("Invalid SLCK Voltage Dependency table entry.\n");
clock = table->entries[table->count - 1].clk;
}
pi->uvd_dpm.hard_min_clk = 0;
cz_send_msg_to_smc(adev, PPSMC_MSG_GetMaxUvdLevel);
level = cz_get_argument(adev);
- if (level < table->count)
+ if (level < table->count) {
clock = table->entries[level].vclk;
- else {
+ } else {
DRM_ERROR("Invalid UVD Voltage Dependency table entry.\n");
clock = table->entries[table->count - 1].vclk;
}
pi->vce_dpm.hard_min_clk = table->entries[0].ecclk;
cz_send_msg_to_smc(adev, PPSMC_MSG_GetMaxEclkLevel);
level = cz_get_argument(adev);
- if (level < table->count)
+ if (level < table->count) {
clock = table->entries[level].ecclk;
- else {
+ } else {
/* future BIOS would fix this error */
DRM_ERROR("Invalid VCE Voltage Dependency table entry.\n");
clock = table->entries[table->count - 1].ecclk;
pi->acp_dpm.hard_min_clk = 0;
cz_send_msg_to_smc(adev, PPSMC_MSG_GetMaxAclkLevel);
level = cz_get_argument(adev);
- if (level < table->count)
+ if (level < table->count) {
clock = table->entries[level].clk;
- else {
+ } else {
DRM_ERROR("Invalid ACP Voltage Dependency table entry.\n");
clock = table->entries[table->count - 1].clk;
}
struct cz_power_info *pi = cz_get_pi(adev);
pi->low_sclk_interrupt_threshold = 0;
-
}
static void cz_dpm_setup_asic(struct amdgpu_device *adev)
int ret;
if (pi->caps_sq_ramping || pi->caps_db_ramping ||
- pi->caps_td_ramping || pi->caps_tcp_ramping) {
+ pi->caps_td_ramping || pi->caps_tcp_ramping) {
if (adev->gfx.gfx_current_status != AMDGPU_GFX_SAFE_MODE) {
ret = cz_disable_cgpg(adev);
if (ret) {
ps->force_high = false;
ps->need_dfs_bypass = true;
pi->video_start = new_rps->dclk || new_rps->vclk ||
- new_rps->evclk || new_rps->ecclk;
+ new_rps->evclk || new_rps->ecclk;
if ((new_rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) ==
ATOM_PPLIB_CLASSIFICATION_UI_BATTERY)
}
cz_reset_acp_boot_level(adev);
-
cz_update_current_ps(adev, adev->pm.dpm.boot_ps);
return 0;
struct amdgpu_ps *ps = &pi->requested_rps;
cz_update_current_ps(adev, ps);
-
}
static int cz_dpm_force_highest(struct amdgpu_device *adev)
/* disable clockgating so we can properly shut down the block */
ret = amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
AMD_CG_STATE_UNGATE);
+ if (ret) {
+ DRM_ERROR("UVD DPM Power Gating failed to set clockgating state\n");
+ return;
+ }
+
/* shutdown the UVD block */
ret = amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
AMD_PG_STATE_GATE);
- /* XXX: check for errors */
+
+ if (ret) {
+ DRM_ERROR("UVD DPM Power Gating failed to set powergating state\n");
+ return;
+ }
}
cz_update_uvd_dpm(adev, gate);
- if (pi->caps_uvd_pg)
+ if (pi->caps_uvd_pg) {
/* power off the UVD block */
- cz_send_msg_to_smc(adev, PPSMC_MSG_UVDPowerOFF);
+ ret = cz_send_msg_to_smc(adev, PPSMC_MSG_UVDPowerOFF);
+ if (ret) {
+ DRM_ERROR("UVD DPM Power Gating failed to send SMU PowerOFF message\n");
+ return;
+ }
+ }
} else {
if (pi->caps_uvd_pg) {
/* power on the UVD block */
if (pi->uvd_dynamic_pg)
- cz_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_UVDPowerON, 1);
+ ret = cz_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_UVDPowerON, 1);
else
- cz_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_UVDPowerON, 0);
+ ret = cz_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_UVDPowerON, 0);
+
+ if (ret) {
+ DRM_ERROR("UVD DPM Power Gating Failed to send SMU PowerON message\n");
+ return;
+ }
+
/* re-init the UVD block */
ret = amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
AMD_PG_STATE_UNGATE);
+
+ if (ret) {
+ DRM_ERROR("UVD DPM Power Gating Failed to set powergating state\n");
+ return;
+ }
+
/* enable clockgating. hw will dynamically gate/ungate clocks on the fly */
ret = amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
AMD_CG_STATE_GATE);
- /* XXX: check for errors */
+ if (ret) {
+ DRM_ERROR("UVD DPM Power Gating Failed to set clockgating state\n");
+ return;
+ }
}
cz_update_uvd_dpm(adev, gate);
}
/* Stable Pstate is enabled and we need to set the VCE DPM to highest level */
if (pi->caps_stable_power_state) {
pi->vce_dpm.hard_min_clk = table->entries[table->count-1].ecclk;
-
} else { /* non-stable p-state cases. without vce.Arbiter.EcclkHardMin */
/* leave it as set by user */
/*pi->vce_dpm.hard_min_clk = table->entries[0].ecclk;*/
#include "cz_smumgr.h"
#include "smu_ucode_xfer_cz.h"
#include "amdgpu_ucode.h"
+#include "cz_dpm.h"
+#include "vi_dpm.h"
#include "smu/smu_8_0_d.h"
#include "smu/smu_8_0_sh_mask.h"
return priv;
}
-int cz_send_msg_to_smc_async(struct amdgpu_device *adev, u16 msg)
+static int cz_send_msg_to_smc_async(struct amdgpu_device *adev, u16 msg)
{
int i;
u32 content = 0, tmp;
return 0;
}
-int cz_write_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address,
+static int cz_write_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address,
u32 value, u32 limit)
{
int ret;
if (save->crtc_enabled[i]) {
tmp = RREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i]);
- if (REG_GET_FIELD(tmp, MASTER_UPDATE_MODE, MASTER_UPDATE_MODE) != 3) {
- tmp = REG_SET_FIELD(tmp, MASTER_UPDATE_MODE, MASTER_UPDATE_MODE, 3);
+ if (REG_GET_FIELD(tmp, MASTER_UPDATE_MODE, MASTER_UPDATE_MODE) != 0) {
+ tmp = REG_SET_FIELD(tmp, MASTER_UPDATE_MODE, MASTER_UPDATE_MODE, 0);
WREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i], tmp);
}
tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]);
WREG32(mmVGA_RENDER_CONTROL, tmp);
}
+static int dce_v10_0_get_num_crtc(struct amdgpu_device *adev)
+{
+ int num_crtc = 0;
+
+ switch (adev->asic_type) {
+ case CHIP_FIJI:
+ case CHIP_TONGA:
+ num_crtc = 6;
+ break;
+ default:
+ num_crtc = 0;
+ }
+ return num_crtc;
+}
+
+void dce_v10_0_disable_dce(struct amdgpu_device *adev)
+{
+ /*Disable VGA render and enabled crtc, if has DCE engine*/
+ if (amdgpu_atombios_has_dce_engine_info(adev)) {
+ u32 tmp;
+ int crtc_enabled, i;
+
+ dce_v10_0_set_vga_render_state(adev, false);
+
+ /*Disable crtc*/
+ for (i = 0; i < dce_v10_0_get_num_crtc(adev); i++) {
+ crtc_enabled = REG_GET_FIELD(RREG32(mmCRTC_CONTROL + crtc_offsets[i]),
+ CRTC_CONTROL, CRTC_MASTER_EN);
+ if (crtc_enabled) {
+ WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1);
+ tmp = RREG32(mmCRTC_CONTROL + crtc_offsets[i]);
+ tmp = REG_SET_FIELD(tmp, CRTC_CONTROL, CRTC_MASTER_EN, 0);
+ WREG32(mmCRTC_CONTROL + crtc_offsets[i], tmp);
+ WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0);
+ }
+ }
+ }
+}
+
static void dce_v10_0_program_fmt(struct drm_encoder *encoder)
{
struct drm_device *dev = encoder->dev;
WREG32(mmVIEWPORT_SIZE + amdgpu_crtc->crtc_offset,
(viewport_w << 16) | viewport_h);
- /* set pageflip to happen only at start of vblank interval (front porch) */
- WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 3);
+ /* set pageflip to happen anywhere in vblank interval */
+ WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
if (!atomic && fb && fb != crtc->primary->fb) {
amdgpu_fb = to_amdgpu_framebuffer(fb);
.gamma_set = dce_v10_0_crtc_gamma_set,
.set_config = amdgpu_crtc_set_config,
.destroy = dce_v10_0_crtc_destroy,
- .page_flip = amdgpu_crtc_page_flip,
+ .page_flip_target = amdgpu_crtc_page_flip_target,
};
static void dce_v10_0_crtc_dpms(struct drm_crtc *crtc, int mode)
dce_v10_0_set_display_funcs(adev);
dce_v10_0_set_irq_funcs(adev);
+ adev->mode_info.num_crtc = dce_v10_0_get_num_crtc(adev);
+
switch (adev->asic_type) {
case CHIP_FIJI:
case CHIP_TONGA:
- adev->mode_info.num_crtc = 6; /* XXX 7??? */
adev->mode_info.num_hpd = 6;
adev->mode_info.num_dig = 7;
break;
return 0;
}
+static int dce_v10_0_check_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (dce_v10_0_is_display_hung(adev))
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_DCE].hang = true;
+ else
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_DCE].hang = false;
+
+ return 0;
+}
+
static int dce_v10_0_soft_reset(void *handle)
{
u32 srbm_soft_reset = 0, tmp;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_DCE].hang)
+ return 0;
+
if (dce_v10_0_is_display_hung(adev))
srbm_soft_reset |= SRBM_SOFT_RESET__SOFT_RESET_DC_MASK;
.resume = dce_v10_0_resume,
.is_idle = dce_v10_0_is_idle,
.wait_for_idle = dce_v10_0_wait_for_idle,
+ .check_soft_reset = dce_v10_0_check_soft_reset,
.soft_reset = dce_v10_0_soft_reset,
.set_clockgating_state = dce_v10_0_set_clockgating_state,
.set_powergating_state = dce_v10_0_set_powergating_state,
extern const struct amd_ip_funcs dce_v10_0_ip_funcs;
+void dce_v10_0_disable_dce(struct amdgpu_device *adev);
+
#endif
WREG32(mmVGA_RENDER_CONTROL, tmp);
}
+static int dce_v11_0_get_num_crtc (struct amdgpu_device *adev)
+{
+ int num_crtc = 0;
+
+ switch (adev->asic_type) {
+ case CHIP_CARRIZO:
+ num_crtc = 3;
+ break;
+ case CHIP_STONEY:
+ num_crtc = 2;
+ break;
+ case CHIP_POLARIS10:
+ num_crtc = 6;
+ break;
+ case CHIP_POLARIS11:
+ num_crtc = 5;
+ break;
+ default:
+ num_crtc = 0;
+ }
+ return num_crtc;
+}
+
+void dce_v11_0_disable_dce(struct amdgpu_device *adev)
+{
+ /*Disable VGA render and enabled crtc, if has DCE engine*/
+ if (amdgpu_atombios_has_dce_engine_info(adev)) {
+ u32 tmp;
+ int crtc_enabled, i;
+
+ dce_v11_0_set_vga_render_state(adev, false);
+
+ /*Disable crtc*/
+ for (i = 0; i < dce_v11_0_get_num_crtc(adev); i++) {
+ crtc_enabled = REG_GET_FIELD(RREG32(mmCRTC_CONTROL + crtc_offsets[i]),
+ CRTC_CONTROL, CRTC_MASTER_EN);
+ if (crtc_enabled) {
+ WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1);
+ tmp = RREG32(mmCRTC_CONTROL + crtc_offsets[i]);
+ tmp = REG_SET_FIELD(tmp, CRTC_CONTROL, CRTC_MASTER_EN, 0);
+ WREG32(mmCRTC_CONTROL + crtc_offsets[i], tmp);
+ WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0);
+ }
+ }
+ }
+}
+
static void dce_v11_0_program_fmt(struct drm_encoder *encoder)
{
struct drm_device *dev = encoder->dev;
WREG32(mmVIEWPORT_SIZE + amdgpu_crtc->crtc_offset,
(viewport_w << 16) | viewport_h);
- /* set pageflip to happen only at start of vblank interval (front porch) */
- WREG32(mmCRTC_MASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 3);
+ /* set pageflip to happen anywhere in vblank interval */
+ WREG32(mmCRTC_MASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
if (!atomic && fb && fb != crtc->primary->fb) {
amdgpu_fb = to_amdgpu_framebuffer(fb);
.gamma_set = dce_v11_0_crtc_gamma_set,
.set_config = amdgpu_crtc_set_config,
.destroy = dce_v11_0_crtc_destroy,
- .page_flip = amdgpu_crtc_page_flip,
+ .page_flip_target = amdgpu_crtc_page_flip_target,
};
static void dce_v11_0_crtc_dpms(struct drm_crtc *crtc, int mode)
dce_v11_0_set_display_funcs(adev);
dce_v11_0_set_irq_funcs(adev);
+ adev->mode_info.num_crtc = dce_v11_0_get_num_crtc(adev);
+
switch (adev->asic_type) {
case CHIP_CARRIZO:
- adev->mode_info.num_crtc = 3;
adev->mode_info.num_hpd = 6;
adev->mode_info.num_dig = 9;
break;
case CHIP_STONEY:
- adev->mode_info.num_crtc = 2;
adev->mode_info.num_hpd = 6;
adev->mode_info.num_dig = 9;
break;
case CHIP_POLARIS10:
- adev->mode_info.num_crtc = 6;
adev->mode_info.num_hpd = 6;
adev->mode_info.num_dig = 6;
break;
case CHIP_POLARIS11:
- adev->mode_info.num_crtc = 5;
adev->mode_info.num_hpd = 5;
adev->mode_info.num_dig = 5;
break;
extern const struct amd_ip_funcs dce_v11_0_ip_funcs;
+void dce_v11_0_disable_dce(struct amdgpu_device *adev);
+
#endif
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "drmP.h"
+#include "amdgpu.h"
+#include "amdgpu_pm.h"
+#include "amdgpu_i2c.h"
+#include "atom.h"
+#include "amdgpu_atombios.h"
+#include "atombios_crtc.h"
+#include "atombios_encoders.h"
+#include "amdgpu_pll.h"
+#include "amdgpu_connectors.h"
+#include "si/si_reg.h"
+#include "si/sid.h"
+
+static void dce_v6_0_set_display_funcs(struct amdgpu_device *adev);
+static void dce_v6_0_set_irq_funcs(struct amdgpu_device *adev);
+
+static const u32 crtc_offsets[6] =
+{
+ SI_CRTC0_REGISTER_OFFSET,
+ SI_CRTC1_REGISTER_OFFSET,
+ SI_CRTC2_REGISTER_OFFSET,
+ SI_CRTC3_REGISTER_OFFSET,
+ SI_CRTC4_REGISTER_OFFSET,
+ SI_CRTC5_REGISTER_OFFSET
+};
+
+static const uint32_t dig_offsets[] = {
+ SI_CRTC0_REGISTER_OFFSET,
+ SI_CRTC1_REGISTER_OFFSET,
+ SI_CRTC2_REGISTER_OFFSET,
+ SI_CRTC3_REGISTER_OFFSET,
+ SI_CRTC4_REGISTER_OFFSET,
+ SI_CRTC5_REGISTER_OFFSET,
+ (0x13830 - 0x7030) >> 2,
+};
+
+static const struct {
+ uint32_t reg;
+ uint32_t vblank;
+ uint32_t vline;
+ uint32_t hpd;
+
+} interrupt_status_offsets[6] = { {
+ .reg = DISP_INTERRUPT_STATUS,
+ .vblank = DISP_INTERRUPT_STATUS__LB_D1_VBLANK_INTERRUPT_MASK,
+ .vline = DISP_INTERRUPT_STATUS__LB_D1_VLINE_INTERRUPT_MASK,
+ .hpd = DISP_INTERRUPT_STATUS__DC_HPD1_INTERRUPT_MASK
+}, {
+ .reg = DISP_INTERRUPT_STATUS_CONTINUE,
+ .vblank = DISP_INTERRUPT_STATUS_CONTINUE__LB_D2_VBLANK_INTERRUPT_MASK,
+ .vline = DISP_INTERRUPT_STATUS_CONTINUE__LB_D2_VLINE_INTERRUPT_MASK,
+ .hpd = DISP_INTERRUPT_STATUS_CONTINUE__DC_HPD2_INTERRUPT_MASK
+}, {
+ .reg = DISP_INTERRUPT_STATUS_CONTINUE2,
+ .vblank = DISP_INTERRUPT_STATUS_CONTINUE2__LB_D3_VBLANK_INTERRUPT_MASK,
+ .vline = DISP_INTERRUPT_STATUS_CONTINUE2__LB_D3_VLINE_INTERRUPT_MASK,
+ .hpd = DISP_INTERRUPT_STATUS_CONTINUE2__DC_HPD3_INTERRUPT_MASK
+}, {
+ .reg = DISP_INTERRUPT_STATUS_CONTINUE3,
+ .vblank = DISP_INTERRUPT_STATUS_CONTINUE3__LB_D4_VBLANK_INTERRUPT_MASK,
+ .vline = DISP_INTERRUPT_STATUS_CONTINUE3__LB_D4_VLINE_INTERRUPT_MASK,
+ .hpd = DISP_INTERRUPT_STATUS_CONTINUE3__DC_HPD4_INTERRUPT_MASK
+}, {
+ .reg = DISP_INTERRUPT_STATUS_CONTINUE4,
+ .vblank = DISP_INTERRUPT_STATUS_CONTINUE4__LB_D5_VBLANK_INTERRUPT_MASK,
+ .vline = DISP_INTERRUPT_STATUS_CONTINUE4__LB_D5_VLINE_INTERRUPT_MASK,
+ .hpd = DISP_INTERRUPT_STATUS_CONTINUE4__DC_HPD5_INTERRUPT_MASK
+}, {
+ .reg = DISP_INTERRUPT_STATUS_CONTINUE5,
+ .vblank = DISP_INTERRUPT_STATUS_CONTINUE5__LB_D6_VBLANK_INTERRUPT_MASK,
+ .vline = DISP_INTERRUPT_STATUS_CONTINUE5__LB_D6_VLINE_INTERRUPT_MASK,
+ .hpd = DISP_INTERRUPT_STATUS_CONTINUE5__DC_HPD6_INTERRUPT_MASK
+} };
+
+static const uint32_t hpd_int_control_offsets[6] = {
+ DC_HPD1_INT_CONTROL,
+ DC_HPD2_INT_CONTROL,
+ DC_HPD3_INT_CONTROL,
+ DC_HPD4_INT_CONTROL,
+ DC_HPD5_INT_CONTROL,
+ DC_HPD6_INT_CONTROL,
+};
+
+static u32 dce_v6_0_audio_endpt_rreg(struct amdgpu_device *adev,
+ u32 block_offset, u32 reg)
+{
+ DRM_INFO("xxxx: dce_v6_0_audio_endpt_rreg ----no impl!!!!\n");
+ return 0;
+}
+
+static void dce_v6_0_audio_endpt_wreg(struct amdgpu_device *adev,
+ u32 block_offset, u32 reg, u32 v)
+{
+ DRM_INFO("xxxx: dce_v6_0_audio_endpt_wreg ----no impl!!!!\n");
+}
+
+static bool dce_v6_0_is_in_vblank(struct amdgpu_device *adev, int crtc)
+{
+ if (RREG32(EVERGREEN_CRTC_STATUS + crtc_offsets[crtc]) & EVERGREEN_CRTC_V_BLANK)
+ return true;
+ else
+ return false;
+}
+
+static bool dce_v6_0_is_counter_moving(struct amdgpu_device *adev, int crtc)
+{
+ u32 pos1, pos2;
+
+ pos1 = RREG32(EVERGREEN_CRTC_STATUS_POSITION + crtc_offsets[crtc]);
+ pos2 = RREG32(EVERGREEN_CRTC_STATUS_POSITION + crtc_offsets[crtc]);
+
+ if (pos1 != pos2)
+ return true;
+ else
+ return false;
+}
+
+/**
+ * dce_v6_0_wait_for_vblank - vblank wait asic callback.
+ *
+ * @crtc: crtc to wait for vblank on
+ *
+ * Wait for vblank on the requested crtc (evergreen+).
+ */
+static void dce_v6_0_vblank_wait(struct amdgpu_device *adev, int crtc)
+{
+ unsigned i = 0;
+
+ if (crtc >= adev->mode_info.num_crtc)
+ return;
+
+ if (!(RREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[crtc]) & EVERGREEN_CRTC_MASTER_EN))
+ return;
+
+ /* depending on when we hit vblank, we may be close to active; if so,
+ * wait for another frame.
+ */
+ while (dce_v6_0_is_in_vblank(adev, crtc)) {
+ if (i++ % 100 == 0) {
+ if (!dce_v6_0_is_counter_moving(adev, crtc))
+ break;
+ }
+ }
+
+ while (!dce_v6_0_is_in_vblank(adev, crtc)) {
+ if (i++ % 100 == 0) {
+ if (!dce_v6_0_is_counter_moving(adev, crtc))
+ break;
+ }
+ }
+}
+
+static u32 dce_v6_0_vblank_get_counter(struct amdgpu_device *adev, int crtc)
+{
+ if (crtc >= adev->mode_info.num_crtc)
+ return 0;
+ else
+ return RREG32(CRTC_STATUS_FRAME_COUNT + crtc_offsets[crtc]);
+}
+
+static void dce_v6_0_pageflip_interrupt_init(struct amdgpu_device *adev)
+{
+ unsigned i;
+
+ /* Enable pflip interrupts */
+ for (i = 0; i <= adev->mode_info.num_crtc; i++)
+ amdgpu_irq_get(adev, &adev->pageflip_irq, i);
+}
+
+static void dce_v6_0_pageflip_interrupt_fini(struct amdgpu_device *adev)
+{
+ unsigned i;
+
+ /* Disable pflip interrupts */
+ for (i = 0; i <= adev->mode_info.num_crtc; i++)
+ amdgpu_irq_put(adev, &adev->pageflip_irq, i);
+}
+
+/**
+ * dce_v6_0_page_flip - pageflip callback.
+ *
+ * @adev: amdgpu_device pointer
+ * @crtc_id: crtc to cleanup pageflip on
+ * @crtc_base: new address of the crtc (GPU MC address)
+ *
+ * Does the actual pageflip (evergreen+).
+ * During vblank we take the crtc lock and wait for the update_pending
+ * bit to go high, when it does, we release the lock, and allow the
+ * double buffered update to take place.
+ * Returns the current update pending status.
+ */
+static void dce_v6_0_page_flip(struct amdgpu_device *adev,
+ int crtc_id, u64 crtc_base, bool async)
+{
+ struct amdgpu_crtc *amdgpu_crtc = adev->mode_info.crtcs[crtc_id];
+
+ /* flip at hsync for async, default is vsync */
+ WREG32(EVERGREEN_GRPH_FLIP_CONTROL + amdgpu_crtc->crtc_offset, async ?
+ EVERGREEN_GRPH_SURFACE_UPDATE_H_RETRACE_EN : 0);
+ /* update the scanout addresses */
+ WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset,
+ upper_32_bits(crtc_base));
+ WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset,
+ (u32)crtc_base);
+
+ /* post the write */
+ RREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset);
+}
+
+static int dce_v6_0_crtc_get_scanoutpos(struct amdgpu_device *adev, int crtc,
+ u32 *vbl, u32 *position)
+{
+ if ((crtc < 0) || (crtc >= adev->mode_info.num_crtc))
+ return -EINVAL;
+ *vbl = RREG32(EVERGREEN_CRTC_V_BLANK_START_END + crtc_offsets[crtc]);
+ *position = RREG32(EVERGREEN_CRTC_STATUS_POSITION + crtc_offsets[crtc]);
+
+ return 0;
+
+}
+
+/**
+ * dce_v6_0_hpd_sense - hpd sense callback.
+ *
+ * @adev: amdgpu_device pointer
+ * @hpd: hpd (hotplug detect) pin
+ *
+ * Checks if a digital monitor is connected (evergreen+).
+ * Returns true if connected, false if not connected.
+ */
+static bool dce_v6_0_hpd_sense(struct amdgpu_device *adev,
+ enum amdgpu_hpd_id hpd)
+{
+ bool connected = false;
+
+ switch (hpd) {
+ case AMDGPU_HPD_1:
+ if (RREG32(DC_HPD1_INT_STATUS) & DC_HPDx_SENSE)
+ connected = true;
+ break;
+ case AMDGPU_HPD_2:
+ if (RREG32(DC_HPD2_INT_STATUS) & DC_HPDx_SENSE)
+ connected = true;
+ break;
+ case AMDGPU_HPD_3:
+ if (RREG32(DC_HPD3_INT_STATUS) & DC_HPDx_SENSE)
+ connected = true;
+ break;
+ case AMDGPU_HPD_4:
+ if (RREG32(DC_HPD4_INT_STATUS) & DC_HPDx_SENSE)
+ connected = true;
+ break;
+ case AMDGPU_HPD_5:
+ if (RREG32(DC_HPD5_INT_STATUS) & DC_HPDx_SENSE)
+ connected = true;
+ break;
+ case AMDGPU_HPD_6:
+ if (RREG32(DC_HPD6_INT_STATUS) & DC_HPDx_SENSE)
+ connected = true;
+ break;
+ default:
+ break;
+ }
+
+ return connected;
+}
+
+/**
+ * dce_v6_0_hpd_set_polarity - hpd set polarity callback.
+ *
+ * @adev: amdgpu_device pointer
+ * @hpd: hpd (hotplug detect) pin
+ *
+ * Set the polarity of the hpd pin (evergreen+).
+ */
+static void dce_v6_0_hpd_set_polarity(struct amdgpu_device *adev,
+ enum amdgpu_hpd_id hpd)
+{
+ u32 tmp;
+ bool connected = dce_v6_0_hpd_sense(adev, hpd);
+
+ switch (hpd) {
+ case AMDGPU_HPD_1:
+ tmp = RREG32(DC_HPD1_INT_CONTROL);
+ if (connected)
+ tmp &= ~DC_HPDx_INT_POLARITY;
+ else
+ tmp |= DC_HPDx_INT_POLARITY;
+ WREG32(DC_HPD1_INT_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_2:
+ tmp = RREG32(DC_HPD2_INT_CONTROL);
+ if (connected)
+ tmp &= ~DC_HPDx_INT_POLARITY;
+ else
+ tmp |= DC_HPDx_INT_POLARITY;
+ WREG32(DC_HPD2_INT_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_3:
+ tmp = RREG32(DC_HPD3_INT_CONTROL);
+ if (connected)
+ tmp &= ~DC_HPDx_INT_POLARITY;
+ else
+ tmp |= DC_HPDx_INT_POLARITY;
+ WREG32(DC_HPD3_INT_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_4:
+ tmp = RREG32(DC_HPD4_INT_CONTROL);
+ if (connected)
+ tmp &= ~DC_HPDx_INT_POLARITY;
+ else
+ tmp |= DC_HPDx_INT_POLARITY;
+ WREG32(DC_HPD4_INT_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_5:
+ tmp = RREG32(DC_HPD5_INT_CONTROL);
+ if (connected)
+ tmp &= ~DC_HPDx_INT_POLARITY;
+ else
+ tmp |= DC_HPDx_INT_POLARITY;
+ WREG32(DC_HPD5_INT_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_6:
+ tmp = RREG32(DC_HPD6_INT_CONTROL);
+ if (connected)
+ tmp &= ~DC_HPDx_INT_POLARITY;
+ else
+ tmp |= DC_HPDx_INT_POLARITY;
+ WREG32(DC_HPD6_INT_CONTROL, tmp);
+ break;
+ default:
+ break;
+ }
+}
+
+/**
+ * dce_v6_0_hpd_init - hpd setup callback.
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Setup the hpd pins used by the card (evergreen+).
+ * Enable the pin, set the polarity, and enable the hpd interrupts.
+ */
+static void dce_v6_0_hpd_init(struct amdgpu_device *adev)
+{
+ struct drm_device *dev = adev->ddev;
+ struct drm_connector *connector;
+ u32 tmp = DC_HPDx_CONNECTION_TIMER(0x9c4) |
+ DC_HPDx_RX_INT_TIMER(0xfa) | DC_HPDx_EN;
+
+ list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+ struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
+
+ if (connector->connector_type == DRM_MODE_CONNECTOR_eDP ||
+ connector->connector_type == DRM_MODE_CONNECTOR_LVDS) {
+ /* don't try to enable hpd on eDP or LVDS avoid breaking the
+ * aux dp channel on imac and help (but not completely fix)
+ * https://bugzilla.redhat.com/show_bug.cgi?id=726143
+ * also avoid interrupt storms during dpms.
+ */
+ continue;
+ }
+ switch (amdgpu_connector->hpd.hpd) {
+ case AMDGPU_HPD_1:
+ WREG32(DC_HPD1_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_2:
+ WREG32(DC_HPD2_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_3:
+ WREG32(DC_HPD3_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_4:
+ WREG32(DC_HPD4_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_5:
+ WREG32(DC_HPD5_CONTROL, tmp);
+ break;
+ case AMDGPU_HPD_6:
+ WREG32(DC_HPD6_CONTROL, tmp);
+ break;
+ default:
+ break;
+ }
+ dce_v6_0_hpd_set_polarity(adev, amdgpu_connector->hpd.hpd);
+ amdgpu_irq_get(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd);
+ }
+
+}
+
+/**
+ * dce_v6_0_hpd_fini - hpd tear down callback.
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Tear down the hpd pins used by the card (evergreen+).
+ * Disable the hpd interrupts.
+ */
+static void dce_v6_0_hpd_fini(struct amdgpu_device *adev)
+{
+ struct drm_device *dev = adev->ddev;
+ struct drm_connector *connector;
+
+ list_for_each_entry(connector, &dev->mode_config.connector_list, head) {
+ struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
+
+ switch (amdgpu_connector->hpd.hpd) {
+ case AMDGPU_HPD_1:
+ WREG32(DC_HPD1_CONTROL, 0);
+ break;
+ case AMDGPU_HPD_2:
+ WREG32(DC_HPD2_CONTROL, 0);
+ break;
+ case AMDGPU_HPD_3:
+ WREG32(DC_HPD3_CONTROL, 0);
+ break;
+ case AMDGPU_HPD_4:
+ WREG32(DC_HPD4_CONTROL, 0);
+ break;
+ case AMDGPU_HPD_5:
+ WREG32(DC_HPD5_CONTROL, 0);
+ break;
+ case AMDGPU_HPD_6:
+ WREG32(DC_HPD6_CONTROL, 0);
+ break;
+ default:
+ break;
+ }
+ amdgpu_irq_put(adev, &adev->hpd_irq, amdgpu_connector->hpd.hpd);
+ }
+}
+
+static u32 dce_v6_0_hpd_get_gpio_reg(struct amdgpu_device *adev)
+{
+ return SI_DC_GPIO_HPD_A;
+}
+
+static bool dce_v6_0_is_display_hung(struct amdgpu_device *adev)
+{
+ DRM_INFO("xxxx: dce_v6_0_is_display_hung ----no imp!!!!!\n");
+
+ return true;
+}
+
+static u32 evergreen_get_vblank_counter(struct amdgpu_device* adev, int crtc)
+{
+ if (crtc >= adev->mode_info.num_crtc)
+ return 0;
+ else
+ return RREG32(CRTC_STATUS_FRAME_COUNT + crtc_offsets[crtc]);
+}
+
+static void dce_v6_0_stop_mc_access(struct amdgpu_device *adev,
+ struct amdgpu_mode_mc_save *save)
+{
+ u32 crtc_enabled, tmp, frame_count;
+ int i, j;
+
+ save->vga_render_control = RREG32(VGA_RENDER_CONTROL);
+ save->vga_hdp_control = RREG32(VGA_HDP_CONTROL);
+
+ /* disable VGA render */
+ WREG32(VGA_RENDER_CONTROL, 0);
+
+ /* blank the display controllers */
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ crtc_enabled = RREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[i]) & EVERGREEN_CRTC_MASTER_EN;
+ if (crtc_enabled) {
+ save->crtc_enabled[i] = true;
+ tmp = RREG32(EVERGREEN_CRTC_BLANK_CONTROL + crtc_offsets[i]);
+
+ if (!(tmp & EVERGREEN_CRTC_BLANK_DATA_EN)) {
+ dce_v6_0_vblank_wait(adev, i);
+ WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 1);
+ tmp |= EVERGREEN_CRTC_BLANK_DATA_EN;
+ WREG32(EVERGREEN_CRTC_BLANK_CONTROL + crtc_offsets[i], tmp);
+ WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 0);
+ }
+ /* wait for the next frame */
+ frame_count = evergreen_get_vblank_counter(adev, i);
+ for (j = 0; j < adev->usec_timeout; j++) {
+ if (evergreen_get_vblank_counter(adev, i) != frame_count)
+ break;
+ udelay(1);
+ }
+
+ /* XXX this is a hack to avoid strange behavior with EFI on certain systems */
+ WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 1);
+ tmp = RREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[i]);
+ tmp &= ~EVERGREEN_CRTC_MASTER_EN;
+ WREG32(EVERGREEN_CRTC_CONTROL + crtc_offsets[i], tmp);
+ WREG32(EVERGREEN_CRTC_UPDATE_LOCK + crtc_offsets[i], 0);
+ save->crtc_enabled[i] = false;
+ /* ***** */
+ } else {
+ save->crtc_enabled[i] = false;
+ }
+ }
+}
+
+static void dce_v6_0_resume_mc_access(struct amdgpu_device *adev,
+ struct amdgpu_mode_mc_save *save)
+{
+ u32 tmp;
+ int i, j;
+
+ /* update crtc base addresses */
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i],
+ upper_32_bits(adev->mc.vram_start));
+ WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i],
+ upper_32_bits(adev->mc.vram_start));
+ WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + crtc_offsets[i],
+ (u32)adev->mc.vram_start);
+ WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS + crtc_offsets[i],
+ (u32)adev->mc.vram_start);
+ }
+
+ WREG32(EVERGREEN_VGA_MEMORY_BASE_ADDRESS_HIGH, upper_32_bits(adev->mc.vram_start));
+ WREG32(EVERGREEN_VGA_MEMORY_BASE_ADDRESS, (u32)adev->mc.vram_start);
+
+ /* unlock regs and wait for update */
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ if (save->crtc_enabled[i]) {
+ tmp = RREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i]);
+ if ((tmp & 0x7) != 3) {
+ tmp &= ~0x7;
+ tmp |= 0x3;
+ WREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i], tmp);
+ }
+ tmp = RREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i]);
+ if (tmp & EVERGREEN_GRPH_UPDATE_LOCK) {
+ tmp &= ~EVERGREEN_GRPH_UPDATE_LOCK;
+ WREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i], tmp);
+ }
+ tmp = RREG32(EVERGREEN_MASTER_UPDATE_LOCK + crtc_offsets[i]);
+ if (tmp & 1) {
+ tmp &= ~1;
+ WREG32(EVERGREEN_MASTER_UPDATE_LOCK + crtc_offsets[i], tmp);
+ }
+ for (j = 0; j < adev->usec_timeout; j++) {
+ tmp = RREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i]);
+ if ((tmp & EVERGREEN_GRPH_SURFACE_UPDATE_PENDING) == 0)
+ break;
+ udelay(1);
+ }
+ }
+ }
+
+ /* Unlock vga access */
+ WREG32(VGA_HDP_CONTROL, save->vga_hdp_control);
+ mdelay(1);
+ WREG32(VGA_RENDER_CONTROL, save->vga_render_control);
+
+}
+
+static void dce_v6_0_set_vga_render_state(struct amdgpu_device *adev,
+ bool render)
+{
+ if (!render)
+ WREG32(R_000300_VGA_RENDER_CONTROL,
+ RREG32(R_000300_VGA_RENDER_CONTROL) & C_000300_VGA_VSTATUS_CNTL);
+
+}
+
+static void dce_v6_0_program_fmt(struct drm_encoder *encoder)
+{
+
+ struct drm_device *dev = encoder->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+ struct drm_connector *connector = amdgpu_get_connector_for_encoder(encoder);
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(encoder->crtc);
+ int bpc = 0;
+ u32 tmp = 0;
+ enum amdgpu_connector_dither dither = AMDGPU_FMT_DITHER_DISABLE;
+
+ if (connector) {
+ struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
+ bpc = amdgpu_connector_get_monitor_bpc(connector);
+ dither = amdgpu_connector->dither;
+ }
+
+ /* LVDS FMT is set up by atom */
+ if (amdgpu_encoder->devices & ATOM_DEVICE_LCD_SUPPORT)
+ return;
+
+ if (bpc == 0)
+ return;
+
+
+ switch (bpc) {
+ case 6:
+ if (dither == AMDGPU_FMT_DITHER_ENABLE)
+ /* XXX sort out optimal dither settings */
+ tmp |= (FMT_FRAME_RANDOM_ENABLE | FMT_HIGHPASS_RANDOM_ENABLE |
+ FMT_SPATIAL_DITHER_EN);
+ else
+ tmp |= FMT_TRUNCATE_EN;
+ break;
+ case 8:
+ if (dither == AMDGPU_FMT_DITHER_ENABLE)
+ /* XXX sort out optimal dither settings */
+ tmp |= (FMT_FRAME_RANDOM_ENABLE | FMT_HIGHPASS_RANDOM_ENABLE |
+ FMT_RGB_RANDOM_ENABLE |
+ FMT_SPATIAL_DITHER_EN | FMT_SPATIAL_DITHER_DEPTH);
+ else
+ tmp |= (FMT_TRUNCATE_EN | FMT_TRUNCATE_DEPTH);
+ break;
+ case 10:
+ default:
+ /* not needed */
+ break;
+ }
+
+ WREG32(FMT_BIT_DEPTH_CONTROL + amdgpu_crtc->crtc_offset, tmp);
+}
+
+/**
+ * cik_get_number_of_dram_channels - get the number of dram channels
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Look up the number of video ram channels (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns the number of dram channels
+ */
+static u32 si_get_number_of_dram_channels(struct amdgpu_device *adev)
+{
+ u32 tmp = RREG32(MC_SHARED_CHMAP);
+
+ switch ((tmp & MC_SHARED_CHMAP__NOOFCHAN_MASK) >> MC_SHARED_CHMAP__NOOFCHAN__SHIFT) {
+ case 0:
+ default:
+ return 1;
+ case 1:
+ return 2;
+ case 2:
+ return 4;
+ case 3:
+ return 8;
+ case 4:
+ return 3;
+ case 5:
+ return 6;
+ case 6:
+ return 10;
+ case 7:
+ return 12;
+ case 8:
+ return 16;
+ }
+}
+
+struct dce6_wm_params {
+ u32 dram_channels; /* number of dram channels */
+ u32 yclk; /* bandwidth per dram data pin in kHz */
+ u32 sclk; /* engine clock in kHz */
+ u32 disp_clk; /* display clock in kHz */
+ u32 src_width; /* viewport width */
+ u32 active_time; /* active display time in ns */
+ u32 blank_time; /* blank time in ns */
+ bool interlaced; /* mode is interlaced */
+ fixed20_12 vsc; /* vertical scale ratio */
+ u32 num_heads; /* number of active crtcs */
+ u32 bytes_per_pixel; /* bytes per pixel display + overlay */
+ u32 lb_size; /* line buffer allocated to pipe */
+ u32 vtaps; /* vertical scaler taps */
+};
+
+/**
+ * dce_v6_0_dram_bandwidth - get the dram bandwidth
+ *
+ * @wm: watermark calculation data
+ *
+ * Calculate the raw dram bandwidth (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns the dram bandwidth in MBytes/s
+ */
+static u32 dce_v6_0_dram_bandwidth(struct dce6_wm_params *wm)
+{
+ /* Calculate raw DRAM Bandwidth */
+ fixed20_12 dram_efficiency; /* 0.7 */
+ fixed20_12 yclk, dram_channels, bandwidth;
+ fixed20_12 a;
+
+ a.full = dfixed_const(1000);
+ yclk.full = dfixed_const(wm->yclk);
+ yclk.full = dfixed_div(yclk, a);
+ dram_channels.full = dfixed_const(wm->dram_channels * 4);
+ a.full = dfixed_const(10);
+ dram_efficiency.full = dfixed_const(7);
+ dram_efficiency.full = dfixed_div(dram_efficiency, a);
+ bandwidth.full = dfixed_mul(dram_channels, yclk);
+ bandwidth.full = dfixed_mul(bandwidth, dram_efficiency);
+
+ return dfixed_trunc(bandwidth);
+}
+
+/**
+ * dce_v6_0_dram_bandwidth_for_display - get the dram bandwidth for display
+ *
+ * @wm: watermark calculation data
+ *
+ * Calculate the dram bandwidth used for display (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns the dram bandwidth for display in MBytes/s
+ */
+static u32 dce_v6_0_dram_bandwidth_for_display(struct dce6_wm_params *wm)
+{
+ /* Calculate DRAM Bandwidth and the part allocated to display. */
+ fixed20_12 disp_dram_allocation; /* 0.3 to 0.7 */
+ fixed20_12 yclk, dram_channels, bandwidth;
+ fixed20_12 a;
+
+ a.full = dfixed_const(1000);
+ yclk.full = dfixed_const(wm->yclk);
+ yclk.full = dfixed_div(yclk, a);
+ dram_channels.full = dfixed_const(wm->dram_channels * 4);
+ a.full = dfixed_const(10);
+ disp_dram_allocation.full = dfixed_const(3); /* XXX worse case value 0.3 */
+ disp_dram_allocation.full = dfixed_div(disp_dram_allocation, a);
+ bandwidth.full = dfixed_mul(dram_channels, yclk);
+ bandwidth.full = dfixed_mul(bandwidth, disp_dram_allocation);
+
+ return dfixed_trunc(bandwidth);
+}
+
+/**
+ * dce_v6_0_data_return_bandwidth - get the data return bandwidth
+ *
+ * @wm: watermark calculation data
+ *
+ * Calculate the data return bandwidth used for display (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns the data return bandwidth in MBytes/s
+ */
+static u32 dce_v6_0_data_return_bandwidth(struct dce6_wm_params *wm)
+{
+ /* Calculate the display Data return Bandwidth */
+ fixed20_12 return_efficiency; /* 0.8 */
+ fixed20_12 sclk, bandwidth;
+ fixed20_12 a;
+
+ a.full = dfixed_const(1000);
+ sclk.full = dfixed_const(wm->sclk);
+ sclk.full = dfixed_div(sclk, a);
+ a.full = dfixed_const(10);
+ return_efficiency.full = dfixed_const(8);
+ return_efficiency.full = dfixed_div(return_efficiency, a);
+ a.full = dfixed_const(32);
+ bandwidth.full = dfixed_mul(a, sclk);
+ bandwidth.full = dfixed_mul(bandwidth, return_efficiency);
+
+ return dfixed_trunc(bandwidth);
+}
+
+/**
+ * dce_v6_0_dmif_request_bandwidth - get the dmif bandwidth
+ *
+ * @wm: watermark calculation data
+ *
+ * Calculate the dmif bandwidth used for display (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns the dmif bandwidth in MBytes/s
+ */
+static u32 dce_v6_0_dmif_request_bandwidth(struct dce6_wm_params *wm)
+{
+ /* Calculate the DMIF Request Bandwidth */
+ fixed20_12 disp_clk_request_efficiency; /* 0.8 */
+ fixed20_12 disp_clk, bandwidth;
+ fixed20_12 a, b;
+
+ a.full = dfixed_const(1000);
+ disp_clk.full = dfixed_const(wm->disp_clk);
+ disp_clk.full = dfixed_div(disp_clk, a);
+ a.full = dfixed_const(32);
+ b.full = dfixed_mul(a, disp_clk);
+
+ a.full = dfixed_const(10);
+ disp_clk_request_efficiency.full = dfixed_const(8);
+ disp_clk_request_efficiency.full = dfixed_div(disp_clk_request_efficiency, a);
+
+ bandwidth.full = dfixed_mul(b, disp_clk_request_efficiency);
+
+ return dfixed_trunc(bandwidth);
+}
+
+/**
+ * dce_v6_0_available_bandwidth - get the min available bandwidth
+ *
+ * @wm: watermark calculation data
+ *
+ * Calculate the min available bandwidth used for display (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns the min available bandwidth in MBytes/s
+ */
+static u32 dce_v6_0_available_bandwidth(struct dce6_wm_params *wm)
+{
+ /* Calculate the Available bandwidth. Display can use this temporarily but not in average. */
+ u32 dram_bandwidth = dce_v6_0_dram_bandwidth(wm);
+ u32 data_return_bandwidth = dce_v6_0_data_return_bandwidth(wm);
+ u32 dmif_req_bandwidth = dce_v6_0_dmif_request_bandwidth(wm);
+
+ return min(dram_bandwidth, min(data_return_bandwidth, dmif_req_bandwidth));
+}
+
+/**
+ * dce_v6_0_average_bandwidth - get the average available bandwidth
+ *
+ * @wm: watermark calculation data
+ *
+ * Calculate the average available bandwidth used for display (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns the average available bandwidth in MBytes/s
+ */
+static u32 dce_v6_0_average_bandwidth(struct dce6_wm_params *wm)
+{
+ /* Calculate the display mode Average Bandwidth
+ * DisplayMode should contain the source and destination dimensions,
+ * timing, etc.
+ */
+ fixed20_12 bpp;
+ fixed20_12 line_time;
+ fixed20_12 src_width;
+ fixed20_12 bandwidth;
+ fixed20_12 a;
+
+ a.full = dfixed_const(1000);
+ line_time.full = dfixed_const(wm->active_time + wm->blank_time);
+ line_time.full = dfixed_div(line_time, a);
+ bpp.full = dfixed_const(wm->bytes_per_pixel);
+ src_width.full = dfixed_const(wm->src_width);
+ bandwidth.full = dfixed_mul(src_width, bpp);
+ bandwidth.full = dfixed_mul(bandwidth, wm->vsc);
+ bandwidth.full = dfixed_div(bandwidth, line_time);
+
+ return dfixed_trunc(bandwidth);
+}
+
+/**
+ * dce_v6_0_latency_watermark - get the latency watermark
+ *
+ * @wm: watermark calculation data
+ *
+ * Calculate the latency watermark (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns the latency watermark in ns
+ */
+static u32 dce_v6_0_latency_watermark(struct dce6_wm_params *wm)
+{
+ /* First calculate the latency in ns */
+ u32 mc_latency = 2000; /* 2000 ns. */
+ u32 available_bandwidth = dce_v6_0_available_bandwidth(wm);
+ u32 worst_chunk_return_time = (512 * 8 * 1000) / available_bandwidth;
+ u32 cursor_line_pair_return_time = (128 * 4 * 1000) / available_bandwidth;
+ u32 dc_latency = 40000000 / wm->disp_clk; /* dc pipe latency */
+ u32 other_heads_data_return_time = ((wm->num_heads + 1) * worst_chunk_return_time) +
+ (wm->num_heads * cursor_line_pair_return_time);
+ u32 latency = mc_latency + other_heads_data_return_time + dc_latency;
+ u32 max_src_lines_per_dst_line, lb_fill_bw, line_fill_time;
+ u32 tmp, dmif_size = 12288;
+ fixed20_12 a, b, c;
+
+ if (wm->num_heads == 0)
+ return 0;
+
+ a.full = dfixed_const(2);
+ b.full = dfixed_const(1);
+ if ((wm->vsc.full > a.full) ||
+ ((wm->vsc.full > b.full) && (wm->vtaps >= 3)) ||
+ (wm->vtaps >= 5) ||
+ ((wm->vsc.full >= a.full) && wm->interlaced))
+ max_src_lines_per_dst_line = 4;
+ else
+ max_src_lines_per_dst_line = 2;
+
+ a.full = dfixed_const(available_bandwidth);
+ b.full = dfixed_const(wm->num_heads);
+ a.full = dfixed_div(a, b);
+
+ b.full = dfixed_const(mc_latency + 512);
+ c.full = dfixed_const(wm->disp_clk);
+ b.full = dfixed_div(b, c);
+
+ c.full = dfixed_const(dmif_size);
+ b.full = dfixed_div(c, b);
+
+ tmp = min(dfixed_trunc(a), dfixed_trunc(b));
+
+ b.full = dfixed_const(1000);
+ c.full = dfixed_const(wm->disp_clk);
+ b.full = dfixed_div(c, b);
+ c.full = dfixed_const(wm->bytes_per_pixel);
+ b.full = dfixed_mul(b, c);
+
+ lb_fill_bw = min(tmp, dfixed_trunc(b));
+
+ a.full = dfixed_const(max_src_lines_per_dst_line * wm->src_width * wm->bytes_per_pixel);
+ b.full = dfixed_const(1000);
+ c.full = dfixed_const(lb_fill_bw);
+ b.full = dfixed_div(c, b);
+ a.full = dfixed_div(a, b);
+ line_fill_time = dfixed_trunc(a);
+
+ if (line_fill_time < wm->active_time)
+ return latency;
+ else
+ return latency + (line_fill_time - wm->active_time);
+
+}
+
+/**
+ * dce_v6_0_average_bandwidth_vs_dram_bandwidth_for_display - check
+ * average and available dram bandwidth
+ *
+ * @wm: watermark calculation data
+ *
+ * Check if the display average bandwidth fits in the display
+ * dram bandwidth (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns true if the display fits, false if not.
+ */
+static bool dce_v6_0_average_bandwidth_vs_dram_bandwidth_for_display(struct dce6_wm_params *wm)
+{
+ if (dce_v6_0_average_bandwidth(wm) <=
+ (dce_v6_0_dram_bandwidth_for_display(wm) / wm->num_heads))
+ return true;
+ else
+ return false;
+}
+
+/**
+ * dce_v6_0_average_bandwidth_vs_available_bandwidth - check
+ * average and available bandwidth
+ *
+ * @wm: watermark calculation data
+ *
+ * Check if the display average bandwidth fits in the display
+ * available bandwidth (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns true if the display fits, false if not.
+ */
+static bool dce_v6_0_average_bandwidth_vs_available_bandwidth(struct dce6_wm_params *wm)
+{
+ if (dce_v6_0_average_bandwidth(wm) <=
+ (dce_v6_0_available_bandwidth(wm) / wm->num_heads))
+ return true;
+ else
+ return false;
+}
+
+/**
+ * dce_v6_0_check_latency_hiding - check latency hiding
+ *
+ * @wm: watermark calculation data
+ *
+ * Check latency hiding (CIK).
+ * Used for display watermark bandwidth calculations
+ * Returns true if the display fits, false if not.
+ */
+static bool dce_v6_0_check_latency_hiding(struct dce6_wm_params *wm)
+{
+ u32 lb_partitions = wm->lb_size / wm->src_width;
+ u32 line_time = wm->active_time + wm->blank_time;
+ u32 latency_tolerant_lines;
+ u32 latency_hiding;
+ fixed20_12 a;
+
+ a.full = dfixed_const(1);
+ if (wm->vsc.full > a.full)
+ latency_tolerant_lines = 1;
+ else {
+ if (lb_partitions <= (wm->vtaps + 1))
+ latency_tolerant_lines = 1;
+ else
+ latency_tolerant_lines = 2;
+ }
+
+ latency_hiding = (latency_tolerant_lines * line_time + wm->blank_time);
+
+ if (dce_v6_0_latency_watermark(wm) <= latency_hiding)
+ return true;
+ else
+ return false;
+}
+
+/**
+ * dce_v6_0_program_watermarks - program display watermarks
+ *
+ * @adev: amdgpu_device pointer
+ * @amdgpu_crtc: the selected display controller
+ * @lb_size: line buffer size
+ * @num_heads: number of display controllers in use
+ *
+ * Calculate and program the display watermarks for the
+ * selected display controller (CIK).
+ */
+static void dce_v6_0_program_watermarks(struct amdgpu_device *adev,
+ struct amdgpu_crtc *amdgpu_crtc,
+ u32 lb_size, u32 num_heads)
+{
+ struct drm_display_mode *mode = &amdgpu_crtc->base.mode;
+ struct dce6_wm_params wm_low, wm_high;
+ u32 dram_channels;
+ u32 pixel_period;
+ u32 line_time = 0;
+ u32 latency_watermark_a = 0, latency_watermark_b = 0;
+ u32 priority_a_mark = 0, priority_b_mark = 0;
+ u32 priority_a_cnt = PRIORITY_OFF;
+ u32 priority_b_cnt = PRIORITY_OFF;
+ u32 tmp, arb_control3;
+ fixed20_12 a, b, c;
+
+ if (amdgpu_crtc->base.enabled && num_heads && mode) {
+ pixel_period = 1000000 / (u32)mode->clock;
+ line_time = min((u32)mode->crtc_htotal * pixel_period, (u32)65535);
+ priority_a_cnt = 0;
+ priority_b_cnt = 0;
+
+ dram_channels = si_get_number_of_dram_channels(adev);
+
+ /* watermark for high clocks */
+ if (adev->pm.dpm_enabled) {
+ wm_high.yclk =
+ amdgpu_dpm_get_mclk(adev, false) * 10;
+ wm_high.sclk =
+ amdgpu_dpm_get_sclk(adev, false) * 10;
+ } else {
+ wm_high.yclk = adev->pm.current_mclk * 10;
+ wm_high.sclk = adev->pm.current_sclk * 10;
+ }
+
+ wm_high.disp_clk = mode->clock;
+ wm_high.src_width = mode->crtc_hdisplay;
+ wm_high.active_time = mode->crtc_hdisplay * pixel_period;
+ wm_high.blank_time = line_time - wm_high.active_time;
+ wm_high.interlaced = false;
+ if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+ wm_high.interlaced = true;
+ wm_high.vsc = amdgpu_crtc->vsc;
+ wm_high.vtaps = 1;
+ if (amdgpu_crtc->rmx_type != RMX_OFF)
+ wm_high.vtaps = 2;
+ wm_high.bytes_per_pixel = 4; /* XXX: get this from fb config */
+ wm_high.lb_size = lb_size;
+ wm_high.dram_channels = dram_channels;
+ wm_high.num_heads = num_heads;
+
+ if (adev->pm.dpm_enabled) {
+ /* watermark for low clocks */
+ wm_low.yclk =
+ amdgpu_dpm_get_mclk(adev, true) * 10;
+ wm_low.sclk =
+ amdgpu_dpm_get_sclk(adev, true) * 10;
+ } else {
+ wm_low.yclk = adev->pm.current_mclk * 10;
+ wm_low.sclk = adev->pm.current_sclk * 10;
+ }
+
+ wm_low.disp_clk = mode->clock;
+ wm_low.src_width = mode->crtc_hdisplay;
+ wm_low.active_time = mode->crtc_hdisplay * pixel_period;
+ wm_low.blank_time = line_time - wm_low.active_time;
+ wm_low.interlaced = false;
+ if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+ wm_low.interlaced = true;
+ wm_low.vsc = amdgpu_crtc->vsc;
+ wm_low.vtaps = 1;
+ if (amdgpu_crtc->rmx_type != RMX_OFF)
+ wm_low.vtaps = 2;
+ wm_low.bytes_per_pixel = 4; /* XXX: get this from fb config */
+ wm_low.lb_size = lb_size;
+ wm_low.dram_channels = dram_channels;
+ wm_low.num_heads = num_heads;
+
+ /* set for high clocks */
+ latency_watermark_a = min(dce_v6_0_latency_watermark(&wm_high), (u32)65535);
+ /* set for low clocks */
+ latency_watermark_b = min(dce_v6_0_latency_watermark(&wm_low), (u32)65535);
+
+ /* possibly force display priority to high */
+ /* should really do this at mode validation time... */
+ if (!dce_v6_0_average_bandwidth_vs_dram_bandwidth_for_display(&wm_high) ||
+ !dce_v6_0_average_bandwidth_vs_available_bandwidth(&wm_high) ||
+ !dce_v6_0_check_latency_hiding(&wm_high) ||
+ (adev->mode_info.disp_priority == 2)) {
+ DRM_DEBUG_KMS("force priority to high\n");
+ priority_a_cnt |= PRIORITY_ALWAYS_ON;
+ priority_b_cnt |= PRIORITY_ALWAYS_ON;
+ }
+ if (!dce_v6_0_average_bandwidth_vs_dram_bandwidth_for_display(&wm_low) ||
+ !dce_v6_0_average_bandwidth_vs_available_bandwidth(&wm_low) ||
+ !dce_v6_0_check_latency_hiding(&wm_low) ||
+ (adev->mode_info.disp_priority == 2)) {
+ DRM_DEBUG_KMS("force priority to high\n");
+ priority_a_cnt |= PRIORITY_ALWAYS_ON;
+ priority_b_cnt |= PRIORITY_ALWAYS_ON;
+ }
+
+ a.full = dfixed_const(1000);
+ b.full = dfixed_const(mode->clock);
+ b.full = dfixed_div(b, a);
+ c.full = dfixed_const(latency_watermark_a);
+ c.full = dfixed_mul(c, b);
+ c.full = dfixed_mul(c, amdgpu_crtc->hsc);
+ c.full = dfixed_div(c, a);
+ a.full = dfixed_const(16);
+ c.full = dfixed_div(c, a);
+ priority_a_mark = dfixed_trunc(c);
+ priority_a_cnt |= priority_a_mark & PRIORITY_MARK_MASK;
+
+ a.full = dfixed_const(1000);
+ b.full = dfixed_const(mode->clock);
+ b.full = dfixed_div(b, a);
+ c.full = dfixed_const(latency_watermark_b);
+ c.full = dfixed_mul(c, b);
+ c.full = dfixed_mul(c, amdgpu_crtc->hsc);
+ c.full = dfixed_div(c, a);
+ a.full = dfixed_const(16);
+ c.full = dfixed_div(c, a);
+ priority_b_mark = dfixed_trunc(c);
+ priority_b_cnt |= priority_b_mark & PRIORITY_MARK_MASK;
+ }
+
+ /* select wm A */
+ arb_control3 = RREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset);
+ tmp = arb_control3;
+ tmp &= ~LATENCY_WATERMARK_MASK(3);
+ tmp |= LATENCY_WATERMARK_MASK(1);
+ WREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset, tmp);
+ WREG32(DPG_PIPE_LATENCY_CONTROL + amdgpu_crtc->crtc_offset,
+ (LATENCY_LOW_WATERMARK(latency_watermark_a) |
+ LATENCY_HIGH_WATERMARK(line_time)));
+ /* select wm B */
+ tmp = RREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset);
+ tmp &= ~LATENCY_WATERMARK_MASK(3);
+ tmp |= LATENCY_WATERMARK_MASK(2);
+ WREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset, tmp);
+ WREG32(DPG_PIPE_LATENCY_CONTROL + amdgpu_crtc->crtc_offset,
+ (LATENCY_LOW_WATERMARK(latency_watermark_b) |
+ LATENCY_HIGH_WATERMARK(line_time)));
+ /* restore original selection */
+ WREG32(DPG_PIPE_ARBITRATION_CONTROL3 + amdgpu_crtc->crtc_offset, arb_control3);
+
+ /* write the priority marks */
+ WREG32(PRIORITY_A_CNT + amdgpu_crtc->crtc_offset, priority_a_cnt);
+ WREG32(PRIORITY_B_CNT + amdgpu_crtc->crtc_offset, priority_b_cnt);
+
+ /* save values for DPM */
+ amdgpu_crtc->line_time = line_time;
+ amdgpu_crtc->wm_high = latency_watermark_a;
+}
+
+/* watermark setup */
+static u32 dce_v6_0_line_buffer_adjust(struct amdgpu_device *adev,
+ struct amdgpu_crtc *amdgpu_crtc,
+ struct drm_display_mode *mode,
+ struct drm_display_mode *other_mode)
+{
+ u32 tmp, buffer_alloc, i;
+ u32 pipe_offset = amdgpu_crtc->crtc_id * 0x8;
+ /*
+ * Line Buffer Setup
+ * There are 3 line buffers, each one shared by 2 display controllers.
+ * DC_LB_MEMORY_SPLIT controls how that line buffer is shared between
+ * the display controllers. The paritioning is done via one of four
+ * preset allocations specified in bits 21:20:
+ * 0 - half lb
+ * 2 - whole lb, other crtc must be disabled
+ */
+ /* this can get tricky if we have two large displays on a paired group
+ * of crtcs. Ideally for multiple large displays we'd assign them to
+ * non-linked crtcs for maximum line buffer allocation.
+ */
+ if (amdgpu_crtc->base.enabled && mode) {
+ if (other_mode) {
+ tmp = 0; /* 1/2 */
+ buffer_alloc = 1;
+ } else {
+ tmp = 2; /* whole */
+ buffer_alloc = 2;
+ }
+ } else {
+ tmp = 0;
+ buffer_alloc = 0;
+ }
+
+ WREG32(DC_LB_MEMORY_SPLIT + amdgpu_crtc->crtc_offset,
+ DC_LB_MEMORY_CONFIG(tmp));
+
+ WREG32(PIPE0_DMIF_BUFFER_CONTROL + pipe_offset,
+ DMIF_BUFFERS_ALLOCATED(buffer_alloc));
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (RREG32(PIPE0_DMIF_BUFFER_CONTROL + pipe_offset) &
+ DMIF_BUFFERS_ALLOCATED_COMPLETED)
+ break;
+ udelay(1);
+ }
+
+ if (amdgpu_crtc->base.enabled && mode) {
+ switch (tmp) {
+ case 0:
+ default:
+ return 4096 * 2;
+ case 2:
+ return 8192 * 2;
+ }
+ }
+
+ /* controller not enabled, so no lb used */
+ return 0;
+}
+
+
+/**
+ *
+ * dce_v6_0_bandwidth_update - program display watermarks
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Calculate and program the display watermarks and line
+ * buffer allocation (CIK).
+ */
+static void dce_v6_0_bandwidth_update(struct amdgpu_device *adev)
+{
+ struct drm_display_mode *mode0 = NULL;
+ struct drm_display_mode *mode1 = NULL;
+ u32 num_heads = 0, lb_size;
+ int i;
+
+ if (!adev->mode_info.mode_config_initialized)
+ return;
+
+ amdgpu_update_display_priority(adev);
+
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ if (adev->mode_info.crtcs[i]->base.enabled)
+ num_heads++;
+ }
+ for (i = 0; i < adev->mode_info.num_crtc; i += 2) {
+ mode0 = &adev->mode_info.crtcs[i]->base.mode;
+ mode1 = &adev->mode_info.crtcs[i+1]->base.mode;
+ lb_size = dce_v6_0_line_buffer_adjust(adev, adev->mode_info.crtcs[i], mode0, mode1);
+ dce_v6_0_program_watermarks(adev, adev->mode_info.crtcs[i], lb_size, num_heads);
+ lb_size = dce_v6_0_line_buffer_adjust(adev, adev->mode_info.crtcs[i+1], mode1, mode0);
+ dce_v6_0_program_watermarks(adev, adev->mode_info.crtcs[i+1], lb_size, num_heads);
+ }
+}
+/*
+static void dce_v6_0_audio_get_connected_pins(struct amdgpu_device *adev)
+{
+ int i;
+ u32 offset, tmp;
+
+ for (i = 0; i < adev->mode_info.audio.num_pins; i++) {
+ offset = adev->mode_info.audio.pin[i].offset;
+ tmp = RREG32_AUDIO_ENDPT(offset,
+ AZ_F0_CODEC_PIN_CONTROL_RESPONSE_CONFIGURATION_DEFAULT);
+ if (((tmp & PORT_CONNECTIVITY_MASK) >> PORT_CONNECTIVITY_SHIFT) == 1)
+ adev->mode_info.audio.pin[i].connected = false;
+ else
+ adev->mode_info.audio.pin[i].connected = true;
+ }
+
+}
+
+static struct amdgpu_audio_pin *dce_v6_0_audio_get_pin(struct amdgpu_device *adev)
+{
+ int i;
+
+ dce_v6_0_audio_get_connected_pins(adev);
+
+ for (i = 0; i < adev->mode_info.audio.num_pins; i++) {
+ if (adev->mode_info.audio.pin[i].connected)
+ return &adev->mode_info.audio.pin[i];
+ }
+ DRM_ERROR("No connected audio pins found!\n");
+ return NULL;
+}
+
+static void dce_v6_0_afmt_audio_select_pin(struct drm_encoder *encoder)
+{
+ struct amdgpu_device *adev = encoder->dev->dev_private;
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+ struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
+ u32 offset;
+
+ if (!dig || !dig->afmt || !dig->afmt->pin)
+ return;
+
+ offset = dig->afmt->offset;
+
+ WREG32(AFMT_AUDIO_SRC_CONTROL + offset,
+ AFMT_AUDIO_SRC_SELECT(dig->afmt->pin->id));
+
+}
+
+static void dce_v6_0_audio_write_latency_fields(struct drm_encoder *encoder,
+ struct drm_display_mode *mode)
+{
+ DRM_INFO("xxxx: dce_v6_0_audio_write_latency_fields---no imp!!!!!\n");
+}
+
+static void dce_v6_0_audio_write_speaker_allocation(struct drm_encoder *encoder)
+{
+ DRM_INFO("xxxx: dce_v6_0_audio_write_speaker_allocation---no imp!!!!!\n");
+}
+
+static void dce_v6_0_audio_write_sad_regs(struct drm_encoder *encoder)
+{
+ DRM_INFO("xxxx: dce_v6_0_audio_write_sad_regs---no imp!!!!!\n");
+
+}
+*/
+static void dce_v6_0_audio_enable(struct amdgpu_device *adev,
+ struct amdgpu_audio_pin *pin,
+ bool enable)
+{
+ DRM_INFO("xxxx: dce_v6_0_audio_enable---no imp!!!!!\n");
+}
+
+static const u32 pin_offsets[7] =
+{
+ (0x1780 - 0x1780),
+ (0x1786 - 0x1780),
+ (0x178c - 0x1780),
+ (0x1792 - 0x1780),
+ (0x1798 - 0x1780),
+ (0x179d - 0x1780),
+ (0x17a4 - 0x1780),
+};
+
+static int dce_v6_0_audio_init(struct amdgpu_device *adev)
+{
+ return 0;
+}
+
+static void dce_v6_0_audio_fini(struct amdgpu_device *adev)
+{
+
+}
+
+/*
+static void dce_v6_0_afmt_update_ACR(struct drm_encoder *encoder, uint32_t clock)
+{
+ DRM_INFO("xxxx: dce_v6_0_afmt_update_ACR---no imp!!!!!\n");
+}
+*/
+/*
+ * build a HDMI Video Info Frame
+ */
+/*
+static void dce_v6_0_afmt_update_avi_infoframe(struct drm_encoder *encoder,
+ void *buffer, size_t size)
+{
+ DRM_INFO("xxxx: dce_v6_0_afmt_update_avi_infoframe---no imp!!!!!\n");
+}
+
+static void dce_v6_0_audio_set_dto(struct drm_encoder *encoder, u32 clock)
+{
+ DRM_INFO("xxxx: dce_v6_0_audio_set_dto---no imp!!!!!\n");
+}
+*/
+/*
+ * update the info frames with the data from the current display mode
+ */
+static void dce_v6_0_afmt_setmode(struct drm_encoder *encoder,
+ struct drm_display_mode *mode)
+{
+ DRM_INFO("xxxx: dce_v6_0_afmt_setmode ----no impl !!!!!!!!\n");
+}
+
+static void dce_v6_0_afmt_enable(struct drm_encoder *encoder, bool enable)
+{
+ struct drm_device *dev = encoder->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+ struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
+
+ if (!dig || !dig->afmt)
+ return;
+
+ /* Silent, r600_hdmi_enable will raise WARN for us */
+ if (enable && dig->afmt->enabled)
+ return;
+ if (!enable && !dig->afmt->enabled)
+ return;
+
+ if (!enable && dig->afmt->pin) {
+ dce_v6_0_audio_enable(adev, dig->afmt->pin, false);
+ dig->afmt->pin = NULL;
+ }
+
+ dig->afmt->enabled = enable;
+
+ DRM_DEBUG("%sabling AFMT interface @ 0x%04X for encoder 0x%x\n",
+ enable ? "En" : "Dis", dig->afmt->offset, amdgpu_encoder->encoder_id);
+}
+
+static void dce_v6_0_afmt_init(struct amdgpu_device *adev)
+{
+ int i;
+
+ for (i = 0; i < adev->mode_info.num_dig; i++)
+ adev->mode_info.afmt[i] = NULL;
+
+ /* DCE8 has audio blocks tied to DIG encoders */
+ for (i = 0; i < adev->mode_info.num_dig; i++) {
+ adev->mode_info.afmt[i] = kzalloc(sizeof(struct amdgpu_afmt), GFP_KERNEL);
+ if (adev->mode_info.afmt[i]) {
+ adev->mode_info.afmt[i]->offset = dig_offsets[i];
+ adev->mode_info.afmt[i]->id = i;
+ }
+ }
+}
+
+static void dce_v6_0_afmt_fini(struct amdgpu_device *adev)
+{
+ int i;
+
+ for (i = 0; i < adev->mode_info.num_dig; i++) {
+ kfree(adev->mode_info.afmt[i]);
+ adev->mode_info.afmt[i] = NULL;
+ }
+}
+
+static const u32 vga_control_regs[6] =
+{
+ AVIVO_D1VGA_CONTROL,
+ AVIVO_D2VGA_CONTROL,
+ EVERGREEN_D3VGA_CONTROL,
+ EVERGREEN_D4VGA_CONTROL,
+ EVERGREEN_D5VGA_CONTROL,
+ EVERGREEN_D6VGA_CONTROL,
+};
+
+static void dce_v6_0_vga_enable(struct drm_crtc *crtc, bool enable)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ u32 vga_control;
+
+ vga_control = RREG32(vga_control_regs[amdgpu_crtc->crtc_id]) & ~1;
+ if (enable)
+ WREG32(vga_control_regs[amdgpu_crtc->crtc_id], vga_control | 1);
+ else
+ WREG32(vga_control_regs[amdgpu_crtc->crtc_id], vga_control);
+}
+
+static void dce_v6_0_grph_enable(struct drm_crtc *crtc, bool enable)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+
+ if (enable)
+ WREG32(EVERGREEN_GRPH_ENABLE + amdgpu_crtc->crtc_offset, 1);
+ else
+ WREG32(EVERGREEN_GRPH_ENABLE + amdgpu_crtc->crtc_offset, 0);
+}
+
+static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc,
+ struct drm_framebuffer *fb,
+ int x, int y, int atomic)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ struct amdgpu_framebuffer *amdgpu_fb;
+ struct drm_framebuffer *target_fb;
+ struct drm_gem_object *obj;
+ struct amdgpu_bo *rbo;
+ uint64_t fb_location, tiling_flags;
+ uint32_t fb_format, fb_pitch_pixels, pipe_config;
+ u32 fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_NONE);
+ u32 viewport_w, viewport_h;
+ int r;
+ bool bypass_lut = false;
+
+ /* no fb bound */
+ if (!atomic && !crtc->primary->fb) {
+ DRM_DEBUG_KMS("No FB bound\n");
+ return 0;
+ }
+
+ if (atomic) {
+ amdgpu_fb = to_amdgpu_framebuffer(fb);
+ target_fb = fb;
+ }
+ else {
+ amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+ target_fb = crtc->primary->fb;
+ }
+
+ /* If atomic, assume fb object is pinned & idle & fenced and
+ * just update base pointers
+ */
+ obj = amdgpu_fb->obj;
+ rbo = gem_to_amdgpu_bo(obj);
+ r = amdgpu_bo_reserve(rbo, false);
+ if (unlikely(r != 0))
+ return r;
+
+ if (atomic)
+ fb_location = amdgpu_bo_gpu_offset(rbo);
+ else {
+ r = amdgpu_bo_pin(rbo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location);
+ if (unlikely(r != 0)) {
+ amdgpu_bo_unreserve(rbo);
+ return -EINVAL;
+ }
+ }
+
+ amdgpu_bo_get_tiling_flags(rbo, &tiling_flags);
+ amdgpu_bo_unreserve(rbo);
+
+ switch (target_fb->pixel_format) {
+ case DRM_FORMAT_C8:
+ fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_8BPP) |
+ EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_INDEXED));
+ break;
+ case DRM_FORMAT_XRGB4444:
+ case DRM_FORMAT_ARGB4444:
+ fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) |
+ EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB4444));
+#ifdef __BIG_ENDIAN
+ fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16);
+#endif
+ break;
+ case DRM_FORMAT_XRGB1555:
+ case DRM_FORMAT_ARGB1555:
+ fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) |
+ EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB1555));
+#ifdef __BIG_ENDIAN
+ fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16);
+#endif
+ break;
+ case DRM_FORMAT_BGRX5551:
+ case DRM_FORMAT_BGRA5551:
+ fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) |
+ EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_BGRA5551));
+#ifdef __BIG_ENDIAN
+ fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16);
+#endif
+ break;
+ case DRM_FORMAT_RGB565:
+ fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_16BPP) |
+ EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB565));
+#ifdef __BIG_ENDIAN
+ fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN16);
+#endif
+ break;
+ case DRM_FORMAT_XRGB8888:
+ case DRM_FORMAT_ARGB8888:
+ fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) |
+ EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB8888));
+#ifdef __BIG_ENDIAN
+ fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32);
+#endif
+ break;
+ case DRM_FORMAT_XRGB2101010:
+ case DRM_FORMAT_ARGB2101010:
+ fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) |
+ EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_ARGB2101010));
+#ifdef __BIG_ENDIAN
+ fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32);
+#endif
+ /* Greater 8 bpc fb needs to bypass hw-lut to retain precision */
+ bypass_lut = true;
+ break;
+ case DRM_FORMAT_BGRX1010102:
+ case DRM_FORMAT_BGRA1010102:
+ fb_format = (EVERGREEN_GRPH_DEPTH(EVERGREEN_GRPH_DEPTH_32BPP) |
+ EVERGREEN_GRPH_FORMAT(EVERGREEN_GRPH_FORMAT_BGRA1010102));
+#ifdef __BIG_ENDIAN
+ fb_swap = EVERGREEN_GRPH_ENDIAN_SWAP(EVERGREEN_GRPH_ENDIAN_8IN32);
+#endif
+ /* Greater 8 bpc fb needs to bypass hw-lut to retain precision */
+ bypass_lut = true;
+ break;
+ default:
+ DRM_ERROR("Unsupported screen format %s\n",
+ drm_get_format_name(target_fb->pixel_format));
+ return -EINVAL;
+ }
+
+ if (AMDGPU_TILING_GET(tiling_flags, ARRAY_MODE) == ARRAY_2D_TILED_THIN1) {
+ unsigned bankw, bankh, mtaspect, tile_split, num_banks;
+
+ bankw = AMDGPU_TILING_GET(tiling_flags, BANK_WIDTH);
+ bankh = AMDGPU_TILING_GET(tiling_flags, BANK_HEIGHT);
+ mtaspect = AMDGPU_TILING_GET(tiling_flags, MACRO_TILE_ASPECT);
+ tile_split = AMDGPU_TILING_GET(tiling_flags, TILE_SPLIT);
+ num_banks = AMDGPU_TILING_GET(tiling_flags, NUM_BANKS);
+
+ fb_format |= EVERGREEN_GRPH_NUM_BANKS(num_banks);
+ fb_format |= EVERGREEN_GRPH_ARRAY_MODE(EVERGREEN_GRPH_ARRAY_2D_TILED_THIN1);
+ fb_format |= EVERGREEN_GRPH_TILE_SPLIT(tile_split);
+ fb_format |= EVERGREEN_GRPH_BANK_WIDTH(bankw);
+ fb_format |= EVERGREEN_GRPH_BANK_HEIGHT(bankh);
+ fb_format |= EVERGREEN_GRPH_MACRO_TILE_ASPECT(mtaspect);
+ } else if (AMDGPU_TILING_GET(tiling_flags, ARRAY_MODE) == ARRAY_1D_TILED_THIN1)
+ fb_format |= EVERGREEN_GRPH_ARRAY_MODE(EVERGREEN_GRPH_ARRAY_1D_TILED_THIN1);
+
+ pipe_config = AMDGPU_TILING_GET(tiling_flags, PIPE_CONFIG);
+ fb_format |= SI_GRPH_PIPE_CONFIG(pipe_config);
+
+ dce_v6_0_vga_enable(crtc, false);
+
+ /* Make sure surface address is updated at vertical blank rather than
+ * horizontal blank
+ */
+ WREG32(EVERGREEN_GRPH_FLIP_CONTROL + amdgpu_crtc->crtc_offset, 0);
+
+ WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset,
+ upper_32_bits(fb_location));
+ WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset,
+ upper_32_bits(fb_location));
+ WREG32(EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset,
+ (u32)fb_location & EVERGREEN_GRPH_SURFACE_ADDRESS_MASK);
+ WREG32(EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset,
+ (u32) fb_location & EVERGREEN_GRPH_SURFACE_ADDRESS_MASK);
+ WREG32(EVERGREEN_GRPH_CONTROL + amdgpu_crtc->crtc_offset, fb_format);
+ WREG32(EVERGREEN_GRPH_SWAP_CONTROL + amdgpu_crtc->crtc_offset, fb_swap);
+
+ /*
+ * The LUT only has 256 slots for indexing by a 8 bpc fb. Bypass the LUT
+ * for > 8 bpc scanout to avoid truncation of fb indices to 8 msb's, to
+ * retain the full precision throughout the pipeline.
+ */
+ WREG32_P(EVERGREEN_GRPH_LUT_10BIT_BYPASS_CONTROL + amdgpu_crtc->crtc_offset,
+ (bypass_lut ? EVERGREEN_LUT_10BIT_BYPASS_EN : 0),
+ ~EVERGREEN_LUT_10BIT_BYPASS_EN);
+
+ if (bypass_lut)
+ DRM_DEBUG_KMS("Bypassing hardware LUT due to 10 bit fb scanout.\n");
+
+ WREG32(EVERGREEN_GRPH_SURFACE_OFFSET_X + amdgpu_crtc->crtc_offset, 0);
+ WREG32(EVERGREEN_GRPH_SURFACE_OFFSET_Y + amdgpu_crtc->crtc_offset, 0);
+ WREG32(EVERGREEN_GRPH_X_START + amdgpu_crtc->crtc_offset, 0);
+ WREG32(EVERGREEN_GRPH_Y_START + amdgpu_crtc->crtc_offset, 0);
+ WREG32(EVERGREEN_GRPH_X_END + amdgpu_crtc->crtc_offset, target_fb->width);
+ WREG32(EVERGREEN_GRPH_Y_END + amdgpu_crtc->crtc_offset, target_fb->height);
+
+ fb_pitch_pixels = target_fb->pitches[0] / (target_fb->bits_per_pixel / 8);
+ WREG32(EVERGREEN_GRPH_PITCH + amdgpu_crtc->crtc_offset, fb_pitch_pixels);
+
+ dce_v6_0_grph_enable(crtc, true);
+
+ WREG32(EVERGREEN_DESKTOP_HEIGHT + amdgpu_crtc->crtc_offset,
+ target_fb->height);
+ x &= ~3;
+ y &= ~1;
+ WREG32(EVERGREEN_VIEWPORT_START + amdgpu_crtc->crtc_offset,
+ (x << 16) | y);
+ viewport_w = crtc->mode.hdisplay;
+ viewport_h = (crtc->mode.vdisplay + 1) & ~1;
+
+ WREG32(EVERGREEN_VIEWPORT_SIZE + amdgpu_crtc->crtc_offset,
+ (viewport_w << 16) | viewport_h);
+
+ /* set pageflip to happen anywhere in vblank interval */
+ WREG32(EVERGREEN_MASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
+
+ if (!atomic && fb && fb != crtc->primary->fb) {
+ amdgpu_fb = to_amdgpu_framebuffer(fb);
+ rbo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+ r = amdgpu_bo_reserve(rbo, false);
+ if (unlikely(r != 0))
+ return r;
+ amdgpu_bo_unpin(rbo);
+ amdgpu_bo_unreserve(rbo);
+ }
+
+ /* Bytes per pixel may have changed */
+ dce_v6_0_bandwidth_update(adev);
+
+ return 0;
+
+}
+
+static void dce_v6_0_set_interleave(struct drm_crtc *crtc,
+ struct drm_display_mode *mode)
+{
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+
+ if (mode->flags & DRM_MODE_FLAG_INTERLACE)
+ WREG32(EVERGREEN_DATA_FORMAT + amdgpu_crtc->crtc_offset,
+ EVERGREEN_INTERLEAVE_EN);
+ else
+ WREG32(EVERGREEN_DATA_FORMAT + amdgpu_crtc->crtc_offset, 0);
+}
+
+static void dce_v6_0_crtc_load_lut(struct drm_crtc *crtc)
+{
+
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ int i;
+
+ DRM_DEBUG_KMS("%d\n", amdgpu_crtc->crtc_id);
+
+ WREG32(NI_INPUT_CSC_CONTROL + amdgpu_crtc->crtc_offset,
+ (NI_INPUT_CSC_GRPH_MODE(NI_INPUT_CSC_BYPASS) |
+ NI_INPUT_CSC_OVL_MODE(NI_INPUT_CSC_BYPASS)));
+ WREG32(NI_PRESCALE_GRPH_CONTROL + amdgpu_crtc->crtc_offset,
+ NI_GRPH_PRESCALE_BYPASS);
+ WREG32(NI_PRESCALE_OVL_CONTROL + amdgpu_crtc->crtc_offset,
+ NI_OVL_PRESCALE_BYPASS);
+ WREG32(NI_INPUT_GAMMA_CONTROL + amdgpu_crtc->crtc_offset,
+ (NI_GRPH_INPUT_GAMMA_MODE(NI_INPUT_GAMMA_USE_LUT) |
+ NI_OVL_INPUT_GAMMA_MODE(NI_INPUT_GAMMA_USE_LUT)));
+
+
+
+ WREG32(EVERGREEN_DC_LUT_CONTROL + amdgpu_crtc->crtc_offset, 0);
+
+ WREG32(EVERGREEN_DC_LUT_BLACK_OFFSET_BLUE + amdgpu_crtc->crtc_offset, 0);
+ WREG32(EVERGREEN_DC_LUT_BLACK_OFFSET_GREEN + amdgpu_crtc->crtc_offset, 0);
+ WREG32(EVERGREEN_DC_LUT_BLACK_OFFSET_RED + amdgpu_crtc->crtc_offset, 0);
+
+ WREG32(EVERGREEN_DC_LUT_WHITE_OFFSET_BLUE + amdgpu_crtc->crtc_offset, 0xffff);
+ WREG32(EVERGREEN_DC_LUT_WHITE_OFFSET_GREEN + amdgpu_crtc->crtc_offset, 0xffff);
+ WREG32(EVERGREEN_DC_LUT_WHITE_OFFSET_RED + amdgpu_crtc->crtc_offset, 0xffff);
+
+ WREG32(EVERGREEN_DC_LUT_RW_MODE + amdgpu_crtc->crtc_offset, 0);
+ WREG32(EVERGREEN_DC_LUT_WRITE_EN_MASK + amdgpu_crtc->crtc_offset, 0x00000007);
+
+ WREG32(EVERGREEN_DC_LUT_RW_INDEX + amdgpu_crtc->crtc_offset, 0);
+ for (i = 0; i < 256; i++) {
+ WREG32(EVERGREEN_DC_LUT_30_COLOR + amdgpu_crtc->crtc_offset,
+ (amdgpu_crtc->lut_r[i] << 20) |
+ (amdgpu_crtc->lut_g[i] << 10) |
+ (amdgpu_crtc->lut_b[i] << 0));
+ }
+
+ WREG32(NI_DEGAMMA_CONTROL + amdgpu_crtc->crtc_offset,
+ (NI_GRPH_DEGAMMA_MODE(NI_DEGAMMA_BYPASS) |
+ NI_OVL_DEGAMMA_MODE(NI_DEGAMMA_BYPASS) |
+ NI_ICON_DEGAMMA_MODE(NI_DEGAMMA_BYPASS) |
+ NI_CURSOR_DEGAMMA_MODE(NI_DEGAMMA_BYPASS)));
+ WREG32(NI_GAMUT_REMAP_CONTROL + amdgpu_crtc->crtc_offset,
+ (NI_GRPH_GAMUT_REMAP_MODE(NI_GAMUT_REMAP_BYPASS) |
+ NI_OVL_GAMUT_REMAP_MODE(NI_GAMUT_REMAP_BYPASS)));
+ WREG32(NI_REGAMMA_CONTROL + amdgpu_crtc->crtc_offset,
+ (NI_GRPH_REGAMMA_MODE(NI_REGAMMA_BYPASS) |
+ NI_OVL_REGAMMA_MODE(NI_REGAMMA_BYPASS)));
+ WREG32(NI_OUTPUT_CSC_CONTROL + amdgpu_crtc->crtc_offset,
+ (NI_OUTPUT_CSC_GRPH_MODE(0) |
+ NI_OUTPUT_CSC_OVL_MODE(NI_OUTPUT_CSC_BYPASS)));
+ /* XXX match this to the depth of the crtc fmt block, move to modeset? */
+ WREG32(0x1a50 + amdgpu_crtc->crtc_offset, 0);
+
+
+}
+
+static int dce_v6_0_pick_dig_encoder(struct drm_encoder *encoder)
+{
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+ struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
+
+ switch (amdgpu_encoder->encoder_id) {
+ case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
+ if (dig->linkb)
+ return 1;
+ else
+ return 0;
+ break;
+ case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
+ if (dig->linkb)
+ return 3;
+ else
+ return 2;
+ break;
+ case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
+ if (dig->linkb)
+ return 5;
+ else
+ return 4;
+ break;
+ case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
+ return 6;
+ break;
+ default:
+ DRM_ERROR("invalid encoder_id: 0x%x\n", amdgpu_encoder->encoder_id);
+ return 0;
+ }
+}
+
+/**
+ * dce_v6_0_pick_pll - Allocate a PPLL for use by the crtc.
+ *
+ * @crtc: drm crtc
+ *
+ * Returns the PPLL (Pixel PLL) to be used by the crtc. For DP monitors
+ * a single PPLL can be used for all DP crtcs/encoders. For non-DP
+ * monitors a dedicated PPLL must be used. If a particular board has
+ * an external DP PLL, return ATOM_PPLL_INVALID to skip PLL programming
+ * as there is no need to program the PLL itself. If we are not able to
+ * allocate a PLL, return ATOM_PPLL_INVALID to skip PLL programming to
+ * avoid messing up an existing monitor.
+ *
+ *
+ */
+static u32 dce_v6_0_pick_pll(struct drm_crtc *crtc)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ u32 pll_in_use;
+ int pll;
+
+ if (ENCODER_MODE_IS_DP(amdgpu_atombios_encoder_get_encoder_mode(amdgpu_crtc->encoder))) {
+ if (adev->clock.dp_extclk)
+ /* skip PPLL programming if using ext clock */
+ return ATOM_PPLL_INVALID;
+ else
+ return ATOM_PPLL0;
+ } else {
+ /* use the same PPLL for all monitors with the same clock */
+ pll = amdgpu_pll_get_shared_nondp_ppll(crtc);
+ if (pll != ATOM_PPLL_INVALID)
+ return pll;
+ }
+
+ /* PPLL1, and PPLL2 */
+ pll_in_use = amdgpu_pll_get_use_mask(crtc);
+ if (!(pll_in_use & (1 << ATOM_PPLL2)))
+ return ATOM_PPLL2;
+ if (!(pll_in_use & (1 << ATOM_PPLL1)))
+ return ATOM_PPLL1;
+ DRM_ERROR("unable to allocate a PPLL\n");
+ return ATOM_PPLL_INVALID;
+}
+
+static void dce_v6_0_lock_cursor(struct drm_crtc *crtc, bool lock)
+{
+ struct amdgpu_device *adev = crtc->dev->dev_private;
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ uint32_t cur_lock;
+
+ cur_lock = RREG32(EVERGREEN_CUR_UPDATE + amdgpu_crtc->crtc_offset);
+ if (lock)
+ cur_lock |= EVERGREEN_CURSOR_UPDATE_LOCK;
+ else
+ cur_lock &= ~EVERGREEN_CURSOR_UPDATE_LOCK;
+ WREG32(EVERGREEN_CUR_UPDATE + amdgpu_crtc->crtc_offset, cur_lock);
+}
+
+static void dce_v6_0_hide_cursor(struct drm_crtc *crtc)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct amdgpu_device *adev = crtc->dev->dev_private;
+
+ WREG32_IDX(EVERGREEN_CUR_CONTROL + amdgpu_crtc->crtc_offset,
+ EVERGREEN_CURSOR_MODE(EVERGREEN_CURSOR_24_8_PRE_MULT) |
+ EVERGREEN_CURSOR_URGENT_CONTROL(EVERGREEN_CURSOR_URGENT_1_2));
+
+
+}
+
+static void dce_v6_0_show_cursor(struct drm_crtc *crtc)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct amdgpu_device *adev = crtc->dev->dev_private;
+
+ WREG32(EVERGREEN_CUR_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset,
+ upper_32_bits(amdgpu_crtc->cursor_addr));
+ WREG32(EVERGREEN_CUR_SURFACE_ADDRESS + amdgpu_crtc->crtc_offset,
+ lower_32_bits(amdgpu_crtc->cursor_addr));
+
+ WREG32_IDX(EVERGREEN_CUR_CONTROL + amdgpu_crtc->crtc_offset,
+ EVERGREEN_CURSOR_EN |
+ EVERGREEN_CURSOR_MODE(EVERGREEN_CURSOR_24_8_PRE_MULT) |
+ EVERGREEN_CURSOR_URGENT_CONTROL(EVERGREEN_CURSOR_URGENT_1_2));
+
+}
+
+static int dce_v6_0_cursor_move_locked(struct drm_crtc *crtc,
+ int x, int y)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct amdgpu_device *adev = crtc->dev->dev_private;
+ int xorigin = 0, yorigin = 0;
+
+ int w = amdgpu_crtc->cursor_width;
+
+ /* avivo cursor are offset into the total surface */
+ x += crtc->x;
+ y += crtc->y;
+ DRM_DEBUG("x %d y %d c->x %d c->y %d\n", x, y, crtc->x, crtc->y);
+
+ if (x < 0) {
+ xorigin = min(-x, amdgpu_crtc->max_cursor_width - 1);
+ x = 0;
+ }
+ if (y < 0) {
+ yorigin = min(-y, amdgpu_crtc->max_cursor_height - 1);
+ y = 0;
+ }
+
+ WREG32(EVERGREEN_CUR_POSITION + amdgpu_crtc->crtc_offset, (x << 16) | y);
+ WREG32(EVERGREEN_CUR_HOT_SPOT + amdgpu_crtc->crtc_offset, (xorigin << 16) | yorigin);
+ WREG32(EVERGREEN_CUR_SIZE + amdgpu_crtc->crtc_offset,
+ ((w - 1) << 16) | (amdgpu_crtc->cursor_height - 1));
+
+ amdgpu_crtc->cursor_x = x;
+ amdgpu_crtc->cursor_y = y;
+ return 0;
+}
+
+static int dce_v6_0_crtc_cursor_move(struct drm_crtc *crtc,
+ int x, int y)
+{
+ int ret;
+
+ dce_v6_0_lock_cursor(crtc, true);
+ ret = dce_v6_0_cursor_move_locked(crtc, x, y);
+ dce_v6_0_lock_cursor(crtc, false);
+
+ return ret;
+}
+
+static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc,
+ struct drm_file *file_priv,
+ uint32_t handle,
+ uint32_t width,
+ uint32_t height,
+ int32_t hot_x,
+ int32_t hot_y)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_gem_object *obj;
+ struct amdgpu_bo *aobj;
+ int ret;
+
+ if (!handle) {
+ /* turn off cursor */
+ dce_v6_0_hide_cursor(crtc);
+ obj = NULL;
+ goto unpin;
+ }
+
+ if ((width > amdgpu_crtc->max_cursor_width) ||
+ (height > amdgpu_crtc->max_cursor_height)) {
+ DRM_ERROR("bad cursor width or height %d x %d\n", width, height);
+ return -EINVAL;
+ }
+
+ obj = drm_gem_object_lookup(file_priv, handle);
+ if (!obj) {
+ DRM_ERROR("Cannot find cursor object %x for crtc %d\n", handle, amdgpu_crtc->crtc_id);
+ return -ENOENT;
+ }
+
+ aobj = gem_to_amdgpu_bo(obj);
+ ret = amdgpu_bo_reserve(aobj, false);
+ if (ret != 0) {
+ drm_gem_object_unreference_unlocked(obj);
+ return ret;
+ }
+
+ ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr);
+ amdgpu_bo_unreserve(aobj);
+ if (ret) {
+ DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret);
+ drm_gem_object_unreference_unlocked(obj);
+ return ret;
+ }
+
+ amdgpu_crtc->cursor_width = width;
+ amdgpu_crtc->cursor_height = height;
+
+ dce_v6_0_lock_cursor(crtc, true);
+
+ if (hot_x != amdgpu_crtc->cursor_hot_x ||
+ hot_y != amdgpu_crtc->cursor_hot_y) {
+ int x, y;
+
+ x = amdgpu_crtc->cursor_x + amdgpu_crtc->cursor_hot_x - hot_x;
+ y = amdgpu_crtc->cursor_y + amdgpu_crtc->cursor_hot_y - hot_y;
+
+ dce_v6_0_cursor_move_locked(crtc, x, y);
+
+ amdgpu_crtc->cursor_hot_x = hot_x;
+ amdgpu_crtc->cursor_hot_y = hot_y;
+ }
+
+ dce_v6_0_show_cursor(crtc);
+ dce_v6_0_lock_cursor(crtc, false);
+
+unpin:
+ if (amdgpu_crtc->cursor_bo) {
+ struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo);
+ ret = amdgpu_bo_reserve(aobj, false);
+ if (likely(ret == 0)) {
+ amdgpu_bo_unpin(aobj);
+ amdgpu_bo_unreserve(aobj);
+ }
+ drm_gem_object_unreference_unlocked(amdgpu_crtc->cursor_bo);
+ }
+
+ amdgpu_crtc->cursor_bo = obj;
+ return 0;
+}
+
+static void dce_v6_0_cursor_reset(struct drm_crtc *crtc)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+
+ if (amdgpu_crtc->cursor_bo) {
+ dce_v6_0_lock_cursor(crtc, true);
+
+ dce_v6_0_cursor_move_locked(crtc, amdgpu_crtc->cursor_x,
+ amdgpu_crtc->cursor_y);
+
+ dce_v6_0_show_cursor(crtc);
+
+ dce_v6_0_lock_cursor(crtc, false);
+ }
+}
+
+static int dce_v6_0_crtc_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green,
+ u16 *blue, uint32_t size)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ int i;
+
+ /* userspace palettes are always correct as is */
+ for (i = 0; i < size; i++) {
+ amdgpu_crtc->lut_r[i] = red[i] >> 6;
+ amdgpu_crtc->lut_g[i] = green[i] >> 6;
+ amdgpu_crtc->lut_b[i] = blue[i] >> 6;
+ }
+ dce_v6_0_crtc_load_lut(crtc);
+
+ return 0;
+}
+
+static void dce_v6_0_crtc_destroy(struct drm_crtc *crtc)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+
+ drm_crtc_cleanup(crtc);
+ kfree(amdgpu_crtc);
+}
+
+static const struct drm_crtc_funcs dce_v6_0_crtc_funcs = {
+ .cursor_set2 = dce_v6_0_crtc_cursor_set2,
+ .cursor_move = dce_v6_0_crtc_cursor_move,
+ .gamma_set = dce_v6_0_crtc_gamma_set,
+ .set_config = amdgpu_crtc_set_config,
+ .destroy = dce_v6_0_crtc_destroy,
+ .page_flip_target = amdgpu_crtc_page_flip_target,
+};
+
+static void dce_v6_0_crtc_dpms(struct drm_crtc *crtc, int mode)
+{
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ unsigned type;
+
+ switch (mode) {
+ case DRM_MODE_DPMS_ON:
+ amdgpu_crtc->enabled = true;
+ amdgpu_atombios_crtc_enable(crtc, ATOM_ENABLE);
+ amdgpu_atombios_crtc_blank(crtc, ATOM_DISABLE);
+ /* Make sure VBLANK and PFLIP interrupts are still enabled */
+ type = amdgpu_crtc_idx_to_irq_type(adev, amdgpu_crtc->crtc_id);
+ amdgpu_irq_update(adev, &adev->crtc_irq, type);
+ amdgpu_irq_update(adev, &adev->pageflip_irq, type);
+ drm_vblank_post_modeset(dev, amdgpu_crtc->crtc_id);
+ dce_v6_0_crtc_load_lut(crtc);
+ break;
+ case DRM_MODE_DPMS_STANDBY:
+ case DRM_MODE_DPMS_SUSPEND:
+ case DRM_MODE_DPMS_OFF:
+ drm_vblank_pre_modeset(dev, amdgpu_crtc->crtc_id);
+ if (amdgpu_crtc->enabled)
+ amdgpu_atombios_crtc_blank(crtc, ATOM_ENABLE);
+ amdgpu_atombios_crtc_enable(crtc, ATOM_DISABLE);
+ amdgpu_crtc->enabled = false;
+ break;
+ }
+ /* adjust pm to dpms */
+ amdgpu_pm_compute_clocks(adev);
+}
+
+static void dce_v6_0_crtc_prepare(struct drm_crtc *crtc)
+{
+ /* disable crtc pair power gating before programming */
+ amdgpu_atombios_crtc_powergate(crtc, ATOM_DISABLE);
+ amdgpu_atombios_crtc_lock(crtc, ATOM_ENABLE);
+ dce_v6_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
+}
+
+static void dce_v6_0_crtc_commit(struct drm_crtc *crtc)
+{
+ dce_v6_0_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
+ amdgpu_atombios_crtc_lock(crtc, ATOM_DISABLE);
+}
+
+static void dce_v6_0_crtc_disable(struct drm_crtc *crtc)
+{
+
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ struct amdgpu_atom_ss ss;
+ int i;
+
+ dce_v6_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
+ if (crtc->primary->fb) {
+ int r;
+ struct amdgpu_framebuffer *amdgpu_fb;
+ struct amdgpu_bo *rbo;
+
+ amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+ rbo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+ r = amdgpu_bo_reserve(rbo, false);
+ if (unlikely(r))
+ DRM_ERROR("failed to reserve rbo before unpin\n");
+ else {
+ amdgpu_bo_unpin(rbo);
+ amdgpu_bo_unreserve(rbo);
+ }
+ }
+ /* disable the GRPH */
+ dce_v6_0_grph_enable(crtc, false);
+
+ amdgpu_atombios_crtc_powergate(crtc, ATOM_ENABLE);
+
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ if (adev->mode_info.crtcs[i] &&
+ adev->mode_info.crtcs[i]->enabled &&
+ i != amdgpu_crtc->crtc_id &&
+ amdgpu_crtc->pll_id == adev->mode_info.crtcs[i]->pll_id) {
+ /* one other crtc is using this pll don't turn
+ * off the pll
+ */
+ goto done;
+ }
+ }
+
+ switch (amdgpu_crtc->pll_id) {
+ case ATOM_PPLL1:
+ case ATOM_PPLL2:
+ /* disable the ppll */
+ amdgpu_atombios_crtc_program_pll(crtc, amdgpu_crtc->crtc_id, amdgpu_crtc->pll_id,
+ 0, 0, ATOM_DISABLE, 0, 0, 0, 0, 0, false, &ss);
+ break;
+ default:
+ break;
+ }
+done:
+ amdgpu_crtc->pll_id = ATOM_PPLL_INVALID;
+ amdgpu_crtc->adjusted_clock = 0;
+ amdgpu_crtc->encoder = NULL;
+ amdgpu_crtc->connector = NULL;
+}
+
+static int dce_v6_0_crtc_mode_set(struct drm_crtc *crtc,
+ struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode,
+ int x, int y, struct drm_framebuffer *old_fb)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+
+ if (!amdgpu_crtc->adjusted_clock)
+ return -EINVAL;
+
+ amdgpu_atombios_crtc_set_pll(crtc, adjusted_mode);
+ amdgpu_atombios_crtc_set_dtd_timing(crtc, adjusted_mode);
+ dce_v6_0_crtc_do_set_base(crtc, old_fb, x, y, 0);
+ amdgpu_atombios_crtc_overscan_setup(crtc, mode, adjusted_mode);
+ amdgpu_atombios_crtc_scaler_setup(crtc);
+ dce_v6_0_cursor_reset(crtc);
+ /* update the hw version fpr dpm */
+ amdgpu_crtc->hw_mode = *adjusted_mode;
+
+ return 0;
+}
+
+static bool dce_v6_0_crtc_mode_fixup(struct drm_crtc *crtc,
+ const struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_device *dev = crtc->dev;
+ struct drm_encoder *encoder;
+
+ /* assign the encoder to the amdgpu crtc to avoid repeated lookups later */
+ list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+ if (encoder->crtc == crtc) {
+ amdgpu_crtc->encoder = encoder;
+ amdgpu_crtc->connector = amdgpu_get_connector_for_encoder(encoder);
+ break;
+ }
+ }
+ if ((amdgpu_crtc->encoder == NULL) || (amdgpu_crtc->connector == NULL)) {
+ amdgpu_crtc->encoder = NULL;
+ amdgpu_crtc->connector = NULL;
+ return false;
+ }
+ if (!amdgpu_crtc_scaling_mode_fixup(crtc, mode, adjusted_mode))
+ return false;
+ if (amdgpu_atombios_crtc_prepare_pll(crtc, adjusted_mode))
+ return false;
+ /* pick pll */
+ amdgpu_crtc->pll_id = dce_v6_0_pick_pll(crtc);
+ /* if we can't get a PPLL for a non-DP encoder, fail */
+ if ((amdgpu_crtc->pll_id == ATOM_PPLL_INVALID) &&
+ !ENCODER_MODE_IS_DP(amdgpu_atombios_encoder_get_encoder_mode(amdgpu_crtc->encoder)))
+ return false;
+
+ return true;
+}
+
+static int dce_v6_0_crtc_set_base(struct drm_crtc *crtc, int x, int y,
+ struct drm_framebuffer *old_fb)
+{
+ return dce_v6_0_crtc_do_set_base(crtc, old_fb, x, y, 0);
+}
+
+static int dce_v6_0_crtc_set_base_atomic(struct drm_crtc *crtc,
+ struct drm_framebuffer *fb,
+ int x, int y, enum mode_set_atomic state)
+{
+ return dce_v6_0_crtc_do_set_base(crtc, fb, x, y, 1);
+}
+
+static const struct drm_crtc_helper_funcs dce_v6_0_crtc_helper_funcs = {
+ .dpms = dce_v6_0_crtc_dpms,
+ .mode_fixup = dce_v6_0_crtc_mode_fixup,
+ .mode_set = dce_v6_0_crtc_mode_set,
+ .mode_set_base = dce_v6_0_crtc_set_base,
+ .mode_set_base_atomic = dce_v6_0_crtc_set_base_atomic,
+ .prepare = dce_v6_0_crtc_prepare,
+ .commit = dce_v6_0_crtc_commit,
+ .load_lut = dce_v6_0_crtc_load_lut,
+ .disable = dce_v6_0_crtc_disable,
+};
+
+static int dce_v6_0_crtc_init(struct amdgpu_device *adev, int index)
+{
+ struct amdgpu_crtc *amdgpu_crtc;
+ int i;
+
+ amdgpu_crtc = kzalloc(sizeof(struct amdgpu_crtc) +
+ (AMDGPUFB_CONN_LIMIT * sizeof(struct drm_connector *)), GFP_KERNEL);
+ if (amdgpu_crtc == NULL)
+ return -ENOMEM;
+
+ drm_crtc_init(adev->ddev, &amdgpu_crtc->base, &dce_v6_0_crtc_funcs);
+
+ drm_mode_crtc_set_gamma_size(&amdgpu_crtc->base, 256);
+ amdgpu_crtc->crtc_id = index;
+ adev->mode_info.crtcs[index] = amdgpu_crtc;
+
+ amdgpu_crtc->max_cursor_width = CURSOR_WIDTH;
+ amdgpu_crtc->max_cursor_height = CURSOR_HEIGHT;
+ adev->ddev->mode_config.cursor_width = amdgpu_crtc->max_cursor_width;
+ adev->ddev->mode_config.cursor_height = amdgpu_crtc->max_cursor_height;
+
+ for (i = 0; i < 256; i++) {
+ amdgpu_crtc->lut_r[i] = i << 2;
+ amdgpu_crtc->lut_g[i] = i << 2;
+ amdgpu_crtc->lut_b[i] = i << 2;
+ }
+
+ amdgpu_crtc->crtc_offset = crtc_offsets[amdgpu_crtc->crtc_id];
+
+ amdgpu_crtc->pll_id = ATOM_PPLL_INVALID;
+ amdgpu_crtc->adjusted_clock = 0;
+ amdgpu_crtc->encoder = NULL;
+ amdgpu_crtc->connector = NULL;
+ drm_crtc_helper_add(&amdgpu_crtc->base, &dce_v6_0_crtc_helper_funcs);
+
+ return 0;
+}
+
+static int dce_v6_0_early_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ adev->audio_endpt_rreg = &dce_v6_0_audio_endpt_rreg;
+ adev->audio_endpt_wreg = &dce_v6_0_audio_endpt_wreg;
+
+ dce_v6_0_set_display_funcs(adev);
+ dce_v6_0_set_irq_funcs(adev);
+
+ switch (adev->asic_type) {
+ case CHIP_TAHITI:
+ case CHIP_PITCAIRN:
+ case CHIP_VERDE:
+ adev->mode_info.num_crtc = 6;
+ adev->mode_info.num_hpd = 6;
+ adev->mode_info.num_dig = 6;
+ break;
+ case CHIP_OLAND:
+ adev->mode_info.num_crtc = 2;
+ adev->mode_info.num_hpd = 2;
+ adev->mode_info.num_dig = 2;
+ break;
+ default:
+ /* FIXME: not supported yet */
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int dce_v6_0_sw_init(void *handle)
+{
+ int r, i;
+ bool ret;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ r = amdgpu_irq_add_id(adev, i + 1, &adev->crtc_irq);
+ if (r)
+ return r;
+ }
+
+ for (i = 8; i < 20; i += 2) {
+ r = amdgpu_irq_add_id(adev, i, &adev->pageflip_irq);
+ if (r)
+ return r;
+ }
+
+ /* HPD hotplug */
+ r = amdgpu_irq_add_id(adev, 42, &adev->hpd_irq);
+ if (r)
+ return r;
+
+ adev->mode_info.mode_config_initialized = true;
+
+ adev->ddev->mode_config.funcs = &amdgpu_mode_funcs;
+
+ adev->ddev->mode_config.async_page_flip = true;
+
+ adev->ddev->mode_config.max_width = 16384;
+ adev->ddev->mode_config.max_height = 16384;
+
+ adev->ddev->mode_config.preferred_depth = 24;
+ adev->ddev->mode_config.prefer_shadow = 1;
+
+ adev->ddev->mode_config.fb_base = adev->mc.aper_base;
+
+ r = amdgpu_modeset_create_props(adev);
+ if (r)
+ return r;
+
+ adev->ddev->mode_config.max_width = 16384;
+ adev->ddev->mode_config.max_height = 16384;
+
+ /* allocate crtcs */
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ r = dce_v6_0_crtc_init(adev, i);
+ if (r)
+ return r;
+ }
+
+ ret = amdgpu_atombios_get_connector_info_from_object_table(adev);
+ if (ret)
+ amdgpu_print_display_setup(adev->ddev);
+ else
+ return -EINVAL;
+
+ /* setup afmt */
+ dce_v6_0_afmt_init(adev);
+
+ r = dce_v6_0_audio_init(adev);
+ if (r)
+ return r;
+
+ drm_kms_helper_poll_init(adev->ddev);
+
+ return r;
+}
+
+static int dce_v6_0_sw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ kfree(adev->mode_info.bios_hardcoded_edid);
+
+ drm_kms_helper_poll_fini(adev->ddev);
+
+ dce_v6_0_audio_fini(adev);
+
+ dce_v6_0_afmt_fini(adev);
+
+ drm_mode_config_cleanup(adev->ddev);
+ adev->mode_info.mode_config_initialized = false;
+
+ return 0;
+}
+
+static int dce_v6_0_hw_init(void *handle)
+{
+ int i;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ /* init dig PHYs, disp eng pll */
+ amdgpu_atombios_encoder_init_dig(adev);
+ amdgpu_atombios_crtc_set_disp_eng_pll(adev, adev->clock.default_dispclk);
+
+ /* initialize hpd */
+ dce_v6_0_hpd_init(adev);
+
+ for (i = 0; i < adev->mode_info.audio.num_pins; i++) {
+ dce_v6_0_audio_enable(adev, &adev->mode_info.audio.pin[i], false);
+ }
+
+ dce_v6_0_pageflip_interrupt_init(adev);
+
+ return 0;
+}
+
+static int dce_v6_0_hw_fini(void *handle)
+{
+ int i;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ dce_v6_0_hpd_fini(adev);
+
+ for (i = 0; i < adev->mode_info.audio.num_pins; i++) {
+ dce_v6_0_audio_enable(adev, &adev->mode_info.audio.pin[i], false);
+ }
+
+ dce_v6_0_pageflip_interrupt_fini(adev);
+
+ return 0;
+}
+
+static int dce_v6_0_suspend(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ amdgpu_atombios_scratch_regs_save(adev);
+
+ return dce_v6_0_hw_fini(handle);
+}
+
+static int dce_v6_0_resume(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ int ret;
+
+ ret = dce_v6_0_hw_init(handle);
+
+ amdgpu_atombios_scratch_regs_restore(adev);
+
+ /* turn on the BL */
+ if (adev->mode_info.bl_encoder) {
+ u8 bl_level = amdgpu_display_backlight_get_level(adev,
+ adev->mode_info.bl_encoder);
+ amdgpu_display_backlight_set_level(adev, adev->mode_info.bl_encoder,
+ bl_level);
+ }
+
+ return ret;
+}
+
+static bool dce_v6_0_is_idle(void *handle)
+{
+ return true;
+}
+
+static int dce_v6_0_wait_for_idle(void *handle)
+{
+ return 0;
+}
+
+static int dce_v6_0_soft_reset(void *handle)
+{
+ DRM_INFO("xxxx: dce_v6_0_soft_reset --- no impl!!\n");
+ return 0;
+}
+
+static void dce_v6_0_set_crtc_vblank_interrupt_state(struct amdgpu_device *adev,
+ int crtc,
+ enum amdgpu_interrupt_state state)
+{
+ u32 reg_block, interrupt_mask;
+
+ if (crtc >= adev->mode_info.num_crtc) {
+ DRM_DEBUG("invalid crtc %d\n", crtc);
+ return;
+ }
+
+ switch (crtc) {
+ case 0:
+ reg_block = SI_CRTC0_REGISTER_OFFSET;
+ break;
+ case 1:
+ reg_block = SI_CRTC1_REGISTER_OFFSET;
+ break;
+ case 2:
+ reg_block = SI_CRTC2_REGISTER_OFFSET;
+ break;
+ case 3:
+ reg_block = SI_CRTC3_REGISTER_OFFSET;
+ break;
+ case 4:
+ reg_block = SI_CRTC4_REGISTER_OFFSET;
+ break;
+ case 5:
+ reg_block = SI_CRTC5_REGISTER_OFFSET;
+ break;
+ default:
+ DRM_DEBUG("invalid crtc %d\n", crtc);
+ return;
+ }
+
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ interrupt_mask = RREG32(INT_MASK + reg_block);
+ interrupt_mask &= ~VBLANK_INT_MASK;
+ WREG32(INT_MASK + reg_block, interrupt_mask);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ interrupt_mask = RREG32(INT_MASK + reg_block);
+ interrupt_mask |= VBLANK_INT_MASK;
+ WREG32(INT_MASK + reg_block, interrupt_mask);
+ break;
+ default:
+ break;
+ }
+}
+
+static void dce_v6_0_set_crtc_vline_interrupt_state(struct amdgpu_device *adev,
+ int crtc,
+ enum amdgpu_interrupt_state state)
+{
+
+}
+
+static int dce_v6_0_set_hpd_interrupt_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ u32 dc_hpd_int_cntl_reg, dc_hpd_int_cntl;
+
+ switch (type) {
+ case AMDGPU_HPD_1:
+ dc_hpd_int_cntl_reg = DC_HPD1_INT_CONTROL;
+ break;
+ case AMDGPU_HPD_2:
+ dc_hpd_int_cntl_reg = DC_HPD2_INT_CONTROL;
+ break;
+ case AMDGPU_HPD_3:
+ dc_hpd_int_cntl_reg = DC_HPD3_INT_CONTROL;
+ break;
+ case AMDGPU_HPD_4:
+ dc_hpd_int_cntl_reg = DC_HPD4_INT_CONTROL;
+ break;
+ case AMDGPU_HPD_5:
+ dc_hpd_int_cntl_reg = DC_HPD5_INT_CONTROL;
+ break;
+ case AMDGPU_HPD_6:
+ dc_hpd_int_cntl_reg = DC_HPD6_INT_CONTROL;
+ break;
+ default:
+ DRM_DEBUG("invalid hdp %d\n", type);
+ return 0;
+ }
+
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ dc_hpd_int_cntl = RREG32(dc_hpd_int_cntl_reg);
+ dc_hpd_int_cntl &= ~(DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
+ WREG32(dc_hpd_int_cntl_reg, dc_hpd_int_cntl);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ dc_hpd_int_cntl = RREG32(dc_hpd_int_cntl_reg);
+ dc_hpd_int_cntl |= (DC_HPDx_INT_EN | DC_HPDx_RX_INT_EN);
+ WREG32(dc_hpd_int_cntl_reg, dc_hpd_int_cntl);
+ break;
+ default:
+ break;
+ }
+
+ return 0;
+}
+
+static int dce_v6_0_set_crtc_interrupt_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ switch (type) {
+ case AMDGPU_CRTC_IRQ_VBLANK1:
+ dce_v6_0_set_crtc_vblank_interrupt_state(adev, 0, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VBLANK2:
+ dce_v6_0_set_crtc_vblank_interrupt_state(adev, 1, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VBLANK3:
+ dce_v6_0_set_crtc_vblank_interrupt_state(adev, 2, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VBLANK4:
+ dce_v6_0_set_crtc_vblank_interrupt_state(adev, 3, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VBLANK5:
+ dce_v6_0_set_crtc_vblank_interrupt_state(adev, 4, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VBLANK6:
+ dce_v6_0_set_crtc_vblank_interrupt_state(adev, 5, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VLINE1:
+ dce_v6_0_set_crtc_vline_interrupt_state(adev, 0, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VLINE2:
+ dce_v6_0_set_crtc_vline_interrupt_state(adev, 1, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VLINE3:
+ dce_v6_0_set_crtc_vline_interrupt_state(adev, 2, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VLINE4:
+ dce_v6_0_set_crtc_vline_interrupt_state(adev, 3, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VLINE5:
+ dce_v6_0_set_crtc_vline_interrupt_state(adev, 4, state);
+ break;
+ case AMDGPU_CRTC_IRQ_VLINE6:
+ dce_v6_0_set_crtc_vline_interrupt_state(adev, 5, state);
+ break;
+ default:
+ break;
+ }
+ return 0;
+}
+
+static int dce_v6_0_crtc_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ unsigned crtc = entry->src_id - 1;
+ uint32_t disp_int = RREG32(interrupt_status_offsets[crtc].reg);
+ unsigned irq_type = amdgpu_crtc_idx_to_irq_type(adev, crtc);
+
+ switch (entry->src_data) {
+ case 0: /* vblank */
+ if (disp_int & interrupt_status_offsets[crtc].vblank)
+ WREG32(VBLANK_STATUS + crtc_offsets[crtc], VBLANK_ACK);
+ else
+ DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+
+ if (amdgpu_irq_enabled(adev, source, irq_type)) {
+ drm_handle_vblank(adev->ddev, crtc);
+ }
+ DRM_DEBUG("IH: D%d vblank\n", crtc + 1);
+ break;
+ case 1: /* vline */
+ if (disp_int & interrupt_status_offsets[crtc].vline)
+ WREG32(VLINE_STATUS + crtc_offsets[crtc], VLINE_ACK);
+ else
+ DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
+
+ DRM_DEBUG("IH: D%d vline\n", crtc + 1);
+ break;
+ default:
+ DRM_DEBUG("Unhandled interrupt: %d %d\n", entry->src_id, entry->src_data);
+ break;
+ }
+
+ return 0;
+}
+
+static int dce_v6_0_set_pageflip_interrupt_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ u32 reg;
+
+ if (type >= adev->mode_info.num_crtc) {
+ DRM_ERROR("invalid pageflip crtc %d\n", type);
+ return -EINVAL;
+ }
+
+ reg = RREG32(GRPH_INT_CONTROL + crtc_offsets[type]);
+ if (state == AMDGPU_IRQ_STATE_DISABLE)
+ WREG32(GRPH_INT_CONTROL + crtc_offsets[type],
+ reg & ~GRPH_INTERRUPT_CONTROL__GRPH_PFLIP_INT_MASK_MASK);
+ else
+ WREG32(GRPH_INT_CONTROL + crtc_offsets[type],
+ reg | GRPH_INTERRUPT_CONTROL__GRPH_PFLIP_INT_MASK_MASK);
+
+ return 0;
+}
+
+static int dce_v6_0_pageflip_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ unsigned long flags;
+ unsigned crtc_id;
+ struct amdgpu_crtc *amdgpu_crtc;
+ struct amdgpu_flip_work *works;
+
+ crtc_id = (entry->src_id - 8) >> 1;
+ amdgpu_crtc = adev->mode_info.crtcs[crtc_id];
+
+ if (crtc_id >= adev->mode_info.num_crtc) {
+ DRM_ERROR("invalid pageflip crtc %d\n", crtc_id);
+ return -EINVAL;
+ }
+
+ if (RREG32(GRPH_INT_STATUS + crtc_offsets[crtc_id]) &
+ GRPH_INTERRUPT_STATUS__GRPH_PFLIP_INT_OCCURRED_MASK)
+ WREG32(GRPH_INT_STATUS + crtc_offsets[crtc_id],
+ GRPH_INTERRUPT_STATUS__GRPH_PFLIP_INT_CLEAR_MASK);
+
+ /* IRQ could occur when in initial stage */
+ if (amdgpu_crtc == NULL)
+ return 0;
+
+ spin_lock_irqsave(&adev->ddev->event_lock, flags);
+ works = amdgpu_crtc->pflip_works;
+ if (amdgpu_crtc->pflip_status != AMDGPU_FLIP_SUBMITTED){
+ DRM_DEBUG_DRIVER("amdgpu_crtc->pflip_status = %d != "
+ "AMDGPU_FLIP_SUBMITTED(%d)\n",
+ amdgpu_crtc->pflip_status,
+ AMDGPU_FLIP_SUBMITTED);
+ spin_unlock_irqrestore(&adev->ddev->event_lock, flags);
+ return 0;
+ }
+
+ /* page flip completed. clean up */
+ amdgpu_crtc->pflip_status = AMDGPU_FLIP_NONE;
+ amdgpu_crtc->pflip_works = NULL;
+
+ /* wakeup usersapce */
+ if (works->event)
+ drm_crtc_send_vblank_event(&amdgpu_crtc->base, works->event);
+
+ spin_unlock_irqrestore(&adev->ddev->event_lock, flags);
+
+ drm_crtc_vblank_put(&amdgpu_crtc->base);
+ schedule_work(&works->unpin_work);
+
+ return 0;
+}
+
+static int dce_v6_0_hpd_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ uint32_t disp_int, mask, int_control, tmp;
+ unsigned hpd;
+
+ if (entry->src_data > 6) {
+ DRM_DEBUG("Unhandled interrupt: %d %d\n", entry->src_id, entry->src_data);
+ return 0;
+ }
+
+ hpd = entry->src_data;
+ disp_int = RREG32(interrupt_status_offsets[hpd].reg);
+ mask = interrupt_status_offsets[hpd].hpd;
+ int_control = hpd_int_control_offsets[hpd];
+
+ if (disp_int & mask) {
+ tmp = RREG32(int_control);
+ tmp |= DC_HPD1_INT_CONTROL__DC_HPD1_INT_ACK_MASK;
+ WREG32(int_control, tmp);
+ schedule_work(&adev->hotplug_work);
+ DRM_INFO("IH: HPD%d\n", hpd + 1);
+ }
+
+ return 0;
+
+}
+
+static int dce_v6_0_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ return 0;
+}
+
+static int dce_v6_0_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ return 0;
+}
+
+const struct amd_ip_funcs dce_v6_0_ip_funcs = {
+ .name = "dce_v6_0",
+ .early_init = dce_v6_0_early_init,
+ .late_init = NULL,
+ .sw_init = dce_v6_0_sw_init,
+ .sw_fini = dce_v6_0_sw_fini,
+ .hw_init = dce_v6_0_hw_init,
+ .hw_fini = dce_v6_0_hw_fini,
+ .suspend = dce_v6_0_suspend,
+ .resume = dce_v6_0_resume,
+ .is_idle = dce_v6_0_is_idle,
+ .wait_for_idle = dce_v6_0_wait_for_idle,
+ .soft_reset = dce_v6_0_soft_reset,
+ .set_clockgating_state = dce_v6_0_set_clockgating_state,
+ .set_powergating_state = dce_v6_0_set_powergating_state,
+};
+
+static void
+dce_v6_0_encoder_mode_set(struct drm_encoder *encoder,
+ struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+
+ amdgpu_encoder->pixel_clock = adjusted_mode->clock;
+
+ /* need to call this here rather than in prepare() since we need some crtc info */
+ amdgpu_atombios_encoder_dpms(encoder, DRM_MODE_DPMS_OFF);
+
+ /* set scaler clears this on some chips */
+ dce_v6_0_set_interleave(encoder->crtc, mode);
+
+ if (amdgpu_atombios_encoder_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI) {
+ dce_v6_0_afmt_enable(encoder, true);
+ dce_v6_0_afmt_setmode(encoder, adjusted_mode);
+ }
+}
+
+static void dce_v6_0_encoder_prepare(struct drm_encoder *encoder)
+{
+
+ struct amdgpu_device *adev = encoder->dev->dev_private;
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+ struct drm_connector *connector = amdgpu_get_connector_for_encoder(encoder);
+
+ if ((amdgpu_encoder->active_device &
+ (ATOM_DEVICE_DFP_SUPPORT | ATOM_DEVICE_LCD_SUPPORT)) ||
+ (amdgpu_encoder_get_dp_bridge_encoder_id(encoder) !=
+ ENCODER_OBJECT_ID_NONE)) {
+ struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
+ if (dig) {
+ dig->dig_encoder = dce_v6_0_pick_dig_encoder(encoder);
+ if (amdgpu_encoder->active_device & ATOM_DEVICE_DFP_SUPPORT)
+ dig->afmt = adev->mode_info.afmt[dig->dig_encoder];
+ }
+ }
+
+ amdgpu_atombios_scratch_regs_lock(adev, true);
+
+ if (connector) {
+ struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
+
+ /* select the clock/data port if it uses a router */
+ if (amdgpu_connector->router.cd_valid)
+ amdgpu_i2c_router_select_cd_port(amdgpu_connector);
+
+ /* turn eDP panel on for mode set */
+ if (connector->connector_type == DRM_MODE_CONNECTOR_eDP)
+ amdgpu_atombios_encoder_set_edp_panel_power(connector,
+ ATOM_TRANSMITTER_ACTION_POWER_ON);
+ }
+
+ /* this is needed for the pll/ss setup to work correctly in some cases */
+ amdgpu_atombios_encoder_set_crtc_source(encoder);
+ /* set up the FMT blocks */
+ dce_v6_0_program_fmt(encoder);
+}
+
+static void dce_v6_0_encoder_commit(struct drm_encoder *encoder)
+{
+
+ struct drm_device *dev = encoder->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+
+ /* need to call this here as we need the crtc set up */
+ amdgpu_atombios_encoder_dpms(encoder, DRM_MODE_DPMS_ON);
+ amdgpu_atombios_scratch_regs_lock(adev, false);
+}
+
+static void dce_v6_0_encoder_disable(struct drm_encoder *encoder)
+{
+
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+ struct amdgpu_encoder_atom_dig *dig;
+
+ amdgpu_atombios_encoder_dpms(encoder, DRM_MODE_DPMS_OFF);
+
+ if (amdgpu_atombios_encoder_is_digital(encoder)) {
+ if (amdgpu_atombios_encoder_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI)
+ dce_v6_0_afmt_enable(encoder, false);
+ dig = amdgpu_encoder->enc_priv;
+ dig->dig_encoder = -1;
+ }
+ amdgpu_encoder->active_device = 0;
+}
+
+/* these are handled by the primary encoders */
+static void dce_v6_0_ext_prepare(struct drm_encoder *encoder)
+{
+
+}
+
+static void dce_v6_0_ext_commit(struct drm_encoder *encoder)
+{
+
+}
+
+static void
+dce_v6_0_ext_mode_set(struct drm_encoder *encoder,
+ struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+
+}
+
+static void dce_v6_0_ext_disable(struct drm_encoder *encoder)
+{
+
+}
+
+static void
+dce_v6_0_ext_dpms(struct drm_encoder *encoder, int mode)
+{
+
+}
+
+static bool dce_v6_0_ext_mode_fixup(struct drm_encoder *encoder,
+ const struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+ return true;
+}
+
+static const struct drm_encoder_helper_funcs dce_v6_0_ext_helper_funcs = {
+ .dpms = dce_v6_0_ext_dpms,
+ .mode_fixup = dce_v6_0_ext_mode_fixup,
+ .prepare = dce_v6_0_ext_prepare,
+ .mode_set = dce_v6_0_ext_mode_set,
+ .commit = dce_v6_0_ext_commit,
+ .disable = dce_v6_0_ext_disable,
+ /* no detect for TMDS/LVDS yet */
+};
+
+static const struct drm_encoder_helper_funcs dce_v6_0_dig_helper_funcs = {
+ .dpms = amdgpu_atombios_encoder_dpms,
+ .mode_fixup = amdgpu_atombios_encoder_mode_fixup,
+ .prepare = dce_v6_0_encoder_prepare,
+ .mode_set = dce_v6_0_encoder_mode_set,
+ .commit = dce_v6_0_encoder_commit,
+ .disable = dce_v6_0_encoder_disable,
+ .detect = amdgpu_atombios_encoder_dig_detect,
+};
+
+static const struct drm_encoder_helper_funcs dce_v6_0_dac_helper_funcs = {
+ .dpms = amdgpu_atombios_encoder_dpms,
+ .mode_fixup = amdgpu_atombios_encoder_mode_fixup,
+ .prepare = dce_v6_0_encoder_prepare,
+ .mode_set = dce_v6_0_encoder_mode_set,
+ .commit = dce_v6_0_encoder_commit,
+ .detect = amdgpu_atombios_encoder_dac_detect,
+};
+
+static void dce_v6_0_encoder_destroy(struct drm_encoder *encoder)
+{
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+ if (amdgpu_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
+ amdgpu_atombios_encoder_fini_backlight(amdgpu_encoder);
+ kfree(amdgpu_encoder->enc_priv);
+ drm_encoder_cleanup(encoder);
+ kfree(amdgpu_encoder);
+}
+
+static const struct drm_encoder_funcs dce_v6_0_encoder_funcs = {
+ .destroy = dce_v6_0_encoder_destroy,
+};
+
+static void dce_v6_0_encoder_add(struct amdgpu_device *adev,
+ uint32_t encoder_enum,
+ uint32_t supported_device,
+ u16 caps)
+{
+ struct drm_device *dev = adev->ddev;
+ struct drm_encoder *encoder;
+ struct amdgpu_encoder *amdgpu_encoder;
+
+ /* see if we already added it */
+ list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+ amdgpu_encoder = to_amdgpu_encoder(encoder);
+ if (amdgpu_encoder->encoder_enum == encoder_enum) {
+ amdgpu_encoder->devices |= supported_device;
+ return;
+ }
+
+ }
+
+ /* add a new one */
+ amdgpu_encoder = kzalloc(sizeof(struct amdgpu_encoder), GFP_KERNEL);
+ if (!amdgpu_encoder)
+ return;
+
+ encoder = &amdgpu_encoder->base;
+ switch (adev->mode_info.num_crtc) {
+ case 1:
+ encoder->possible_crtcs = 0x1;
+ break;
+ case 2:
+ default:
+ encoder->possible_crtcs = 0x3;
+ break;
+ case 4:
+ encoder->possible_crtcs = 0xf;
+ break;
+ case 6:
+ encoder->possible_crtcs = 0x3f;
+ break;
+ }
+
+ amdgpu_encoder->enc_priv = NULL;
+
+ amdgpu_encoder->encoder_enum = encoder_enum;
+ amdgpu_encoder->encoder_id = (encoder_enum & OBJECT_ID_MASK) >> OBJECT_ID_SHIFT;
+ amdgpu_encoder->devices = supported_device;
+ amdgpu_encoder->rmx_type = RMX_OFF;
+ amdgpu_encoder->underscan_type = UNDERSCAN_OFF;
+ amdgpu_encoder->is_ext_encoder = false;
+ amdgpu_encoder->caps = caps;
+
+ switch (amdgpu_encoder->encoder_id) {
+ case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC1:
+ case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC2:
+ drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs,
+ DRM_MODE_ENCODER_DAC, NULL);
+ drm_encoder_helper_add(encoder, &dce_v6_0_dac_helper_funcs);
+ break;
+ case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
+ case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
+ case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
+ case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
+ case ENCODER_OBJECT_ID_INTERNAL_UNIPHY3:
+ if (amdgpu_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) {
+ amdgpu_encoder->rmx_type = RMX_FULL;
+ drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs,
+ DRM_MODE_ENCODER_LVDS, NULL);
+ amdgpu_encoder->enc_priv = amdgpu_atombios_encoder_get_lcd_info(amdgpu_encoder);
+ } else if (amdgpu_encoder->devices & (ATOM_DEVICE_CRT_SUPPORT)) {
+ drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs,
+ DRM_MODE_ENCODER_DAC, NULL);
+ amdgpu_encoder->enc_priv = amdgpu_atombios_encoder_get_dig_info(amdgpu_encoder);
+ } else {
+ drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs,
+ DRM_MODE_ENCODER_TMDS, NULL);
+ amdgpu_encoder->enc_priv = amdgpu_atombios_encoder_get_dig_info(amdgpu_encoder);
+ }
+ drm_encoder_helper_add(encoder, &dce_v6_0_dig_helper_funcs);
+ break;
+ case ENCODER_OBJECT_ID_SI170B:
+ case ENCODER_OBJECT_ID_CH7303:
+ case ENCODER_OBJECT_ID_EXTERNAL_SDVOA:
+ case ENCODER_OBJECT_ID_EXTERNAL_SDVOB:
+ case ENCODER_OBJECT_ID_TITFP513:
+ case ENCODER_OBJECT_ID_VT1623:
+ case ENCODER_OBJECT_ID_HDMI_SI1930:
+ case ENCODER_OBJECT_ID_TRAVIS:
+ case ENCODER_OBJECT_ID_NUTMEG:
+ /* these are handled by the primary encoders */
+ amdgpu_encoder->is_ext_encoder = true;
+ if (amdgpu_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
+ drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs,
+ DRM_MODE_ENCODER_LVDS, NULL);
+ else if (amdgpu_encoder->devices & (ATOM_DEVICE_CRT_SUPPORT))
+ drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs,
+ DRM_MODE_ENCODER_DAC, NULL);
+ else
+ drm_encoder_init(dev, encoder, &dce_v6_0_encoder_funcs,
+ DRM_MODE_ENCODER_TMDS, NULL);
+ drm_encoder_helper_add(encoder, &dce_v6_0_ext_helper_funcs);
+ break;
+ }
+}
+
+static const struct amdgpu_display_funcs dce_v6_0_display_funcs = {
+ .set_vga_render_state = &dce_v6_0_set_vga_render_state,
+ .bandwidth_update = &dce_v6_0_bandwidth_update,
+ .vblank_get_counter = &dce_v6_0_vblank_get_counter,
+ .vblank_wait = &dce_v6_0_vblank_wait,
+ .is_display_hung = &dce_v6_0_is_display_hung,
+ .backlight_set_level = &amdgpu_atombios_encoder_set_backlight_level,
+ .backlight_get_level = &amdgpu_atombios_encoder_get_backlight_level,
+ .hpd_sense = &dce_v6_0_hpd_sense,
+ .hpd_set_polarity = &dce_v6_0_hpd_set_polarity,
+ .hpd_get_gpio_reg = &dce_v6_0_hpd_get_gpio_reg,
+ .page_flip = &dce_v6_0_page_flip,
+ .page_flip_get_scanoutpos = &dce_v6_0_crtc_get_scanoutpos,
+ .add_encoder = &dce_v6_0_encoder_add,
+ .add_connector = &amdgpu_connector_add,
+ .stop_mc_access = &dce_v6_0_stop_mc_access,
+ .resume_mc_access = &dce_v6_0_resume_mc_access,
+};
+
+static void dce_v6_0_set_display_funcs(struct amdgpu_device *adev)
+{
+ if (adev->mode_info.funcs == NULL)
+ adev->mode_info.funcs = &dce_v6_0_display_funcs;
+}
+
+static const struct amdgpu_irq_src_funcs dce_v6_0_crtc_irq_funcs = {
+ .set = dce_v6_0_set_crtc_interrupt_state,
+ .process = dce_v6_0_crtc_irq,
+};
+
+static const struct amdgpu_irq_src_funcs dce_v6_0_pageflip_irq_funcs = {
+ .set = dce_v6_0_set_pageflip_interrupt_state,
+ .process = dce_v6_0_pageflip_irq,
+};
+
+static const struct amdgpu_irq_src_funcs dce_v6_0_hpd_irq_funcs = {
+ .set = dce_v6_0_set_hpd_interrupt_state,
+ .process = dce_v6_0_hpd_irq,
+};
+
+static void dce_v6_0_set_irq_funcs(struct amdgpu_device *adev)
+{
+ adev->crtc_irq.num_types = AMDGPU_CRTC_IRQ_LAST;
+ adev->crtc_irq.funcs = &dce_v6_0_crtc_irq_funcs;
+
+ adev->pageflip_irq.num_types = AMDGPU_PAGEFLIP_IRQ_LAST;
+ adev->pageflip_irq.funcs = &dce_v6_0_pageflip_irq_funcs;
+
+ adev->hpd_irq.num_types = AMDGPU_HPD_LAST;
+ adev->hpd_irq.funcs = &dce_v6_0_hpd_irq_funcs;
+}
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __DCE_V6_0_H__
+#define __DCE_V6_0_H__
+
+extern const struct amd_ip_funcs dce_v6_0_ip_funcs;
+
+#endif
WREG32(mmVGA_RENDER_CONTROL, tmp);
}
+static int dce_v8_0_get_num_crtc(struct amdgpu_device *adev)
+{
+ int num_crtc = 0;
+
+ switch (adev->asic_type) {
+ case CHIP_BONAIRE:
+ case CHIP_HAWAII:
+ num_crtc = 6;
+ break;
+ case CHIP_KAVERI:
+ num_crtc = 4;
+ break;
+ case CHIP_KABINI:
+ case CHIP_MULLINS:
+ num_crtc = 2;
+ break;
+ default:
+ num_crtc = 0;
+ }
+ return num_crtc;
+}
+
+void dce_v8_0_disable_dce(struct amdgpu_device *adev)
+{
+ /*Disable VGA render and enabled crtc, if has DCE engine*/
+ if (amdgpu_atombios_has_dce_engine_info(adev)) {
+ u32 tmp;
+ int crtc_enabled, i;
+
+ dce_v8_0_set_vga_render_state(adev, false);
+
+ /*Disable crtc*/
+ for (i = 0; i < dce_v8_0_get_num_crtc(adev); i++) {
+ crtc_enabled = REG_GET_FIELD(RREG32(mmCRTC_CONTROL + crtc_offsets[i]),
+ CRTC_CONTROL, CRTC_MASTER_EN);
+ if (crtc_enabled) {
+ WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1);
+ tmp = RREG32(mmCRTC_CONTROL + crtc_offsets[i]);
+ tmp = REG_SET_FIELD(tmp, CRTC_CONTROL, CRTC_MASTER_EN, 0);
+ WREG32(mmCRTC_CONTROL + crtc_offsets[i], tmp);
+ WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0);
+ }
+ }
+ }
+}
+
static void dce_v8_0_program_fmt(struct drm_encoder *encoder)
{
struct drm_device *dev = encoder->dev;
if (sad->format == eld_reg_to_type[i][1]) {
if (sad->channels > max_channels) {
- value = (sad->channels <<
- AZALIA_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0__MAX_CHANNELS__SHIFT) |
- (sad->byte2 <<
- AZALIA_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0__DESCRIPTOR_BYTE_2__SHIFT) |
- (sad->freq <<
- AZALIA_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0__SUPPORTED_FREQUENCIES__SHIFT);
- max_channels = sad->channels;
+ value = (sad->channels <<
+ AZALIA_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0__MAX_CHANNELS__SHIFT) |
+ (sad->byte2 <<
+ AZALIA_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0__DESCRIPTOR_BYTE_2__SHIFT) |
+ (sad->freq <<
+ AZALIA_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0__SUPPORTED_FREQUENCIES__SHIFT);
+ max_channels = sad->channels;
}
if (sad->format == HDMI_AUDIO_CODING_TYPE_PCM)
struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
uint32_t offset = dig->afmt->offset;
- WREG32(mmHDMI_ACR_32_0 + offset, (acr.cts_32khz << HDMI_ACR_44_0__HDMI_ACR_CTS_44__SHIFT));
+ WREG32(mmHDMI_ACR_32_0 + offset, (acr.cts_32khz << HDMI_ACR_32_0__HDMI_ACR_CTS_32__SHIFT));
WREG32(mmHDMI_ACR_32_1 + offset, acr.n_32khz);
WREG32(mmHDMI_ACR_44_0 + offset, (acr.cts_44_1khz << HDMI_ACR_44_0__HDMI_ACR_CTS_44__SHIFT));
/* Silent, r600_hdmi_enable will raise WARN for us */
if (!dig->afmt->enabled)
return;
+
offset = dig->afmt->offset;
/* hdmi deep color mode general control packets setup, if bpc > 8 */
WREG32_OR(mmHDMI_INFOFRAME_CONTROL0 + offset,
HDMI_INFOFRAME_CONTROL0__HDMI_AVI_INFO_SEND_MASK | /* enable AVI info frames */
- HDMI_INFOFRAME_CONTROL0__HDMI_AVI_INFO_SEND_MASK); /* required for audio info values to be updated */
+ HDMI_INFOFRAME_CONTROL0__HDMI_AVI_INFO_CONT_MASK); /* required for audio info values to be updated */
WREG32_P(mmHDMI_INFOFRAME_CONTROL1 + offset,
(2 << HDMI_INFOFRAME_CONTROL1__HDMI_AVI_INFO_LINE__SHIFT), /* anything other than 0 */
WREG32_OR(mmAFMT_AUDIO_PACKET_CONTROL + offset,
AFMT_AUDIO_PACKET_CONTROL__AFMT_AUDIO_SAMPLE_SEND_MASK); /* send audio packets */
- /* it's unknown what these bits do excatly, but it's indeed quite useful for debugging */
WREG32(mmAFMT_RAMP_CONTROL0 + offset, 0x00FFFFFF);
WREG32(mmAFMT_RAMP_CONTROL1 + offset, 0x007FFFFF);
WREG32(mmAFMT_RAMP_CONTROL2 + offset, 0x00000001);
WREG32(mmAFMT_RAMP_CONTROL3 + offset, 0x00000001);
- /* enable audio after to setting up hw */
+ /* enable audio after setting up hw */
dce_v8_0_audio_enable(adev, dig->afmt->pin, true);
}
case DRM_FORMAT_XRGB4444:
case DRM_FORMAT_ARGB4444:
fb_format = ((GRPH_DEPTH_16BPP << GRPH_CONTROL__GRPH_DEPTH__SHIFT) |
- (GRPH_FORMAT_ARGB1555 << GRPH_CONTROL__GRPH_FORMAT__SHIFT));
+ (GRPH_FORMAT_ARGB4444 << GRPH_CONTROL__GRPH_FORMAT__SHIFT));
#ifdef __BIG_ENDIAN
fb_swap = (GRPH_ENDIAN_8IN16 << GRPH_SWAP_CNTL__GRPH_ENDIAN_SWAP__SHIFT);
#endif
WREG32(mmVIEWPORT_SIZE + amdgpu_crtc->crtc_offset,
(viewport_w << 16) | viewport_h);
- /* set pageflip to happen only at start of vblank interval (front porch) */
- WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 3);
+ /* set pageflip to happen anywhere in vblank interval */
+ WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
if (!atomic && fb && fb != crtc->primary->fb) {
amdgpu_fb = to_amdgpu_framebuffer(fb);
.gamma_set = dce_v8_0_crtc_gamma_set,
.set_config = amdgpu_crtc_set_config,
.destroy = dce_v8_0_crtc_destroy,
- .page_flip = amdgpu_crtc_page_flip,
+ .page_flip_target = amdgpu_crtc_page_flip_target,
};
static void dce_v8_0_crtc_dpms(struct drm_crtc *crtc, int mode)
case ATOM_PPLL2:
/* disable the ppll */
amdgpu_atombios_crtc_program_pll(crtc, amdgpu_crtc->crtc_id, amdgpu_crtc->pll_id,
- 0, 0, ATOM_DISABLE, 0, 0, 0, 0, 0, false, &ss);
+ 0, 0, ATOM_DISABLE, 0, 0, 0, 0, 0, false, &ss);
break;
case ATOM_PPLL0:
/* disable the ppll */
dce_v8_0_set_display_funcs(adev);
dce_v8_0_set_irq_funcs(adev);
+ adev->mode_info.num_crtc = dce_v8_0_get_num_crtc(adev);
+
switch (adev->asic_type) {
case CHIP_BONAIRE:
case CHIP_HAWAII:
- adev->mode_info.num_crtc = 6;
adev->mode_info.num_hpd = 6;
adev->mode_info.num_dig = 6;
break;
case CHIP_KAVERI:
- adev->mode_info.num_crtc = 4;
adev->mode_info.num_hpd = 6;
adev->mode_info.num_dig = 7;
break;
case CHIP_KABINI:
case CHIP_MULLINS:
- adev->mode_info.num_crtc = 2;
adev->mode_info.num_hpd = 6;
adev->mode_info.num_dig = 6; /* ? */
break;
drm_handle_vblank(adev->ddev, crtc);
}
DRM_DEBUG("IH: D%d vblank\n", crtc + 1);
-
break;
case 1: /* vline */
if (disp_int & interrupt_status_offsets[crtc].vline)
DRM_DEBUG("IH: IH event w/o asserted irq bit?\n");
DRM_DEBUG("IH: D%d vline\n", crtc + 1);
-
break;
default:
DRM_DEBUG("Unhandled interrupt: %d %d\n", entry->src_id, entry->src_data);
extern const struct amd_ip_funcs dce_v8_0_ip_funcs;
+void dce_v8_0_disable_dce(struct amdgpu_device *adev);
+
#endif
--- /dev/null
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "drmP.h"
+#include "amdgpu.h"
+#include "amdgpu_pm.h"
+#include "amdgpu_i2c.h"
+#include "atom.h"
+#include "amdgpu_pll.h"
+#include "amdgpu_connectors.h"
+#ifdef CONFIG_DRM_AMDGPU_CIK
+#include "dce_v8_0.h"
+#endif
+#include "dce_v10_0.h"
+#include "dce_v11_0.h"
+#include "dce_virtual.h"
+
+static void dce_virtual_set_display_funcs(struct amdgpu_device *adev);
+static void dce_virtual_set_irq_funcs(struct amdgpu_device *adev);
+static int dce_virtual_pageflip_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry);
+
+/**
+ * dce_virtual_vblank_wait - vblank wait asic callback.
+ *
+ * @adev: amdgpu_device pointer
+ * @crtc: crtc to wait for vblank on
+ *
+ * Wait for vblank on the requested crtc (evergreen+).
+ */
+static void dce_virtual_vblank_wait(struct amdgpu_device *adev, int crtc)
+{
+ return;
+}
+
+static u32 dce_virtual_vblank_get_counter(struct amdgpu_device *adev, int crtc)
+{
+ return 0;
+}
+
+static void dce_virtual_page_flip(struct amdgpu_device *adev,
+ int crtc_id, u64 crtc_base, bool async)
+{
+ return;
+}
+
+static int dce_virtual_crtc_get_scanoutpos(struct amdgpu_device *adev, int crtc,
+ u32 *vbl, u32 *position)
+{
+ *vbl = 0;
+ *position = 0;
+
+ return -EINVAL;
+}
+
+static bool dce_virtual_hpd_sense(struct amdgpu_device *adev,
+ enum amdgpu_hpd_id hpd)
+{
+ return true;
+}
+
+static void dce_virtual_hpd_set_polarity(struct amdgpu_device *adev,
+ enum amdgpu_hpd_id hpd)
+{
+ return;
+}
+
+static u32 dce_virtual_hpd_get_gpio_reg(struct amdgpu_device *adev)
+{
+ return 0;
+}
+
+static bool dce_virtual_is_display_hung(struct amdgpu_device *adev)
+{
+ return false;
+}
+
+void dce_virtual_stop_mc_access(struct amdgpu_device *adev,
+ struct amdgpu_mode_mc_save *save)
+{
+ switch (adev->asic_type) {
+ case CHIP_BONAIRE:
+ case CHIP_HAWAII:
+ case CHIP_KAVERI:
+ case CHIP_KABINI:
+ case CHIP_MULLINS:
+#ifdef CONFIG_DRM_AMDGPU_CIK
+ dce_v8_0_disable_dce(adev);
+#endif
+ break;
+ case CHIP_FIJI:
+ case CHIP_TONGA:
+ dce_v10_0_disable_dce(adev);
+ break;
+ case CHIP_CARRIZO:
+ case CHIP_STONEY:
+ case CHIP_POLARIS11:
+ case CHIP_POLARIS10:
+ dce_v11_0_disable_dce(adev);
+ break;
+ case CHIP_TOPAZ:
+ /* no DCE */
+ return;
+ default:
+ DRM_ERROR("Virtual display unsupported ASIC type: 0x%X\n", adev->asic_type);
+ }
+
+ return;
+}
+void dce_virtual_resume_mc_access(struct amdgpu_device *adev,
+ struct amdgpu_mode_mc_save *save)
+{
+ return;
+}
+
+void dce_virtual_set_vga_render_state(struct amdgpu_device *adev,
+ bool render)
+{
+ return;
+}
+
+/**
+ * dce_virtual_bandwidth_update - program display watermarks
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Calculate and program the display watermarks and line
+ * buffer allocation (CIK).
+ */
+static void dce_virtual_bandwidth_update(struct amdgpu_device *adev)
+{
+ return;
+}
+
+static int dce_virtual_crtc_gamma_set(struct drm_crtc *crtc, u16 *red,
+ u16 *green, u16 *blue, uint32_t size)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ int i;
+
+ /* userspace palettes are always correct as is */
+ for (i = 0; i < size; i++) {
+ amdgpu_crtc->lut_r[i] = red[i] >> 6;
+ amdgpu_crtc->lut_g[i] = green[i] >> 6;
+ amdgpu_crtc->lut_b[i] = blue[i] >> 6;
+ }
+
+ return 0;
+}
+
+static void dce_virtual_crtc_destroy(struct drm_crtc *crtc)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+
+ drm_crtc_cleanup(crtc);
+ kfree(amdgpu_crtc);
+}
+
+static const struct drm_crtc_funcs dce_virtual_crtc_funcs = {
+ .cursor_set2 = NULL,
+ .cursor_move = NULL,
+ .gamma_set = dce_virtual_crtc_gamma_set,
+ .set_config = amdgpu_crtc_set_config,
+ .destroy = dce_virtual_crtc_destroy,
+ .page_flip_target = amdgpu_crtc_page_flip_target,
+};
+
+static void dce_virtual_crtc_dpms(struct drm_crtc *crtc, int mode)
+{
+ struct drm_device *dev = crtc->dev;
+ struct amdgpu_device *adev = dev->dev_private;
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ unsigned type;
+
+ switch (mode) {
+ case DRM_MODE_DPMS_ON:
+ amdgpu_crtc->enabled = true;
+ /* Make sure VBLANK and PFLIP interrupts are still enabled */
+ type = amdgpu_crtc_idx_to_irq_type(adev, amdgpu_crtc->crtc_id);
+ amdgpu_irq_update(adev, &adev->crtc_irq, type);
+ amdgpu_irq_update(adev, &adev->pageflip_irq, type);
+ drm_vblank_on(dev, amdgpu_crtc->crtc_id);
+ break;
+ case DRM_MODE_DPMS_STANDBY:
+ case DRM_MODE_DPMS_SUSPEND:
+ case DRM_MODE_DPMS_OFF:
+ drm_vblank_off(dev, amdgpu_crtc->crtc_id);
+ amdgpu_crtc->enabled = false;
+ break;
+ }
+}
+
+
+static void dce_virtual_crtc_prepare(struct drm_crtc *crtc)
+{
+ dce_virtual_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
+}
+
+static void dce_virtual_crtc_commit(struct drm_crtc *crtc)
+{
+ dce_virtual_crtc_dpms(crtc, DRM_MODE_DPMS_ON);
+}
+
+static void dce_virtual_crtc_disable(struct drm_crtc *crtc)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+
+ dce_virtual_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
+ if (crtc->primary->fb) {
+ int r;
+ struct amdgpu_framebuffer *amdgpu_fb;
+ struct amdgpu_bo *rbo;
+
+ amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+ rbo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+ r = amdgpu_bo_reserve(rbo, false);
+ if (unlikely(r))
+ DRM_ERROR("failed to reserve rbo before unpin\n");
+ else {
+ amdgpu_bo_unpin(rbo);
+ amdgpu_bo_unreserve(rbo);
+ }
+ }
+
+ amdgpu_crtc->pll_id = ATOM_PPLL_INVALID;
+ amdgpu_crtc->encoder = NULL;
+ amdgpu_crtc->connector = NULL;
+}
+
+static int dce_virtual_crtc_mode_set(struct drm_crtc *crtc,
+ struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode,
+ int x, int y, struct drm_framebuffer *old_fb)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+
+ /* update the hw version fpr dpm */
+ amdgpu_crtc->hw_mode = *adjusted_mode;
+
+ return 0;
+}
+
+static bool dce_virtual_crtc_mode_fixup(struct drm_crtc *crtc,
+ const struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+ struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
+ struct drm_device *dev = crtc->dev;
+ struct drm_encoder *encoder;
+
+ /* assign the encoder to the amdgpu crtc to avoid repeated lookups later */
+ list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+ if (encoder->crtc == crtc) {
+ amdgpu_crtc->encoder = encoder;
+ amdgpu_crtc->connector = amdgpu_get_connector_for_encoder(encoder);
+ break;
+ }
+ }
+ if ((amdgpu_crtc->encoder == NULL) || (amdgpu_crtc->connector == NULL)) {
+ amdgpu_crtc->encoder = NULL;
+ amdgpu_crtc->connector = NULL;
+ return false;
+ }
+
+ return true;
+}
+
+
+static int dce_virtual_crtc_set_base(struct drm_crtc *crtc, int x, int y,
+ struct drm_framebuffer *old_fb)
+{
+ return 0;
+}
+
+static void dce_virtual_crtc_load_lut(struct drm_crtc *crtc)
+{
+ return;
+}
+
+static int dce_virtual_crtc_set_base_atomic(struct drm_crtc *crtc,
+ struct drm_framebuffer *fb,
+ int x, int y, enum mode_set_atomic state)
+{
+ return 0;
+}
+
+static const struct drm_crtc_helper_funcs dce_virtual_crtc_helper_funcs = {
+ .dpms = dce_virtual_crtc_dpms,
+ .mode_fixup = dce_virtual_crtc_mode_fixup,
+ .mode_set = dce_virtual_crtc_mode_set,
+ .mode_set_base = dce_virtual_crtc_set_base,
+ .mode_set_base_atomic = dce_virtual_crtc_set_base_atomic,
+ .prepare = dce_virtual_crtc_prepare,
+ .commit = dce_virtual_crtc_commit,
+ .load_lut = dce_virtual_crtc_load_lut,
+ .disable = dce_virtual_crtc_disable,
+};
+
+static int dce_virtual_crtc_init(struct amdgpu_device *adev, int index)
+{
+ struct amdgpu_crtc *amdgpu_crtc;
+ int i;
+
+ amdgpu_crtc = kzalloc(sizeof(struct amdgpu_crtc) +
+ (AMDGPUFB_CONN_LIMIT * sizeof(struct drm_connector *)), GFP_KERNEL);
+ if (amdgpu_crtc == NULL)
+ return -ENOMEM;
+
+ drm_crtc_init(adev->ddev, &amdgpu_crtc->base, &dce_virtual_crtc_funcs);
+
+ drm_mode_crtc_set_gamma_size(&amdgpu_crtc->base, 256);
+ amdgpu_crtc->crtc_id = index;
+ adev->mode_info.crtcs[index] = amdgpu_crtc;
+
+ for (i = 0; i < 256; i++) {
+ amdgpu_crtc->lut_r[i] = i << 2;
+ amdgpu_crtc->lut_g[i] = i << 2;
+ amdgpu_crtc->lut_b[i] = i << 2;
+ }
+
+ amdgpu_crtc->pll_id = ATOM_PPLL_INVALID;
+ amdgpu_crtc->encoder = NULL;
+ amdgpu_crtc->connector = NULL;
+ drm_crtc_helper_add(&amdgpu_crtc->base, &dce_virtual_crtc_helper_funcs);
+
+ return 0;
+}
+
+static int dce_virtual_early_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ adev->mode_info.vsync_timer_enabled = AMDGPU_IRQ_STATE_DISABLE;
+ dce_virtual_set_display_funcs(adev);
+ dce_virtual_set_irq_funcs(adev);
+
+ adev->mode_info.num_crtc = 1;
+ adev->mode_info.num_hpd = 1;
+ adev->mode_info.num_dig = 1;
+ return 0;
+}
+
+static bool dce_virtual_get_connector_info(struct amdgpu_device *adev)
+{
+ struct amdgpu_i2c_bus_rec ddc_bus;
+ struct amdgpu_router router;
+ struct amdgpu_hpd hpd;
+
+ /* look up gpio for ddc, hpd */
+ ddc_bus.valid = false;
+ hpd.hpd = AMDGPU_HPD_NONE;
+ /* needed for aux chan transactions */
+ ddc_bus.hpd = hpd.hpd;
+
+ memset(&router, 0, sizeof(router));
+ router.ddc_valid = false;
+ router.cd_valid = false;
+ amdgpu_display_add_connector(adev,
+ 0,
+ ATOM_DEVICE_CRT1_SUPPORT,
+ DRM_MODE_CONNECTOR_VIRTUAL, &ddc_bus,
+ CONNECTOR_OBJECT_ID_VIRTUAL,
+ &hpd,
+ &router);
+
+ amdgpu_display_add_encoder(adev, ENCODER_VIRTUAL_ENUM_VIRTUAL,
+ ATOM_DEVICE_CRT1_SUPPORT,
+ 0);
+
+ amdgpu_link_encoder_connector(adev->ddev);
+
+ return true;
+}
+
+static int dce_virtual_sw_init(void *handle)
+{
+ int r, i;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ r = amdgpu_irq_add_id(adev, 229, &adev->crtc_irq);
+ if (r)
+ return r;
+
+ adev->ddev->max_vblank_count = 0;
+
+ adev->ddev->mode_config.funcs = &amdgpu_mode_funcs;
+
+ adev->ddev->mode_config.max_width = 16384;
+ adev->ddev->mode_config.max_height = 16384;
+
+ adev->ddev->mode_config.preferred_depth = 24;
+ adev->ddev->mode_config.prefer_shadow = 1;
+
+ adev->ddev->mode_config.fb_base = adev->mc.aper_base;
+
+ r = amdgpu_modeset_create_props(adev);
+ if (r)
+ return r;
+
+ adev->ddev->mode_config.max_width = 16384;
+ adev->ddev->mode_config.max_height = 16384;
+
+ /* allocate crtcs */
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ r = dce_virtual_crtc_init(adev, i);
+ if (r)
+ return r;
+ }
+
+ dce_virtual_get_connector_info(adev);
+ amdgpu_print_display_setup(adev->ddev);
+
+ drm_kms_helper_poll_init(adev->ddev);
+
+ adev->mode_info.mode_config_initialized = true;
+ return 0;
+}
+
+static int dce_virtual_sw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ kfree(adev->mode_info.bios_hardcoded_edid);
+
+ drm_kms_helper_poll_fini(adev->ddev);
+
+ drm_mode_config_cleanup(adev->ddev);
+ adev->mode_info.mode_config_initialized = false;
+ return 0;
+}
+
+static int dce_virtual_hw_init(void *handle)
+{
+ return 0;
+}
+
+static int dce_virtual_hw_fini(void *handle)
+{
+ return 0;
+}
+
+static int dce_virtual_suspend(void *handle)
+{
+ return dce_virtual_hw_fini(handle);
+}
+
+static int dce_virtual_resume(void *handle)
+{
+ return dce_virtual_hw_init(handle);
+}
+
+static bool dce_virtual_is_idle(void *handle)
+{
+ return true;
+}
+
+static int dce_virtual_wait_for_idle(void *handle)
+{
+ return 0;
+}
+
+static int dce_virtual_soft_reset(void *handle)
+{
+ return 0;
+}
+
+static int dce_virtual_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ return 0;
+}
+
+static int dce_virtual_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ return 0;
+}
+
+const struct amd_ip_funcs dce_virtual_ip_funcs = {
+ .name = "dce_virtual",
+ .early_init = dce_virtual_early_init,
+ .late_init = NULL,
+ .sw_init = dce_virtual_sw_init,
+ .sw_fini = dce_virtual_sw_fini,
+ .hw_init = dce_virtual_hw_init,
+ .hw_fini = dce_virtual_hw_fini,
+ .suspend = dce_virtual_suspend,
+ .resume = dce_virtual_resume,
+ .is_idle = dce_virtual_is_idle,
+ .wait_for_idle = dce_virtual_wait_for_idle,
+ .soft_reset = dce_virtual_soft_reset,
+ .set_clockgating_state = dce_virtual_set_clockgating_state,
+ .set_powergating_state = dce_virtual_set_powergating_state,
+};
+
+/* these are handled by the primary encoders */
+static void dce_virtual_encoder_prepare(struct drm_encoder *encoder)
+{
+ return;
+}
+
+static void dce_virtual_encoder_commit(struct drm_encoder *encoder)
+{
+ return;
+}
+
+static void
+dce_virtual_encoder_mode_set(struct drm_encoder *encoder,
+ struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+ return;
+}
+
+static void dce_virtual_encoder_disable(struct drm_encoder *encoder)
+{
+ return;
+}
+
+static void
+dce_virtual_encoder_dpms(struct drm_encoder *encoder, int mode)
+{
+ return;
+}
+
+static bool dce_virtual_encoder_mode_fixup(struct drm_encoder *encoder,
+ const struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+
+ /* set the active encoder to connector routing */
+ amdgpu_encoder_set_active_device(encoder);
+
+ return true;
+}
+
+static const struct drm_encoder_helper_funcs dce_virtual_encoder_helper_funcs = {
+ .dpms = dce_virtual_encoder_dpms,
+ .mode_fixup = dce_virtual_encoder_mode_fixup,
+ .prepare = dce_virtual_encoder_prepare,
+ .mode_set = dce_virtual_encoder_mode_set,
+ .commit = dce_virtual_encoder_commit,
+ .disable = dce_virtual_encoder_disable,
+};
+
+static void dce_virtual_encoder_destroy(struct drm_encoder *encoder)
+{
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+
+ kfree(amdgpu_encoder->enc_priv);
+ drm_encoder_cleanup(encoder);
+ kfree(amdgpu_encoder);
+}
+
+static const struct drm_encoder_funcs dce_virtual_encoder_funcs = {
+ .destroy = dce_virtual_encoder_destroy,
+};
+
+static void dce_virtual_encoder_add(struct amdgpu_device *adev,
+ uint32_t encoder_enum,
+ uint32_t supported_device,
+ u16 caps)
+{
+ struct drm_device *dev = adev->ddev;
+ struct drm_encoder *encoder;
+ struct amdgpu_encoder *amdgpu_encoder;
+
+ /* see if we already added it */
+ list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+ amdgpu_encoder = to_amdgpu_encoder(encoder);
+ if (amdgpu_encoder->encoder_enum == encoder_enum) {
+ amdgpu_encoder->devices |= supported_device;
+ return;
+ }
+
+ }
+
+ /* add a new one */
+ amdgpu_encoder = kzalloc(sizeof(struct amdgpu_encoder), GFP_KERNEL);
+ if (!amdgpu_encoder)
+ return;
+
+ encoder = &amdgpu_encoder->base;
+ encoder->possible_crtcs = 0x1;
+ amdgpu_encoder->enc_priv = NULL;
+ amdgpu_encoder->encoder_enum = encoder_enum;
+ amdgpu_encoder->encoder_id = (encoder_enum & OBJECT_ID_MASK) >> OBJECT_ID_SHIFT;
+ amdgpu_encoder->devices = supported_device;
+ amdgpu_encoder->rmx_type = RMX_OFF;
+ amdgpu_encoder->underscan_type = UNDERSCAN_OFF;
+ amdgpu_encoder->is_ext_encoder = false;
+ amdgpu_encoder->caps = caps;
+
+ drm_encoder_init(dev, encoder, &dce_virtual_encoder_funcs,
+ DRM_MODE_ENCODER_VIRTUAL, NULL);
+ drm_encoder_helper_add(encoder, &dce_virtual_encoder_helper_funcs);
+ DRM_INFO("[FM]encoder: %d is VIRTUAL\n", amdgpu_encoder->encoder_id);
+}
+
+static const struct amdgpu_display_funcs dce_virtual_display_funcs = {
+ .set_vga_render_state = &dce_virtual_set_vga_render_state,
+ .bandwidth_update = &dce_virtual_bandwidth_update,
+ .vblank_get_counter = &dce_virtual_vblank_get_counter,
+ .vblank_wait = &dce_virtual_vblank_wait,
+ .is_display_hung = &dce_virtual_is_display_hung,
+ .backlight_set_level = NULL,
+ .backlight_get_level = NULL,
+ .hpd_sense = &dce_virtual_hpd_sense,
+ .hpd_set_polarity = &dce_virtual_hpd_set_polarity,
+ .hpd_get_gpio_reg = &dce_virtual_hpd_get_gpio_reg,
+ .page_flip = &dce_virtual_page_flip,
+ .page_flip_get_scanoutpos = &dce_virtual_crtc_get_scanoutpos,
+ .add_encoder = &dce_virtual_encoder_add,
+ .add_connector = &amdgpu_connector_add,
+ .stop_mc_access = &dce_virtual_stop_mc_access,
+ .resume_mc_access = &dce_virtual_resume_mc_access,
+};
+
+static void dce_virtual_set_display_funcs(struct amdgpu_device *adev)
+{
+ if (adev->mode_info.funcs == NULL)
+ adev->mode_info.funcs = &dce_virtual_display_funcs;
+}
+
+static enum hrtimer_restart dce_virtual_vblank_timer_handle(struct hrtimer *vblank_timer)
+{
+ struct amdgpu_mode_info *mode_info = container_of(vblank_timer, struct amdgpu_mode_info ,vblank_timer);
+ struct amdgpu_device *adev = container_of(mode_info, struct amdgpu_device ,mode_info);
+ unsigned crtc = 0;
+ drm_handle_vblank(adev->ddev, crtc);
+ dce_virtual_pageflip_irq(adev, NULL, NULL);
+ hrtimer_start(vblank_timer, ktime_set(0, DCE_VIRTUAL_VBLANK_PERIOD), HRTIMER_MODE_REL);
+ return HRTIMER_NORESTART;
+}
+
+static void dce_virtual_set_crtc_vblank_interrupt_state(struct amdgpu_device *adev,
+ int crtc,
+ enum amdgpu_interrupt_state state)
+{
+ if (crtc >= adev->mode_info.num_crtc) {
+ DRM_DEBUG("invalid crtc %d\n", crtc);
+ return;
+ }
+
+ if (state && !adev->mode_info.vsync_timer_enabled) {
+ DRM_DEBUG("Enable software vsync timer\n");
+ hrtimer_init(&adev->mode_info.vblank_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+ hrtimer_set_expires(&adev->mode_info.vblank_timer, ktime_set(0, DCE_VIRTUAL_VBLANK_PERIOD));
+ adev->mode_info.vblank_timer.function = dce_virtual_vblank_timer_handle;
+ hrtimer_start(&adev->mode_info.vblank_timer, ktime_set(0, DCE_VIRTUAL_VBLANK_PERIOD), HRTIMER_MODE_REL);
+ } else if (!state && adev->mode_info.vsync_timer_enabled) {
+ DRM_DEBUG("Disable software vsync timer\n");
+ hrtimer_cancel(&adev->mode_info.vblank_timer);
+ }
+
+ adev->mode_info.vsync_timer_enabled = state;
+ DRM_DEBUG("[FM]set crtc %d vblank interrupt state %d\n", crtc, state);
+}
+
+
+static int dce_virtual_set_crtc_irq_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ switch (type) {
+ case AMDGPU_CRTC_IRQ_VBLANK1:
+ dce_virtual_set_crtc_vblank_interrupt_state(adev, 0, state);
+ break;
+ default:
+ break;
+ }
+ return 0;
+}
+
+static void dce_virtual_crtc_vblank_int_ack(struct amdgpu_device *adev,
+ int crtc)
+{
+ if (crtc >= adev->mode_info.num_crtc) {
+ DRM_DEBUG("invalid crtc %d\n", crtc);
+ return;
+ }
+}
+
+static int dce_virtual_crtc_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ unsigned crtc = 0;
+ unsigned irq_type = AMDGPU_CRTC_IRQ_VBLANK1;
+
+ dce_virtual_crtc_vblank_int_ack(adev, crtc);
+
+ if (amdgpu_irq_enabled(adev, source, irq_type)) {
+ drm_handle_vblank(adev->ddev, crtc);
+ }
+ dce_virtual_pageflip_irq(adev, NULL, NULL);
+ DRM_DEBUG("IH: D%d vblank\n", crtc + 1);
+ return 0;
+}
+
+static int dce_virtual_set_pageflip_irq_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ if (type >= adev->mode_info.num_crtc) {
+ DRM_ERROR("invalid pageflip crtc %d\n", type);
+ return -EINVAL;
+ }
+ DRM_DEBUG("[FM]set pageflip irq type %d state %d\n", type, state);
+
+ return 0;
+}
+
+static int dce_virtual_pageflip_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ unsigned long flags;
+ unsigned crtc_id = 0;
+ struct amdgpu_crtc *amdgpu_crtc;
+ struct amdgpu_flip_work *works;
+
+ crtc_id = 0;
+ amdgpu_crtc = adev->mode_info.crtcs[crtc_id];
+
+ if (crtc_id >= adev->mode_info.num_crtc) {
+ DRM_ERROR("invalid pageflip crtc %d\n", crtc_id);
+ return -EINVAL;
+ }
+
+ /* IRQ could occur when in initial stage */
+ if (amdgpu_crtc == NULL)
+ return 0;
+
+ spin_lock_irqsave(&adev->ddev->event_lock, flags);
+ works = amdgpu_crtc->pflip_works;
+ if (amdgpu_crtc->pflip_status != AMDGPU_FLIP_SUBMITTED) {
+ DRM_DEBUG_DRIVER("amdgpu_crtc->pflip_status = %d != "
+ "AMDGPU_FLIP_SUBMITTED(%d)\n",
+ amdgpu_crtc->pflip_status,
+ AMDGPU_FLIP_SUBMITTED);
+ spin_unlock_irqrestore(&adev->ddev->event_lock, flags);
+ return 0;
+ }
+
+ /* page flip completed. clean up */
+ amdgpu_crtc->pflip_status = AMDGPU_FLIP_NONE;
+ amdgpu_crtc->pflip_works = NULL;
+
+ /* wakeup usersapce */
+ if (works->event)
+ drm_crtc_send_vblank_event(&amdgpu_crtc->base, works->event);
+
+ spin_unlock_irqrestore(&adev->ddev->event_lock, flags);
+
+ drm_crtc_vblank_put(&amdgpu_crtc->base);
+ schedule_work(&works->unpin_work);
+
+ return 0;
+}
+
+static const struct amdgpu_irq_src_funcs dce_virtual_crtc_irq_funcs = {
+ .set = dce_virtual_set_crtc_irq_state,
+ .process = dce_virtual_crtc_irq,
+};
+
+static const struct amdgpu_irq_src_funcs dce_virtual_pageflip_irq_funcs = {
+ .set = dce_virtual_set_pageflip_irq_state,
+ .process = dce_virtual_pageflip_irq,
+};
+
+static void dce_virtual_set_irq_funcs(struct amdgpu_device *adev)
+{
+ adev->crtc_irq.num_types = AMDGPU_CRTC_IRQ_LAST;
+ adev->crtc_irq.funcs = &dce_virtual_crtc_irq_funcs;
+
+ adev->pageflip_irq.num_types = AMDGPU_PAGEFLIP_IRQ_LAST;
+ adev->pageflip_irq.funcs = &dce_virtual_pageflip_irq_funcs;
+}
+
--- /dev/null
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __DCE_VIRTUAL_H__
+#define __DCE_VIRTUAL_H__
+
+extern const struct amd_ip_funcs dce_virtual_ip_funcs;
+#define DCE_VIRTUAL_VBLANK_PERIOD 16666666
+
+#endif
+
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include <linux/firmware.h>
+#include "amdgpu.h"
+#include "amdgpu_ih.h"
+#include "amdgpu_gfx.h"
+#include "amdgpu_ucode.h"
+#include "si/clearstate_si.h"
+#include "si/sid.h"
+
+#define GFX6_NUM_GFX_RINGS 1
+#define GFX6_NUM_COMPUTE_RINGS 2
+#define STATIC_PER_CU_PG_ENABLE (1 << 3)
+#define DYN_PER_CU_PG_ENABLE (1 << 2)
+#define RLC_SAVE_AND_RESTORE_STARTING_OFFSET 0x90
+#define RLC_CLEAR_STATE_DESCRIPTOR_OFFSET 0x3D
+
+
+static void gfx_v6_0_set_ring_funcs(struct amdgpu_device *adev);
+static void gfx_v6_0_set_irq_funcs(struct amdgpu_device *adev);
+static void gfx_v6_0_get_cu_info(struct amdgpu_device *adev);
+
+MODULE_FIRMWARE("radeon/tahiti_pfp.bin");
+MODULE_FIRMWARE("radeon/tahiti_me.bin");
+MODULE_FIRMWARE("radeon/tahiti_ce.bin");
+MODULE_FIRMWARE("radeon/tahiti_rlc.bin");
+
+MODULE_FIRMWARE("radeon/pitcairn_pfp.bin");
+MODULE_FIRMWARE("radeon/pitcairn_me.bin");
+MODULE_FIRMWARE("radeon/pitcairn_ce.bin");
+MODULE_FIRMWARE("radeon/pitcairn_rlc.bin");
+
+MODULE_FIRMWARE("radeon/verde_pfp.bin");
+MODULE_FIRMWARE("radeon/verde_me.bin");
+MODULE_FIRMWARE("radeon/verde_ce.bin");
+MODULE_FIRMWARE("radeon/verde_rlc.bin");
+
+MODULE_FIRMWARE("radeon/oland_pfp.bin");
+MODULE_FIRMWARE("radeon/oland_me.bin");
+MODULE_FIRMWARE("radeon/oland_ce.bin");
+MODULE_FIRMWARE("radeon/oland_rlc.bin");
+
+MODULE_FIRMWARE("radeon/hainan_pfp.bin");
+MODULE_FIRMWARE("radeon/hainan_me.bin");
+MODULE_FIRMWARE("radeon/hainan_ce.bin");
+MODULE_FIRMWARE("radeon/hainan_rlc.bin");
+
+static u32 gfx_v6_0_get_csb_size(struct amdgpu_device *adev);
+static void gfx_v6_0_get_csb_buffer(struct amdgpu_device *adev, volatile u32 *buffer);
+//static void gfx_v6_0_init_cp_pg_table(struct amdgpu_device *adev);
+static void gfx_v6_0_init_pg(struct amdgpu_device *adev);
+
+
+static const u32 verde_rlc_save_restore_register_list[] =
+{
+ (0x8000 << 16) | (0x98f4 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x98f4 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0xe80 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0xe80 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x89bc >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x89bc >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x8c1c >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x8c1c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x98f0 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0xe7c >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x9148 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x9148 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9150 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x897c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8d8c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0xac54 >> 2),
+ 0X00000000,
+ 0x3,
+ (0x9c00 << 16) | (0x98f8 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9910 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9914 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9918 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x991c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9920 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9924 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9928 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x992c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9930 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9934 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9938 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x993c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9940 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9944 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9948 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x994c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9950 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9954 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9958 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x995c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9960 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9964 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9968 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x996c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9970 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9974 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9978 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x997c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9980 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9984 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9988 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x998c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8c00 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8c14 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8c04 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8c08 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x9b7c >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x9b7c >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0xe84 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0xe84 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x89c0 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x89c0 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x914c >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x914c >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x8c20 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x8c20 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x9354 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x9354 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9060 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9364 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9100 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x913c >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x90e0 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x90e4 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x90e8 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x90e0 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x90e4 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x90e8 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8bcc >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8b24 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x88c4 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8e50 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8c0c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8e58 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8e5c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9508 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x950c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9494 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0xac0c >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0xac10 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0xac14 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0xae00 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0xac08 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x88d4 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x88c8 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x88cc >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x89b0 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8b10 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x8a14 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9830 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9834 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9838 >> 2),
+ 0x00000000,
+ (0x9c00 << 16) | (0x9a10 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x9870 >> 2),
+ 0x00000000,
+ (0x8000 << 16) | (0x9874 >> 2),
+ 0x00000000,
+ (0x8001 << 16) | (0x9870 >> 2),
+ 0x00000000,
+ (0x8001 << 16) | (0x9874 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x9870 >> 2),
+ 0x00000000,
+ (0x8040 << 16) | (0x9874 >> 2),
+ 0x00000000,
+ (0x8041 << 16) | (0x9870 >> 2),
+ 0x00000000,
+ (0x8041 << 16) | (0x9874 >> 2),
+ 0x00000000,
+ 0x00000000
+};
+
+static int gfx_v6_0_init_microcode(struct amdgpu_device *adev)
+{
+ const char *chip_name;
+ char fw_name[30];
+ int err;
+ const struct gfx_firmware_header_v1_0 *cp_hdr;
+ const struct rlc_firmware_header_v1_0 *rlc_hdr;
+
+ DRM_DEBUG("\n");
+
+ switch (adev->asic_type) {
+ case CHIP_TAHITI:
+ chip_name = "tahiti";
+ break;
+ case CHIP_PITCAIRN:
+ chip_name = "pitcairn";
+ break;
+ case CHIP_VERDE:
+ chip_name = "verde";
+ break;
+ case CHIP_OLAND:
+ chip_name = "oland";
+ break;
+ case CHIP_HAINAN:
+ chip_name = "hainan";
+ break;
+ default: BUG();
+ }
+
+ snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", chip_name);
+ err = request_firmware(&adev->gfx.pfp_fw, fw_name, adev->dev);
+ if (err)
+ goto out;
+ err = amdgpu_ucode_validate(adev->gfx.pfp_fw);
+ if (err)
+ goto out;
+ cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data;
+ adev->gfx.pfp_fw_version = le32_to_cpu(cp_hdr->header.ucode_version);
+ adev->gfx.pfp_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version);
+
+ snprintf(fw_name, sizeof(fw_name), "radeon/%s_me.bin", chip_name);
+ err = request_firmware(&adev->gfx.me_fw, fw_name, adev->dev);
+ if (err)
+ goto out;
+ err = amdgpu_ucode_validate(adev->gfx.me_fw);
+ if (err)
+ goto out;
+ cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data;
+ adev->gfx.me_fw_version = le32_to_cpu(cp_hdr->header.ucode_version);
+ adev->gfx.me_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version);
+
+ snprintf(fw_name, sizeof(fw_name), "radeon/%s_ce.bin", chip_name);
+ err = request_firmware(&adev->gfx.ce_fw, fw_name, adev->dev);
+ if (err)
+ goto out;
+ err = amdgpu_ucode_validate(adev->gfx.ce_fw);
+ if (err)
+ goto out;
+ cp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data;
+ adev->gfx.ce_fw_version = le32_to_cpu(cp_hdr->header.ucode_version);
+ adev->gfx.ce_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version);
+
+ snprintf(fw_name, sizeof(fw_name), "radeon/%s_rlc.bin", chip_name);
+ err = request_firmware(&adev->gfx.rlc_fw, fw_name, adev->dev);
+ if (err)
+ goto out;
+ err = amdgpu_ucode_validate(adev->gfx.rlc_fw);
+ rlc_hdr = (const struct rlc_firmware_header_v1_0 *)adev->gfx.rlc_fw->data;
+ adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version);
+ adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version);
+
+out:
+ if (err) {
+ printk(KERN_ERR
+ "gfx6: Failed to load firmware \"%s\"\n",
+ fw_name);
+ release_firmware(adev->gfx.pfp_fw);
+ adev->gfx.pfp_fw = NULL;
+ release_firmware(adev->gfx.me_fw);
+ adev->gfx.me_fw = NULL;
+ release_firmware(adev->gfx.ce_fw);
+ adev->gfx.ce_fw = NULL;
+ release_firmware(adev->gfx.rlc_fw);
+ adev->gfx.rlc_fw = NULL;
+ }
+ return err;
+}
+
+static void gfx_v6_0_tiling_mode_table_init(struct amdgpu_device *adev)
+{
+ const u32 num_tile_mode_states = 32;
+ u32 reg_offset, gb_tile_moden, split_equal_to_row_size;
+
+ switch (adev->gfx.config.mem_row_size_in_kb) {
+ case 1:
+ split_equal_to_row_size = ADDR_SURF_TILE_SPLIT_1KB;
+ break;
+ case 2:
+ default:
+ split_equal_to_row_size = ADDR_SURF_TILE_SPLIT_2KB;
+ break;
+ case 4:
+ split_equal_to_row_size = ADDR_SURF_TILE_SPLIT_4KB;
+ break;
+ }
+
+ if (adev->asic_type == CHIP_VERDE ||
+ adev->asic_type == CHIP_OLAND ||
+ adev->asic_type == CHIP_HAINAN) {
+ for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
+ switch (reg_offset) {
+ case 0:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+ break;
+ case 1:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+ break;
+ case 2:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+ break;
+ case 3:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+ break;
+ case 4:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 5:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(split_equal_to_row_size) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 6:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(split_equal_to_row_size) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 7:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(split_equal_to_row_size) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+ break;
+ case 8:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 9:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 10:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+ break;
+ case 11:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 12:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 13:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 14:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 15:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 16:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 17:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P4_8x16) |
+ TILE_SPLIT(split_equal_to_row_size) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 21:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 22:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+ break;
+ case 23:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 24:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 25:
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_1KB) |
+ NUM_BANKS(ADDR_SURF_8_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+ break;
+ default:
+ gb_tile_moden = 0;
+ break;
+ }
+ adev->gfx.config.tile_mode_array[reg_offset] = gb_tile_moden;
+ WREG32(GB_TILE_MODE0 + reg_offset, gb_tile_moden);
+ }
+ } else if ((adev->asic_type == CHIP_TAHITI) || (adev->asic_type == CHIP_PITCAIRN)) {
+ for (reg_offset = 0; reg_offset < num_tile_mode_states; reg_offset++) {
+ switch (reg_offset) {
+ case 0: /* non-AA compressed depth or any compressed stencil */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 1: /* 2xAA/4xAA compressed depth only */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 2: /* 8xAA compressed depth only */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 3: /* 2xAA/4xAA compressed depth with stencil (for depth buffer) */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_128B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 4: /* Maps w/ a dimension less than the 2D macro-tile dimensions (for mipmapped depth textures) */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 5: /* Uncompressed 16bpp depth - and stencil buffer allocated with it */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(split_equal_to_row_size) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 6: /* Uncompressed 32bpp depth - and stencil buffer allocated with it */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(split_equal_to_row_size) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+ break;
+ case 7: /* Uncompressed 8bpp stencil without depth (drivers typically do not use) */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DEPTH_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(split_equal_to_row_size) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 8: /* 1D and 1D Array Surfaces */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_LINEAR_ALIGNED) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 9: /* Displayable maps. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 10: /* Display 8bpp. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 11: /* Display 16bpp. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 12: /* Display 32bpp. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_DISPLAY_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+ break;
+ case 13: /* Thin. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_1D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 14: /* Thin 8 bpp. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+ break;
+ case 15: /* Thin 16 bpp. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+ break;
+ case 16: /* Thin 32 bpp. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+ break;
+ case 17: /* Thin 64 bpp. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(split_equal_to_row_size) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+ break;
+ case 21: /* 8 bpp PRT. */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_2) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 22: /* 16 bpp PRT */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_4) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_4));
+ break;
+ case 23: /* 32 bpp PRT */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_256B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_2) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 24: /* 64 bpp PRT */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_512B) |
+ NUM_BANKS(ADDR_SURF_16_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_2));
+ break;
+ case 25: /* 128 bpp PRT */
+ gb_tile_moden = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
+ MICRO_TILE_MODE(ADDR_SURF_THIN_MICRO_TILING) |
+ PIPE_CONFIG(ADDR_SURF_P8_32x32_8x16) |
+ TILE_SPLIT(ADDR_SURF_TILE_SPLIT_1KB) |
+ NUM_BANKS(ADDR_SURF_8_BANK) |
+ BANK_WIDTH(ADDR_SURF_BANK_WIDTH_1) |
+ BANK_HEIGHT(ADDR_SURF_BANK_HEIGHT_1) |
+ MACRO_TILE_ASPECT(ADDR_SURF_MACRO_ASPECT_1));
+ break;
+ default:
+ gb_tile_moden = 0;
+ break;
+ }
+ adev->gfx.config.tile_mode_array[reg_offset] = gb_tile_moden;
+ WREG32(GB_TILE_MODE0 + reg_offset, gb_tile_moden);
+ }
+ } else{
+
+ DRM_ERROR("unknown asic: 0x%x\n", adev->asic_type);
+ }
+
+}
+
+static void gfx_v6_0_select_se_sh(struct amdgpu_device *adev, u32 se_num,
+ u32 sh_num, u32 instance)
+{
+ u32 data;
+
+ if (instance == 0xffffffff)
+ data = INSTANCE_BROADCAST_WRITES;
+ else
+ data = INSTANCE_INDEX(instance);
+
+ if ((se_num == 0xffffffff) && (sh_num == 0xffffffff))
+ data |= SH_BROADCAST_WRITES | SE_BROADCAST_WRITES;
+ else if (se_num == 0xffffffff)
+ data |= SE_BROADCAST_WRITES | SH_INDEX(sh_num);
+ else if (sh_num == 0xffffffff)
+ data |= SH_BROADCAST_WRITES | SE_INDEX(se_num);
+ else
+ data |= SH_INDEX(sh_num) | SE_INDEX(se_num);
+ WREG32(GRBM_GFX_INDEX, data);
+}
+
+static u32 gfx_v6_0_create_bitmask(u32 bit_width)
+{
+ return (u32)(((u64)1 << bit_width) - 1);
+}
+
+static u32 gfx_v6_0_get_rb_disabled(struct amdgpu_device *adev,
+ u32 max_rb_num_per_se,
+ u32 sh_per_se)
+{
+ u32 data, mask;
+
+ data = RREG32(CC_RB_BACKEND_DISABLE);
+ data &= BACKEND_DISABLE_MASK;
+ data |= RREG32(GC_USER_RB_BACKEND_DISABLE);
+
+ data >>= BACKEND_DISABLE_SHIFT;
+
+ mask = gfx_v6_0_create_bitmask(max_rb_num_per_se / sh_per_se);
+
+ return data & mask;
+}
+
+static void gfx_v6_0_setup_rb(struct amdgpu_device *adev,
+ u32 se_num, u32 sh_per_se,
+ u32 max_rb_num_per_se)
+{
+ int i, j;
+ u32 data, mask;
+ u32 disabled_rbs = 0;
+ u32 enabled_rbs = 0;
+
+ mutex_lock(&adev->grbm_idx_mutex);
+ for (i = 0; i < se_num; i++) {
+ for (j = 0; j < sh_per_se; j++) {
+ gfx_v6_0_select_se_sh(adev, i, j, 0xffffffff);
+ data = gfx_v6_0_get_rb_disabled(adev, max_rb_num_per_se, sh_per_se);
+ disabled_rbs |= data << ((i * sh_per_se + j) * TAHITI_RB_BITMAP_WIDTH_PER_SH);
+ }
+ }
+ gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+ mutex_unlock(&adev->grbm_idx_mutex);
+
+ mask = 1;
+ for (i = 0; i < max_rb_num_per_se * se_num; i++) {
+ if (!(disabled_rbs & mask))
+ enabled_rbs |= mask;
+ mask <<= 1;
+ }
+
+ adev->gfx.config.backend_enable_mask = enabled_rbs;
+ adev->gfx.config.num_rbs = hweight32(enabled_rbs);
+
+ mutex_lock(&adev->grbm_idx_mutex);
+ for (i = 0; i < se_num; i++) {
+ gfx_v6_0_select_se_sh(adev, i, 0xffffffff, 0xffffffff);
+ data = 0;
+ for (j = 0; j < sh_per_se; j++) {
+ switch (enabled_rbs & 3) {
+ case 1:
+ data |= (RASTER_CONFIG_RB_MAP_0 << (i * sh_per_se + j) * 2);
+ break;
+ case 2:
+ data |= (RASTER_CONFIG_RB_MAP_3 << (i * sh_per_se + j) * 2);
+ break;
+ case 3:
+ default:
+ data |= (RASTER_CONFIG_RB_MAP_2 << (i * sh_per_se + j) * 2);
+ break;
+ }
+ enabled_rbs >>= 2;
+ }
+ WREG32(PA_SC_RASTER_CONFIG, data);
+ }
+ gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+ mutex_unlock(&adev->grbm_idx_mutex);
+}
+/*
+static void gmc_v6_0_init_compute_vmid(struct amdgpu_device *adev)
+{
+}
+*/
+
+static u32 gfx_v6_0_get_cu_enabled(struct amdgpu_device *adev, u32 cu_per_sh)
+{
+ u32 data, mask;
+
+ data = RREG32(CC_GC_SHADER_ARRAY_CONFIG);
+ data &= INACTIVE_CUS_MASK;
+ data |= RREG32(GC_USER_SHADER_ARRAY_CONFIG);
+
+ data >>= INACTIVE_CUS_SHIFT;
+
+ mask = gfx_v6_0_create_bitmask(cu_per_sh);
+
+ return ~data & mask;
+}
+
+
+static void gfx_v6_0_setup_spi(struct amdgpu_device *adev,
+ u32 se_num, u32 sh_per_se,
+ u32 cu_per_sh)
+{
+ int i, j, k;
+ u32 data, mask;
+ u32 active_cu = 0;
+
+ mutex_lock(&adev->grbm_idx_mutex);
+ for (i = 0; i < se_num; i++) {
+ for (j = 0; j < sh_per_se; j++) {
+ gfx_v6_0_select_se_sh(adev, i, j, 0xffffffff);
+ data = RREG32(SPI_STATIC_THREAD_MGMT_3);
+ active_cu = gfx_v6_0_get_cu_enabled(adev, cu_per_sh);
+
+ mask = 1;
+ for (k = 0; k < 16; k++) {
+ mask <<= k;
+ if (active_cu & mask) {
+ data &= ~mask;
+ WREG32(SPI_STATIC_THREAD_MGMT_3, data);
+ break;
+ }
+ }
+ }
+ }
+ gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+ mutex_unlock(&adev->grbm_idx_mutex);
+}
+
+static void gfx_v6_0_gpu_init(struct amdgpu_device *adev)
+{
+ u32 gb_addr_config = 0;
+ u32 mc_shared_chmap, mc_arb_ramcfg;
+ u32 sx_debug_1;
+ u32 hdp_host_path_cntl;
+ u32 tmp;
+
+ switch (adev->asic_type) {
+ case CHIP_TAHITI:
+ adev->gfx.config.max_shader_engines = 2;
+ adev->gfx.config.max_tile_pipes = 12;
+ adev->gfx.config.max_cu_per_sh = 8;
+ adev->gfx.config.max_sh_per_se = 2;
+ adev->gfx.config.max_backends_per_se = 4;
+ adev->gfx.config.max_texture_channel_caches = 12;
+ adev->gfx.config.max_gprs = 256;
+ adev->gfx.config.max_gs_threads = 32;
+ adev->gfx.config.max_hw_contexts = 8;
+
+ adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
+ adev->gfx.config.sc_prim_fifo_size_backend = 0x100;
+ adev->gfx.config.sc_hiz_tile_fifo_size = 0x30;
+ adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = TAHITI_GB_ADDR_CONFIG_GOLDEN;
+ break;
+ case CHIP_PITCAIRN:
+ adev->gfx.config.max_shader_engines = 2;
+ adev->gfx.config.max_tile_pipes = 8;
+ adev->gfx.config.max_cu_per_sh = 5;
+ adev->gfx.config.max_sh_per_se = 2;
+ adev->gfx.config.max_backends_per_se = 4;
+ adev->gfx.config.max_texture_channel_caches = 8;
+ adev->gfx.config.max_gprs = 256;
+ adev->gfx.config.max_gs_threads = 32;
+ adev->gfx.config.max_hw_contexts = 8;
+
+ adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
+ adev->gfx.config.sc_prim_fifo_size_backend = 0x100;
+ adev->gfx.config.sc_hiz_tile_fifo_size = 0x30;
+ adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = TAHITI_GB_ADDR_CONFIG_GOLDEN;
+ break;
+
+ case CHIP_VERDE:
+ adev->gfx.config.max_shader_engines = 1;
+ adev->gfx.config.max_tile_pipes = 4;
+ adev->gfx.config.max_cu_per_sh = 5;
+ adev->gfx.config.max_sh_per_se = 2;
+ adev->gfx.config.max_backends_per_se = 4;
+ adev->gfx.config.max_texture_channel_caches = 4;
+ adev->gfx.config.max_gprs = 256;
+ adev->gfx.config.max_gs_threads = 32;
+ adev->gfx.config.max_hw_contexts = 8;
+
+ adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
+ adev->gfx.config.sc_prim_fifo_size_backend = 0x40;
+ adev->gfx.config.sc_hiz_tile_fifo_size = 0x30;
+ adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = VERDE_GB_ADDR_CONFIG_GOLDEN;
+ break;
+ case CHIP_OLAND:
+ adev->gfx.config.max_shader_engines = 1;
+ adev->gfx.config.max_tile_pipes = 4;
+ adev->gfx.config.max_cu_per_sh = 6;
+ adev->gfx.config.max_sh_per_se = 1;
+ adev->gfx.config.max_backends_per_se = 2;
+ adev->gfx.config.max_texture_channel_caches = 4;
+ adev->gfx.config.max_gprs = 256;
+ adev->gfx.config.max_gs_threads = 16;
+ adev->gfx.config.max_hw_contexts = 8;
+
+ adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
+ adev->gfx.config.sc_prim_fifo_size_backend = 0x40;
+ adev->gfx.config.sc_hiz_tile_fifo_size = 0x30;
+ adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = VERDE_GB_ADDR_CONFIG_GOLDEN;
+ break;
+ case CHIP_HAINAN:
+ adev->gfx.config.max_shader_engines = 1;
+ adev->gfx.config.max_tile_pipes = 4;
+ adev->gfx.config.max_cu_per_sh = 5;
+ adev->gfx.config.max_sh_per_se = 1;
+ adev->gfx.config.max_backends_per_se = 1;
+ adev->gfx.config.max_texture_channel_caches = 2;
+ adev->gfx.config.max_gprs = 256;
+ adev->gfx.config.max_gs_threads = 16;
+ adev->gfx.config.max_hw_contexts = 8;
+
+ adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
+ adev->gfx.config.sc_prim_fifo_size_backend = 0x40;
+ adev->gfx.config.sc_hiz_tile_fifo_size = 0x30;
+ adev->gfx.config.sc_earlyz_tile_fifo_size = 0x130;
+ gb_addr_config = HAINAN_GB_ADDR_CONFIG_GOLDEN;
+ break;
+ default:
+ BUG();
+ break;
+ }
+
+ WREG32(GRBM_CNTL, GRBM_READ_TIMEOUT(0xff));
+ WREG32(SRBM_INT_CNTL, 1);
+ WREG32(SRBM_INT_ACK, 1);
+
+ WREG32(BIF_FB_EN, FB_READ_EN | FB_WRITE_EN);
+
+ mc_shared_chmap = RREG32(MC_SHARED_CHMAP);
+ mc_arb_ramcfg = RREG32(MC_ARB_RAMCFG);
+
+ adev->gfx.config.num_tile_pipes = adev->gfx.config.max_tile_pipes;
+ adev->gfx.config.mem_max_burst_length_bytes = 256;
+ tmp = (mc_arb_ramcfg & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT;
+ adev->gfx.config.mem_row_size_in_kb = (4 * (1 << (8 + tmp))) / 1024;
+ if (adev->gfx.config.mem_row_size_in_kb > 4)
+ adev->gfx.config.mem_row_size_in_kb = 4;
+ adev->gfx.config.shader_engine_tile_size = 32;
+ adev->gfx.config.num_gpus = 1;
+ adev->gfx.config.multi_gpu_tile_size = 64;
+
+ gb_addr_config &= ~ROW_SIZE_MASK;
+ switch (adev->gfx.config.mem_row_size_in_kb) {
+ case 1:
+ default:
+ gb_addr_config |= ROW_SIZE(0);
+ break;
+ case 2:
+ gb_addr_config |= ROW_SIZE(1);
+ break;
+ case 4:
+ gb_addr_config |= ROW_SIZE(2);
+ break;
+ }
+ adev->gfx.config.gb_addr_config = gb_addr_config;
+
+ WREG32(GB_ADDR_CONFIG, gb_addr_config);
+ WREG32(DMIF_ADDR_CONFIG, gb_addr_config);
+ WREG32(DMIF_ADDR_CALC, gb_addr_config);
+ WREG32(HDP_ADDR_CONFIG, gb_addr_config);
+ WREG32(DMA_TILING_CONFIG + DMA0_REGISTER_OFFSET, gb_addr_config);
+ WREG32(DMA_TILING_CONFIG + DMA1_REGISTER_OFFSET, gb_addr_config);
+#if 0
+ if (adev->has_uvd) {
+ WREG32(UVD_UDEC_ADDR_CONFIG, gb_addr_config);
+ WREG32(UVD_UDEC_DB_ADDR_CONFIG, gb_addr_config);
+ WREG32(UVD_UDEC_DBW_ADDR_CONFIG, gb_addr_config);
+ }
+#endif
+ gfx_v6_0_tiling_mode_table_init(adev);
+
+ gfx_v6_0_setup_rb(adev, adev->gfx.config.max_shader_engines,
+ adev->gfx.config.max_sh_per_se,
+ adev->gfx.config.max_backends_per_se);
+
+ gfx_v6_0_setup_spi(adev, adev->gfx.config.max_shader_engines,
+ adev->gfx.config.max_sh_per_se,
+ adev->gfx.config.max_cu_per_sh);
+
+ gfx_v6_0_get_cu_info(adev);
+
+ WREG32(CP_QUEUE_THRESHOLDS, (ROQ_IB1_START(0x16) |
+ ROQ_IB2_START(0x2b)));
+ WREG32(CP_MEQ_THRESHOLDS, MEQ1_START(0x30) | MEQ2_START(0x60));
+
+ sx_debug_1 = RREG32(SX_DEBUG_1);
+ WREG32(SX_DEBUG_1, sx_debug_1);
+
+ WREG32(SPI_CONFIG_CNTL_1, VTX_DONE_DELAY(4));
+
+ WREG32(PA_SC_FIFO_SIZE, (SC_FRONTEND_PRIM_FIFO_SIZE(adev->gfx.config.sc_prim_fifo_size_frontend) |
+ SC_BACKEND_PRIM_FIFO_SIZE(adev->gfx.config.sc_prim_fifo_size_backend) |
+ SC_HIZ_TILE_FIFO_SIZE(adev->gfx.config.sc_hiz_tile_fifo_size) |
+ SC_EARLYZ_TILE_FIFO_SIZE(adev->gfx.config.sc_earlyz_tile_fifo_size)));
+
+ WREG32(VGT_NUM_INSTANCES, 1);
+ WREG32(CP_PERFMON_CNTL, 0);
+ WREG32(SQ_CONFIG, 0);
+ WREG32(PA_SC_FORCE_EOV_MAX_CNTS, (FORCE_EOV_MAX_CLK_CNT(4095) |
+ FORCE_EOV_MAX_REZ_CNT(255)));
+
+ WREG32(VGT_CACHE_INVALIDATION, CACHE_INVALIDATION(VC_AND_TC) |
+ AUTO_INVLD_EN(ES_AND_GS_AUTO));
+
+ WREG32(VGT_GS_VERTEX_REUSE, 16);
+ WREG32(PA_SC_LINE_STIPPLE_STATE, 0);
+
+ WREG32(CB_PERFCOUNTER0_SELECT0, 0);
+ WREG32(CB_PERFCOUNTER0_SELECT1, 0);
+ WREG32(CB_PERFCOUNTER1_SELECT0, 0);
+ WREG32(CB_PERFCOUNTER1_SELECT1, 0);
+ WREG32(CB_PERFCOUNTER2_SELECT0, 0);
+ WREG32(CB_PERFCOUNTER2_SELECT1, 0);
+ WREG32(CB_PERFCOUNTER3_SELECT0, 0);
+ WREG32(CB_PERFCOUNTER3_SELECT1, 0);
+
+ hdp_host_path_cntl = RREG32(HDP_HOST_PATH_CNTL);
+ WREG32(HDP_HOST_PATH_CNTL, hdp_host_path_cntl);
+
+ WREG32(PA_CL_ENHANCE, CLIP_VTX_REORDER_ENA | NUM_CLIP_SEQ(3));
+
+ udelay(50);
+}
+
+
+static void gfx_v6_0_scratch_init(struct amdgpu_device *adev)
+{
+ int i;
+
+ adev->gfx.scratch.num_reg = 7;
+ adev->gfx.scratch.reg_base = SCRATCH_REG0;
+ for (i = 0; i < adev->gfx.scratch.num_reg; i++) {
+ adev->gfx.scratch.free[i] = true;
+ adev->gfx.scratch.reg[i] = adev->gfx.scratch.reg_base + i;
+ }
+}
+
+static int gfx_v6_0_ring_test_ring(struct amdgpu_ring *ring)
+{
+ struct amdgpu_device *adev = ring->adev;
+ uint32_t scratch;
+ uint32_t tmp = 0;
+ unsigned i;
+ int r;
+
+ r = amdgpu_gfx_scratch_get(adev, &scratch);
+ if (r) {
+ DRM_ERROR("amdgpu: cp failed to get scratch reg (%d).\n", r);
+ return r;
+ }
+ WREG32(scratch, 0xCAFEDEAD);
+
+ r = amdgpu_ring_alloc(ring, 3);
+ if (r) {
+ DRM_ERROR("amdgpu: cp failed to lock ring %d (%d).\n", ring->idx, r);
+ amdgpu_gfx_scratch_free(adev, scratch);
+ return r;
+ }
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
+ amdgpu_ring_write(ring, (scratch - PACKET3_SET_CONFIG_REG_START));
+ amdgpu_ring_write(ring, 0xDEADBEEF);
+ amdgpu_ring_commit(ring);
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ tmp = RREG32(scratch);
+ if (tmp == 0xDEADBEEF)
+ break;
+ DRM_UDELAY(1);
+ }
+ if (i < adev->usec_timeout) {
+ DRM_INFO("ring test on %d succeeded in %d usecs\n", ring->idx, i);
+ } else {
+ DRM_ERROR("amdgpu: ring %d test failed (scratch(0x%04X)=0x%08X)\n",
+ ring->idx, scratch, tmp);
+ r = -EINVAL;
+ }
+ amdgpu_gfx_scratch_free(adev, scratch);
+ return r;
+}
+
+static void gfx_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
+{
+ /* flush hdp cache */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
+ amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) |
+ WRITE_DATA_DST_SEL(0)));
+ amdgpu_ring_write(ring, HDP_MEM_COHERENCY_FLUSH_CNTL);
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, 0x1);
+}
+
+/**
+ * gfx_v6_0_ring_emit_hdp_invalidate - emit an hdp invalidate on the cp
+ *
+ * @adev: amdgpu_device pointer
+ * @ridx: amdgpu ring index
+ *
+ * Emits an hdp invalidate on the cp.
+ */
+static void gfx_v6_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring)
+{
+ amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
+ amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) |
+ WRITE_DATA_DST_SEL(0)));
+ amdgpu_ring_write(ring, HDP_DEBUG0);
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, 0x1);
+}
+
+static void gfx_v6_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
+ u64 seq, unsigned flags)
+{
+ bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT;
+ bool int_sel = flags & AMDGPU_FENCE_FLAG_INT;
+ /* flush read cache over gart */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
+ amdgpu_ring_write(ring, (CP_COHER_CNTL2 - PACKET3_SET_CONFIG_REG_START));
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SURFACE_SYNC, 3));
+ amdgpu_ring_write(ring, PACKET3_TCL1_ACTION_ENA |
+ PACKET3_TC_ACTION_ENA |
+ PACKET3_SH_KCACHE_ACTION_ENA |
+ PACKET3_SH_ICACHE_ACTION_ENA);
+ amdgpu_ring_write(ring, 0xFFFFFFFF);
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, 10); /* poll interval */
+ /* EVENT_WRITE_EOP - flush caches, send int */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_EVENT_WRITE_EOP, 4));
+ amdgpu_ring_write(ring, EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) | EVENT_INDEX(5));
+ amdgpu_ring_write(ring, addr & 0xfffffffc);
+ amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xffff) |
+ DATA_SEL(write64bit ? 2 : 1) | INT_SEL(int_sel ? 2 : 0));
+ amdgpu_ring_write(ring, lower_32_bits(seq));
+ amdgpu_ring_write(ring, upper_32_bits(seq));
+}
+
+static void gfx_v6_0_ring_emit_ib(struct amdgpu_ring *ring,
+ struct amdgpu_ib *ib,
+ unsigned vm_id, bool ctx_switch)
+{
+ u32 header, control = 0;
+
+ /* insert SWITCH_BUFFER packet before first IB in the ring frame */
+ if (ctx_switch) {
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
+ amdgpu_ring_write(ring, 0);
+ }
+
+ if (ib->flags & AMDGPU_IB_FLAG_CE)
+ header = PACKET3(PACKET3_INDIRECT_BUFFER_CONST, 2);
+ else
+ header = PACKET3(PACKET3_INDIRECT_BUFFER, 2);
+
+ control |= ib->length_dw | (vm_id << 24);
+
+ amdgpu_ring_write(ring, header);
+ amdgpu_ring_write(ring,
+#ifdef __BIG_ENDIAN
+ (2 << 0) |
+#endif
+ (ib->gpu_addr & 0xFFFFFFFC));
+ amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr) & 0xFFFF);
+ amdgpu_ring_write(ring, control);
+}
+
+/**
+ * gfx_v6_0_ring_test_ib - basic ring IB test
+ *
+ * @ring: amdgpu_ring structure holding ring information
+ *
+ * Allocate an IB and execute it on the gfx ring (SI).
+ * Provides a basic gfx ring test to verify that IBs are working.
+ * Returns 0 on success, error on failure.
+ */
+static int gfx_v6_0_ring_test_ib(struct amdgpu_ring *ring, long timeout)
+{
+ struct amdgpu_device *adev = ring->adev;
+ struct amdgpu_ib ib;
+ struct fence *f = NULL;
+ uint32_t scratch;
+ uint32_t tmp = 0;
+ long r;
+
+ r = amdgpu_gfx_scratch_get(adev, &scratch);
+ if (r) {
+ DRM_ERROR("amdgpu: failed to get scratch reg (%ld).\n", r);
+ return r;
+ }
+ WREG32(scratch, 0xCAFEDEAD);
+ memset(&ib, 0, sizeof(ib));
+ r = amdgpu_ib_get(adev, NULL, 256, &ib);
+ if (r) {
+ DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r);
+ goto err1;
+ }
+ ib.ptr[0] = PACKET3(PACKET3_SET_CONFIG_REG, 1);
+ ib.ptr[1] = ((scratch - PACKET3_SET_CONFIG_REG_START));
+ ib.ptr[2] = 0xDEADBEEF;
+ ib.length_dw = 3;
+
+ r = amdgpu_ib_schedule(ring, 1, &ib, NULL, NULL, &f);
+ if (r)
+ goto err2;
+
+ r = fence_wait_timeout(f, false, timeout);
+ if (r == 0) {
+ DRM_ERROR("amdgpu: IB test timed out\n");
+ r = -ETIMEDOUT;
+ goto err2;
+ } else if (r < 0) {
+ DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
+ goto err2;
+ }
+ tmp = RREG32(scratch);
+ if (tmp == 0xDEADBEEF) {
+ DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
+ r = 0;
+ } else {
+ DRM_ERROR("amdgpu: ib test failed (scratch(0x%04X)=0x%08X)\n",
+ scratch, tmp);
+ r = -EINVAL;
+ }
+
+err2:
+ amdgpu_ib_free(adev, &ib, NULL);
+ fence_put(f);
+err1:
+ amdgpu_gfx_scratch_free(adev, scratch);
+ return r;
+}
+
+static void gfx_v6_0_cp_gfx_enable(struct amdgpu_device *adev, bool enable)
+{
+ int i;
+ if (enable)
+ WREG32(CP_ME_CNTL, 0);
+ else {
+ WREG32(CP_ME_CNTL, (CP_ME_HALT | CP_PFP_HALT | CP_CE_HALT));
+ WREG32(SCRATCH_UMSK, 0);
+ for (i = 0; i < adev->gfx.num_gfx_rings; i++)
+ adev->gfx.gfx_ring[i].ready = false;
+ for (i = 0; i < adev->gfx.num_compute_rings; i++)
+ adev->gfx.compute_ring[i].ready = false;
+ }
+ udelay(50);
+}
+
+static int gfx_v6_0_cp_gfx_load_microcode(struct amdgpu_device *adev)
+{
+ unsigned i;
+ const struct gfx_firmware_header_v1_0 *pfp_hdr;
+ const struct gfx_firmware_header_v1_0 *ce_hdr;
+ const struct gfx_firmware_header_v1_0 *me_hdr;
+ const __le32 *fw_data;
+ u32 fw_size;
+
+ if (!adev->gfx.me_fw || !adev->gfx.pfp_fw || !adev->gfx.ce_fw)
+ return -EINVAL;
+
+ gfx_v6_0_cp_gfx_enable(adev, false);
+ pfp_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data;
+ ce_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data;
+ me_hdr = (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data;
+
+ amdgpu_ucode_print_gfx_hdr(&pfp_hdr->header);
+ amdgpu_ucode_print_gfx_hdr(&ce_hdr->header);
+ amdgpu_ucode_print_gfx_hdr(&me_hdr->header);
+
+ /* PFP */
+ fw_data = (const __le32 *)
+ (adev->gfx.pfp_fw->data + le32_to_cpu(pfp_hdr->header.ucode_array_offset_bytes));
+ fw_size = le32_to_cpu(pfp_hdr->header.ucode_size_bytes) / 4;
+ WREG32(CP_PFP_UCODE_ADDR, 0);
+ for (i = 0; i < fw_size; i++)
+ WREG32(CP_PFP_UCODE_DATA, le32_to_cpup(fw_data++));
+ WREG32(CP_PFP_UCODE_ADDR, 0);
+
+ /* CE */
+ fw_data = (const __le32 *)
+ (adev->gfx.ce_fw->data + le32_to_cpu(ce_hdr->header.ucode_array_offset_bytes));
+ fw_size = le32_to_cpu(ce_hdr->header.ucode_size_bytes) / 4;
+ WREG32(CP_CE_UCODE_ADDR, 0);
+ for (i = 0; i < fw_size; i++)
+ WREG32(CP_CE_UCODE_DATA, le32_to_cpup(fw_data++));
+ WREG32(CP_CE_UCODE_ADDR, 0);
+
+ /* ME */
+ fw_data = (const __be32 *)
+ (adev->gfx.me_fw->data + le32_to_cpu(me_hdr->header.ucode_array_offset_bytes));
+ fw_size = le32_to_cpu(me_hdr->header.ucode_size_bytes) / 4;
+ WREG32(CP_ME_RAM_WADDR, 0);
+ for (i = 0; i < fw_size; i++)
+ WREG32(CP_ME_RAM_DATA, le32_to_cpup(fw_data++));
+ WREG32(CP_ME_RAM_WADDR, 0);
+
+
+ WREG32(CP_PFP_UCODE_ADDR, 0);
+ WREG32(CP_CE_UCODE_ADDR, 0);
+ WREG32(CP_ME_RAM_WADDR, 0);
+ WREG32(CP_ME_RAM_RADDR, 0);
+ return 0;
+}
+
+static int gfx_v6_0_cp_gfx_start(struct amdgpu_device *adev)
+{
+ const struct cs_section_def *sect = NULL;
+ const struct cs_extent_def *ext = NULL;
+ struct amdgpu_ring *ring = &adev->gfx.gfx_ring[0];
+ int r, i;
+
+ r = amdgpu_ring_alloc(ring, 7 + 4);
+ if (r) {
+ DRM_ERROR("amdgpu: cp failed to lock ring (%d).\n", r);
+ return r;
+ }
+ amdgpu_ring_write(ring, PACKET3(PACKET3_ME_INITIALIZE, 5));
+ amdgpu_ring_write(ring, 0x1);
+ amdgpu_ring_write(ring, 0x0);
+ amdgpu_ring_write(ring, adev->gfx.config.max_hw_contexts - 1);
+ amdgpu_ring_write(ring, PACKET3_ME_INITIALIZE_DEVICE_ID(1));
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, 0);
+
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SET_BASE, 2));
+ amdgpu_ring_write(ring, PACKET3_BASE_INDEX(CE_PARTITION_BASE));
+ amdgpu_ring_write(ring, 0xc000);
+ amdgpu_ring_write(ring, 0xe000);
+ amdgpu_ring_commit(ring);
+
+ gfx_v6_0_cp_gfx_enable(adev, true);
+
+ r = amdgpu_ring_alloc(ring, gfx_v6_0_get_csb_size(adev) + 10);
+ if (r) {
+ DRM_ERROR("amdgpu: cp failed to lock ring (%d).\n", r);
+ return r;
+ }
+
+ amdgpu_ring_write(ring, PACKET3(PACKET3_PREAMBLE_CNTL, 0));
+ amdgpu_ring_write(ring, PACKET3_PREAMBLE_BEGIN_CLEAR_STATE);
+
+ for (sect = adev->gfx.rlc.cs_data; sect->section != NULL; ++sect) {
+ for (ext = sect->section; ext->extent != NULL; ++ext) {
+ if (sect->id == SECT_CONTEXT) {
+ amdgpu_ring_write(ring,
+ PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count));
+ amdgpu_ring_write(ring, ext->reg_index - PACKET3_SET_CONTEXT_REG_START);
+ for (i = 0; i < ext->reg_count; i++)
+ amdgpu_ring_write(ring, ext->extent[i]);
+ }
+ }
+ }
+
+ amdgpu_ring_write(ring, PACKET3(PACKET3_PREAMBLE_CNTL, 0));
+ amdgpu_ring_write(ring, PACKET3_PREAMBLE_END_CLEAR_STATE);
+
+ amdgpu_ring_write(ring, PACKET3(PACKET3_CLEAR_STATE, 0));
+ amdgpu_ring_write(ring, 0);
+
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONTEXT_REG, 2));
+ amdgpu_ring_write(ring, 0x00000316);
+ amdgpu_ring_write(ring, 0x0000000e);
+ amdgpu_ring_write(ring, 0x00000010);
+
+ amdgpu_ring_commit(ring);
+
+ return 0;
+}
+
+static int gfx_v6_0_cp_gfx_resume(struct amdgpu_device *adev)
+{
+ struct amdgpu_ring *ring;
+ u32 tmp;
+ u32 rb_bufsz;
+ int r;
+ u64 rptr_addr;
+
+ WREG32(CP_SEM_WAIT_TIMER, 0x0);
+ WREG32(CP_SEM_INCOMPLETE_TIMER_CNTL, 0x0);
+
+ /* Set the write pointer delay */
+ WREG32(CP_RB_WPTR_DELAY, 0);
+
+ WREG32(CP_DEBUG, 0);
+ WREG32(SCRATCH_ADDR, 0);
+
+ /* ring 0 - compute and gfx */
+ /* Set ring buffer size */
+ ring = &adev->gfx.gfx_ring[0];
+ rb_bufsz = order_base_2(ring->ring_size / 8);
+ tmp = (order_base_2(AMDGPU_GPU_PAGE_SIZE/8) << 8) | rb_bufsz;
+
+#ifdef __BIG_ENDIAN
+ tmp |= BUF_SWAP_32BIT;
+#endif
+ WREG32(CP_RB0_CNTL, tmp);
+
+ /* Initialize the ring buffer's read and write pointers */
+ WREG32(CP_RB0_CNTL, tmp | RB_RPTR_WR_ENA);
+ ring->wptr = 0;
+ WREG32(CP_RB0_WPTR, ring->wptr);
+
+ /* set the wb address whether it's enabled or not */
+ rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4);
+ WREG32(CP_RB0_RPTR_ADDR, lower_32_bits(rptr_addr));
+ WREG32(CP_RB0_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF);
+
+ WREG32(SCRATCH_UMSK, 0);
+
+ mdelay(1);
+ WREG32(CP_RB0_CNTL, tmp);
+
+ WREG32(CP_RB0_BASE, ring->gpu_addr >> 8);
+
+ /* start the rings */
+ gfx_v6_0_cp_gfx_start(adev);
+ ring->ready = true;
+ r = amdgpu_ring_test_ring(ring);
+ if (r) {
+ ring->ready = false;
+ return r;
+ }
+
+ return 0;
+}
+
+static u32 gfx_v6_0_ring_get_rptr(struct amdgpu_ring *ring)
+{
+ return ring->adev->wb.wb[ring->rptr_offs];
+}
+
+static u32 gfx_v6_0_ring_get_wptr(struct amdgpu_ring *ring)
+{
+ struct amdgpu_device *adev = ring->adev;
+
+ if (ring == &adev->gfx.gfx_ring[0])
+ return RREG32(CP_RB0_WPTR);
+ else if (ring == &adev->gfx.compute_ring[0])
+ return RREG32(CP_RB1_WPTR);
+ else if (ring == &adev->gfx.compute_ring[1])
+ return RREG32(CP_RB2_WPTR);
+ else
+ BUG();
+}
+
+static void gfx_v6_0_ring_set_wptr_gfx(struct amdgpu_ring *ring)
+{
+ struct amdgpu_device *adev = ring->adev;
+
+ WREG32(CP_RB0_WPTR, ring->wptr);
+ (void)RREG32(CP_RB0_WPTR);
+}
+
+static void gfx_v6_0_ring_set_wptr_compute(struct amdgpu_ring *ring)
+{
+ struct amdgpu_device *adev = ring->adev;
+
+ if (ring == &adev->gfx.compute_ring[0]) {
+ WREG32(CP_RB1_WPTR, ring->wptr);
+ (void)RREG32(CP_RB1_WPTR);
+ } else if (ring == &adev->gfx.compute_ring[1]) {
+ WREG32(CP_RB2_WPTR, ring->wptr);
+ (void)RREG32(CP_RB2_WPTR);
+ } else {
+ BUG();
+ }
+
+}
+
+static int gfx_v6_0_cp_compute_resume(struct amdgpu_device *adev)
+{
+ struct amdgpu_ring *ring;
+ u32 tmp;
+ u32 rb_bufsz;
+ int r;
+ u64 rptr_addr;
+
+ /* ring1 - compute only */
+ /* Set ring buffer size */
+
+ ring = &adev->gfx.compute_ring[0];
+ rb_bufsz = order_base_2(ring->ring_size / 8);
+ tmp = (order_base_2(AMDGPU_GPU_PAGE_SIZE/8) << 8) | rb_bufsz;
+#ifdef __BIG_ENDIAN
+ tmp |= BUF_SWAP_32BIT;
+#endif
+ WREG32(CP_RB1_CNTL, tmp);
+
+ WREG32(CP_RB1_CNTL, tmp | RB_RPTR_WR_ENA);
+ ring->wptr = 0;
+ WREG32(CP_RB1_WPTR, ring->wptr);
+
+ rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4);
+ WREG32(CP_RB1_RPTR_ADDR, lower_32_bits(rptr_addr));
+ WREG32(CP_RB1_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF);
+
+ mdelay(1);
+ WREG32(CP_RB1_CNTL, tmp);
+ WREG32(CP_RB1_BASE, ring->gpu_addr >> 8);
+
+ ring = &adev->gfx.compute_ring[1];
+ rb_bufsz = order_base_2(ring->ring_size / 8);
+ tmp = (order_base_2(AMDGPU_GPU_PAGE_SIZE/8) << 8) | rb_bufsz;
+#ifdef __BIG_ENDIAN
+ tmp |= BUF_SWAP_32BIT;
+#endif
+ WREG32(CP_RB2_CNTL, tmp);
+
+ WREG32(CP_RB2_CNTL, tmp | RB_RPTR_WR_ENA);
+ ring->wptr = 0;
+ WREG32(CP_RB2_WPTR, ring->wptr);
+ rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4);
+ WREG32(CP_RB2_RPTR_ADDR, lower_32_bits(rptr_addr));
+ WREG32(CP_RB2_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF);
+
+ mdelay(1);
+ WREG32(CP_RB2_CNTL, tmp);
+ WREG32(CP_RB2_BASE, ring->gpu_addr >> 8);
+
+ adev->gfx.compute_ring[0].ready = true;
+ adev->gfx.compute_ring[1].ready = true;
+
+ r = amdgpu_ring_test_ring(&adev->gfx.compute_ring[0]);
+ if (r) {
+ adev->gfx.compute_ring[0].ready = false;
+ return r;
+ }
+
+ r = amdgpu_ring_test_ring(&adev->gfx.compute_ring[1]);
+ if (r) {
+ adev->gfx.compute_ring[1].ready = false;
+ return r;
+ }
+
+ return 0;
+}
+
+static void gfx_v6_0_cp_enable(struct amdgpu_device *adev, bool enable)
+{
+ gfx_v6_0_cp_gfx_enable(adev, enable);
+}
+
+static int gfx_v6_0_cp_load_microcode(struct amdgpu_device *adev)
+{
+ return gfx_v6_0_cp_gfx_load_microcode(adev);
+}
+
+static void gfx_v6_0_enable_gui_idle_interrupt(struct amdgpu_device *adev,
+ bool enable)
+{
+ u32 tmp = RREG32(CP_INT_CNTL_RING0);
+ u32 mask;
+ int i;
+
+ if (enable)
+ tmp |= (CNTX_BUSY_INT_ENABLE | CNTX_EMPTY_INT_ENABLE);
+ else
+ tmp &= ~(CNTX_BUSY_INT_ENABLE | CNTX_EMPTY_INT_ENABLE);
+ WREG32(CP_INT_CNTL_RING0, tmp);
+
+ if (!enable) {
+ /* read a gfx register */
+ tmp = RREG32(DB_DEPTH_INFO);
+
+ mask = RLC_BUSY_STATUS | GFX_POWER_STATUS | GFX_CLOCK_STATUS | GFX_LS_STATUS;
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if ((RREG32(RLC_STAT) & mask) == (GFX_CLOCK_STATUS | GFX_POWER_STATUS))
+ break;
+ udelay(1);
+ }
+ }
+}
+
+static int gfx_v6_0_cp_resume(struct amdgpu_device *adev)
+{
+ int r;
+
+ gfx_v6_0_enable_gui_idle_interrupt(adev, false);
+
+ r = gfx_v6_0_cp_load_microcode(adev);
+ if (r)
+ return r;
+
+ r = gfx_v6_0_cp_gfx_resume(adev);
+ if (r)
+ return r;
+ r = gfx_v6_0_cp_compute_resume(adev);
+ if (r)
+ return r;
+
+ gfx_v6_0_enable_gui_idle_interrupt(adev, true);
+
+ return 0;
+}
+
+static void gfx_v6_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
+{
+ int usepfp = (ring->type == AMDGPU_RING_TYPE_GFX);
+ uint32_t seq = ring->fence_drv.sync_seq;
+ uint64_t addr = ring->fence_drv.gpu_addr;
+
+ amdgpu_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5));
+ amdgpu_ring_write(ring, (WAIT_REG_MEM_MEM_SPACE(1) | /* memory */
+ WAIT_REG_MEM_FUNCTION(3) | /* equal */
+ WAIT_REG_MEM_ENGINE(usepfp))); /* pfp or me */
+ amdgpu_ring_write(ring, addr & 0xfffffffc);
+ amdgpu_ring_write(ring, upper_32_bits(addr) & 0xffffffff);
+ amdgpu_ring_write(ring, seq);
+ amdgpu_ring_write(ring, 0xffffffff);
+ amdgpu_ring_write(ring, 4); /* poll interval */
+
+ if (usepfp) {
+ /* synce CE with ME to prevent CE fetch CEIB before context switch done */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
+ amdgpu_ring_write(ring, 0);
+ }
+}
+
+static void gfx_v6_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
+ unsigned vm_id, uint64_t pd_addr)
+{
+ int usepfp = (ring->type == AMDGPU_RING_TYPE_GFX);
+
+ /* write new base address */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
+ amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) |
+ WRITE_DATA_DST_SEL(0)));
+ if (vm_id < 8) {
+ amdgpu_ring_write(ring, (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vm_id ));
+ } else {
+ amdgpu_ring_write(ring, (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (vm_id - 8)));
+ }
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, pd_addr >> 12);
+
+ /* bits 0-15 are the VM contexts0-15 */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
+ amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(1) |
+ WRITE_DATA_DST_SEL(0)));
+ amdgpu_ring_write(ring, VM_INVALIDATE_REQUEST);
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, 1 << vm_id);
+
+ /* wait for the invalidate to complete */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_WAIT_REG_MEM, 5));
+ amdgpu_ring_write(ring, (WAIT_REG_MEM_FUNCTION(0) | /* always */
+ WAIT_REG_MEM_ENGINE(0))); /* me */
+ amdgpu_ring_write(ring, VM_INVALIDATE_REQUEST);
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, 0); /* ref */
+ amdgpu_ring_write(ring, 0); /* mask */
+ amdgpu_ring_write(ring, 0x20); /* poll interval */
+
+ if (usepfp) {
+ /* sync PFP to ME, otherwise we might get invalid PFP reads */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0));
+ amdgpu_ring_write(ring, 0x0);
+
+ /* synce CE with ME to prevent CE fetch CEIB before context switch done */
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
+ amdgpu_ring_write(ring, 0);
+ }
+}
+
+
+static void gfx_v6_0_rlc_fini(struct amdgpu_device *adev)
+{
+ int r;
+
+ if (adev->gfx.rlc.save_restore_obj) {
+ r = amdgpu_bo_reserve(adev->gfx.rlc.save_restore_obj, false);
+ if (unlikely(r != 0))
+ dev_warn(adev->dev, "(%d) reserve RLC sr bo failed\n", r);
+ amdgpu_bo_unpin(adev->gfx.rlc.save_restore_obj);
+ amdgpu_bo_unreserve(adev->gfx.rlc.save_restore_obj);
+
+ amdgpu_bo_unref(&adev->gfx.rlc.save_restore_obj);
+ adev->gfx.rlc.save_restore_obj = NULL;
+ }
+
+ if (adev->gfx.rlc.clear_state_obj) {
+ r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false);
+ if (unlikely(r != 0))
+ dev_warn(adev->dev, "(%d) reserve RLC c bo failed\n", r);
+ amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj);
+ amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
+
+ amdgpu_bo_unref(&adev->gfx.rlc.clear_state_obj);
+ adev->gfx.rlc.clear_state_obj = NULL;
+ }
+
+ if (adev->gfx.rlc.cp_table_obj) {
+ r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, false);
+ if (unlikely(r != 0))
+ dev_warn(adev->dev, "(%d) reserve RLC cp table bo failed\n", r);
+ amdgpu_bo_unpin(adev->gfx.rlc.cp_table_obj);
+ amdgpu_bo_unreserve(adev->gfx.rlc.cp_table_obj);
+
+ amdgpu_bo_unref(&adev->gfx.rlc.cp_table_obj);
+ adev->gfx.rlc.cp_table_obj = NULL;
+ }
+}
+
+static int gfx_v6_0_rlc_init(struct amdgpu_device *adev)
+{
+ const u32 *src_ptr;
+ volatile u32 *dst_ptr;
+ u32 dws, i;
+ u64 reg_list_mc_addr;
+ const struct cs_section_def *cs_data;
+ int r;
+
+ adev->gfx.rlc.reg_list = verde_rlc_save_restore_register_list;
+ adev->gfx.rlc.reg_list_size =
+ (u32)ARRAY_SIZE(verde_rlc_save_restore_register_list);
+
+ adev->gfx.rlc.cs_data = si_cs_data;
+ src_ptr = adev->gfx.rlc.reg_list;
+ dws = adev->gfx.rlc.reg_list_size;
+ cs_data = adev->gfx.rlc.cs_data;
+
+ if (src_ptr) {
+ /* save restore block */
+ if (adev->gfx.rlc.save_restore_obj == NULL) {
+
+ r = amdgpu_bo_create(adev, dws * 4, PAGE_SIZE, true,
+ AMDGPU_GEM_DOMAIN_VRAM,
+ AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
+ NULL, NULL,
+ &adev->gfx.rlc.save_restore_obj);
+
+ if (r) {
+ dev_warn(adev->dev, "(%d) create RLC sr bo failed\n", r);
+ return r;
+ }
+ }
+
+ r = amdgpu_bo_reserve(adev->gfx.rlc.save_restore_obj, false);
+ if (unlikely(r != 0)) {
+ gfx_v6_0_rlc_fini(adev);
+ return r;
+ }
+ r = amdgpu_bo_pin(adev->gfx.rlc.save_restore_obj, AMDGPU_GEM_DOMAIN_VRAM,
+ &adev->gfx.rlc.save_restore_gpu_addr);
+ if (r) {
+ amdgpu_bo_unreserve(adev->gfx.rlc.save_restore_obj);
+ dev_warn(adev->dev, "(%d) pin RLC sr bo failed\n", r);
+ gfx_v6_0_rlc_fini(adev);
+ return r;
+ }
+
+ r = amdgpu_bo_kmap(adev->gfx.rlc.save_restore_obj, (void **)&adev->gfx.rlc.sr_ptr);
+ if (r) {
+ dev_warn(adev->dev, "(%d) map RLC sr bo failed\n", r);
+ gfx_v6_0_rlc_fini(adev);
+ return r;
+ }
+ /* write the sr buffer */
+ dst_ptr = adev->gfx.rlc.sr_ptr;
+ for (i = 0; i < adev->gfx.rlc.reg_list_size; i++)
+ dst_ptr[i] = cpu_to_le32(src_ptr[i]);
+ amdgpu_bo_kunmap(adev->gfx.rlc.save_restore_obj);
+ amdgpu_bo_unreserve(adev->gfx.rlc.save_restore_obj);
+ }
+
+ if (cs_data) {
+ /* clear state block */
+ adev->gfx.rlc.clear_state_size = gfx_v6_0_get_csb_size(adev);
+ dws = adev->gfx.rlc.clear_state_size + (256 / 4);
+
+ if (adev->gfx.rlc.clear_state_obj == NULL) {
+ r = amdgpu_bo_create(adev, dws * 4, PAGE_SIZE, true,
+ AMDGPU_GEM_DOMAIN_VRAM,
+ AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
+ NULL, NULL,
+ &adev->gfx.rlc.clear_state_obj);
+
+ if (r) {
+ dev_warn(adev->dev, "(%d) create RLC c bo failed\n", r);
+ gfx_v6_0_rlc_fini(adev);
+ return r;
+ }
+ }
+ r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false);
+ if (unlikely(r != 0)) {
+ gfx_v6_0_rlc_fini(adev);
+ return r;
+ }
+ r = amdgpu_bo_pin(adev->gfx.rlc.clear_state_obj, AMDGPU_GEM_DOMAIN_VRAM,
+ &adev->gfx.rlc.clear_state_gpu_addr);
+ if (r) {
+ amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
+ dev_warn(adev->dev, "(%d) pin RLC c bo failed\n", r);
+ gfx_v6_0_rlc_fini(adev);
+ return r;
+ }
+
+ r = amdgpu_bo_kmap(adev->gfx.rlc.clear_state_obj, (void **)&adev->gfx.rlc.cs_ptr);
+ if (r) {
+ dev_warn(adev->dev, "(%d) map RLC c bo failed\n", r);
+ gfx_v6_0_rlc_fini(adev);
+ return r;
+ }
+ /* set up the cs buffer */
+ dst_ptr = adev->gfx.rlc.cs_ptr;
+ reg_list_mc_addr = adev->gfx.rlc.clear_state_gpu_addr + 256;
+ dst_ptr[0] = cpu_to_le32(upper_32_bits(reg_list_mc_addr));
+ dst_ptr[1] = cpu_to_le32(lower_32_bits(reg_list_mc_addr));
+ dst_ptr[2] = cpu_to_le32(adev->gfx.rlc.clear_state_size);
+ gfx_v6_0_get_csb_buffer(adev, &dst_ptr[(256/4)]);
+ amdgpu_bo_kunmap(adev->gfx.rlc.clear_state_obj);
+ amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
+ }
+
+ return 0;
+}
+
+static void gfx_v6_0_enable_lbpw(struct amdgpu_device *adev, bool enable)
+{
+ u32 tmp;
+
+ tmp = RREG32(RLC_LB_CNTL);
+ if (enable)
+ tmp |= LOAD_BALANCE_ENABLE;
+ else
+ tmp &= ~LOAD_BALANCE_ENABLE;
+ WREG32(RLC_LB_CNTL, tmp);
+
+ if (!enable) {
+ gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+ WREG32(SPI_LB_CU_MASK, 0x00ff);
+ }
+
+}
+
+static void gfx_v6_0_wait_for_rlc_serdes(struct amdgpu_device *adev)
+{
+ int i;
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (RREG32(RLC_SERDES_MASTER_BUSY_0) == 0)
+ break;
+ udelay(1);
+ }
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (RREG32(RLC_SERDES_MASTER_BUSY_1) == 0)
+ break;
+ udelay(1);
+ }
+}
+
+static void gfx_v6_0_update_rlc(struct amdgpu_device *adev, u32 rlc)
+{
+ u32 tmp;
+
+ tmp = RREG32(RLC_CNTL);
+ if (tmp != rlc)
+ WREG32(RLC_CNTL, rlc);
+}
+
+static u32 gfx_v6_0_halt_rlc(struct amdgpu_device *adev)
+{
+ u32 data, orig;
+
+ orig = data = RREG32(RLC_CNTL);
+
+ if (data & RLC_ENABLE) {
+ data &= ~RLC_ENABLE;
+ WREG32(RLC_CNTL, data);
+
+ gfx_v6_0_wait_for_rlc_serdes(adev);
+ }
+
+ return orig;
+}
+
+static void gfx_v6_0_rlc_stop(struct amdgpu_device *adev)
+{
+ WREG32(RLC_CNTL, 0);
+
+ gfx_v6_0_enable_gui_idle_interrupt(adev, false);
+ gfx_v6_0_wait_for_rlc_serdes(adev);
+}
+
+static void gfx_v6_0_rlc_start(struct amdgpu_device *adev)
+{
+ WREG32(RLC_CNTL, RLC_ENABLE);
+
+ gfx_v6_0_enable_gui_idle_interrupt(adev, true);
+
+ udelay(50);
+}
+
+static void gfx_v6_0_rlc_reset(struct amdgpu_device *adev)
+{
+ u32 tmp = RREG32(GRBM_SOFT_RESET);
+
+ tmp |= SOFT_RESET_RLC;
+ WREG32(GRBM_SOFT_RESET, tmp);
+ udelay(50);
+ tmp &= ~SOFT_RESET_RLC;
+ WREG32(GRBM_SOFT_RESET, tmp);
+ udelay(50);
+}
+
+static bool gfx_v6_0_lbpw_supported(struct amdgpu_device *adev)
+{
+ u32 tmp;
+
+ /* Enable LBPW only for DDR3 */
+ tmp = RREG32(MC_SEQ_MISC0);
+ if ((tmp & 0xF0000000) == 0xB0000000)
+ return true;
+ return false;
+}
+static void gfx_v6_0_init_cg(struct amdgpu_device *adev)
+{
+}
+
+static int gfx_v6_0_rlc_resume(struct amdgpu_device *adev)
+{
+ u32 i;
+ const struct rlc_firmware_header_v1_0 *hdr;
+ const __le32 *fw_data;
+ u32 fw_size;
+
+
+ if (!adev->gfx.rlc_fw)
+ return -EINVAL;
+
+ gfx_v6_0_rlc_stop(adev);
+ gfx_v6_0_rlc_reset(adev);
+ gfx_v6_0_init_pg(adev);
+ gfx_v6_0_init_cg(adev);
+
+ WREG32(RLC_RL_BASE, 0);
+ WREG32(RLC_RL_SIZE, 0);
+ WREG32(RLC_LB_CNTL, 0);
+ WREG32(RLC_LB_CNTR_MAX, 0xffffffff);
+ WREG32(RLC_LB_CNTR_INIT, 0);
+ WREG32(RLC_LB_INIT_CU_MASK, 0xffffffff);
+
+ WREG32(RLC_MC_CNTL, 0);
+ WREG32(RLC_UCODE_CNTL, 0);
+
+ hdr = (const struct rlc_firmware_header_v1_0 *)adev->gfx.rlc_fw->data;
+ fw_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4;
+ fw_data = (const __le32 *)
+ (adev->gfx.rlc_fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes));
+
+ amdgpu_ucode_print_rlc_hdr(&hdr->header);
+
+ for (i = 0; i < fw_size; i++) {
+ WREG32(RLC_UCODE_ADDR, i);
+ WREG32(RLC_UCODE_DATA, le32_to_cpup(fw_data++));
+ }
+ WREG32(RLC_UCODE_ADDR, 0);
+
+ gfx_v6_0_enable_lbpw(adev, gfx_v6_0_lbpw_supported(adev));
+ gfx_v6_0_rlc_start(adev);
+
+ return 0;
+}
+
+static void gfx_v6_0_enable_cgcg(struct amdgpu_device *adev, bool enable)
+{
+ u32 data, orig, tmp;
+
+ orig = data = RREG32(RLC_CGCG_CGLS_CTRL);
+
+ if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGCG)) {
+ gfx_v6_0_enable_gui_idle_interrupt(adev, true);
+
+ WREG32(RLC_GCPM_GENERAL_3, 0x00000080);
+
+ tmp = gfx_v6_0_halt_rlc(adev);
+
+ WREG32(RLC_SERDES_WR_MASTER_MASK_0, 0xffffffff);
+ WREG32(RLC_SERDES_WR_MASTER_MASK_1, 0xffffffff);
+ WREG32(RLC_SERDES_WR_CTRL, 0x00b000ff);
+
+ gfx_v6_0_wait_for_rlc_serdes(adev);
+ gfx_v6_0_update_rlc(adev, tmp);
+
+ WREG32(RLC_SERDES_WR_CTRL, 0x007000ff);
+
+ data |= CGCG_EN | CGLS_EN;
+ } else {
+ gfx_v6_0_enable_gui_idle_interrupt(adev, false);
+
+ RREG32(CB_CGTT_SCLK_CTRL);
+ RREG32(CB_CGTT_SCLK_CTRL);
+ RREG32(CB_CGTT_SCLK_CTRL);
+ RREG32(CB_CGTT_SCLK_CTRL);
+
+ data &= ~(CGCG_EN | CGLS_EN);
+ }
+
+ if (orig != data)
+ WREG32(RLC_CGCG_CGLS_CTRL, data);
+
+}
+
+static void gfx_v6_0_enable_mgcg(struct amdgpu_device *adev, bool enable)
+{
+
+ u32 data, orig, tmp = 0;
+
+ if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_MGCG)) {
+ orig = data = RREG32(CGTS_SM_CTRL_REG);
+ data = 0x96940200;
+ if (orig != data)
+ WREG32(CGTS_SM_CTRL_REG, data);
+
+ if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CP_LS) {
+ orig = data = RREG32(CP_MEM_SLP_CNTL);
+ data |= CP_MEM_LS_EN;
+ if (orig != data)
+ WREG32(CP_MEM_SLP_CNTL, data);
+ }
+
+ orig = data = RREG32(RLC_CGTT_MGCG_OVERRIDE);
+ data &= 0xffffffc0;
+ if (orig != data)
+ WREG32(RLC_CGTT_MGCG_OVERRIDE, data);
+
+ tmp = gfx_v6_0_halt_rlc(adev);
+
+ WREG32(RLC_SERDES_WR_MASTER_MASK_0, 0xffffffff);
+ WREG32(RLC_SERDES_WR_MASTER_MASK_1, 0xffffffff);
+ WREG32(RLC_SERDES_WR_CTRL, 0x00d000ff);
+
+ gfx_v6_0_update_rlc(adev, tmp);
+ } else {
+ orig = data = RREG32(RLC_CGTT_MGCG_OVERRIDE);
+ data |= 0x00000003;
+ if (orig != data)
+ WREG32(RLC_CGTT_MGCG_OVERRIDE, data);
+
+ data = RREG32(CP_MEM_SLP_CNTL);
+ if (data & CP_MEM_LS_EN) {
+ data &= ~CP_MEM_LS_EN;
+ WREG32(CP_MEM_SLP_CNTL, data);
+ }
+ orig = data = RREG32(CGTS_SM_CTRL_REG);
+ data |= LS_OVERRIDE | OVERRIDE;
+ if (orig != data)
+ WREG32(CGTS_SM_CTRL_REG, data);
+
+ tmp = gfx_v6_0_halt_rlc(adev);
+
+ WREG32(RLC_SERDES_WR_MASTER_MASK_0, 0xffffffff);
+ WREG32(RLC_SERDES_WR_MASTER_MASK_1, 0xffffffff);
+ WREG32(RLC_SERDES_WR_CTRL, 0x00e000ff);
+
+ gfx_v6_0_update_rlc(adev, tmp);
+ }
+}
+/*
+static void gfx_v6_0_update_cg(struct amdgpu_device *adev,
+ bool enable)
+{
+ gfx_v6_0_enable_gui_idle_interrupt(adev, false);
+ if (enable) {
+ gfx_v6_0_enable_mgcg(adev, true);
+ gfx_v6_0_enable_cgcg(adev, true);
+ } else {
+ gfx_v6_0_enable_cgcg(adev, false);
+ gfx_v6_0_enable_mgcg(adev, false);
+ }
+ gfx_v6_0_enable_gui_idle_interrupt(adev, true);
+}
+*/
+static void gfx_v6_0_enable_sclk_slowdown_on_pu(struct amdgpu_device *adev,
+ bool enable)
+{
+}
+
+static void gfx_v6_0_enable_sclk_slowdown_on_pd(struct amdgpu_device *adev,
+ bool enable)
+{
+}
+
+static void gfx_v6_0_enable_cp_pg(struct amdgpu_device *adev, bool enable)
+{
+ u32 data, orig;
+
+ orig = data = RREG32(RLC_PG_CNTL);
+ if (enable && (adev->pg_flags & AMD_PG_SUPPORT_CP))
+ data &= ~0x8000;
+ else
+ data |= 0x8000;
+ if (orig != data)
+ WREG32(RLC_PG_CNTL, data);
+}
+
+static void gfx_v6_0_enable_gds_pg(struct amdgpu_device *adev, bool enable)
+{
+}
+/*
+static void gfx_v6_0_init_cp_pg_table(struct amdgpu_device *adev)
+{
+ const __le32 *fw_data;
+ volatile u32 *dst_ptr;
+ int me, i, max_me = 4;
+ u32 bo_offset = 0;
+ u32 table_offset, table_size;
+
+ if (adev->asic_type == CHIP_KAVERI)
+ max_me = 5;
+
+ if (adev->gfx.rlc.cp_table_ptr == NULL)
+ return;
+
+ dst_ptr = adev->gfx.rlc.cp_table_ptr;
+ for (me = 0; me < max_me; me++) {
+ if (me == 0) {
+ const struct gfx_firmware_header_v1_0 *hdr =
+ (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data;
+ fw_data = (const __le32 *)
+ (adev->gfx.ce_fw->data +
+ le32_to_cpu(hdr->header.ucode_array_offset_bytes));
+ table_offset = le32_to_cpu(hdr->jt_offset);
+ table_size = le32_to_cpu(hdr->jt_size);
+ } else if (me == 1) {
+ const struct gfx_firmware_header_v1_0 *hdr =
+ (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data;
+ fw_data = (const __le32 *)
+ (adev->gfx.pfp_fw->data +
+ le32_to_cpu(hdr->header.ucode_array_offset_bytes));
+ table_offset = le32_to_cpu(hdr->jt_offset);
+ table_size = le32_to_cpu(hdr->jt_size);
+ } else if (me == 2) {
+ const struct gfx_firmware_header_v1_0 *hdr =
+ (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data;
+ fw_data = (const __le32 *)
+ (adev->gfx.me_fw->data +
+ le32_to_cpu(hdr->header.ucode_array_offset_bytes));
+ table_offset = le32_to_cpu(hdr->jt_offset);
+ table_size = le32_to_cpu(hdr->jt_size);
+ } else if (me == 3) {
+ const struct gfx_firmware_header_v1_0 *hdr =
+ (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec_fw->data;
+ fw_data = (const __le32 *)
+ (adev->gfx.mec_fw->data +
+ le32_to_cpu(hdr->header.ucode_array_offset_bytes));
+ table_offset = le32_to_cpu(hdr->jt_offset);
+ table_size = le32_to_cpu(hdr->jt_size);
+ } else {
+ const struct gfx_firmware_header_v1_0 *hdr =
+ (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec2_fw->data;
+ fw_data = (const __le32 *)
+ (adev->gfx.mec2_fw->data +
+ le32_to_cpu(hdr->header.ucode_array_offset_bytes));
+ table_offset = le32_to_cpu(hdr->jt_offset);
+ table_size = le32_to_cpu(hdr->jt_size);
+ }
+
+ for (i = 0; i < table_size; i ++) {
+ dst_ptr[bo_offset + i] =
+ cpu_to_le32(le32_to_cpu(fw_data[table_offset + i]));
+ }
+
+ bo_offset += table_size;
+ }
+}
+*/
+static void gfx_v6_0_enable_gfx_cgpg(struct amdgpu_device *adev,
+ bool enable)
+{
+
+ u32 tmp;
+
+ if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)) {
+ tmp = RLC_PUD(0x10) | RLC_PDD(0x10) | RLC_TTPD(0x10) | RLC_MSD(0x10);
+ WREG32(RLC_TTOP_D, tmp);
+
+ tmp = RREG32(RLC_PG_CNTL);
+ tmp |= GFX_PG_ENABLE;
+ WREG32(RLC_PG_CNTL, tmp);
+
+ tmp = RREG32(RLC_AUTO_PG_CTRL);
+ tmp |= AUTO_PG_EN;
+ WREG32(RLC_AUTO_PG_CTRL, tmp);
+ } else {
+ tmp = RREG32(RLC_AUTO_PG_CTRL);
+ tmp &= ~AUTO_PG_EN;
+ WREG32(RLC_AUTO_PG_CTRL, tmp);
+
+ tmp = RREG32(DB_RENDER_CONTROL);
+ }
+}
+
+static u32 gfx_v6_0_get_cu_active_bitmap(struct amdgpu_device *adev,
+ u32 se, u32 sh)
+{
+
+ u32 mask = 0, tmp, tmp1;
+ int i;
+
+ mutex_lock(&adev->grbm_idx_mutex);
+ gfx_v6_0_select_se_sh(adev, se, sh, 0xffffffff);
+ tmp = RREG32(CC_GC_SHADER_ARRAY_CONFIG);
+ tmp1 = RREG32(GC_USER_SHADER_ARRAY_CONFIG);
+ gfx_v6_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+ mutex_unlock(&adev->grbm_idx_mutex);
+
+ tmp &= 0xffff0000;
+
+ tmp |= tmp1;
+ tmp >>= 16;
+
+ for (i = 0; i < adev->gfx.config.max_cu_per_sh; i ++) {
+ mask <<= 1;
+ mask |= 1;
+ }
+
+ return (~tmp) & mask;
+}
+
+static void gfx_v6_0_init_ao_cu_mask(struct amdgpu_device *adev)
+{
+ u32 i, j, k, active_cu_number = 0;
+
+ u32 mask, counter, cu_bitmap;
+ u32 tmp = 0;
+
+ for (i = 0; i < adev->gfx.config.max_shader_engines; i++) {
+ for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) {
+ mask = 1;
+ cu_bitmap = 0;
+ counter = 0;
+ for (k = 0; k < adev->gfx.config.max_cu_per_sh; k++) {
+ if (gfx_v6_0_get_cu_active_bitmap(adev, i, j) & mask) {
+ if (counter < 2)
+ cu_bitmap |= mask;
+ counter++;
+ }
+ mask <<= 1;
+ }
+
+ active_cu_number += counter;
+ tmp |= (cu_bitmap << (i * 16 + j * 8));
+ }
+ }
+
+ WREG32(RLC_PG_AO_CU_MASK, tmp);
+
+ tmp = RREG32(RLC_MAX_PG_CU);
+ tmp &= ~MAX_PU_CU_MASK;
+ tmp |= MAX_PU_CU(active_cu_number);
+ WREG32(RLC_MAX_PG_CU, tmp);
+}
+
+static void gfx_v6_0_enable_gfx_static_mgpg(struct amdgpu_device *adev,
+ bool enable)
+{
+ u32 data, orig;
+
+ orig = data = RREG32(RLC_PG_CNTL);
+ if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_SMG))
+ data |= STATIC_PER_CU_PG_ENABLE;
+ else
+ data &= ~STATIC_PER_CU_PG_ENABLE;
+ if (orig != data)
+ WREG32(RLC_PG_CNTL, data);
+}
+
+static void gfx_v6_0_enable_gfx_dynamic_mgpg(struct amdgpu_device *adev,
+ bool enable)
+{
+ u32 data, orig;
+
+ orig = data = RREG32(RLC_PG_CNTL);
+ if (enable && (adev->pg_flags & AMD_PG_SUPPORT_GFX_DMG))
+ data |= DYN_PER_CU_PG_ENABLE;
+ else
+ data &= ~DYN_PER_CU_PG_ENABLE;
+ if (orig != data)
+ WREG32(RLC_PG_CNTL, data);
+}
+
+static void gfx_v6_0_init_gfx_cgpg(struct amdgpu_device *adev)
+{
+ u32 tmp;
+
+ WREG32(RLC_SAVE_AND_RESTORE_BASE, adev->gfx.rlc.save_restore_gpu_addr >> 8);
+
+ tmp = RREG32(RLC_PG_CNTL);
+ tmp |= GFX_PG_SRC;
+ WREG32(RLC_PG_CNTL, tmp);
+
+ WREG32(RLC_CLEAR_STATE_RESTORE_BASE, adev->gfx.rlc.clear_state_gpu_addr >> 8);
+
+ tmp = RREG32(RLC_AUTO_PG_CTRL);
+
+ tmp &= ~GRBM_REG_SGIT_MASK;
+ tmp |= GRBM_REG_SGIT(0x700);
+ tmp &= ~PG_AFTER_GRBM_REG_ST_MASK;
+ WREG32(RLC_AUTO_PG_CTRL, tmp);
+}
+
+static void gfx_v6_0_update_gfx_pg(struct amdgpu_device *adev, bool enable)
+{
+ gfx_v6_0_enable_gfx_cgpg(adev, enable);
+ gfx_v6_0_enable_gfx_static_mgpg(adev, enable);
+ gfx_v6_0_enable_gfx_dynamic_mgpg(adev, enable);
+}
+
+static u32 gfx_v6_0_get_csb_size(struct amdgpu_device *adev)
+{
+ u32 count = 0;
+ const struct cs_section_def *sect = NULL;
+ const struct cs_extent_def *ext = NULL;
+
+ if (adev->gfx.rlc.cs_data == NULL)
+ return 0;
+
+ /* begin clear state */
+ count += 2;
+ /* context control state */
+ count += 3;
+
+ for (sect = adev->gfx.rlc.cs_data; sect->section != NULL; ++sect) {
+ for (ext = sect->section; ext->extent != NULL; ++ext) {
+ if (sect->id == SECT_CONTEXT)
+ count += 2 + ext->reg_count;
+ else
+ return 0;
+ }
+ }
+ /* pa_sc_raster_config */
+ count += 3;
+ /* end clear state */
+ count += 2;
+ /* clear state */
+ count += 2;
+
+ return count;
+}
+
+static void gfx_v6_0_get_csb_buffer(struct amdgpu_device *adev,
+ volatile u32 *buffer)
+{
+ u32 count = 0, i;
+ const struct cs_section_def *sect = NULL;
+ const struct cs_extent_def *ext = NULL;
+
+ if (adev->gfx.rlc.cs_data == NULL)
+ return;
+ if (buffer == NULL)
+ return;
+
+ buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0));
+ buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_BEGIN_CLEAR_STATE);
+
+ buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CONTEXT_CONTROL, 1));
+ buffer[count++] = cpu_to_le32(0x80000000);
+ buffer[count++] = cpu_to_le32(0x80000000);
+
+ for (sect = adev->gfx.rlc.cs_data; sect->section != NULL; ++sect) {
+ for (ext = sect->section; ext->extent != NULL; ++ext) {
+ if (sect->id == SECT_CONTEXT) {
+ buffer[count++] =
+ cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, ext->reg_count));
+ buffer[count++] = cpu_to_le32(ext->reg_index - 0xa000);
+ for (i = 0; i < ext->reg_count; i++)
+ buffer[count++] = cpu_to_le32(ext->extent[i]);
+ } else {
+ return;
+ }
+ }
+ }
+
+ buffer[count++] = cpu_to_le32(PACKET3(PACKET3_SET_CONTEXT_REG, 1));
+ buffer[count++] = cpu_to_le32(PA_SC_RASTER_CONFIG - PACKET3_SET_CONTEXT_REG_START);
+
+ switch (adev->asic_type) {
+ case CHIP_TAHITI:
+ case CHIP_PITCAIRN:
+ buffer[count++] = cpu_to_le32(0x2a00126a);
+ break;
+ case CHIP_VERDE:
+ buffer[count++] = cpu_to_le32(0x0000124a);
+ break;
+ case CHIP_OLAND:
+ buffer[count++] = cpu_to_le32(0x00000082);
+ break;
+ case CHIP_HAINAN:
+ buffer[count++] = cpu_to_le32(0x00000000);
+ break;
+ default:
+ buffer[count++] = cpu_to_le32(0x00000000);
+ break;
+ }
+
+ buffer[count++] = cpu_to_le32(PACKET3(PACKET3_PREAMBLE_CNTL, 0));
+ buffer[count++] = cpu_to_le32(PACKET3_PREAMBLE_END_CLEAR_STATE);
+
+ buffer[count++] = cpu_to_le32(PACKET3(PACKET3_CLEAR_STATE, 0));
+ buffer[count++] = cpu_to_le32(0);
+}
+
+static void gfx_v6_0_init_pg(struct amdgpu_device *adev)
+{
+ if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG |
+ AMD_PG_SUPPORT_GFX_SMG |
+ AMD_PG_SUPPORT_GFX_DMG |
+ AMD_PG_SUPPORT_CP |
+ AMD_PG_SUPPORT_GDS |
+ AMD_PG_SUPPORT_RLC_SMU_HS)) {
+ gfx_v6_0_enable_sclk_slowdown_on_pu(adev, true);
+ gfx_v6_0_enable_sclk_slowdown_on_pd(adev, true);
+ if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) {
+ gfx_v6_0_init_gfx_cgpg(adev);
+ gfx_v6_0_enable_cp_pg(adev, true);
+ gfx_v6_0_enable_gds_pg(adev, true);
+ } else {
+ WREG32(RLC_SAVE_AND_RESTORE_BASE, adev->gfx.rlc.save_restore_gpu_addr >> 8);
+ WREG32(RLC_CLEAR_STATE_RESTORE_BASE, adev->gfx.rlc.clear_state_gpu_addr >> 8);
+
+ }
+ gfx_v6_0_init_ao_cu_mask(adev);
+ gfx_v6_0_update_gfx_pg(adev, true);
+ } else {
+
+ WREG32(RLC_SAVE_AND_RESTORE_BASE, adev->gfx.rlc.save_restore_gpu_addr >> 8);
+ WREG32(RLC_CLEAR_STATE_RESTORE_BASE, adev->gfx.rlc.clear_state_gpu_addr >> 8);
+ }
+}
+
+static void gfx_v6_0_fini_pg(struct amdgpu_device *adev)
+{
+ if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG |
+ AMD_PG_SUPPORT_GFX_SMG |
+ AMD_PG_SUPPORT_GFX_DMG |
+ AMD_PG_SUPPORT_CP |
+ AMD_PG_SUPPORT_GDS |
+ AMD_PG_SUPPORT_RLC_SMU_HS)) {
+ gfx_v6_0_update_gfx_pg(adev, false);
+ if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) {
+ gfx_v6_0_enable_cp_pg(adev, false);
+ gfx_v6_0_enable_gds_pg(adev, false);
+ }
+ }
+}
+
+static uint64_t gfx_v6_0_get_gpu_clock_counter(struct amdgpu_device *adev)
+{
+ uint64_t clock;
+
+ mutex_lock(&adev->gfx.gpu_clock_mutex);
+ WREG32(RLC_CAPTURE_GPU_CLOCK_COUNT, 1);
+ clock = (uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_LSB) |
+ ((uint64_t)RREG32(RLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
+ mutex_unlock(&adev->gfx.gpu_clock_mutex);
+ return clock;
+}
+
+static void gfx_v6_ring_emit_cntxcntl(struct amdgpu_ring *ring, uint32_t flags)
+{
+ amdgpu_ring_write(ring, PACKET3(PACKET3_CONTEXT_CONTROL, 1));
+ amdgpu_ring_write(ring, 0x80000000);
+ amdgpu_ring_write(ring, 0);
+}
+
+static unsigned gfx_v6_0_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 6; /* gfx_v6_0_ring_emit_ib */
+}
+
+static unsigned gfx_v6_0_ring_get_dma_frame_size_gfx(struct amdgpu_ring *ring)
+{
+ return
+ 5 + /* gfx_v6_0_ring_emit_hdp_flush */
+ 5 + /* gfx_v6_0_ring_emit_hdp_invalidate */
+ 14 + 14 + 14 + /* gfx_v6_0_ring_emit_fence x3 for user fence, vm fence */
+ 7 + 4 + /* gfx_v6_0_ring_emit_pipeline_sync */
+ 17 + 6 + /* gfx_v6_0_ring_emit_vm_flush */
+ 3; /* gfx_v6_ring_emit_cntxcntl */
+}
+
+static unsigned gfx_v6_0_ring_get_dma_frame_size_compute(struct amdgpu_ring *ring)
+{
+ return
+ 5 + /* gfx_v6_0_ring_emit_hdp_flush */
+ 5 + /* gfx_v6_0_ring_emit_hdp_invalidate */
+ 7 + /* gfx_v6_0_ring_emit_pipeline_sync */
+ 17 + /* gfx_v6_0_ring_emit_vm_flush */
+ 14 + 14 + 14; /* gfx_v6_0_ring_emit_fence x3 for user fence, vm fence */
+}
+
+static const struct amdgpu_gfx_funcs gfx_v6_0_gfx_funcs = {
+ .get_gpu_clock_counter = &gfx_v6_0_get_gpu_clock_counter,
+ .select_se_sh = &gfx_v6_0_select_se_sh,
+};
+
+static int gfx_v6_0_early_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ adev->gfx.num_gfx_rings = GFX6_NUM_GFX_RINGS;
+ adev->gfx.num_compute_rings = GFX6_NUM_COMPUTE_RINGS;
+ adev->gfx.funcs = &gfx_v6_0_gfx_funcs;
+ gfx_v6_0_set_ring_funcs(adev);
+ gfx_v6_0_set_irq_funcs(adev);
+
+ return 0;
+}
+
+static int gfx_v6_0_sw_init(void *handle)
+{
+ struct amdgpu_ring *ring;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ int i, r;
+
+ r = amdgpu_irq_add_id(adev, 181, &adev->gfx.eop_irq);
+ if (r)
+ return r;
+
+ r = amdgpu_irq_add_id(adev, 184, &adev->gfx.priv_reg_irq);
+ if (r)
+ return r;
+
+ r = amdgpu_irq_add_id(adev, 185, &adev->gfx.priv_inst_irq);
+ if (r)
+ return r;
+
+ gfx_v6_0_scratch_init(adev);
+
+ r = gfx_v6_0_init_microcode(adev);
+ if (r) {
+ DRM_ERROR("Failed to load gfx firmware!\n");
+ return r;
+ }
+
+ r = gfx_v6_0_rlc_init(adev);
+ if (r) {
+ DRM_ERROR("Failed to init rlc BOs!\n");
+ return r;
+ }
+
+ for (i = 0; i < adev->gfx.num_gfx_rings; i++) {
+ ring = &adev->gfx.gfx_ring[i];
+ ring->ring_obj = NULL;
+ sprintf(ring->name, "gfx");
+ r = amdgpu_ring_init(adev, ring, 1024,
+ 0x80000000, 0xf,
+ &adev->gfx.eop_irq, AMDGPU_CP_IRQ_GFX_EOP,
+ AMDGPU_RING_TYPE_GFX);
+ if (r)
+ return r;
+ }
+
+ for (i = 0; i < adev->gfx.num_compute_rings; i++) {
+ unsigned irq_type;
+
+ if ((i >= 32) || (i >= AMDGPU_MAX_COMPUTE_RINGS)) {
+ DRM_ERROR("Too many (%d) compute rings!\n", i);
+ break;
+ }
+ ring = &adev->gfx.compute_ring[i];
+ ring->ring_obj = NULL;
+ ring->use_doorbell = false;
+ ring->doorbell_index = 0;
+ ring->me = 1;
+ ring->pipe = i;
+ ring->queue = i;
+ sprintf(ring->name, "comp %d.%d.%d", ring->me, ring->pipe, ring->queue);
+ irq_type = AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE0_EOP + ring->pipe;
+ r = amdgpu_ring_init(adev, ring, 1024,
+ 0x80000000, 0xf,
+ &adev->gfx.eop_irq, irq_type,
+ AMDGPU_RING_TYPE_COMPUTE);
+ if (r)
+ return r;
+ }
+
+ return r;
+}
+
+static int gfx_v6_0_sw_fini(void *handle)
+{
+ int i;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ amdgpu_bo_unref(&adev->gds.oa_gfx_bo);
+ amdgpu_bo_unref(&adev->gds.gws_gfx_bo);
+ amdgpu_bo_unref(&adev->gds.gds_gfx_bo);
+
+ for (i = 0; i < adev->gfx.num_gfx_rings; i++)
+ amdgpu_ring_fini(&adev->gfx.gfx_ring[i]);
+ for (i = 0; i < adev->gfx.num_compute_rings; i++)
+ amdgpu_ring_fini(&adev->gfx.compute_ring[i]);
+
+ gfx_v6_0_rlc_fini(adev);
+
+ return 0;
+}
+
+static int gfx_v6_0_hw_init(void *handle)
+{
+ int r;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ gfx_v6_0_gpu_init(adev);
+
+ r = gfx_v6_0_rlc_resume(adev);
+ if (r)
+ return r;
+
+ r = gfx_v6_0_cp_resume(adev);
+ if (r)
+ return r;
+
+ adev->gfx.ce_ram_size = 0x8000;
+
+ return r;
+}
+
+static int gfx_v6_0_hw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ gfx_v6_0_cp_enable(adev, false);
+ gfx_v6_0_rlc_stop(adev);
+ gfx_v6_0_fini_pg(adev);
+
+ return 0;
+}
+
+static int gfx_v6_0_suspend(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return gfx_v6_0_hw_fini(adev);
+}
+
+static int gfx_v6_0_resume(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return gfx_v6_0_hw_init(adev);
+}
+
+static bool gfx_v6_0_is_idle(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (RREG32(GRBM_STATUS) & GRBM_STATUS__GUI_ACTIVE_MASK)
+ return false;
+ else
+ return true;
+}
+
+static int gfx_v6_0_wait_for_idle(void *handle)
+{
+ unsigned i;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (gfx_v6_0_is_idle(handle))
+ return 0;
+ udelay(1);
+ }
+ return -ETIMEDOUT;
+}
+
+static int gfx_v6_0_soft_reset(void *handle)
+{
+ return 0;
+}
+
+static void gfx_v6_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
+ enum amdgpu_interrupt_state state)
+{
+ u32 cp_int_cntl;
+
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING0);
+ cp_int_cntl &= ~CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING0, cp_int_cntl);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING0);
+ cp_int_cntl |= CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING0, cp_int_cntl);
+ break;
+ default:
+ break;
+ }
+}
+
+static void gfx_v6_0_set_compute_eop_interrupt_state(struct amdgpu_device *adev,
+ int ring,
+ enum amdgpu_interrupt_state state)
+{
+ u32 cp_int_cntl;
+ switch (state){
+ case AMDGPU_IRQ_STATE_DISABLE:
+ if (ring == 0) {
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING1);
+ cp_int_cntl &= ~CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING1, cp_int_cntl);
+ break;
+ } else {
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING2);
+ cp_int_cntl &= ~CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING2, cp_int_cntl);
+ break;
+
+ }
+ case AMDGPU_IRQ_STATE_ENABLE:
+ if (ring == 0) {
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING1);
+ cp_int_cntl |= CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING1, cp_int_cntl);
+ break;
+ } else {
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING2);
+ cp_int_cntl |= CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING2, cp_int_cntl);
+ break;
+
+ }
+
+ default:
+ BUG();
+ break;
+
+ }
+}
+
+static int gfx_v6_0_set_priv_reg_fault_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ u32 cp_int_cntl;
+
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING0);
+ cp_int_cntl &= ~CP_INT_CNTL_RING0__PRIV_REG_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING0, cp_int_cntl);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING0);
+ cp_int_cntl |= CP_INT_CNTL_RING0__PRIV_REG_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING0, cp_int_cntl);
+ break;
+ default:
+ break;
+ }
+
+ return 0;
+}
+
+static int gfx_v6_0_set_priv_inst_fault_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ u32 cp_int_cntl;
+
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING0);
+ cp_int_cntl &= ~CP_INT_CNTL_RING0__PRIV_INSTR_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING0, cp_int_cntl);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ cp_int_cntl = RREG32(CP_INT_CNTL_RING0);
+ cp_int_cntl |= CP_INT_CNTL_RING0__PRIV_INSTR_INT_ENABLE_MASK;
+ WREG32(CP_INT_CNTL_RING0, cp_int_cntl);
+ break;
+ default:
+ break;
+ }
+
+ return 0;
+}
+
+static int gfx_v6_0_set_eop_interrupt_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ switch (type) {
+ case AMDGPU_CP_IRQ_GFX_EOP:
+ gfx_v6_0_set_gfx_eop_interrupt_state(adev, state);
+ break;
+ case AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE0_EOP:
+ gfx_v6_0_set_compute_eop_interrupt_state(adev, 0, state);
+ break;
+ case AMDGPU_CP_IRQ_COMPUTE_MEC1_PIPE1_EOP:
+ gfx_v6_0_set_compute_eop_interrupt_state(adev, 1, state);
+ break;
+ default:
+ break;
+ }
+ return 0;
+}
+
+static int gfx_v6_0_eop_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ switch (entry->ring_id) {
+ case 0:
+ amdgpu_fence_process(&adev->gfx.gfx_ring[0]);
+ break;
+ case 1:
+ case 2:
+ amdgpu_fence_process(&adev->gfx.compute_ring[entry->ring_id -1]);
+ break;
+ default:
+ break;
+ }
+ return 0;
+}
+
+static int gfx_v6_0_priv_reg_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ DRM_ERROR("Illegal register access in command stream\n");
+ schedule_work(&adev->reset_work);
+ return 0;
+}
+
+static int gfx_v6_0_priv_inst_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ DRM_ERROR("Illegal instruction in command stream\n");
+ schedule_work(&adev->reset_work);
+ return 0;
+}
+
+static int gfx_v6_0_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ bool gate = false;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (state == AMD_CG_STATE_GATE)
+ gate = true;
+
+ gfx_v6_0_enable_gui_idle_interrupt(adev, false);
+ if (gate) {
+ gfx_v6_0_enable_mgcg(adev, true);
+ gfx_v6_0_enable_cgcg(adev, true);
+ } else {
+ gfx_v6_0_enable_cgcg(adev, false);
+ gfx_v6_0_enable_mgcg(adev, false);
+ }
+ gfx_v6_0_enable_gui_idle_interrupt(adev, true);
+
+ return 0;
+}
+
+static int gfx_v6_0_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ bool gate = false;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (state == AMD_PG_STATE_GATE)
+ gate = true;
+
+ if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG |
+ AMD_PG_SUPPORT_GFX_SMG |
+ AMD_PG_SUPPORT_GFX_DMG |
+ AMD_PG_SUPPORT_CP |
+ AMD_PG_SUPPORT_GDS |
+ AMD_PG_SUPPORT_RLC_SMU_HS)) {
+ gfx_v6_0_update_gfx_pg(adev, gate);
+ if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) {
+ gfx_v6_0_enable_cp_pg(adev, gate);
+ gfx_v6_0_enable_gds_pg(adev, gate);
+ }
+ }
+
+ return 0;
+}
+
+const struct amd_ip_funcs gfx_v6_0_ip_funcs = {
+ .name = "gfx_v6_0",
+ .early_init = gfx_v6_0_early_init,
+ .late_init = NULL,
+ .sw_init = gfx_v6_0_sw_init,
+ .sw_fini = gfx_v6_0_sw_fini,
+ .hw_init = gfx_v6_0_hw_init,
+ .hw_fini = gfx_v6_0_hw_fini,
+ .suspend = gfx_v6_0_suspend,
+ .resume = gfx_v6_0_resume,
+ .is_idle = gfx_v6_0_is_idle,
+ .wait_for_idle = gfx_v6_0_wait_for_idle,
+ .soft_reset = gfx_v6_0_soft_reset,
+ .set_clockgating_state = gfx_v6_0_set_clockgating_state,
+ .set_powergating_state = gfx_v6_0_set_powergating_state,
+};
+
+static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_gfx = {
+ .get_rptr = gfx_v6_0_ring_get_rptr,
+ .get_wptr = gfx_v6_0_ring_get_wptr,
+ .set_wptr = gfx_v6_0_ring_set_wptr_gfx,
+ .parse_cs = NULL,
+ .emit_ib = gfx_v6_0_ring_emit_ib,
+ .emit_fence = gfx_v6_0_ring_emit_fence,
+ .emit_pipeline_sync = gfx_v6_0_ring_emit_pipeline_sync,
+ .emit_vm_flush = gfx_v6_0_ring_emit_vm_flush,
+ .emit_hdp_flush = gfx_v6_0_ring_emit_hdp_flush,
+ .emit_hdp_invalidate = gfx_v6_0_ring_emit_hdp_invalidate,
+ .test_ring = gfx_v6_0_ring_test_ring,
+ .test_ib = gfx_v6_0_ring_test_ib,
+ .insert_nop = amdgpu_ring_insert_nop,
+ .emit_cntxcntl = gfx_v6_ring_emit_cntxcntl,
+ .get_emit_ib_size = gfx_v6_0_ring_get_emit_ib_size,
+ .get_dma_frame_size = gfx_v6_0_ring_get_dma_frame_size_gfx,
+};
+
+static const struct amdgpu_ring_funcs gfx_v6_0_ring_funcs_compute = {
+ .get_rptr = gfx_v6_0_ring_get_rptr,
+ .get_wptr = gfx_v6_0_ring_get_wptr,
+ .set_wptr = gfx_v6_0_ring_set_wptr_compute,
+ .parse_cs = NULL,
+ .emit_ib = gfx_v6_0_ring_emit_ib,
+ .emit_fence = gfx_v6_0_ring_emit_fence,
+ .emit_pipeline_sync = gfx_v6_0_ring_emit_pipeline_sync,
+ .emit_vm_flush = gfx_v6_0_ring_emit_vm_flush,
+ .emit_hdp_flush = gfx_v6_0_ring_emit_hdp_flush,
+ .emit_hdp_invalidate = gfx_v6_0_ring_emit_hdp_invalidate,
+ .test_ring = gfx_v6_0_ring_test_ring,
+ .test_ib = gfx_v6_0_ring_test_ib,
+ .insert_nop = amdgpu_ring_insert_nop,
+ .get_emit_ib_size = gfx_v6_0_ring_get_emit_ib_size,
+ .get_dma_frame_size = gfx_v6_0_ring_get_dma_frame_size_compute,
+};
+
+static void gfx_v6_0_set_ring_funcs(struct amdgpu_device *adev)
+{
+ int i;
+
+ for (i = 0; i < adev->gfx.num_gfx_rings; i++)
+ adev->gfx.gfx_ring[i].funcs = &gfx_v6_0_ring_funcs_gfx;
+ for (i = 0; i < adev->gfx.num_compute_rings; i++)
+ adev->gfx.compute_ring[i].funcs = &gfx_v6_0_ring_funcs_compute;
+}
+
+static const struct amdgpu_irq_src_funcs gfx_v6_0_eop_irq_funcs = {
+ .set = gfx_v6_0_set_eop_interrupt_state,
+ .process = gfx_v6_0_eop_irq,
+};
+
+static const struct amdgpu_irq_src_funcs gfx_v6_0_priv_reg_irq_funcs = {
+ .set = gfx_v6_0_set_priv_reg_fault_state,
+ .process = gfx_v6_0_priv_reg_irq,
+};
+
+static const struct amdgpu_irq_src_funcs gfx_v6_0_priv_inst_irq_funcs = {
+ .set = gfx_v6_0_set_priv_inst_fault_state,
+ .process = gfx_v6_0_priv_inst_irq,
+};
+
+static void gfx_v6_0_set_irq_funcs(struct amdgpu_device *adev)
+{
+ adev->gfx.eop_irq.num_types = AMDGPU_CP_IRQ_LAST;
+ adev->gfx.eop_irq.funcs = &gfx_v6_0_eop_irq_funcs;
+
+ adev->gfx.priv_reg_irq.num_types = 1;
+ adev->gfx.priv_reg_irq.funcs = &gfx_v6_0_priv_reg_irq_funcs;
+
+ adev->gfx.priv_inst_irq.num_types = 1;
+ adev->gfx.priv_inst_irq.funcs = &gfx_v6_0_priv_inst_irq_funcs;
+}
+
+static void gfx_v6_0_get_cu_info(struct amdgpu_device *adev)
+{
+ int i, j, k, counter, active_cu_number = 0;
+ u32 mask, bitmap, ao_bitmap, ao_cu_mask = 0;
+ struct amdgpu_cu_info *cu_info = &adev->gfx.cu_info;
+
+ memset(cu_info, 0, sizeof(*cu_info));
+
+ for (i = 0; i < adev->gfx.config.max_shader_engines; i++) {
+ for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) {
+ mask = 1;
+ ao_bitmap = 0;
+ counter = 0;
+ bitmap = gfx_v6_0_get_cu_active_bitmap(adev, i, j);
+ cu_info->bitmap[i][j] = bitmap;
+
+ for (k = 0; k < adev->gfx.config.max_cu_per_sh; k ++) {
+ if (bitmap & mask) {
+ if (counter < 2)
+ ao_bitmap |= mask;
+ counter ++;
+ }
+ mask <<= 1;
+ }
+ active_cu_number += counter;
+ ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8));
+ }
+ }
+
+ cu_info->number = active_cu_number;
+ cu_info->ao_cu_mask = ao_cu_mask;
+}
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __GFX_V6_0_H__
+#define __GFX_V6_0_H__
+
+extern const struct amd_ip_funcs gfx_v6_0_ip_funcs;
+
+#endif
amdgpu_ring_write(ring, control);
}
+static void gfx_v7_ring_emit_cntxcntl(struct amdgpu_ring *ring, uint32_t flags)
+{
+ uint32_t dw2 = 0;
+
+ dw2 |= 0x80000000; /* set load_enable otherwise this package is just NOPs */
+ if (flags & AMDGPU_HAVE_CTX_SWITCH) {
+ /* set load_global_config & load_global_uconfig */
+ dw2 |= 0x8001;
+ /* set load_cs_sh_regs */
+ dw2 |= 0x01000000;
+ /* set load_per_context_state & load_gfx_sh_regs */
+ dw2 |= 0x10002;
+ }
+
+ amdgpu_ring_write(ring, PACKET3(PACKET3_CONTEXT_CONTROL, 1));
+ amdgpu_ring_write(ring, dw2);
+ amdgpu_ring_write(ring, 0);
+}
+
/**
* gfx_v7_0_ring_test_ib - basic ring IB test
*
return 0;
}
-static u32 gfx_v7_0_ring_get_rptr_gfx(struct amdgpu_ring *ring)
+static u32 gfx_v7_0_ring_get_rptr(struct amdgpu_ring *ring)
{
return ring->adev->wb.wb[ring->rptr_offs];
}
(void)RREG32(mmCP_RB0_WPTR);
}
-static u32 gfx_v7_0_ring_get_rptr_compute(struct amdgpu_ring *ring)
-{
- return ring->adev->wb.wb[ring->rptr_offs];
-}
-
static u32 gfx_v7_0_ring_get_wptr_compute(struct amdgpu_ring *ring)
{
/* XXX check if swapping is necessary on BE */
amdgpu_ring_write(ring, (1 << (oa_size + oa_base)) - (1 << oa_base));
}
+static unsigned gfx_v7_0_ring_get_emit_ib_size_gfx(struct amdgpu_ring *ring)
+{
+ return
+ 4; /* gfx_v7_0_ring_emit_ib_gfx */
+}
+
+static unsigned gfx_v7_0_ring_get_dma_frame_size_gfx(struct amdgpu_ring *ring)
+{
+ return
+ 20 + /* gfx_v7_0_ring_emit_gds_switch */
+ 7 + /* gfx_v7_0_ring_emit_hdp_flush */
+ 5 + /* gfx_v7_0_ring_emit_hdp_invalidate */
+ 12 + 12 + 12 + /* gfx_v7_0_ring_emit_fence_gfx x3 for user fence, vm fence */
+ 7 + 4 + /* gfx_v7_0_ring_emit_pipeline_sync */
+ 17 + 6 + /* gfx_v7_0_ring_emit_vm_flush */
+ 3; /* gfx_v7_ring_emit_cntxcntl */
+}
+
+static unsigned gfx_v7_0_ring_get_emit_ib_size_compute(struct amdgpu_ring *ring)
+{
+ return
+ 4; /* gfx_v7_0_ring_emit_ib_compute */
+}
+
+static unsigned gfx_v7_0_ring_get_dma_frame_size_compute(struct amdgpu_ring *ring)
+{
+ return
+ 20 + /* gfx_v7_0_ring_emit_gds_switch */
+ 7 + /* gfx_v7_0_ring_emit_hdp_flush */
+ 5 + /* gfx_v7_0_ring_emit_hdp_invalidate */
+ 7 + /* gfx_v7_0_ring_emit_pipeline_sync */
+ 17 + /* gfx_v7_0_ring_emit_vm_flush */
+ 7 + 7 + 7; /* gfx_v7_0_ring_emit_fence_compute x3 for user fence, vm fence */
+}
+
static const struct amdgpu_gfx_funcs gfx_v7_0_gfx_funcs = {
.get_gpu_clock_counter = &gfx_v7_0_get_gpu_clock_counter,
.select_se_sh = &gfx_v7_0_select_se_sh,
}
/* reserve GDS, GWS and OA resource for gfx */
- r = amdgpu_bo_create(adev, adev->gds.mem.gfx_partition_size,
- PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_GDS, 0,
- NULL, NULL, &adev->gds.gds_gfx_bo);
+ r = amdgpu_bo_create_kernel(adev, adev->gds.mem.gfx_partition_size,
+ PAGE_SIZE, AMDGPU_GEM_DOMAIN_GDS,
+ &adev->gds.gds_gfx_bo, NULL, NULL);
if (r)
return r;
- r = amdgpu_bo_create(adev, adev->gds.gws.gfx_partition_size,
- PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_GWS, 0,
- NULL, NULL, &adev->gds.gws_gfx_bo);
+ r = amdgpu_bo_create_kernel(adev, adev->gds.gws.gfx_partition_size,
+ PAGE_SIZE, AMDGPU_GEM_DOMAIN_GWS,
+ &adev->gds.gws_gfx_bo, NULL, NULL);
if (r)
return r;
- r = amdgpu_bo_create(adev, adev->gds.oa.gfx_partition_size,
- PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_OA, 0,
- NULL, NULL, &adev->gds.oa_gfx_bo);
+ r = amdgpu_bo_create_kernel(adev, adev->gds.oa.gfx_partition_size,
+ PAGE_SIZE, AMDGPU_GEM_DOMAIN_OA,
+ &adev->gds.oa_gfx_bo, NULL, NULL);
if (r)
return r;
int i;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
- amdgpu_bo_unref(&adev->gds.oa_gfx_bo);
- amdgpu_bo_unref(&adev->gds.gws_gfx_bo);
- amdgpu_bo_unref(&adev->gds.gds_gfx_bo);
+ amdgpu_bo_free_kernel(&adev->gds.oa_gfx_bo, NULL, NULL);
+ amdgpu_bo_free_kernel(&adev->gds.gws_gfx_bo, NULL, NULL);
+ amdgpu_bo_free_kernel(&adev->gds.gds_gfx_bo, NULL, NULL);
for (i = 0; i < adev->gfx.num_gfx_rings; i++)
amdgpu_ring_fini(&adev->gfx.gfx_ring[i]);
};
static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_gfx = {
- .get_rptr = gfx_v7_0_ring_get_rptr_gfx,
+ .get_rptr = gfx_v7_0_ring_get_rptr,
.get_wptr = gfx_v7_0_ring_get_wptr_gfx,
.set_wptr = gfx_v7_0_ring_set_wptr_gfx,
.parse_cs = NULL,
.test_ib = gfx_v7_0_ring_test_ib,
.insert_nop = amdgpu_ring_insert_nop,
.pad_ib = amdgpu_ring_generic_pad_ib,
+ .emit_cntxcntl = gfx_v7_ring_emit_cntxcntl,
+ .get_emit_ib_size = gfx_v7_0_ring_get_emit_ib_size_gfx,
+ .get_dma_frame_size = gfx_v7_0_ring_get_dma_frame_size_gfx,
};
static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = {
- .get_rptr = gfx_v7_0_ring_get_rptr_compute,
+ .get_rptr = gfx_v7_0_ring_get_rptr,
.get_wptr = gfx_v7_0_ring_get_wptr_compute,
.set_wptr = gfx_v7_0_ring_set_wptr_compute,
.parse_cs = NULL,
.test_ib = gfx_v7_0_ring_test_ib,
.insert_nop = amdgpu_ring_insert_nop,
.pad_ib = amdgpu_ring_generic_pad_ib,
+ .get_emit_ib_size = gfx_v7_0_ring_get_emit_ib_size_compute,
+ .get_dma_frame_size = gfx_v7_0_ring_get_dma_frame_size_compute,
};
static void gfx_v7_0_set_ring_funcs(struct amdgpu_device *adev)
polaris10_golden_common_all,
(const u32)ARRAY_SIZE(polaris10_golden_common_all));
WREG32_SMC(ixCG_ACLK_CNTL, 0x0000001C);
- if (adev->pdev->revision == 0xc7) {
+ if (adev->pdev->revision == 0xc7 &&
+ ((adev->pdev->subsystem_device == 0xb37 && adev->pdev->subsystem_vendor == 0x1002) ||
+ (adev->pdev->subsystem_device == 0x4a8 && adev->pdev->subsystem_vendor == 0x1043) ||
+ (adev->pdev->subsystem_device == 0x9480 && adev->pdev->subsystem_vendor == 0x1682))) {
amdgpu_atombios_i2c_channel_trans(adev, 0x10, 0x96, 0x1E, 0xDD);
amdgpu_atombios_i2c_channel_trans(adev, 0x10, 0x96, 0x1F, 0xD0);
}
if (adev->gfx.rlc.clear_state_obj) {
r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false);
if (unlikely(r != 0))
- dev_warn(adev->dev, "(%d) reserve RLC c bo failed\n", r);
+ dev_warn(adev->dev, "(%d) reserve RLC cbs bo failed\n", r);
amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj);
amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
-
amdgpu_bo_unref(&adev->gfx.rlc.clear_state_obj);
adev->gfx.rlc.clear_state_obj = NULL;
}
dev_warn(adev->dev, "(%d) reserve RLC cp table bo failed\n", r);
amdgpu_bo_unpin(adev->gfx.rlc.cp_table_obj);
amdgpu_bo_unreserve(adev->gfx.rlc.cp_table_obj);
-
amdgpu_bo_unref(&adev->gfx.rlc.cp_table_obj);
adev->gfx.rlc.cp_table_obj = NULL;
}
&adev->gfx.rlc.clear_state_gpu_addr);
if (r) {
amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
- dev_warn(adev->dev, "(%d) pin RLC c bo failed\n", r);
+ dev_warn(adev->dev, "(%d) pin RLC cbs bo failed\n", r);
gfx_v8_0_rlc_fini(adev);
return r;
}
r = amdgpu_bo_kmap(adev->gfx.rlc.clear_state_obj, (void **)&adev->gfx.rlc.cs_ptr);
if (r) {
- dev_warn(adev->dev, "(%d) map RLC c bo failed\n", r);
+ dev_warn(adev->dev, "(%d) map RLC cbs bo failed\n", r);
gfx_v8_0_rlc_fini(adev);
return r;
}
&adev->gfx.rlc.cp_table_gpu_addr);
if (r) {
amdgpu_bo_unreserve(adev->gfx.rlc.cp_table_obj);
- dev_warn(adev->dev, "(%d) pin RLC cp_table bo failed\n", r);
+ dev_warn(adev->dev, "(%d) pin RLC cp table bo failed\n", r);
return r;
}
r = amdgpu_bo_kmap(adev->gfx.rlc.cp_table_obj, (void **)&adev->gfx.rlc.cp_table_ptr);
amdgpu_bo_kunmap(adev->gfx.rlc.cp_table_obj);
amdgpu_bo_unreserve(adev->gfx.rlc.cp_table_obj);
-
}
return 0;
dev_warn(adev->dev, "(%d) reserve HPD EOP bo failed\n", r);
amdgpu_bo_unpin(adev->gfx.mec.hpd_eop_obj);
amdgpu_bo_unreserve(adev->gfx.mec.hpd_eop_obj);
-
amdgpu_bo_unref(&adev->gfx.mec.hpd_eop_obj);
adev->gfx.mec.hpd_eop_obj = NULL;
}
}
/* reserve GDS, GWS and OA resource for gfx */
- r = amdgpu_bo_create(adev, adev->gds.mem.gfx_partition_size,
- PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_GDS, 0, NULL,
- NULL, &adev->gds.gds_gfx_bo);
+ r = amdgpu_bo_create_kernel(adev, adev->gds.mem.gfx_partition_size,
+ PAGE_SIZE, AMDGPU_GEM_DOMAIN_GDS,
+ &adev->gds.gds_gfx_bo, NULL, NULL);
if (r)
return r;
- r = amdgpu_bo_create(adev, adev->gds.gws.gfx_partition_size,
- PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_GWS, 0, NULL,
- NULL, &adev->gds.gws_gfx_bo);
+ r = amdgpu_bo_create_kernel(adev, adev->gds.gws.gfx_partition_size,
+ PAGE_SIZE, AMDGPU_GEM_DOMAIN_GWS,
+ &adev->gds.gws_gfx_bo, NULL, NULL);
if (r)
return r;
- r = amdgpu_bo_create(adev, adev->gds.oa.gfx_partition_size,
- PAGE_SIZE, true,
- AMDGPU_GEM_DOMAIN_OA, 0, NULL,
- NULL, &adev->gds.oa_gfx_bo);
+ r = amdgpu_bo_create_kernel(adev, adev->gds.oa.gfx_partition_size,
+ PAGE_SIZE, AMDGPU_GEM_DOMAIN_OA,
+ &adev->gds.oa_gfx_bo, NULL, NULL);
if (r)
return r;
int i;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
- amdgpu_bo_unref(&adev->gds.oa_gfx_bo);
- amdgpu_bo_unref(&adev->gds.gws_gfx_bo);
- amdgpu_bo_unref(&adev->gds.gds_gfx_bo);
+ amdgpu_bo_free_kernel(&adev->gds.oa_gfx_bo, NULL, NULL);
+ amdgpu_bo_free_kernel(&adev->gds.gws_gfx_bo, NULL, NULL);
+ amdgpu_bo_free_kernel(&adev->gds.gds_gfx_bo, NULL, NULL);
for (i = 0; i < adev->gfx.num_gfx_rings; i++)
amdgpu_ring_fini(&adev->gfx.gfx_ring[i]);
amdgpu_ring_fini(&adev->gfx.compute_ring[i]);
gfx_v8_0_mec_fini(adev);
-
gfx_v8_0_rlc_fini(adev);
-
gfx_v8_0_free_microcode(adev);
return 0;
else
data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_INDEX, instance);
- if ((se_num == 0xffffffff) && (sh_num == 0xffffffff)) {
- data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_BROADCAST_WRITES, 1);
+ if (se_num == 0xffffffff)
data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_BROADCAST_WRITES, 1);
- } else if (se_num == 0xffffffff) {
- data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_INDEX, sh_num);
- data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_BROADCAST_WRITES, 1);
- } else if (sh_num == 0xffffffff) {
- data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_BROADCAST_WRITES, 1);
+ else
data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_INDEX, se_num);
- } else {
+
+ if (sh_num == 0xffffffff)
+ data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_BROADCAST_WRITES, 1);
+ else
data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_INDEX, sh_num);
- data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SE_INDEX, se_num);
- }
+
WREG32(mmGRBM_GFX_INDEX, data);
}
{
u32 data, mask;
- data = RREG32(mmCC_RB_BACKEND_DISABLE);
- data |= RREG32(mmGC_USER_RB_BACKEND_DISABLE);
+ data = RREG32(mmCC_RB_BACKEND_DISABLE) |
+ RREG32(mmGC_USER_RB_BACKEND_DISABLE);
- data &= CC_RB_BACKEND_DISABLE__BACKEND_DISABLE_MASK;
- data >>= GC_USER_RB_BACKEND_DISABLE__BACKEND_DISABLE__SHIFT;
+ data = REG_GET_FIELD(data, GC_USER_RB_BACKEND_DISABLE, BACKEND_DISABLE);
mask = gfx_v8_0_create_bitmask(adev->gfx.config.max_backends_per_se /
adev->gfx.config.max_sh_per_se);
u32 tmp;
int i;
- tmp = RREG32(mmGRBM_CNTL);
- tmp = REG_SET_FIELD(tmp, GRBM_CNTL, READ_TIMEOUT, 0xff);
- WREG32(mmGRBM_CNTL, tmp);
-
+ WREG32_FIELD(GRBM_CNTL, READ_TIMEOUT, 0xFF);
WREG32(mmGB_ADDR_CONFIG, adev->gfx.config.gb_addr_config);
WREG32(mmHDP_ADDR_CONFIG, adev->gfx.config.gb_addr_config);
WREG32(mmDMIF_ADDR_CALC, adev->gfx.config.gb_addr_config);
gfx_v8_0_tiling_mode_table_init(adev);
-
gfx_v8_0_setup_rb(adev);
gfx_v8_0_get_cu_info(adev);
sizeof(indirect_start_offsets)/sizeof(int));
/* save and restore list */
- temp = RREG32(mmRLC_SRM_CNTL);
- temp |= RLC_SRM_CNTL__AUTO_INCR_ADDR_MASK;
- WREG32(mmRLC_SRM_CNTL, temp);
+ WREG32_FIELD(RLC_SRM_CNTL, AUTO_INCR_ADDR, 1);
WREG32(mmRLC_SRM_ARAM_ADDR, 0);
for (i = 0; i < adev->gfx.rlc.reg_list_size_bytes >> 2; i++)
static void gfx_v8_0_enable_save_restore_machine(struct amdgpu_device *adev)
{
- uint32_t data;
-
- data = RREG32(mmRLC_SRM_CNTL);
- data |= RLC_SRM_CNTL__SRM_ENABLE_MASK;
- WREG32(mmRLC_SRM_CNTL, data);
+ WREG32_FIELD(RLC_SRM_CNTL, SRM_ENABLE, 1);
}
static void gfx_v8_0_init_power_gating(struct amdgpu_device *adev)
if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG |
AMD_PG_SUPPORT_GFX_SMG |
AMD_PG_SUPPORT_GFX_DMG)) {
- data = RREG32(mmCP_RB_WPTR_POLL_CNTL);
- data &= ~CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT_MASK;
- data |= (0x60 << CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT__SHIFT);
- WREG32(mmCP_RB_WPTR_POLL_CNTL, data);
-
- data = 0;
- data |= (0x10 << RLC_PG_DELAY__POWER_UP_DELAY__SHIFT);
- data |= (0x10 << RLC_PG_DELAY__POWER_DOWN_DELAY__SHIFT);
- data |= (0x10 << RLC_PG_DELAY__CMD_PROPAGATE_DELAY__SHIFT);
- data |= (0x10 << RLC_PG_DELAY__MEM_SLEEP_DELAY__SHIFT);
- WREG32(mmRLC_PG_DELAY, data);
+ WREG32_FIELD(CP_RB_WPTR_POLL_CNTL, IDLE_POLL_COUNT, 0x60);
- data = RREG32(mmRLC_PG_DELAY_2);
- data &= ~RLC_PG_DELAY_2__SERDES_CMD_DELAY_MASK;
- data |= (0x3 << RLC_PG_DELAY_2__SERDES_CMD_DELAY__SHIFT);
- WREG32(mmRLC_PG_DELAY_2, data);
+ data = REG_SET_FIELD(0, RLC_PG_DELAY, POWER_UP_DELAY, 0x10);
+ data = REG_SET_FIELD(data, RLC_PG_DELAY, POWER_DOWN_DELAY, 0x10);
+ data = REG_SET_FIELD(data, RLC_PG_DELAY, CMD_PROPAGATE_DELAY, 0x10);
+ data = REG_SET_FIELD(data, RLC_PG_DELAY, MEM_SLEEP_DELAY, 0x10);
+ WREG32(mmRLC_PG_DELAY, data);
- data = RREG32(mmRLC_AUTO_PG_CTRL);
- data &= ~RLC_AUTO_PG_CTRL__GRBM_REG_SAVE_GFX_IDLE_THRESHOLD_MASK;
- data |= (0x55f0 << RLC_AUTO_PG_CTRL__GRBM_REG_SAVE_GFX_IDLE_THRESHOLD__SHIFT);
- WREG32(mmRLC_AUTO_PG_CTRL, data);
+ WREG32_FIELD(RLC_PG_DELAY_2, SERDES_CMD_DELAY, 0x3);
+ WREG32_FIELD(RLC_AUTO_PG_CTRL, GRBM_REG_SAVE_GFX_IDLE_THRESHOLD, 0x55f0);
}
}
static void cz_enable_sck_slow_down_on_power_up(struct amdgpu_device *adev,
bool enable)
{
- u32 data, orig;
-
- orig = data = RREG32(mmRLC_PG_CNTL);
-
- if (enable)
- data |= RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PU_ENABLE_MASK;
- else
- data &= ~RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PU_ENABLE_MASK;
-
- if (orig != data)
- WREG32(mmRLC_PG_CNTL, data);
+ WREG32_FIELD(RLC_PG_CNTL, SMU_CLK_SLOWDOWN_ON_PU_ENABLE, enable ? 1 : 0);
}
static void cz_enable_sck_slow_down_on_power_down(struct amdgpu_device *adev,
bool enable)
{
- u32 data, orig;
-
- orig = data = RREG32(mmRLC_PG_CNTL);
-
- if (enable)
- data |= RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PD_ENABLE_MASK;
- else
- data &= ~RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PD_ENABLE_MASK;
-
- if (orig != data)
- WREG32(mmRLC_PG_CNTL, data);
+ WREG32_FIELD(RLC_PG_CNTL, SMU_CLK_SLOWDOWN_ON_PD_ENABLE, enable ? 1 : 0);
}
static void cz_enable_cp_power_gating(struct amdgpu_device *adev, bool enable)
{
- u32 data, orig;
-
- orig = data = RREG32(mmRLC_PG_CNTL);
-
- if (enable)
- data &= ~RLC_PG_CNTL__CP_PG_DISABLE_MASK;
- else
- data |= RLC_PG_CNTL__CP_PG_DISABLE_MASK;
-
- if (orig != data)
- WREG32(mmRLC_PG_CNTL, data);
+ WREG32_FIELD(RLC_PG_CNTL, CP_PG_DISABLE, enable ? 1 : 0);
}
static void gfx_v8_0_init_pg(struct amdgpu_device *adev)
}
}
-void gfx_v8_0_rlc_stop(struct amdgpu_device *adev)
+static void gfx_v8_0_rlc_stop(struct amdgpu_device *adev)
{
- u32 tmp = RREG32(mmRLC_CNTL);
-
- tmp = REG_SET_FIELD(tmp, RLC_CNTL, RLC_ENABLE_F32, 0);
- WREG32(mmRLC_CNTL, tmp);
+ WREG32_FIELD(RLC_CNTL, RLC_ENABLE_F32, 0);
gfx_v8_0_enable_gui_idle_interrupt(adev, false);
-
gfx_v8_0_wait_for_rlc_serdes(adev);
}
static void gfx_v8_0_rlc_reset(struct amdgpu_device *adev)
{
- u32 tmp = RREG32(mmGRBM_SOFT_RESET);
-
- tmp = REG_SET_FIELD(tmp, GRBM_SOFT_RESET, SOFT_RESET_RLC, 1);
- WREG32(mmGRBM_SOFT_RESET, tmp);
+ WREG32_FIELD(GRBM_SOFT_RESET, SOFT_RESET_RLC, 1);
udelay(50);
- tmp = REG_SET_FIELD(tmp, GRBM_SOFT_RESET, SOFT_RESET_RLC, 0);
- WREG32(mmGRBM_SOFT_RESET, tmp);
+
+ WREG32_FIELD(GRBM_SOFT_RESET, SOFT_RESET_RLC, 0);
udelay(50);
}
static void gfx_v8_0_rlc_start(struct amdgpu_device *adev)
{
- u32 tmp = RREG32(mmRLC_CNTL);
-
- tmp = REG_SET_FIELD(tmp, RLC_CNTL, RLC_ENABLE_F32, 1);
- WREG32(mmRLC_CNTL, tmp);
+ WREG32_FIELD(RLC_CNTL, RLC_ENABLE_F32, 1);
/* carrizo do enable cp interrupt after cp inited */
if (!(adev->flags & AMD_IS_APU))
/* disable CG */
WREG32(mmRLC_CGCG_CGLS_CTRL, 0);
if (adev->asic_type == CHIP_POLARIS11 ||
- adev->asic_type == CHIP_POLARIS10)
+ adev->asic_type == CHIP_POLARIS10)
WREG32(mmRLC_CGCG_CGLS_CTRL_3D, 0);
/* disable PG */
WREG32(mmRLC_PG_CNTL, 0);
gfx_v8_0_rlc_reset(adev);
-
gfx_v8_0_init_pg(adev);
if (!adev->pp_enabled) {
gfx_v8_0_cp_gfx_start(adev);
ring->ready = true;
r = amdgpu_ring_test_ring(ring);
- if (r) {
+ if (r)
ring->ready = false;
- return r;
- }
- return 0;
+ return r;
}
static void gfx_v8_0_cp_compute_enable(struct amdgpu_device *adev, bool enable)
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
gfx_v8_0_init_golden_registers(adev);
-
gfx_v8_0_gpu_init(adev);
r = gfx_v8_0_rlc_resume(adev);
return r;
r = gfx_v8_0_cp_resume(adev);
- if (r)
- return r;
return r;
}
static int gfx_v8_0_wait_for_idle(void *handle)
{
unsigned i;
- u32 tmp;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
for (i = 0; i < adev->usec_timeout; i++) {
- /* read MC_STATUS */
- tmp = RREG32(mmGRBM_STATUS) & GRBM_STATUS__GUI_ACTIVE_MASK;
-
- if (!REG_GET_FIELD(tmp, GRBM_STATUS, GUI_ACTIVE))
+ if (gfx_v8_0_is_idle(handle))
return 0;
+
udelay(1);
}
return -ETIMEDOUT;
}
-static int gfx_v8_0_soft_reset(void *handle)
+static int gfx_v8_0_check_soft_reset(void *handle)
{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
u32 grbm_soft_reset = 0, srbm_soft_reset = 0;
u32 tmp;
- struct amdgpu_device *adev = (struct amdgpu_device *)handle;
/* GRBM_STATUS */
tmp = RREG32(mmGRBM_STATUS);
GRBM_STATUS__TA_BUSY_MASK | GRBM_STATUS__VGT_BUSY_MASK |
GRBM_STATUS__DB_BUSY_MASK | GRBM_STATUS__CB_BUSY_MASK |
GRBM_STATUS__GDS_BUSY_MASK | GRBM_STATUS__SPI_BUSY_MASK |
- GRBM_STATUS__IA_BUSY_MASK | GRBM_STATUS__IA_BUSY_NO_DMA_MASK)) {
+ GRBM_STATUS__IA_BUSY_MASK | GRBM_STATUS__IA_BUSY_NO_DMA_MASK |
+ GRBM_STATUS__CP_BUSY_MASK | GRBM_STATUS__CP_COHERENCY_BUSY_MASK)) {
grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset,
GRBM_SOFT_RESET, SOFT_RESET_CP, 1);
grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset,
GRBM_SOFT_RESET, SOFT_RESET_GFX, 1);
- }
-
- if (tmp & (GRBM_STATUS__CP_BUSY_MASK | GRBM_STATUS__CP_COHERENCY_BUSY_MASK)) {
- grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset,
- GRBM_SOFT_RESET, SOFT_RESET_CP, 1);
srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset,
SRBM_SOFT_RESET, SOFT_RESET_GRBM, 1);
}
grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset,
GRBM_SOFT_RESET, SOFT_RESET_RLC, 1);
+ if (REG_GET_FIELD(tmp, GRBM_STATUS2, CPF_BUSY) ||
+ REG_GET_FIELD(tmp, GRBM_STATUS2, CPC_BUSY) ||
+ REG_GET_FIELD(tmp, GRBM_STATUS2, CPG_BUSY)) {
+ grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET,
+ SOFT_RESET_CPF, 1);
+ grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET,
+ SOFT_RESET_CPC, 1);
+ grbm_soft_reset = REG_SET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET,
+ SOFT_RESET_CPG, 1);
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET,
+ SOFT_RESET_GRBM, 1);
+ }
+
/* SRBM_STATUS */
tmp = RREG32(mmSRBM_STATUS);
if (REG_GET_FIELD(tmp, SRBM_STATUS, GRBM_RQ_PENDING))
srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset,
SRBM_SOFT_RESET, SOFT_RESET_GRBM, 1);
+ if (REG_GET_FIELD(tmp, SRBM_STATUS, SEM_BUSY))
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset,
+ SRBM_SOFT_RESET, SOFT_RESET_SEM, 1);
if (grbm_soft_reset || srbm_soft_reset) {
- /* stop the rlc */
- gfx_v8_0_rlc_stop(adev);
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_GFX].hang = true;
+ adev->gfx.grbm_soft_reset = grbm_soft_reset;
+ adev->gfx.srbm_soft_reset = srbm_soft_reset;
+ } else {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_GFX].hang = false;
+ adev->gfx.grbm_soft_reset = 0;
+ adev->gfx.srbm_soft_reset = 0;
+ }
+
+ return 0;
+}
+
+static void gfx_v8_0_inactive_hqd(struct amdgpu_device *adev,
+ struct amdgpu_ring *ring)
+{
+ int i;
+
+ vi_srbm_select(adev, ring->me, ring->pipe, ring->queue, 0);
+ if (RREG32(mmCP_HQD_ACTIVE) & CP_HQD_ACTIVE__ACTIVE_MASK) {
+ u32 tmp;
+ tmp = RREG32(mmCP_HQD_DEQUEUE_REQUEST);
+ tmp = REG_SET_FIELD(tmp, CP_HQD_DEQUEUE_REQUEST,
+ DEQUEUE_REQ, 2);
+ WREG32(mmCP_HQD_DEQUEUE_REQUEST, tmp);
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (!(RREG32(mmCP_HQD_ACTIVE) & CP_HQD_ACTIVE__ACTIVE_MASK))
+ break;
+ udelay(1);
+ }
+ }
+}
+
+static int gfx_v8_0_pre_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 grbm_soft_reset = 0, srbm_soft_reset = 0;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_GFX].hang)
+ return 0;
+
+ grbm_soft_reset = adev->gfx.grbm_soft_reset;
+ srbm_soft_reset = adev->gfx.srbm_soft_reset;
+
+ /* stop the rlc */
+ gfx_v8_0_rlc_stop(adev);
+ if (REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CP) ||
+ REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_GFX))
/* Disable GFX parsing/prefetching */
gfx_v8_0_cp_gfx_enable(adev, false);
+ if (REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CP) ||
+ REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CPF) ||
+ REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CPC) ||
+ REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CPG)) {
+ int i;
+
+ for (i = 0; i < adev->gfx.num_compute_rings; i++) {
+ struct amdgpu_ring *ring = &adev->gfx.compute_ring[i];
+
+ gfx_v8_0_inactive_hqd(adev, ring);
+ }
/* Disable MEC parsing/prefetching */
gfx_v8_0_cp_compute_enable(adev, false);
+ }
- if (grbm_soft_reset || srbm_soft_reset) {
- tmp = RREG32(mmGMCON_DEBUG);
- tmp = REG_SET_FIELD(tmp,
- GMCON_DEBUG, GFX_STALL, 1);
- tmp = REG_SET_FIELD(tmp,
- GMCON_DEBUG, GFX_CLEAR, 1);
- WREG32(mmGMCON_DEBUG, tmp);
+ return 0;
+}
- udelay(50);
- }
+static int gfx_v8_0_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 grbm_soft_reset = 0, srbm_soft_reset = 0;
+ u32 tmp;
- if (grbm_soft_reset) {
- tmp = RREG32(mmGRBM_SOFT_RESET);
- tmp |= grbm_soft_reset;
- dev_info(adev->dev, "GRBM_SOFT_RESET=0x%08X\n", tmp);
- WREG32(mmGRBM_SOFT_RESET, tmp);
- tmp = RREG32(mmGRBM_SOFT_RESET);
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_GFX].hang)
+ return 0;
- udelay(50);
+ grbm_soft_reset = adev->gfx.grbm_soft_reset;
+ srbm_soft_reset = adev->gfx.srbm_soft_reset;
- tmp &= ~grbm_soft_reset;
- WREG32(mmGRBM_SOFT_RESET, tmp);
- tmp = RREG32(mmGRBM_SOFT_RESET);
- }
+ if (grbm_soft_reset || srbm_soft_reset) {
+ tmp = RREG32(mmGMCON_DEBUG);
+ tmp = REG_SET_FIELD(tmp, GMCON_DEBUG, GFX_STALL, 1);
+ tmp = REG_SET_FIELD(tmp, GMCON_DEBUG, GFX_CLEAR, 1);
+ WREG32(mmGMCON_DEBUG, tmp);
+ udelay(50);
+ }
- if (srbm_soft_reset) {
- tmp = RREG32(mmSRBM_SOFT_RESET);
- tmp |= srbm_soft_reset;
- dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp);
- WREG32(mmSRBM_SOFT_RESET, tmp);
- tmp = RREG32(mmSRBM_SOFT_RESET);
+ if (grbm_soft_reset) {
+ tmp = RREG32(mmGRBM_SOFT_RESET);
+ tmp |= grbm_soft_reset;
+ dev_info(adev->dev, "GRBM_SOFT_RESET=0x%08X\n", tmp);
+ WREG32(mmGRBM_SOFT_RESET, tmp);
+ tmp = RREG32(mmGRBM_SOFT_RESET);
- udelay(50);
+ udelay(50);
- tmp &= ~srbm_soft_reset;
- WREG32(mmSRBM_SOFT_RESET, tmp);
- tmp = RREG32(mmSRBM_SOFT_RESET);
- }
+ tmp &= ~grbm_soft_reset;
+ WREG32(mmGRBM_SOFT_RESET, tmp);
+ tmp = RREG32(mmGRBM_SOFT_RESET);
+ }
- if (grbm_soft_reset || srbm_soft_reset) {
- tmp = RREG32(mmGMCON_DEBUG);
- tmp = REG_SET_FIELD(tmp,
- GMCON_DEBUG, GFX_STALL, 0);
- tmp = REG_SET_FIELD(tmp,
- GMCON_DEBUG, GFX_CLEAR, 0);
- WREG32(mmGMCON_DEBUG, tmp);
- }
+ if (srbm_soft_reset) {
+ tmp = RREG32(mmSRBM_SOFT_RESET);
+ tmp |= srbm_soft_reset;
+ dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp);
+ WREG32(mmSRBM_SOFT_RESET, tmp);
+ tmp = RREG32(mmSRBM_SOFT_RESET);
- /* Wait a little for things to settle down */
udelay(50);
+
+ tmp &= ~srbm_soft_reset;
+ WREG32(mmSRBM_SOFT_RESET, tmp);
+ tmp = RREG32(mmSRBM_SOFT_RESET);
}
+
+ if (grbm_soft_reset || srbm_soft_reset) {
+ tmp = RREG32(mmGMCON_DEBUG);
+ tmp = REG_SET_FIELD(tmp, GMCON_DEBUG, GFX_STALL, 0);
+ tmp = REG_SET_FIELD(tmp, GMCON_DEBUG, GFX_CLEAR, 0);
+ WREG32(mmGMCON_DEBUG, tmp);
+ }
+
+ /* Wait a little for things to settle down */
+ udelay(50);
+
+ return 0;
+}
+
+static void gfx_v8_0_init_hqd(struct amdgpu_device *adev,
+ struct amdgpu_ring *ring)
+{
+ vi_srbm_select(adev, ring->me, ring->pipe, ring->queue, 0);
+ WREG32(mmCP_HQD_DEQUEUE_REQUEST, 0);
+ WREG32(mmCP_HQD_PQ_RPTR, 0);
+ WREG32(mmCP_HQD_PQ_WPTR, 0);
+ vi_srbm_select(adev, 0, 0, 0, 0);
+}
+
+static int gfx_v8_0_post_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 grbm_soft_reset = 0, srbm_soft_reset = 0;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_GFX].hang)
+ return 0;
+
+ grbm_soft_reset = adev->gfx.grbm_soft_reset;
+ srbm_soft_reset = adev->gfx.srbm_soft_reset;
+
+ if (REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CP) ||
+ REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_GFX))
+ gfx_v8_0_cp_gfx_resume(adev);
+
+ if (REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CP) ||
+ REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CPF) ||
+ REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CPC) ||
+ REG_GET_FIELD(grbm_soft_reset, GRBM_SOFT_RESET, SOFT_RESET_CPG)) {
+ int i;
+
+ for (i = 0; i < adev->gfx.num_compute_rings; i++) {
+ struct amdgpu_ring *ring = &adev->gfx.compute_ring[i];
+
+ gfx_v8_0_init_hqd(adev, ring);
+ }
+ gfx_v8_0_cp_compute_resume(adev);
+ }
+ gfx_v8_0_rlc_start(adev);
+
return 0;
}
static void gfx_v8_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev,
bool enable)
{
- uint32_t data, temp;
-
if (adev->asic_type == CHIP_POLARIS11)
/* Send msg to SMU via Powerplay */
amdgpu_set_powergating_state(adev,
enable ?
AMD_PG_STATE_GATE : AMD_PG_STATE_UNGATE);
- temp = data = RREG32(mmRLC_PG_CNTL);
- /* Enable static MGPG */
- if (enable)
- data |= RLC_PG_CNTL__STATIC_PER_CU_PG_ENABLE_MASK;
- else
- data &= ~RLC_PG_CNTL__STATIC_PER_CU_PG_ENABLE_MASK;
-
- if (temp != data)
- WREG32(mmRLC_PG_CNTL, data);
+ WREG32_FIELD(RLC_PG_CNTL, STATIC_PER_CU_PG_ENABLE, enable ? 1 : 0);
}
static void gfx_v8_0_enable_gfx_dynamic_mg_power_gating(struct amdgpu_device *adev,
bool enable)
{
- uint32_t data, temp;
-
- temp = data = RREG32(mmRLC_PG_CNTL);
- /* Enable dynamic MGPG */
- if (enable)
- data |= RLC_PG_CNTL__DYN_PER_CU_PG_ENABLE_MASK;
- else
- data &= ~RLC_PG_CNTL__DYN_PER_CU_PG_ENABLE_MASK;
-
- if (temp != data)
- WREG32(mmRLC_PG_CNTL, data);
+ WREG32_FIELD(RLC_PG_CNTL, DYN_PER_CU_PG_ENABLE, enable ? 1 : 0);
}
static void polaris11_enable_gfx_quick_mg_power_gating(struct amdgpu_device *adev,
bool enable)
{
- uint32_t data, temp;
-
- temp = data = RREG32(mmRLC_PG_CNTL);
- /* Enable quick PG */
- if (enable)
- data |= RLC_PG_CNTL__QUICK_PG_ENABLE_MASK;
- else
- data &= ~RLC_PG_CNTL__QUICK_PG_ENABLE_MASK;
-
- if (temp != data)
- WREG32(mmRLC_PG_CNTL, data);
+ WREG32_FIELD(RLC_PG_CNTL, QUICK_PG_ENABLE, enable ? 1 : 0);
}
static void cz_enable_gfx_cg_power_gating(struct amdgpu_device *adev,
bool enable)
{
- u32 data, orig;
-
- orig = data = RREG32(mmRLC_PG_CNTL);
-
- if (enable)
- data |= RLC_PG_CNTL__GFX_POWER_GATING_ENABLE_MASK;
- else
- data &= ~RLC_PG_CNTL__GFX_POWER_GATING_ENABLE_MASK;
-
- if (orig != data)
- WREG32(mmRLC_PG_CNTL, data);
+ WREG32_FIELD(RLC_PG_CNTL, GFX_POWER_GATING_ENABLE, enable ? 1 : 0);
}
static void cz_enable_gfx_pipeline_power_gating(struct amdgpu_device *adev,
bool enable)
{
- u32 data, orig;
-
- orig = data = RREG32(mmRLC_PG_CNTL);
-
- if (enable)
- data |= RLC_PG_CNTL__GFX_PIPELINE_PG_ENABLE_MASK;
- else
- data &= ~RLC_PG_CNTL__GFX_PIPELINE_PG_ENABLE_MASK;
-
- if (orig != data)
- WREG32(mmRLC_PG_CNTL, data);
+ WREG32_FIELD(RLC_PG_CNTL, GFX_PIPELINE_PG_ENABLE, enable ? 1 : 0);
/* Read any GFX register to wake up GFX. */
if (!enable)
- data = RREG32(mmDB_RENDER_CONTROL);
+ RREG32(mmDB_RENDER_CONTROL);
}
static void cz_update_gfx_cg_power_gating(struct amdgpu_device *adev,
data = RREG32(mmRLC_SERDES_WR_CTRL);
if (adev->asic_type == CHIP_STONEY)
- data &= ~(RLC_SERDES_WR_CTRL__WRITE_COMMAND_MASK |
- RLC_SERDES_WR_CTRL__READ_COMMAND_MASK |
- RLC_SERDES_WR_CTRL__P1_SELECT_MASK |
- RLC_SERDES_WR_CTRL__P2_SELECT_MASK |
- RLC_SERDES_WR_CTRL__RDDATA_RESET_MASK |
- RLC_SERDES_WR_CTRL__POWER_DOWN_MASK |
- RLC_SERDES_WR_CTRL__POWER_UP_MASK |
- RLC_SERDES_WR_CTRL__SHORT_FORMAT_MASK |
- RLC_SERDES_WR_CTRL__SRBM_OVERRIDE_MASK);
+ data &= ~(RLC_SERDES_WR_CTRL__WRITE_COMMAND_MASK |
+ RLC_SERDES_WR_CTRL__READ_COMMAND_MASK |
+ RLC_SERDES_WR_CTRL__P1_SELECT_MASK |
+ RLC_SERDES_WR_CTRL__P2_SELECT_MASK |
+ RLC_SERDES_WR_CTRL__RDDATA_RESET_MASK |
+ RLC_SERDES_WR_CTRL__POWER_DOWN_MASK |
+ RLC_SERDES_WR_CTRL__POWER_UP_MASK |
+ RLC_SERDES_WR_CTRL__SHORT_FORMAT_MASK |
+ RLC_SERDES_WR_CTRL__SRBM_OVERRIDE_MASK);
else
data &= ~(RLC_SERDES_WR_CTRL__WRITE_COMMAND_MASK |
RLC_SERDES_WR_CTRL__READ_COMMAND_MASK |
#define MSG_ENTER_RLC_SAFE_MODE 1
#define MSG_EXIT_RLC_SAFE_MODE 0
-
-#define RLC_GPR_REG2__REQ_MASK 0x00000001
-#define RLC_GPR_REG2__MESSAGE__SHIFT 0x00000001
-#define RLC_GPR_REG2__MESSAGE_MASK 0x0000001e
+#define RLC_GPR_REG2__REQ_MASK 0x00000001
+#define RLC_GPR_REG2__REQ__SHIFT 0
+#define RLC_GPR_REG2__MESSAGE__SHIFT 0x00000001
+#define RLC_GPR_REG2__MESSAGE_MASK 0x0000001e
static void cz_enter_rlc_safe_mode(struct amdgpu_device *adev)
{
}
for (i = 0; i < adev->usec_timeout; i++) {
- if ((RREG32(mmRLC_GPR_REG2) & RLC_GPR_REG2__REQ_MASK) == 0)
+ if (!REG_GET_FIELD(RREG32(mmRLC_GPR_REG2), RLC_GPR_REG2, REQ))
break;
udelay(1);
}
}
for (i = 0; i < adev->usec_timeout; i++) {
- if ((RREG32(mmRLC_GPR_REG2) & RLC_GPR_REG2__REQ_MASK) == 0)
+ if (!REG_GET_FIELD(RREG32(mmRLC_GPR_REG2), RLC_GPR_REG2, REQ))
break;
udelay(1);
}
}
for (i = 0; i < adev->usec_timeout; i++) {
- if ((RREG32(mmRLC_SAFE_MODE) & RLC_SAFE_MODE__CMD_MASK) == 0)
+ if (!REG_GET_FIELD(RREG32(mmRLC_SAFE_MODE), RLC_SAFE_MODE, CMD))
break;
udelay(1);
}
}
for (i = 0; i < adev->usec_timeout; i++) {
- if ((RREG32(mmRLC_SAFE_MODE) & RLC_SAFE_MODE__CMD_MASK) == 0)
+ if (!REG_GET_FIELD(RREG32(mmRLC_SAFE_MODE), RLC_SAFE_MODE, CMD))
break;
udelay(1);
}
/* It is disabled by HW by default */
if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_MGCG)) {
if (adev->cg_flags & AMD_CG_SUPPORT_GFX_MGLS) {
- if (adev->cg_flags & AMD_CG_SUPPORT_GFX_RLC_LS) {
+ if (adev->cg_flags & AMD_CG_SUPPORT_GFX_RLC_LS)
/* 1 - RLC memory Light sleep */
- temp = data = RREG32(mmRLC_MEM_SLP_CNTL);
- data |= RLC_MEM_SLP_CNTL__RLC_MEM_LS_EN_MASK;
- if (temp != data)
- WREG32(mmRLC_MEM_SLP_CNTL, data);
- }
+ WREG32_FIELD(RLC_MEM_SLP_CNTL, RLC_MEM_LS_EN, 1);
- if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CP_LS) {
- /* 2 - CP memory Light sleep */
- temp = data = RREG32(mmCP_MEM_SLP_CNTL);
- data |= CP_MEM_SLP_CNTL__CP_MEM_LS_EN_MASK;
- if (temp != data)
- WREG32(mmCP_MEM_SLP_CNTL, data);
- }
+ if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CP_LS)
+ WREG32_FIELD(CP_MEM_SLP_CNTL, CP_MEM_LS_EN, 1);
}
/* 3 - RLC_CGTT_MGCG_OVERRIDE */
return 0;
}
-static u32 gfx_v8_0_ring_get_rptr_gfx(struct amdgpu_ring *ring)
+static u32 gfx_v8_0_ring_get_rptr(struct amdgpu_ring *ring)
{
- u32 rptr;
-
- rptr = ring->adev->wb.wb[ring->rptr_offs];
-
- return rptr;
+ return ring->adev->wb.wb[ring->rptr_offs];
}
static u32 gfx_v8_0_ring_get_wptr_gfx(struct amdgpu_ring *ring)
{
struct amdgpu_device *adev = ring->adev;
- u32 wptr;
if (ring->use_doorbell)
/* XXX check if swapping is necessary on BE */
- wptr = ring->adev->wb.wb[ring->wptr_offs];
+ return ring->adev->wb.wb[ring->wptr_offs];
else
- wptr = RREG32(mmCP_RB0_WPTR);
-
- return wptr;
+ return RREG32(mmCP_RB0_WPTR);
}
static void gfx_v8_0_ring_set_wptr_gfx(struct amdgpu_ring *ring)
{
u32 header, control = 0;
- /* insert SWITCH_BUFFER packet before first IB in the ring frame */
- if (ctx_switch) {
- amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
- amdgpu_ring_write(ring, 0);
- }
-
if (ib->flags & AMDGPU_IB_FLAG_CE)
header = PACKET3(PACKET3_INDIRECT_BUFFER_CONST, 2);
else
amdgpu_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2));
amdgpu_ring_write(ring,
#ifdef __BIG_ENDIAN
- (2 << 0) |
+ (2 << 0) |
#endif
- (ib->gpu_addr & 0xFFFFFFFC));
+ (ib->gpu_addr & 0xFFFFFFFC));
amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr) & 0xFFFF);
amdgpu_ring_write(ring, control);
}
amdgpu_ring_write(ring, seq);
amdgpu_ring_write(ring, 0xffffffff);
amdgpu_ring_write(ring, 4); /* poll interval */
-
- if (usepfp) {
- /* synce CE with ME to prevent CE fetch CEIB before context switch done */
- amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
- amdgpu_ring_write(ring, 0);
- amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
- amdgpu_ring_write(ring, 0);
- }
}
static void gfx_v8_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
{
int usepfp = (ring->type == AMDGPU_RING_TYPE_GFX);
+ /* GFX8 emits 128 dw nop to prevent DE do vm_flush before CE finish CEIB */
+ if (usepfp)
+ amdgpu_ring_insert_nop(ring, 128);
+
amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3));
amdgpu_ring_write(ring, (WRITE_DATA_ENGINE_SEL(usepfp) |
WRITE_DATA_DST_SEL(0)) |
/* sync PFP to ME, otherwise we might get invalid PFP reads */
amdgpu_ring_write(ring, PACKET3(PACKET3_PFP_SYNC_ME, 0));
amdgpu_ring_write(ring, 0x0);
- amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
- amdgpu_ring_write(ring, 0);
- amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
- amdgpu_ring_write(ring, 0);
+ /* GFX8 emits 128 dw nop to prevent CE access VM before vm_flush finish */
+ amdgpu_ring_insert_nop(ring, 128);
}
}
-static u32 gfx_v8_0_ring_get_rptr_compute(struct amdgpu_ring *ring)
-{
- return ring->adev->wb.wb[ring->rptr_offs];
-}
-
static u32 gfx_v8_0_ring_get_wptr_compute(struct amdgpu_ring *ring)
{
return ring->adev->wb.wb[ring->wptr_offs];
amdgpu_ring_write(ring, upper_32_bits(seq));
}
-static void gfx_v8_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
- enum amdgpu_interrupt_state state)
+static void gfx_v8_ring_emit_sb(struct amdgpu_ring *ring)
{
- u32 cp_int_cntl;
+ amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
+ amdgpu_ring_write(ring, 0);
+}
- switch (state) {
- case AMDGPU_IRQ_STATE_DISABLE:
- cp_int_cntl = RREG32(mmCP_INT_CNTL_RING0);
- cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0,
- TIME_STAMP_INT_ENABLE, 0);
- WREG32(mmCP_INT_CNTL_RING0, cp_int_cntl);
- break;
- case AMDGPU_IRQ_STATE_ENABLE:
- cp_int_cntl = RREG32(mmCP_INT_CNTL_RING0);
- cp_int_cntl =
- REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0,
- TIME_STAMP_INT_ENABLE, 1);
- WREG32(mmCP_INT_CNTL_RING0, cp_int_cntl);
- break;
- default:
- break;
+static void gfx_v8_ring_emit_cntxcntl(struct amdgpu_ring *ring, uint32_t flags)
+{
+ uint32_t dw2 = 0;
+
+ dw2 |= 0x80000000; /* set load_enable otherwise this package is just NOPs */
+ if (flags & AMDGPU_HAVE_CTX_SWITCH) {
+ /* set load_global_config & load_global_uconfig */
+ dw2 |= 0x8001;
+ /* set load_cs_sh_regs */
+ dw2 |= 0x01000000;
+ /* set load_per_context_state & load_gfx_sh_regs for GFX */
+ dw2 |= 0x10002;
+
+ /* set load_ce_ram if preamble presented */
+ if (AMDGPU_PREAMBLE_IB_PRESENT & flags)
+ dw2 |= 0x10000000;
+ } else {
+ /* still load_ce_ram if this is the first time preamble presented
+ * although there is no context switch happens.
+ */
+ if (AMDGPU_PREAMBLE_IB_PRESENT_FIRST & flags)
+ dw2 |= 0x10000000;
}
+
+ amdgpu_ring_write(ring, PACKET3(PACKET3_CONTEXT_CONTROL, 1));
+ amdgpu_ring_write(ring, dw2);
+ amdgpu_ring_write(ring, 0);
+}
+
+static unsigned gfx_v8_0_ring_get_emit_ib_size_gfx(struct amdgpu_ring *ring)
+{
+ return
+ 4; /* gfx_v8_0_ring_emit_ib_gfx */
+}
+
+static unsigned gfx_v8_0_ring_get_dma_frame_size_gfx(struct amdgpu_ring *ring)
+{
+ return
+ 20 + /* gfx_v8_0_ring_emit_gds_switch */
+ 7 + /* gfx_v8_0_ring_emit_hdp_flush */
+ 5 + /* gfx_v8_0_ring_emit_hdp_invalidate */
+ 6 + 6 + 6 +/* gfx_v8_0_ring_emit_fence_gfx x3 for user fence, vm fence */
+ 7 + /* gfx_v8_0_ring_emit_pipeline_sync */
+ 256 + 19 + /* gfx_v8_0_ring_emit_vm_flush */
+ 2 + /* gfx_v8_ring_emit_sb */
+ 3; /* gfx_v8_ring_emit_cntxcntl */
+}
+
+static unsigned gfx_v8_0_ring_get_emit_ib_size_compute(struct amdgpu_ring *ring)
+{
+ return
+ 4; /* gfx_v8_0_ring_emit_ib_compute */
+}
+
+static unsigned gfx_v8_0_ring_get_dma_frame_size_compute(struct amdgpu_ring *ring)
+{
+ return
+ 20 + /* gfx_v8_0_ring_emit_gds_switch */
+ 7 + /* gfx_v8_0_ring_emit_hdp_flush */
+ 5 + /* gfx_v8_0_ring_emit_hdp_invalidate */
+ 7 + /* gfx_v8_0_ring_emit_pipeline_sync */
+ 17 + /* gfx_v8_0_ring_emit_vm_flush */
+ 7 + 7 + 7; /* gfx_v8_0_ring_emit_fence_compute x3 for user fence, vm fence */
+}
+
+static void gfx_v8_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
+ enum amdgpu_interrupt_state state)
+{
+ WREG32_FIELD(CP_INT_CNTL_RING0, TIME_STAMP_INT_ENABLE,
+ state == AMDGPU_IRQ_STATE_DISABLE ? 0 : 1);
}
static void gfx_v8_0_set_compute_eop_interrupt_state(struct amdgpu_device *adev,
int me, int pipe,
enum amdgpu_interrupt_state state)
{
- u32 mec_int_cntl, mec_int_cntl_reg;
-
/*
* amdgpu controls only pipe 0 of MEC1. That's why this function only
* handles the setting of interrupts for this specific pipe. All other
if (me == 1) {
switch (pipe) {
case 0:
- mec_int_cntl_reg = mmCP_ME1_PIPE0_INT_CNTL;
break;
default:
DRM_DEBUG("invalid pipe %d\n", pipe);
return;
}
- switch (state) {
- case AMDGPU_IRQ_STATE_DISABLE:
- mec_int_cntl = RREG32(mec_int_cntl_reg);
- mec_int_cntl = REG_SET_FIELD(mec_int_cntl, CP_ME1_PIPE0_INT_CNTL,
- TIME_STAMP_INT_ENABLE, 0);
- WREG32(mec_int_cntl_reg, mec_int_cntl);
- break;
- case AMDGPU_IRQ_STATE_ENABLE:
- mec_int_cntl = RREG32(mec_int_cntl_reg);
- mec_int_cntl = REG_SET_FIELD(mec_int_cntl, CP_ME1_PIPE0_INT_CNTL,
- TIME_STAMP_INT_ENABLE, 1);
- WREG32(mec_int_cntl_reg, mec_int_cntl);
- break;
- default:
- break;
- }
+ WREG32_FIELD(CP_ME1_PIPE0_INT_CNTL, TIME_STAMP_INT_ENABLE,
+ state == AMDGPU_IRQ_STATE_DISABLE ? 0 : 1);
}
static int gfx_v8_0_set_priv_reg_fault_state(struct amdgpu_device *adev,
unsigned type,
enum amdgpu_interrupt_state state)
{
- u32 cp_int_cntl;
-
- switch (state) {
- case AMDGPU_IRQ_STATE_DISABLE:
- cp_int_cntl = RREG32(mmCP_INT_CNTL_RING0);
- cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0,
- PRIV_REG_INT_ENABLE, 0);
- WREG32(mmCP_INT_CNTL_RING0, cp_int_cntl);
- break;
- case AMDGPU_IRQ_STATE_ENABLE:
- cp_int_cntl = RREG32(mmCP_INT_CNTL_RING0);
- cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0,
- PRIV_REG_INT_ENABLE, 1);
- WREG32(mmCP_INT_CNTL_RING0, cp_int_cntl);
- break;
- default:
- break;
- }
+ WREG32_FIELD(CP_INT_CNTL_RING0, PRIV_REG_INT_ENABLE,
+ state == AMDGPU_IRQ_STATE_DISABLE ? 0 : 1);
return 0;
}
unsigned type,
enum amdgpu_interrupt_state state)
{
- u32 cp_int_cntl;
-
- switch (state) {
- case AMDGPU_IRQ_STATE_DISABLE:
- cp_int_cntl = RREG32(mmCP_INT_CNTL_RING0);
- cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0,
- PRIV_INSTR_INT_ENABLE, 0);
- WREG32(mmCP_INT_CNTL_RING0, cp_int_cntl);
- break;
- case AMDGPU_IRQ_STATE_ENABLE:
- cp_int_cntl = RREG32(mmCP_INT_CNTL_RING0);
- cp_int_cntl = REG_SET_FIELD(cp_int_cntl, CP_INT_CNTL_RING0,
- PRIV_INSTR_INT_ENABLE, 1);
- WREG32(mmCP_INT_CNTL_RING0, cp_int_cntl);
- break;
- default:
- break;
- }
+ WREG32_FIELD(CP_INT_CNTL_RING0, PRIV_INSTR_INT_ENABLE,
+ state == AMDGPU_IRQ_STATE_DISABLE ? 0 : 1);
return 0;
}
.resume = gfx_v8_0_resume,
.is_idle = gfx_v8_0_is_idle,
.wait_for_idle = gfx_v8_0_wait_for_idle,
+ .check_soft_reset = gfx_v8_0_check_soft_reset,
+ .pre_soft_reset = gfx_v8_0_pre_soft_reset,
.soft_reset = gfx_v8_0_soft_reset,
+ .post_soft_reset = gfx_v8_0_post_soft_reset,
.set_clockgating_state = gfx_v8_0_set_clockgating_state,
.set_powergating_state = gfx_v8_0_set_powergating_state,
};
static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = {
- .get_rptr = gfx_v8_0_ring_get_rptr_gfx,
+ .get_rptr = gfx_v8_0_ring_get_rptr,
.get_wptr = gfx_v8_0_ring_get_wptr_gfx,
.set_wptr = gfx_v8_0_ring_set_wptr_gfx,
.parse_cs = NULL,
.test_ib = gfx_v8_0_ring_test_ib,
.insert_nop = amdgpu_ring_insert_nop,
.pad_ib = amdgpu_ring_generic_pad_ib,
+ .emit_switch_buffer = gfx_v8_ring_emit_sb,
+ .emit_cntxcntl = gfx_v8_ring_emit_cntxcntl,
+ .get_emit_ib_size = gfx_v8_0_ring_get_emit_ib_size_gfx,
+ .get_dma_frame_size = gfx_v8_0_ring_get_dma_frame_size_gfx,
};
static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_compute = {
- .get_rptr = gfx_v8_0_ring_get_rptr_compute,
+ .get_rptr = gfx_v8_0_ring_get_rptr,
.get_wptr = gfx_v8_0_ring_get_wptr_compute,
.set_wptr = gfx_v8_0_ring_set_wptr_compute,
.parse_cs = NULL,
.test_ib = gfx_v8_0_ring_test_ib,
.insert_nop = amdgpu_ring_insert_nop,
.pad_ib = amdgpu_ring_generic_pad_ib,
+ .get_emit_ib_size = gfx_v8_0_ring_get_emit_ib_size_compute,
+ .get_dma_frame_size = gfx_v8_0_ring_get_dma_frame_size_compute,
};
static void gfx_v8_0_set_ring_funcs(struct amdgpu_device *adev)
{
u32 data, mask;
- data = RREG32(mmCC_GC_SHADER_ARRAY_CONFIG);
- data |= RREG32(mmGC_USER_SHADER_ARRAY_CONFIG);
-
- data &= CC_GC_SHADER_ARRAY_CONFIG__INACTIVE_CUS_MASK;
- data >>= CC_GC_SHADER_ARRAY_CONFIG__INACTIVE_CUS__SHIFT;
+ data = RREG32(mmCC_GC_SHADER_ARRAY_CONFIG) |
+ RREG32(mmGC_USER_SHADER_ARRAY_CONFIG);
mask = gfx_v8_0_create_bitmask(adev->gfx.config.max_cu_per_sh);
- return (~data) & mask;
+ return ~REG_GET_FIELD(data, CC_GC_SHADER_ARRAY_CONFIG, INACTIVE_CUS) & mask;
}
static void gfx_v8_0_get_cu_info(struct amdgpu_device *adev)
extern const struct amd_ip_funcs gfx_v8_0_ip_funcs;
-void gfx_v8_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, u32 sh_num);
-
#endif
--- /dev/null
+
+/*
+ * Copyright 2014 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include <linux/firmware.h>
+#include "drmP.h"
+#include "amdgpu.h"
+#include "gmc_v6_0.h"
+#include "amdgpu_ucode.h"
+#include "si/sid.h"
+
+static void gmc_v6_0_set_gart_funcs(struct amdgpu_device *adev);
+static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev);
+static int gmc_v6_0_wait_for_idle(void *handle);
+
+MODULE_FIRMWARE("radeon/tahiti_mc.bin");
+MODULE_FIRMWARE("radeon/pitcairn_mc.bin");
+MODULE_FIRMWARE("radeon/verde_mc.bin");
+MODULE_FIRMWARE("radeon/oland_mc.bin");
+
+static const u32 crtc_offsets[6] =
+{
+ SI_CRTC0_REGISTER_OFFSET,
+ SI_CRTC1_REGISTER_OFFSET,
+ SI_CRTC2_REGISTER_OFFSET,
+ SI_CRTC3_REGISTER_OFFSET,
+ SI_CRTC4_REGISTER_OFFSET,
+ SI_CRTC5_REGISTER_OFFSET
+};
+
+static void gmc_v6_0_mc_stop(struct amdgpu_device *adev,
+ struct amdgpu_mode_mc_save *save)
+{
+ u32 blackout;
+
+ if (adev->mode_info.num_crtc)
+ amdgpu_display_stop_mc_access(adev, save);
+
+ gmc_v6_0_wait_for_idle((void *)adev);
+
+ blackout = RREG32(MC_SHARED_BLACKOUT_CNTL);
+ if (REG_GET_FIELD(blackout, mmMC_SHARED_BLACKOUT_CNTL, xxBLACKOUT_MODE) != 1) {
+ /* Block CPU access */
+ WREG32(BIF_FB_EN, 0);
+ /* blackout the MC */
+ blackout = REG_SET_FIELD(blackout,
+ mmMC_SHARED_BLACKOUT_CNTL, xxBLACKOUT_MODE, 0);
+ WREG32(MC_SHARED_BLACKOUT_CNTL, blackout | 1);
+ }
+ /* wait for the MC to settle */
+ udelay(100);
+
+}
+
+static void gmc_v6_0_mc_resume(struct amdgpu_device *adev,
+ struct amdgpu_mode_mc_save *save)
+{
+ u32 tmp;
+
+ /* unblackout the MC */
+ tmp = RREG32(MC_SHARED_BLACKOUT_CNTL);
+ tmp = REG_SET_FIELD(tmp, mmMC_SHARED_BLACKOUT_CNTL, xxBLACKOUT_MODE, 0);
+ WREG32(MC_SHARED_BLACKOUT_CNTL, tmp);
+ /* allow CPU access */
+ tmp = REG_SET_FIELD(0, mmBIF_FB_EN, xxFB_READ_EN, 1);
+ tmp = REG_SET_FIELD(tmp, mmBIF_FB_EN, xxFB_WRITE_EN, 1);
+ WREG32(BIF_FB_EN, tmp);
+
+ if (adev->mode_info.num_crtc)
+ amdgpu_display_resume_mc_access(adev, save);
+
+}
+
+static int gmc_v6_0_init_microcode(struct amdgpu_device *adev)
+{
+ const char *chip_name;
+ char fw_name[30];
+ int err;
+
+ DRM_DEBUG("\n");
+
+ switch (adev->asic_type) {
+ case CHIP_TAHITI:
+ chip_name = "tahiti";
+ break;
+ case CHIP_PITCAIRN:
+ chip_name = "pitcairn";
+ break;
+ case CHIP_VERDE:
+ chip_name = "verde";
+ break;
+ case CHIP_OLAND:
+ chip_name = "oland";
+ break;
+ case CHIP_HAINAN:
+ chip_name = "hainan";
+ break;
+ default: BUG();
+ }
+
+ snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
+ err = request_firmware(&adev->mc.fw, fw_name, adev->dev);
+ if (err)
+ goto out;
+
+ err = amdgpu_ucode_validate(adev->mc.fw);
+
+out:
+ if (err) {
+ dev_err(adev->dev,
+ "si_mc: Failed to load firmware \"%s\"\n",
+ fw_name);
+ release_firmware(adev->mc.fw);
+ adev->mc.fw = NULL;
+ }
+ return err;
+}
+
+static int gmc_v6_0_mc_load_microcode(struct amdgpu_device *adev)
+{
+ const __le32 *new_fw_data = NULL;
+ u32 running;
+ const __le32 *new_io_mc_regs = NULL;
+ int i, regs_size, ucode_size;
+ const struct mc_firmware_header_v1_0 *hdr;
+
+ if (!adev->mc.fw)
+ return -EINVAL;
+
+ hdr = (const struct mc_firmware_header_v1_0 *)adev->mc.fw->data;
+
+ amdgpu_ucode_print_mc_hdr(&hdr->header);
+
+ adev->mc.fw_version = le32_to_cpu(hdr->header.ucode_version);
+ regs_size = le32_to_cpu(hdr->io_debug_size_bytes) / (4 * 2);
+ new_io_mc_regs = (const __le32 *)
+ (adev->mc.fw->data + le32_to_cpu(hdr->io_debug_array_offset_bytes));
+ ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes) / 4;
+ new_fw_data = (const __le32 *)
+ (adev->mc.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes));
+
+ running = RREG32(MC_SEQ_SUP_CNTL) & RUN_MASK;
+
+ if (running == 0) {
+
+ /* reset the engine and set to writable */
+ WREG32(MC_SEQ_SUP_CNTL, 0x00000008);
+ WREG32(MC_SEQ_SUP_CNTL, 0x00000010);
+
+ /* load mc io regs */
+ for (i = 0; i < regs_size; i++) {
+ WREG32(MC_SEQ_IO_DEBUG_INDEX, le32_to_cpup(new_io_mc_regs++));
+ WREG32(MC_SEQ_IO_DEBUG_DATA, le32_to_cpup(new_io_mc_regs++));
+ }
+ /* load the MC ucode */
+ for (i = 0; i < ucode_size; i++) {
+ WREG32(MC_SEQ_SUP_PGM, le32_to_cpup(new_fw_data++));
+ }
+
+ /* put the engine back into the active state */
+ WREG32(MC_SEQ_SUP_CNTL, 0x00000008);
+ WREG32(MC_SEQ_SUP_CNTL, 0x00000004);
+ WREG32(MC_SEQ_SUP_CNTL, 0x00000001);
+
+ /* wait for training to complete */
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (RREG32(MC_SEQ_TRAIN_WAKEUP_CNTL) & TRAIN_DONE_D0)
+ break;
+ udelay(1);
+ }
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (RREG32(MC_SEQ_TRAIN_WAKEUP_CNTL) & TRAIN_DONE_D1)
+ break;
+ udelay(1);
+ }
+
+ }
+
+ return 0;
+}
+
+static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev,
+ struct amdgpu_mc *mc)
+{
+ if (mc->mc_vram_size > 0xFFC0000000ULL) {
+ dev_warn(adev->dev, "limiting VRAM\n");
+ mc->real_vram_size = 0xFFC0000000ULL;
+ mc->mc_vram_size = 0xFFC0000000ULL;
+ }
+ amdgpu_vram_location(adev, &adev->mc, 0);
+ adev->mc.gtt_base_align = 0;
+ amdgpu_gtt_location(adev, mc);
+}
+
+static void gmc_v6_0_mc_program(struct amdgpu_device *adev)
+{
+ struct amdgpu_mode_mc_save save;
+ u32 tmp;
+ int i, j;
+
+ /* Initialize HDP */
+ for (i = 0, j = 0; i < 32; i++, j += 0x6) {
+ WREG32((0xb05 + j), 0x00000000);
+ WREG32((0xb06 + j), 0x00000000);
+ WREG32((0xb07 + j), 0x00000000);
+ WREG32((0xb08 + j), 0x00000000);
+ WREG32((0xb09 + j), 0x00000000);
+ }
+ WREG32(HDP_REG_COHERENCY_FLUSH_CNTL, 0);
+
+ gmc_v6_0_mc_stop(adev, &save);
+
+ if (gmc_v6_0_wait_for_idle((void *)adev)) {
+ dev_warn(adev->dev, "Wait for MC idle timedout !\n");
+ }
+
+ WREG32(VGA_HDP_CONTROL, VGA_MEMORY_DISABLE);
+ /* Update configuration */
+ WREG32(MC_VM_SYSTEM_APERTURE_LOW_ADDR,
+ adev->mc.vram_start >> 12);
+ WREG32(MC_VM_SYSTEM_APERTURE_HIGH_ADDR,
+ adev->mc.vram_end >> 12);
+ WREG32(MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR,
+ adev->vram_scratch.gpu_addr >> 12);
+ tmp = ((adev->mc.vram_end >> 24) & 0xFFFF) << 16;
+ tmp |= ((adev->mc.vram_start >> 24) & 0xFFFF);
+ WREG32(MC_VM_FB_LOCATION, tmp);
+ /* XXX double check these! */
+ WREG32(HDP_NONSURFACE_BASE, (adev->mc.vram_start >> 8));
+ WREG32(HDP_NONSURFACE_INFO, (2 << 7) | (1 << 30));
+ WREG32(HDP_NONSURFACE_SIZE, 0x3FFFFFFF);
+ WREG32(MC_VM_AGP_BASE, 0);
+ WREG32(MC_VM_AGP_TOP, 0x0FFFFFFF);
+ WREG32(MC_VM_AGP_BOT, 0x0FFFFFFF);
+
+ if (gmc_v6_0_wait_for_idle((void *)adev)) {
+ dev_warn(adev->dev, "Wait for MC idle timedout !\n");
+ }
+ gmc_v6_0_mc_resume(adev, &save);
+ amdgpu_display_set_vga_render_state(adev, false);
+}
+
+static int gmc_v6_0_mc_init(struct amdgpu_device *adev)
+{
+
+ u32 tmp;
+ int chansize, numchan;
+
+ tmp = RREG32(MC_ARB_RAMCFG);
+ if (tmp & CHANSIZE_OVERRIDE) {
+ chansize = 16;
+ } else if (tmp & CHANSIZE_MASK) {
+ chansize = 64;
+ } else {
+ chansize = 32;
+ }
+ tmp = RREG32(MC_SHARED_CHMAP);
+ switch ((tmp & NOOFCHAN_MASK) >> NOOFCHAN_SHIFT) {
+ case 0:
+ default:
+ numchan = 1;
+ break;
+ case 1:
+ numchan = 2;
+ break;
+ case 2:
+ numchan = 4;
+ break;
+ case 3:
+ numchan = 8;
+ break;
+ case 4:
+ numchan = 3;
+ break;
+ case 5:
+ numchan = 6;
+ break;
+ case 6:
+ numchan = 10;
+ break;
+ case 7:
+ numchan = 12;
+ break;
+ case 8:
+ numchan = 16;
+ break;
+ }
+ adev->mc.vram_width = numchan * chansize;
+ /* Could aper size report 0 ? */
+ adev->mc.aper_base = pci_resource_start(adev->pdev, 0);
+ adev->mc.aper_size = pci_resource_len(adev->pdev, 0);
+ /* size in MB on si */
+ adev->mc.mc_vram_size = RREG32(CONFIG_MEMSIZE) * 1024ULL * 1024ULL;
+ adev->mc.real_vram_size = RREG32(CONFIG_MEMSIZE) * 1024ULL * 1024ULL;
+ adev->mc.visible_vram_size = adev->mc.aper_size;
+
+ /* unless the user had overridden it, set the gart
+ * size equal to the 1024 or vram, whichever is larger.
+ */
+ if (amdgpu_gart_size == -1)
+ adev->mc.gtt_size = amdgpu_ttm_get_gtt_mem_size(adev);
+ else
+ adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20;
+
+ gmc_v6_0_vram_gtt_location(adev, &adev->mc);
+
+ return 0;
+}
+
+static void gmc_v6_0_gart_flush_gpu_tlb(struct amdgpu_device *adev,
+ uint32_t vmid)
+{
+ WREG32(HDP_MEM_COHERENCY_FLUSH_CNTL, 0);
+
+ WREG32(VM_INVALIDATE_REQUEST, 1 << vmid);
+}
+
+static int gmc_v6_0_gart_set_pte_pde(struct amdgpu_device *adev,
+ void *cpu_pt_addr,
+ uint32_t gpu_page_idx,
+ uint64_t addr,
+ uint32_t flags)
+{
+ void __iomem *ptr = (void *)cpu_pt_addr;
+ uint64_t value;
+
+ value = addr & 0xFFFFFFFFFFFFF000ULL;
+ value |= flags;
+ writeq(value, ptr + (gpu_page_idx * 8));
+
+ return 0;
+}
+
+static void gmc_v6_0_set_fault_enable_default(struct amdgpu_device *adev,
+ bool value)
+{
+ u32 tmp;
+
+ tmp = RREG32(VM_CONTEXT1_CNTL);
+ tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL,
+ xxRANGE_PROTECTION_FAULT_ENABLE_DEFAULT, value);
+ tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL,
+ xxDUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT, value);
+ tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL,
+ xxPDE0_PROTECTION_FAULT_ENABLE_DEFAULT, value);
+ tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL,
+ xxVALID_PROTECTION_FAULT_ENABLE_DEFAULT, value);
+ tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL,
+ xxREAD_PROTECTION_FAULT_ENABLE_DEFAULT, value);
+ tmp = REG_SET_FIELD(tmp, mmVM_CONTEXT1_CNTL,
+ xxWRITE_PROTECTION_FAULT_ENABLE_DEFAULT, value);
+ WREG32(VM_CONTEXT1_CNTL, tmp);
+}
+
+static int gmc_v6_0_gart_enable(struct amdgpu_device *adev)
+{
+ int r, i;
+
+ if (adev->gart.robj == NULL) {
+ dev_err(adev->dev, "No VRAM object for PCIE GART.\n");
+ return -EINVAL;
+ }
+ r = amdgpu_gart_table_vram_pin(adev);
+ if (r)
+ return r;
+ /* Setup TLB control */
+ WREG32(MC_VM_MX_L1_TLB_CNTL,
+ (0xA << 7) |
+ ENABLE_L1_TLB |
+ ENABLE_L1_FRAGMENT_PROCESSING |
+ SYSTEM_ACCESS_MODE_NOT_IN_SYS |
+ ENABLE_ADVANCED_DRIVER_MODEL |
+ SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU);
+ /* Setup L2 cache */
+ WREG32(VM_L2_CNTL, ENABLE_L2_CACHE |
+ ENABLE_L2_FRAGMENT_PROCESSING |
+ ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE |
+ ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE |
+ EFFECTIVE_L2_QUEUE_SIZE(7) |
+ CONTEXT1_IDENTITY_ACCESS_MODE(1));
+ WREG32(VM_L2_CNTL2, INVALIDATE_ALL_L1_TLBS | INVALIDATE_L2_CACHE);
+ WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY |
+ BANK_SELECT(4) |
+ L2_CACHE_BIGK_FRAGMENT_SIZE(4));
+ /* setup context0 */
+ WREG32(VM_CONTEXT0_PAGE_TABLE_START_ADDR, adev->mc.gtt_start >> 12);
+ WREG32(VM_CONTEXT0_PAGE_TABLE_END_ADDR, adev->mc.gtt_end >> 12);
+ WREG32(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR, adev->gart.table_addr >> 12);
+ WREG32(VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR,
+ (u32)(adev->dummy_page.addr >> 12));
+ WREG32(VM_CONTEXT0_CNTL2, 0);
+ WREG32(VM_CONTEXT0_CNTL, (ENABLE_CONTEXT | PAGE_TABLE_DEPTH(0) |
+ RANGE_PROTECTION_FAULT_ENABLE_DEFAULT));
+
+ WREG32(0x575, 0);
+ WREG32(0x576, 0);
+ WREG32(0x577, 0);
+
+ /* empty context1-15 */
+ /* set vm size, must be a multiple of 4 */
+ WREG32(VM_CONTEXT1_PAGE_TABLE_START_ADDR, 0);
+ WREG32(VM_CONTEXT1_PAGE_TABLE_END_ADDR, adev->vm_manager.max_pfn - 1);
+ /* Assign the pt base to something valid for now; the pts used for
+ * the VMs are determined by the application and setup and assigned
+ * on the fly in the vm part of radeon_gart.c
+ */
+ for (i = 1; i < 16; i++) {
+ if (i < 8)
+ WREG32(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + i,
+ adev->gart.table_addr >> 12);
+ else
+ WREG32(VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + i - 8,
+ adev->gart.table_addr >> 12);
+ }
+
+ /* enable context1-15 */
+ WREG32(VM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR,
+ (u32)(adev->dummy_page.addr >> 12));
+ WREG32(VM_CONTEXT1_CNTL2, 4);
+ WREG32(VM_CONTEXT1_CNTL, ENABLE_CONTEXT | PAGE_TABLE_DEPTH(1) |
+ PAGE_TABLE_BLOCK_SIZE(amdgpu_vm_block_size - 9) |
+ RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT |
+ RANGE_PROTECTION_FAULT_ENABLE_DEFAULT |
+ DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT |
+ DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT |
+ PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT |
+ PDE0_PROTECTION_FAULT_ENABLE_DEFAULT |
+ VALID_PROTECTION_FAULT_ENABLE_INTERRUPT |
+ VALID_PROTECTION_FAULT_ENABLE_DEFAULT |
+ READ_PROTECTION_FAULT_ENABLE_INTERRUPT |
+ READ_PROTECTION_FAULT_ENABLE_DEFAULT |
+ WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT |
+ WRITE_PROTECTION_FAULT_ENABLE_DEFAULT);
+
+ gmc_v6_0_gart_flush_gpu_tlb(adev, 0);
+ dev_info(adev->dev, "PCIE GART of %uM enabled (table at 0x%016llX).\n",
+ (unsigned)(adev->mc.gtt_size >> 20),
+ (unsigned long long)adev->gart.table_addr);
+ adev->gart.ready = true;
+ return 0;
+}
+
+static int gmc_v6_0_gart_init(struct amdgpu_device *adev)
+{
+ int r;
+
+ if (adev->gart.robj) {
+ dev_warn(adev->dev, "gmc_v6_0 PCIE GART already initialized\n");
+ return 0;
+ }
+ r = amdgpu_gart_init(adev);
+ if (r)
+ return r;
+ adev->gart.table_size = adev->gart.num_gpu_pages * 8;
+ return amdgpu_gart_table_vram_alloc(adev);
+}
+
+static void gmc_v6_0_gart_disable(struct amdgpu_device *adev)
+{
+ /*unsigned i;
+
+ for (i = 1; i < 16; ++i) {
+ uint32_t reg;
+ if (i < 8)
+ reg = VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + i ;
+ else
+ reg = VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (i - 8);
+ adev->vm_manager.saved_table_addr[i] = RREG32(reg);
+ }*/
+
+ /* Disable all tables */
+ WREG32(VM_CONTEXT0_CNTL, 0);
+ WREG32(VM_CONTEXT1_CNTL, 0);
+ /* Setup TLB control */
+ WREG32(MC_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE_NOT_IN_SYS |
+ SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU);
+ /* Setup L2 cache */
+ WREG32(VM_L2_CNTL, ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE |
+ ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE |
+ EFFECTIVE_L2_QUEUE_SIZE(7) |
+ CONTEXT1_IDENTITY_ACCESS_MODE(1));
+ WREG32(VM_L2_CNTL2, 0);
+ WREG32(VM_L2_CNTL3, L2_CACHE_BIGK_ASSOCIATIVITY |
+ L2_CACHE_BIGK_FRAGMENT_SIZE(0));
+ amdgpu_gart_table_vram_unpin(adev);
+}
+
+static void gmc_v6_0_gart_fini(struct amdgpu_device *adev)
+{
+ amdgpu_gart_table_vram_free(adev);
+ amdgpu_gart_fini(adev);
+}
+
+static int gmc_v6_0_vm_init(struct amdgpu_device *adev)
+{
+ /*
+ * number of VMs
+ * VMID 0 is reserved for System
+ * amdgpu graphics/compute will use VMIDs 1-7
+ * amdkfd will use VMIDs 8-15
+ */
+ adev->vm_manager.num_ids = AMDGPU_NUM_OF_VMIDS;
+ amdgpu_vm_manager_init(adev);
+
+ /* base offset of vram pages */
+ if (adev->flags & AMD_IS_APU) {
+ u64 tmp = RREG32(MC_VM_FB_OFFSET);
+ tmp <<= 22;
+ adev->vm_manager.vram_base_offset = tmp;
+ } else
+ adev->vm_manager.vram_base_offset = 0;
+
+ return 0;
+}
+
+static void gmc_v6_0_vm_fini(struct amdgpu_device *adev)
+{
+}
+
+static void gmc_v6_0_vm_decode_fault(struct amdgpu_device *adev,
+ u32 status, u32 addr, u32 mc_client)
+{
+ u32 mc_id;
+ u32 vmid = REG_GET_FIELD(status, mmVM_CONTEXT1_PROTECTION_FAULT_STATUS, xxVMID);
+ u32 protections = REG_GET_FIELD(status, mmVM_CONTEXT1_PROTECTION_FAULT_STATUS,
+ xxPROTECTIONS);
+ char block[5] = { mc_client >> 24, (mc_client >> 16) & 0xff,
+ (mc_client >> 8) & 0xff, mc_client & 0xff, 0 };
+
+ mc_id = REG_GET_FIELD(status, mmVM_CONTEXT1_PROTECTION_FAULT_STATUS,
+ xxMEMORY_CLIENT_ID);
+
+ dev_err(adev->dev, "VM fault (0x%02x, vmid %d) at page %u, %s from '%s' (0x%08x) (%d)\n",
+ protections, vmid, addr,
+ REG_GET_FIELD(status, mmVM_CONTEXT1_PROTECTION_FAULT_STATUS,
+ xxMEMORY_CLIENT_RW) ?
+ "write" : "read", block, mc_client, mc_id);
+}
+
+/*
+static const u32 mc_cg_registers[] = {
+ MC_HUB_MISC_HUB_CG,
+ MC_HUB_MISC_SIP_CG,
+ MC_HUB_MISC_VM_CG,
+ MC_XPB_CLK_GAT,
+ ATC_MISC_CG,
+ MC_CITF_MISC_WR_CG,
+ MC_CITF_MISC_RD_CG,
+ MC_CITF_MISC_VM_CG,
+ VM_L2_CG,
+};
+
+static const u32 mc_cg_ls_en[] = {
+ MC_HUB_MISC_HUB_CG__MEM_LS_ENABLE_MASK,
+ MC_HUB_MISC_SIP_CG__MEM_LS_ENABLE_MASK,
+ MC_HUB_MISC_VM_CG__MEM_LS_ENABLE_MASK,
+ MC_XPB_CLK_GAT__MEM_LS_ENABLE_MASK,
+ ATC_MISC_CG__MEM_LS_ENABLE_MASK,
+ MC_CITF_MISC_WR_CG__MEM_LS_ENABLE_MASK,
+ MC_CITF_MISC_RD_CG__MEM_LS_ENABLE_MASK,
+ MC_CITF_MISC_VM_CG__MEM_LS_ENABLE_MASK,
+ VM_L2_CG__MEM_LS_ENABLE_MASK,
+};
+
+static const u32 mc_cg_en[] = {
+ MC_HUB_MISC_HUB_CG__ENABLE_MASK,
+ MC_HUB_MISC_SIP_CG__ENABLE_MASK,
+ MC_HUB_MISC_VM_CG__ENABLE_MASK,
+ MC_XPB_CLK_GAT__ENABLE_MASK,
+ ATC_MISC_CG__ENABLE_MASK,
+ MC_CITF_MISC_WR_CG__ENABLE_MASK,
+ MC_CITF_MISC_RD_CG__ENABLE_MASK,
+ MC_CITF_MISC_VM_CG__ENABLE_MASK,
+ VM_L2_CG__ENABLE_MASK,
+};
+
+static void gmc_v6_0_enable_mc_ls(struct amdgpu_device *adev,
+ bool enable)
+{
+ int i;
+ u32 orig, data;
+
+ for (i = 0; i < ARRAY_SIZE(mc_cg_registers); i++) {
+ orig = data = RREG32(mc_cg_registers[i]);
+ if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_MC_LS))
+ data |= mc_cg_ls_en[i];
+ else
+ data &= ~mc_cg_ls_en[i];
+ if (data != orig)
+ WREG32(mc_cg_registers[i], data);
+ }
+}
+
+static void gmc_v6_0_enable_mc_mgcg(struct amdgpu_device *adev,
+ bool enable)
+{
+ int i;
+ u32 orig, data;
+
+ for (i = 0; i < ARRAY_SIZE(mc_cg_registers); i++) {
+ orig = data = RREG32(mc_cg_registers[i]);
+ if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_MC_MGCG))
+ data |= mc_cg_en[i];
+ else
+ data &= ~mc_cg_en[i];
+ if (data != orig)
+ WREG32(mc_cg_registers[i], data);
+ }
+}
+
+static void gmc_v6_0_enable_bif_mgls(struct amdgpu_device *adev,
+ bool enable)
+{
+ u32 orig, data;
+
+ orig = data = RREG32_PCIE(ixPCIE_CNTL2);
+
+ if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_BIF_LS)) {
+ data = REG_SET_FIELD(data, PCIE_CNTL2, SLV_MEM_LS_EN, 1);
+ data = REG_SET_FIELD(data, PCIE_CNTL2, MST_MEM_LS_EN, 1);
+ data = REG_SET_FIELD(data, PCIE_CNTL2, REPLAY_MEM_LS_EN, 1);
+ data = REG_SET_FIELD(data, PCIE_CNTL2, SLV_MEM_AGGRESSIVE_LS_EN, 1);
+ } else {
+ data = REG_SET_FIELD(data, PCIE_CNTL2, SLV_MEM_LS_EN, 0);
+ data = REG_SET_FIELD(data, PCIE_CNTL2, MST_MEM_LS_EN, 0);
+ data = REG_SET_FIELD(data, PCIE_CNTL2, REPLAY_MEM_LS_EN, 0);
+ data = REG_SET_FIELD(data, PCIE_CNTL2, SLV_MEM_AGGRESSIVE_LS_EN, 0);
+ }
+
+ if (orig != data)
+ WREG32_PCIE(ixPCIE_CNTL2, data);
+}
+
+static void gmc_v6_0_enable_hdp_mgcg(struct amdgpu_device *adev,
+ bool enable)
+{
+ u32 orig, data;
+
+ orig = data = RREG32(HDP_HOST_PATH_CNTL);
+
+ if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_HDP_MGCG))
+ data = REG_SET_FIELD(data, HDP_HOST_PATH_CNTL, CLOCK_GATING_DIS, 0);
+ else
+ data = REG_SET_FIELD(data, HDP_HOST_PATH_CNTL, CLOCK_GATING_DIS, 1);
+
+ if (orig != data)
+ WREG32(HDP_HOST_PATH_CNTL, data);
+}
+
+static void gmc_v6_0_enable_hdp_ls(struct amdgpu_device *adev,
+ bool enable)
+{
+ u32 orig, data;
+
+ orig = data = RREG32(HDP_MEM_POWER_LS);
+
+ if (enable && (adev->cg_flags & AMDGPU_CG_SUPPORT_HDP_LS))
+ data = REG_SET_FIELD(data, HDP_MEM_POWER_LS, LS_ENABLE, 1);
+ else
+ data = REG_SET_FIELD(data, HDP_MEM_POWER_LS, LS_ENABLE, 0);
+
+ if (orig != data)
+ WREG32(HDP_MEM_POWER_LS, data);
+}
+*/
+
+static int gmc_v6_0_convert_vram_type(int mc_seq_vram_type)
+{
+ switch (mc_seq_vram_type) {
+ case MC_SEQ_MISC0__MT__GDDR1:
+ return AMDGPU_VRAM_TYPE_GDDR1;
+ case MC_SEQ_MISC0__MT__DDR2:
+ return AMDGPU_VRAM_TYPE_DDR2;
+ case MC_SEQ_MISC0__MT__GDDR3:
+ return AMDGPU_VRAM_TYPE_GDDR3;
+ case MC_SEQ_MISC0__MT__GDDR4:
+ return AMDGPU_VRAM_TYPE_GDDR4;
+ case MC_SEQ_MISC0__MT__GDDR5:
+ return AMDGPU_VRAM_TYPE_GDDR5;
+ case MC_SEQ_MISC0__MT__DDR3:
+ return AMDGPU_VRAM_TYPE_DDR3;
+ default:
+ return AMDGPU_VRAM_TYPE_UNKNOWN;
+ }
+}
+
+static int gmc_v6_0_early_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ gmc_v6_0_set_gart_funcs(adev);
+ gmc_v6_0_set_irq_funcs(adev);
+
+ if (adev->flags & AMD_IS_APU) {
+ adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN;
+ } else {
+ u32 tmp = RREG32(MC_SEQ_MISC0);
+ tmp &= MC_SEQ_MISC0__MT__MASK;
+ adev->mc.vram_type = gmc_v6_0_convert_vram_type(tmp);
+ }
+
+ return 0;
+}
+
+static int gmc_v6_0_late_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return amdgpu_irq_get(adev, &adev->mc.vm_fault, 0);
+}
+
+static int gmc_v6_0_sw_init(void *handle)
+{
+ int r;
+ int dma_bits;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ r = amdgpu_irq_add_id(adev, 146, &adev->mc.vm_fault);
+ if (r)
+ return r;
+
+ r = amdgpu_irq_add_id(adev, 147, &adev->mc.vm_fault);
+ if (r)
+ return r;
+
+ adev->vm_manager.max_pfn = amdgpu_vm_size << 18;
+
+ adev->mc.mc_mask = 0xffffffffffULL;
+
+ adev->need_dma32 = false;
+ dma_bits = adev->need_dma32 ? 32 : 40;
+ r = pci_set_dma_mask(adev->pdev, DMA_BIT_MASK(dma_bits));
+ if (r) {
+ adev->need_dma32 = true;
+ dma_bits = 32;
+ dev_warn(adev->dev, "amdgpu: No suitable DMA available.\n");
+ }
+ r = pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(dma_bits));
+ if (r) {
+ pci_set_consistent_dma_mask(adev->pdev, DMA_BIT_MASK(32));
+ dev_warn(adev->dev, "amdgpu: No coherent DMA available.\n");
+ }
+
+ r = gmc_v6_0_init_microcode(adev);
+ if (r) {
+ dev_err(adev->dev, "Failed to load mc firmware!\n");
+ return r;
+ }
+
+ r = amdgpu_ttm_global_init(adev);
+ if (r) {
+ return r;
+ }
+
+ r = gmc_v6_0_mc_init(adev);
+ if (r)
+ return r;
+
+ r = amdgpu_bo_init(adev);
+ if (r)
+ return r;
+
+ r = gmc_v6_0_gart_init(adev);
+ if (r)
+ return r;
+
+ if (!adev->vm_manager.enabled) {
+ r = gmc_v6_0_vm_init(adev);
+ if (r) {
+ dev_err(adev->dev, "vm manager initialization failed (%d).\n", r);
+ return r;
+ }
+ adev->vm_manager.enabled = true;
+ }
+
+ return r;
+}
+
+static int gmc_v6_0_sw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (adev->vm_manager.enabled) {
+ gmc_v6_0_vm_fini(adev);
+ adev->vm_manager.enabled = false;
+ }
+ gmc_v6_0_gart_fini(adev);
+ amdgpu_gem_force_release(adev);
+ amdgpu_bo_fini(adev);
+
+ return 0;
+}
+
+static int gmc_v6_0_hw_init(void *handle)
+{
+ int r;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ gmc_v6_0_mc_program(adev);
+
+ if (!(adev->flags & AMD_IS_APU)) {
+ r = gmc_v6_0_mc_load_microcode(adev);
+ if (r) {
+ dev_err(adev->dev, "Failed to load MC firmware!\n");
+ return r;
+ }
+ }
+
+ r = gmc_v6_0_gart_enable(adev);
+ if (r)
+ return r;
+
+ return r;
+}
+
+static int gmc_v6_0_hw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ amdgpu_irq_put(adev, &adev->mc.vm_fault, 0);
+ gmc_v6_0_gart_disable(adev);
+
+ return 0;
+}
+
+static int gmc_v6_0_suspend(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (adev->vm_manager.enabled) {
+ gmc_v6_0_vm_fini(adev);
+ adev->vm_manager.enabled = false;
+ }
+ gmc_v6_0_hw_fini(adev);
+
+ return 0;
+}
+
+static int gmc_v6_0_resume(void *handle)
+{
+ int r;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ r = gmc_v6_0_hw_init(adev);
+ if (r)
+ return r;
+
+ if (!adev->vm_manager.enabled) {
+ r = gmc_v6_0_vm_init(adev);
+ if (r) {
+ dev_err(adev->dev, "vm manager initialization failed (%d).\n", r);
+ return r;
+ }
+ adev->vm_manager.enabled = true;
+ }
+
+ return r;
+}
+
+static bool gmc_v6_0_is_idle(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 tmp = RREG32(SRBM_STATUS);
+
+ if (tmp & (SRBM_STATUS__MCB_BUSY_MASK | SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK |
+ SRBM_STATUS__MCC_BUSY_MASK | SRBM_STATUS__MCD_BUSY_MASK | SRBM_STATUS__VMC_BUSY_MASK))
+ return false;
+
+ return true;
+}
+
+static int gmc_v6_0_wait_for_idle(void *handle)
+{
+ unsigned i;
+ u32 tmp;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ tmp = RREG32(SRBM_STATUS) & (SRBM_STATUS__MCB_BUSY_MASK |
+ SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK |
+ SRBM_STATUS__MCC_BUSY_MASK |
+ SRBM_STATUS__MCD_BUSY_MASK |
+ SRBM_STATUS__VMC_BUSY_MASK);
+ if (!tmp)
+ return 0;
+ udelay(1);
+ }
+ return -ETIMEDOUT;
+
+}
+
+static int gmc_v6_0_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ struct amdgpu_mode_mc_save save;
+ u32 srbm_soft_reset = 0;
+ u32 tmp = RREG32(SRBM_STATUS);
+
+ if (tmp & SRBM_STATUS__VMC_BUSY_MASK)
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset,
+ mmSRBM_SOFT_RESET, xxSOFT_RESET_VMC, 1);
+
+ if (tmp & (SRBM_STATUS__MCB_BUSY_MASK | SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK |
+ SRBM_STATUS__MCC_BUSY_MASK | SRBM_STATUS__MCD_BUSY_MASK)) {
+ if (!(adev->flags & AMD_IS_APU))
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset,
+ mmSRBM_SOFT_RESET, xxSOFT_RESET_MC, 1);
+ }
+
+ if (srbm_soft_reset) {
+ gmc_v6_0_mc_stop(adev, &save);
+ if (gmc_v6_0_wait_for_idle(adev)) {
+ dev_warn(adev->dev, "Wait for GMC idle timed out !\n");
+ }
+
+
+ tmp = RREG32(SRBM_SOFT_RESET);
+ tmp |= srbm_soft_reset;
+ dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp);
+ WREG32(SRBM_SOFT_RESET, tmp);
+ tmp = RREG32(SRBM_SOFT_RESET);
+
+ udelay(50);
+
+ tmp &= ~srbm_soft_reset;
+ WREG32(SRBM_SOFT_RESET, tmp);
+ tmp = RREG32(SRBM_SOFT_RESET);
+
+ udelay(50);
+
+ gmc_v6_0_mc_resume(adev, &save);
+ udelay(50);
+ }
+
+ return 0;
+}
+
+static int gmc_v6_0_vm_fault_interrupt_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ u32 tmp;
+ u32 bits = (VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
+ VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
+ VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
+ VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
+ VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK |
+ VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK);
+
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ tmp = RREG32(VM_CONTEXT0_CNTL);
+ tmp &= ~bits;
+ WREG32(VM_CONTEXT0_CNTL, tmp);
+ tmp = RREG32(VM_CONTEXT1_CNTL);
+ tmp &= ~bits;
+ WREG32(VM_CONTEXT1_CNTL, tmp);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ tmp = RREG32(VM_CONTEXT0_CNTL);
+ tmp |= bits;
+ WREG32(VM_CONTEXT0_CNTL, tmp);
+ tmp = RREG32(VM_CONTEXT1_CNTL);
+ tmp |= bits;
+ WREG32(VM_CONTEXT1_CNTL, tmp);
+ break;
+ default:
+ break;
+ }
+
+ return 0;
+}
+
+static int gmc_v6_0_process_interrupt(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ u32 addr, status;
+
+ addr = RREG32(VM_CONTEXT1_PROTECTION_FAULT_ADDR);
+ status = RREG32(VM_CONTEXT1_PROTECTION_FAULT_STATUS);
+ WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1);
+
+ if (!addr && !status)
+ return 0;
+
+ if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_FIRST)
+ gmc_v6_0_set_fault_enable_default(adev, false);
+
+ dev_err(adev->dev, "GPU fault detected: %d 0x%08x\n",
+ entry->src_id, entry->src_data);
+ dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x%08X\n",
+ addr);
+ dev_err(adev->dev, " VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n",
+ status);
+ gmc_v6_0_vm_decode_fault(adev, status, addr, 0);
+
+ return 0;
+}
+
+static int gmc_v6_0_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ return 0;
+}
+
+static int gmc_v6_0_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ return 0;
+}
+
+const struct amd_ip_funcs gmc_v6_0_ip_funcs = {
+ .name = "gmc_v6_0",
+ .early_init = gmc_v6_0_early_init,
+ .late_init = gmc_v6_0_late_init,
+ .sw_init = gmc_v6_0_sw_init,
+ .sw_fini = gmc_v6_0_sw_fini,
+ .hw_init = gmc_v6_0_hw_init,
+ .hw_fini = gmc_v6_0_hw_fini,
+ .suspend = gmc_v6_0_suspend,
+ .resume = gmc_v6_0_resume,
+ .is_idle = gmc_v6_0_is_idle,
+ .wait_for_idle = gmc_v6_0_wait_for_idle,
+ .soft_reset = gmc_v6_0_soft_reset,
+ .set_clockgating_state = gmc_v6_0_set_clockgating_state,
+ .set_powergating_state = gmc_v6_0_set_powergating_state,
+};
+
+static const struct amdgpu_gart_funcs gmc_v6_0_gart_funcs = {
+ .flush_gpu_tlb = gmc_v6_0_gart_flush_gpu_tlb,
+ .set_pte_pde = gmc_v6_0_gart_set_pte_pde,
+};
+
+static const struct amdgpu_irq_src_funcs gmc_v6_0_irq_funcs = {
+ .set = gmc_v6_0_vm_fault_interrupt_state,
+ .process = gmc_v6_0_process_interrupt,
+};
+
+static void gmc_v6_0_set_gart_funcs(struct amdgpu_device *adev)
+{
+ if (adev->gart.gart_funcs == NULL)
+ adev->gart.gart_funcs = &gmc_v6_0_gart_funcs;
+}
+
+static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev)
+{
+ adev->mc.vm_fault.num_types = 1;
+ adev->mc.vm_fault.funcs = &gmc_v6_0_irq_funcs;
+}
+
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __GMC_V6_0_H__
+#define __GMC_V6_0_H__
+
+extern const struct amd_ip_funcs gmc_v6_0_ip_funcs;
+
+#endif
const struct mc_firmware_header_v1_0 *hdr;
const __le32 *fw_data = NULL;
const __le32 *io_mc_regs = NULL;
- u32 running, blackout = 0;
+ u32 running;
int i, ucode_size, regs_size;
if (!adev->mc.fw)
running = REG_GET_FIELD(RREG32(mmMC_SEQ_SUP_CNTL), MC_SEQ_SUP_CNTL, RUN);
if (running == 0) {
- if (running) {
- blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL);
- WREG32(mmMC_SHARED_BLACKOUT_CNTL, blackout | 1);
- }
-
/* reset the engine and set to writable */
WREG32(mmMC_SEQ_SUP_CNTL, 0x00000008);
WREG32(mmMC_SEQ_SUP_CNTL, 0x00000010);
break;
udelay(1);
}
-
- if (running)
- WREG32(mmMC_SHARED_BLACKOUT_CNTL, blackout);
}
return 0;
* size equal to the 1024 or vram, whichever is larger.
*/
if (amdgpu_gart_size == -1)
- adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size);
+ adev->mc.gtt_size = amdgpu_ttm_get_gtt_mem_size(adev);
else
adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20;
return r;
}
+ r = amdgpu_ttm_global_init(adev);
+ if (r) {
+ return r;
+ }
+
r = gmc_v7_0_mc_init(adev);
if (r)
return r;
const struct mc_firmware_header_v1_0 *hdr;
const __le32 *fw_data = NULL;
const __le32 *io_mc_regs = NULL;
- u32 running, blackout = 0;
+ u32 running;
int i, ucode_size, regs_size;
if (!adev->mc.fw)
running = REG_GET_FIELD(RREG32(mmMC_SEQ_SUP_CNTL), MC_SEQ_SUP_CNTL, RUN);
if (running == 0) {
- if (running) {
- blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL);
- WREG32(mmMC_SHARED_BLACKOUT_CNTL, blackout | 1);
- }
-
/* reset the engine and set to writable */
WREG32(mmMC_SEQ_SUP_CNTL, 0x00000008);
WREG32(mmMC_SEQ_SUP_CNTL, 0x00000010);
break;
udelay(1);
}
-
- if (running)
- WREG32(mmMC_SHARED_BLACKOUT_CNTL, blackout);
}
return 0;
* size equal to the 1024 or vram, whichever is larger.
*/
if (amdgpu_gart_size == -1)
- adev->mc.gtt_size = max((1024ULL << 20), adev->mc.mc_vram_size);
+ adev->mc.gtt_size = amdgpu_ttm_get_gtt_mem_size(adev);
else
adev->mc.gtt_size = (uint64_t)amdgpu_gart_size << 20;
return r;
}
+ r = amdgpu_ttm_global_init(adev);
+ if (r) {
+ return r;
+ }
+
r = gmc_v8_0_mc_init(adev);
if (r)
return r;
}
-static int gmc_v8_0_soft_reset(void *handle)
+static int gmc_v8_0_check_soft_reset(void *handle)
{
- struct amdgpu_mode_mc_save save;
u32 srbm_soft_reset = 0;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
u32 tmp = RREG32(mmSRBM_STATUS);
srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset,
SRBM_SOFT_RESET, SOFT_RESET_MC, 1);
}
-
if (srbm_soft_reset) {
- gmc_v8_0_mc_stop(adev, &save);
- if (gmc_v8_0_wait_for_idle((void *)adev)) {
- dev_warn(adev->dev, "Wait for GMC idle timed out !\n");
- }
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_GMC].hang = true;
+ adev->mc.srbm_soft_reset = srbm_soft_reset;
+ } else {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_GMC].hang = false;
+ adev->mc.srbm_soft_reset = 0;
+ }
+ return 0;
+}
+static int gmc_v8_0_pre_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_GMC].hang)
+ return 0;
+
+ gmc_v8_0_mc_stop(adev, &adev->mc.save);
+ if (gmc_v8_0_wait_for_idle(adev)) {
+ dev_warn(adev->dev, "Wait for GMC idle timed out !\n");
+ }
+
+ return 0;
+}
+
+static int gmc_v8_0_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_GMC].hang)
+ return 0;
+ srbm_soft_reset = adev->mc.srbm_soft_reset;
+
+ if (srbm_soft_reset) {
+ u32 tmp;
tmp = RREG32(mmSRBM_SOFT_RESET);
tmp |= srbm_soft_reset;
/* Wait a little for things to settle down */
udelay(50);
-
- gmc_v8_0_mc_resume(adev, &save);
- udelay(50);
}
return 0;
}
+static int gmc_v8_0_post_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_GMC].hang)
+ return 0;
+
+ gmc_v8_0_mc_resume(adev, &adev->mc.save);
+ return 0;
+}
+
static int gmc_v8_0_vm_fault_interrupt_state(struct amdgpu_device *adev,
struct amdgpu_irq_src *src,
unsigned type,
.resume = gmc_v8_0_resume,
.is_idle = gmc_v8_0_is_idle,
.wait_for_idle = gmc_v8_0_wait_for_idle,
+ .check_soft_reset = gmc_v8_0_check_soft_reset,
+ .pre_soft_reset = gmc_v8_0_pre_soft_reset,
.soft_reset = gmc_v8_0_soft_reset,
+ .post_soft_reset = gmc_v8_0_post_soft_reset,
.set_clockgating_state = gmc_v8_0_set_clockgating_state,
.set_powergating_state = gmc_v8_0_set_powergating_state,
};
return result;
}
-void iceland_start_smc(struct amdgpu_device *adev)
+static void iceland_start_smc(struct amdgpu_device *adev)
{
uint32_t val = RREG32_SMC(ixSMC_SYSCON_RESET_CNTL);
WREG32_SMC(ixSMC_SYSCON_RESET_CNTL, val);
}
-void iceland_reset_smc(struct amdgpu_device *adev)
+static void iceland_reset_smc(struct amdgpu_device *adev)
{
uint32_t val = RREG32_SMC(ixSMC_SYSCON_RESET_CNTL);
return 0;
}
-void iceland_stop_smc_clock(struct amdgpu_device *adev)
+static void iceland_stop_smc_clock(struct amdgpu_device *adev)
{
uint32_t val = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0);
WREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0, val);
}
-void iceland_start_smc_clock(struct amdgpu_device *adev)
+static void iceland_start_smc_clock(struct amdgpu_device *adev)
{
uint32_t val = RREG32_SMC(ixSMC_SYSCON_CLOCK_CNTL_0);
pi->caps_tcp_ramping = true;
}
- pi->caps_sclk_ds = true;
+ if (amdgpu_sclk_deep_sleep_en)
+ pi->caps_sclk_ds = true;
+ else
+ pi->caps_sclk_ds = false;
+
pi->enable_auto_thermal_throttling = true;
pi->disable_nb_ps3_in_battery = false;
if (amdgpu_bapm == 0)
--- /dev/null
+/*
+ * Copyright 2011 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __R600_DPM_H__
+#define __R600_DPM_H__
+
+#define R600_ASI_DFLT 10000
+#define R600_BSP_DFLT 0x41EB
+#define R600_BSU_DFLT 0x2
+#define R600_AH_DFLT 5
+#define R600_RLP_DFLT 25
+#define R600_RMP_DFLT 65
+#define R600_LHP_DFLT 40
+#define R600_LMP_DFLT 15
+#define R600_TD_DFLT 0
+#define R600_UTC_DFLT_00 0x24
+#define R600_UTC_DFLT_01 0x22
+#define R600_UTC_DFLT_02 0x22
+#define R600_UTC_DFLT_03 0x22
+#define R600_UTC_DFLT_04 0x22
+#define R600_UTC_DFLT_05 0x22
+#define R600_UTC_DFLT_06 0x22
+#define R600_UTC_DFLT_07 0x22
+#define R600_UTC_DFLT_08 0x22
+#define R600_UTC_DFLT_09 0x22
+#define R600_UTC_DFLT_10 0x22
+#define R600_UTC_DFLT_11 0x22
+#define R600_UTC_DFLT_12 0x22
+#define R600_UTC_DFLT_13 0x22
+#define R600_UTC_DFLT_14 0x22
+#define R600_DTC_DFLT_00 0x24
+#define R600_DTC_DFLT_01 0x22
+#define R600_DTC_DFLT_02 0x22
+#define R600_DTC_DFLT_03 0x22
+#define R600_DTC_DFLT_04 0x22
+#define R600_DTC_DFLT_05 0x22
+#define R600_DTC_DFLT_06 0x22
+#define R600_DTC_DFLT_07 0x22
+#define R600_DTC_DFLT_08 0x22
+#define R600_DTC_DFLT_09 0x22
+#define R600_DTC_DFLT_10 0x22
+#define R600_DTC_DFLT_11 0x22
+#define R600_DTC_DFLT_12 0x22
+#define R600_DTC_DFLT_13 0x22
+#define R600_DTC_DFLT_14 0x22
+#define R600_VRC_DFLT 0x0000C003
+#define R600_VOLTAGERESPONSETIME_DFLT 1000
+#define R600_BACKBIASRESPONSETIME_DFLT 1000
+#define R600_VRU_DFLT 0x3
+#define R600_SPLLSTEPTIME_DFLT 0x1000
+#define R600_SPLLSTEPUNIT_DFLT 0x3
+#define R600_TPU_DFLT 0
+#define R600_TPC_DFLT 0x200
+#define R600_SSTU_DFLT 0
+#define R600_SST_DFLT 0x00C8
+#define R600_GICST_DFLT 0x200
+#define R600_FCT_DFLT 0x0400
+#define R600_FCTU_DFLT 0
+#define R600_CTXCGTT3DRPHC_DFLT 0x20
+#define R600_CTXCGTT3DRSDC_DFLT 0x40
+#define R600_VDDC3DOORPHC_DFLT 0x100
+#define R600_VDDC3DOORSDC_DFLT 0x7
+#define R600_VDDC3DOORSU_DFLT 0
+#define R600_MPLLLOCKTIME_DFLT 100
+#define R600_MPLLRESETTIME_DFLT 150
+#define R600_VCOSTEPPCT_DFLT 20
+#define R600_ENDINGVCOSTEPPCT_DFLT 5
+#define R600_REFERENCEDIVIDER_DFLT 4
+
+#define R600_PM_NUMBER_OF_TC 15
+#define R600_PM_NUMBER_OF_SCLKS 20
+#define R600_PM_NUMBER_OF_MCLKS 4
+#define R600_PM_NUMBER_OF_VOLTAGE_LEVELS 4
+#define R600_PM_NUMBER_OF_ACTIVITY_LEVELS 3
+
+/* XXX are these ok? */
+#define R600_TEMP_RANGE_MIN (90 * 1000)
+#define R600_TEMP_RANGE_MAX (120 * 1000)
+
+#define FDO_PWM_MODE_STATIC 1
+#define FDO_PWM_MODE_STATIC_RPM 5
+
+enum r600_power_level {
+ R600_POWER_LEVEL_LOW = 0,
+ R600_POWER_LEVEL_MEDIUM = 1,
+ R600_POWER_LEVEL_HIGH = 2,
+ R600_POWER_LEVEL_CTXSW = 3,
+};
+
+enum r600_td {
+ R600_TD_AUTO,
+ R600_TD_UP,
+ R600_TD_DOWN,
+};
+
+enum r600_display_watermark {
+ R600_DISPLAY_WATERMARK_LOW = 0,
+ R600_DISPLAY_WATERMARK_HIGH = 1,
+};
+
+enum r600_display_gap
+{
+ R600_PM_DISPLAY_GAP_VBLANK_OR_WM = 0,
+ R600_PM_DISPLAY_GAP_VBLANK = 1,
+ R600_PM_DISPLAY_GAP_WATERMARK = 2,
+ R600_PM_DISPLAY_GAP_IGNORE = 3,
+};
+#endif
*/
static uint32_t sdma_v2_4_ring_get_rptr(struct amdgpu_ring *ring)
{
- u32 rptr;
-
/* XXX check if swapping is necessary on BE */
- rptr = ring->adev->wb.wb[ring->rptr_offs] >> 2;
-
- return rptr;
+ return ring->adev->wb.wb[ring->rptr_offs] >> 2;
}
/**
uint64_t pe, uint64_t src,
unsigned count)
{
- while (count) {
- unsigned bytes = count * 8;
- if (bytes > 0x1FFFF8)
- bytes = 0x1FFFF8;
-
- ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) |
- SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR);
- ib->ptr[ib->length_dw++] = bytes;
- ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */
- ib->ptr[ib->length_dw++] = lower_32_bits(src);
- ib->ptr[ib->length_dw++] = upper_32_bits(src);
- ib->ptr[ib->length_dw++] = lower_32_bits(pe);
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
-
- pe += bytes;
- src += bytes;
- count -= bytes / 8;
- }
+ unsigned bytes = count * 8;
+
+ ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) |
+ SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR);
+ ib->ptr[ib->length_dw++] = bytes;
+ ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */
+ ib->ptr[ib->length_dw++] = lower_32_bits(src);
+ ib->ptr[ib->length_dw++] = upper_32_bits(src);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
}
/**
*
* @ib: indirect buffer to fill with commands
* @pe: addr of the page entry
- * @addr: dst addr to write into pe
+ * @value: dst addr to write into pe
* @count: number of page entries to update
* @incr: increase next addr by incr bytes
- * @flags: access flags
*
* Update PTEs by writing them manually using sDMA (CIK).
*/
-static void sdma_v2_4_vm_write_pte(struct amdgpu_ib *ib,
- const dma_addr_t *pages_addr, uint64_t pe,
- uint64_t addr, unsigned count,
- uint32_t incr, uint32_t flags)
+static void sdma_v2_4_vm_write_pte(struct amdgpu_ib *ib, uint64_t pe,
+ uint64_t value, unsigned count,
+ uint32_t incr)
{
- uint64_t value;
- unsigned ndw;
-
- while (count) {
- ndw = count * 2;
- if (ndw > 0xFFFFE)
- ndw = 0xFFFFE;
-
- /* for non-physically contiguous pages (system) */
- ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_WRITE) |
- SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR);
- ib->ptr[ib->length_dw++] = pe;
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
- ib->ptr[ib->length_dw++] = ndw;
- for (; ndw > 0; ndw -= 2, --count, pe += 8) {
- value = amdgpu_vm_map_gart(pages_addr, addr);
- addr += incr;
- value |= flags;
- ib->ptr[ib->length_dw++] = value;
- ib->ptr[ib->length_dw++] = upper_32_bits(value);
- }
+ unsigned ndw = count * 2;
+
+ ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_WRITE) |
+ SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR);
+ ib->ptr[ib->length_dw++] = pe;
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
+ ib->ptr[ib->length_dw++] = ndw;
+ for (; ndw > 0; ndw -= 2, --count, pe += 8) {
+ ib->ptr[ib->length_dw++] = lower_32_bits(value);
+ ib->ptr[ib->length_dw++] = upper_32_bits(value);
+ value += incr;
}
}
*
* Update the page tables using sDMA (CIK).
*/
-static void sdma_v2_4_vm_set_pte_pde(struct amdgpu_ib *ib,
- uint64_t pe,
+static void sdma_v2_4_vm_set_pte_pde(struct amdgpu_ib *ib, uint64_t pe,
uint64_t addr, unsigned count,
uint32_t incr, uint32_t flags)
{
- uint64_t value;
- unsigned ndw;
-
- while (count) {
- ndw = count;
- if (ndw > 0x7FFFF)
- ndw = 0x7FFFF;
-
- if (flags & AMDGPU_PTE_VALID)
- value = addr;
- else
- value = 0;
-
- /* for physically contiguous pages (vram) */
- ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_GEN_PTEPDE);
- ib->ptr[ib->length_dw++] = pe; /* dst addr */
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
- ib->ptr[ib->length_dw++] = flags; /* mask */
- ib->ptr[ib->length_dw++] = 0;
- ib->ptr[ib->length_dw++] = value; /* value */
- ib->ptr[ib->length_dw++] = upper_32_bits(value);
- ib->ptr[ib->length_dw++] = incr; /* increment size */
- ib->ptr[ib->length_dw++] = 0;
- ib->ptr[ib->length_dw++] = ndw; /* number of entries */
-
- pe += ndw * 8;
- addr += ndw * incr;
- count -= ndw;
- }
+ /* for physically contiguous pages (vram) */
+ ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_GEN_PTEPDE);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe); /* dst addr */
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
+ ib->ptr[ib->length_dw++] = flags; /* mask */
+ ib->ptr[ib->length_dw++] = 0;
+ ib->ptr[ib->length_dw++] = lower_32_bits(addr); /* value */
+ ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+ ib->ptr[ib->length_dw++] = incr; /* increment size */
+ ib->ptr[ib->length_dw++] = 0;
+ ib->ptr[ib->length_dw++] = count; /* number of entries */
}
/**
SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */
}
+static unsigned sdma_v2_4_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 7 + 6; /* sdma_v2_4_ring_emit_ib */
+}
+
+static unsigned sdma_v2_4_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 6 + /* sdma_v2_4_ring_emit_hdp_flush */
+ 3 + /* sdma_v2_4_ring_emit_hdp_invalidate */
+ 6 + /* sdma_v2_4_ring_emit_pipeline_sync */
+ 12 + /* sdma_v2_4_ring_emit_vm_flush */
+ 10 + 10 + 10; /* sdma_v2_4_ring_emit_fence x3 for user fence, vm fence */
+}
+
static int sdma_v2_4_early_init(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
.test_ib = sdma_v2_4_ring_test_ib,
.insert_nop = sdma_v2_4_ring_insert_nop,
.pad_ib = sdma_v2_4_ring_pad_ib,
+ .get_emit_ib_size = sdma_v2_4_ring_get_emit_ib_size,
+ .get_dma_frame_size = sdma_v2_4_ring_get_dma_frame_size,
};
static void sdma_v2_4_set_ring_funcs(struct amdgpu_device *adev)
*/
static uint32_t sdma_v3_0_ring_get_rptr(struct amdgpu_ring *ring)
{
- u32 rptr;
-
/* XXX check if swapping is necessary on BE */
- rptr = ring->adev->wb.wb[ring->rptr_offs] >> 2;
-
- return rptr;
+ return ring->adev->wb.wb[ring->rptr_offs] >> 2;
}
/**
amdgpu_ring_write(ring, SDMA_PKT_TRAP_INT_CONTEXT_INT_CONTEXT(0));
}
-unsigned init_cond_exec(struct amdgpu_ring *ring)
-{
- unsigned ret;
- amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_COND_EXE));
- amdgpu_ring_write(ring, lower_32_bits(ring->cond_exe_gpu_addr));
- amdgpu_ring_write(ring, upper_32_bits(ring->cond_exe_gpu_addr));
- amdgpu_ring_write(ring, 1);
- ret = ring->wptr;/* this is the offset we need patch later */
- amdgpu_ring_write(ring, 0x55aa55aa);/* insert dummy here and patch it later */
- return ret;
-}
-
-void patch_cond_exec(struct amdgpu_ring *ring, unsigned offset)
-{
- unsigned cur;
- BUG_ON(ring->ring[offset] != 0x55aa55aa);
-
- cur = ring->wptr - 1;
- if (likely(cur > offset))
- ring->ring[offset] = cur - offset;
- else
- ring->ring[offset] = (ring->ring_size>>2) - offset + cur;
-}
-
-
/**
* sdma_v3_0_gfx_stop - stop the gfx async dma engines
*
uint64_t pe, uint64_t src,
unsigned count)
{
- while (count) {
- unsigned bytes = count * 8;
- if (bytes > 0x1FFFF8)
- bytes = 0x1FFFF8;
-
- ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) |
- SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR);
- ib->ptr[ib->length_dw++] = bytes;
- ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */
- ib->ptr[ib->length_dw++] = lower_32_bits(src);
- ib->ptr[ib->length_dw++] = upper_32_bits(src);
- ib->ptr[ib->length_dw++] = lower_32_bits(pe);
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
-
- pe += bytes;
- src += bytes;
- count -= bytes / 8;
- }
+ unsigned bytes = count * 8;
+
+ ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_COPY) |
+ SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR);
+ ib->ptr[ib->length_dw++] = bytes;
+ ib->ptr[ib->length_dw++] = 0; /* src/dst endian swap */
+ ib->ptr[ib->length_dw++] = lower_32_bits(src);
+ ib->ptr[ib->length_dw++] = upper_32_bits(src);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
}
/**
*
* @ib: indirect buffer to fill with commands
* @pe: addr of the page entry
- * @addr: dst addr to write into pe
+ * @value: dst addr to write into pe
* @count: number of page entries to update
* @incr: increase next addr by incr bytes
- * @flags: access flags
*
* Update PTEs by writing them manually using sDMA (CIK).
*/
-static void sdma_v3_0_vm_write_pte(struct amdgpu_ib *ib,
- const dma_addr_t *pages_addr, uint64_t pe,
- uint64_t addr, unsigned count,
- uint32_t incr, uint32_t flags)
-{
- uint64_t value;
- unsigned ndw;
-
- while (count) {
- ndw = count * 2;
- if (ndw > 0xFFFFE)
- ndw = 0xFFFFE;
-
- /* for non-physically contiguous pages (system) */
- ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_WRITE) |
- SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR);
- ib->ptr[ib->length_dw++] = pe;
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
- ib->ptr[ib->length_dw++] = ndw;
- for (; ndw > 0; ndw -= 2, --count, pe += 8) {
- value = amdgpu_vm_map_gart(pages_addr, addr);
- addr += incr;
- value |= flags;
- ib->ptr[ib->length_dw++] = value;
- ib->ptr[ib->length_dw++] = upper_32_bits(value);
- }
+static void sdma_v3_0_vm_write_pte(struct amdgpu_ib *ib, uint64_t pe,
+ uint64_t value, unsigned count,
+ uint32_t incr)
+{
+ unsigned ndw = count * 2;
+
+ ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_WRITE) |
+ SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_COPY_LINEAR);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
+ ib->ptr[ib->length_dw++] = ndw;
+ for (; ndw > 0; ndw -= 2, --count, pe += 8) {
+ ib->ptr[ib->length_dw++] = lower_32_bits(value);
+ ib->ptr[ib->length_dw++] = upper_32_bits(value);
+ value += incr;
}
}
*
* Update the page tables using sDMA (CIK).
*/
-static void sdma_v3_0_vm_set_pte_pde(struct amdgpu_ib *ib,
- uint64_t pe,
+static void sdma_v3_0_vm_set_pte_pde(struct amdgpu_ib *ib, uint64_t pe,
uint64_t addr, unsigned count,
uint32_t incr, uint32_t flags)
{
- uint64_t value;
- unsigned ndw;
-
- while (count) {
- ndw = count;
- if (ndw > 0x7FFFF)
- ndw = 0x7FFFF;
-
- if (flags & AMDGPU_PTE_VALID)
- value = addr;
- else
- value = 0;
-
- /* for physically contiguous pages (vram) */
- ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_GEN_PTEPDE);
- ib->ptr[ib->length_dw++] = pe; /* dst addr */
- ib->ptr[ib->length_dw++] = upper_32_bits(pe);
- ib->ptr[ib->length_dw++] = flags; /* mask */
- ib->ptr[ib->length_dw++] = 0;
- ib->ptr[ib->length_dw++] = value; /* value */
- ib->ptr[ib->length_dw++] = upper_32_bits(value);
- ib->ptr[ib->length_dw++] = incr; /* increment size */
- ib->ptr[ib->length_dw++] = 0;
- ib->ptr[ib->length_dw++] = ndw; /* number of entries */
-
- pe += ndw * 8;
- addr += ndw * incr;
- count -= ndw;
- }
+ /* for physically contiguous pages (vram) */
+ ib->ptr[ib->length_dw++] = SDMA_PKT_HEADER_OP(SDMA_OP_GEN_PTEPDE);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe); /* dst addr */
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
+ ib->ptr[ib->length_dw++] = flags; /* mask */
+ ib->ptr[ib->length_dw++] = 0;
+ ib->ptr[ib->length_dw++] = lower_32_bits(addr); /* value */
+ ib->ptr[ib->length_dw++] = upper_32_bits(addr);
+ ib->ptr[ib->length_dw++] = incr; /* increment size */
+ ib->ptr[ib->length_dw++] = 0;
+ ib->ptr[ib->length_dw++] = count; /* number of entries */
}
/**
SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */
}
+static unsigned sdma_v3_0_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 7 + 6; /* sdma_v3_0_ring_emit_ib */
+}
+
+static unsigned sdma_v3_0_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 6 + /* sdma_v3_0_ring_emit_hdp_flush */
+ 3 + /* sdma_v3_0_ring_emit_hdp_invalidate */
+ 6 + /* sdma_v3_0_ring_emit_pipeline_sync */
+ 12 + /* sdma_v3_0_ring_emit_vm_flush */
+ 10 + 10 + 10; /* sdma_v3_0_ring_emit_fence x3 for user fence, vm fence */
+}
+
static int sdma_v3_0_early_init(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
return -ETIMEDOUT;
}
-static int sdma_v3_0_soft_reset(void *handle)
+static int sdma_v3_0_check_soft_reset(void *handle)
{
- u32 srbm_soft_reset = 0;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset = 0;
u32 tmp = RREG32(mmSRBM_STATUS2);
- if (tmp & SRBM_STATUS2__SDMA_BUSY_MASK) {
- /* sdma0 */
- tmp = RREG32(mmSDMA0_F32_CNTL + SDMA0_REGISTER_OFFSET);
- tmp = REG_SET_FIELD(tmp, SDMA0_F32_CNTL, HALT, 0);
- WREG32(mmSDMA0_F32_CNTL + SDMA0_REGISTER_OFFSET, tmp);
+ if ((tmp & SRBM_STATUS2__SDMA_BUSY_MASK) ||
+ (tmp & SRBM_STATUS2__SDMA1_BUSY_MASK)) {
srbm_soft_reset |= SRBM_SOFT_RESET__SOFT_RESET_SDMA_MASK;
- }
- if (tmp & SRBM_STATUS2__SDMA1_BUSY_MASK) {
- /* sdma1 */
- tmp = RREG32(mmSDMA0_F32_CNTL + SDMA1_REGISTER_OFFSET);
- tmp = REG_SET_FIELD(tmp, SDMA0_F32_CNTL, HALT, 0);
- WREG32(mmSDMA0_F32_CNTL + SDMA1_REGISTER_OFFSET, tmp);
srbm_soft_reset |= SRBM_SOFT_RESET__SOFT_RESET_SDMA1_MASK;
}
+ if (srbm_soft_reset) {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_SDMA].hang = true;
+ adev->sdma.srbm_soft_reset = srbm_soft_reset;
+ } else {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_SDMA].hang = false;
+ adev->sdma.srbm_soft_reset = 0;
+ }
+
+ return 0;
+}
+
+static int sdma_v3_0_pre_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset = 0;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_SDMA].hang)
+ return 0;
+
+ srbm_soft_reset = adev->sdma.srbm_soft_reset;
+
+ if (REG_GET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_SDMA) ||
+ REG_GET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_SDMA1)) {
+ sdma_v3_0_ctx_switch_enable(adev, false);
+ sdma_v3_0_enable(adev, false);
+ }
+
+ return 0;
+}
+
+static int sdma_v3_0_post_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset = 0;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_SDMA].hang)
+ return 0;
+
+ srbm_soft_reset = adev->sdma.srbm_soft_reset;
+
+ if (REG_GET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_SDMA) ||
+ REG_GET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_SDMA1)) {
+ sdma_v3_0_gfx_resume(adev);
+ sdma_v3_0_rlc_resume(adev);
+ }
+
+ return 0;
+}
+
+static int sdma_v3_0_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset = 0;
+ u32 tmp;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_SDMA].hang)
+ return 0;
+
+ srbm_soft_reset = adev->sdma.srbm_soft_reset;
+
if (srbm_soft_reset) {
tmp = RREG32(mmSRBM_SOFT_RESET);
tmp |= srbm_soft_reset;
.resume = sdma_v3_0_resume,
.is_idle = sdma_v3_0_is_idle,
.wait_for_idle = sdma_v3_0_wait_for_idle,
+ .check_soft_reset = sdma_v3_0_check_soft_reset,
+ .pre_soft_reset = sdma_v3_0_pre_soft_reset,
+ .post_soft_reset = sdma_v3_0_post_soft_reset,
.soft_reset = sdma_v3_0_soft_reset,
.set_clockgating_state = sdma_v3_0_set_clockgating_state,
.set_powergating_state = sdma_v3_0_set_powergating_state,
.test_ib = sdma_v3_0_ring_test_ib,
.insert_nop = sdma_v3_0_ring_insert_nop,
.pad_ib = sdma_v3_0_ring_pad_ib,
+ .get_emit_ib_size = sdma_v3_0_ring_get_emit_ib_size,
+ .get_dma_frame_size = sdma_v3_0_ring_get_dma_frame_size,
};
static void sdma_v3_0_set_ring_funcs(struct amdgpu_device *adev)
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/firmware.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include "drmP.h"
+#include "amdgpu.h"
+#include "amdgpu_atombios.h"
+#include "amdgpu_ih.h"
+#include "amdgpu_uvd.h"
+#include "amdgpu_vce.h"
+#include "atom.h"
+#include "amdgpu_powerplay.h"
+#include "si/sid.h"
+#include "si_ih.h"
+#include "gfx_v6_0.h"
+#include "gmc_v6_0.h"
+#include "si_dma.h"
+#include "dce_v6_0.h"
+#include "si.h"
+
+static const u32 tahiti_golden_registers[] =
+{
+ 0x2684, 0x00010000, 0x00018208,
+ 0x260c, 0xffffffff, 0x00000000,
+ 0x260d, 0xf00fffff, 0x00000400,
+ 0x260e, 0x0002021c, 0x00020200,
+ 0x031e, 0x00000080, 0x00000000,
+ 0x340c, 0x000300c0, 0x00800040,
+ 0x360c, 0x000300c0, 0x00800040,
+ 0x16ec, 0x000000f0, 0x00000070,
+ 0x16f0, 0x00200000, 0x50100000,
+ 0x1c0c, 0x31000311, 0x00000011,
+ 0x09df, 0x00000003, 0x000007ff,
+ 0x0903, 0x000007ff, 0x00000000,
+ 0x2285, 0xf000001f, 0x00000007,
+ 0x22c9, 0xffffffff, 0x00ffffff,
+ 0x22c4, 0x0000ff0f, 0x00000000,
+ 0xa293, 0x07ffffff, 0x4e000000,
+ 0xa0d4, 0x3f3f3fff, 0x2a00126a,
+ 0x000c, 0x000000ff, 0x0040,
+ 0x000d, 0x00000040, 0x00004040,
+ 0x2440, 0x07ffffff, 0x03000000,
+ 0x23a2, 0x01ff1f3f, 0x00000000,
+ 0x23a1, 0x01ff1f3f, 0x00000000,
+ 0x2418, 0x0000007f, 0x00000020,
+ 0x2542, 0x00010000, 0x00010000,
+ 0x2b05, 0x00000200, 0x000002fb,
+ 0x2b04, 0xffffffff, 0x0000543b,
+ 0x2b03, 0xffffffff, 0xa9210876,
+ 0x2234, 0xffffffff, 0x000fff40,
+ 0x2235, 0x0000001f, 0x00000010,
+ 0x0504, 0x20000000, 0x20fffed8,
+ 0x0570, 0x000c0fc0, 0x000c0400
+};
+
+static const u32 tahiti_golden_registers2[] =
+{
+ 0x0319, 0x00000001, 0x00000001
+};
+
+static const u32 tahiti_golden_rlc_registers[] =
+{
+ 0x3109, 0xffffffff, 0x00601005,
+ 0x311f, 0xffffffff, 0x10104040,
+ 0x3122, 0xffffffff, 0x0100000a,
+ 0x30c5, 0xffffffff, 0x00000800,
+ 0x30c3, 0xffffffff, 0x800000f4,
+ 0x3d2a, 0xffffffff, 0x00000000
+};
+
+static const u32 pitcairn_golden_registers[] =
+{
+ 0x2684, 0x00010000, 0x00018208,
+ 0x260c, 0xffffffff, 0x00000000,
+ 0x260d, 0xf00fffff, 0x00000400,
+ 0x260e, 0x0002021c, 0x00020200,
+ 0x031e, 0x00000080, 0x00000000,
+ 0x340c, 0x000300c0, 0x00800040,
+ 0x360c, 0x000300c0, 0x00800040,
+ 0x16ec, 0x000000f0, 0x00000070,
+ 0x16f0, 0x00200000, 0x50100000,
+ 0x1c0c, 0x31000311, 0x00000011,
+ 0x0ab9, 0x00073ffe, 0x000022a2,
+ 0x0903, 0x000007ff, 0x00000000,
+ 0x2285, 0xf000001f, 0x00000007,
+ 0x22c9, 0xffffffff, 0x00ffffff,
+ 0x22c4, 0x0000ff0f, 0x00000000,
+ 0xa293, 0x07ffffff, 0x4e000000,
+ 0xa0d4, 0x3f3f3fff, 0x2a00126a,
+ 0x000c, 0x000000ff, 0x0040,
+ 0x000d, 0x00000040, 0x00004040,
+ 0x2440, 0x07ffffff, 0x03000000,
+ 0x2418, 0x0000007f, 0x00000020,
+ 0x2542, 0x00010000, 0x00010000,
+ 0x2b05, 0x000003ff, 0x000000f7,
+ 0x2b04, 0xffffffff, 0x00000000,
+ 0x2b03, 0xffffffff, 0x32761054,
+ 0x2235, 0x0000001f, 0x00000010,
+ 0x0570, 0x000c0fc0, 0x000c0400
+};
+
+static const u32 pitcairn_golden_rlc_registers[] =
+{
+ 0x3109, 0xffffffff, 0x00601004,
+ 0x311f, 0xffffffff, 0x10102020,
+ 0x3122, 0xffffffff, 0x01000020,
+ 0x30c5, 0xffffffff, 0x00000800,
+ 0x30c3, 0xffffffff, 0x800000a4
+};
+
+static const u32 verde_pg_init[] =
+{
+ 0xd4f, 0xffffffff, 0x40000,
+ 0xd4e, 0xffffffff, 0x200010ff,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x7007,
+ 0xd4e, 0xffffffff, 0x300010ff,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x400000,
+ 0xd4e, 0xffffffff, 0x100010ff,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x120200,
+ 0xd4e, 0xffffffff, 0x500010ff,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x1e1e16,
+ 0xd4e, 0xffffffff, 0x600010ff,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x171f1e,
+ 0xd4e, 0xffffffff, 0x700010ff,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4f, 0xffffffff, 0x0,
+ 0xd4e, 0xffffffff, 0x9ff,
+ 0xd40, 0xffffffff, 0x0,
+ 0xd41, 0xffffffff, 0x10000800,
+ 0xd41, 0xffffffff, 0xf,
+ 0xd41, 0xffffffff, 0xf,
+ 0xd40, 0xffffffff, 0x4,
+ 0xd41, 0xffffffff, 0x1000051e,
+ 0xd41, 0xffffffff, 0xffff,
+ 0xd41, 0xffffffff, 0xffff,
+ 0xd40, 0xffffffff, 0x8,
+ 0xd41, 0xffffffff, 0x80500,
+ 0xd40, 0xffffffff, 0x12,
+ 0xd41, 0xffffffff, 0x9050c,
+ 0xd40, 0xffffffff, 0x1d,
+ 0xd41, 0xffffffff, 0xb052c,
+ 0xd40, 0xffffffff, 0x2a,
+ 0xd41, 0xffffffff, 0x1053e,
+ 0xd40, 0xffffffff, 0x2d,
+ 0xd41, 0xffffffff, 0x10546,
+ 0xd40, 0xffffffff, 0x30,
+ 0xd41, 0xffffffff, 0xa054e,
+ 0xd40, 0xffffffff, 0x3c,
+ 0xd41, 0xffffffff, 0x1055f,
+ 0xd40, 0xffffffff, 0x3f,
+ 0xd41, 0xffffffff, 0x10567,
+ 0xd40, 0xffffffff, 0x42,
+ 0xd41, 0xffffffff, 0x1056f,
+ 0xd40, 0xffffffff, 0x45,
+ 0xd41, 0xffffffff, 0x10572,
+ 0xd40, 0xffffffff, 0x48,
+ 0xd41, 0xffffffff, 0x20575,
+ 0xd40, 0xffffffff, 0x4c,
+ 0xd41, 0xffffffff, 0x190801,
+ 0xd40, 0xffffffff, 0x67,
+ 0xd41, 0xffffffff, 0x1082a,
+ 0xd40, 0xffffffff, 0x6a,
+ 0xd41, 0xffffffff, 0x1b082d,
+ 0xd40, 0xffffffff, 0x87,
+ 0xd41, 0xffffffff, 0x310851,
+ 0xd40, 0xffffffff, 0xba,
+ 0xd41, 0xffffffff, 0x891,
+ 0xd40, 0xffffffff, 0xbc,
+ 0xd41, 0xffffffff, 0x893,
+ 0xd40, 0xffffffff, 0xbe,
+ 0xd41, 0xffffffff, 0x20895,
+ 0xd40, 0xffffffff, 0xc2,
+ 0xd41, 0xffffffff, 0x20899,
+ 0xd40, 0xffffffff, 0xc6,
+ 0xd41, 0xffffffff, 0x2089d,
+ 0xd40, 0xffffffff, 0xca,
+ 0xd41, 0xffffffff, 0x8a1,
+ 0xd40, 0xffffffff, 0xcc,
+ 0xd41, 0xffffffff, 0x8a3,
+ 0xd40, 0xffffffff, 0xce,
+ 0xd41, 0xffffffff, 0x308a5,
+ 0xd40, 0xffffffff, 0xd3,
+ 0xd41, 0xffffffff, 0x6d08cd,
+ 0xd40, 0xffffffff, 0x142,
+ 0xd41, 0xffffffff, 0x2000095a,
+ 0xd41, 0xffffffff, 0x1,
+ 0xd40, 0xffffffff, 0x144,
+ 0xd41, 0xffffffff, 0x301f095b,
+ 0xd40, 0xffffffff, 0x165,
+ 0xd41, 0xffffffff, 0xc094d,
+ 0xd40, 0xffffffff, 0x173,
+ 0xd41, 0xffffffff, 0xf096d,
+ 0xd40, 0xffffffff, 0x184,
+ 0xd41, 0xffffffff, 0x15097f,
+ 0xd40, 0xffffffff, 0x19b,
+ 0xd41, 0xffffffff, 0xc0998,
+ 0xd40, 0xffffffff, 0x1a9,
+ 0xd41, 0xffffffff, 0x409a7,
+ 0xd40, 0xffffffff, 0x1af,
+ 0xd41, 0xffffffff, 0xcdc,
+ 0xd40, 0xffffffff, 0x1b1,
+ 0xd41, 0xffffffff, 0x800,
+ 0xd42, 0xffffffff, 0x6c9b2000,
+ 0xd44, 0xfc00, 0x2000,
+ 0xd51, 0xffffffff, 0xfc0,
+ 0xa35, 0x00000100, 0x100
+};
+
+static const u32 verde_golden_rlc_registers[] =
+{
+ 0x3109, 0xffffffff, 0x033f1005,
+ 0x311f, 0xffffffff, 0x10808020,
+ 0x3122, 0xffffffff, 0x00800008,
+ 0x30c5, 0xffffffff, 0x00001000,
+ 0x30c3, 0xffffffff, 0x80010014
+};
+
+static const u32 verde_golden_registers[] =
+{
+ 0x2684, 0x00010000, 0x00018208,
+ 0x260c, 0xffffffff, 0x00000000,
+ 0x260d, 0xf00fffff, 0x00000400,
+ 0x260e, 0x0002021c, 0x00020200,
+ 0x031e, 0x00000080, 0x00000000,
+ 0x340c, 0x000300c0, 0x00800040,
+ 0x340c, 0x000300c0, 0x00800040,
+ 0x360c, 0x000300c0, 0x00800040,
+ 0x360c, 0x000300c0, 0x00800040,
+ 0x16ec, 0x000000f0, 0x00000070,
+ 0x16f0, 0x00200000, 0x50100000,
+
+ 0x1c0c, 0x31000311, 0x00000011,
+ 0x0ab9, 0x00073ffe, 0x000022a2,
+ 0x0ab9, 0x00073ffe, 0x000022a2,
+ 0x0ab9, 0x00073ffe, 0x000022a2,
+ 0x0903, 0x000007ff, 0x00000000,
+ 0x0903, 0x000007ff, 0x00000000,
+ 0x0903, 0x000007ff, 0x00000000,
+ 0x2285, 0xf000001f, 0x00000007,
+ 0x2285, 0xf000001f, 0x00000007,
+ 0x2285, 0xf000001f, 0x00000007,
+ 0x2285, 0xffffffff, 0x00ffffff,
+ 0x22c4, 0x0000ff0f, 0x00000000,
+
+ 0xa293, 0x07ffffff, 0x4e000000,
+ 0xa0d4, 0x3f3f3fff, 0x0000124a,
+ 0xa0d4, 0x3f3f3fff, 0x0000124a,
+ 0xa0d4, 0x3f3f3fff, 0x0000124a,
+ 0x000c, 0x000000ff, 0x0040,
+ 0x000d, 0x00000040, 0x00004040,
+ 0x2440, 0x07ffffff, 0x03000000,
+ 0x2440, 0x07ffffff, 0x03000000,
+ 0x23a2, 0x01ff1f3f, 0x00000000,
+ 0x23a3, 0x01ff1f3f, 0x00000000,
+ 0x23a2, 0x01ff1f3f, 0x00000000,
+ 0x23a1, 0x01ff1f3f, 0x00000000,
+ 0x23a1, 0x01ff1f3f, 0x00000000,
+
+ 0x23a1, 0x01ff1f3f, 0x00000000,
+ 0x2418, 0x0000007f, 0x00000020,
+ 0x2542, 0x00010000, 0x00010000,
+ 0x2b01, 0x000003ff, 0x00000003,
+ 0x2b05, 0x000003ff, 0x00000003,
+ 0x2b05, 0x000003ff, 0x00000003,
+ 0x2b04, 0xffffffff, 0x00000000,
+ 0x2b04, 0xffffffff, 0x00000000,
+ 0x2b04, 0xffffffff, 0x00000000,
+ 0x2b03, 0xffffffff, 0x00001032,
+ 0x2b03, 0xffffffff, 0x00001032,
+ 0x2b03, 0xffffffff, 0x00001032,
+ 0x2235, 0x0000001f, 0x00000010,
+ 0x2235, 0x0000001f, 0x00000010,
+ 0x2235, 0x0000001f, 0x00000010,
+ 0x0570, 0x000c0fc0, 0x000c0400
+};
+
+static const u32 oland_golden_registers[] =
+{
+ 0x2684, 0x00010000, 0x00018208,
+ 0x260c, 0xffffffff, 0x00000000,
+ 0x260d, 0xf00fffff, 0x00000400,
+ 0x260e, 0x0002021c, 0x00020200,
+ 0x031e, 0x00000080, 0x00000000,
+ 0x340c, 0x000300c0, 0x00800040,
+ 0x360c, 0x000300c0, 0x00800040,
+ 0x16ec, 0x000000f0, 0x00000070,
+ 0x16f9, 0x00200000, 0x50100000,
+ 0x1c0c, 0x31000311, 0x00000011,
+ 0x0ab9, 0x00073ffe, 0x000022a2,
+ 0x0903, 0x000007ff, 0x00000000,
+ 0x2285, 0xf000001f, 0x00000007,
+ 0x22c9, 0xffffffff, 0x00ffffff,
+ 0x22c4, 0x0000ff0f, 0x00000000,
+ 0xa293, 0x07ffffff, 0x4e000000,
+ 0xa0d4, 0x3f3f3fff, 0x00000082,
+ 0x000c, 0x000000ff, 0x0040,
+ 0x000d, 0x00000040, 0x00004040,
+ 0x2440, 0x07ffffff, 0x03000000,
+ 0x2418, 0x0000007f, 0x00000020,
+ 0x2542, 0x00010000, 0x00010000,
+ 0x2b05, 0x000003ff, 0x000000f3,
+ 0x2b04, 0xffffffff, 0x00000000,
+ 0x2b03, 0xffffffff, 0x00003210,
+ 0x2235, 0x0000001f, 0x00000010,
+ 0x0570, 0x000c0fc0, 0x000c0400
+};
+
+static const u32 oland_golden_rlc_registers[] =
+{
+ 0x3109, 0xffffffff, 0x00601005,
+ 0x311f, 0xffffffff, 0x10104040,
+ 0x3122, 0xffffffff, 0x0100000a,
+ 0x30c5, 0xffffffff, 0x00000800,
+ 0x30c3, 0xffffffff, 0x800000f4
+};
+
+static const u32 hainan_golden_registers[] =
+{
+ 0x2684, 0x00010000, 0x00018208,
+ 0x260c, 0xffffffff, 0x00000000,
+ 0x260d, 0xf00fffff, 0x00000400,
+ 0x260e, 0x0002021c, 0x00020200,
+ 0x4595, 0xff000fff, 0x00000100,
+ 0x340c, 0x000300c0, 0x00800040,
+ 0x3630, 0xff000fff, 0x00000100,
+ 0x360c, 0x000300c0, 0x00800040,
+ 0x0ab9, 0x00073ffe, 0x000022a2,
+ 0x0903, 0x000007ff, 0x00000000,
+ 0x2285, 0xf000001f, 0x00000007,
+ 0x22c9, 0xffffffff, 0x00ffffff,
+ 0x22c4, 0x0000ff0f, 0x00000000,
+ 0xa393, 0x07ffffff, 0x4e000000,
+ 0xa0d4, 0x3f3f3fff, 0x00000000,
+ 0x000c, 0x000000ff, 0x0040,
+ 0x000d, 0x00000040, 0x00004040,
+ 0x2440, 0x03e00000, 0x03600000,
+ 0x2418, 0x0000007f, 0x00000020,
+ 0x2542, 0x00010000, 0x00010000,
+ 0x2b05, 0x000003ff, 0x000000f1,
+ 0x2b04, 0xffffffff, 0x00000000,
+ 0x2b03, 0xffffffff, 0x00003210,
+ 0x2235, 0x0000001f, 0x00000010,
+ 0x0570, 0x000c0fc0, 0x000c0400
+};
+
+static const u32 hainan_golden_registers2[] =
+{
+ 0x263e, 0xffffffff, 0x02010001
+};
+
+static const u32 tahiti_mgcg_cgcg_init[] =
+{
+ 0x3100, 0xffffffff, 0xfffffffc,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2698, 0xffffffff, 0x00000100,
+ 0x24a9, 0xffffffff, 0x00000100,
+ 0x3059, 0xffffffff, 0x00000100,
+ 0x25dd, 0xffffffff, 0x00000100,
+ 0x2261, 0xffffffff, 0x06000100,
+ 0x2286, 0xffffffff, 0x00000100,
+ 0x24a8, 0xffffffff, 0x00000100,
+ 0x30e0, 0xffffffff, 0x00000100,
+ 0x22ca, 0xffffffff, 0x00000100,
+ 0x2451, 0xffffffff, 0x00000100,
+ 0x2362, 0xffffffff, 0x00000100,
+ 0x2363, 0xffffffff, 0x00000100,
+ 0x240c, 0xffffffff, 0x00000100,
+ 0x240d, 0xffffffff, 0x00000100,
+ 0x240e, 0xffffffff, 0x00000100,
+ 0x240f, 0xffffffff, 0x00000100,
+ 0x2b60, 0xffffffff, 0x00000100,
+ 0x2b15, 0xffffffff, 0x00000100,
+ 0x225f, 0xffffffff, 0x06000100,
+ 0x261a, 0xffffffff, 0x00000100,
+ 0x2544, 0xffffffff, 0x00000100,
+ 0x2bc1, 0xffffffff, 0x00000100,
+ 0x2b81, 0xffffffff, 0x00000100,
+ 0x2527, 0xffffffff, 0x00000100,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2458, 0xffffffff, 0x00010000,
+ 0x2459, 0xffffffff, 0x00030002,
+ 0x245a, 0xffffffff, 0x00040007,
+ 0x245b, 0xffffffff, 0x00060005,
+ 0x245c, 0xffffffff, 0x00090008,
+ 0x245d, 0xffffffff, 0x00020001,
+ 0x245e, 0xffffffff, 0x00040003,
+ 0x245f, 0xffffffff, 0x00000007,
+ 0x2460, 0xffffffff, 0x00060005,
+ 0x2461, 0xffffffff, 0x00090008,
+ 0x2462, 0xffffffff, 0x00030002,
+ 0x2463, 0xffffffff, 0x00050004,
+ 0x2464, 0xffffffff, 0x00000008,
+ 0x2465, 0xffffffff, 0x00070006,
+ 0x2466, 0xffffffff, 0x000a0009,
+ 0x2467, 0xffffffff, 0x00040003,
+ 0x2468, 0xffffffff, 0x00060005,
+ 0x2469, 0xffffffff, 0x00000009,
+ 0x246a, 0xffffffff, 0x00080007,
+ 0x246b, 0xffffffff, 0x000b000a,
+ 0x246c, 0xffffffff, 0x00050004,
+ 0x246d, 0xffffffff, 0x00070006,
+ 0x246e, 0xffffffff, 0x0008000b,
+ 0x246f, 0xffffffff, 0x000a0009,
+ 0x2470, 0xffffffff, 0x000d000c,
+ 0x2471, 0xffffffff, 0x00060005,
+ 0x2472, 0xffffffff, 0x00080007,
+ 0x2473, 0xffffffff, 0x0000000b,
+ 0x2474, 0xffffffff, 0x000a0009,
+ 0x2475, 0xffffffff, 0x000d000c,
+ 0x2476, 0xffffffff, 0x00070006,
+ 0x2477, 0xffffffff, 0x00090008,
+ 0x2478, 0xffffffff, 0x0000000c,
+ 0x2479, 0xffffffff, 0x000b000a,
+ 0x247a, 0xffffffff, 0x000e000d,
+ 0x247b, 0xffffffff, 0x00080007,
+ 0x247c, 0xffffffff, 0x000a0009,
+ 0x247d, 0xffffffff, 0x0000000d,
+ 0x247e, 0xffffffff, 0x000c000b,
+ 0x247f, 0xffffffff, 0x000f000e,
+ 0x2480, 0xffffffff, 0x00090008,
+ 0x2481, 0xffffffff, 0x000b000a,
+ 0x2482, 0xffffffff, 0x000c000f,
+ 0x2483, 0xffffffff, 0x000e000d,
+ 0x2484, 0xffffffff, 0x00110010,
+ 0x2485, 0xffffffff, 0x000a0009,
+ 0x2486, 0xffffffff, 0x000c000b,
+ 0x2487, 0xffffffff, 0x0000000f,
+ 0x2488, 0xffffffff, 0x000e000d,
+ 0x2489, 0xffffffff, 0x00110010,
+ 0x248a, 0xffffffff, 0x000b000a,
+ 0x248b, 0xffffffff, 0x000d000c,
+ 0x248c, 0xffffffff, 0x00000010,
+ 0x248d, 0xffffffff, 0x000f000e,
+ 0x248e, 0xffffffff, 0x00120011,
+ 0x248f, 0xffffffff, 0x000c000b,
+ 0x2490, 0xffffffff, 0x000e000d,
+ 0x2491, 0xffffffff, 0x00000011,
+ 0x2492, 0xffffffff, 0x0010000f,
+ 0x2493, 0xffffffff, 0x00130012,
+ 0x2494, 0xffffffff, 0x000d000c,
+ 0x2495, 0xffffffff, 0x000f000e,
+ 0x2496, 0xffffffff, 0x00100013,
+ 0x2497, 0xffffffff, 0x00120011,
+ 0x2498, 0xffffffff, 0x00150014,
+ 0x2499, 0xffffffff, 0x000e000d,
+ 0x249a, 0xffffffff, 0x0010000f,
+ 0x249b, 0xffffffff, 0x00000013,
+ 0x249c, 0xffffffff, 0x00120011,
+ 0x249d, 0xffffffff, 0x00150014,
+ 0x249e, 0xffffffff, 0x000f000e,
+ 0x249f, 0xffffffff, 0x00110010,
+ 0x24a0, 0xffffffff, 0x00000014,
+ 0x24a1, 0xffffffff, 0x00130012,
+ 0x24a2, 0xffffffff, 0x00160015,
+ 0x24a3, 0xffffffff, 0x0010000f,
+ 0x24a4, 0xffffffff, 0x00120011,
+ 0x24a5, 0xffffffff, 0x00000015,
+ 0x24a6, 0xffffffff, 0x00140013,
+ 0x24a7, 0xffffffff, 0x00170016,
+ 0x2454, 0xffffffff, 0x96940200,
+ 0x21c2, 0xffffffff, 0x00900100,
+ 0x311e, 0xffffffff, 0x00000080,
+ 0x3101, 0xffffffff, 0x0020003f,
+ 0xc, 0xffffffff, 0x0000001c,
+ 0xd, 0x000f0000, 0x000f0000,
+ 0x583, 0xffffffff, 0x00000100,
+ 0x409, 0xffffffff, 0x00000100,
+ 0x40b, 0x00000101, 0x00000000,
+ 0x82a, 0xffffffff, 0x00000104,
+ 0x993, 0x000c0000, 0x000c0000,
+ 0x992, 0x000c0000, 0x000c0000,
+ 0x1579, 0xff000fff, 0x00000100,
+ 0x157a, 0x00000001, 0x00000001,
+ 0xbd4, 0x00000001, 0x00000001,
+ 0xc33, 0xc0000fff, 0x00000104,
+ 0x3079, 0x00000001, 0x00000001,
+ 0x3430, 0xfffffff0, 0x00000100,
+ 0x3630, 0xfffffff0, 0x00000100
+};
+static const u32 pitcairn_mgcg_cgcg_init[] =
+{
+ 0x3100, 0xffffffff, 0xfffffffc,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2698, 0xffffffff, 0x00000100,
+ 0x24a9, 0xffffffff, 0x00000100,
+ 0x3059, 0xffffffff, 0x00000100,
+ 0x25dd, 0xffffffff, 0x00000100,
+ 0x2261, 0xffffffff, 0x06000100,
+ 0x2286, 0xffffffff, 0x00000100,
+ 0x24a8, 0xffffffff, 0x00000100,
+ 0x30e0, 0xffffffff, 0x00000100,
+ 0x22ca, 0xffffffff, 0x00000100,
+ 0x2451, 0xffffffff, 0x00000100,
+ 0x2362, 0xffffffff, 0x00000100,
+ 0x2363, 0xffffffff, 0x00000100,
+ 0x240c, 0xffffffff, 0x00000100,
+ 0x240d, 0xffffffff, 0x00000100,
+ 0x240e, 0xffffffff, 0x00000100,
+ 0x240f, 0xffffffff, 0x00000100,
+ 0x2b60, 0xffffffff, 0x00000100,
+ 0x2b15, 0xffffffff, 0x00000100,
+ 0x225f, 0xffffffff, 0x06000100,
+ 0x261a, 0xffffffff, 0x00000100,
+ 0x2544, 0xffffffff, 0x00000100,
+ 0x2bc1, 0xffffffff, 0x00000100,
+ 0x2b81, 0xffffffff, 0x00000100,
+ 0x2527, 0xffffffff, 0x00000100,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2458, 0xffffffff, 0x00010000,
+ 0x2459, 0xffffffff, 0x00030002,
+ 0x245a, 0xffffffff, 0x00040007,
+ 0x245b, 0xffffffff, 0x00060005,
+ 0x245c, 0xffffffff, 0x00090008,
+ 0x245d, 0xffffffff, 0x00020001,
+ 0x245e, 0xffffffff, 0x00040003,
+ 0x245f, 0xffffffff, 0x00000007,
+ 0x2460, 0xffffffff, 0x00060005,
+ 0x2461, 0xffffffff, 0x00090008,
+ 0x2462, 0xffffffff, 0x00030002,
+ 0x2463, 0xffffffff, 0x00050004,
+ 0x2464, 0xffffffff, 0x00000008,
+ 0x2465, 0xffffffff, 0x00070006,
+ 0x2466, 0xffffffff, 0x000a0009,
+ 0x2467, 0xffffffff, 0x00040003,
+ 0x2468, 0xffffffff, 0x00060005,
+ 0x2469, 0xffffffff, 0x00000009,
+ 0x246a, 0xffffffff, 0x00080007,
+ 0x246b, 0xffffffff, 0x000b000a,
+ 0x246c, 0xffffffff, 0x00050004,
+ 0x246d, 0xffffffff, 0x00070006,
+ 0x246e, 0xffffffff, 0x0008000b,
+ 0x246f, 0xffffffff, 0x000a0009,
+ 0x2470, 0xffffffff, 0x000d000c,
+ 0x2480, 0xffffffff, 0x00090008,
+ 0x2481, 0xffffffff, 0x000b000a,
+ 0x2482, 0xffffffff, 0x000c000f,
+ 0x2483, 0xffffffff, 0x000e000d,
+ 0x2484, 0xffffffff, 0x00110010,
+ 0x2485, 0xffffffff, 0x000a0009,
+ 0x2486, 0xffffffff, 0x000c000b,
+ 0x2487, 0xffffffff, 0x0000000f,
+ 0x2488, 0xffffffff, 0x000e000d,
+ 0x2489, 0xffffffff, 0x00110010,
+ 0x248a, 0xffffffff, 0x000b000a,
+ 0x248b, 0xffffffff, 0x000d000c,
+ 0x248c, 0xffffffff, 0x00000010,
+ 0x248d, 0xffffffff, 0x000f000e,
+ 0x248e, 0xffffffff, 0x00120011,
+ 0x248f, 0xffffffff, 0x000c000b,
+ 0x2490, 0xffffffff, 0x000e000d,
+ 0x2491, 0xffffffff, 0x00000011,
+ 0x2492, 0xffffffff, 0x0010000f,
+ 0x2493, 0xffffffff, 0x00130012,
+ 0x2494, 0xffffffff, 0x000d000c,
+ 0x2495, 0xffffffff, 0x000f000e,
+ 0x2496, 0xffffffff, 0x00100013,
+ 0x2497, 0xffffffff, 0x00120011,
+ 0x2498, 0xffffffff, 0x00150014,
+ 0x2454, 0xffffffff, 0x96940200,
+ 0x21c2, 0xffffffff, 0x00900100,
+ 0x311e, 0xffffffff, 0x00000080,
+ 0x3101, 0xffffffff, 0x0020003f,
+ 0xc, 0xffffffff, 0x0000001c,
+ 0xd, 0x000f0000, 0x000f0000,
+ 0x583, 0xffffffff, 0x00000100,
+ 0x409, 0xffffffff, 0x00000100,
+ 0x40b, 0x00000101, 0x00000000,
+ 0x82a, 0xffffffff, 0x00000104,
+ 0x1579, 0xff000fff, 0x00000100,
+ 0x157a, 0x00000001, 0x00000001,
+ 0xbd4, 0x00000001, 0x00000001,
+ 0xc33, 0xc0000fff, 0x00000104,
+ 0x3079, 0x00000001, 0x00000001,
+ 0x3430, 0xfffffff0, 0x00000100,
+ 0x3630, 0xfffffff0, 0x00000100
+};
+static const u32 verde_mgcg_cgcg_init[] =
+{
+ 0x3100, 0xffffffff, 0xfffffffc,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2698, 0xffffffff, 0x00000100,
+ 0x24a9, 0xffffffff, 0x00000100,
+ 0x3059, 0xffffffff, 0x00000100,
+ 0x25dd, 0xffffffff, 0x00000100,
+ 0x2261, 0xffffffff, 0x06000100,
+ 0x2286, 0xffffffff, 0x00000100,
+ 0x24a8, 0xffffffff, 0x00000100,
+ 0x30e0, 0xffffffff, 0x00000100,
+ 0x22ca, 0xffffffff, 0x00000100,
+ 0x2451, 0xffffffff, 0x00000100,
+ 0x2362, 0xffffffff, 0x00000100,
+ 0x2363, 0xffffffff, 0x00000100,
+ 0x240c, 0xffffffff, 0x00000100,
+ 0x240d, 0xffffffff, 0x00000100,
+ 0x240e, 0xffffffff, 0x00000100,
+ 0x240f, 0xffffffff, 0x00000100,
+ 0x2b60, 0xffffffff, 0x00000100,
+ 0x2b15, 0xffffffff, 0x00000100,
+ 0x225f, 0xffffffff, 0x06000100,
+ 0x261a, 0xffffffff, 0x00000100,
+ 0x2544, 0xffffffff, 0x00000100,
+ 0x2bc1, 0xffffffff, 0x00000100,
+ 0x2b81, 0xffffffff, 0x00000100,
+ 0x2527, 0xffffffff, 0x00000100,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2458, 0xffffffff, 0x00010000,
+ 0x2459, 0xffffffff, 0x00030002,
+ 0x245a, 0xffffffff, 0x00040007,
+ 0x245b, 0xffffffff, 0x00060005,
+ 0x245c, 0xffffffff, 0x00090008,
+ 0x245d, 0xffffffff, 0x00020001,
+ 0x245e, 0xffffffff, 0x00040003,
+ 0x245f, 0xffffffff, 0x00000007,
+ 0x2460, 0xffffffff, 0x00060005,
+ 0x2461, 0xffffffff, 0x00090008,
+ 0x2462, 0xffffffff, 0x00030002,
+ 0x2463, 0xffffffff, 0x00050004,
+ 0x2464, 0xffffffff, 0x00000008,
+ 0x2465, 0xffffffff, 0x00070006,
+ 0x2466, 0xffffffff, 0x000a0009,
+ 0x2467, 0xffffffff, 0x00040003,
+ 0x2468, 0xffffffff, 0x00060005,
+ 0x2469, 0xffffffff, 0x00000009,
+ 0x246a, 0xffffffff, 0x00080007,
+ 0x246b, 0xffffffff, 0x000b000a,
+ 0x246c, 0xffffffff, 0x00050004,
+ 0x246d, 0xffffffff, 0x00070006,
+ 0x246e, 0xffffffff, 0x0008000b,
+ 0x246f, 0xffffffff, 0x000a0009,
+ 0x2470, 0xffffffff, 0x000d000c,
+ 0x2480, 0xffffffff, 0x00090008,
+ 0x2481, 0xffffffff, 0x000b000a,
+ 0x2482, 0xffffffff, 0x000c000f,
+ 0x2483, 0xffffffff, 0x000e000d,
+ 0x2484, 0xffffffff, 0x00110010,
+ 0x2485, 0xffffffff, 0x000a0009,
+ 0x2486, 0xffffffff, 0x000c000b,
+ 0x2487, 0xffffffff, 0x0000000f,
+ 0x2488, 0xffffffff, 0x000e000d,
+ 0x2489, 0xffffffff, 0x00110010,
+ 0x248a, 0xffffffff, 0x000b000a,
+ 0x248b, 0xffffffff, 0x000d000c,
+ 0x248c, 0xffffffff, 0x00000010,
+ 0x248d, 0xffffffff, 0x000f000e,
+ 0x248e, 0xffffffff, 0x00120011,
+ 0x248f, 0xffffffff, 0x000c000b,
+ 0x2490, 0xffffffff, 0x000e000d,
+ 0x2491, 0xffffffff, 0x00000011,
+ 0x2492, 0xffffffff, 0x0010000f,
+ 0x2493, 0xffffffff, 0x00130012,
+ 0x2494, 0xffffffff, 0x000d000c,
+ 0x2495, 0xffffffff, 0x000f000e,
+ 0x2496, 0xffffffff, 0x00100013,
+ 0x2497, 0xffffffff, 0x00120011,
+ 0x2498, 0xffffffff, 0x00150014,
+ 0x2454, 0xffffffff, 0x96940200,
+ 0x21c2, 0xffffffff, 0x00900100,
+ 0x311e, 0xffffffff, 0x00000080,
+ 0x3101, 0xffffffff, 0x0020003f,
+ 0xc, 0xffffffff, 0x0000001c,
+ 0xd, 0x000f0000, 0x000f0000,
+ 0x583, 0xffffffff, 0x00000100,
+ 0x409, 0xffffffff, 0x00000100,
+ 0x40b, 0x00000101, 0x00000000,
+ 0x82a, 0xffffffff, 0x00000104,
+ 0x993, 0x000c0000, 0x000c0000,
+ 0x992, 0x000c0000, 0x000c0000,
+ 0x1579, 0xff000fff, 0x00000100,
+ 0x157a, 0x00000001, 0x00000001,
+ 0xbd4, 0x00000001, 0x00000001,
+ 0xc33, 0xc0000fff, 0x00000104,
+ 0x3079, 0x00000001, 0x00000001,
+ 0x3430, 0xfffffff0, 0x00000100,
+ 0x3630, 0xfffffff0, 0x00000100
+};
+static const u32 oland_mgcg_cgcg_init[] =
+{
+ 0x3100, 0xffffffff, 0xfffffffc,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2698, 0xffffffff, 0x00000100,
+ 0x24a9, 0xffffffff, 0x00000100,
+ 0x3059, 0xffffffff, 0x00000100,
+ 0x25dd, 0xffffffff, 0x00000100,
+ 0x2261, 0xffffffff, 0x06000100,
+ 0x2286, 0xffffffff, 0x00000100,
+ 0x24a8, 0xffffffff, 0x00000100,
+ 0x30e0, 0xffffffff, 0x00000100,
+ 0x22ca, 0xffffffff, 0x00000100,
+ 0x2451, 0xffffffff, 0x00000100,
+ 0x2362, 0xffffffff, 0x00000100,
+ 0x2363, 0xffffffff, 0x00000100,
+ 0x240c, 0xffffffff, 0x00000100,
+ 0x240d, 0xffffffff, 0x00000100,
+ 0x240e, 0xffffffff, 0x00000100,
+ 0x240f, 0xffffffff, 0x00000100,
+ 0x2b60, 0xffffffff, 0x00000100,
+ 0x2b15, 0xffffffff, 0x00000100,
+ 0x225f, 0xffffffff, 0x06000100,
+ 0x261a, 0xffffffff, 0x00000100,
+ 0x2544, 0xffffffff, 0x00000100,
+ 0x2bc1, 0xffffffff, 0x00000100,
+ 0x2b81, 0xffffffff, 0x00000100,
+ 0x2527, 0xffffffff, 0x00000100,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2458, 0xffffffff, 0x00010000,
+ 0x2459, 0xffffffff, 0x00030002,
+ 0x245a, 0xffffffff, 0x00040007,
+ 0x245b, 0xffffffff, 0x00060005,
+ 0x245c, 0xffffffff, 0x00090008,
+ 0x245d, 0xffffffff, 0x00020001,
+ 0x245e, 0xffffffff, 0x00040003,
+ 0x245f, 0xffffffff, 0x00000007,
+ 0x2460, 0xffffffff, 0x00060005,
+ 0x2461, 0xffffffff, 0x00090008,
+ 0x2462, 0xffffffff, 0x00030002,
+ 0x2463, 0xffffffff, 0x00050004,
+ 0x2464, 0xffffffff, 0x00000008,
+ 0x2465, 0xffffffff, 0x00070006,
+ 0x2466, 0xffffffff, 0x000a0009,
+ 0x2467, 0xffffffff, 0x00040003,
+ 0x2468, 0xffffffff, 0x00060005,
+ 0x2469, 0xffffffff, 0x00000009,
+ 0x246a, 0xffffffff, 0x00080007,
+ 0x246b, 0xffffffff, 0x000b000a,
+ 0x246c, 0xffffffff, 0x00050004,
+ 0x246d, 0xffffffff, 0x00070006,
+ 0x246e, 0xffffffff, 0x0008000b,
+ 0x246f, 0xffffffff, 0x000a0009,
+ 0x2470, 0xffffffff, 0x000d000c,
+ 0x2471, 0xffffffff, 0x00060005,
+ 0x2472, 0xffffffff, 0x00080007,
+ 0x2473, 0xffffffff, 0x0000000b,
+ 0x2474, 0xffffffff, 0x000a0009,
+ 0x2475, 0xffffffff, 0x000d000c,
+ 0x2454, 0xffffffff, 0x96940200,
+ 0x21c2, 0xffffffff, 0x00900100,
+ 0x311e, 0xffffffff, 0x00000080,
+ 0x3101, 0xffffffff, 0x0020003f,
+ 0xc, 0xffffffff, 0x0000001c,
+ 0xd, 0x000f0000, 0x000f0000,
+ 0x583, 0xffffffff, 0x00000100,
+ 0x409, 0xffffffff, 0x00000100,
+ 0x40b, 0x00000101, 0x00000000,
+ 0x82a, 0xffffffff, 0x00000104,
+ 0x993, 0x000c0000, 0x000c0000,
+ 0x992, 0x000c0000, 0x000c0000,
+ 0x1579, 0xff000fff, 0x00000100,
+ 0x157a, 0x00000001, 0x00000001,
+ 0xbd4, 0x00000001, 0x00000001,
+ 0xc33, 0xc0000fff, 0x00000104,
+ 0x3079, 0x00000001, 0x00000001,
+ 0x3430, 0xfffffff0, 0x00000100,
+ 0x3630, 0xfffffff0, 0x00000100
+};
+static const u32 hainan_mgcg_cgcg_init[] =
+{
+ 0x3100, 0xffffffff, 0xfffffffc,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2698, 0xffffffff, 0x00000100,
+ 0x24a9, 0xffffffff, 0x00000100,
+ 0x3059, 0xffffffff, 0x00000100,
+ 0x25dd, 0xffffffff, 0x00000100,
+ 0x2261, 0xffffffff, 0x06000100,
+ 0x2286, 0xffffffff, 0x00000100,
+ 0x24a8, 0xffffffff, 0x00000100,
+ 0x30e0, 0xffffffff, 0x00000100,
+ 0x22ca, 0xffffffff, 0x00000100,
+ 0x2451, 0xffffffff, 0x00000100,
+ 0x2362, 0xffffffff, 0x00000100,
+ 0x2363, 0xffffffff, 0x00000100,
+ 0x240c, 0xffffffff, 0x00000100,
+ 0x240d, 0xffffffff, 0x00000100,
+ 0x240e, 0xffffffff, 0x00000100,
+ 0x240f, 0xffffffff, 0x00000100,
+ 0x2b60, 0xffffffff, 0x00000100,
+ 0x2b15, 0xffffffff, 0x00000100,
+ 0x225f, 0xffffffff, 0x06000100,
+ 0x261a, 0xffffffff, 0x00000100,
+ 0x2544, 0xffffffff, 0x00000100,
+ 0x2bc1, 0xffffffff, 0x00000100,
+ 0x2b81, 0xffffffff, 0x00000100,
+ 0x2527, 0xffffffff, 0x00000100,
+ 0x200b, 0xffffffff, 0xe0000000,
+ 0x2458, 0xffffffff, 0x00010000,
+ 0x2459, 0xffffffff, 0x00030002,
+ 0x245a, 0xffffffff, 0x00040007,
+ 0x245b, 0xffffffff, 0x00060005,
+ 0x245c, 0xffffffff, 0x00090008,
+ 0x245d, 0xffffffff, 0x00020001,
+ 0x245e, 0xffffffff, 0x00040003,
+ 0x245f, 0xffffffff, 0x00000007,
+ 0x2460, 0xffffffff, 0x00060005,
+ 0x2461, 0xffffffff, 0x00090008,
+ 0x2462, 0xffffffff, 0x00030002,
+ 0x2463, 0xffffffff, 0x00050004,
+ 0x2464, 0xffffffff, 0x00000008,
+ 0x2465, 0xffffffff, 0x00070006,
+ 0x2466, 0xffffffff, 0x000a0009,
+ 0x2467, 0xffffffff, 0x00040003,
+ 0x2468, 0xffffffff, 0x00060005,
+ 0x2469, 0xffffffff, 0x00000009,
+ 0x246a, 0xffffffff, 0x00080007,
+ 0x246b, 0xffffffff, 0x000b000a,
+ 0x246c, 0xffffffff, 0x00050004,
+ 0x246d, 0xffffffff, 0x00070006,
+ 0x246e, 0xffffffff, 0x0008000b,
+ 0x246f, 0xffffffff, 0x000a0009,
+ 0x2470, 0xffffffff, 0x000d000c,
+ 0x2471, 0xffffffff, 0x00060005,
+ 0x2472, 0xffffffff, 0x00080007,
+ 0x2473, 0xffffffff, 0x0000000b,
+ 0x2474, 0xffffffff, 0x000a0009,
+ 0x2475, 0xffffffff, 0x000d000c,
+ 0x2454, 0xffffffff, 0x96940200,
+ 0x21c2, 0xffffffff, 0x00900100,
+ 0x311e, 0xffffffff, 0x00000080,
+ 0x3101, 0xffffffff, 0x0020003f,
+ 0xc, 0xffffffff, 0x0000001c,
+ 0xd, 0x000f0000, 0x000f0000,
+ 0x583, 0xffffffff, 0x00000100,
+ 0x409, 0xffffffff, 0x00000100,
+ 0x82a, 0xffffffff, 0x00000104,
+ 0x993, 0x000c0000, 0x000c0000,
+ 0x992, 0x000c0000, 0x000c0000,
+ 0xbd4, 0x00000001, 0x00000001,
+ 0xc33, 0xc0000fff, 0x00000104,
+ 0x3079, 0x00000001, 0x00000001,
+ 0x3430, 0xfffffff0, 0x00000100,
+ 0x3630, 0xfffffff0, 0x00000100
+};
+
+static u32 si_pcie_rreg(struct amdgpu_device *adev, u32 reg)
+{
+ unsigned long flags;
+ u32 r;
+
+ spin_lock_irqsave(&adev->pcie_idx_lock, flags);
+ WREG32(AMDGPU_PCIE_INDEX, reg);
+ (void)RREG32(AMDGPU_PCIE_INDEX);
+ r = RREG32(AMDGPU_PCIE_DATA);
+ spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
+ return r;
+}
+
+static void si_pcie_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&adev->pcie_idx_lock, flags);
+ WREG32(AMDGPU_PCIE_INDEX, reg);
+ (void)RREG32(AMDGPU_PCIE_INDEX);
+ WREG32(AMDGPU_PCIE_DATA, v);
+ (void)RREG32(AMDGPU_PCIE_DATA);
+ spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
+}
+
+u32 si_pciep_rreg(struct amdgpu_device *adev, u32 reg)
+{
+ unsigned long flags;
+ u32 r;
+
+ spin_lock_irqsave(&adev->pcie_idx_lock, flags);
+ WREG32(PCIE_PORT_INDEX, ((reg) & 0xff));
+ (void)RREG32(PCIE_PORT_INDEX);
+ r = RREG32(PCIE_PORT_DATA);
+ spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
+ return r;
+}
+
+void si_pciep_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&adev->pcie_idx_lock, flags);
+ WREG32(PCIE_PORT_INDEX, ((reg) & 0xff));
+ (void)RREG32(PCIE_PORT_INDEX);
+ WREG32(PCIE_PORT_DATA, (v));
+ (void)RREG32(PCIE_PORT_DATA);
+ spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
+}
+
+static u32 si_smc_rreg(struct amdgpu_device *adev, u32 reg)
+{
+ unsigned long flags;
+ u32 r;
+
+ spin_lock_irqsave(&adev->smc_idx_lock, flags);
+ WREG32(SMC_IND_INDEX_0, (reg));
+ r = RREG32(SMC_IND_DATA_0);
+ spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
+ return r;
+}
+
+static void si_smc_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&adev->smc_idx_lock, flags);
+ WREG32(SMC_IND_INDEX_0, (reg));
+ WREG32(SMC_IND_DATA_0, (v));
+ spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
+}
+
+static u32 si_get_virtual_caps(struct amdgpu_device *adev)
+{
+ /* SI does not support SR-IOV */
+ return 0;
+}
+
+static struct amdgpu_allowed_register_entry si_allowed_read_registers[] = {
+ {GRBM_STATUS, false},
+ {GB_ADDR_CONFIG, false},
+ {MC_ARB_RAMCFG, false},
+ {GB_TILE_MODE0, false},
+ {GB_TILE_MODE1, false},
+ {GB_TILE_MODE2, false},
+ {GB_TILE_MODE3, false},
+ {GB_TILE_MODE4, false},
+ {GB_TILE_MODE5, false},
+ {GB_TILE_MODE6, false},
+ {GB_TILE_MODE7, false},
+ {GB_TILE_MODE8, false},
+ {GB_TILE_MODE9, false},
+ {GB_TILE_MODE10, false},
+ {GB_TILE_MODE11, false},
+ {GB_TILE_MODE12, false},
+ {GB_TILE_MODE13, false},
+ {GB_TILE_MODE14, false},
+ {GB_TILE_MODE15, false},
+ {GB_TILE_MODE16, false},
+ {GB_TILE_MODE17, false},
+ {GB_TILE_MODE18, false},
+ {GB_TILE_MODE19, false},
+ {GB_TILE_MODE20, false},
+ {GB_TILE_MODE21, false},
+ {GB_TILE_MODE22, false},
+ {GB_TILE_MODE23, false},
+ {GB_TILE_MODE24, false},
+ {GB_TILE_MODE25, false},
+ {GB_TILE_MODE26, false},
+ {GB_TILE_MODE27, false},
+ {GB_TILE_MODE28, false},
+ {GB_TILE_MODE29, false},
+ {GB_TILE_MODE30, false},
+ {GB_TILE_MODE31, false},
+ {CC_RB_BACKEND_DISABLE, false, true},
+ {GC_USER_RB_BACKEND_DISABLE, false, true},
+ {PA_SC_RASTER_CONFIG, false, true},
+};
+
+static uint32_t si_read_indexed_register(struct amdgpu_device *adev,
+ u32 se_num, u32 sh_num,
+ u32 reg_offset)
+{
+ uint32_t val;
+
+ mutex_lock(&adev->grbm_idx_mutex);
+ if (se_num != 0xffffffff || sh_num != 0xffffffff)
+ amdgpu_gfx_select_se_sh(adev, se_num, sh_num, 0xffffffff);
+
+ val = RREG32(reg_offset);
+
+ if (se_num != 0xffffffff || sh_num != 0xffffffff)
+ amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+ mutex_unlock(&adev->grbm_idx_mutex);
+ return val;
+}
+
+static int si_read_register(struct amdgpu_device *adev, u32 se_num,
+ u32 sh_num, u32 reg_offset, u32 *value)
+{
+ uint32_t i;
+
+ *value = 0;
+ for (i = 0; i < ARRAY_SIZE(si_allowed_read_registers); i++) {
+ if (reg_offset != si_allowed_read_registers[i].reg_offset)
+ continue;
+
+ if (!si_allowed_read_registers[i].untouched)
+ *value = si_allowed_read_registers[i].grbm_indexed ?
+ si_read_indexed_register(adev, se_num,
+ sh_num, reg_offset) :
+ RREG32(reg_offset);
+ return 0;
+ }
+ return -EINVAL;
+}
+
+static bool si_read_disabled_bios(struct amdgpu_device *adev)
+{
+ u32 bus_cntl;
+ u32 d1vga_control = 0;
+ u32 d2vga_control = 0;
+ u32 vga_render_control = 0;
+ u32 rom_cntl;
+ bool r;
+
+ bus_cntl = RREG32(R600_BUS_CNTL);
+ if (adev->mode_info.num_crtc) {
+ d1vga_control = RREG32(AVIVO_D1VGA_CONTROL);
+ d2vga_control = RREG32(AVIVO_D2VGA_CONTROL);
+ vga_render_control = RREG32(VGA_RENDER_CONTROL);
+ }
+ rom_cntl = RREG32(R600_ROM_CNTL);
+
+ /* enable the rom */
+ WREG32(R600_BUS_CNTL, (bus_cntl & ~R600_BIOS_ROM_DIS));
+ if (adev->mode_info.num_crtc) {
+ /* Disable VGA mode */
+ WREG32(AVIVO_D1VGA_CONTROL,
+ (d1vga_control & ~(AVIVO_DVGA_CONTROL_MODE_ENABLE |
+ AVIVO_DVGA_CONTROL_TIMING_SELECT)));
+ WREG32(AVIVO_D2VGA_CONTROL,
+ (d2vga_control & ~(AVIVO_DVGA_CONTROL_MODE_ENABLE |
+ AVIVO_DVGA_CONTROL_TIMING_SELECT)));
+ WREG32(VGA_RENDER_CONTROL,
+ (vga_render_control & C_000300_VGA_VSTATUS_CNTL));
+ }
+ WREG32(R600_ROM_CNTL, rom_cntl | R600_SCK_OVERWRITE);
+
+ r = amdgpu_read_bios(adev);
+
+ /* restore regs */
+ WREG32(R600_BUS_CNTL, bus_cntl);
+ if (adev->mode_info.num_crtc) {
+ WREG32(AVIVO_D1VGA_CONTROL, d1vga_control);
+ WREG32(AVIVO_D2VGA_CONTROL, d2vga_control);
+ WREG32(VGA_RENDER_CONTROL, vga_render_control);
+ }
+ WREG32(R600_ROM_CNTL, rom_cntl);
+ return r;
+}
+
+//xxx: not implemented
+static int si_asic_reset(struct amdgpu_device *adev)
+{
+ return 0;
+}
+
+static void si_vga_set_state(struct amdgpu_device *adev, bool state)
+{
+ uint32_t temp;
+
+ temp = RREG32(CONFIG_CNTL);
+ if (state == false) {
+ temp &= ~(1<<0);
+ temp |= (1<<1);
+ } else {
+ temp &= ~(1<<1);
+ }
+ WREG32(CONFIG_CNTL, temp);
+}
+
+static u32 si_get_xclk(struct amdgpu_device *adev)
+{
+ u32 reference_clock = adev->clock.spll.reference_freq;
+ u32 tmp;
+
+ tmp = RREG32(CG_CLKPIN_CNTL_2);
+ if (tmp & MUX_TCLK_TO_XCLK)
+ return TCLK;
+
+ tmp = RREG32(CG_CLKPIN_CNTL);
+ if (tmp & XTALIN_DIVIDE)
+ return reference_clock / 4;
+
+ return reference_clock;
+}
+
+//xxx:not implemented
+static int si_set_uvd_clocks(struct amdgpu_device *adev, u32 vclk, u32 dclk)
+{
+ return 0;
+}
+
+static const struct amdgpu_asic_funcs si_asic_funcs =
+{
+ .read_disabled_bios = &si_read_disabled_bios,
+ .read_register = &si_read_register,
+ .reset = &si_asic_reset,
+ .set_vga_state = &si_vga_set_state,
+ .get_xclk = &si_get_xclk,
+ .set_uvd_clocks = &si_set_uvd_clocks,
+ .set_vce_clocks = NULL,
+ .get_virtual_caps = &si_get_virtual_caps,
+};
+
+static uint32_t si_get_rev_id(struct amdgpu_device *adev)
+{
+ return (RREG32(CC_DRM_ID_STRAPS) & CC_DRM_ID_STRAPS__ATI_REV_ID_MASK)
+ >> CC_DRM_ID_STRAPS__ATI_REV_ID__SHIFT;
+}
+
+static int si_common_early_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ adev->smc_rreg = &si_smc_rreg;
+ adev->smc_wreg = &si_smc_wreg;
+ adev->pcie_rreg = &si_pcie_rreg;
+ adev->pcie_wreg = &si_pcie_wreg;
+ adev->pciep_rreg = &si_pciep_rreg;
+ adev->pciep_wreg = &si_pciep_wreg;
+ adev->uvd_ctx_rreg = NULL;
+ adev->uvd_ctx_wreg = NULL;
+ adev->didt_rreg = NULL;
+ adev->didt_wreg = NULL;
+
+ adev->asic_funcs = &si_asic_funcs;
+
+ adev->rev_id = si_get_rev_id(adev);
+ adev->external_rev_id = 0xFF;
+ switch (adev->asic_type) {
+ case CHIP_TAHITI:
+ adev->cg_flags =
+ AMD_CG_SUPPORT_GFX_MGCG |
+ AMD_CG_SUPPORT_GFX_MGLS |
+ /*AMD_CG_SUPPORT_GFX_CGCG |*/
+ AMD_CG_SUPPORT_GFX_CGLS |
+ AMD_CG_SUPPORT_GFX_CGTS |
+ AMD_CG_SUPPORT_GFX_CP_LS |
+ AMD_CG_SUPPORT_MC_MGCG |
+ AMD_CG_SUPPORT_SDMA_MGCG |
+ AMD_CG_SUPPORT_BIF_LS |
+ AMD_CG_SUPPORT_VCE_MGCG |
+ AMD_CG_SUPPORT_UVD_MGCG |
+ AMD_CG_SUPPORT_HDP_LS |
+ AMD_CG_SUPPORT_HDP_MGCG;
+ adev->pg_flags = 0;
+ break;
+ case CHIP_PITCAIRN:
+ adev->cg_flags =
+ AMD_CG_SUPPORT_GFX_MGCG |
+ AMD_CG_SUPPORT_GFX_MGLS |
+ /*AMD_CG_SUPPORT_GFX_CGCG |*/
+ AMD_CG_SUPPORT_GFX_CGLS |
+ AMD_CG_SUPPORT_GFX_CGTS |
+ AMD_CG_SUPPORT_GFX_CP_LS |
+ AMD_CG_SUPPORT_GFX_RLC_LS |
+ AMD_CG_SUPPORT_MC_LS |
+ AMD_CG_SUPPORT_MC_MGCG |
+ AMD_CG_SUPPORT_SDMA_MGCG |
+ AMD_CG_SUPPORT_BIF_LS |
+ AMD_CG_SUPPORT_VCE_MGCG |
+ AMD_CG_SUPPORT_UVD_MGCG |
+ AMD_CG_SUPPORT_HDP_LS |
+ AMD_CG_SUPPORT_HDP_MGCG;
+ adev->pg_flags = 0;
+ break;
+
+ case CHIP_VERDE:
+ adev->cg_flags =
+ AMD_CG_SUPPORT_GFX_MGCG |
+ AMD_CG_SUPPORT_GFX_MGLS |
+ AMD_CG_SUPPORT_GFX_CGLS |
+ AMD_CG_SUPPORT_GFX_CGTS |
+ AMD_CG_SUPPORT_GFX_CGTS_LS |
+ AMD_CG_SUPPORT_GFX_CP_LS |
+ AMD_CG_SUPPORT_MC_LS |
+ AMD_CG_SUPPORT_MC_MGCG |
+ AMD_CG_SUPPORT_SDMA_MGCG |
+ AMD_CG_SUPPORT_SDMA_LS |
+ AMD_CG_SUPPORT_BIF_LS |
+ AMD_CG_SUPPORT_VCE_MGCG |
+ AMD_CG_SUPPORT_UVD_MGCG |
+ AMD_CG_SUPPORT_HDP_LS |
+ AMD_CG_SUPPORT_HDP_MGCG;
+ adev->pg_flags = 0;
+ //???
+ adev->external_rev_id = adev->rev_id + 0x14;
+ break;
+ case CHIP_OLAND:
+ adev->cg_flags =
+ AMD_CG_SUPPORT_GFX_MGCG |
+ AMD_CG_SUPPORT_GFX_MGLS |
+ /*AMD_CG_SUPPORT_GFX_CGCG |*/
+ AMD_CG_SUPPORT_GFX_CGLS |
+ AMD_CG_SUPPORT_GFX_CGTS |
+ AMD_CG_SUPPORT_GFX_CP_LS |
+ AMD_CG_SUPPORT_GFX_RLC_LS |
+ AMD_CG_SUPPORT_MC_LS |
+ AMD_CG_SUPPORT_MC_MGCG |
+ AMD_CG_SUPPORT_SDMA_MGCG |
+ AMD_CG_SUPPORT_BIF_LS |
+ AMD_CG_SUPPORT_UVD_MGCG |
+ AMD_CG_SUPPORT_HDP_LS |
+ AMD_CG_SUPPORT_HDP_MGCG;
+ adev->pg_flags = 0;
+ break;
+ case CHIP_HAINAN:
+ adev->cg_flags =
+ AMD_CG_SUPPORT_GFX_MGCG |
+ AMD_CG_SUPPORT_GFX_MGLS |
+ /*AMD_CG_SUPPORT_GFX_CGCG |*/
+ AMD_CG_SUPPORT_GFX_CGLS |
+ AMD_CG_SUPPORT_GFX_CGTS |
+ AMD_CG_SUPPORT_GFX_CP_LS |
+ AMD_CG_SUPPORT_GFX_RLC_LS |
+ AMD_CG_SUPPORT_MC_LS |
+ AMD_CG_SUPPORT_MC_MGCG |
+ AMD_CG_SUPPORT_SDMA_MGCG |
+ AMD_CG_SUPPORT_BIF_LS |
+ AMD_CG_SUPPORT_HDP_LS |
+ AMD_CG_SUPPORT_HDP_MGCG;
+ adev->pg_flags = 0;
+ break;
+
+ default:
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int si_common_sw_init(void *handle)
+{
+ return 0;
+}
+
+static int si_common_sw_fini(void *handle)
+{
+ return 0;
+}
+
+
+static void si_init_golden_registers(struct amdgpu_device *adev)
+{
+ switch (adev->asic_type) {
+ case CHIP_TAHITI:
+ amdgpu_program_register_sequence(adev,
+ tahiti_golden_registers,
+ (const u32)ARRAY_SIZE(tahiti_golden_registers));
+ amdgpu_program_register_sequence(adev,
+ tahiti_golden_rlc_registers,
+ (const u32)ARRAY_SIZE(tahiti_golden_rlc_registers));
+ amdgpu_program_register_sequence(adev,
+ tahiti_mgcg_cgcg_init,
+ (const u32)ARRAY_SIZE(tahiti_mgcg_cgcg_init));
+ amdgpu_program_register_sequence(adev,
+ tahiti_golden_registers2,
+ (const u32)ARRAY_SIZE(tahiti_golden_registers2));
+ break;
+ case CHIP_PITCAIRN:
+ amdgpu_program_register_sequence(adev,
+ pitcairn_golden_registers,
+ (const u32)ARRAY_SIZE(pitcairn_golden_registers));
+ amdgpu_program_register_sequence(adev,
+ pitcairn_golden_rlc_registers,
+ (const u32)ARRAY_SIZE(pitcairn_golden_rlc_registers));
+ amdgpu_program_register_sequence(adev,
+ pitcairn_mgcg_cgcg_init,
+ (const u32)ARRAY_SIZE(pitcairn_mgcg_cgcg_init));
+ case CHIP_VERDE:
+ amdgpu_program_register_sequence(adev,
+ verde_golden_registers,
+ (const u32)ARRAY_SIZE(verde_golden_registers));
+ amdgpu_program_register_sequence(adev,
+ verde_golden_rlc_registers,
+ (const u32)ARRAY_SIZE(verde_golden_rlc_registers));
+ amdgpu_program_register_sequence(adev,
+ verde_mgcg_cgcg_init,
+ (const u32)ARRAY_SIZE(verde_mgcg_cgcg_init));
+ amdgpu_program_register_sequence(adev,
+ verde_pg_init,
+ (const u32)ARRAY_SIZE(verde_pg_init));
+ break;
+ case CHIP_OLAND:
+ amdgpu_program_register_sequence(adev,
+ oland_golden_registers,
+ (const u32)ARRAY_SIZE(oland_golden_registers));
+ amdgpu_program_register_sequence(adev,
+ oland_golden_rlc_registers,
+ (const u32)ARRAY_SIZE(oland_golden_rlc_registers));
+ amdgpu_program_register_sequence(adev,
+ oland_mgcg_cgcg_init,
+ (const u32)ARRAY_SIZE(oland_mgcg_cgcg_init));
+ case CHIP_HAINAN:
+ amdgpu_program_register_sequence(adev,
+ hainan_golden_registers,
+ (const u32)ARRAY_SIZE(hainan_golden_registers));
+ amdgpu_program_register_sequence(adev,
+ hainan_golden_registers2,
+ (const u32)ARRAY_SIZE(hainan_golden_registers2));
+ amdgpu_program_register_sequence(adev,
+ hainan_mgcg_cgcg_init,
+ (const u32)ARRAY_SIZE(hainan_mgcg_cgcg_init));
+ break;
+
+
+ default:
+ BUG();
+ }
+}
+
+static void si_pcie_gen3_enable(struct amdgpu_device *adev)
+{
+ struct pci_dev *root = adev->pdev->bus->self;
+ int bridge_pos, gpu_pos;
+ u32 speed_cntl, mask, current_data_rate;
+ int ret, i;
+ u16 tmp16;
+
+ if (pci_is_root_bus(adev->pdev->bus))
+ return;
+
+ if (amdgpu_pcie_gen2 == 0)
+ return;
+
+ if (adev->flags & AMD_IS_APU)
+ return;
+
+ ret = drm_pcie_get_speed_cap_mask(adev->ddev, &mask);
+ if (ret != 0)
+ return;
+
+ if (!(mask & (DRM_PCIE_SPEED_50 | DRM_PCIE_SPEED_80)))
+ return;
+
+ speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
+ current_data_rate = (speed_cntl & LC_CURRENT_DATA_RATE_MASK) >>
+ LC_CURRENT_DATA_RATE_SHIFT;
+ if (mask & DRM_PCIE_SPEED_80) {
+ if (current_data_rate == 2) {
+ DRM_INFO("PCIE gen 3 link speeds already enabled\n");
+ return;
+ }
+ DRM_INFO("enabling PCIE gen 3 link speeds, disable with amdgpu.pcie_gen2=0\n");
+ } else if (mask & DRM_PCIE_SPEED_50) {
+ if (current_data_rate == 1) {
+ DRM_INFO("PCIE gen 2 link speeds already enabled\n");
+ return;
+ }
+ DRM_INFO("enabling PCIE gen 2 link speeds, disable with amdgpu.pcie_gen2=0\n");
+ }
+
+ bridge_pos = pci_pcie_cap(root);
+ if (!bridge_pos)
+ return;
+
+ gpu_pos = pci_pcie_cap(adev->pdev);
+ if (!gpu_pos)
+ return;
+
+ if (mask & DRM_PCIE_SPEED_80) {
+ if (current_data_rate != 2) {
+ u16 bridge_cfg, gpu_cfg;
+ u16 bridge_cfg2, gpu_cfg2;
+ u32 max_lw, current_lw, tmp;
+
+ pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
+ pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+
+ tmp16 = bridge_cfg | PCI_EXP_LNKCTL_HAWD;
+ pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+
+ tmp16 = gpu_cfg | PCI_EXP_LNKCTL_HAWD;
+ pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+
+ tmp = RREG32_PCIE(PCIE_LC_STATUS1);
+ max_lw = (tmp & LC_DETECTED_LINK_WIDTH_MASK) >> LC_DETECTED_LINK_WIDTH_SHIFT;
+ current_lw = (tmp & LC_OPERATING_LINK_WIDTH_MASK) >> LC_OPERATING_LINK_WIDTH_SHIFT;
+
+ if (current_lw < max_lw) {
+ tmp = RREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL);
+ if (tmp & LC_RENEGOTIATION_SUPPORT) {
+ tmp &= ~(LC_LINK_WIDTH_MASK | LC_UPCONFIGURE_DIS);
+ tmp |= (max_lw << LC_LINK_WIDTH_SHIFT);
+ tmp |= LC_UPCONFIGURE_SUPPORT | LC_RENEGOTIATE_EN | LC_RECONFIG_NOW;
+ WREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL, tmp);
+ }
+ }
+
+ for (i = 0; i < 10; i++) {
+ pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_DEVSTA, &tmp16);
+ if (tmp16 & PCI_EXP_DEVSTA_TRPND)
+ break;
+
+ pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &bridge_cfg);
+ pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &gpu_cfg);
+
+ pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &bridge_cfg2);
+ pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &gpu_cfg2);
+
+ tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
+ tmp |= LC_SET_QUIESCE;
+ WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp);
+
+ tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
+ tmp |= LC_REDO_EQ;
+ WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp);
+
+ mdelay(100);
+
+ pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL, &tmp16);
+ tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
+ tmp16 |= (bridge_cfg & PCI_EXP_LNKCTL_HAWD);
+ pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL, tmp16);
+
+ pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, &tmp16);
+ tmp16 &= ~PCI_EXP_LNKCTL_HAWD;
+ tmp16 |= (gpu_cfg & PCI_EXP_LNKCTL_HAWD);
+ pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL, tmp16);
+
+ pci_read_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, &tmp16);
+ tmp16 &= ~((1 << 4) | (7 << 9));
+ tmp16 |= (bridge_cfg2 & ((1 << 4) | (7 << 9)));
+ pci_write_config_word(root, bridge_pos + PCI_EXP_LNKCTL2, tmp16);
+
+ pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
+ tmp16 &= ~((1 << 4) | (7 << 9));
+ tmp16 |= (gpu_cfg2 & ((1 << 4) | (7 << 9)));
+ pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+
+ tmp = RREG32_PCIE_PORT(PCIE_LC_CNTL4);
+ tmp &= ~LC_SET_QUIESCE;
+ WREG32_PCIE_PORT(PCIE_LC_CNTL4, tmp);
+ }
+ }
+ }
+
+ speed_cntl |= LC_FORCE_EN_SW_SPEED_CHANGE | LC_FORCE_DIS_HW_SPEED_CHANGE;
+ speed_cntl &= ~LC_FORCE_DIS_SW_SPEED_CHANGE;
+ WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl);
+
+ pci_read_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
+ tmp16 &= ~0xf;
+ if (mask & DRM_PCIE_SPEED_80)
+ tmp16 |= 3;
+ else if (mask & DRM_PCIE_SPEED_50)
+ tmp16 |= 2;
+ else
+ tmp16 |= 1;
+ pci_write_config_word(adev->pdev, gpu_pos + PCI_EXP_LNKCTL2, tmp16);
+
+ speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
+ speed_cntl |= LC_INITIATE_LINK_SPEED_CHANGE;
+ WREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL, speed_cntl);
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
+ if ((speed_cntl & LC_INITIATE_LINK_SPEED_CHANGE) == 0)
+ break;
+ udelay(1);
+ }
+}
+
+static inline u32 si_pif_phy0_rreg(struct amdgpu_device *adev, u32 reg)
+{
+ unsigned long flags;
+ u32 r;
+
+ spin_lock_irqsave(&adev->pcie_idx_lock, flags);
+ WREG32(EVERGREEN_PIF_PHY0_INDEX, ((reg) & 0xffff));
+ r = RREG32(EVERGREEN_PIF_PHY0_DATA);
+ spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
+ return r;
+}
+
+static inline void si_pif_phy0_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&adev->pcie_idx_lock, flags);
+ WREG32(EVERGREEN_PIF_PHY0_INDEX, ((reg) & 0xffff));
+ WREG32(EVERGREEN_PIF_PHY0_DATA, (v));
+ spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
+}
+
+static inline u32 si_pif_phy1_rreg(struct amdgpu_device *adev, u32 reg)
+{
+ unsigned long flags;
+ u32 r;
+
+ spin_lock_irqsave(&adev->pcie_idx_lock, flags);
+ WREG32(EVERGREEN_PIF_PHY1_INDEX, ((reg) & 0xffff));
+ r = RREG32(EVERGREEN_PIF_PHY1_DATA);
+ spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
+ return r;
+}
+
+static inline void si_pif_phy1_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&adev->pcie_idx_lock, flags);
+ WREG32(EVERGREEN_PIF_PHY1_INDEX, ((reg) & 0xffff));
+ WREG32(EVERGREEN_PIF_PHY1_DATA, (v));
+ spin_unlock_irqrestore(&adev->pcie_idx_lock, flags);
+}
+static void si_program_aspm(struct amdgpu_device *adev)
+{
+ u32 data, orig;
+ bool disable_l0s = false, disable_l1 = false, disable_plloff_in_l1 = false;
+ bool disable_clkreq = false;
+
+ if (amdgpu_aspm == 0)
+ return;
+
+ if (adev->flags & AMD_IS_APU)
+ return;
+ orig = data = RREG32_PCIE_PORT(PCIE_LC_N_FTS_CNTL);
+ data &= ~LC_XMIT_N_FTS_MASK;
+ data |= LC_XMIT_N_FTS(0x24) | LC_XMIT_N_FTS_OVERRIDE_EN;
+ if (orig != data)
+ WREG32_PCIE_PORT(PCIE_LC_N_FTS_CNTL, data);
+
+ orig = data = RREG32_PCIE_PORT(PCIE_LC_CNTL3);
+ data |= LC_GO_TO_RECOVERY;
+ if (orig != data)
+ WREG32_PCIE_PORT(PCIE_LC_CNTL3, data);
+
+ orig = data = RREG32_PCIE(PCIE_P_CNTL);
+ data |= P_IGNORE_EDB_ERR;
+ if (orig != data)
+ WREG32_PCIE(PCIE_P_CNTL, data);
+
+ orig = data = RREG32_PCIE_PORT(PCIE_LC_CNTL);
+ data &= ~(LC_L0S_INACTIVITY_MASK | LC_L1_INACTIVITY_MASK);
+ data |= LC_PMI_TO_L1_DIS;
+ if (!disable_l0s)
+ data |= LC_L0S_INACTIVITY(7);
+
+ if (!disable_l1) {
+ data |= LC_L1_INACTIVITY(7);
+ data &= ~LC_PMI_TO_L1_DIS;
+ if (orig != data)
+ WREG32_PCIE_PORT(PCIE_LC_CNTL, data);
+
+ if (!disable_plloff_in_l1) {
+ bool clk_req_support;
+
+ orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_0);
+ data &= ~(PLL_POWER_STATE_IN_OFF_0_MASK | PLL_POWER_STATE_IN_TXS2_0_MASK);
+ data |= PLL_POWER_STATE_IN_OFF_0(7) | PLL_POWER_STATE_IN_TXS2_0(7);
+ if (orig != data)
+ si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_0, data);
+
+ orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_1);
+ data &= ~(PLL_POWER_STATE_IN_OFF_1_MASK | PLL_POWER_STATE_IN_TXS2_1_MASK);
+ data |= PLL_POWER_STATE_IN_OFF_1(7) | PLL_POWER_STATE_IN_TXS2_1(7);
+ if (orig != data)
+ si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_1, data);
+
+ orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_0);
+ data &= ~(PLL_POWER_STATE_IN_OFF_0_MASK | PLL_POWER_STATE_IN_TXS2_0_MASK);
+ data |= PLL_POWER_STATE_IN_OFF_0(7) | PLL_POWER_STATE_IN_TXS2_0(7);
+ if (orig != data)
+ si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_0, data);
+
+ orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_1);
+ data &= ~(PLL_POWER_STATE_IN_OFF_1_MASK | PLL_POWER_STATE_IN_TXS2_1_MASK);
+ data |= PLL_POWER_STATE_IN_OFF_1(7) | PLL_POWER_STATE_IN_TXS2_1(7);
+ if (orig != data)
+ si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_1, data);
+
+ if ((adev->family != CHIP_OLAND) && (adev->family != CHIP_HAINAN)) {
+ orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_0);
+ data &= ~PLL_RAMP_UP_TIME_0_MASK;
+ if (orig != data)
+ si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_0, data);
+
+ orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_1);
+ data &= ~PLL_RAMP_UP_TIME_1_MASK;
+ if (orig != data)
+ si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_1, data);
+
+ orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_2);
+ data &= ~PLL_RAMP_UP_TIME_2_MASK;
+ if (orig != data)
+ si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_2, data);
+
+ orig = data = si_pif_phy0_rreg(adev,PB0_PIF_PWRDOWN_3);
+ data &= ~PLL_RAMP_UP_TIME_3_MASK;
+ if (orig != data)
+ si_pif_phy0_wreg(adev,PB0_PIF_PWRDOWN_3, data);
+
+ orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_0);
+ data &= ~PLL_RAMP_UP_TIME_0_MASK;
+ if (orig != data)
+ si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_0, data);
+
+ orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_1);
+ data &= ~PLL_RAMP_UP_TIME_1_MASK;
+ if (orig != data)
+ si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_1, data);
+
+ orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_2);
+ data &= ~PLL_RAMP_UP_TIME_2_MASK;
+ if (orig != data)
+ si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_2, data);
+
+ orig = data = si_pif_phy1_rreg(adev,PB1_PIF_PWRDOWN_3);
+ data &= ~PLL_RAMP_UP_TIME_3_MASK;
+ if (orig != data)
+ si_pif_phy1_wreg(adev,PB1_PIF_PWRDOWN_3, data);
+ }
+ orig = data = RREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL);
+ data &= ~LC_DYN_LANES_PWR_STATE_MASK;
+ data |= LC_DYN_LANES_PWR_STATE(3);
+ if (orig != data)
+ WREG32_PCIE_PORT(PCIE_LC_LINK_WIDTH_CNTL, data);
+
+ orig = data = si_pif_phy0_rreg(adev,PB0_PIF_CNTL);
+ data &= ~LS2_EXIT_TIME_MASK;
+ if ((adev->family == CHIP_OLAND) || (adev->family == CHIP_HAINAN))
+ data |= LS2_EXIT_TIME(5);
+ if (orig != data)
+ si_pif_phy0_wreg(adev,PB0_PIF_CNTL, data);
+
+ orig = data = si_pif_phy1_rreg(adev,PB1_PIF_CNTL);
+ data &= ~LS2_EXIT_TIME_MASK;
+ if ((adev->family == CHIP_OLAND) || (adev->family == CHIP_HAINAN))
+ data |= LS2_EXIT_TIME(5);
+ if (orig != data)
+ si_pif_phy1_wreg(adev,PB1_PIF_CNTL, data);
+
+ if (!disable_clkreq &&
+ !pci_is_root_bus(adev->pdev->bus)) {
+ struct pci_dev *root = adev->pdev->bus->self;
+ u32 lnkcap;
+
+ clk_req_support = false;
+ pcie_capability_read_dword(root, PCI_EXP_LNKCAP, &lnkcap);
+ if (lnkcap & PCI_EXP_LNKCAP_CLKPM)
+ clk_req_support = true;
+ } else {
+ clk_req_support = false;
+ }
+
+ if (clk_req_support) {
+ orig = data = RREG32_PCIE_PORT(PCIE_LC_CNTL2);
+ data |= LC_ALLOW_PDWN_IN_L1 | LC_ALLOW_PDWN_IN_L23;
+ if (orig != data)
+ WREG32_PCIE_PORT(PCIE_LC_CNTL2, data);
+
+ orig = data = RREG32(THM_CLK_CNTL);
+ data &= ~(CMON_CLK_SEL_MASK | TMON_CLK_SEL_MASK);
+ data |= CMON_CLK_SEL(1) | TMON_CLK_SEL(1);
+ if (orig != data)
+ WREG32(THM_CLK_CNTL, data);
+
+ orig = data = RREG32(MISC_CLK_CNTL);
+ data &= ~(DEEP_SLEEP_CLK_SEL_MASK | ZCLK_SEL_MASK);
+ data |= DEEP_SLEEP_CLK_SEL(1) | ZCLK_SEL(1);
+ if (orig != data)
+ WREG32(MISC_CLK_CNTL, data);
+
+ orig = data = RREG32(CG_CLKPIN_CNTL);
+ data &= ~BCLK_AS_XCLK;
+ if (orig != data)
+ WREG32(CG_CLKPIN_CNTL, data);
+
+ orig = data = RREG32(CG_CLKPIN_CNTL_2);
+ data &= ~FORCE_BIF_REFCLK_EN;
+ if (orig != data)
+ WREG32(CG_CLKPIN_CNTL_2, data);
+
+ orig = data = RREG32(MPLL_BYPASSCLK_SEL);
+ data &= ~MPLL_CLKOUT_SEL_MASK;
+ data |= MPLL_CLKOUT_SEL(4);
+ if (orig != data)
+ WREG32(MPLL_BYPASSCLK_SEL, data);
+
+ orig = data = RREG32(SPLL_CNTL_MODE);
+ data &= ~SPLL_REFCLK_SEL_MASK;
+ if (orig != data)
+ WREG32(SPLL_CNTL_MODE, data);
+ }
+ }
+ } else {
+ if (orig != data)
+ WREG32_PCIE_PORT(PCIE_LC_CNTL, data);
+ }
+
+ orig = data = RREG32_PCIE(PCIE_CNTL2);
+ data |= SLV_MEM_LS_EN | MST_MEM_LS_EN | REPLAY_MEM_LS_EN;
+ if (orig != data)
+ WREG32_PCIE(PCIE_CNTL2, data);
+
+ if (!disable_l0s) {
+ data = RREG32_PCIE_PORT(PCIE_LC_N_FTS_CNTL);
+ if((data & LC_N_FTS_MASK) == LC_N_FTS_MASK) {
+ data = RREG32_PCIE(PCIE_LC_STATUS1);
+ if ((data & LC_REVERSE_XMIT) && (data & LC_REVERSE_RCVR)) {
+ orig = data = RREG32_PCIE_PORT(PCIE_LC_CNTL);
+ data &= ~LC_L0S_INACTIVITY_MASK;
+ if (orig != data)
+ WREG32_PCIE_PORT(PCIE_LC_CNTL, data);
+ }
+ }
+ }
+}
+
+static void si_fix_pci_max_read_req_size(struct amdgpu_device *adev)
+{
+ int readrq;
+ u16 v;
+
+ readrq = pcie_get_readrq(adev->pdev);
+ v = ffs(readrq) - 8;
+ if ((v == 0) || (v == 6) || (v == 7))
+ pcie_set_readrq(adev->pdev, 512);
+}
+
+static int si_common_hw_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ si_fix_pci_max_read_req_size(adev);
+ si_init_golden_registers(adev);
+ si_pcie_gen3_enable(adev);
+ si_program_aspm(adev);
+
+ return 0;
+}
+
+static int si_common_hw_fini(void *handle)
+{
+ return 0;
+}
+
+static int si_common_suspend(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return si_common_hw_fini(adev);
+}
+
+static int si_common_resume(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return si_common_hw_init(adev);
+}
+
+static bool si_common_is_idle(void *handle)
+{
+ return true;
+}
+
+static int si_common_wait_for_idle(void *handle)
+{
+ return 0;
+}
+
+static int si_common_soft_reset(void *handle)
+{
+ return 0;
+}
+
+static int si_common_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ return 0;
+}
+
+static int si_common_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ return 0;
+}
+
+const struct amd_ip_funcs si_common_ip_funcs = {
+ .name = "si_common",
+ .early_init = si_common_early_init,
+ .late_init = NULL,
+ .sw_init = si_common_sw_init,
+ .sw_fini = si_common_sw_fini,
+ .hw_init = si_common_hw_init,
+ .hw_fini = si_common_hw_fini,
+ .suspend = si_common_suspend,
+ .resume = si_common_resume,
+ .is_idle = si_common_is_idle,
+ .wait_for_idle = si_common_wait_for_idle,
+ .soft_reset = si_common_soft_reset,
+ .set_clockgating_state = si_common_set_clockgating_state,
+ .set_powergating_state = si_common_set_powergating_state,
+};
+
+static const struct amdgpu_ip_block_version verde_ip_blocks[] =
+{
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &si_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v6_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &si_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &dce_v6_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gfx_v6_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &si_dma_ip_funcs,
+ },
+/* {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 3,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &si_null_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &si_null_ip_funcs,
+ },
+ */
+};
+
+
+static const struct amdgpu_ip_block_version hainan_ip_blocks[] =
+{
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &si_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v6_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &si_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gfx_v6_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &si_dma_ip_funcs,
+ },
+};
+
+int si_set_ip_blocks(struct amdgpu_device *adev)
+{
+ switch (adev->asic_type) {
+ case CHIP_VERDE:
+ case CHIP_TAHITI:
+ case CHIP_PITCAIRN:
+ case CHIP_OLAND:
+ adev->ip_blocks = verde_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(verde_ip_blocks);
+ break;
+ case CHIP_HAINAN:
+ adev->ip_blocks = hainan_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(hainan_ip_blocks);
+ break;
+ default:
+ BUG();
+ }
+ return 0;
+}
+
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __SI_H__
+#define __SI_H__
+
+extern const struct amd_ip_funcs si_common_ip_funcs;
+
+void si_srbm_select(struct amdgpu_device *adev,
+ u32 me, u32 pipe, u32 queue, u32 vmid);
+int si_set_ip_blocks(struct amdgpu_device *adev);
+
+#endif
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Alex Deucher
+ */
+#include <drm/drmP.h>
+#include "amdgpu.h"
+#include "amdgpu_trace.h"
+#include "si/sid.h"
+
+const u32 sdma_offsets[SDMA_MAX_INSTANCE] =
+{
+ DMA0_REGISTER_OFFSET,
+ DMA1_REGISTER_OFFSET
+};
+
+static void si_dma_set_ring_funcs(struct amdgpu_device *adev);
+static void si_dma_set_buffer_funcs(struct amdgpu_device *adev);
+static void si_dma_set_vm_pte_funcs(struct amdgpu_device *adev);
+static void si_dma_set_irq_funcs(struct amdgpu_device *adev);
+
+static uint32_t si_dma_ring_get_rptr(struct amdgpu_ring *ring)
+{
+ return ring->adev->wb.wb[ring->rptr_offs>>2];
+}
+
+static uint32_t si_dma_ring_get_wptr(struct amdgpu_ring *ring)
+{
+ struct amdgpu_device *adev = ring->adev;
+ u32 me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1;
+
+ return (RREG32(DMA_RB_WPTR + sdma_offsets[me]) & 0x3fffc) >> 2;
+}
+
+static void si_dma_ring_set_wptr(struct amdgpu_ring *ring)
+{
+ struct amdgpu_device *adev = ring->adev;
+ u32 me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1;
+
+ WREG32(DMA_RB_WPTR + sdma_offsets[me], (ring->wptr << 2) & 0x3fffc);
+}
+
+static void si_dma_ring_emit_ib(struct amdgpu_ring *ring,
+ struct amdgpu_ib *ib,
+ unsigned vm_id, bool ctx_switch)
+{
+ /* The indirect buffer packet must end on an 8 DW boundary in the DMA ring.
+ * Pad as necessary with NOPs.
+ */
+ while ((ring->wptr & 7) != 5)
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0));
+ amdgpu_ring_write(ring, DMA_IB_PACKET(DMA_PACKET_INDIRECT_BUFFER, vm_id, 0));
+ amdgpu_ring_write(ring, (ib->gpu_addr & 0xFFFFFFE0));
+ amdgpu_ring_write(ring, (ib->length_dw << 12) | (upper_32_bits(ib->gpu_addr) & 0xFF));
+
+}
+
+static void si_dma_ring_emit_hdp_flush(struct amdgpu_ring *ring)
+{
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
+ amdgpu_ring_write(ring, (0xf << 16) | (HDP_MEM_COHERENCY_FLUSH_CNTL));
+ amdgpu_ring_write(ring, 1);
+}
+
+static void si_dma_ring_emit_hdp_invalidate(struct amdgpu_ring *ring)
+{
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
+ amdgpu_ring_write(ring, (0xf << 16) | (HDP_DEBUG0));
+ amdgpu_ring_write(ring, 1);
+}
+
+/**
+ * si_dma_ring_emit_fence - emit a fence on the DMA ring
+ *
+ * @ring: amdgpu ring pointer
+ * @fence: amdgpu fence object
+ *
+ * Add a DMA fence packet to the ring to write
+ * the fence seq number and DMA trap packet to generate
+ * an interrupt if needed (VI).
+ */
+static void si_dma_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
+ unsigned flags)
+{
+
+ bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT;
+ /* write the fence */
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_FENCE, 0, 0, 0, 0));
+ amdgpu_ring_write(ring, addr & 0xfffffffc);
+ amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xff));
+ amdgpu_ring_write(ring, seq);
+ /* optionally write high bits as well */
+ if (write64bit) {
+ addr += 4;
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_FENCE, 0, 0, 0, 0));
+ amdgpu_ring_write(ring, addr & 0xfffffffc);
+ amdgpu_ring_write(ring, (upper_32_bits(addr) & 0xff));
+ amdgpu_ring_write(ring, upper_32_bits(seq));
+ }
+ /* generate an interrupt */
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_TRAP, 0, 0, 0, 0));
+}
+
+static void si_dma_stop(struct amdgpu_device *adev)
+{
+ struct amdgpu_ring *ring;
+ u32 rb_cntl;
+ unsigned i;
+
+ for (i = 0; i < adev->sdma.num_instances; i++) {
+ ring = &adev->sdma.instance[i].ring;
+ /* dma0 */
+ rb_cntl = RREG32(DMA_RB_CNTL + sdma_offsets[i]);
+ rb_cntl &= ~DMA_RB_ENABLE;
+ WREG32(DMA_RB_CNTL + sdma_offsets[i], rb_cntl);
+
+ if (adev->mman.buffer_funcs_ring == ring)
+ amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size);
+ ring->ready = false;
+ }
+}
+
+static int si_dma_start(struct amdgpu_device *adev)
+{
+ struct amdgpu_ring *ring;
+ u32 rb_cntl, dma_cntl, ib_cntl, rb_bufsz;
+ int i, r;
+ uint64_t rptr_addr;
+
+ for (i = 0; i < adev->sdma.num_instances; i++) {
+ ring = &adev->sdma.instance[i].ring;
+
+ WREG32(DMA_SEM_INCOMPLETE_TIMER_CNTL + sdma_offsets[i], 0);
+ WREG32(DMA_SEM_WAIT_FAIL_TIMER_CNTL + sdma_offsets[i], 0);
+
+ /* Set ring buffer size in dwords */
+ rb_bufsz = order_base_2(ring->ring_size / 4);
+ rb_cntl = rb_bufsz << 1;
+#ifdef __BIG_ENDIAN
+ rb_cntl |= DMA_RB_SWAP_ENABLE | DMA_RPTR_WRITEBACK_SWAP_ENABLE;
+#endif
+ WREG32(DMA_RB_CNTL + sdma_offsets[i], rb_cntl);
+
+ /* Initialize the ring buffer's read and write pointers */
+ WREG32(DMA_RB_RPTR + sdma_offsets[i], 0);
+ WREG32(DMA_RB_WPTR + sdma_offsets[i], 0);
+
+ rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4);
+
+ WREG32(DMA_RB_RPTR_ADDR_LO + sdma_offsets[i], lower_32_bits(rptr_addr));
+ WREG32(DMA_RB_RPTR_ADDR_HI + sdma_offsets[i], upper_32_bits(rptr_addr) & 0xFF);
+
+ rb_cntl |= DMA_RPTR_WRITEBACK_ENABLE;
+
+ WREG32(DMA_RB_BASE + sdma_offsets[i], ring->gpu_addr >> 8);
+
+ /* enable DMA IBs */
+ ib_cntl = DMA_IB_ENABLE | CMD_VMID_FORCE;
+#ifdef __BIG_ENDIAN
+ ib_cntl |= DMA_IB_SWAP_ENABLE;
+#endif
+ WREG32(DMA_IB_CNTL + sdma_offsets[i], ib_cntl);
+
+ dma_cntl = RREG32(DMA_CNTL + sdma_offsets[i]);
+ dma_cntl &= ~CTXEMPTY_INT_ENABLE;
+ WREG32(DMA_CNTL + sdma_offsets[i], dma_cntl);
+
+ ring->wptr = 0;
+ WREG32(DMA_RB_WPTR + sdma_offsets[i], ring->wptr << 2);
+ WREG32(DMA_RB_CNTL + sdma_offsets[i], rb_cntl | DMA_RB_ENABLE);
+
+ ring->ready = true;
+
+ r = amdgpu_ring_test_ring(ring);
+ if (r) {
+ ring->ready = false;
+ return r;
+ }
+
+ if (adev->mman.buffer_funcs_ring == ring)
+ amdgpu_ttm_set_active_vram_size(adev, adev->mc.real_vram_size);
+ }
+
+ return 0;
+}
+
+/**
+ * si_dma_ring_test_ring - simple async dma engine test
+ *
+ * @ring: amdgpu_ring structure holding ring information
+ *
+ * Test the DMA engine by writing using it to write an
+ * value to memory. (VI).
+ * Returns 0 for success, error for failure.
+ */
+static int si_dma_ring_test_ring(struct amdgpu_ring *ring)
+{
+ struct amdgpu_device *adev = ring->adev;
+ unsigned i;
+ unsigned index;
+ int r;
+ u32 tmp;
+ u64 gpu_addr;
+
+ r = amdgpu_wb_get(adev, &index);
+ if (r) {
+ dev_err(adev->dev, "(%d) failed to allocate wb slot\n", r);
+ return r;
+ }
+
+ gpu_addr = adev->wb.gpu_addr + (index * 4);
+ tmp = 0xCAFEDEAD;
+ adev->wb.wb[index] = cpu_to_le32(tmp);
+
+ r = amdgpu_ring_alloc(ring, 4);
+ if (r) {
+ DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r);
+ amdgpu_wb_free(adev, index);
+ return r;
+ }
+
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 0, 1));
+ amdgpu_ring_write(ring, lower_32_bits(gpu_addr));
+ amdgpu_ring_write(ring, upper_32_bits(gpu_addr) & 0xff);
+ amdgpu_ring_write(ring, 0xDEADBEEF);
+ amdgpu_ring_commit(ring);
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ tmp = le32_to_cpu(adev->wb.wb[index]);
+ if (tmp == 0xDEADBEEF)
+ break;
+ DRM_UDELAY(1);
+ }
+
+ if (i < adev->usec_timeout) {
+ DRM_INFO("ring test on %d succeeded in %d usecs\n", ring->idx, i);
+ } else {
+ DRM_ERROR("amdgpu: ring %d test failed (0x%08X)\n",
+ ring->idx, tmp);
+ r = -EINVAL;
+ }
+ amdgpu_wb_free(adev, index);
+
+ return r;
+}
+
+/**
+ * si_dma_ring_test_ib - test an IB on the DMA engine
+ *
+ * @ring: amdgpu_ring structure holding ring information
+ *
+ * Test a simple IB in the DMA ring (VI).
+ * Returns 0 on success, error on failure.
+ */
+static int si_dma_ring_test_ib(struct amdgpu_ring *ring, long timeout)
+{
+ struct amdgpu_device *adev = ring->adev;
+ struct amdgpu_ib ib;
+ struct fence *f = NULL;
+ unsigned index;
+ u32 tmp = 0;
+ u64 gpu_addr;
+ long r;
+
+ r = amdgpu_wb_get(adev, &index);
+ if (r) {
+ dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r);
+ return r;
+ }
+
+ gpu_addr = adev->wb.gpu_addr + (index * 4);
+ tmp = 0xCAFEDEAD;
+ adev->wb.wb[index] = cpu_to_le32(tmp);
+ memset(&ib, 0, sizeof(ib));
+ r = amdgpu_ib_get(adev, NULL, 256, &ib);
+ if (r) {
+ DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r);
+ goto err0;
+ }
+
+ ib.ptr[0] = DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 0, 1);
+ ib.ptr[1] = lower_32_bits(gpu_addr);
+ ib.ptr[2] = upper_32_bits(gpu_addr) & 0xff;
+ ib.ptr[3] = 0xDEADBEEF;
+ ib.length_dw = 4;
+ r = amdgpu_ib_schedule(ring, 1, &ib, NULL, NULL, &f);
+ if (r)
+ goto err1;
+
+ r = fence_wait_timeout(f, false, timeout);
+ if (r == 0) {
+ DRM_ERROR("amdgpu: IB test timed out\n");
+ r = -ETIMEDOUT;
+ goto err1;
+ } else if (r < 0) {
+ DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
+ goto err1;
+ }
+ tmp = le32_to_cpu(adev->wb.wb[index]);
+ if (tmp == 0xDEADBEEF) {
+ DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
+ r = 0;
+ } else {
+ DRM_ERROR("amdgpu: ib test failed (0x%08X)\n", tmp);
+ r = -EINVAL;
+ }
+
+err1:
+ amdgpu_ib_free(adev, &ib, NULL);
+ fence_put(f);
+err0:
+ amdgpu_wb_free(adev, index);
+ return r;
+}
+
+/**
+ * cik_dma_vm_copy_pte - update PTEs by copying them from the GART
+ *
+ * @ib: indirect buffer to fill with commands
+ * @pe: addr of the page entry
+ * @src: src addr to copy from
+ * @count: number of page entries to update
+ *
+ * Update PTEs by copying them from the GART using DMA (SI).
+ */
+static void si_dma_vm_copy_pte(struct amdgpu_ib *ib,
+ uint64_t pe, uint64_t src,
+ unsigned count)
+{
+ unsigned bytes = count * 8;
+
+ ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_COPY,
+ 1, 0, 0, bytes);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+ ib->ptr[ib->length_dw++] = lower_32_bits(src);
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff;
+ ib->ptr[ib->length_dw++] = upper_32_bits(src) & 0xff;
+}
+
+/**
+ * si_dma_vm_write_pte - update PTEs by writing them manually
+ *
+ * @ib: indirect buffer to fill with commands
+ * @pe: addr of the page entry
+ * @value: dst addr to write into pe
+ * @count: number of page entries to update
+ * @incr: increase next addr by incr bytes
+ *
+ * Update PTEs by writing them manually using DMA (SI).
+ */
+static void si_dma_vm_write_pte(struct amdgpu_ib *ib, uint64_t pe,
+ uint64_t value, unsigned count,
+ uint32_t incr)
+{
+ unsigned ndw = count * 2;
+
+ ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_WRITE, 0, 0, 0, ndw);
+ ib->ptr[ib->length_dw++] = lower_32_bits(pe);
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe);
+ for (; ndw > 0; ndw -= 2) {
+ ib->ptr[ib->length_dw++] = lower_32_bits(value);
+ ib->ptr[ib->length_dw++] = upper_32_bits(value);
+ value += incr;
+ }
+}
+
+/**
+ * si_dma_vm_set_pte_pde - update the page tables using sDMA
+ *
+ * @ib: indirect buffer to fill with commands
+ * @pe: addr of the page entry
+ * @addr: dst addr to write into pe
+ * @count: number of page entries to update
+ * @incr: increase next addr by incr bytes
+ * @flags: access flags
+ *
+ * Update the page tables using sDMA (CIK).
+ */
+static void si_dma_vm_set_pte_pde(struct amdgpu_ib *ib,
+ uint64_t pe,
+ uint64_t addr, unsigned count,
+ uint32_t incr, uint32_t flags)
+{
+ uint64_t value;
+ unsigned ndw;
+
+ while (count) {
+ ndw = count * 2;
+ if (ndw > 0xFFFFE)
+ ndw = 0xFFFFE;
+
+ if (flags & AMDGPU_PTE_VALID)
+ value = addr;
+ else
+ value = 0;
+
+ /* for physically contiguous pages (vram) */
+ ib->ptr[ib->length_dw++] = DMA_PTE_PDE_PACKET(ndw);
+ ib->ptr[ib->length_dw++] = pe; /* dst addr */
+ ib->ptr[ib->length_dw++] = upper_32_bits(pe) & 0xff;
+ ib->ptr[ib->length_dw++] = flags; /* mask */
+ ib->ptr[ib->length_dw++] = 0;
+ ib->ptr[ib->length_dw++] = value; /* value */
+ ib->ptr[ib->length_dw++] = upper_32_bits(value);
+ ib->ptr[ib->length_dw++] = incr; /* increment size */
+ ib->ptr[ib->length_dw++] = 0;
+ pe += ndw * 4;
+ addr += (ndw / 2) * incr;
+ count -= ndw / 2;
+ }
+}
+
+/**
+ * si_dma_pad_ib - pad the IB to the required number of dw
+ *
+ * @ib: indirect buffer to fill with padding
+ *
+ */
+static void si_dma_ring_pad_ib(struct amdgpu_ring *ring, struct amdgpu_ib *ib)
+{
+ while (ib->length_dw & 0x7)
+ ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0);
+}
+
+/**
+ * cik_sdma_ring_emit_pipeline_sync - sync the pipeline
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Make sure all previous operations are completed (CIK).
+ */
+static void si_dma_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
+{
+ uint32_t seq = ring->fence_drv.sync_seq;
+ uint64_t addr = ring->fence_drv.gpu_addr;
+
+ /* wait for idle */
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_POLL_REG_MEM, 0, 0, 0, 0) |
+ (1 << 27)); /* Poll memory */
+ amdgpu_ring_write(ring, lower_32_bits(addr));
+ amdgpu_ring_write(ring, (0xff << 16) | upper_32_bits(addr)); /* retry, addr_hi */
+ amdgpu_ring_write(ring, 0xffffffff); /* mask */
+ amdgpu_ring_write(ring, seq); /* value */
+ amdgpu_ring_write(ring, (3 << 28) | 0x20); /* func(equal) | poll interval */
+}
+
+/**
+ * si_dma_ring_emit_vm_flush - cik vm flush using sDMA
+ *
+ * @ring: amdgpu_ring pointer
+ * @vm: amdgpu_vm pointer
+ *
+ * Update the page table base and flush the VM TLB
+ * using sDMA (VI).
+ */
+static void si_dma_ring_emit_vm_flush(struct amdgpu_ring *ring,
+ unsigned vm_id, uint64_t pd_addr)
+{
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
+ if (vm_id < 8)
+ amdgpu_ring_write(ring, (0xf << 16) | (VM_CONTEXT0_PAGE_TABLE_BASE_ADDR + vm_id));
+ else
+ amdgpu_ring_write(ring, (0xf << 16) | (VM_CONTEXT8_PAGE_TABLE_BASE_ADDR + (vm_id - 8)));
+ amdgpu_ring_write(ring, pd_addr >> 12);
+
+ /* bits 0-7 are the VM contexts0-7 */
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_SRBM_WRITE, 0, 0, 0, 0));
+ amdgpu_ring_write(ring, (0xf << 16) | (VM_INVALIDATE_REQUEST));
+ amdgpu_ring_write(ring, 1 << vm_id);
+
+ /* wait for invalidate to complete */
+ amdgpu_ring_write(ring, DMA_PACKET(DMA_PACKET_POLL_REG_MEM, 0, 0, 0, 0));
+ amdgpu_ring_write(ring, VM_INVALIDATE_REQUEST);
+ amdgpu_ring_write(ring, 0xff << 16); /* retry */
+ amdgpu_ring_write(ring, 1 << vm_id); /* mask */
+ amdgpu_ring_write(ring, 0); /* value */
+ amdgpu_ring_write(ring, (0 << 28) | 0x20); /* func(always) | poll interval */
+}
+
+static unsigned si_dma_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 7 + 3; /* si_dma_ring_emit_ib */
+}
+
+static unsigned si_dma_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 3 + /* si_dma_ring_emit_hdp_flush */
+ 3 + /* si_dma_ring_emit_hdp_invalidate */
+ 6 + /* si_dma_ring_emit_pipeline_sync */
+ 12 + /* si_dma_ring_emit_vm_flush */
+ 9 + 9 + 9; /* si_dma_ring_emit_fence x3 for user fence, vm fence */
+}
+
+static int si_dma_early_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ adev->sdma.num_instances = 2;
+
+ si_dma_set_ring_funcs(adev);
+ si_dma_set_buffer_funcs(adev);
+ si_dma_set_vm_pte_funcs(adev);
+ si_dma_set_irq_funcs(adev);
+
+ return 0;
+}
+
+static int si_dma_sw_init(void *handle)
+{
+ struct amdgpu_ring *ring;
+ int r, i;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ /* DMA0 trap event */
+ r = amdgpu_irq_add_id(adev, 224, &adev->sdma.trap_irq);
+ if (r)
+ return r;
+
+ /* DMA1 trap event */
+ r = amdgpu_irq_add_id(adev, 244, &adev->sdma.trap_irq_1);
+ if (r)
+ return r;
+
+ for (i = 0; i < adev->sdma.num_instances; i++) {
+ ring = &adev->sdma.instance[i].ring;
+ ring->ring_obj = NULL;
+ ring->use_doorbell = false;
+ sprintf(ring->name, "sdma%d", i);
+ r = amdgpu_ring_init(adev, ring, 1024,
+ DMA_PACKET(DMA_PACKET_NOP, 0, 0, 0, 0), 0xf,
+ &adev->sdma.trap_irq,
+ (i == 0) ?
+ AMDGPU_SDMA_IRQ_TRAP0 : AMDGPU_SDMA_IRQ_TRAP1,
+ AMDGPU_RING_TYPE_SDMA);
+ if (r)
+ return r;
+ }
+
+ return r;
+}
+
+static int si_dma_sw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ int i;
+
+ for (i = 0; i < adev->sdma.num_instances; i++)
+ amdgpu_ring_fini(&adev->sdma.instance[i].ring);
+
+ return 0;
+}
+
+static int si_dma_hw_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return si_dma_start(adev);
+}
+
+static int si_dma_hw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ si_dma_stop(adev);
+
+ return 0;
+}
+
+static int si_dma_suspend(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return si_dma_hw_fini(adev);
+}
+
+static int si_dma_resume(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return si_dma_hw_init(adev);
+}
+
+static bool si_dma_is_idle(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 tmp = RREG32(SRBM_STATUS2);
+
+ if (tmp & (DMA_BUSY_MASK | DMA1_BUSY_MASK))
+ return false;
+
+ return true;
+}
+
+static int si_dma_wait_for_idle(void *handle)
+{
+ unsigned i;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (si_dma_is_idle(handle))
+ return 0;
+ udelay(1);
+ }
+ return -ETIMEDOUT;
+}
+
+static int si_dma_soft_reset(void *handle)
+{
+ DRM_INFO("si_dma_soft_reset --- not implemented !!!!!!!\n");
+ return 0;
+}
+
+static int si_dma_set_trap_irq_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *src,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ u32 sdma_cntl;
+
+ switch (type) {
+ case AMDGPU_SDMA_IRQ_TRAP0:
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ sdma_cntl = RREG32(DMA_CNTL + DMA0_REGISTER_OFFSET);
+ sdma_cntl &= ~TRAP_ENABLE;
+ WREG32(DMA_CNTL + DMA0_REGISTER_OFFSET, sdma_cntl);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ sdma_cntl = RREG32(DMA_CNTL + DMA0_REGISTER_OFFSET);
+ sdma_cntl |= TRAP_ENABLE;
+ WREG32(DMA_CNTL + DMA0_REGISTER_OFFSET, sdma_cntl);
+ break;
+ default:
+ break;
+ }
+ break;
+ case AMDGPU_SDMA_IRQ_TRAP1:
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ sdma_cntl = RREG32(DMA_CNTL + DMA1_REGISTER_OFFSET);
+ sdma_cntl &= ~TRAP_ENABLE;
+ WREG32(DMA_CNTL + DMA1_REGISTER_OFFSET, sdma_cntl);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ sdma_cntl = RREG32(DMA_CNTL + DMA1_REGISTER_OFFSET);
+ sdma_cntl |= TRAP_ENABLE;
+ WREG32(DMA_CNTL + DMA1_REGISTER_OFFSET, sdma_cntl);
+ break;
+ default:
+ break;
+ }
+ break;
+ default:
+ break;
+ }
+ return 0;
+}
+
+static int si_dma_process_trap_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ amdgpu_fence_process(&adev->sdma.instance[0].ring);
+
+ return 0;
+}
+
+static int si_dma_process_trap_irq_1(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ amdgpu_fence_process(&adev->sdma.instance[1].ring);
+
+ return 0;
+}
+
+static int si_dma_process_illegal_inst_irq(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ DRM_ERROR("Illegal instruction in SDMA command stream\n");
+ schedule_work(&adev->reset_work);
+ return 0;
+}
+
+static int si_dma_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ u32 orig, data, offset;
+ int i;
+ bool enable;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ enable = (state == AMD_CG_STATE_GATE) ? true : false;
+
+ if (enable && (adev->cg_flags & AMD_CG_SUPPORT_SDMA_MGCG)) {
+ for (i = 0; i < adev->sdma.num_instances; i++) {
+ if (i == 0)
+ offset = DMA0_REGISTER_OFFSET;
+ else
+ offset = DMA1_REGISTER_OFFSET;
+ orig = data = RREG32(DMA_POWER_CNTL + offset);
+ data &= ~MEM_POWER_OVERRIDE;
+ if (data != orig)
+ WREG32(DMA_POWER_CNTL + offset, data);
+ WREG32(DMA_CLK_CTRL + offset, 0x00000100);
+ }
+ } else {
+ for (i = 0; i < adev->sdma.num_instances; i++) {
+ if (i == 0)
+ offset = DMA0_REGISTER_OFFSET;
+ else
+ offset = DMA1_REGISTER_OFFSET;
+ orig = data = RREG32(DMA_POWER_CNTL + offset);
+ data |= MEM_POWER_OVERRIDE;
+ if (data != orig)
+ WREG32(DMA_POWER_CNTL + offset, data);
+
+ orig = data = RREG32(DMA_CLK_CTRL + offset);
+ data = 0xff000000;
+ if (data != orig)
+ WREG32(DMA_CLK_CTRL + offset, data);
+ }
+ }
+
+ return 0;
+}
+
+static int si_dma_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ u32 tmp;
+
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ WREG32(DMA_PGFSM_WRITE, 0x00002000);
+ WREG32(DMA_PGFSM_CONFIG, 0x100010ff);
+
+ for (tmp = 0; tmp < 5; tmp++)
+ WREG32(DMA_PGFSM_WRITE, 0);
+
+ return 0;
+}
+
+const struct amd_ip_funcs si_dma_ip_funcs = {
+ .name = "si_dma",
+ .early_init = si_dma_early_init,
+ .late_init = NULL,
+ .sw_init = si_dma_sw_init,
+ .sw_fini = si_dma_sw_fini,
+ .hw_init = si_dma_hw_init,
+ .hw_fini = si_dma_hw_fini,
+ .suspend = si_dma_suspend,
+ .resume = si_dma_resume,
+ .is_idle = si_dma_is_idle,
+ .wait_for_idle = si_dma_wait_for_idle,
+ .soft_reset = si_dma_soft_reset,
+ .set_clockgating_state = si_dma_set_clockgating_state,
+ .set_powergating_state = si_dma_set_powergating_state,
+};
+
+static const struct amdgpu_ring_funcs si_dma_ring_funcs = {
+ .get_rptr = si_dma_ring_get_rptr,
+ .get_wptr = si_dma_ring_get_wptr,
+ .set_wptr = si_dma_ring_set_wptr,
+ .parse_cs = NULL,
+ .emit_ib = si_dma_ring_emit_ib,
+ .emit_fence = si_dma_ring_emit_fence,
+ .emit_pipeline_sync = si_dma_ring_emit_pipeline_sync,
+ .emit_vm_flush = si_dma_ring_emit_vm_flush,
+ .emit_hdp_flush = si_dma_ring_emit_hdp_flush,
+ .emit_hdp_invalidate = si_dma_ring_emit_hdp_invalidate,
+ .test_ring = si_dma_ring_test_ring,
+ .test_ib = si_dma_ring_test_ib,
+ .insert_nop = amdgpu_ring_insert_nop,
+ .pad_ib = si_dma_ring_pad_ib,
+ .get_emit_ib_size = si_dma_ring_get_emit_ib_size,
+ .get_dma_frame_size = si_dma_ring_get_dma_frame_size,
+};
+
+static void si_dma_set_ring_funcs(struct amdgpu_device *adev)
+{
+ int i;
+
+ for (i = 0; i < adev->sdma.num_instances; i++)
+ adev->sdma.instance[i].ring.funcs = &si_dma_ring_funcs;
+}
+
+static const struct amdgpu_irq_src_funcs si_dma_trap_irq_funcs = {
+ .set = si_dma_set_trap_irq_state,
+ .process = si_dma_process_trap_irq,
+};
+
+static const struct amdgpu_irq_src_funcs si_dma_trap_irq_funcs_1 = {
+ .set = si_dma_set_trap_irq_state,
+ .process = si_dma_process_trap_irq_1,
+};
+
+static const struct amdgpu_irq_src_funcs si_dma_illegal_inst_irq_funcs = {
+ .process = si_dma_process_illegal_inst_irq,
+};
+
+static void si_dma_set_irq_funcs(struct amdgpu_device *adev)
+{
+ adev->sdma.trap_irq.num_types = AMDGPU_SDMA_IRQ_LAST;
+ adev->sdma.trap_irq.funcs = &si_dma_trap_irq_funcs;
+ adev->sdma.trap_irq_1.funcs = &si_dma_trap_irq_funcs_1;
+ adev->sdma.illegal_inst_irq.funcs = &si_dma_illegal_inst_irq_funcs;
+}
+
+/**
+ * si_dma_emit_copy_buffer - copy buffer using the sDMA engine
+ *
+ * @ring: amdgpu_ring structure holding ring information
+ * @src_offset: src GPU address
+ * @dst_offset: dst GPU address
+ * @byte_count: number of bytes to xfer
+ *
+ * Copy GPU buffers using the DMA engine (VI).
+ * Used by the amdgpu ttm implementation to move pages if
+ * registered as the asic copy callback.
+ */
+static void si_dma_emit_copy_buffer(struct amdgpu_ib *ib,
+ uint64_t src_offset,
+ uint64_t dst_offset,
+ uint32_t byte_count)
+{
+ ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_COPY,
+ 1, 0, 0, byte_count);
+ ib->ptr[ib->length_dw++] = lower_32_bits(dst_offset);
+ ib->ptr[ib->length_dw++] = lower_32_bits(src_offset);
+ ib->ptr[ib->length_dw++] = upper_32_bits(dst_offset) & 0xff;
+ ib->ptr[ib->length_dw++] = upper_32_bits(src_offset) & 0xff;
+}
+
+/**
+ * si_dma_emit_fill_buffer - fill buffer using the sDMA engine
+ *
+ * @ring: amdgpu_ring structure holding ring information
+ * @src_data: value to write to buffer
+ * @dst_offset: dst GPU address
+ * @byte_count: number of bytes to xfer
+ *
+ * Fill GPU buffers using the DMA engine (VI).
+ */
+static void si_dma_emit_fill_buffer(struct amdgpu_ib *ib,
+ uint32_t src_data,
+ uint64_t dst_offset,
+ uint32_t byte_count)
+{
+ ib->ptr[ib->length_dw++] = DMA_PACKET(DMA_PACKET_CONSTANT_FILL,
+ 0, 0, 0, byte_count / 4);
+ ib->ptr[ib->length_dw++] = lower_32_bits(dst_offset);
+ ib->ptr[ib->length_dw++] = src_data;
+ ib->ptr[ib->length_dw++] = upper_32_bits(dst_offset) << 16;
+}
+
+
+static const struct amdgpu_buffer_funcs si_dma_buffer_funcs = {
+ .copy_max_bytes = 0xffff8,
+ .copy_num_dw = 5,
+ .emit_copy_buffer = si_dma_emit_copy_buffer,
+
+ .fill_max_bytes = 0xffff8,
+ .fill_num_dw = 4,
+ .emit_fill_buffer = si_dma_emit_fill_buffer,
+};
+
+static void si_dma_set_buffer_funcs(struct amdgpu_device *adev)
+{
+ if (adev->mman.buffer_funcs == NULL) {
+ adev->mman.buffer_funcs = &si_dma_buffer_funcs;
+ adev->mman.buffer_funcs_ring = &adev->sdma.instance[0].ring;
+ }
+}
+
+static const struct amdgpu_vm_pte_funcs si_dma_vm_pte_funcs = {
+ .copy_pte = si_dma_vm_copy_pte,
+ .write_pte = si_dma_vm_write_pte,
+ .set_pte_pde = si_dma_vm_set_pte_pde,
+};
+
+static void si_dma_set_vm_pte_funcs(struct amdgpu_device *adev)
+{
+ unsigned i;
+
+ if (adev->vm_manager.vm_pte_funcs == NULL) {
+ adev->vm_manager.vm_pte_funcs = &si_dma_vm_pte_funcs;
+ for (i = 0; i < adev->sdma.num_instances; i++)
+ adev->vm_manager.vm_pte_rings[i] =
+ &adev->sdma.instance[i].ring;
+
+ adev->vm_manager.vm_pte_num_rings = adev->sdma.num_instances;
+ }
+}
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __SI_DMA_H__
+#define __SI_DMA_H__
+
+extern const struct amd_ip_funcs si_dma_ip_funcs;
+
+#endif
--- /dev/null
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "drmP.h"
+#include "amdgpu.h"
+#include "amdgpu_pm.h"
+#include "amdgpu_dpm.h"
+#include "amdgpu_atombios.h"
+#include "si/sid.h"
+#include "r600_dpm.h"
+#include "si_dpm.h"
+#include "atom.h"
+#include "../include/pptable.h"
+#include <linux/math64.h>
+#include <linux/seq_file.h>
+#include <linux/firmware.h>
+
+#define MC_CG_ARB_FREQ_F0 0x0a
+#define MC_CG_ARB_FREQ_F1 0x0b
+#define MC_CG_ARB_FREQ_F2 0x0c
+#define MC_CG_ARB_FREQ_F3 0x0d
+
+#define SMC_RAM_END 0x20000
+
+#define SCLK_MIN_DEEPSLEEP_FREQ 1350
+
+
+/* sizeof(ATOM_PPLIB_EXTENDEDHEADER) */
+#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V2 12
+#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V3 14
+#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V4 16
+#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V5 18
+#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V6 20
+#define SIZE_OF_ATOM_PPLIB_EXTENDEDHEADER_V7 22
+
+#define BIOS_SCRATCH_4 0x5cd
+
+MODULE_FIRMWARE("radeon/tahiti_smc.bin");
+MODULE_FIRMWARE("radeon/tahiti_k_smc.bin");
+MODULE_FIRMWARE("radeon/pitcairn_smc.bin");
+MODULE_FIRMWARE("radeon/pitcairn_k_smc.bin");
+MODULE_FIRMWARE("radeon/verde_smc.bin");
+MODULE_FIRMWARE("radeon/verde_k_smc.bin");
+MODULE_FIRMWARE("radeon/oland_smc.bin");
+MODULE_FIRMWARE("radeon/oland_k_smc.bin");
+MODULE_FIRMWARE("radeon/hainan_smc.bin");
+MODULE_FIRMWARE("radeon/hainan_k_smc.bin");
+
+union power_info {
+ struct _ATOM_POWERPLAY_INFO info;
+ struct _ATOM_POWERPLAY_INFO_V2 info_2;
+ struct _ATOM_POWERPLAY_INFO_V3 info_3;
+ struct _ATOM_PPLIB_POWERPLAYTABLE pplib;
+ struct _ATOM_PPLIB_POWERPLAYTABLE2 pplib2;
+ struct _ATOM_PPLIB_POWERPLAYTABLE3 pplib3;
+ struct _ATOM_PPLIB_POWERPLAYTABLE4 pplib4;
+ struct _ATOM_PPLIB_POWERPLAYTABLE5 pplib5;
+};
+
+union fan_info {
+ struct _ATOM_PPLIB_FANTABLE fan;
+ struct _ATOM_PPLIB_FANTABLE2 fan2;
+ struct _ATOM_PPLIB_FANTABLE3 fan3;
+};
+
+union pplib_clock_info {
+ struct _ATOM_PPLIB_R600_CLOCK_INFO r600;
+ struct _ATOM_PPLIB_RS780_CLOCK_INFO rs780;
+ struct _ATOM_PPLIB_EVERGREEN_CLOCK_INFO evergreen;
+ struct _ATOM_PPLIB_SUMO_CLOCK_INFO sumo;
+ struct _ATOM_PPLIB_SI_CLOCK_INFO si;
+};
+
+static const u32 r600_utc[R600_PM_NUMBER_OF_TC] =
+{
+ R600_UTC_DFLT_00,
+ R600_UTC_DFLT_01,
+ R600_UTC_DFLT_02,
+ R600_UTC_DFLT_03,
+ R600_UTC_DFLT_04,
+ R600_UTC_DFLT_05,
+ R600_UTC_DFLT_06,
+ R600_UTC_DFLT_07,
+ R600_UTC_DFLT_08,
+ R600_UTC_DFLT_09,
+ R600_UTC_DFLT_10,
+ R600_UTC_DFLT_11,
+ R600_UTC_DFLT_12,
+ R600_UTC_DFLT_13,
+ R600_UTC_DFLT_14,
+};
+
+static const u32 r600_dtc[R600_PM_NUMBER_OF_TC] =
+{
+ R600_DTC_DFLT_00,
+ R600_DTC_DFLT_01,
+ R600_DTC_DFLT_02,
+ R600_DTC_DFLT_03,
+ R600_DTC_DFLT_04,
+ R600_DTC_DFLT_05,
+ R600_DTC_DFLT_06,
+ R600_DTC_DFLT_07,
+ R600_DTC_DFLT_08,
+ R600_DTC_DFLT_09,
+ R600_DTC_DFLT_10,
+ R600_DTC_DFLT_11,
+ R600_DTC_DFLT_12,
+ R600_DTC_DFLT_13,
+ R600_DTC_DFLT_14,
+};
+
+static const struct si_cac_config_reg cac_weights_tahiti[] =
+{
+ { 0x0, 0x0000ffff, 0, 0xc, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x101, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0xc, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x8fc, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x95, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x34e, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x1a1, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0xda, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x46, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x208, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0xe7, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x948, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x167, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x31, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x18e, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg lcac_tahiti[] =
+{
+ { 0x143, 0x0001fffe, 1, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x146, 0x0001fffe, 1, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0x146, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x149, 0x0001fffe, 1, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0x149, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x14c, 0x0001fffe, 1, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0x14c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x9b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x9b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x9e, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x9e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x101, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x101, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x107, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x107, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10a, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x10a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x10d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8f, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x8f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x92, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x92, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x95, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x95, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x152, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x152, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x155, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x155, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x158, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x158, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x113, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x113, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x116, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x116, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x119, 0x0001fffe, 1, 0x8, SISLANDS_CACCONFIG_CGIND },
+ { 0x119, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x122, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x122, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x125, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x125, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x128, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x128, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x12b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x12b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+
+};
+
+static const struct si_cac_config_reg cac_override_tahiti[] =
+{
+ { 0xFFFFFFFF }
+};
+
+static const struct si_powertune_data powertune_data_tahiti =
+{
+ ((1 << 16) | 27027),
+ 6,
+ 0,
+ 4,
+ 95,
+ {
+ 0UL,
+ 0UL,
+ 4521550UL,
+ 309631529UL,
+ -1270850L,
+ 4513710L,
+ 40
+ },
+ 595000000UL,
+ 12,
+ {
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0
+ },
+ true
+};
+
+static const struct si_dte_data dte_data_tahiti =
+{
+ { 1159409, 0, 0, 0, 0 },
+ { 777, 0, 0, 0, 0 },
+ 2,
+ 54000,
+ 127000,
+ 25,
+ 2,
+ 10,
+ 13,
+ { 27, 31, 35, 39, 43, 47, 54, 61, 67, 74, 81, 88, 95, 0, 0, 0 },
+ { 240888759, 221057860, 235370597, 162287531, 158510299, 131423027, 116673180, 103067515, 87941937, 76209048, 68209175, 64090048, 58301890, 0, 0, 0 },
+ { 12024, 11189, 11451, 8411, 7939, 6666, 5681, 4905, 4241, 3720, 3354, 3122, 2890, 0, 0, 0 },
+ 85,
+ false
+};
+
+#if 0
+static const struct si_dte_data dte_data_tahiti_le =
+{
+ { 0x1E8480, 0x7A1200, 0x2160EC0, 0x3938700, 0 },
+ { 0x7D, 0x7D, 0x4E4, 0xB00, 0 },
+ 0x5,
+ 0xAFC8,
+ 0x64,
+ 0x32,
+ 1,
+ 0,
+ 0x10,
+ { 0x78, 0x7C, 0x82, 0x88, 0x8E, 0x94, 0x9A, 0xA0, 0xA6, 0xAC, 0xB0, 0xB4, 0xB8, 0xBC, 0xC0, 0xC4 },
+ { 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700, 0x3938700 },
+ { 0x2AF8, 0x2AF8, 0x29BB, 0x27F9, 0x2637, 0x2475, 0x22B3, 0x20F1, 0x1F2F, 0x1D6D, 0x1734, 0x1414, 0x10F4, 0xDD4, 0xAB4, 0x794 },
+ 85,
+ true
+};
+#endif
+
+static const struct si_dte_data dte_data_tahiti_pro =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 5,
+ 45000,
+ 100,
+ 0xA,
+ 1,
+ 0,
+ 0x10,
+ { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0x7D0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_dte_data dte_data_new_zealand =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0 },
+ { 0x29B, 0x3E9, 0x537, 0x7D2, 0 },
+ 0x5,
+ 0xAFC8,
+ 0x69,
+ 0x32,
+ 1,
+ 0,
+ 0x10,
+ { 0x82, 0xA0, 0xB4, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE, 0xFE },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0xDAC, 0x1388, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685, 0x685 },
+ 85,
+ true
+};
+
+static const struct si_dte_data dte_data_aruba_pro =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 5,
+ 45000,
+ 100,
+ 0xA,
+ 1,
+ 0,
+ 0x10,
+ { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0x1000, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_dte_data dte_data_malta =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 5,
+ 45000,
+ 100,
+ 0xA,
+ 1,
+ 0,
+ 0x10,
+ { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_cac_config_reg cac_weights_pitcairn[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x8a, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x24d, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x19, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0xc11, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0x7f3, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x403, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x367, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x4c9, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x45d, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0x36d, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x534, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x5da, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x880, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0x201, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x9f, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x1f, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x5de, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x7b, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x13, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0xf9, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x66, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x13, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x186, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg lcac_pitcairn[] =
+{
+ { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x9b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x9b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x107, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x107, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x113, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x113, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x152, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x152, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8f, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x8f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x146, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x146, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x9e, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x9e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10a, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x10a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x116, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x116, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x155, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x155, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x92, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x92, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x149, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x149, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x101, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x101, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x10d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x119, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x119, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x158, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x158, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x95, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x95, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x14c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x14c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x122, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x122, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x125, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x125, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x128, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x128, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x12b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x12b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_override_pitcairn[] =
+{
+ { 0xFFFFFFFF }
+};
+
+static const struct si_powertune_data powertune_data_pitcairn =
+{
+ ((1 << 16) | 27027),
+ 5,
+ 0,
+ 6,
+ 100,
+ {
+ 51600000UL,
+ 1800000UL,
+ 7194395UL,
+ 309631529UL,
+ -1270850L,
+ 4513710L,
+ 100
+ },
+ 117830498UL,
+ 12,
+ {
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0
+ },
+ true
+};
+
+static const struct si_dte_data dte_data_pitcairn =
+{
+ { 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0 },
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ 0,
+ false
+};
+
+static const struct si_dte_data dte_data_curacao_xt =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 5,
+ 45000,
+ 100,
+ 0xA,
+ 1,
+ 0,
+ 0x10,
+ { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0x1D17, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_dte_data dte_data_curacao_pro =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 5,
+ 45000,
+ 100,
+ 0xA,
+ 1,
+ 0,
+ 0x10,
+ { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0x1D17, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_dte_data dte_data_neptune_xt =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 5,
+ 45000,
+ 100,
+ 0xA,
+ 1,
+ 0,
+ 0x10,
+ { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0x3A2F, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_cac_config_reg cac_weights_chelsea_pro[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x2BD, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_weights_chelsea_xt[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x30A, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_weights_heathrow[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x362, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_weights_cape_verde_pro[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x315, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_weights_cape_verde[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x3BA, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg lcac_cape_verde[] =
+{
+ { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x9b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x9b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x107, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x107, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x113, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x113, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x152, 0x0001fffe, 1, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x152, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8f, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x146, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x146, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_override_cape_verde[] =
+{
+ { 0xFFFFFFFF }
+};
+
+static const struct si_powertune_data powertune_data_cape_verde =
+{
+ ((1 << 16) | 0x6993),
+ 5,
+ 0,
+ 7,
+ 105,
+ {
+ 0UL,
+ 0UL,
+ 7194395UL,
+ 309631529UL,
+ -1270850L,
+ 4513710L,
+ 100
+ },
+ 117830498UL,
+ 12,
+ {
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0
+ },
+ true
+};
+
+static const struct si_dte_data dte_data_cape_verde =
+{
+ { 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0 },
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ 0,
+ false
+};
+
+static const struct si_dte_data dte_data_venus_xtx =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x71C, 0xAAB, 0xE39, 0x11C7, 0x0 },
+ 5,
+ 55000,
+ 0x69,
+ 0xA,
+ 1,
+ 0,
+ 0x3,
+ { 0x96, 0xB4, 0xFF, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ { 0x895440, 0x3D0900, 0x989680, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ { 0xD6D8, 0x88B8, 0x1555, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_dte_data dte_data_venus_xt =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0xBDA, 0x11C7, 0x17B4, 0x1DA1, 0x0 },
+ 5,
+ 55000,
+ 0x69,
+ 0xA,
+ 1,
+ 0,
+ 0x3,
+ { 0x96, 0xB4, 0xFF, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ { 0x895440, 0x3D0900, 0x989680, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ { 0xAFC8, 0x88B8, 0x238E, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_dte_data dte_data_venus_pro =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x11C7, 0x1AAB, 0x238E, 0x2C72, 0x0 },
+ 5,
+ 55000,
+ 0x69,
+ 0xA,
+ 1,
+ 0,
+ 0x3,
+ { 0x96, 0xB4, 0xFF, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ { 0x895440, 0x3D0900, 0x989680, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ { 0x88B8, 0x88B8, 0x3555, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_cac_config_reg cac_weights_oland[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x82, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x153, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x52, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x4F, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x135, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0xAC, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x118, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0xBE, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x110, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x4CD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x37, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x27, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0xC3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x35, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0x28, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x26C, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3B2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x99D, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA3F, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0xA, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x34, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x3BA, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x30, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7A, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x100, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_weights_mars_pro[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x43, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0xAF, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x2A, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0x59, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x1A5, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0x1D6, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x2A3, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x8FD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x8A, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0xA3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0x71, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x36, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0xA6, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x81, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3D2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x27C, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA96, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0xB, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x36, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x10, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x10, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x32, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7E, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0x280, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0x3C, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0x203, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0xB4, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_weights_mars_xt[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x43, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0xAF, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x2A, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0x59, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x1A5, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0x1D6, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x2A3, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x8FD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x8A, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0xA3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0x71, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x36, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0xA6, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x81, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3D2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x27C, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA96, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0xB, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x36, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x10, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x10, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x60, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x32, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7E, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0x280, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0x3C, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0x203, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0xB4, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_weights_oland_pro[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x43, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0xAF, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x2A, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0x59, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x1A5, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0x1D6, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x2A3, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x8FD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x8A, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0xA3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0x71, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x36, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0xA6, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x81, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3D2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x27C, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA96, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0xB, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x36, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x10, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x10, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x90, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x32, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7E, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0x280, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0x3C, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0x203, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0xB4, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_weights_oland_xt[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x43, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0xAF, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x2A, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x29, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0xA0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0x59, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x1A5, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0x1D6, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0x2A3, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x8FD, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x76, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x8A, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0xA3, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0x71, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0x36, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0xA6, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x81, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0x3D2, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0x27C, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xA96, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0x5, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0xB, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x15, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x36, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x10, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x10, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x120, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x32, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x7E, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0x280, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0x7, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0x3C, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0x203, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0xB4, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg lcac_oland[] =
+{
+ { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x0001fffe, 1, 0x4, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg lcac_mars_pro[] =
+{
+ { 0x98, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x98, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x104, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x110, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x14f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x0001fffe, 1, 0x6, SISLANDS_CACCONFIG_CGIND },
+ { 0x8c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x143, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x11f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x164, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x167, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16a, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15e, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x161, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x15b, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x0001fffe, 1, 0x2, SISLANDS_CACCONFIG_CGIND },
+ { 0x16d, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x170, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x173, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x176, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x179, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17c, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x0001fffe, 1, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0x17f, 0x00000001, 0, 0x1, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_cac_config_reg cac_override_oland[] =
+{
+ { 0xFFFFFFFF }
+};
+
+static const struct si_powertune_data powertune_data_oland =
+{
+ ((1 << 16) | 0x6993),
+ 5,
+ 0,
+ 7,
+ 105,
+ {
+ 0UL,
+ 0UL,
+ 7194395UL,
+ 309631529UL,
+ -1270850L,
+ 4513710L,
+ 100
+ },
+ 117830498UL,
+ 12,
+ {
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0
+ },
+ true
+};
+
+static const struct si_powertune_data powertune_data_mars_pro =
+{
+ ((1 << 16) | 0x6993),
+ 5,
+ 0,
+ 7,
+ 105,
+ {
+ 0UL,
+ 0UL,
+ 7194395UL,
+ 309631529UL,
+ -1270850L,
+ 4513710L,
+ 100
+ },
+ 117830498UL,
+ 12,
+ {
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0
+ },
+ true
+};
+
+static const struct si_dte_data dte_data_oland =
+{
+ { 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0 },
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
+ 0,
+ false
+};
+
+static const struct si_dte_data dte_data_mars_pro =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 5,
+ 55000,
+ 105,
+ 0xA,
+ 1,
+ 0,
+ 0x10,
+ { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0xF627, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+static const struct si_dte_data dte_data_sun_xt =
+{
+ { 0x1E8480, 0x3D0900, 0x989680, 0x2625A00, 0x0 },
+ { 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 5,
+ 55000,
+ 105,
+ 0xA,
+ 1,
+ 0,
+ 0x10,
+ { 0x96, 0xB4, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF },
+ { 0x895440, 0x3D0900, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680, 0x989680 },
+ { 0xD555, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 },
+ 90,
+ true
+};
+
+
+static const struct si_cac_config_reg cac_weights_hainan[] =
+{
+ { 0x0, 0x0000ffff, 0, 0x2d9, SISLANDS_CACCONFIG_CGIND },
+ { 0x0, 0xffff0000, 16, 0x22b, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0x0000ffff, 0, 0x21c, SISLANDS_CACCONFIG_CGIND },
+ { 0x1, 0xffff0000, 16, 0x1dc, SISLANDS_CACCONFIG_CGIND },
+ { 0x2, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0x0000ffff, 0, 0x24e, SISLANDS_CACCONFIG_CGIND },
+ { 0x3, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0x0000ffff, 0, 0x35e, SISLANDS_CACCONFIG_CGIND },
+ { 0x5, 0xffff0000, 16, 0x1143, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0x0000ffff, 0, 0xe17, SISLANDS_CACCONFIG_CGIND },
+ { 0x6, 0xffff0000, 16, 0x441, SISLANDS_CACCONFIG_CGIND },
+ { 0x18f, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0x0000ffff, 0, 0x28b, SISLANDS_CACCONFIG_CGIND },
+ { 0x7, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x8, 0xffff0000, 16, 0xabe, SISLANDS_CACCONFIG_CGIND },
+ { 0x9, 0x0000ffff, 0, 0xf11, SISLANDS_CACCONFIG_CGIND },
+ { 0xa, 0x0000ffff, 0, 0x907, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0x0000ffff, 0, 0xb45, SISLANDS_CACCONFIG_CGIND },
+ { 0xb, 0xffff0000, 16, 0xd1e, SISLANDS_CACCONFIG_CGIND },
+ { 0xc, 0x0000ffff, 0, 0xa2c, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0x0000ffff, 0, 0x62, SISLANDS_CACCONFIG_CGIND },
+ { 0xd, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0xe, 0x0000ffff, 0, 0x1f3, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0x0000ffff, 0, 0x42, SISLANDS_CACCONFIG_CGIND },
+ { 0xf, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x10, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0x0000ffff, 0, 0x709, SISLANDS_CACCONFIG_CGIND },
+ { 0x11, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x12, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x13, 0xffff0000, 16, 0x3a, SISLANDS_CACCONFIG_CGIND },
+ { 0x14, 0x0000ffff, 0, 0x357, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0x0000ffff, 0, 0x9f, SISLANDS_CACCONFIG_CGIND },
+ { 0x15, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x4e, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0x0000ffff, 0, 0x314, SISLANDS_CACCONFIG_CGIND },
+ { 0x16, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x17, 0x0000ffff, 0, 0x6d, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x18, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0x0000ffff, 0, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x19, 0xffff0000, 16, 0x0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1a, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1b, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1c, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1d, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1e, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x1f, 0xffff0000, 16, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x20, 0x0000ffff, 0, 0, SISLANDS_CACCONFIG_CGIND },
+ { 0x6d, 0x0000ffff, 0, 0x1b9, SISLANDS_CACCONFIG_CGIND },
+ { 0xFFFFFFFF }
+};
+
+static const struct si_powertune_data powertune_data_hainan =
+{
+ ((1 << 16) | 0x6993),
+ 5,
+ 0,
+ 9,
+ 105,
+ {
+ 0UL,
+ 0UL,
+ 7194395UL,
+ 309631529UL,
+ -1270850L,
+ 4513710L,
+ 100
+ },
+ 117830498UL,
+ 12,
+ {
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0
+ },
+ true
+};
+
+static struct rv7xx_power_info *rv770_get_pi(struct amdgpu_device *adev);
+static struct evergreen_power_info *evergreen_get_pi(struct amdgpu_device *adev);
+static struct ni_power_info *ni_get_pi(struct amdgpu_device *adev);
+static struct si_ps *si_get_ps(struct amdgpu_ps *rps);
+
+static int si_populate_voltage_value(struct amdgpu_device *adev,
+ const struct atom_voltage_table *table,
+ u16 value, SISLANDS_SMC_VOLTAGE_VALUE *voltage);
+static int si_get_std_voltage_value(struct amdgpu_device *adev,
+ SISLANDS_SMC_VOLTAGE_VALUE *voltage,
+ u16 *std_voltage);
+static int si_write_smc_soft_register(struct amdgpu_device *adev,
+ u16 reg_offset, u32 value);
+static int si_convert_power_level_to_smc(struct amdgpu_device *adev,
+ struct rv7xx_pl *pl,
+ SISLANDS_SMC_HW_PERFORMANCE_LEVEL *level);
+static int si_calculate_sclk_params(struct amdgpu_device *adev,
+ u32 engine_clock,
+ SISLANDS_SMC_SCLK_VALUE *sclk);
+
+static void si_thermal_start_smc_fan_control(struct amdgpu_device *adev);
+static void si_fan_ctrl_set_default_mode(struct amdgpu_device *adev);
+static void si_dpm_set_dpm_funcs(struct amdgpu_device *adev);
+static void si_dpm_set_irq_funcs(struct amdgpu_device *adev);
+
+static struct si_power_info *si_get_pi(struct amdgpu_device *adev)
+{
+ struct si_power_info *pi = adev->pm.dpm.priv;
+ return pi;
+}
+
+static void si_calculate_leakage_for_v_and_t_formula(const struct ni_leakage_coeffients *coeff,
+ u16 v, s32 t, u32 ileakage, u32 *leakage)
+{
+ s64 kt, kv, leakage_w, i_leakage, vddc;
+ s64 temperature, t_slope, t_intercept, av, bv, t_ref;
+ s64 tmp;
+
+ i_leakage = div64_s64(drm_int2fixp(ileakage), 100);
+ vddc = div64_s64(drm_int2fixp(v), 1000);
+ temperature = div64_s64(drm_int2fixp(t), 1000);
+
+ t_slope = div64_s64(drm_int2fixp(coeff->t_slope), 100000000);
+ t_intercept = div64_s64(drm_int2fixp(coeff->t_intercept), 100000000);
+ av = div64_s64(drm_int2fixp(coeff->av), 100000000);
+ bv = div64_s64(drm_int2fixp(coeff->bv), 100000000);
+ t_ref = drm_int2fixp(coeff->t_ref);
+
+ tmp = drm_fixp_mul(t_slope, vddc) + t_intercept;
+ kt = drm_fixp_exp(drm_fixp_mul(tmp, temperature));
+ kt = drm_fixp_div(kt, drm_fixp_exp(drm_fixp_mul(tmp, t_ref)));
+ kv = drm_fixp_mul(av, drm_fixp_exp(drm_fixp_mul(bv, vddc)));
+
+ leakage_w = drm_fixp_mul(drm_fixp_mul(drm_fixp_mul(i_leakage, kt), kv), vddc);
+
+ *leakage = drm_fixp2int(leakage_w * 1000);
+}
+
+static void si_calculate_leakage_for_v_and_t(struct amdgpu_device *adev,
+ const struct ni_leakage_coeffients *coeff,
+ u16 v,
+ s32 t,
+ u32 i_leakage,
+ u32 *leakage)
+{
+ si_calculate_leakage_for_v_and_t_formula(coeff, v, t, i_leakage, leakage);
+}
+
+static void si_calculate_leakage_for_v_formula(const struct ni_leakage_coeffients *coeff,
+ const u32 fixed_kt, u16 v,
+ u32 ileakage, u32 *leakage)
+{
+ s64 kt, kv, leakage_w, i_leakage, vddc;
+
+ i_leakage = div64_s64(drm_int2fixp(ileakage), 100);
+ vddc = div64_s64(drm_int2fixp(v), 1000);
+
+ kt = div64_s64(drm_int2fixp(fixed_kt), 100000000);
+ kv = drm_fixp_mul(div64_s64(drm_int2fixp(coeff->av), 100000000),
+ drm_fixp_exp(drm_fixp_mul(div64_s64(drm_int2fixp(coeff->bv), 100000000), vddc)));
+
+ leakage_w = drm_fixp_mul(drm_fixp_mul(drm_fixp_mul(i_leakage, kt), kv), vddc);
+
+ *leakage = drm_fixp2int(leakage_w * 1000);
+}
+
+static void si_calculate_leakage_for_v(struct amdgpu_device *adev,
+ const struct ni_leakage_coeffients *coeff,
+ const u32 fixed_kt,
+ u16 v,
+ u32 i_leakage,
+ u32 *leakage)
+{
+ si_calculate_leakage_for_v_formula(coeff, fixed_kt, v, i_leakage, leakage);
+}
+
+
+static void si_update_dte_from_pl2(struct amdgpu_device *adev,
+ struct si_dte_data *dte_data)
+{
+ u32 p_limit1 = adev->pm.dpm.tdp_limit;
+ u32 p_limit2 = adev->pm.dpm.near_tdp_limit;
+ u32 k = dte_data->k;
+ u32 t_max = dte_data->max_t;
+ u32 t_split[5] = { 10, 15, 20, 25, 30 };
+ u32 t_0 = dte_data->t0;
+ u32 i;
+
+ if (p_limit2 != 0 && p_limit2 <= p_limit1) {
+ dte_data->tdep_count = 3;
+
+ for (i = 0; i < k; i++) {
+ dte_data->r[i] =
+ (t_split[i] * (t_max - t_0/(u32)1000) * (1 << 14)) /
+ (p_limit2 * (u32)100);
+ }
+
+ dte_data->tdep_r[1] = dte_data->r[4] * 2;
+
+ for (i = 2; i < SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE; i++) {
+ dte_data->tdep_r[i] = dte_data->r[4];
+ }
+ } else {
+ DRM_ERROR("Invalid PL2! DTE will not be updated.\n");
+ }
+}
+
+static struct rv7xx_power_info *rv770_get_pi(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = adev->pm.dpm.priv;
+
+ return pi;
+}
+
+static struct ni_power_info *ni_get_pi(struct amdgpu_device *adev)
+{
+ struct ni_power_info *pi = adev->pm.dpm.priv;
+
+ return pi;
+}
+
+static struct si_ps *si_get_ps(struct amdgpu_ps *aps)
+{
+ struct si_ps *ps = aps->ps_priv;
+
+ return ps;
+}
+
+static void si_initialize_powertune_defaults(struct amdgpu_device *adev)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ bool update_dte_from_pl2 = false;
+
+ if (adev->asic_type == CHIP_TAHITI) {
+ si_pi->cac_weights = cac_weights_tahiti;
+ si_pi->lcac_config = lcac_tahiti;
+ si_pi->cac_override = cac_override_tahiti;
+ si_pi->powertune_data = &powertune_data_tahiti;
+ si_pi->dte_data = dte_data_tahiti;
+
+ switch (adev->pdev->device) {
+ case 0x6798:
+ si_pi->dte_data.enable_dte_by_default = true;
+ break;
+ case 0x6799:
+ si_pi->dte_data = dte_data_new_zealand;
+ break;
+ case 0x6790:
+ case 0x6791:
+ case 0x6792:
+ case 0x679E:
+ si_pi->dte_data = dte_data_aruba_pro;
+ update_dte_from_pl2 = true;
+ break;
+ case 0x679B:
+ si_pi->dte_data = dte_data_malta;
+ update_dte_from_pl2 = true;
+ break;
+ case 0x679A:
+ si_pi->dte_data = dte_data_tahiti_pro;
+ update_dte_from_pl2 = true;
+ break;
+ default:
+ if (si_pi->dte_data.enable_dte_by_default == true)
+ DRM_ERROR("DTE is not enabled!\n");
+ break;
+ }
+ } else if (adev->asic_type == CHIP_PITCAIRN) {
+ si_pi->cac_weights = cac_weights_pitcairn;
+ si_pi->lcac_config = lcac_pitcairn;
+ si_pi->cac_override = cac_override_pitcairn;
+ si_pi->powertune_data = &powertune_data_pitcairn;
+
+ switch (adev->pdev->device) {
+ case 0x6810:
+ case 0x6818:
+ si_pi->dte_data = dte_data_curacao_xt;
+ update_dte_from_pl2 = true;
+ break;
+ case 0x6819:
+ case 0x6811:
+ si_pi->dte_data = dte_data_curacao_pro;
+ update_dte_from_pl2 = true;
+ break;
+ case 0x6800:
+ case 0x6806:
+ si_pi->dte_data = dte_data_neptune_xt;
+ update_dte_from_pl2 = true;
+ break;
+ default:
+ si_pi->dte_data = dte_data_pitcairn;
+ break;
+ }
+ } else if (adev->asic_type == CHIP_VERDE) {
+ si_pi->lcac_config = lcac_cape_verde;
+ si_pi->cac_override = cac_override_cape_verde;
+ si_pi->powertune_data = &powertune_data_cape_verde;
+
+ switch (adev->pdev->device) {
+ case 0x683B:
+ case 0x683F:
+ case 0x6829:
+ case 0x6835:
+ si_pi->cac_weights = cac_weights_cape_verde_pro;
+ si_pi->dte_data = dte_data_cape_verde;
+ break;
+ case 0x682C:
+ si_pi->cac_weights = cac_weights_cape_verde_pro;
+ si_pi->dte_data = dte_data_sun_xt;
+ break;
+ case 0x6825:
+ case 0x6827:
+ si_pi->cac_weights = cac_weights_heathrow;
+ si_pi->dte_data = dte_data_cape_verde;
+ break;
+ case 0x6824:
+ case 0x682D:
+ si_pi->cac_weights = cac_weights_chelsea_xt;
+ si_pi->dte_data = dte_data_cape_verde;
+ break;
+ case 0x682F:
+ si_pi->cac_weights = cac_weights_chelsea_pro;
+ si_pi->dte_data = dte_data_cape_verde;
+ break;
+ case 0x6820:
+ si_pi->cac_weights = cac_weights_heathrow;
+ si_pi->dte_data = dte_data_venus_xtx;
+ break;
+ case 0x6821:
+ si_pi->cac_weights = cac_weights_heathrow;
+ si_pi->dte_data = dte_data_venus_xt;
+ break;
+ case 0x6823:
+ case 0x682B:
+ case 0x6822:
+ case 0x682A:
+ si_pi->cac_weights = cac_weights_chelsea_pro;
+ si_pi->dte_data = dte_data_venus_pro;
+ break;
+ default:
+ si_pi->cac_weights = cac_weights_cape_verde;
+ si_pi->dte_data = dte_data_cape_verde;
+ break;
+ }
+ } else if (adev->asic_type == CHIP_OLAND) {
+ si_pi->lcac_config = lcac_mars_pro;
+ si_pi->cac_override = cac_override_oland;
+ si_pi->powertune_data = &powertune_data_mars_pro;
+ si_pi->dte_data = dte_data_mars_pro;
+
+ switch (adev->pdev->device) {
+ case 0x6601:
+ case 0x6621:
+ case 0x6603:
+ case 0x6605:
+ si_pi->cac_weights = cac_weights_mars_pro;
+ update_dte_from_pl2 = true;
+ break;
+ case 0x6600:
+ case 0x6606:
+ case 0x6620:
+ case 0x6604:
+ si_pi->cac_weights = cac_weights_mars_xt;
+ update_dte_from_pl2 = true;
+ break;
+ case 0x6611:
+ case 0x6613:
+ case 0x6608:
+ si_pi->cac_weights = cac_weights_oland_pro;
+ update_dte_from_pl2 = true;
+ break;
+ case 0x6610:
+ si_pi->cac_weights = cac_weights_oland_xt;
+ update_dte_from_pl2 = true;
+ break;
+ default:
+ si_pi->cac_weights = cac_weights_oland;
+ si_pi->lcac_config = lcac_oland;
+ si_pi->cac_override = cac_override_oland;
+ si_pi->powertune_data = &powertune_data_oland;
+ si_pi->dte_data = dte_data_oland;
+ break;
+ }
+ } else if (adev->asic_type == CHIP_HAINAN) {
+ si_pi->cac_weights = cac_weights_hainan;
+ si_pi->lcac_config = lcac_oland;
+ si_pi->cac_override = cac_override_oland;
+ si_pi->powertune_data = &powertune_data_hainan;
+ si_pi->dte_data = dte_data_sun_xt;
+ update_dte_from_pl2 = true;
+ } else {
+ DRM_ERROR("Unknown SI asic revision, failed to initialize PowerTune!\n");
+ return;
+ }
+
+ ni_pi->enable_power_containment = false;
+ ni_pi->enable_cac = false;
+ ni_pi->enable_sq_ramping = false;
+ si_pi->enable_dte = false;
+
+ if (si_pi->powertune_data->enable_powertune_by_default) {
+ ni_pi->enable_power_containment = true;
+ ni_pi->enable_cac = true;
+ if (si_pi->dte_data.enable_dte_by_default) {
+ si_pi->enable_dte = true;
+ if (update_dte_from_pl2)
+ si_update_dte_from_pl2(adev, &si_pi->dte_data);
+
+ }
+ ni_pi->enable_sq_ramping = true;
+ }
+
+ ni_pi->driver_calculate_cac_leakage = true;
+ ni_pi->cac_configuration_required = true;
+
+ if (ni_pi->cac_configuration_required) {
+ ni_pi->support_cac_long_term_average = true;
+ si_pi->dyn_powertune_data.l2_lta_window_size =
+ si_pi->powertune_data->l2_lta_window_size_default;
+ si_pi->dyn_powertune_data.lts_truncate =
+ si_pi->powertune_data->lts_truncate_default;
+ } else {
+ ni_pi->support_cac_long_term_average = false;
+ si_pi->dyn_powertune_data.l2_lta_window_size = 0;
+ si_pi->dyn_powertune_data.lts_truncate = 0;
+ }
+
+ si_pi->dyn_powertune_data.disable_uvd_powertune = false;
+}
+
+static u32 si_get_smc_power_scaling_factor(struct amdgpu_device *adev)
+{
+ return 1;
+}
+
+static u32 si_calculate_cac_wintime(struct amdgpu_device *adev)
+{
+ u32 xclk;
+ u32 wintime;
+ u32 cac_window;
+ u32 cac_window_size;
+
+ xclk = amdgpu_asic_get_xclk(adev);
+
+ if (xclk == 0)
+ return 0;
+
+ cac_window = RREG32(CG_CAC_CTRL) & CAC_WINDOW_MASK;
+ cac_window_size = ((cac_window & 0xFFFF0000) >> 16) * (cac_window & 0x0000FFFF);
+
+ wintime = (cac_window_size * 100) / xclk;
+
+ return wintime;
+}
+
+static u32 si_scale_power_for_smc(u32 power_in_watts, u32 scaling_factor)
+{
+ return power_in_watts;
+}
+
+static int si_calculate_adjusted_tdp_limits(struct amdgpu_device *adev,
+ bool adjust_polarity,
+ u32 tdp_adjustment,
+ u32 *tdp_limit,
+ u32 *near_tdp_limit)
+{
+ u32 adjustment_delta, max_tdp_limit;
+
+ if (tdp_adjustment > (u32)adev->pm.dpm.tdp_od_limit)
+ return -EINVAL;
+
+ max_tdp_limit = ((100 + 100) * adev->pm.dpm.tdp_limit) / 100;
+
+ if (adjust_polarity) {
+ *tdp_limit = ((100 + tdp_adjustment) * adev->pm.dpm.tdp_limit) / 100;
+ *near_tdp_limit = adev->pm.dpm.near_tdp_limit_adjusted + (*tdp_limit - adev->pm.dpm.tdp_limit);
+ } else {
+ *tdp_limit = ((100 - tdp_adjustment) * adev->pm.dpm.tdp_limit) / 100;
+ adjustment_delta = adev->pm.dpm.tdp_limit - *tdp_limit;
+ if (adjustment_delta < adev->pm.dpm.near_tdp_limit_adjusted)
+ *near_tdp_limit = adev->pm.dpm.near_tdp_limit_adjusted - adjustment_delta;
+ else
+ *near_tdp_limit = 0;
+ }
+
+ if ((*tdp_limit <= 0) || (*tdp_limit > max_tdp_limit))
+ return -EINVAL;
+ if ((*near_tdp_limit <= 0) || (*near_tdp_limit > *tdp_limit))
+ return -EINVAL;
+
+ return 0;
+}
+
+static int si_populate_smc_tdp_limits(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ if (ni_pi->enable_power_containment) {
+ SISLANDS_SMC_STATETABLE *smc_table = &si_pi->smc_statetable;
+ PP_SIslands_PAPMParameters *papm_parm;
+ struct amdgpu_ppm_table *ppm = adev->pm.dpm.dyn_state.ppm_table;
+ u32 scaling_factor = si_get_smc_power_scaling_factor(adev);
+ u32 tdp_limit;
+ u32 near_tdp_limit;
+ int ret;
+
+ if (scaling_factor == 0)
+ return -EINVAL;
+
+ memset(smc_table, 0, sizeof(SISLANDS_SMC_STATETABLE));
+
+ ret = si_calculate_adjusted_tdp_limits(adev,
+ false, /* ??? */
+ adev->pm.dpm.tdp_adjustment,
+ &tdp_limit,
+ &near_tdp_limit);
+ if (ret)
+ return ret;
+
+ smc_table->dpm2Params.TDPLimit =
+ cpu_to_be32(si_scale_power_for_smc(tdp_limit, scaling_factor) * 1000);
+ smc_table->dpm2Params.NearTDPLimit =
+ cpu_to_be32(si_scale_power_for_smc(near_tdp_limit, scaling_factor) * 1000);
+ smc_table->dpm2Params.SafePowerLimit =
+ cpu_to_be32(si_scale_power_for_smc((near_tdp_limit * SISLANDS_DPM2_TDP_SAFE_LIMIT_PERCENT) / 100, scaling_factor) * 1000);
+
+ ret = amdgpu_si_copy_bytes_to_smc(adev,
+ (si_pi->state_table_start + offsetof(SISLANDS_SMC_STATETABLE, dpm2Params) +
+ offsetof(PP_SIslands_DPM2Parameters, TDPLimit)),
+ (u8 *)(&(smc_table->dpm2Params.TDPLimit)),
+ sizeof(u32) * 3,
+ si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ if (si_pi->enable_ppm) {
+ papm_parm = &si_pi->papm_parm;
+ memset(papm_parm, 0, sizeof(PP_SIslands_PAPMParameters));
+ papm_parm->NearTDPLimitTherm = cpu_to_be32(ppm->dgpu_tdp);
+ papm_parm->dGPU_T_Limit = cpu_to_be32(ppm->tj_max);
+ papm_parm->dGPU_T_Warning = cpu_to_be32(95);
+ papm_parm->dGPU_T_Hysteresis = cpu_to_be32(5);
+ papm_parm->PlatformPowerLimit = 0xffffffff;
+ papm_parm->NearTDPLimitPAPM = 0xffffffff;
+
+ ret = amdgpu_si_copy_bytes_to_smc(adev, si_pi->papm_cfg_table_start,
+ (u8 *)papm_parm,
+ sizeof(PP_SIslands_PAPMParameters),
+ si_pi->sram_end);
+ if (ret)
+ return ret;
+ }
+ }
+ return 0;
+}
+
+static int si_populate_smc_tdp_limits_2(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ if (ni_pi->enable_power_containment) {
+ SISLANDS_SMC_STATETABLE *smc_table = &si_pi->smc_statetable;
+ u32 scaling_factor = si_get_smc_power_scaling_factor(adev);
+ int ret;
+
+ memset(smc_table, 0, sizeof(SISLANDS_SMC_STATETABLE));
+
+ smc_table->dpm2Params.NearTDPLimit =
+ cpu_to_be32(si_scale_power_for_smc(adev->pm.dpm.near_tdp_limit_adjusted, scaling_factor) * 1000);
+ smc_table->dpm2Params.SafePowerLimit =
+ cpu_to_be32(si_scale_power_for_smc((adev->pm.dpm.near_tdp_limit_adjusted * SISLANDS_DPM2_TDP_SAFE_LIMIT_PERCENT) / 100, scaling_factor) * 1000);
+
+ ret = amdgpu_si_copy_bytes_to_smc(adev,
+ (si_pi->state_table_start +
+ offsetof(SISLANDS_SMC_STATETABLE, dpm2Params) +
+ offsetof(PP_SIslands_DPM2Parameters, NearTDPLimit)),
+ (u8 *)(&(smc_table->dpm2Params.NearTDPLimit)),
+ sizeof(u32) * 2,
+ si_pi->sram_end);
+ if (ret)
+ return ret;
+ }
+
+ return 0;
+}
+
+static u16 si_calculate_power_efficiency_ratio(struct amdgpu_device *adev,
+ const u16 prev_std_vddc,
+ const u16 curr_std_vddc)
+{
+ u64 margin = (u64)SISLANDS_DPM2_PWREFFICIENCYRATIO_MARGIN;
+ u64 prev_vddc = (u64)prev_std_vddc;
+ u64 curr_vddc = (u64)curr_std_vddc;
+ u64 pwr_efficiency_ratio, n, d;
+
+ if ((prev_vddc == 0) || (curr_vddc == 0))
+ return 0;
+
+ n = div64_u64((u64)1024 * curr_vddc * curr_vddc * ((u64)1000 + margin), (u64)1000);
+ d = prev_vddc * prev_vddc;
+ pwr_efficiency_ratio = div64_u64(n, d);
+
+ if (pwr_efficiency_ratio > (u64)0xFFFF)
+ return 0;
+
+ return (u16)pwr_efficiency_ratio;
+}
+
+static bool si_should_disable_uvd_powertune(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ if (si_pi->dyn_powertune_data.disable_uvd_powertune &&
+ amdgpu_state->vclk && amdgpu_state->dclk)
+ return true;
+
+ return false;
+}
+
+struct evergreen_power_info *evergreen_get_pi(struct amdgpu_device *adev)
+{
+ struct evergreen_power_info *pi = adev->pm.dpm.priv;
+
+ return pi;
+}
+
+static int si_populate_power_containment_values(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state,
+ SISLANDS_SMC_SWSTATE *smc_state)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_ps *state = si_get_ps(amdgpu_state);
+ SISLANDS_SMC_VOLTAGE_VALUE vddc;
+ u32 prev_sclk;
+ u32 max_sclk;
+ u32 min_sclk;
+ u16 prev_std_vddc;
+ u16 curr_std_vddc;
+ int i;
+ u16 pwr_efficiency_ratio;
+ u8 max_ps_percent;
+ bool disable_uvd_power_tune;
+ int ret;
+
+ if (ni_pi->enable_power_containment == false)
+ return 0;
+
+ if (state->performance_level_count == 0)
+ return -EINVAL;
+
+ if (smc_state->levelCount != state->performance_level_count)
+ return -EINVAL;
+
+ disable_uvd_power_tune = si_should_disable_uvd_powertune(adev, amdgpu_state);
+
+ smc_state->levels[0].dpm2.MaxPS = 0;
+ smc_state->levels[0].dpm2.NearTDPDec = 0;
+ smc_state->levels[0].dpm2.AboveSafeInc = 0;
+ smc_state->levels[0].dpm2.BelowSafeInc = 0;
+ smc_state->levels[0].dpm2.PwrEfficiencyRatio = 0;
+
+ for (i = 1; i < state->performance_level_count; i++) {
+ prev_sclk = state->performance_levels[i-1].sclk;
+ max_sclk = state->performance_levels[i].sclk;
+ if (i == 1)
+ max_ps_percent = SISLANDS_DPM2_MAXPS_PERCENT_M;
+ else
+ max_ps_percent = SISLANDS_DPM2_MAXPS_PERCENT_H;
+
+ if (prev_sclk > max_sclk)
+ return -EINVAL;
+
+ if ((max_ps_percent == 0) ||
+ (prev_sclk == max_sclk) ||
+ disable_uvd_power_tune)
+ min_sclk = max_sclk;
+ else if (i == 1)
+ min_sclk = prev_sclk;
+ else
+ min_sclk = (prev_sclk * (u32)max_ps_percent) / 100;
+
+ if (min_sclk < state->performance_levels[0].sclk)
+ min_sclk = state->performance_levels[0].sclk;
+
+ if (min_sclk == 0)
+ return -EINVAL;
+
+ ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table,
+ state->performance_levels[i-1].vddc, &vddc);
+ if (ret)
+ return ret;
+
+ ret = si_get_std_voltage_value(adev, &vddc, &prev_std_vddc);
+ if (ret)
+ return ret;
+
+ ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table,
+ state->performance_levels[i].vddc, &vddc);
+ if (ret)
+ return ret;
+
+ ret = si_get_std_voltage_value(adev, &vddc, &curr_std_vddc);
+ if (ret)
+ return ret;
+
+ pwr_efficiency_ratio = si_calculate_power_efficiency_ratio(adev,
+ prev_std_vddc, curr_std_vddc);
+
+ smc_state->levels[i].dpm2.MaxPS = (u8)((SISLANDS_DPM2_MAX_PULSE_SKIP * (max_sclk - min_sclk)) / max_sclk);
+ smc_state->levels[i].dpm2.NearTDPDec = SISLANDS_DPM2_NEAR_TDP_DEC;
+ smc_state->levels[i].dpm2.AboveSafeInc = SISLANDS_DPM2_ABOVE_SAFE_INC;
+ smc_state->levels[i].dpm2.BelowSafeInc = SISLANDS_DPM2_BELOW_SAFE_INC;
+ smc_state->levels[i].dpm2.PwrEfficiencyRatio = cpu_to_be16(pwr_efficiency_ratio);
+ }
+
+ return 0;
+}
+
+static int si_populate_sq_ramping_values(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state,
+ SISLANDS_SMC_SWSTATE *smc_state)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_ps *state = si_get_ps(amdgpu_state);
+ u32 sq_power_throttle, sq_power_throttle2;
+ bool enable_sq_ramping = ni_pi->enable_sq_ramping;
+ int i;
+
+ if (state->performance_level_count == 0)
+ return -EINVAL;
+
+ if (smc_state->levelCount != state->performance_level_count)
+ return -EINVAL;
+
+ if (adev->pm.dpm.sq_ramping_threshold == 0)
+ return -EINVAL;
+
+ if (SISLANDS_DPM2_SQ_RAMP_MAX_POWER > (MAX_POWER_MASK >> MAX_POWER_SHIFT))
+ enable_sq_ramping = false;
+
+ if (SISLANDS_DPM2_SQ_RAMP_MIN_POWER > (MIN_POWER_MASK >> MIN_POWER_SHIFT))
+ enable_sq_ramping = false;
+
+ if (SISLANDS_DPM2_SQ_RAMP_MAX_POWER_DELTA > (MAX_POWER_DELTA_MASK >> MAX_POWER_DELTA_SHIFT))
+ enable_sq_ramping = false;
+
+ if (SISLANDS_DPM2_SQ_RAMP_STI_SIZE > (STI_SIZE_MASK >> STI_SIZE_SHIFT))
+ enable_sq_ramping = false;
+
+ if (SISLANDS_DPM2_SQ_RAMP_LTI_RATIO > (LTI_RATIO_MASK >> LTI_RATIO_SHIFT))
+ enable_sq_ramping = false;
+
+ for (i = 0; i < state->performance_level_count; i++) {
+ sq_power_throttle = 0;
+ sq_power_throttle2 = 0;
+
+ if ((state->performance_levels[i].sclk >= adev->pm.dpm.sq_ramping_threshold) &&
+ enable_sq_ramping) {
+ sq_power_throttle |= MAX_POWER(SISLANDS_DPM2_SQ_RAMP_MAX_POWER);
+ sq_power_throttle |= MIN_POWER(SISLANDS_DPM2_SQ_RAMP_MIN_POWER);
+ sq_power_throttle2 |= MAX_POWER_DELTA(SISLANDS_DPM2_SQ_RAMP_MAX_POWER_DELTA);
+ sq_power_throttle2 |= STI_SIZE(SISLANDS_DPM2_SQ_RAMP_STI_SIZE);
+ sq_power_throttle2 |= LTI_RATIO(SISLANDS_DPM2_SQ_RAMP_LTI_RATIO);
+ } else {
+ sq_power_throttle |= MAX_POWER_MASK | MIN_POWER_MASK;
+ sq_power_throttle2 |= MAX_POWER_DELTA_MASK | STI_SIZE_MASK | LTI_RATIO_MASK;
+ }
+
+ smc_state->levels[i].SQPowerThrottle = cpu_to_be32(sq_power_throttle);
+ smc_state->levels[i].SQPowerThrottle_2 = cpu_to_be32(sq_power_throttle2);
+ }
+
+ return 0;
+}
+
+static int si_enable_power_containment(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state,
+ bool enable)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ PPSMC_Result smc_result;
+ int ret = 0;
+
+ if (ni_pi->enable_power_containment) {
+ if (enable) {
+ if (!si_should_disable_uvd_powertune(adev, amdgpu_new_state)) {
+ smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_TDPClampingActive);
+ if (smc_result != PPSMC_Result_OK) {
+ ret = -EINVAL;
+ ni_pi->pc_enabled = false;
+ } else {
+ ni_pi->pc_enabled = true;
+ }
+ }
+ } else {
+ smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_TDPClampingInactive);
+ if (smc_result != PPSMC_Result_OK)
+ ret = -EINVAL;
+ ni_pi->pc_enabled = false;
+ }
+ }
+
+ return ret;
+}
+
+static int si_initialize_smc_dte_tables(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ int ret = 0;
+ struct si_dte_data *dte_data = &si_pi->dte_data;
+ Smc_SIslands_DTE_Configuration *dte_tables = NULL;
+ u32 table_size;
+ u8 tdep_count;
+ u32 i;
+
+ if (dte_data == NULL)
+ si_pi->enable_dte = false;
+
+ if (si_pi->enable_dte == false)
+ return 0;
+
+ if (dte_data->k <= 0)
+ return -EINVAL;
+
+ dte_tables = kzalloc(sizeof(Smc_SIslands_DTE_Configuration), GFP_KERNEL);
+ if (dte_tables == NULL) {
+ si_pi->enable_dte = false;
+ return -ENOMEM;
+ }
+
+ table_size = dte_data->k;
+
+ if (table_size > SMC_SISLANDS_DTE_MAX_FILTER_STAGES)
+ table_size = SMC_SISLANDS_DTE_MAX_FILTER_STAGES;
+
+ tdep_count = dte_data->tdep_count;
+ if (tdep_count > SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE)
+ tdep_count = SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE;
+
+ dte_tables->K = cpu_to_be32(table_size);
+ dte_tables->T0 = cpu_to_be32(dte_data->t0);
+ dte_tables->MaxT = cpu_to_be32(dte_data->max_t);
+ dte_tables->WindowSize = dte_data->window_size;
+ dte_tables->temp_select = dte_data->temp_select;
+ dte_tables->DTE_mode = dte_data->dte_mode;
+ dte_tables->Tthreshold = cpu_to_be32(dte_data->t_threshold);
+
+ if (tdep_count > 0)
+ table_size--;
+
+ for (i = 0; i < table_size; i++) {
+ dte_tables->tau[i] = cpu_to_be32(dte_data->tau[i]);
+ dte_tables->R[i] = cpu_to_be32(dte_data->r[i]);
+ }
+
+ dte_tables->Tdep_count = tdep_count;
+
+ for (i = 0; i < (u32)tdep_count; i++) {
+ dte_tables->T_limits[i] = dte_data->t_limits[i];
+ dte_tables->Tdep_tau[i] = cpu_to_be32(dte_data->tdep_tau[i]);
+ dte_tables->Tdep_R[i] = cpu_to_be32(dte_data->tdep_r[i]);
+ }
+
+ ret = amdgpu_si_copy_bytes_to_smc(adev, si_pi->dte_table_start,
+ (u8 *)dte_tables,
+ sizeof(Smc_SIslands_DTE_Configuration),
+ si_pi->sram_end);
+ kfree(dte_tables);
+
+ return ret;
+}
+
+static int si_get_cac_std_voltage_max_min(struct amdgpu_device *adev,
+ u16 *max, u16 *min)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct amdgpu_cac_leakage_table *table =
+ &adev->pm.dpm.dyn_state.cac_leakage_table;
+ u32 i;
+ u32 v0_loadline;
+
+ if (table == NULL)
+ return -EINVAL;
+
+ *max = 0;
+ *min = 0xFFFF;
+
+ for (i = 0; i < table->count; i++) {
+ if (table->entries[i].vddc > *max)
+ *max = table->entries[i].vddc;
+ if (table->entries[i].vddc < *min)
+ *min = table->entries[i].vddc;
+ }
+
+ if (si_pi->powertune_data->lkge_lut_v0_percent > 100)
+ return -EINVAL;
+
+ v0_loadline = (*min) * (100 - si_pi->powertune_data->lkge_lut_v0_percent) / 100;
+
+ if (v0_loadline > 0xFFFFUL)
+ return -EINVAL;
+
+ *min = (u16)v0_loadline;
+
+ if ((*min > *max) || (*max == 0) || (*min == 0))
+ return -EINVAL;
+
+ return 0;
+}
+
+static u16 si_get_cac_std_voltage_step(u16 max, u16 min)
+{
+ return ((max - min) + (SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES - 1)) /
+ SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES;
+}
+
+static int si_init_dte_leakage_table(struct amdgpu_device *adev,
+ PP_SIslands_CacConfig *cac_tables,
+ u16 vddc_max, u16 vddc_min, u16 vddc_step,
+ u16 t0, u16 t_step)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 leakage;
+ unsigned int i, j;
+ s32 t;
+ u32 smc_leakage;
+ u32 scaling_factor;
+ u16 voltage;
+
+ scaling_factor = si_get_smc_power_scaling_factor(adev);
+
+ for (i = 0; i < SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES ; i++) {
+ t = (1000 * (i * t_step + t0));
+
+ for (j = 0; j < SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES; j++) {
+ voltage = vddc_max - (vddc_step * j);
+
+ si_calculate_leakage_for_v_and_t(adev,
+ &si_pi->powertune_data->leakage_coefficients,
+ voltage,
+ t,
+ si_pi->dyn_powertune_data.cac_leakage,
+ &leakage);
+
+ smc_leakage = si_scale_power_for_smc(leakage, scaling_factor) / 4;
+
+ if (smc_leakage > 0xFFFF)
+ smc_leakage = 0xFFFF;
+
+ cac_tables->cac_lkge_lut[i][SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES-1-j] =
+ cpu_to_be16((u16)smc_leakage);
+ }
+ }
+ return 0;
+}
+
+static int si_init_simplified_leakage_table(struct amdgpu_device *adev,
+ PP_SIslands_CacConfig *cac_tables,
+ u16 vddc_max, u16 vddc_min, u16 vddc_step)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 leakage;
+ unsigned int i, j;
+ u32 smc_leakage;
+ u32 scaling_factor;
+ u16 voltage;
+
+ scaling_factor = si_get_smc_power_scaling_factor(adev);
+
+ for (j = 0; j < SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES; j++) {
+ voltage = vddc_max - (vddc_step * j);
+
+ si_calculate_leakage_for_v(adev,
+ &si_pi->powertune_data->leakage_coefficients,
+ si_pi->powertune_data->fixed_kt,
+ voltage,
+ si_pi->dyn_powertune_data.cac_leakage,
+ &leakage);
+
+ smc_leakage = si_scale_power_for_smc(leakage, scaling_factor) / 4;
+
+ if (smc_leakage > 0xFFFF)
+ smc_leakage = 0xFFFF;
+
+ for (i = 0; i < SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES ; i++)
+ cac_tables->cac_lkge_lut[i][SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES-1-j] =
+ cpu_to_be16((u16)smc_leakage);
+ }
+ return 0;
+}
+
+static int si_initialize_smc_cac_tables(struct amdgpu_device *adev)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ PP_SIslands_CacConfig *cac_tables = NULL;
+ u16 vddc_max, vddc_min, vddc_step;
+ u16 t0, t_step;
+ u32 load_line_slope, reg;
+ int ret = 0;
+ u32 ticks_per_us = amdgpu_asic_get_xclk(adev) / 100;
+
+ if (ni_pi->enable_cac == false)
+ return 0;
+
+ cac_tables = kzalloc(sizeof(PP_SIslands_CacConfig), GFP_KERNEL);
+ if (!cac_tables)
+ return -ENOMEM;
+
+ reg = RREG32(CG_CAC_CTRL) & ~CAC_WINDOW_MASK;
+ reg |= CAC_WINDOW(si_pi->powertune_data->cac_window);
+ WREG32(CG_CAC_CTRL, reg);
+
+ si_pi->dyn_powertune_data.cac_leakage = adev->pm.dpm.cac_leakage;
+ si_pi->dyn_powertune_data.dc_pwr_value =
+ si_pi->powertune_data->dc_cac[NISLANDS_DCCAC_LEVEL_0];
+ si_pi->dyn_powertune_data.wintime = si_calculate_cac_wintime(adev);
+ si_pi->dyn_powertune_data.shift_n = si_pi->powertune_data->shift_n_default;
+
+ si_pi->dyn_powertune_data.leakage_minimum_temperature = 80 * 1000;
+
+ ret = si_get_cac_std_voltage_max_min(adev, &vddc_max, &vddc_min);
+ if (ret)
+ goto done_free;
+
+ vddc_step = si_get_cac_std_voltage_step(vddc_max, vddc_min);
+ vddc_min = vddc_max - (vddc_step * (SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES - 1));
+ t_step = 4;
+ t0 = 60;
+
+ if (si_pi->enable_dte || ni_pi->driver_calculate_cac_leakage)
+ ret = si_init_dte_leakage_table(adev, cac_tables,
+ vddc_max, vddc_min, vddc_step,
+ t0, t_step);
+ else
+ ret = si_init_simplified_leakage_table(adev, cac_tables,
+ vddc_max, vddc_min, vddc_step);
+ if (ret)
+ goto done_free;
+
+ load_line_slope = ((u32)adev->pm.dpm.load_line_slope << SMC_SISLANDS_SCALE_R) / 100;
+
+ cac_tables->l2numWin_TDP = cpu_to_be32(si_pi->dyn_powertune_data.l2_lta_window_size);
+ cac_tables->lts_truncate_n = si_pi->dyn_powertune_data.lts_truncate;
+ cac_tables->SHIFT_N = si_pi->dyn_powertune_data.shift_n;
+ cac_tables->lkge_lut_V0 = cpu_to_be32((u32)vddc_min);
+ cac_tables->lkge_lut_Vstep = cpu_to_be32((u32)vddc_step);
+ cac_tables->R_LL = cpu_to_be32(load_line_slope);
+ cac_tables->WinTime = cpu_to_be32(si_pi->dyn_powertune_data.wintime);
+ cac_tables->calculation_repeats = cpu_to_be32(2);
+ cac_tables->dc_cac = cpu_to_be32(0);
+ cac_tables->log2_PG_LKG_SCALE = 12;
+ cac_tables->cac_temp = si_pi->powertune_data->operating_temp;
+ cac_tables->lkge_lut_T0 = cpu_to_be32((u32)t0);
+ cac_tables->lkge_lut_Tstep = cpu_to_be32((u32)t_step);
+
+ ret = amdgpu_si_copy_bytes_to_smc(adev, si_pi->cac_table_start,
+ (u8 *)cac_tables,
+ sizeof(PP_SIslands_CacConfig),
+ si_pi->sram_end);
+
+ if (ret)
+ goto done_free;
+
+ ret = si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_ticks_per_us, ticks_per_us);
+
+done_free:
+ if (ret) {
+ ni_pi->enable_cac = false;
+ ni_pi->enable_power_containment = false;
+ }
+
+ kfree(cac_tables);
+
+ return ret;
+}
+
+static int si_program_cac_config_registers(struct amdgpu_device *adev,
+ const struct si_cac_config_reg *cac_config_regs)
+{
+ const struct si_cac_config_reg *config_regs = cac_config_regs;
+ u32 data = 0, offset;
+
+ if (!config_regs)
+ return -EINVAL;
+
+ while (config_regs->offset != 0xFFFFFFFF) {
+ switch (config_regs->type) {
+ case SISLANDS_CACCONFIG_CGIND:
+ offset = SMC_CG_IND_START + config_regs->offset;
+ if (offset < SMC_CG_IND_END)
+ data = RREG32_SMC(offset);
+ break;
+ default:
+ data = RREG32(config_regs->offset);
+ break;
+ }
+
+ data &= ~config_regs->mask;
+ data |= ((config_regs->value << config_regs->shift) & config_regs->mask);
+
+ switch (config_regs->type) {
+ case SISLANDS_CACCONFIG_CGIND:
+ offset = SMC_CG_IND_START + config_regs->offset;
+ if (offset < SMC_CG_IND_END)
+ WREG32_SMC(offset, data);
+ break;
+ default:
+ WREG32(config_regs->offset, data);
+ break;
+ }
+ config_regs++;
+ }
+ return 0;
+}
+
+static int si_initialize_hardware_cac_manager(struct amdgpu_device *adev)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ int ret;
+
+ if ((ni_pi->enable_cac == false) ||
+ (ni_pi->cac_configuration_required == false))
+ return 0;
+
+ ret = si_program_cac_config_registers(adev, si_pi->lcac_config);
+ if (ret)
+ return ret;
+ ret = si_program_cac_config_registers(adev, si_pi->cac_override);
+ if (ret)
+ return ret;
+ ret = si_program_cac_config_registers(adev, si_pi->cac_weights);
+ if (ret)
+ return ret;
+
+ return 0;
+}
+
+static int si_enable_smc_cac(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state,
+ bool enable)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ PPSMC_Result smc_result;
+ int ret = 0;
+
+ if (ni_pi->enable_cac) {
+ if (enable) {
+ if (!si_should_disable_uvd_powertune(adev, amdgpu_new_state)) {
+ if (ni_pi->support_cac_long_term_average) {
+ smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_CACLongTermAvgEnable);
+ if (smc_result != PPSMC_Result_OK)
+ ni_pi->support_cac_long_term_average = false;
+ }
+
+ smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableCac);
+ if (smc_result != PPSMC_Result_OK) {
+ ret = -EINVAL;
+ ni_pi->cac_enabled = false;
+ } else {
+ ni_pi->cac_enabled = true;
+ }
+
+ if (si_pi->enable_dte) {
+ smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableDTE);
+ if (smc_result != PPSMC_Result_OK)
+ ret = -EINVAL;
+ }
+ }
+ } else if (ni_pi->cac_enabled) {
+ if (si_pi->enable_dte)
+ smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_DisableDTE);
+
+ smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_DisableCac);
+
+ ni_pi->cac_enabled = false;
+
+ if (ni_pi->support_cac_long_term_average)
+ smc_result = amdgpu_si_send_msg_to_smc(adev, PPSMC_CACLongTermAvgDisable);
+ }
+ }
+ return ret;
+}
+
+static int si_init_smc_spll_table(struct amdgpu_device *adev)
+{
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ SMC_SISLANDS_SPLL_DIV_TABLE *spll_table;
+ SISLANDS_SMC_SCLK_VALUE sclk_params;
+ u32 fb_div, p_div;
+ u32 clk_s, clk_v;
+ u32 sclk = 0;
+ int ret = 0;
+ u32 tmp;
+ int i;
+
+ if (si_pi->spll_table_start == 0)
+ return -EINVAL;
+
+ spll_table = kzalloc(sizeof(SMC_SISLANDS_SPLL_DIV_TABLE), GFP_KERNEL);
+ if (spll_table == NULL)
+ return -ENOMEM;
+
+ for (i = 0; i < 256; i++) {
+ ret = si_calculate_sclk_params(adev, sclk, &sclk_params);
+ if (ret)
+ break;
+ p_div = (sclk_params.vCG_SPLL_FUNC_CNTL & SPLL_PDIV_A_MASK) >> SPLL_PDIV_A_SHIFT;
+ fb_div = (sclk_params.vCG_SPLL_FUNC_CNTL_3 & SPLL_FB_DIV_MASK) >> SPLL_FB_DIV_SHIFT;
+ clk_s = (sclk_params.vCG_SPLL_SPREAD_SPECTRUM & CLK_S_MASK) >> CLK_S_SHIFT;
+ clk_v = (sclk_params.vCG_SPLL_SPREAD_SPECTRUM_2 & CLK_V_MASK) >> CLK_V_SHIFT;
+
+ fb_div &= ~0x00001FFF;
+ fb_div >>= 1;
+ clk_v >>= 6;
+
+ if (p_div & ~(SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_MASK >> SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_SHIFT))
+ ret = -EINVAL;
+ if (fb_div & ~(SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_MASK >> SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_SHIFT))
+ ret = -EINVAL;
+ if (clk_s & ~(SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_MASK >> SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_SHIFT))
+ ret = -EINVAL;
+ if (clk_v & ~(SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_MASK >> SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_SHIFT))
+ ret = -EINVAL;
+
+ if (ret)
+ break;
+
+ tmp = ((fb_div << SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_SHIFT) & SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_MASK) |
+ ((p_div << SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_SHIFT) & SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_MASK);
+ spll_table->freq[i] = cpu_to_be32(tmp);
+
+ tmp = ((clk_v << SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_SHIFT) & SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_MASK) |
+ ((clk_s << SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_SHIFT) & SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_MASK);
+ spll_table->ss[i] = cpu_to_be32(tmp);
+
+ sclk += 512;
+ }
+
+
+ if (!ret)
+ ret = amdgpu_si_copy_bytes_to_smc(adev, si_pi->spll_table_start,
+ (u8 *)spll_table,
+ sizeof(SMC_SISLANDS_SPLL_DIV_TABLE),
+ si_pi->sram_end);
+
+ if (ret)
+ ni_pi->enable_power_containment = false;
+
+ kfree(spll_table);
+
+ return ret;
+}
+
+struct si_dpm_quirk {
+ u32 chip_vendor;
+ u32 chip_device;
+ u32 subsys_vendor;
+ u32 subsys_device;
+ u32 max_sclk;
+ u32 max_mclk;
+};
+
+/* cards with dpm stability problems */
+static struct si_dpm_quirk si_dpm_quirk_list[] = {
+ /* PITCAIRN - https://bugs.freedesktop.org/show_bug.cgi?id=76490 */
+ { PCI_VENDOR_ID_ATI, 0x6810, 0x1462, 0x3036, 0, 120000 },
+ { PCI_VENDOR_ID_ATI, 0x6811, 0x174b, 0xe271, 0, 120000 },
+ { PCI_VENDOR_ID_ATI, 0x6810, 0x174b, 0xe271, 85000, 90000 },
+ { PCI_VENDOR_ID_ATI, 0x6811, 0x1462, 0x2015, 0, 120000 },
+ { PCI_VENDOR_ID_ATI, 0x6811, 0x1043, 0x2015, 0, 120000 },
+ { 0, 0, 0, 0 },
+};
+
+static u16 si_get_lower_of_leakage_and_vce_voltage(struct amdgpu_device *adev,
+ u16 vce_voltage)
+{
+ u16 highest_leakage = 0;
+ struct si_power_info *si_pi = si_get_pi(adev);
+ int i;
+
+ for (i = 0; i < si_pi->leakage_voltage.count; i++){
+ if (highest_leakage < si_pi->leakage_voltage.entries[i].voltage)
+ highest_leakage = si_pi->leakage_voltage.entries[i].voltage;
+ }
+
+ if (si_pi->leakage_voltage.count && (highest_leakage < vce_voltage))
+ return highest_leakage;
+
+ return vce_voltage;
+}
+
+static int si_get_vce_clock_voltage(struct amdgpu_device *adev,
+ u32 evclk, u32 ecclk, u16 *voltage)
+{
+ u32 i;
+ int ret = -EINVAL;
+ struct amdgpu_vce_clock_voltage_dependency_table *table =
+ &adev->pm.dpm.dyn_state.vce_clock_voltage_dependency_table;
+
+ if (((evclk == 0) && (ecclk == 0)) ||
+ (table && (table->count == 0))) {
+ *voltage = 0;
+ return 0;
+ }
+
+ for (i = 0; i < table->count; i++) {
+ if ((evclk <= table->entries[i].evclk) &&
+ (ecclk <= table->entries[i].ecclk)) {
+ *voltage = table->entries[i].v;
+ ret = 0;
+ break;
+ }
+ }
+
+ /* if no match return the highest voltage */
+ if (ret)
+ *voltage = table->entries[table->count - 1].v;
+
+ *voltage = si_get_lower_of_leakage_and_vce_voltage(adev, *voltage);
+
+ return ret;
+}
+
+static bool si_dpm_vblank_too_short(struct amdgpu_device *adev)
+{
+
+ u32 vblank_time = amdgpu_dpm_get_vblank_time(adev);
+ /* we never hit the non-gddr5 limit so disable it */
+ u32 switch_limit = adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5 ? 450 : 0;
+
+ if (vblank_time < switch_limit)
+ return true;
+ else
+ return false;
+
+}
+
+static int ni_copy_and_switch_arb_sets(struct amdgpu_device *adev,
+ u32 arb_freq_src, u32 arb_freq_dest)
+{
+ u32 mc_arb_dram_timing;
+ u32 mc_arb_dram_timing2;
+ u32 burst_time;
+ u32 mc_cg_config;
+
+ switch (arb_freq_src) {
+ case MC_CG_ARB_FREQ_F0:
+ mc_arb_dram_timing = RREG32(MC_ARB_DRAM_TIMING);
+ mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2);
+ burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE0_MASK) >> STATE0_SHIFT;
+ break;
+ case MC_CG_ARB_FREQ_F1:
+ mc_arb_dram_timing = RREG32(MC_ARB_DRAM_TIMING_1);
+ mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_1);
+ burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE1_MASK) >> STATE1_SHIFT;
+ break;
+ case MC_CG_ARB_FREQ_F2:
+ mc_arb_dram_timing = RREG32(MC_ARB_DRAM_TIMING_2);
+ mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_2);
+ burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE2_MASK) >> STATE2_SHIFT;
+ break;
+ case MC_CG_ARB_FREQ_F3:
+ mc_arb_dram_timing = RREG32(MC_ARB_DRAM_TIMING_3);
+ mc_arb_dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2_3);
+ burst_time = (RREG32(MC_ARB_BURST_TIME) & STATE3_MASK) >> STATE3_SHIFT;
+ break;
+ default:
+ return -EINVAL;
+ }
+
+ switch (arb_freq_dest) {
+ case MC_CG_ARB_FREQ_F0:
+ WREG32(MC_ARB_DRAM_TIMING, mc_arb_dram_timing);
+ WREG32(MC_ARB_DRAM_TIMING2, mc_arb_dram_timing2);
+ WREG32_P(MC_ARB_BURST_TIME, STATE0(burst_time), ~STATE0_MASK);
+ break;
+ case MC_CG_ARB_FREQ_F1:
+ WREG32(MC_ARB_DRAM_TIMING_1, mc_arb_dram_timing);
+ WREG32(MC_ARB_DRAM_TIMING2_1, mc_arb_dram_timing2);
+ WREG32_P(MC_ARB_BURST_TIME, STATE1(burst_time), ~STATE1_MASK);
+ break;
+ case MC_CG_ARB_FREQ_F2:
+ WREG32(MC_ARB_DRAM_TIMING_2, mc_arb_dram_timing);
+ WREG32(MC_ARB_DRAM_TIMING2_2, mc_arb_dram_timing2);
+ WREG32_P(MC_ARB_BURST_TIME, STATE2(burst_time), ~STATE2_MASK);
+ break;
+ case MC_CG_ARB_FREQ_F3:
+ WREG32(MC_ARB_DRAM_TIMING_3, mc_arb_dram_timing);
+ WREG32(MC_ARB_DRAM_TIMING2_3, mc_arb_dram_timing2);
+ WREG32_P(MC_ARB_BURST_TIME, STATE3(burst_time), ~STATE3_MASK);
+ break;
+ default:
+ return -EINVAL;
+ }
+
+ mc_cg_config = RREG32(MC_CG_CONFIG) | 0x0000000F;
+ WREG32(MC_CG_CONFIG, mc_cg_config);
+ WREG32_P(MC_ARB_CG, CG_ARB_REQ(arb_freq_dest), ~CG_ARB_REQ_MASK);
+
+ return 0;
+}
+
+static void ni_update_current_ps(struct amdgpu_device *adev,
+ struct amdgpu_ps *rps)
+{
+ struct si_ps *new_ps = si_get_ps(rps);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+
+ eg_pi->current_rps = *rps;
+ ni_pi->current_ps = *new_ps;
+ eg_pi->current_rps.ps_priv = &ni_pi->current_ps;
+}
+
+static void ni_update_requested_ps(struct amdgpu_device *adev,
+ struct amdgpu_ps *rps)
+{
+ struct si_ps *new_ps = si_get_ps(rps);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+
+ eg_pi->requested_rps = *rps;
+ ni_pi->requested_ps = *new_ps;
+ eg_pi->requested_rps.ps_priv = &ni_pi->requested_ps;
+}
+
+static void ni_set_uvd_clock_before_set_eng_clock(struct amdgpu_device *adev,
+ struct amdgpu_ps *new_ps,
+ struct amdgpu_ps *old_ps)
+{
+ struct si_ps *new_state = si_get_ps(new_ps);
+ struct si_ps *current_state = si_get_ps(old_ps);
+
+ if ((new_ps->vclk == old_ps->vclk) &&
+ (new_ps->dclk == old_ps->dclk))
+ return;
+
+ if (new_state->performance_levels[new_state->performance_level_count - 1].sclk >=
+ current_state->performance_levels[current_state->performance_level_count - 1].sclk)
+ return;
+
+ amdgpu_asic_set_uvd_clocks(adev, new_ps->vclk, new_ps->dclk);
+}
+
+static void ni_set_uvd_clock_after_set_eng_clock(struct amdgpu_device *adev,
+ struct amdgpu_ps *new_ps,
+ struct amdgpu_ps *old_ps)
+{
+ struct si_ps *new_state = si_get_ps(new_ps);
+ struct si_ps *current_state = si_get_ps(old_ps);
+
+ if ((new_ps->vclk == old_ps->vclk) &&
+ (new_ps->dclk == old_ps->dclk))
+ return;
+
+ if (new_state->performance_levels[new_state->performance_level_count - 1].sclk <
+ current_state->performance_levels[current_state->performance_level_count - 1].sclk)
+ return;
+
+ amdgpu_asic_set_uvd_clocks(adev, new_ps->vclk, new_ps->dclk);
+}
+
+static u16 btc_find_voltage(struct atom_voltage_table *table, u16 voltage)
+{
+ unsigned int i;
+
+ for (i = 0; i < table->count; i++)
+ if (voltage <= table->entries[i].value)
+ return table->entries[i].value;
+
+ return table->entries[table->count - 1].value;
+}
+
+static u32 btc_find_valid_clock(struct amdgpu_clock_array *clocks,
+ u32 max_clock, u32 requested_clock)
+{
+ unsigned int i;
+
+ if ((clocks == NULL) || (clocks->count == 0))
+ return (requested_clock < max_clock) ? requested_clock : max_clock;
+
+ for (i = 0; i < clocks->count; i++) {
+ if (clocks->values[i] >= requested_clock)
+ return (clocks->values[i] < max_clock) ? clocks->values[i] : max_clock;
+ }
+
+ return (clocks->values[clocks->count - 1] < max_clock) ?
+ clocks->values[clocks->count - 1] : max_clock;
+}
+
+static u32 btc_get_valid_mclk(struct amdgpu_device *adev,
+ u32 max_mclk, u32 requested_mclk)
+{
+ return btc_find_valid_clock(&adev->pm.dpm.dyn_state.valid_mclk_values,
+ max_mclk, requested_mclk);
+}
+
+static u32 btc_get_valid_sclk(struct amdgpu_device *adev,
+ u32 max_sclk, u32 requested_sclk)
+{
+ return btc_find_valid_clock(&adev->pm.dpm.dyn_state.valid_sclk_values,
+ max_sclk, requested_sclk);
+}
+
+static void btc_get_max_clock_from_voltage_dependency_table(struct amdgpu_clock_voltage_dependency_table *table,
+ u32 *max_clock)
+{
+ u32 i, clock = 0;
+
+ if ((table == NULL) || (table->count == 0)) {
+ *max_clock = clock;
+ return;
+ }
+
+ for (i = 0; i < table->count; i++) {
+ if (clock < table->entries[i].clk)
+ clock = table->entries[i].clk;
+ }
+ *max_clock = clock;
+}
+
+static void btc_apply_voltage_dependency_rules(struct amdgpu_clock_voltage_dependency_table *table,
+ u32 clock, u16 max_voltage, u16 *voltage)
+{
+ u32 i;
+
+ if ((table == NULL) || (table->count == 0))
+ return;
+
+ for (i= 0; i < table->count; i++) {
+ if (clock <= table->entries[i].clk) {
+ if (*voltage < table->entries[i].v)
+ *voltage = (u16)((table->entries[i].v < max_voltage) ?
+ table->entries[i].v : max_voltage);
+ return;
+ }
+ }
+
+ *voltage = (*voltage > max_voltage) ? *voltage : max_voltage;
+}
+
+static void btc_adjust_clock_combinations(struct amdgpu_device *adev,
+ const struct amdgpu_clock_and_voltage_limits *max_limits,
+ struct rv7xx_pl *pl)
+{
+
+ if ((pl->mclk == 0) || (pl->sclk == 0))
+ return;
+
+ if (pl->mclk == pl->sclk)
+ return;
+
+ if (pl->mclk > pl->sclk) {
+ if (((pl->mclk + (pl->sclk - 1)) / pl->sclk) > adev->pm.dpm.dyn_state.mclk_sclk_ratio)
+ pl->sclk = btc_get_valid_sclk(adev,
+ max_limits->sclk,
+ (pl->mclk +
+ (adev->pm.dpm.dyn_state.mclk_sclk_ratio - 1)) /
+ adev->pm.dpm.dyn_state.mclk_sclk_ratio);
+ } else {
+ if ((pl->sclk - pl->mclk) > adev->pm.dpm.dyn_state.sclk_mclk_delta)
+ pl->mclk = btc_get_valid_mclk(adev,
+ max_limits->mclk,
+ pl->sclk -
+ adev->pm.dpm.dyn_state.sclk_mclk_delta);
+ }
+}
+
+static void btc_apply_voltage_delta_rules(struct amdgpu_device *adev,
+ u16 max_vddc, u16 max_vddci,
+ u16 *vddc, u16 *vddci)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ u16 new_voltage;
+
+ if ((0 == *vddc) || (0 == *vddci))
+ return;
+
+ if (*vddc > *vddci) {
+ if ((*vddc - *vddci) > adev->pm.dpm.dyn_state.vddc_vddci_delta) {
+ new_voltage = btc_find_voltage(&eg_pi->vddci_voltage_table,
+ (*vddc - adev->pm.dpm.dyn_state.vddc_vddci_delta));
+ *vddci = (new_voltage < max_vddci) ? new_voltage : max_vddci;
+ }
+ } else {
+ if ((*vddci - *vddc) > adev->pm.dpm.dyn_state.vddc_vddci_delta) {
+ new_voltage = btc_find_voltage(&eg_pi->vddc_voltage_table,
+ (*vddci - adev->pm.dpm.dyn_state.vddc_vddci_delta));
+ *vddc = (new_voltage < max_vddc) ? new_voltage : max_vddc;
+ }
+ }
+}
+
+static enum amdgpu_pcie_gen r600_get_pcie_gen_support(struct amdgpu_device *adev,
+ u32 sys_mask,
+ enum amdgpu_pcie_gen asic_gen,
+ enum amdgpu_pcie_gen default_gen)
+{
+ switch (asic_gen) {
+ case AMDGPU_PCIE_GEN1:
+ return AMDGPU_PCIE_GEN1;
+ case AMDGPU_PCIE_GEN2:
+ return AMDGPU_PCIE_GEN2;
+ case AMDGPU_PCIE_GEN3:
+ return AMDGPU_PCIE_GEN3;
+ default:
+ if ((sys_mask & DRM_PCIE_SPEED_80) && (default_gen == AMDGPU_PCIE_GEN3))
+ return AMDGPU_PCIE_GEN3;
+ else if ((sys_mask & DRM_PCIE_SPEED_50) && (default_gen == AMDGPU_PCIE_GEN2))
+ return AMDGPU_PCIE_GEN2;
+ else
+ return AMDGPU_PCIE_GEN1;
+ }
+ return AMDGPU_PCIE_GEN1;
+}
+
+static void r600_calculate_u_and_p(u32 i, u32 r_c, u32 p_b,
+ u32 *p, u32 *u)
+{
+ u32 b_c = 0;
+ u32 i_c;
+ u32 tmp;
+
+ i_c = (i * r_c) / 100;
+ tmp = i_c >> p_b;
+
+ while (tmp) {
+ b_c++;
+ tmp >>= 1;
+ }
+
+ *u = (b_c + 1) / 2;
+ *p = i_c / (1 << (2 * (*u)));
+}
+
+static int r600_calculate_at(u32 t, u32 h, u32 fh, u32 fl, u32 *tl, u32 *th)
+{
+ u32 k, a, ah, al;
+ u32 t1;
+
+ if ((fl == 0) || (fh == 0) || (fl > fh))
+ return -EINVAL;
+
+ k = (100 * fh) / fl;
+ t1 = (t * (k - 100));
+ a = (1000 * (100 * h + t1)) / (10000 + (t1 / 100));
+ a = (a + 5) / 10;
+ ah = ((a * t) + 5000) / 10000;
+ al = a - ah;
+
+ *th = t - ah;
+ *tl = t + al;
+
+ return 0;
+}
+
+static bool r600_is_uvd_state(u32 class, u32 class2)
+{
+ if (class & ATOM_PPLIB_CLASSIFICATION_UVDSTATE)
+ return true;
+ if (class & ATOM_PPLIB_CLASSIFICATION_HD2STATE)
+ return true;
+ if (class & ATOM_PPLIB_CLASSIFICATION_HDSTATE)
+ return true;
+ if (class & ATOM_PPLIB_CLASSIFICATION_SDSTATE)
+ return true;
+ if (class2 & ATOM_PPLIB_CLASSIFICATION2_MVC)
+ return true;
+ return false;
+}
+
+static u8 rv770_get_memory_module_index(struct amdgpu_device *adev)
+{
+ return (u8) ((RREG32(BIOS_SCRATCH_4) >> 16) & 0xff);
+}
+
+static void rv770_get_max_vddc(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ u16 vddc;
+
+ if (amdgpu_atombios_get_max_vddc(adev, 0, 0, &vddc))
+ pi->max_vddc = 0;
+ else
+ pi->max_vddc = vddc;
+}
+
+static void rv770_get_engine_memory_ss(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct amdgpu_atom_ss ss;
+
+ pi->sclk_ss = amdgpu_atombios_get_asic_ss_info(adev, &ss,
+ ASIC_INTERNAL_ENGINE_SS, 0);
+ pi->mclk_ss = amdgpu_atombios_get_asic_ss_info(adev, &ss,
+ ASIC_INTERNAL_MEMORY_SS, 0);
+
+ if (pi->sclk_ss || pi->mclk_ss)
+ pi->dynamic_ss = true;
+ else
+ pi->dynamic_ss = false;
+}
+
+
+static void si_apply_state_adjust_rules(struct amdgpu_device *adev,
+ struct amdgpu_ps *rps)
+{
+ struct si_ps *ps = si_get_ps(rps);
+ struct amdgpu_clock_and_voltage_limits *max_limits;
+ bool disable_mclk_switching = false;
+ bool disable_sclk_switching = false;
+ u32 mclk, sclk;
+ u16 vddc, vddci, min_vce_voltage = 0;
+ u32 max_sclk_vddc, max_mclk_vddci, max_mclk_vddc;
+ u32 max_sclk = 0, max_mclk = 0;
+ int i;
+ struct si_dpm_quirk *p = si_dpm_quirk_list;
+
+ /* Apply dpm quirks */
+ while (p && p->chip_device != 0) {
+ if (adev->pdev->vendor == p->chip_vendor &&
+ adev->pdev->device == p->chip_device &&
+ adev->pdev->subsystem_vendor == p->subsys_vendor &&
+ adev->pdev->subsystem_device == p->subsys_device) {
+ max_sclk = p->max_sclk;
+ max_mclk = p->max_mclk;
+ break;
+ }
+ ++p;
+ }
+
+ if (rps->vce_active) {
+ rps->evclk = adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].evclk;
+ rps->ecclk = adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].ecclk;
+ si_get_vce_clock_voltage(adev, rps->evclk, rps->ecclk,
+ &min_vce_voltage);
+ } else {
+ rps->evclk = 0;
+ rps->ecclk = 0;
+ }
+
+ if ((adev->pm.dpm.new_active_crtc_count > 1) ||
+ si_dpm_vblank_too_short(adev))
+ disable_mclk_switching = true;
+
+ if (rps->vclk || rps->dclk) {
+ disable_mclk_switching = true;
+ disable_sclk_switching = true;
+ }
+
+ if (adev->pm.dpm.ac_power)
+ max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
+ else
+ max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc;
+
+ for (i = ps->performance_level_count - 2; i >= 0; i--) {
+ if (ps->performance_levels[i].vddc > ps->performance_levels[i+1].vddc)
+ ps->performance_levels[i].vddc = ps->performance_levels[i+1].vddc;
+ }
+ if (adev->pm.dpm.ac_power == false) {
+ for (i = 0; i < ps->performance_level_count; i++) {
+ if (ps->performance_levels[i].mclk > max_limits->mclk)
+ ps->performance_levels[i].mclk = max_limits->mclk;
+ if (ps->performance_levels[i].sclk > max_limits->sclk)
+ ps->performance_levels[i].sclk = max_limits->sclk;
+ if (ps->performance_levels[i].vddc > max_limits->vddc)
+ ps->performance_levels[i].vddc = max_limits->vddc;
+ if (ps->performance_levels[i].vddci > max_limits->vddci)
+ ps->performance_levels[i].vddci = max_limits->vddci;
+ }
+ }
+
+ /* limit clocks to max supported clocks based on voltage dependency tables */
+ btc_get_max_clock_from_voltage_dependency_table(&adev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
+ &max_sclk_vddc);
+ btc_get_max_clock_from_voltage_dependency_table(&adev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
+ &max_mclk_vddci);
+ btc_get_max_clock_from_voltage_dependency_table(&adev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
+ &max_mclk_vddc);
+
+ for (i = 0; i < ps->performance_level_count; i++) {
+ if (max_sclk_vddc) {
+ if (ps->performance_levels[i].sclk > max_sclk_vddc)
+ ps->performance_levels[i].sclk = max_sclk_vddc;
+ }
+ if (max_mclk_vddci) {
+ if (ps->performance_levels[i].mclk > max_mclk_vddci)
+ ps->performance_levels[i].mclk = max_mclk_vddci;
+ }
+ if (max_mclk_vddc) {
+ if (ps->performance_levels[i].mclk > max_mclk_vddc)
+ ps->performance_levels[i].mclk = max_mclk_vddc;
+ }
+ if (max_mclk) {
+ if (ps->performance_levels[i].mclk > max_mclk)
+ ps->performance_levels[i].mclk = max_mclk;
+ }
+ if (max_sclk) {
+ if (ps->performance_levels[i].sclk > max_sclk)
+ ps->performance_levels[i].sclk = max_sclk;
+ }
+ }
+
+ /* XXX validate the min clocks required for display */
+
+ if (disable_mclk_switching) {
+ mclk = ps->performance_levels[ps->performance_level_count - 1].mclk;
+ vddci = ps->performance_levels[ps->performance_level_count - 1].vddci;
+ } else {
+ mclk = ps->performance_levels[0].mclk;
+ vddci = ps->performance_levels[0].vddci;
+ }
+
+ if (disable_sclk_switching) {
+ sclk = ps->performance_levels[ps->performance_level_count - 1].sclk;
+ vddc = ps->performance_levels[ps->performance_level_count - 1].vddc;
+ } else {
+ sclk = ps->performance_levels[0].sclk;
+ vddc = ps->performance_levels[0].vddc;
+ }
+
+ if (rps->vce_active) {
+ if (sclk < adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].sclk)
+ sclk = adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].sclk;
+ if (mclk < adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].mclk)
+ mclk = adev->pm.dpm.vce_states[adev->pm.dpm.vce_level].mclk;
+ }
+
+ /* adjusted low state */
+ ps->performance_levels[0].sclk = sclk;
+ ps->performance_levels[0].mclk = mclk;
+ ps->performance_levels[0].vddc = vddc;
+ ps->performance_levels[0].vddci = vddci;
+
+ if (disable_sclk_switching) {
+ sclk = ps->performance_levels[0].sclk;
+ for (i = 1; i < ps->performance_level_count; i++) {
+ if (sclk < ps->performance_levels[i].sclk)
+ sclk = ps->performance_levels[i].sclk;
+ }
+ for (i = 0; i < ps->performance_level_count; i++) {
+ ps->performance_levels[i].sclk = sclk;
+ ps->performance_levels[i].vddc = vddc;
+ }
+ } else {
+ for (i = 1; i < ps->performance_level_count; i++) {
+ if (ps->performance_levels[i].sclk < ps->performance_levels[i - 1].sclk)
+ ps->performance_levels[i].sclk = ps->performance_levels[i - 1].sclk;
+ if (ps->performance_levels[i].vddc < ps->performance_levels[i - 1].vddc)
+ ps->performance_levels[i].vddc = ps->performance_levels[i - 1].vddc;
+ }
+ }
+
+ if (disable_mclk_switching) {
+ mclk = ps->performance_levels[0].mclk;
+ for (i = 1; i < ps->performance_level_count; i++) {
+ if (mclk < ps->performance_levels[i].mclk)
+ mclk = ps->performance_levels[i].mclk;
+ }
+ for (i = 0; i < ps->performance_level_count; i++) {
+ ps->performance_levels[i].mclk = mclk;
+ ps->performance_levels[i].vddci = vddci;
+ }
+ } else {
+ for (i = 1; i < ps->performance_level_count; i++) {
+ if (ps->performance_levels[i].mclk < ps->performance_levels[i - 1].mclk)
+ ps->performance_levels[i].mclk = ps->performance_levels[i - 1].mclk;
+ if (ps->performance_levels[i].vddci < ps->performance_levels[i - 1].vddci)
+ ps->performance_levels[i].vddci = ps->performance_levels[i - 1].vddci;
+ }
+ }
+
+ for (i = 0; i < ps->performance_level_count; i++)
+ btc_adjust_clock_combinations(adev, max_limits,
+ &ps->performance_levels[i]);
+
+ for (i = 0; i < ps->performance_level_count; i++) {
+ if (ps->performance_levels[i].vddc < min_vce_voltage)
+ ps->performance_levels[i].vddc = min_vce_voltage;
+ btc_apply_voltage_dependency_rules(&adev->pm.dpm.dyn_state.vddc_dependency_on_sclk,
+ ps->performance_levels[i].sclk,
+ max_limits->vddc, &ps->performance_levels[i].vddc);
+ btc_apply_voltage_dependency_rules(&adev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
+ ps->performance_levels[i].mclk,
+ max_limits->vddci, &ps->performance_levels[i].vddci);
+ btc_apply_voltage_dependency_rules(&adev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
+ ps->performance_levels[i].mclk,
+ max_limits->vddc, &ps->performance_levels[i].vddc);
+ btc_apply_voltage_dependency_rules(&adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk,
+ adev->clock.current_dispclk,
+ max_limits->vddc, &ps->performance_levels[i].vddc);
+ }
+
+ for (i = 0; i < ps->performance_level_count; i++) {
+ btc_apply_voltage_delta_rules(adev,
+ max_limits->vddc, max_limits->vddci,
+ &ps->performance_levels[i].vddc,
+ &ps->performance_levels[i].vddci);
+ }
+
+ ps->dc_compatible = true;
+ for (i = 0; i < ps->performance_level_count; i++) {
+ if (ps->performance_levels[i].vddc > adev->pm.dpm.dyn_state.max_clock_voltage_on_dc.vddc)
+ ps->dc_compatible = false;
+ }
+}
+
+#if 0
+static int si_read_smc_soft_register(struct amdgpu_device *adev,
+ u16 reg_offset, u32 *value)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ return amdgpu_si_read_smc_sram_dword(adev,
+ si_pi->soft_regs_start + reg_offset, value,
+ si_pi->sram_end);
+}
+#endif
+
+static int si_write_smc_soft_register(struct amdgpu_device *adev,
+ u16 reg_offset, u32 value)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ return amdgpu_si_write_smc_sram_dword(adev,
+ si_pi->soft_regs_start + reg_offset,
+ value, si_pi->sram_end);
+}
+
+static bool si_is_special_1gb_platform(struct amdgpu_device *adev)
+{
+ bool ret = false;
+ u32 tmp, width, row, column, bank, density;
+ bool is_memory_gddr5, is_special;
+
+ tmp = RREG32(MC_SEQ_MISC0);
+ is_memory_gddr5 = (MC_SEQ_MISC0_GDDR5_VALUE == ((tmp & MC_SEQ_MISC0_GDDR5_MASK) >> MC_SEQ_MISC0_GDDR5_SHIFT));
+ is_special = (MC_SEQ_MISC0_REV_ID_VALUE == ((tmp & MC_SEQ_MISC0_REV_ID_MASK) >> MC_SEQ_MISC0_REV_ID_SHIFT))
+ & (MC_SEQ_MISC0_VEN_ID_VALUE == ((tmp & MC_SEQ_MISC0_VEN_ID_MASK) >> MC_SEQ_MISC0_VEN_ID_SHIFT));
+
+ WREG32(MC_SEQ_IO_DEBUG_INDEX, 0xb);
+ width = ((RREG32(MC_SEQ_IO_DEBUG_DATA) >> 1) & 1) ? 16 : 32;
+
+ tmp = RREG32(MC_ARB_RAMCFG);
+ row = ((tmp & NOOFROWS_MASK) >> NOOFROWS_SHIFT) + 10;
+ column = ((tmp & NOOFCOLS_MASK) >> NOOFCOLS_SHIFT) + 8;
+ bank = ((tmp & NOOFBANK_MASK) >> NOOFBANK_SHIFT) + 2;
+
+ density = (1 << (row + column - 20 + bank)) * width;
+
+ if ((adev->pdev->device == 0x6819) &&
+ is_memory_gddr5 && is_special && (density == 0x400))
+ ret = true;
+
+ return ret;
+}
+
+static void si_get_leakage_vddc(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u16 vddc, count = 0;
+ int i, ret;
+
+ for (i = 0; i < SISLANDS_MAX_LEAKAGE_COUNT; i++) {
+ ret = amdgpu_atombios_get_leakage_vddc_based_on_leakage_idx(adev, &vddc, SISLANDS_LEAKAGE_INDEX0 + i);
+
+ if (!ret && (vddc > 0) && (vddc != (SISLANDS_LEAKAGE_INDEX0 + i))) {
+ si_pi->leakage_voltage.entries[count].voltage = vddc;
+ si_pi->leakage_voltage.entries[count].leakage_index =
+ SISLANDS_LEAKAGE_INDEX0 + i;
+ count++;
+ }
+ }
+ si_pi->leakage_voltage.count = count;
+}
+
+static int si_get_leakage_voltage_from_leakage_index(struct amdgpu_device *adev,
+ u32 index, u16 *leakage_voltage)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ int i;
+
+ if (leakage_voltage == NULL)
+ return -EINVAL;
+
+ if ((index & 0xff00) != 0xff00)
+ return -EINVAL;
+
+ if ((index & 0xff) > SISLANDS_MAX_LEAKAGE_COUNT + 1)
+ return -EINVAL;
+
+ if (index < SISLANDS_LEAKAGE_INDEX0)
+ return -EINVAL;
+
+ for (i = 0; i < si_pi->leakage_voltage.count; i++) {
+ if (si_pi->leakage_voltage.entries[i].leakage_index == index) {
+ *leakage_voltage = si_pi->leakage_voltage.entries[i].voltage;
+ return 0;
+ }
+ }
+ return -EAGAIN;
+}
+
+static void si_set_dpm_event_sources(struct amdgpu_device *adev, u32 sources)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ bool want_thermal_protection;
+ enum amdgpu_dpm_event_src dpm_event_src;
+
+ switch (sources) {
+ case 0:
+ default:
+ want_thermal_protection = false;
+ break;
+ case (1 << AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL):
+ want_thermal_protection = true;
+ dpm_event_src = AMDGPU_DPM_EVENT_SRC_DIGITAL;
+ break;
+ case (1 << AMDGPU_DPM_AUTO_THROTTLE_SRC_EXTERNAL):
+ want_thermal_protection = true;
+ dpm_event_src = AMDGPU_DPM_EVENT_SRC_EXTERNAL;
+ break;
+ case ((1 << AMDGPU_DPM_AUTO_THROTTLE_SRC_EXTERNAL) |
+ (1 << AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL)):
+ want_thermal_protection = true;
+ dpm_event_src = AMDGPU_DPM_EVENT_SRC_DIGIAL_OR_EXTERNAL;
+ break;
+ }
+
+ if (want_thermal_protection) {
+ WREG32_P(CG_THERMAL_CTRL, DPM_EVENT_SRC(dpm_event_src), ~DPM_EVENT_SRC_MASK);
+ if (pi->thermal_protection)
+ WREG32_P(GENERAL_PWRMGT, 0, ~THERMAL_PROTECTION_DIS);
+ } else {
+ WREG32_P(GENERAL_PWRMGT, THERMAL_PROTECTION_DIS, ~THERMAL_PROTECTION_DIS);
+ }
+}
+
+static void si_enable_auto_throttle_source(struct amdgpu_device *adev,
+ enum amdgpu_dpm_auto_throttle_src source,
+ bool enable)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+
+ if (enable) {
+ if (!(pi->active_auto_throttle_sources & (1 << source))) {
+ pi->active_auto_throttle_sources |= 1 << source;
+ si_set_dpm_event_sources(adev, pi->active_auto_throttle_sources);
+ }
+ } else {
+ if (pi->active_auto_throttle_sources & (1 << source)) {
+ pi->active_auto_throttle_sources &= ~(1 << source);
+ si_set_dpm_event_sources(adev, pi->active_auto_throttle_sources);
+ }
+ }
+}
+
+static void si_start_dpm(struct amdgpu_device *adev)
+{
+ WREG32_P(GENERAL_PWRMGT, GLOBAL_PWRMGT_EN, ~GLOBAL_PWRMGT_EN);
+}
+
+static void si_stop_dpm(struct amdgpu_device *adev)
+{
+ WREG32_P(GENERAL_PWRMGT, 0, ~GLOBAL_PWRMGT_EN);
+}
+
+static void si_enable_sclk_control(struct amdgpu_device *adev, bool enable)
+{
+ if (enable)
+ WREG32_P(SCLK_PWRMGT_CNTL, 0, ~SCLK_PWRMGT_OFF);
+ else
+ WREG32_P(SCLK_PWRMGT_CNTL, SCLK_PWRMGT_OFF, ~SCLK_PWRMGT_OFF);
+
+}
+
+#if 0
+static int si_notify_hardware_of_thermal_state(struct amdgpu_device *adev,
+ u32 thermal_level)
+{
+ PPSMC_Result ret;
+
+ if (thermal_level == 0) {
+ ret = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableThermalInterrupt);
+ if (ret == PPSMC_Result_OK)
+ return 0;
+ else
+ return -EINVAL;
+ }
+ return 0;
+}
+
+static void si_notify_hardware_vpu_recovery_event(struct amdgpu_device *adev)
+{
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_tdr_is_about_to_happen, true);
+}
+#endif
+
+#if 0
+static int si_notify_hw_of_powersource(struct amdgpu_device *adev, bool ac_power)
+{
+ if (ac_power)
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_RunningOnAC) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+
+ return 0;
+}
+#endif
+
+static PPSMC_Result si_send_msg_to_smc_with_parameter(struct amdgpu_device *adev,
+ PPSMC_Msg msg, u32 parameter)
+{
+ WREG32(SMC_SCRATCH0, parameter);
+ return amdgpu_si_send_msg_to_smc(adev, msg);
+}
+
+static int si_restrict_performance_levels_before_switch(struct amdgpu_device *adev)
+{
+ if (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_NoForcedLevel) != PPSMC_Result_OK)
+ return -EINVAL;
+
+ return (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetEnabledLevels, 1) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+}
+
+static int si_dpm_force_performance_level(struct amdgpu_device *adev,
+ enum amdgpu_dpm_forced_level level)
+{
+ struct amdgpu_ps *rps = adev->pm.dpm.current_ps;
+ struct si_ps *ps = si_get_ps(rps);
+ u32 levels = ps->performance_level_count;
+
+ if (level == AMDGPU_DPM_FORCED_LEVEL_HIGH) {
+ if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetEnabledLevels, levels) != PPSMC_Result_OK)
+ return -EINVAL;
+
+ if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetForcedLevels, 1) != PPSMC_Result_OK)
+ return -EINVAL;
+ } else if (level == AMDGPU_DPM_FORCED_LEVEL_LOW) {
+ if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetForcedLevels, 0) != PPSMC_Result_OK)
+ return -EINVAL;
+
+ if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetEnabledLevels, 1) != PPSMC_Result_OK)
+ return -EINVAL;
+ } else if (level == AMDGPU_DPM_FORCED_LEVEL_AUTO) {
+ if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetForcedLevels, 0) != PPSMC_Result_OK)
+ return -EINVAL;
+
+ if (si_send_msg_to_smc_with_parameter(adev, PPSMC_MSG_SetEnabledLevels, levels) != PPSMC_Result_OK)
+ return -EINVAL;
+ }
+
+ adev->pm.dpm.forced_level = level;
+
+ return 0;
+}
+
+#if 0
+static int si_set_boot_state(struct amdgpu_device *adev)
+{
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_SwitchToInitialState) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+}
+#endif
+
+static int si_set_sw_state(struct amdgpu_device *adev)
+{
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_SwitchToSwState) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+}
+
+static int si_halt_smc(struct amdgpu_device *adev)
+{
+ if (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_Halt) != PPSMC_Result_OK)
+ return -EINVAL;
+
+ return (amdgpu_si_wait_for_smc_inactive(adev) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+}
+
+static int si_resume_smc(struct amdgpu_device *adev)
+{
+ if (amdgpu_si_send_msg_to_smc(adev, PPSMC_FlushDataCache) != PPSMC_Result_OK)
+ return -EINVAL;
+
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_Resume) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+}
+
+static void si_dpm_start_smc(struct amdgpu_device *adev)
+{
+ amdgpu_si_program_jump_on_start(adev);
+ amdgpu_si_start_smc(adev);
+ amdgpu_si_smc_clock(adev, true);
+}
+
+static void si_dpm_stop_smc(struct amdgpu_device *adev)
+{
+ amdgpu_si_reset_smc(adev);
+ amdgpu_si_smc_clock(adev, false);
+}
+
+static int si_process_firmware_header(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 tmp;
+ int ret;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_stateTable,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->state_table_start = tmp;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_softRegisters,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->soft_regs_start = tmp;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_mcRegisterTable,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->mc_reg_table_start = tmp;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_fanTable,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->fan_table_start = tmp;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_mcArbDramAutoRefreshTable,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->arb_table_start = tmp;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_CacConfigTable,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->cac_table_start = tmp;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_DteConfiguration,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->dte_table_start = tmp;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_spllTable,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->spll_table_start = tmp;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev,
+ SISLANDS_SMC_FIRMWARE_HEADER_LOCATION +
+ SISLANDS_SMC_FIRMWARE_HEADER_PAPMParameters,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ si_pi->papm_cfg_table_start = tmp;
+
+ return ret;
+}
+
+static void si_read_clock_registers(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ si_pi->clock_registers.cg_spll_func_cntl = RREG32(CG_SPLL_FUNC_CNTL);
+ si_pi->clock_registers.cg_spll_func_cntl_2 = RREG32(CG_SPLL_FUNC_CNTL_2);
+ si_pi->clock_registers.cg_spll_func_cntl_3 = RREG32(CG_SPLL_FUNC_CNTL_3);
+ si_pi->clock_registers.cg_spll_func_cntl_4 = RREG32(CG_SPLL_FUNC_CNTL_4);
+ si_pi->clock_registers.cg_spll_spread_spectrum = RREG32(CG_SPLL_SPREAD_SPECTRUM);
+ si_pi->clock_registers.cg_spll_spread_spectrum_2 = RREG32(CG_SPLL_SPREAD_SPECTRUM_2);
+ si_pi->clock_registers.dll_cntl = RREG32(DLL_CNTL);
+ si_pi->clock_registers.mclk_pwrmgt_cntl = RREG32(MCLK_PWRMGT_CNTL);
+ si_pi->clock_registers.mpll_ad_func_cntl = RREG32(MPLL_AD_FUNC_CNTL);
+ si_pi->clock_registers.mpll_dq_func_cntl = RREG32(MPLL_DQ_FUNC_CNTL);
+ si_pi->clock_registers.mpll_func_cntl = RREG32(MPLL_FUNC_CNTL);
+ si_pi->clock_registers.mpll_func_cntl_1 = RREG32(MPLL_FUNC_CNTL_1);
+ si_pi->clock_registers.mpll_func_cntl_2 = RREG32(MPLL_FUNC_CNTL_2);
+ si_pi->clock_registers.mpll_ss1 = RREG32(MPLL_SS1);
+ si_pi->clock_registers.mpll_ss2 = RREG32(MPLL_SS2);
+}
+
+static void si_enable_thermal_protection(struct amdgpu_device *adev,
+ bool enable)
+{
+ if (enable)
+ WREG32_P(GENERAL_PWRMGT, 0, ~THERMAL_PROTECTION_DIS);
+ else
+ WREG32_P(GENERAL_PWRMGT, THERMAL_PROTECTION_DIS, ~THERMAL_PROTECTION_DIS);
+}
+
+static void si_enable_acpi_power_management(struct amdgpu_device *adev)
+{
+ WREG32_P(GENERAL_PWRMGT, STATIC_PM_EN, ~STATIC_PM_EN);
+}
+
+#if 0
+static int si_enter_ulp_state(struct amdgpu_device *adev)
+{
+ WREG32(SMC_MESSAGE_0, PPSMC_MSG_SwitchToMinimumPower);
+
+ udelay(25000);
+
+ return 0;
+}
+
+static int si_exit_ulp_state(struct amdgpu_device *adev)
+{
+ int i;
+
+ WREG32(SMC_MESSAGE_0, PPSMC_MSG_ResumeFromMinimumPower);
+
+ udelay(7000);
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (RREG32(SMC_RESP_0) == 1)
+ break;
+ udelay(1000);
+ }
+
+ return 0;
+}
+#endif
+
+static int si_notify_smc_display_change(struct amdgpu_device *adev,
+ bool has_display)
+{
+ PPSMC_Msg msg = has_display ?
+ PPSMC_MSG_HasDisplay : PPSMC_MSG_NoDisplay;
+
+ return (amdgpu_si_send_msg_to_smc(adev, msg) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+}
+
+static void si_program_response_times(struct amdgpu_device *adev)
+{
+ u32 voltage_response_time, backbias_response_time, acpi_delay_time, vbi_time_out;
+ u32 vddc_dly, acpi_dly, vbi_dly;
+ u32 reference_clock;
+
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_mvdd_chg_time, 1);
+
+ voltage_response_time = (u32)adev->pm.dpm.voltage_response_time;
+ backbias_response_time = (u32)adev->pm.dpm.backbias_response_time;
+
+ if (voltage_response_time == 0)
+ voltage_response_time = 1000;
+
+ acpi_delay_time = 15000;
+ vbi_time_out = 100000;
+
+ reference_clock = amdgpu_asic_get_xclk(adev);
+
+ vddc_dly = (voltage_response_time * reference_clock) / 100;
+ acpi_dly = (acpi_delay_time * reference_clock) / 100;
+ vbi_dly = (vbi_time_out * reference_clock) / 100;
+
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_delay_vreg, vddc_dly);
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_delay_acpi, acpi_dly);
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_mclk_chg_timeout, vbi_dly);
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_mc_block_delay, 0xAA);
+}
+
+static void si_program_ds_registers(struct amdgpu_device *adev)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ u32 tmp;
+
+ /* DEEP_SLEEP_CLK_SEL field should be 0x10 on tahiti A0 */
+ if (adev->asic_type == CHIP_TAHITI && adev->rev_id == 0x0)
+ tmp = 0x10;
+ else
+ tmp = 0x1;
+
+ if (eg_pi->sclk_deep_sleep) {
+ WREG32_P(MISC_CLK_CNTL, DEEP_SLEEP_CLK_SEL(tmp), ~DEEP_SLEEP_CLK_SEL_MASK);
+ WREG32_P(CG_SPLL_AUTOSCALE_CNTL, AUTOSCALE_ON_SS_CLEAR,
+ ~AUTOSCALE_ON_SS_CLEAR);
+ }
+}
+
+static void si_program_display_gap(struct amdgpu_device *adev)
+{
+ u32 tmp, pipe;
+ int i;
+
+ tmp = RREG32(CG_DISPLAY_GAP_CNTL) & ~(DISP1_GAP_MASK | DISP2_GAP_MASK);
+ if (adev->pm.dpm.new_active_crtc_count > 0)
+ tmp |= DISP1_GAP(R600_PM_DISPLAY_GAP_VBLANK_OR_WM);
+ else
+ tmp |= DISP1_GAP(R600_PM_DISPLAY_GAP_IGNORE);
+
+ if (adev->pm.dpm.new_active_crtc_count > 1)
+ tmp |= DISP2_GAP(R600_PM_DISPLAY_GAP_VBLANK_OR_WM);
+ else
+ tmp |= DISP2_GAP(R600_PM_DISPLAY_GAP_IGNORE);
+
+ WREG32(CG_DISPLAY_GAP_CNTL, tmp);
+
+ tmp = RREG32(DCCG_DISP_SLOW_SELECT_REG);
+ pipe = (tmp & DCCG_DISP1_SLOW_SELECT_MASK) >> DCCG_DISP1_SLOW_SELECT_SHIFT;
+
+ if ((adev->pm.dpm.new_active_crtc_count > 0) &&
+ (!(adev->pm.dpm.new_active_crtcs & (1 << pipe)))) {
+ /* find the first active crtc */
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ if (adev->pm.dpm.new_active_crtcs & (1 << i))
+ break;
+ }
+ if (i == adev->mode_info.num_crtc)
+ pipe = 0;
+ else
+ pipe = i;
+
+ tmp &= ~DCCG_DISP1_SLOW_SELECT_MASK;
+ tmp |= DCCG_DISP1_SLOW_SELECT(pipe);
+ WREG32(DCCG_DISP_SLOW_SELECT_REG, tmp);
+ }
+
+ /* Setting this to false forces the performance state to low if the crtcs are disabled.
+ * This can be a problem on PowerXpress systems or if you want to use the card
+ * for offscreen rendering or compute if there are no crtcs enabled.
+ */
+ si_notify_smc_display_change(adev, adev->pm.dpm.new_active_crtc_count > 0);
+}
+
+static void si_enable_spread_spectrum(struct amdgpu_device *adev, bool enable)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+
+ if (enable) {
+ if (pi->sclk_ss)
+ WREG32_P(GENERAL_PWRMGT, DYN_SPREAD_SPECTRUM_EN, ~DYN_SPREAD_SPECTRUM_EN);
+ } else {
+ WREG32_P(CG_SPLL_SPREAD_SPECTRUM, 0, ~SSEN);
+ WREG32_P(GENERAL_PWRMGT, 0, ~DYN_SPREAD_SPECTRUM_EN);
+ }
+}
+
+static void si_setup_bsp(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ u32 xclk = amdgpu_asic_get_xclk(adev);
+
+ r600_calculate_u_and_p(pi->asi,
+ xclk,
+ 16,
+ &pi->bsp,
+ &pi->bsu);
+
+ r600_calculate_u_and_p(pi->pasi,
+ xclk,
+ 16,
+ &pi->pbsp,
+ &pi->pbsu);
+
+
+ pi->dsp = BSP(pi->bsp) | BSU(pi->bsu);
+ pi->psp = BSP(pi->pbsp) | BSU(pi->pbsu);
+
+ WREG32(CG_BSP, pi->dsp);
+}
+
+static void si_program_git(struct amdgpu_device *adev)
+{
+ WREG32_P(CG_GIT, CG_GICST(R600_GICST_DFLT), ~CG_GICST_MASK);
+}
+
+static void si_program_tp(struct amdgpu_device *adev)
+{
+ int i;
+ enum r600_td td = R600_TD_DFLT;
+
+ for (i = 0; i < R600_PM_NUMBER_OF_TC; i++)
+ WREG32(CG_FFCT_0 + i, (UTC_0(r600_utc[i]) | DTC_0(r600_dtc[i])));
+
+ if (td == R600_TD_AUTO)
+ WREG32_P(SCLK_PWRMGT_CNTL, 0, ~FIR_FORCE_TREND_SEL);
+ else
+ WREG32_P(SCLK_PWRMGT_CNTL, FIR_FORCE_TREND_SEL, ~FIR_FORCE_TREND_SEL);
+
+ if (td == R600_TD_UP)
+ WREG32_P(SCLK_PWRMGT_CNTL, 0, ~FIR_TREND_MODE);
+
+ if (td == R600_TD_DOWN)
+ WREG32_P(SCLK_PWRMGT_CNTL, FIR_TREND_MODE, ~FIR_TREND_MODE);
+}
+
+static void si_program_tpp(struct amdgpu_device *adev)
+{
+ WREG32(CG_TPC, R600_TPC_DFLT);
+}
+
+static void si_program_sstp(struct amdgpu_device *adev)
+{
+ WREG32(CG_SSP, (SSTU(R600_SSTU_DFLT) | SST(R600_SST_DFLT)));
+}
+
+static void si_enable_display_gap(struct amdgpu_device *adev)
+{
+ u32 tmp = RREG32(CG_DISPLAY_GAP_CNTL);
+
+ tmp &= ~(DISP1_GAP_MASK | DISP2_GAP_MASK);
+ tmp |= (DISP1_GAP(R600_PM_DISPLAY_GAP_IGNORE) |
+ DISP2_GAP(R600_PM_DISPLAY_GAP_IGNORE));
+
+ tmp &= ~(DISP1_GAP_MCHG_MASK | DISP2_GAP_MCHG_MASK);
+ tmp |= (DISP1_GAP_MCHG(R600_PM_DISPLAY_GAP_VBLANK) |
+ DISP2_GAP_MCHG(R600_PM_DISPLAY_GAP_IGNORE));
+ WREG32(CG_DISPLAY_GAP_CNTL, tmp);
+}
+
+static void si_program_vc(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+
+ WREG32(CG_FTV, pi->vrc);
+}
+
+static void si_clear_vc(struct amdgpu_device *adev)
+{
+ WREG32(CG_FTV, 0);
+}
+
+static u8 si_get_ddr3_mclk_frequency_ratio(u32 memory_clock)
+{
+ u8 mc_para_index;
+
+ if (memory_clock < 10000)
+ mc_para_index = 0;
+ else if (memory_clock >= 80000)
+ mc_para_index = 0x0f;
+ else
+ mc_para_index = (u8)((memory_clock - 10000) / 5000 + 1);
+ return mc_para_index;
+}
+
+static u8 si_get_mclk_frequency_ratio(u32 memory_clock, bool strobe_mode)
+{
+ u8 mc_para_index;
+
+ if (strobe_mode) {
+ if (memory_clock < 12500)
+ mc_para_index = 0x00;
+ else if (memory_clock > 47500)
+ mc_para_index = 0x0f;
+ else
+ mc_para_index = (u8)((memory_clock - 10000) / 2500);
+ } else {
+ if (memory_clock < 65000)
+ mc_para_index = 0x00;
+ else if (memory_clock > 135000)
+ mc_para_index = 0x0f;
+ else
+ mc_para_index = (u8)((memory_clock - 60000) / 5000);
+ }
+ return mc_para_index;
+}
+
+static u8 si_get_strobe_mode_settings(struct amdgpu_device *adev, u32 mclk)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ bool strobe_mode = false;
+ u8 result = 0;
+
+ if (mclk <= pi->mclk_strobe_mode_threshold)
+ strobe_mode = true;
+
+ if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5)
+ result = si_get_mclk_frequency_ratio(mclk, strobe_mode);
+ else
+ result = si_get_ddr3_mclk_frequency_ratio(mclk);
+
+ if (strobe_mode)
+ result |= SISLANDS_SMC_STROBE_ENABLE;
+
+ return result;
+}
+
+static int si_upload_firmware(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ amdgpu_si_reset_smc(adev);
+ amdgpu_si_smc_clock(adev, false);
+
+ return amdgpu_si_load_smc_ucode(adev, si_pi->sram_end);
+}
+
+static bool si_validate_phase_shedding_tables(struct amdgpu_device *adev,
+ const struct atom_voltage_table *table,
+ const struct amdgpu_phase_shedding_limits_table *limits)
+{
+ u32 data, num_bits, num_levels;
+
+ if ((table == NULL) || (limits == NULL))
+ return false;
+
+ data = table->mask_low;
+
+ num_bits = hweight32(data);
+
+ if (num_bits == 0)
+ return false;
+
+ num_levels = (1 << num_bits);
+
+ if (table->count != num_levels)
+ return false;
+
+ if (limits->count != (num_levels - 1))
+ return false;
+
+ return true;
+}
+
+static void si_trim_voltage_table_to_fit_state_table(struct amdgpu_device *adev,
+ u32 max_voltage_steps,
+ struct atom_voltage_table *voltage_table)
+{
+ unsigned int i, diff;
+
+ if (voltage_table->count <= max_voltage_steps)
+ return;
+
+ diff = voltage_table->count - max_voltage_steps;
+
+ for (i= 0; i < max_voltage_steps; i++)
+ voltage_table->entries[i] = voltage_table->entries[i + diff];
+
+ voltage_table->count = max_voltage_steps;
+}
+
+static int si_get_svi2_voltage_table(struct amdgpu_device *adev,
+ struct amdgpu_clock_voltage_dependency_table *voltage_dependency_table,
+ struct atom_voltage_table *voltage_table)
+{
+ u32 i;
+
+ if (voltage_dependency_table == NULL)
+ return -EINVAL;
+
+ voltage_table->mask_low = 0;
+ voltage_table->phase_delay = 0;
+
+ voltage_table->count = voltage_dependency_table->count;
+ for (i = 0; i < voltage_table->count; i++) {
+ voltage_table->entries[i].value = voltage_dependency_table->entries[i].v;
+ voltage_table->entries[i].smio_low = 0;
+ }
+
+ return 0;
+}
+
+static int si_construct_voltage_tables(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ int ret;
+
+ if (pi->voltage_control) {
+ ret = amdgpu_atombios_get_voltage_table(adev, VOLTAGE_TYPE_VDDC,
+ VOLTAGE_OBJ_GPIO_LUT, &eg_pi->vddc_voltage_table);
+ if (ret)
+ return ret;
+
+ if (eg_pi->vddc_voltage_table.count > SISLANDS_MAX_NO_VREG_STEPS)
+ si_trim_voltage_table_to_fit_state_table(adev,
+ SISLANDS_MAX_NO_VREG_STEPS,
+ &eg_pi->vddc_voltage_table);
+ } else if (si_pi->voltage_control_svi2) {
+ ret = si_get_svi2_voltage_table(adev,
+ &adev->pm.dpm.dyn_state.vddc_dependency_on_mclk,
+ &eg_pi->vddc_voltage_table);
+ if (ret)
+ return ret;
+ } else {
+ return -EINVAL;
+ }
+
+ if (eg_pi->vddci_control) {
+ ret = amdgpu_atombios_get_voltage_table(adev, VOLTAGE_TYPE_VDDCI,
+ VOLTAGE_OBJ_GPIO_LUT, &eg_pi->vddci_voltage_table);
+ if (ret)
+ return ret;
+
+ if (eg_pi->vddci_voltage_table.count > SISLANDS_MAX_NO_VREG_STEPS)
+ si_trim_voltage_table_to_fit_state_table(adev,
+ SISLANDS_MAX_NO_VREG_STEPS,
+ &eg_pi->vddci_voltage_table);
+ }
+ if (si_pi->vddci_control_svi2) {
+ ret = si_get_svi2_voltage_table(adev,
+ &adev->pm.dpm.dyn_state.vddci_dependency_on_mclk,
+ &eg_pi->vddci_voltage_table);
+ if (ret)
+ return ret;
+ }
+
+ if (pi->mvdd_control) {
+ ret = amdgpu_atombios_get_voltage_table(adev, VOLTAGE_TYPE_MVDDC,
+ VOLTAGE_OBJ_GPIO_LUT, &si_pi->mvdd_voltage_table);
+
+ if (ret) {
+ pi->mvdd_control = false;
+ return ret;
+ }
+
+ if (si_pi->mvdd_voltage_table.count == 0) {
+ pi->mvdd_control = false;
+ return -EINVAL;
+ }
+
+ if (si_pi->mvdd_voltage_table.count > SISLANDS_MAX_NO_VREG_STEPS)
+ si_trim_voltage_table_to_fit_state_table(adev,
+ SISLANDS_MAX_NO_VREG_STEPS,
+ &si_pi->mvdd_voltage_table);
+ }
+
+ if (si_pi->vddc_phase_shed_control) {
+ ret = amdgpu_atombios_get_voltage_table(adev, VOLTAGE_TYPE_VDDC,
+ VOLTAGE_OBJ_PHASE_LUT, &si_pi->vddc_phase_shed_table);
+ if (ret)
+ si_pi->vddc_phase_shed_control = false;
+
+ if ((si_pi->vddc_phase_shed_table.count == 0) ||
+ (si_pi->vddc_phase_shed_table.count > SISLANDS_MAX_NO_VREG_STEPS))
+ si_pi->vddc_phase_shed_control = false;
+ }
+
+ return 0;
+}
+
+static void si_populate_smc_voltage_table(struct amdgpu_device *adev,
+ const struct atom_voltage_table *voltage_table,
+ SISLANDS_SMC_STATETABLE *table)
+{
+ unsigned int i;
+
+ for (i = 0; i < voltage_table->count; i++)
+ table->lowSMIO[i] |= cpu_to_be32(voltage_table->entries[i].smio_low);
+}
+
+static int si_populate_smc_voltage_tables(struct amdgpu_device *adev,
+ SISLANDS_SMC_STATETABLE *table)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u8 i;
+
+ if (si_pi->voltage_control_svi2) {
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svc,
+ si_pi->svc_gpio_id);
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svd,
+ si_pi->svd_gpio_id);
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_svi_rework_plat_type,
+ 2);
+ } else {
+ if (eg_pi->vddc_voltage_table.count) {
+ si_populate_smc_voltage_table(adev, &eg_pi->vddc_voltage_table, table);
+ table->voltageMaskTable.lowMask[SISLANDS_SMC_VOLTAGEMASK_VDDC] =
+ cpu_to_be32(eg_pi->vddc_voltage_table.mask_low);
+
+ for (i = 0; i < eg_pi->vddc_voltage_table.count; i++) {
+ if (pi->max_vddc_in_table <= eg_pi->vddc_voltage_table.entries[i].value) {
+ table->maxVDDCIndexInPPTable = i;
+ break;
+ }
+ }
+ }
+
+ if (eg_pi->vddci_voltage_table.count) {
+ si_populate_smc_voltage_table(adev, &eg_pi->vddci_voltage_table, table);
+
+ table->voltageMaskTable.lowMask[SISLANDS_SMC_VOLTAGEMASK_VDDCI] =
+ cpu_to_be32(eg_pi->vddci_voltage_table.mask_low);
+ }
+
+
+ if (si_pi->mvdd_voltage_table.count) {
+ si_populate_smc_voltage_table(adev, &si_pi->mvdd_voltage_table, table);
+
+ table->voltageMaskTable.lowMask[SISLANDS_SMC_VOLTAGEMASK_MVDD] =
+ cpu_to_be32(si_pi->mvdd_voltage_table.mask_low);
+ }
+
+ if (si_pi->vddc_phase_shed_control) {
+ if (si_validate_phase_shedding_tables(adev, &si_pi->vddc_phase_shed_table,
+ &adev->pm.dpm.dyn_state.phase_shedding_limits_table)) {
+ si_populate_smc_voltage_table(adev, &si_pi->vddc_phase_shed_table, table);
+
+ table->phaseMaskTable.lowMask[SISLANDS_SMC_VOLTAGEMASK_VDDC] =
+ cpu_to_be32(si_pi->vddc_phase_shed_table.mask_low);
+
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_phase_shedding_delay,
+ (u32)si_pi->vddc_phase_shed_table.phase_delay);
+ } else {
+ si_pi->vddc_phase_shed_control = false;
+ }
+ }
+ }
+
+ return 0;
+}
+
+static int si_populate_voltage_value(struct amdgpu_device *adev,
+ const struct atom_voltage_table *table,
+ u16 value, SISLANDS_SMC_VOLTAGE_VALUE *voltage)
+{
+ unsigned int i;
+
+ for (i = 0; i < table->count; i++) {
+ if (value <= table->entries[i].value) {
+ voltage->index = (u8)i;
+ voltage->value = cpu_to_be16(table->entries[i].value);
+ break;
+ }
+ }
+
+ if (i >= table->count)
+ return -EINVAL;
+
+ return 0;
+}
+
+static int si_populate_mvdd_value(struct amdgpu_device *adev, u32 mclk,
+ SISLANDS_SMC_VOLTAGE_VALUE *voltage)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ if (pi->mvdd_control) {
+ if (mclk <= pi->mvdd_split_frequency)
+ voltage->index = 0;
+ else
+ voltage->index = (u8)(si_pi->mvdd_voltage_table.count) - 1;
+
+ voltage->value = cpu_to_be16(si_pi->mvdd_voltage_table.entries[voltage->index].value);
+ }
+ return 0;
+}
+
+static int si_get_std_voltage_value(struct amdgpu_device *adev,
+ SISLANDS_SMC_VOLTAGE_VALUE *voltage,
+ u16 *std_voltage)
+{
+ u16 v_index;
+ bool voltage_found = false;
+ *std_voltage = be16_to_cpu(voltage->value);
+
+ if (adev->pm.dpm.dyn_state.cac_leakage_table.entries) {
+ if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_NEW_CAC_VOLTAGE) {
+ if (adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.entries == NULL)
+ return -EINVAL;
+
+ for (v_index = 0; (u32)v_index < adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.count; v_index++) {
+ if (be16_to_cpu(voltage->value) ==
+ (u16)adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.entries[v_index].v) {
+ voltage_found = true;
+ if ((u32)v_index < adev->pm.dpm.dyn_state.cac_leakage_table.count)
+ *std_voltage =
+ adev->pm.dpm.dyn_state.cac_leakage_table.entries[v_index].vddc;
+ else
+ *std_voltage =
+ adev->pm.dpm.dyn_state.cac_leakage_table.entries[adev->pm.dpm.dyn_state.cac_leakage_table.count-1].vddc;
+ break;
+ }
+ }
+
+ if (!voltage_found) {
+ for (v_index = 0; (u32)v_index < adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.count; v_index++) {
+ if (be16_to_cpu(voltage->value) <=
+ (u16)adev->pm.dpm.dyn_state.vddc_dependency_on_sclk.entries[v_index].v) {
+ voltage_found = true;
+ if ((u32)v_index < adev->pm.dpm.dyn_state.cac_leakage_table.count)
+ *std_voltage =
+ adev->pm.dpm.dyn_state.cac_leakage_table.entries[v_index].vddc;
+ else
+ *std_voltage =
+ adev->pm.dpm.dyn_state.cac_leakage_table.entries[adev->pm.dpm.dyn_state.cac_leakage_table.count-1].vddc;
+ break;
+ }
+ }
+ }
+ } else {
+ if ((u32)voltage->index < adev->pm.dpm.dyn_state.cac_leakage_table.count)
+ *std_voltage = adev->pm.dpm.dyn_state.cac_leakage_table.entries[voltage->index].vddc;
+ }
+ }
+
+ return 0;
+}
+
+static int si_populate_std_voltage_value(struct amdgpu_device *adev,
+ u16 value, u8 index,
+ SISLANDS_SMC_VOLTAGE_VALUE *voltage)
+{
+ voltage->index = index;
+ voltage->value = cpu_to_be16(value);
+
+ return 0;
+}
+
+static int si_populate_phase_shedding_value(struct amdgpu_device *adev,
+ const struct amdgpu_phase_shedding_limits_table *limits,
+ u16 voltage, u32 sclk, u32 mclk,
+ SISLANDS_SMC_VOLTAGE_VALUE *smc_voltage)
+{
+ unsigned int i;
+
+ for (i = 0; i < limits->count; i++) {
+ if ((voltage <= limits->entries[i].voltage) &&
+ (sclk <= limits->entries[i].sclk) &&
+ (mclk <= limits->entries[i].mclk))
+ break;
+ }
+
+ smc_voltage->phase_settings = (u8)i;
+
+ return 0;
+}
+
+static int si_init_arb_table_index(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 tmp;
+ int ret;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev, si_pi->arb_table_start,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ tmp &= 0x00FFFFFF;
+ tmp |= MC_CG_ARB_FREQ_F1 << 24;
+
+ return amdgpu_si_write_smc_sram_dword(adev, si_pi->arb_table_start,
+ tmp, si_pi->sram_end);
+}
+
+static int si_initial_switch_from_arb_f0_to_f1(struct amdgpu_device *adev)
+{
+ return ni_copy_and_switch_arb_sets(adev, MC_CG_ARB_FREQ_F0, MC_CG_ARB_FREQ_F1);
+}
+
+static int si_reset_to_default(struct amdgpu_device *adev)
+{
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_ResetToDefaults) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+}
+
+static int si_force_switch_to_arb_f0(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 tmp;
+ int ret;
+
+ ret = amdgpu_si_read_smc_sram_dword(adev, si_pi->arb_table_start,
+ &tmp, si_pi->sram_end);
+ if (ret)
+ return ret;
+
+ tmp = (tmp >> 24) & 0xff;
+
+ if (tmp == MC_CG_ARB_FREQ_F0)
+ return 0;
+
+ return ni_copy_and_switch_arb_sets(adev, tmp, MC_CG_ARB_FREQ_F0);
+}
+
+static u32 si_calculate_memory_refresh_rate(struct amdgpu_device *adev,
+ u32 engine_clock)
+{
+ u32 dram_rows;
+ u32 dram_refresh_rate;
+ u32 mc_arb_rfsh_rate;
+ u32 tmp = (RREG32(MC_ARB_RAMCFG) & NOOFROWS_MASK) >> NOOFROWS_SHIFT;
+
+ if (tmp >= 4)
+ dram_rows = 16384;
+ else
+ dram_rows = 1 << (tmp + 10);
+
+ dram_refresh_rate = 1 << ((RREG32(MC_SEQ_MISC0) & 0x3) + 3);
+ mc_arb_rfsh_rate = ((engine_clock * 10) * dram_refresh_rate / dram_rows - 32) / 64;
+
+ return mc_arb_rfsh_rate;
+}
+
+static int si_populate_memory_timing_parameters(struct amdgpu_device *adev,
+ struct rv7xx_pl *pl,
+ SMC_SIslands_MCArbDramTimingRegisterSet *arb_regs)
+{
+ u32 dram_timing;
+ u32 dram_timing2;
+ u32 burst_time;
+
+ arb_regs->mc_arb_rfsh_rate =
+ (u8)si_calculate_memory_refresh_rate(adev, pl->sclk);
+
+ amdgpu_atombios_set_engine_dram_timings(adev,
+ pl->sclk,
+ pl->mclk);
+
+ dram_timing = RREG32(MC_ARB_DRAM_TIMING);
+ dram_timing2 = RREG32(MC_ARB_DRAM_TIMING2);
+ burst_time = RREG32(MC_ARB_BURST_TIME) & STATE0_MASK;
+
+ arb_regs->mc_arb_dram_timing = cpu_to_be32(dram_timing);
+ arb_regs->mc_arb_dram_timing2 = cpu_to_be32(dram_timing2);
+ arb_regs->mc_arb_burst_time = (u8)burst_time;
+
+ return 0;
+}
+
+static int si_do_program_memory_timing_parameters(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state,
+ unsigned int first_arb_set)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ps *state = si_get_ps(amdgpu_state);
+ SMC_SIslands_MCArbDramTimingRegisterSet arb_regs = { 0 };
+ int i, ret = 0;
+
+ for (i = 0; i < state->performance_level_count; i++) {
+ ret = si_populate_memory_timing_parameters(adev, &state->performance_levels[i], &arb_regs);
+ if (ret)
+ break;
+ ret = amdgpu_si_copy_bytes_to_smc(adev,
+ si_pi->arb_table_start +
+ offsetof(SMC_SIslands_MCArbDramTimingRegisters, data) +
+ sizeof(SMC_SIslands_MCArbDramTimingRegisterSet) * (first_arb_set + i),
+ (u8 *)&arb_regs,
+ sizeof(SMC_SIslands_MCArbDramTimingRegisterSet),
+ si_pi->sram_end);
+ if (ret)
+ break;
+ }
+
+ return ret;
+}
+
+static int si_program_memory_timing_parameters(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state)
+{
+ return si_do_program_memory_timing_parameters(adev, amdgpu_new_state,
+ SISLANDS_DRIVER_STATE_ARB_INDEX);
+}
+
+static int si_populate_initial_mvdd_value(struct amdgpu_device *adev,
+ struct SISLANDS_SMC_VOLTAGE_VALUE *voltage)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ if (pi->mvdd_control)
+ return si_populate_voltage_value(adev, &si_pi->mvdd_voltage_table,
+ si_pi->mvdd_bootup_value, voltage);
+
+ return 0;
+}
+
+static int si_populate_smc_initial_state(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_initial_state,
+ SISLANDS_SMC_STATETABLE *table)
+{
+ struct si_ps *initial_state = si_get_ps(amdgpu_initial_state);
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 reg;
+ int ret;
+
+ table->initialState.levels[0].mclk.vDLL_CNTL =
+ cpu_to_be32(si_pi->clock_registers.dll_cntl);
+ table->initialState.levels[0].mclk.vMCLK_PWRMGT_CNTL =
+ cpu_to_be32(si_pi->clock_registers.mclk_pwrmgt_cntl);
+ table->initialState.levels[0].mclk.vMPLL_AD_FUNC_CNTL =
+ cpu_to_be32(si_pi->clock_registers.mpll_ad_func_cntl);
+ table->initialState.levels[0].mclk.vMPLL_DQ_FUNC_CNTL =
+ cpu_to_be32(si_pi->clock_registers.mpll_dq_func_cntl);
+ table->initialState.levels[0].mclk.vMPLL_FUNC_CNTL =
+ cpu_to_be32(si_pi->clock_registers.mpll_func_cntl);
+ table->initialState.levels[0].mclk.vMPLL_FUNC_CNTL_1 =
+ cpu_to_be32(si_pi->clock_registers.mpll_func_cntl_1);
+ table->initialState.levels[0].mclk.vMPLL_FUNC_CNTL_2 =
+ cpu_to_be32(si_pi->clock_registers.mpll_func_cntl_2);
+ table->initialState.levels[0].mclk.vMPLL_SS =
+ cpu_to_be32(si_pi->clock_registers.mpll_ss1);
+ table->initialState.levels[0].mclk.vMPLL_SS2 =
+ cpu_to_be32(si_pi->clock_registers.mpll_ss2);
+
+ table->initialState.levels[0].mclk.mclk_value =
+ cpu_to_be32(initial_state->performance_levels[0].mclk);
+
+ table->initialState.levels[0].sclk.vCG_SPLL_FUNC_CNTL =
+ cpu_to_be32(si_pi->clock_registers.cg_spll_func_cntl);
+ table->initialState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_2 =
+ cpu_to_be32(si_pi->clock_registers.cg_spll_func_cntl_2);
+ table->initialState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_3 =
+ cpu_to_be32(si_pi->clock_registers.cg_spll_func_cntl_3);
+ table->initialState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_4 =
+ cpu_to_be32(si_pi->clock_registers.cg_spll_func_cntl_4);
+ table->initialState.levels[0].sclk.vCG_SPLL_SPREAD_SPECTRUM =
+ cpu_to_be32(si_pi->clock_registers.cg_spll_spread_spectrum);
+ table->initialState.levels[0].sclk.vCG_SPLL_SPREAD_SPECTRUM_2 =
+ cpu_to_be32(si_pi->clock_registers.cg_spll_spread_spectrum_2);
+
+ table->initialState.levels[0].sclk.sclk_value =
+ cpu_to_be32(initial_state->performance_levels[0].sclk);
+
+ table->initialState.levels[0].arbRefreshState =
+ SISLANDS_INITIAL_STATE_ARB_INDEX;
+
+ table->initialState.levels[0].ACIndex = 0;
+
+ ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table,
+ initial_state->performance_levels[0].vddc,
+ &table->initialState.levels[0].vddc);
+
+ if (!ret) {
+ u16 std_vddc;
+
+ ret = si_get_std_voltage_value(adev,
+ &table->initialState.levels[0].vddc,
+ &std_vddc);
+ if (!ret)
+ si_populate_std_voltage_value(adev, std_vddc,
+ table->initialState.levels[0].vddc.index,
+ &table->initialState.levels[0].std_vddc);
+ }
+
+ if (eg_pi->vddci_control)
+ si_populate_voltage_value(adev,
+ &eg_pi->vddci_voltage_table,
+ initial_state->performance_levels[0].vddci,
+ &table->initialState.levels[0].vddci);
+
+ if (si_pi->vddc_phase_shed_control)
+ si_populate_phase_shedding_value(adev,
+ &adev->pm.dpm.dyn_state.phase_shedding_limits_table,
+ initial_state->performance_levels[0].vddc,
+ initial_state->performance_levels[0].sclk,
+ initial_state->performance_levels[0].mclk,
+ &table->initialState.levels[0].vddc);
+
+ si_populate_initial_mvdd_value(adev, &table->initialState.levels[0].mvdd);
+
+ reg = CG_R(0xffff) | CG_L(0);
+ table->initialState.levels[0].aT = cpu_to_be32(reg);
+ table->initialState.levels[0].bSP = cpu_to_be32(pi->dsp);
+ table->initialState.levels[0].gen2PCIE = (u8)si_pi->boot_pcie_gen;
+
+ if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) {
+ table->initialState.levels[0].strobeMode =
+ si_get_strobe_mode_settings(adev,
+ initial_state->performance_levels[0].mclk);
+
+ if (initial_state->performance_levels[0].mclk > pi->mclk_edc_enable_threshold)
+ table->initialState.levels[0].mcFlags = SISLANDS_SMC_MC_EDC_RD_FLAG | SISLANDS_SMC_MC_EDC_WR_FLAG;
+ else
+ table->initialState.levels[0].mcFlags = 0;
+ }
+
+ table->initialState.levelCount = 1;
+
+ table->initialState.flags |= PPSMC_SWSTATE_FLAG_DC;
+
+ table->initialState.levels[0].dpm2.MaxPS = 0;
+ table->initialState.levels[0].dpm2.NearTDPDec = 0;
+ table->initialState.levels[0].dpm2.AboveSafeInc = 0;
+ table->initialState.levels[0].dpm2.BelowSafeInc = 0;
+ table->initialState.levels[0].dpm2.PwrEfficiencyRatio = 0;
+
+ reg = MIN_POWER_MASK | MAX_POWER_MASK;
+ table->initialState.levels[0].SQPowerThrottle = cpu_to_be32(reg);
+
+ reg = MAX_POWER_DELTA_MASK | STI_SIZE_MASK | LTI_RATIO_MASK;
+ table->initialState.levels[0].SQPowerThrottle_2 = cpu_to_be32(reg);
+
+ return 0;
+}
+
+static int si_populate_smc_acpi_state(struct amdgpu_device *adev,
+ SISLANDS_SMC_STATETABLE *table)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 spll_func_cntl = si_pi->clock_registers.cg_spll_func_cntl;
+ u32 spll_func_cntl_2 = si_pi->clock_registers.cg_spll_func_cntl_2;
+ u32 spll_func_cntl_3 = si_pi->clock_registers.cg_spll_func_cntl_3;
+ u32 spll_func_cntl_4 = si_pi->clock_registers.cg_spll_func_cntl_4;
+ u32 dll_cntl = si_pi->clock_registers.dll_cntl;
+ u32 mclk_pwrmgt_cntl = si_pi->clock_registers.mclk_pwrmgt_cntl;
+ u32 mpll_ad_func_cntl = si_pi->clock_registers.mpll_ad_func_cntl;
+ u32 mpll_dq_func_cntl = si_pi->clock_registers.mpll_dq_func_cntl;
+ u32 mpll_func_cntl = si_pi->clock_registers.mpll_func_cntl;
+ u32 mpll_func_cntl_1 = si_pi->clock_registers.mpll_func_cntl_1;
+ u32 mpll_func_cntl_2 = si_pi->clock_registers.mpll_func_cntl_2;
+ u32 reg;
+ int ret;
+
+ table->ACPIState = table->initialState;
+
+ table->ACPIState.flags &= ~PPSMC_SWSTATE_FLAG_DC;
+
+ if (pi->acpi_vddc) {
+ ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table,
+ pi->acpi_vddc, &table->ACPIState.levels[0].vddc);
+ if (!ret) {
+ u16 std_vddc;
+
+ ret = si_get_std_voltage_value(adev,
+ &table->ACPIState.levels[0].vddc, &std_vddc);
+ if (!ret)
+ si_populate_std_voltage_value(adev, std_vddc,
+ table->ACPIState.levels[0].vddc.index,
+ &table->ACPIState.levels[0].std_vddc);
+ }
+ table->ACPIState.levels[0].gen2PCIE = si_pi->acpi_pcie_gen;
+
+ if (si_pi->vddc_phase_shed_control) {
+ si_populate_phase_shedding_value(adev,
+ &adev->pm.dpm.dyn_state.phase_shedding_limits_table,
+ pi->acpi_vddc,
+ 0,
+ 0,
+ &table->ACPIState.levels[0].vddc);
+ }
+ } else {
+ ret = si_populate_voltage_value(adev, &eg_pi->vddc_voltage_table,
+ pi->min_vddc_in_table, &table->ACPIState.levels[0].vddc);
+ if (!ret) {
+ u16 std_vddc;
+
+ ret = si_get_std_voltage_value(adev,
+ &table->ACPIState.levels[0].vddc, &std_vddc);
+
+ if (!ret)
+ si_populate_std_voltage_value(adev, std_vddc,
+ table->ACPIState.levels[0].vddc.index,
+ &table->ACPIState.levels[0].std_vddc);
+ }
+ table->ACPIState.levels[0].gen2PCIE = (u8)r600_get_pcie_gen_support(adev,
+ si_pi->sys_pcie_mask,
+ si_pi->boot_pcie_gen,
+ AMDGPU_PCIE_GEN1);
+
+ if (si_pi->vddc_phase_shed_control)
+ si_populate_phase_shedding_value(adev,
+ &adev->pm.dpm.dyn_state.phase_shedding_limits_table,
+ pi->min_vddc_in_table,
+ 0,
+ 0,
+ &table->ACPIState.levels[0].vddc);
+ }
+
+ if (pi->acpi_vddc) {
+ if (eg_pi->acpi_vddci)
+ si_populate_voltage_value(adev, &eg_pi->vddci_voltage_table,
+ eg_pi->acpi_vddci,
+ &table->ACPIState.levels[0].vddci);
+ }
+
+ mclk_pwrmgt_cntl |= MRDCK0_RESET | MRDCK1_RESET;
+ mclk_pwrmgt_cntl &= ~(MRDCK0_PDNB | MRDCK1_PDNB);
+
+ dll_cntl &= ~(MRDCK0_BYPASS | MRDCK1_BYPASS);
+
+ spll_func_cntl_2 &= ~SCLK_MUX_SEL_MASK;
+ spll_func_cntl_2 |= SCLK_MUX_SEL(4);
+
+ table->ACPIState.levels[0].mclk.vDLL_CNTL =
+ cpu_to_be32(dll_cntl);
+ table->ACPIState.levels[0].mclk.vMCLK_PWRMGT_CNTL =
+ cpu_to_be32(mclk_pwrmgt_cntl);
+ table->ACPIState.levels[0].mclk.vMPLL_AD_FUNC_CNTL =
+ cpu_to_be32(mpll_ad_func_cntl);
+ table->ACPIState.levels[0].mclk.vMPLL_DQ_FUNC_CNTL =
+ cpu_to_be32(mpll_dq_func_cntl);
+ table->ACPIState.levels[0].mclk.vMPLL_FUNC_CNTL =
+ cpu_to_be32(mpll_func_cntl);
+ table->ACPIState.levels[0].mclk.vMPLL_FUNC_CNTL_1 =
+ cpu_to_be32(mpll_func_cntl_1);
+ table->ACPIState.levels[0].mclk.vMPLL_FUNC_CNTL_2 =
+ cpu_to_be32(mpll_func_cntl_2);
+ table->ACPIState.levels[0].mclk.vMPLL_SS =
+ cpu_to_be32(si_pi->clock_registers.mpll_ss1);
+ table->ACPIState.levels[0].mclk.vMPLL_SS2 =
+ cpu_to_be32(si_pi->clock_registers.mpll_ss2);
+
+ table->ACPIState.levels[0].sclk.vCG_SPLL_FUNC_CNTL =
+ cpu_to_be32(spll_func_cntl);
+ table->ACPIState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_2 =
+ cpu_to_be32(spll_func_cntl_2);
+ table->ACPIState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_3 =
+ cpu_to_be32(spll_func_cntl_3);
+ table->ACPIState.levels[0].sclk.vCG_SPLL_FUNC_CNTL_4 =
+ cpu_to_be32(spll_func_cntl_4);
+
+ table->ACPIState.levels[0].mclk.mclk_value = 0;
+ table->ACPIState.levels[0].sclk.sclk_value = 0;
+
+ si_populate_mvdd_value(adev, 0, &table->ACPIState.levels[0].mvdd);
+
+ if (eg_pi->dynamic_ac_timing)
+ table->ACPIState.levels[0].ACIndex = 0;
+
+ table->ACPIState.levels[0].dpm2.MaxPS = 0;
+ table->ACPIState.levels[0].dpm2.NearTDPDec = 0;
+ table->ACPIState.levels[0].dpm2.AboveSafeInc = 0;
+ table->ACPIState.levels[0].dpm2.BelowSafeInc = 0;
+ table->ACPIState.levels[0].dpm2.PwrEfficiencyRatio = 0;
+
+ reg = MIN_POWER_MASK | MAX_POWER_MASK;
+ table->ACPIState.levels[0].SQPowerThrottle = cpu_to_be32(reg);
+
+ reg = MAX_POWER_DELTA_MASK | STI_SIZE_MASK | LTI_RATIO_MASK;
+ table->ACPIState.levels[0].SQPowerThrottle_2 = cpu_to_be32(reg);
+
+ return 0;
+}
+
+static int si_populate_ulv_state(struct amdgpu_device *adev,
+ SISLANDS_SMC_SWSTATE *state)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ulv_param *ulv = &si_pi->ulv;
+ u32 sclk_in_sr = 1350; /* ??? */
+ int ret;
+
+ ret = si_convert_power_level_to_smc(adev, &ulv->pl,
+ &state->levels[0]);
+ if (!ret) {
+ if (eg_pi->sclk_deep_sleep) {
+ if (sclk_in_sr <= SCLK_MIN_DEEPSLEEP_FREQ)
+ state->levels[0].stateFlags |= PPSMC_STATEFLAG_DEEPSLEEP_BYPASS;
+ else
+ state->levels[0].stateFlags |= PPSMC_STATEFLAG_DEEPSLEEP_THROTTLE;
+ }
+ if (ulv->one_pcie_lane_in_ulv)
+ state->flags |= PPSMC_SWSTATE_FLAG_PCIE_X1;
+ state->levels[0].arbRefreshState = (u8)(SISLANDS_ULV_STATE_ARB_INDEX);
+ state->levels[0].ACIndex = 1;
+ state->levels[0].std_vddc = state->levels[0].vddc;
+ state->levelCount = 1;
+
+ state->flags |= PPSMC_SWSTATE_FLAG_DC;
+ }
+
+ return ret;
+}
+
+static int si_program_ulv_memory_timing_parameters(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ulv_param *ulv = &si_pi->ulv;
+ SMC_SIslands_MCArbDramTimingRegisterSet arb_regs = { 0 };
+ int ret;
+
+ ret = si_populate_memory_timing_parameters(adev, &ulv->pl,
+ &arb_regs);
+ if (ret)
+ return ret;
+
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_ulv_volt_change_delay,
+ ulv->volt_change_delay);
+
+ ret = amdgpu_si_copy_bytes_to_smc(adev,
+ si_pi->arb_table_start +
+ offsetof(SMC_SIslands_MCArbDramTimingRegisters, data) +
+ sizeof(SMC_SIslands_MCArbDramTimingRegisterSet) * SISLANDS_ULV_STATE_ARB_INDEX,
+ (u8 *)&arb_regs,
+ sizeof(SMC_SIslands_MCArbDramTimingRegisterSet),
+ si_pi->sram_end);
+
+ return ret;
+}
+
+static void si_get_mvdd_configuration(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+
+ pi->mvdd_split_frequency = 30000;
+}
+
+static int si_init_smc_table(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct amdgpu_ps *amdgpu_boot_state = adev->pm.dpm.boot_ps;
+ const struct si_ulv_param *ulv = &si_pi->ulv;
+ SISLANDS_SMC_STATETABLE *table = &si_pi->smc_statetable;
+ int ret;
+ u32 lane_width;
+ u32 vr_hot_gpio;
+
+ si_populate_smc_voltage_tables(adev, table);
+
+ switch (adev->pm.int_thermal_type) {
+ case THERMAL_TYPE_SI:
+ case THERMAL_TYPE_EMC2103_WITH_INTERNAL:
+ table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_INTERNAL;
+ break;
+ case THERMAL_TYPE_NONE:
+ table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_NONE;
+ break;
+ default:
+ table->thermalProtectType = PPSMC_THERMAL_PROTECT_TYPE_EXTERNAL;
+ break;
+ }
+
+ if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_HARDWAREDC)
+ table->systemFlags |= PPSMC_SYSTEMFLAG_GPIO_DC;
+
+ if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_REGULATOR_HOT) {
+ if ((adev->pdev->device != 0x6818) && (adev->pdev->device != 0x6819))
+ table->systemFlags |= PPSMC_SYSTEMFLAG_REGULATOR_HOT;
+ }
+
+ if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_STEPVDDC)
+ table->systemFlags |= PPSMC_SYSTEMFLAG_STEPVDDC;
+
+ if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5)
+ table->systemFlags |= PPSMC_SYSTEMFLAG_GDDR5;
+
+ if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_REVERT_GPIO5_POLARITY)
+ table->extraFlags |= PPSMC_EXTRAFLAGS_AC2DC_GPIO5_POLARITY_HIGH;
+
+ if (adev->pm.dpm.platform_caps & ATOM_PP_PLATFORM_CAP_VRHOT_GPIO_CONFIGURABLE) {
+ table->systemFlags |= PPSMC_SYSTEMFLAG_REGULATOR_HOT_PROG_GPIO;
+ vr_hot_gpio = adev->pm.dpm.backbias_response_time;
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_vr_hot_gpio,
+ vr_hot_gpio);
+ }
+
+ ret = si_populate_smc_initial_state(adev, amdgpu_boot_state, table);
+ if (ret)
+ return ret;
+
+ ret = si_populate_smc_acpi_state(adev, table);
+ if (ret)
+ return ret;
+
+ table->driverState = table->initialState;
+
+ ret = si_do_program_memory_timing_parameters(adev, amdgpu_boot_state,
+ SISLANDS_INITIAL_STATE_ARB_INDEX);
+ if (ret)
+ return ret;
+
+ if (ulv->supported && ulv->pl.vddc) {
+ ret = si_populate_ulv_state(adev, &table->ULVState);
+ if (ret)
+ return ret;
+
+ ret = si_program_ulv_memory_timing_parameters(adev);
+ if (ret)
+ return ret;
+
+ WREG32(CG_ULV_CONTROL, ulv->cg_ulv_control);
+ WREG32(CG_ULV_PARAMETER, ulv->cg_ulv_parameter);
+
+ lane_width = amdgpu_get_pcie_lanes(adev);
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_non_ulv_pcie_link_width, lane_width);
+ } else {
+ table->ULVState = table->initialState;
+ }
+
+ return amdgpu_si_copy_bytes_to_smc(adev, si_pi->state_table_start,
+ (u8 *)table, sizeof(SISLANDS_SMC_STATETABLE),
+ si_pi->sram_end);
+}
+
+static int si_calculate_sclk_params(struct amdgpu_device *adev,
+ u32 engine_clock,
+ SISLANDS_SMC_SCLK_VALUE *sclk)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct atom_clock_dividers dividers;
+ u32 spll_func_cntl = si_pi->clock_registers.cg_spll_func_cntl;
+ u32 spll_func_cntl_2 = si_pi->clock_registers.cg_spll_func_cntl_2;
+ u32 spll_func_cntl_3 = si_pi->clock_registers.cg_spll_func_cntl_3;
+ u32 spll_func_cntl_4 = si_pi->clock_registers.cg_spll_func_cntl_4;
+ u32 cg_spll_spread_spectrum = si_pi->clock_registers.cg_spll_spread_spectrum;
+ u32 cg_spll_spread_spectrum_2 = si_pi->clock_registers.cg_spll_spread_spectrum_2;
+ u64 tmp;
+ u32 reference_clock = adev->clock.spll.reference_freq;
+ u32 reference_divider;
+ u32 fbdiv;
+ int ret;
+
+ ret = amdgpu_atombios_get_clock_dividers(adev, COMPUTE_ENGINE_PLL_PARAM,
+ engine_clock, false, ÷rs);
+ if (ret)
+ return ret;
+
+ reference_divider = 1 + dividers.ref_div;
+
+ tmp = (u64) engine_clock * reference_divider * dividers.post_div * 16384;
+ do_div(tmp, reference_clock);
+ fbdiv = (u32) tmp;
+
+ spll_func_cntl &= ~(SPLL_PDIV_A_MASK | SPLL_REF_DIV_MASK);
+ spll_func_cntl |= SPLL_REF_DIV(dividers.ref_div);
+ spll_func_cntl |= SPLL_PDIV_A(dividers.post_div);
+
+ spll_func_cntl_2 &= ~SCLK_MUX_SEL_MASK;
+ spll_func_cntl_2 |= SCLK_MUX_SEL(2);
+
+ spll_func_cntl_3 &= ~SPLL_FB_DIV_MASK;
+ spll_func_cntl_3 |= SPLL_FB_DIV(fbdiv);
+ spll_func_cntl_3 |= SPLL_DITHEN;
+
+ if (pi->sclk_ss) {
+ struct amdgpu_atom_ss ss;
+ u32 vco_freq = engine_clock * dividers.post_div;
+
+ if (amdgpu_atombios_get_asic_ss_info(adev, &ss,
+ ASIC_INTERNAL_ENGINE_SS, vco_freq)) {
+ u32 clk_s = reference_clock * 5 / (reference_divider * ss.rate);
+ u32 clk_v = 4 * ss.percentage * fbdiv / (clk_s * 10000);
+
+ cg_spll_spread_spectrum &= ~CLK_S_MASK;
+ cg_spll_spread_spectrum |= CLK_S(clk_s);
+ cg_spll_spread_spectrum |= SSEN;
+
+ cg_spll_spread_spectrum_2 &= ~CLK_V_MASK;
+ cg_spll_spread_spectrum_2 |= CLK_V(clk_v);
+ }
+ }
+
+ sclk->sclk_value = engine_clock;
+ sclk->vCG_SPLL_FUNC_CNTL = spll_func_cntl;
+ sclk->vCG_SPLL_FUNC_CNTL_2 = spll_func_cntl_2;
+ sclk->vCG_SPLL_FUNC_CNTL_3 = spll_func_cntl_3;
+ sclk->vCG_SPLL_FUNC_CNTL_4 = spll_func_cntl_4;
+ sclk->vCG_SPLL_SPREAD_SPECTRUM = cg_spll_spread_spectrum;
+ sclk->vCG_SPLL_SPREAD_SPECTRUM_2 = cg_spll_spread_spectrum_2;
+
+ return 0;
+}
+
+static int si_populate_sclk_value(struct amdgpu_device *adev,
+ u32 engine_clock,
+ SISLANDS_SMC_SCLK_VALUE *sclk)
+{
+ SISLANDS_SMC_SCLK_VALUE sclk_tmp;
+ int ret;
+
+ ret = si_calculate_sclk_params(adev, engine_clock, &sclk_tmp);
+ if (!ret) {
+ sclk->sclk_value = cpu_to_be32(sclk_tmp.sclk_value);
+ sclk->vCG_SPLL_FUNC_CNTL = cpu_to_be32(sclk_tmp.vCG_SPLL_FUNC_CNTL);
+ sclk->vCG_SPLL_FUNC_CNTL_2 = cpu_to_be32(sclk_tmp.vCG_SPLL_FUNC_CNTL_2);
+ sclk->vCG_SPLL_FUNC_CNTL_3 = cpu_to_be32(sclk_tmp.vCG_SPLL_FUNC_CNTL_3);
+ sclk->vCG_SPLL_FUNC_CNTL_4 = cpu_to_be32(sclk_tmp.vCG_SPLL_FUNC_CNTL_4);
+ sclk->vCG_SPLL_SPREAD_SPECTRUM = cpu_to_be32(sclk_tmp.vCG_SPLL_SPREAD_SPECTRUM);
+ sclk->vCG_SPLL_SPREAD_SPECTRUM_2 = cpu_to_be32(sclk_tmp.vCG_SPLL_SPREAD_SPECTRUM_2);
+ }
+
+ return ret;
+}
+
+static int si_populate_mclk_value(struct amdgpu_device *adev,
+ u32 engine_clock,
+ u32 memory_clock,
+ SISLANDS_SMC_MCLK_VALUE *mclk,
+ bool strobe_mode,
+ bool dll_state_on)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 dll_cntl = si_pi->clock_registers.dll_cntl;
+ u32 mclk_pwrmgt_cntl = si_pi->clock_registers.mclk_pwrmgt_cntl;
+ u32 mpll_ad_func_cntl = si_pi->clock_registers.mpll_ad_func_cntl;
+ u32 mpll_dq_func_cntl = si_pi->clock_registers.mpll_dq_func_cntl;
+ u32 mpll_func_cntl = si_pi->clock_registers.mpll_func_cntl;
+ u32 mpll_func_cntl_1 = si_pi->clock_registers.mpll_func_cntl_1;
+ u32 mpll_func_cntl_2 = si_pi->clock_registers.mpll_func_cntl_2;
+ u32 mpll_ss1 = si_pi->clock_registers.mpll_ss1;
+ u32 mpll_ss2 = si_pi->clock_registers.mpll_ss2;
+ struct atom_mpll_param mpll_param;
+ int ret;
+
+ ret = amdgpu_atombios_get_memory_pll_dividers(adev, memory_clock, strobe_mode, &mpll_param);
+ if (ret)
+ return ret;
+
+ mpll_func_cntl &= ~BWCTRL_MASK;
+ mpll_func_cntl |= BWCTRL(mpll_param.bwcntl);
+
+ mpll_func_cntl_1 &= ~(CLKF_MASK | CLKFRAC_MASK | VCO_MODE_MASK);
+ mpll_func_cntl_1 |= CLKF(mpll_param.clkf) |
+ CLKFRAC(mpll_param.clkfrac) | VCO_MODE(mpll_param.vco_mode);
+
+ mpll_ad_func_cntl &= ~YCLK_POST_DIV_MASK;
+ mpll_ad_func_cntl |= YCLK_POST_DIV(mpll_param.post_div);
+
+ if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) {
+ mpll_dq_func_cntl &= ~(YCLK_SEL_MASK | YCLK_POST_DIV_MASK);
+ mpll_dq_func_cntl |= YCLK_SEL(mpll_param.yclk_sel) |
+ YCLK_POST_DIV(mpll_param.post_div);
+ }
+
+ if (pi->mclk_ss) {
+ struct amdgpu_atom_ss ss;
+ u32 freq_nom;
+ u32 tmp;
+ u32 reference_clock = adev->clock.mpll.reference_freq;
+
+ if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5)
+ freq_nom = memory_clock * 4;
+ else
+ freq_nom = memory_clock * 2;
+
+ tmp = freq_nom / reference_clock;
+ tmp = tmp * tmp;
+ if (amdgpu_atombios_get_asic_ss_info(adev, &ss,
+ ASIC_INTERNAL_MEMORY_SS, freq_nom)) {
+ u32 clks = reference_clock * 5 / ss.rate;
+ u32 clkv = (u32)((((131 * ss.percentage * ss.rate) / 100) * tmp) / freq_nom);
+
+ mpll_ss1 &= ~CLKV_MASK;
+ mpll_ss1 |= CLKV(clkv);
+
+ mpll_ss2 &= ~CLKS_MASK;
+ mpll_ss2 |= CLKS(clks);
+ }
+ }
+
+ mclk_pwrmgt_cntl &= ~DLL_SPEED_MASK;
+ mclk_pwrmgt_cntl |= DLL_SPEED(mpll_param.dll_speed);
+
+ if (dll_state_on)
+ mclk_pwrmgt_cntl |= MRDCK0_PDNB | MRDCK1_PDNB;
+ else
+ mclk_pwrmgt_cntl &= ~(MRDCK0_PDNB | MRDCK1_PDNB);
+
+ mclk->mclk_value = cpu_to_be32(memory_clock);
+ mclk->vMPLL_FUNC_CNTL = cpu_to_be32(mpll_func_cntl);
+ mclk->vMPLL_FUNC_CNTL_1 = cpu_to_be32(mpll_func_cntl_1);
+ mclk->vMPLL_FUNC_CNTL_2 = cpu_to_be32(mpll_func_cntl_2);
+ mclk->vMPLL_AD_FUNC_CNTL = cpu_to_be32(mpll_ad_func_cntl);
+ mclk->vMPLL_DQ_FUNC_CNTL = cpu_to_be32(mpll_dq_func_cntl);
+ mclk->vMCLK_PWRMGT_CNTL = cpu_to_be32(mclk_pwrmgt_cntl);
+ mclk->vDLL_CNTL = cpu_to_be32(dll_cntl);
+ mclk->vMPLL_SS = cpu_to_be32(mpll_ss1);
+ mclk->vMPLL_SS2 = cpu_to_be32(mpll_ss2);
+
+ return 0;
+}
+
+static void si_populate_smc_sp(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state,
+ SISLANDS_SMC_SWSTATE *smc_state)
+{
+ struct si_ps *ps = si_get_ps(amdgpu_state);
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ int i;
+
+ for (i = 0; i < ps->performance_level_count - 1; i++)
+ smc_state->levels[i].bSP = cpu_to_be32(pi->dsp);
+
+ smc_state->levels[ps->performance_level_count - 1].bSP =
+ cpu_to_be32(pi->psp);
+}
+
+static int si_convert_power_level_to_smc(struct amdgpu_device *adev,
+ struct rv7xx_pl *pl,
+ SISLANDS_SMC_HW_PERFORMANCE_LEVEL *level)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ int ret;
+ bool dll_state_on;
+ u16 std_vddc;
+ bool gmc_pg = false;
+
+ if (eg_pi->pcie_performance_request &&
+ (si_pi->force_pcie_gen != AMDGPU_PCIE_GEN_INVALID))
+ level->gen2PCIE = (u8)si_pi->force_pcie_gen;
+ else
+ level->gen2PCIE = (u8)pl->pcie_gen;
+
+ ret = si_populate_sclk_value(adev, pl->sclk, &level->sclk);
+ if (ret)
+ return ret;
+
+ level->mcFlags = 0;
+
+ if (pi->mclk_stutter_mode_threshold &&
+ (pl->mclk <= pi->mclk_stutter_mode_threshold) &&
+ !eg_pi->uvd_enabled &&
+ (RREG32(DPG_PIPE_STUTTER_CONTROL) & STUTTER_ENABLE) &&
+ (adev->pm.dpm.new_active_crtc_count <= 2)) {
+ level->mcFlags |= SISLANDS_SMC_MC_STUTTER_EN;
+
+ if (gmc_pg)
+ level->mcFlags |= SISLANDS_SMC_MC_PG_EN;
+ }
+
+ if (adev->mc.vram_type == AMDGPU_VRAM_TYPE_GDDR5) {
+ if (pl->mclk > pi->mclk_edc_enable_threshold)
+ level->mcFlags |= SISLANDS_SMC_MC_EDC_RD_FLAG;
+
+ if (pl->mclk > eg_pi->mclk_edc_wr_enable_threshold)
+ level->mcFlags |= SISLANDS_SMC_MC_EDC_WR_FLAG;
+
+ level->strobeMode = si_get_strobe_mode_settings(adev, pl->mclk);
+
+ if (level->strobeMode & SISLANDS_SMC_STROBE_ENABLE) {
+ if (si_get_mclk_frequency_ratio(pl->mclk, true) >=
+ ((RREG32(MC_SEQ_MISC7) >> 16) & 0xf))
+ dll_state_on = ((RREG32(MC_SEQ_MISC5) >> 1) & 0x1) ? true : false;
+ else
+ dll_state_on = ((RREG32(MC_SEQ_MISC6) >> 1) & 0x1) ? true : false;
+ } else {
+ dll_state_on = false;
+ }
+ } else {
+ level->strobeMode = si_get_strobe_mode_settings(adev,
+ pl->mclk);
+
+ dll_state_on = ((RREG32(MC_SEQ_MISC5) >> 1) & 0x1) ? true : false;
+ }
+
+ ret = si_populate_mclk_value(adev,
+ pl->sclk,
+ pl->mclk,
+ &level->mclk,
+ (level->strobeMode & SISLANDS_SMC_STROBE_ENABLE) != 0, dll_state_on);
+ if (ret)
+ return ret;
+
+ ret = si_populate_voltage_value(adev,
+ &eg_pi->vddc_voltage_table,
+ pl->vddc, &level->vddc);
+ if (ret)
+ return ret;
+
+
+ ret = si_get_std_voltage_value(adev, &level->vddc, &std_vddc);
+ if (ret)
+ return ret;
+
+ ret = si_populate_std_voltage_value(adev, std_vddc,
+ level->vddc.index, &level->std_vddc);
+ if (ret)
+ return ret;
+
+ if (eg_pi->vddci_control) {
+ ret = si_populate_voltage_value(adev, &eg_pi->vddci_voltage_table,
+ pl->vddci, &level->vddci);
+ if (ret)
+ return ret;
+ }
+
+ if (si_pi->vddc_phase_shed_control) {
+ ret = si_populate_phase_shedding_value(adev,
+ &adev->pm.dpm.dyn_state.phase_shedding_limits_table,
+ pl->vddc,
+ pl->sclk,
+ pl->mclk,
+ &level->vddc);
+ if (ret)
+ return ret;
+ }
+
+ level->MaxPoweredUpCU = si_pi->max_cu;
+
+ ret = si_populate_mvdd_value(adev, pl->mclk, &level->mvdd);
+
+ return ret;
+}
+
+static int si_populate_smc_t(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state,
+ SISLANDS_SMC_SWSTATE *smc_state)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct si_ps *state = si_get_ps(amdgpu_state);
+ u32 a_t;
+ u32 t_l, t_h;
+ u32 high_bsp;
+ int i, ret;
+
+ if (state->performance_level_count >= 9)
+ return -EINVAL;
+
+ if (state->performance_level_count < 2) {
+ a_t = CG_R(0xffff) | CG_L(0);
+ smc_state->levels[0].aT = cpu_to_be32(a_t);
+ return 0;
+ }
+
+ smc_state->levels[0].aT = cpu_to_be32(0);
+
+ for (i = 0; i <= state->performance_level_count - 2; i++) {
+ ret = r600_calculate_at(
+ (50 / SISLANDS_MAX_HARDWARE_POWERLEVELS) * 100 * (i + 1),
+ 100 * R600_AH_DFLT,
+ state->performance_levels[i + 1].sclk,
+ state->performance_levels[i].sclk,
+ &t_l,
+ &t_h);
+
+ if (ret) {
+ t_h = (i + 1) * 1000 - 50 * R600_AH_DFLT;
+ t_l = (i + 1) * 1000 + 50 * R600_AH_DFLT;
+ }
+
+ a_t = be32_to_cpu(smc_state->levels[i].aT) & ~CG_R_MASK;
+ a_t |= CG_R(t_l * pi->bsp / 20000);
+ smc_state->levels[i].aT = cpu_to_be32(a_t);
+
+ high_bsp = (i == state->performance_level_count - 2) ?
+ pi->pbsp : pi->bsp;
+ a_t = CG_R(0xffff) | CG_L(t_h * high_bsp / 20000);
+ smc_state->levels[i + 1].aT = cpu_to_be32(a_t);
+ }
+
+ return 0;
+}
+
+static int si_disable_ulv(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ulv_param *ulv = &si_pi->ulv;
+
+ if (ulv->supported)
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_DisableULV) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+
+ return 0;
+}
+
+static bool si_is_state_ulv_compatible(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state)
+{
+ const struct si_power_info *si_pi = si_get_pi(adev);
+ const struct si_ulv_param *ulv = &si_pi->ulv;
+ const struct si_ps *state = si_get_ps(amdgpu_state);
+ int i;
+
+ if (state->performance_levels[0].mclk != ulv->pl.mclk)
+ return false;
+
+ /* XXX validate against display requirements! */
+
+ for (i = 0; i < adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.count; i++) {
+ if (adev->clock.current_dispclk <=
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[i].clk) {
+ if (ulv->pl.vddc <
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[i].v)
+ return false;
+ }
+ }
+
+ if ((amdgpu_state->vclk != 0) || (amdgpu_state->dclk != 0))
+ return false;
+
+ return true;
+}
+
+static int si_set_power_state_conditionally_enable_ulv(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state)
+{
+ const struct si_power_info *si_pi = si_get_pi(adev);
+ const struct si_ulv_param *ulv = &si_pi->ulv;
+
+ if (ulv->supported) {
+ if (si_is_state_ulv_compatible(adev, amdgpu_new_state))
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableULV) == PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+ }
+ return 0;
+}
+
+static int si_convert_power_state_to_smc(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state,
+ SISLANDS_SMC_SWSTATE *smc_state)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct ni_power_info *ni_pi = ni_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ps *state = si_get_ps(amdgpu_state);
+ int i, ret;
+ u32 threshold;
+ u32 sclk_in_sr = 1350; /* ??? */
+
+ if (state->performance_level_count > SISLANDS_MAX_HARDWARE_POWERLEVELS)
+ return -EINVAL;
+
+ threshold = state->performance_levels[state->performance_level_count-1].sclk * 100 / 100;
+
+ if (amdgpu_state->vclk && amdgpu_state->dclk) {
+ eg_pi->uvd_enabled = true;
+ if (eg_pi->smu_uvd_hs)
+ smc_state->flags |= PPSMC_SWSTATE_FLAG_UVD;
+ } else {
+ eg_pi->uvd_enabled = false;
+ }
+
+ if (state->dc_compatible)
+ smc_state->flags |= PPSMC_SWSTATE_FLAG_DC;
+
+ smc_state->levelCount = 0;
+ for (i = 0; i < state->performance_level_count; i++) {
+ if (eg_pi->sclk_deep_sleep) {
+ if ((i == 0) || si_pi->sclk_deep_sleep_above_low) {
+ if (sclk_in_sr <= SCLK_MIN_DEEPSLEEP_FREQ)
+ smc_state->levels[i].stateFlags |= PPSMC_STATEFLAG_DEEPSLEEP_BYPASS;
+ else
+ smc_state->levels[i].stateFlags |= PPSMC_STATEFLAG_DEEPSLEEP_THROTTLE;
+ }
+ }
+
+ ret = si_convert_power_level_to_smc(adev, &state->performance_levels[i],
+ &smc_state->levels[i]);
+ smc_state->levels[i].arbRefreshState =
+ (u8)(SISLANDS_DRIVER_STATE_ARB_INDEX + i);
+
+ if (ret)
+ return ret;
+
+ if (ni_pi->enable_power_containment)
+ smc_state->levels[i].displayWatermark =
+ (state->performance_levels[i].sclk < threshold) ?
+ PPSMC_DISPLAY_WATERMARK_LOW : PPSMC_DISPLAY_WATERMARK_HIGH;
+ else
+ smc_state->levels[i].displayWatermark = (i < 2) ?
+ PPSMC_DISPLAY_WATERMARK_LOW : PPSMC_DISPLAY_WATERMARK_HIGH;
+
+ if (eg_pi->dynamic_ac_timing)
+ smc_state->levels[i].ACIndex = SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT + i;
+ else
+ smc_state->levels[i].ACIndex = 0;
+
+ smc_state->levelCount++;
+ }
+
+ si_write_smc_soft_register(adev,
+ SI_SMC_SOFT_REGISTER_watermark_threshold,
+ threshold / 512);
+
+ si_populate_smc_sp(adev, amdgpu_state, smc_state);
+
+ ret = si_populate_power_containment_values(adev, amdgpu_state, smc_state);
+ if (ret)
+ ni_pi->enable_power_containment = false;
+
+ ret = si_populate_sq_ramping_values(adev, amdgpu_state, smc_state);
+ if (ret)
+ ni_pi->enable_sq_ramping = false;
+
+ return si_populate_smc_t(adev, amdgpu_state, smc_state);
+}
+
+static int si_upload_sw_state(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ps *new_state = si_get_ps(amdgpu_new_state);
+ int ret;
+ u32 address = si_pi->state_table_start +
+ offsetof(SISLANDS_SMC_STATETABLE, driverState);
+ u32 state_size = sizeof(SISLANDS_SMC_SWSTATE) +
+ ((new_state->performance_level_count - 1) *
+ sizeof(SISLANDS_SMC_HW_PERFORMANCE_LEVEL));
+ SISLANDS_SMC_SWSTATE *smc_state = &si_pi->smc_statetable.driverState;
+
+ memset(smc_state, 0, state_size);
+
+ ret = si_convert_power_state_to_smc(adev, amdgpu_new_state, smc_state);
+ if (ret)
+ return ret;
+
+ return amdgpu_si_copy_bytes_to_smc(adev, address, (u8 *)smc_state,
+ state_size, si_pi->sram_end);
+}
+
+static int si_upload_ulv_state(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ulv_param *ulv = &si_pi->ulv;
+ int ret = 0;
+
+ if (ulv->supported && ulv->pl.vddc) {
+ u32 address = si_pi->state_table_start +
+ offsetof(SISLANDS_SMC_STATETABLE, ULVState);
+ SISLANDS_SMC_SWSTATE *smc_state = &si_pi->smc_statetable.ULVState;
+ u32 state_size = sizeof(SISLANDS_SMC_SWSTATE);
+
+ memset(smc_state, 0, state_size);
+
+ ret = si_populate_ulv_state(adev, smc_state);
+ if (!ret)
+ ret = amdgpu_si_copy_bytes_to_smc(adev, address, (u8 *)smc_state,
+ state_size, si_pi->sram_end);
+ }
+
+ return ret;
+}
+
+static int si_upload_smc_data(struct amdgpu_device *adev)
+{
+ struct amdgpu_crtc *amdgpu_crtc = NULL;
+ int i;
+
+ if (adev->pm.dpm.new_active_crtc_count == 0)
+ return 0;
+
+ for (i = 0; i < adev->mode_info.num_crtc; i++) {
+ if (adev->pm.dpm.new_active_crtcs & (1 << i)) {
+ amdgpu_crtc = adev->mode_info.crtcs[i];
+ break;
+ }
+ }
+
+ if (amdgpu_crtc == NULL)
+ return 0;
+
+ if (amdgpu_crtc->line_time <= 0)
+ return 0;
+
+ if (si_write_smc_soft_register(adev,
+ SI_SMC_SOFT_REGISTER_crtc_index,
+ amdgpu_crtc->crtc_id) != PPSMC_Result_OK)
+ return 0;
+
+ if (si_write_smc_soft_register(adev,
+ SI_SMC_SOFT_REGISTER_mclk_change_block_cp_min,
+ amdgpu_crtc->wm_high / amdgpu_crtc->line_time) != PPSMC_Result_OK)
+ return 0;
+
+ if (si_write_smc_soft_register(adev,
+ SI_SMC_SOFT_REGISTER_mclk_change_block_cp_max,
+ amdgpu_crtc->wm_low / amdgpu_crtc->line_time) != PPSMC_Result_OK)
+ return 0;
+
+ return 0;
+}
+
+static int si_set_mc_special_registers(struct amdgpu_device *adev,
+ struct si_mc_reg_table *table)
+{
+ u8 i, j, k;
+ u32 temp_reg;
+
+ for (i = 0, j = table->last; i < table->last; i++) {
+ if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
+ return -EINVAL;
+ switch (table->mc_reg_address[i].s1) {
+ case MC_SEQ_MISC1:
+ temp_reg = RREG32(MC_PMG_CMD_EMRS);
+ table->mc_reg_address[j].s1 = MC_PMG_CMD_EMRS;
+ table->mc_reg_address[j].s0 = MC_SEQ_PMG_CMD_EMRS_LP;
+ for (k = 0; k < table->num_entries; k++)
+ table->mc_reg_table_entry[k].mc_data[j] =
+ ((temp_reg & 0xffff0000)) |
+ ((table->mc_reg_table_entry[k].mc_data[i] & 0xffff0000) >> 16);
+ j++;
+ if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
+ return -EINVAL;
+
+ temp_reg = RREG32(MC_PMG_CMD_MRS);
+ table->mc_reg_address[j].s1 = MC_PMG_CMD_MRS;
+ table->mc_reg_address[j].s0 = MC_SEQ_PMG_CMD_MRS_LP;
+ for (k = 0; k < table->num_entries; k++) {
+ table->mc_reg_table_entry[k].mc_data[j] =
+ (temp_reg & 0xffff0000) |
+ (table->mc_reg_table_entry[k].mc_data[i] & 0x0000ffff);
+ if (adev->mc.vram_type != AMDGPU_VRAM_TYPE_GDDR5)
+ table->mc_reg_table_entry[k].mc_data[j] |= 0x100;
+ }
+ j++;
+ if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
+ return -EINVAL;
+
+ if (adev->mc.vram_type != AMDGPU_VRAM_TYPE_GDDR5) {
+ table->mc_reg_address[j].s1 = MC_PMG_AUTO_CMD;
+ table->mc_reg_address[j].s0 = MC_PMG_AUTO_CMD;
+ for (k = 0; k < table->num_entries; k++)
+ table->mc_reg_table_entry[k].mc_data[j] =
+ (table->mc_reg_table_entry[k].mc_data[i] & 0xffff0000) >> 16;
+ j++;
+ if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
+ return -EINVAL;
+ }
+ break;
+ case MC_SEQ_RESERVE_M:
+ temp_reg = RREG32(MC_PMG_CMD_MRS1);
+ table->mc_reg_address[j].s1 = MC_PMG_CMD_MRS1;
+ table->mc_reg_address[j].s0 = MC_SEQ_PMG_CMD_MRS1_LP;
+ for(k = 0; k < table->num_entries; k++)
+ table->mc_reg_table_entry[k].mc_data[j] =
+ (temp_reg & 0xffff0000) |
+ (table->mc_reg_table_entry[k].mc_data[i] & 0x0000ffff);
+ j++;
+ if (j >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
+ return -EINVAL;
+ break;
+ default:
+ break;
+ }
+ }
+
+ table->last = j;
+
+ return 0;
+}
+
+static bool si_check_s0_mc_reg_index(u16 in_reg, u16 *out_reg)
+{
+ bool result = true;
+ switch (in_reg) {
+ case MC_SEQ_RAS_TIMING:
+ *out_reg = MC_SEQ_RAS_TIMING_LP;
+ break;
+ case MC_SEQ_CAS_TIMING:
+ *out_reg = MC_SEQ_CAS_TIMING_LP;
+ break;
+ case MC_SEQ_MISC_TIMING:
+ *out_reg = MC_SEQ_MISC_TIMING_LP;
+ break;
+ case MC_SEQ_MISC_TIMING2:
+ *out_reg = MC_SEQ_MISC_TIMING2_LP;
+ break;
+ case MC_SEQ_RD_CTL_D0:
+ *out_reg = MC_SEQ_RD_CTL_D0_LP;
+ break;
+ case MC_SEQ_RD_CTL_D1:
+ *out_reg = MC_SEQ_RD_CTL_D1_LP;
+ break;
+ case MC_SEQ_WR_CTL_D0:
+ *out_reg = MC_SEQ_WR_CTL_D0_LP;
+ break;
+ case MC_SEQ_WR_CTL_D1:
+ *out_reg = MC_SEQ_WR_CTL_D1_LP;
+ break;
+ case MC_PMG_CMD_EMRS:
+ *out_reg = MC_SEQ_PMG_CMD_EMRS_LP;
+ break;
+ case MC_PMG_CMD_MRS:
+ *out_reg = MC_SEQ_PMG_CMD_MRS_LP;
+ break;
+ case MC_PMG_CMD_MRS1:
+ *out_reg = MC_SEQ_PMG_CMD_MRS1_LP;
+ break;
+ case MC_SEQ_PMG_TIMING:
+ *out_reg = MC_SEQ_PMG_TIMING_LP;
+ break;
+ case MC_PMG_CMD_MRS2:
+ *out_reg = MC_SEQ_PMG_CMD_MRS2_LP;
+ break;
+ case MC_SEQ_WR_CTL_2:
+ *out_reg = MC_SEQ_WR_CTL_2_LP;
+ break;
+ default:
+ result = false;
+ break;
+ }
+
+ return result;
+}
+
+static void si_set_valid_flag(struct si_mc_reg_table *table)
+{
+ u8 i, j;
+
+ for (i = 0; i < table->last; i++) {
+ for (j = 1; j < table->num_entries; j++) {
+ if (table->mc_reg_table_entry[j-1].mc_data[i] != table->mc_reg_table_entry[j].mc_data[i]) {
+ table->valid_flag |= 1 << i;
+ break;
+ }
+ }
+ }
+}
+
+static void si_set_s0_mc_reg_index(struct si_mc_reg_table *table)
+{
+ u32 i;
+ u16 address;
+
+ for (i = 0; i < table->last; i++)
+ table->mc_reg_address[i].s0 = si_check_s0_mc_reg_index(table->mc_reg_address[i].s1, &address) ?
+ address : table->mc_reg_address[i].s1;
+
+}
+
+static int si_copy_vbios_mc_reg_table(struct atom_mc_reg_table *table,
+ struct si_mc_reg_table *si_table)
+{
+ u8 i, j;
+
+ if (table->last > SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
+ return -EINVAL;
+ if (table->num_entries > MAX_AC_TIMING_ENTRIES)
+ return -EINVAL;
+
+ for (i = 0; i < table->last; i++)
+ si_table->mc_reg_address[i].s1 = table->mc_reg_address[i].s1;
+ si_table->last = table->last;
+
+ for (i = 0; i < table->num_entries; i++) {
+ si_table->mc_reg_table_entry[i].mclk_max =
+ table->mc_reg_table_entry[i].mclk_max;
+ for (j = 0; j < table->last; j++) {
+ si_table->mc_reg_table_entry[i].mc_data[j] =
+ table->mc_reg_table_entry[i].mc_data[j];
+ }
+ }
+ si_table->num_entries = table->num_entries;
+
+ return 0;
+}
+
+static int si_initialize_mc_reg_table(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct atom_mc_reg_table *table;
+ struct si_mc_reg_table *si_table = &si_pi->mc_reg_table;
+ u8 module_index = rv770_get_memory_module_index(adev);
+ int ret;
+
+ table = kzalloc(sizeof(struct atom_mc_reg_table), GFP_KERNEL);
+ if (!table)
+ return -ENOMEM;
+
+ WREG32(MC_SEQ_RAS_TIMING_LP, RREG32(MC_SEQ_RAS_TIMING));
+ WREG32(MC_SEQ_CAS_TIMING_LP, RREG32(MC_SEQ_CAS_TIMING));
+ WREG32(MC_SEQ_MISC_TIMING_LP, RREG32(MC_SEQ_MISC_TIMING));
+ WREG32(MC_SEQ_MISC_TIMING2_LP, RREG32(MC_SEQ_MISC_TIMING2));
+ WREG32(MC_SEQ_PMG_CMD_EMRS_LP, RREG32(MC_PMG_CMD_EMRS));
+ WREG32(MC_SEQ_PMG_CMD_MRS_LP, RREG32(MC_PMG_CMD_MRS));
+ WREG32(MC_SEQ_PMG_CMD_MRS1_LP, RREG32(MC_PMG_CMD_MRS1));
+ WREG32(MC_SEQ_WR_CTL_D0_LP, RREG32(MC_SEQ_WR_CTL_D0));
+ WREG32(MC_SEQ_WR_CTL_D1_LP, RREG32(MC_SEQ_WR_CTL_D1));
+ WREG32(MC_SEQ_RD_CTL_D0_LP, RREG32(MC_SEQ_RD_CTL_D0));
+ WREG32(MC_SEQ_RD_CTL_D1_LP, RREG32(MC_SEQ_RD_CTL_D1));
+ WREG32(MC_SEQ_PMG_TIMING_LP, RREG32(MC_SEQ_PMG_TIMING));
+ WREG32(MC_SEQ_PMG_CMD_MRS2_LP, RREG32(MC_PMG_CMD_MRS2));
+ WREG32(MC_SEQ_WR_CTL_2_LP, RREG32(MC_SEQ_WR_CTL_2));
+
+ ret = amdgpu_atombios_init_mc_reg_table(adev, module_index, table);
+ if (ret)
+ goto init_mc_done;
+
+ ret = si_copy_vbios_mc_reg_table(table, si_table);
+ if (ret)
+ goto init_mc_done;
+
+ si_set_s0_mc_reg_index(si_table);
+
+ ret = si_set_mc_special_registers(adev, si_table);
+ if (ret)
+ goto init_mc_done;
+
+ si_set_valid_flag(si_table);
+
+init_mc_done:
+ kfree(table);
+
+ return ret;
+
+}
+
+static void si_populate_mc_reg_addresses(struct amdgpu_device *adev,
+ SMC_SIslands_MCRegisters *mc_reg_table)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 i, j;
+
+ for (i = 0, j = 0; j < si_pi->mc_reg_table.last; j++) {
+ if (si_pi->mc_reg_table.valid_flag & (1 << j)) {
+ if (i >= SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE)
+ break;
+ mc_reg_table->address[i].s0 =
+ cpu_to_be16(si_pi->mc_reg_table.mc_reg_address[j].s0);
+ mc_reg_table->address[i].s1 =
+ cpu_to_be16(si_pi->mc_reg_table.mc_reg_address[j].s1);
+ i++;
+ }
+ }
+ mc_reg_table->last = (u8)i;
+}
+
+static void si_convert_mc_registers(const struct si_mc_reg_entry *entry,
+ SMC_SIslands_MCRegisterSet *data,
+ u32 num_entries, u32 valid_flag)
+{
+ u32 i, j;
+
+ for(i = 0, j = 0; j < num_entries; j++) {
+ if (valid_flag & (1 << j)) {
+ data->value[i] = cpu_to_be32(entry->mc_data[j]);
+ i++;
+ }
+ }
+}
+
+static void si_convert_mc_reg_table_entry_to_smc(struct amdgpu_device *adev,
+ struct rv7xx_pl *pl,
+ SMC_SIslands_MCRegisterSet *mc_reg_table_data)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 i = 0;
+
+ for (i = 0; i < si_pi->mc_reg_table.num_entries; i++) {
+ if (pl->mclk <= si_pi->mc_reg_table.mc_reg_table_entry[i].mclk_max)
+ break;
+ }
+
+ if ((i == si_pi->mc_reg_table.num_entries) && (i > 0))
+ --i;
+
+ si_convert_mc_registers(&si_pi->mc_reg_table.mc_reg_table_entry[i],
+ mc_reg_table_data, si_pi->mc_reg_table.last,
+ si_pi->mc_reg_table.valid_flag);
+}
+
+static void si_convert_mc_reg_table_to_smc(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state,
+ SMC_SIslands_MCRegisters *mc_reg_table)
+{
+ struct si_ps *state = si_get_ps(amdgpu_state);
+ int i;
+
+ for (i = 0; i < state->performance_level_count; i++) {
+ si_convert_mc_reg_table_entry_to_smc(adev,
+ &state->performance_levels[i],
+ &mc_reg_table->data[SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT + i]);
+ }
+}
+
+static int si_populate_mc_reg_table(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_boot_state)
+{
+ struct si_ps *boot_state = si_get_ps(amdgpu_boot_state);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ulv_param *ulv = &si_pi->ulv;
+ SMC_SIslands_MCRegisters *smc_mc_reg_table = &si_pi->smc_mc_reg_table;
+
+ memset(smc_mc_reg_table, 0, sizeof(SMC_SIslands_MCRegisters));
+
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_seq_index, 1);
+
+ si_populate_mc_reg_addresses(adev, smc_mc_reg_table);
+
+ si_convert_mc_reg_table_entry_to_smc(adev, &boot_state->performance_levels[0],
+ &smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_INITIAL_SLOT]);
+
+ si_convert_mc_registers(&si_pi->mc_reg_table.mc_reg_table_entry[0],
+ &smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_ACPI_SLOT],
+ si_pi->mc_reg_table.last,
+ si_pi->mc_reg_table.valid_flag);
+
+ if (ulv->supported && ulv->pl.vddc != 0)
+ si_convert_mc_reg_table_entry_to_smc(adev, &ulv->pl,
+ &smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_ULV_SLOT]);
+ else
+ si_convert_mc_registers(&si_pi->mc_reg_table.mc_reg_table_entry[0],
+ &smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_ULV_SLOT],
+ si_pi->mc_reg_table.last,
+ si_pi->mc_reg_table.valid_flag);
+
+ si_convert_mc_reg_table_to_smc(adev, amdgpu_boot_state, smc_mc_reg_table);
+
+ return amdgpu_si_copy_bytes_to_smc(adev, si_pi->mc_reg_table_start,
+ (u8 *)smc_mc_reg_table,
+ sizeof(SMC_SIslands_MCRegisters), si_pi->sram_end);
+}
+
+static int si_upload_mc_reg_table(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state)
+{
+ struct si_ps *new_state = si_get_ps(amdgpu_new_state);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 address = si_pi->mc_reg_table_start +
+ offsetof(SMC_SIslands_MCRegisters,
+ data[SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT]);
+ SMC_SIslands_MCRegisters *smc_mc_reg_table = &si_pi->smc_mc_reg_table;
+
+ memset(smc_mc_reg_table, 0, sizeof(SMC_SIslands_MCRegisters));
+
+ si_convert_mc_reg_table_to_smc(adev, amdgpu_new_state, smc_mc_reg_table);
+
+ return amdgpu_si_copy_bytes_to_smc(adev, address,
+ (u8 *)&smc_mc_reg_table->data[SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT],
+ sizeof(SMC_SIslands_MCRegisterSet) * new_state->performance_level_count,
+ si_pi->sram_end);
+}
+
+static void si_enable_voltage_control(struct amdgpu_device *adev, bool enable)
+{
+ if (enable)
+ WREG32_P(GENERAL_PWRMGT, VOLT_PWRMGT_EN, ~VOLT_PWRMGT_EN);
+ else
+ WREG32_P(GENERAL_PWRMGT, 0, ~VOLT_PWRMGT_EN);
+}
+
+static enum amdgpu_pcie_gen si_get_maximum_link_speed(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_state)
+{
+ struct si_ps *state = si_get_ps(amdgpu_state);
+ int i;
+ u16 pcie_speed, max_speed = 0;
+
+ for (i = 0; i < state->performance_level_count; i++) {
+ pcie_speed = state->performance_levels[i].pcie_gen;
+ if (max_speed < pcie_speed)
+ max_speed = pcie_speed;
+ }
+ return max_speed;
+}
+
+static u16 si_get_current_pcie_speed(struct amdgpu_device *adev)
+{
+ u32 speed_cntl;
+
+ speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL) & LC_CURRENT_DATA_RATE_MASK;
+ speed_cntl >>= LC_CURRENT_DATA_RATE_SHIFT;
+
+ return (u16)speed_cntl;
+}
+
+static void si_request_link_speed_change_before_state_change(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state,
+ struct amdgpu_ps *amdgpu_current_state)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ enum amdgpu_pcie_gen target_link_speed = si_get_maximum_link_speed(adev, amdgpu_new_state);
+ enum amdgpu_pcie_gen current_link_speed;
+
+ if (si_pi->force_pcie_gen == AMDGPU_PCIE_GEN_INVALID)
+ current_link_speed = si_get_maximum_link_speed(adev, amdgpu_current_state);
+ else
+ current_link_speed = si_pi->force_pcie_gen;
+
+ si_pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID;
+ si_pi->pspp_notify_required = false;
+ if (target_link_speed > current_link_speed) {
+ switch (target_link_speed) {
+#if defined(CONFIG_ACPI)
+ case AMDGPU_PCIE_GEN3:
+ if (amdgpu_acpi_pcie_performance_request(adev, PCIE_PERF_REQ_PECI_GEN3, false) == 0)
+ break;
+ si_pi->force_pcie_gen = AMDGPU_PCIE_GEN2;
+ if (current_link_speed == AMDGPU_PCIE_GEN2)
+ break;
+ case AMDGPU_PCIE_GEN2:
+ if (amdgpu_acpi_pcie_performance_request(adev, PCIE_PERF_REQ_PECI_GEN2, false) == 0)
+ break;
+#endif
+ default:
+ si_pi->force_pcie_gen = si_get_current_pcie_speed(adev);
+ break;
+ }
+ } else {
+ if (target_link_speed < current_link_speed)
+ si_pi->pspp_notify_required = true;
+ }
+}
+
+static void si_notify_link_speed_change_after_state_change(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state,
+ struct amdgpu_ps *amdgpu_current_state)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ enum amdgpu_pcie_gen target_link_speed = si_get_maximum_link_speed(adev, amdgpu_new_state);
+ u8 request;
+
+ if (si_pi->pspp_notify_required) {
+ if (target_link_speed == AMDGPU_PCIE_GEN3)
+ request = PCIE_PERF_REQ_PECI_GEN3;
+ else if (target_link_speed == AMDGPU_PCIE_GEN2)
+ request = PCIE_PERF_REQ_PECI_GEN2;
+ else
+ request = PCIE_PERF_REQ_PECI_GEN1;
+
+ if ((request == PCIE_PERF_REQ_PECI_GEN1) &&
+ (si_get_current_pcie_speed(adev) > 0))
+ return;
+
+#if defined(CONFIG_ACPI)
+ amdgpu_acpi_pcie_performance_request(adev, request, false);
+#endif
+ }
+}
+
+#if 0
+static int si_ds_request(struct amdgpu_device *adev,
+ bool ds_status_on, u32 count_write)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+
+ if (eg_pi->sclk_deep_sleep) {
+ if (ds_status_on)
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_CancelThrottleOVRDSCLKDS) ==
+ PPSMC_Result_OK) ?
+ 0 : -EINVAL;
+ else
+ return (amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_ThrottleOVRDSCLKDS) ==
+ PPSMC_Result_OK) ? 0 : -EINVAL;
+ }
+ return 0;
+}
+#endif
+
+static void si_set_max_cu_value(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+
+ if (adev->asic_type == CHIP_VERDE) {
+ switch (adev->pdev->device) {
+ case 0x6820:
+ case 0x6825:
+ case 0x6821:
+ case 0x6823:
+ case 0x6827:
+ si_pi->max_cu = 10;
+ break;
+ case 0x682D:
+ case 0x6824:
+ case 0x682F:
+ case 0x6826:
+ si_pi->max_cu = 8;
+ break;
+ case 0x6828:
+ case 0x6830:
+ case 0x6831:
+ case 0x6838:
+ case 0x6839:
+ case 0x683D:
+ si_pi->max_cu = 10;
+ break;
+ case 0x683B:
+ case 0x683F:
+ case 0x6829:
+ si_pi->max_cu = 8;
+ break;
+ default:
+ si_pi->max_cu = 0;
+ break;
+ }
+ } else {
+ si_pi->max_cu = 0;
+ }
+}
+
+static int si_patch_single_dependency_table_based_on_leakage(struct amdgpu_device *adev,
+ struct amdgpu_clock_voltage_dependency_table *table)
+{
+ u32 i;
+ int j;
+ u16 leakage_voltage;
+
+ if (table) {
+ for (i = 0; i < table->count; i++) {
+ switch (si_get_leakage_voltage_from_leakage_index(adev,
+ table->entries[i].v,
+ &leakage_voltage)) {
+ case 0:
+ table->entries[i].v = leakage_voltage;
+ break;
+ case -EAGAIN:
+ return -EINVAL;
+ case -EINVAL:
+ default:
+ break;
+ }
+ }
+
+ for (j = (table->count - 2); j >= 0; j--) {
+ table->entries[j].v = (table->entries[j].v <= table->entries[j + 1].v) ?
+ table->entries[j].v : table->entries[j + 1].v;
+ }
+ }
+ return 0;
+}
+
+static int si_patch_dependency_tables_based_on_leakage(struct amdgpu_device *adev)
+{
+ int ret = 0;
+
+ ret = si_patch_single_dependency_table_based_on_leakage(adev,
+ &adev->pm.dpm.dyn_state.vddc_dependency_on_sclk);
+ if (ret)
+ DRM_ERROR("Could not patch vddc_on_sclk leakage table\n");
+ ret = si_patch_single_dependency_table_based_on_leakage(adev,
+ &adev->pm.dpm.dyn_state.vddc_dependency_on_mclk);
+ if (ret)
+ DRM_ERROR("Could not patch vddc_on_mclk leakage table\n");
+ ret = si_patch_single_dependency_table_based_on_leakage(adev,
+ &adev->pm.dpm.dyn_state.vddci_dependency_on_mclk);
+ if (ret)
+ DRM_ERROR("Could not patch vddci_on_mclk leakage table\n");
+ return ret;
+}
+
+static void si_set_pcie_lane_width_in_smc(struct amdgpu_device *adev,
+ struct amdgpu_ps *amdgpu_new_state,
+ struct amdgpu_ps *amdgpu_current_state)
+{
+ u32 lane_width;
+ u32 new_lane_width =
+ (amdgpu_new_state->caps & ATOM_PPLIB_PCIE_LINK_WIDTH_MASK) >> ATOM_PPLIB_PCIE_LINK_WIDTH_SHIFT;
+ u32 current_lane_width =
+ (amdgpu_current_state->caps & ATOM_PPLIB_PCIE_LINK_WIDTH_MASK) >> ATOM_PPLIB_PCIE_LINK_WIDTH_SHIFT;
+
+ if (new_lane_width != current_lane_width) {
+ amdgpu_set_pcie_lanes(adev, new_lane_width);
+ lane_width = amdgpu_get_pcie_lanes(adev);
+ si_write_smc_soft_register(adev, SI_SMC_SOFT_REGISTER_non_ulv_pcie_link_width, lane_width);
+ }
+}
+
+static void si_dpm_setup_asic(struct amdgpu_device *adev)
+{
+ si_read_clock_registers(adev);
+ si_enable_acpi_power_management(adev);
+}
+
+static int si_thermal_enable_alert(struct amdgpu_device *adev,
+ bool enable)
+{
+ u32 thermal_int = RREG32(CG_THERMAL_INT);
+
+ if (enable) {
+ PPSMC_Result result;
+
+ thermal_int &= ~(THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW);
+ WREG32(CG_THERMAL_INT, thermal_int);
+ result = amdgpu_si_send_msg_to_smc(adev, PPSMC_MSG_EnableThermalInterrupt);
+ if (result != PPSMC_Result_OK) {
+ DRM_DEBUG_KMS("Could not enable thermal interrupts.\n");
+ return -EINVAL;
+ }
+ } else {
+ thermal_int |= THERM_INT_MASK_HIGH | THERM_INT_MASK_LOW;
+ WREG32(CG_THERMAL_INT, thermal_int);
+ }
+
+ return 0;
+}
+
+static int si_thermal_set_temperature_range(struct amdgpu_device *adev,
+ int min_temp, int max_temp)
+{
+ int low_temp = 0 * 1000;
+ int high_temp = 255 * 1000;
+
+ if (low_temp < min_temp)
+ low_temp = min_temp;
+ if (high_temp > max_temp)
+ high_temp = max_temp;
+ if (high_temp < low_temp) {
+ DRM_ERROR("invalid thermal range: %d - %d\n", low_temp, high_temp);
+ return -EINVAL;
+ }
+
+ WREG32_P(CG_THERMAL_INT, DIG_THERM_INTH(high_temp / 1000), ~DIG_THERM_INTH_MASK);
+ WREG32_P(CG_THERMAL_INT, DIG_THERM_INTL(low_temp / 1000), ~DIG_THERM_INTL_MASK);
+ WREG32_P(CG_THERMAL_CTRL, DIG_THERM_DPM(high_temp / 1000), ~DIG_THERM_DPM_MASK);
+
+ adev->pm.dpm.thermal.min_temp = low_temp;
+ adev->pm.dpm.thermal.max_temp = high_temp;
+
+ return 0;
+}
+
+static void si_fan_ctrl_set_static_mode(struct amdgpu_device *adev, u32 mode)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 tmp;
+
+ if (si_pi->fan_ctrl_is_in_default_mode) {
+ tmp = (RREG32(CG_FDO_CTRL2) & FDO_PWM_MODE_MASK) >> FDO_PWM_MODE_SHIFT;
+ si_pi->fan_ctrl_default_mode = tmp;
+ tmp = (RREG32(CG_FDO_CTRL2) & TMIN_MASK) >> TMIN_SHIFT;
+ si_pi->t_min = tmp;
+ si_pi->fan_ctrl_is_in_default_mode = false;
+ }
+
+ tmp = RREG32(CG_FDO_CTRL2) & ~TMIN_MASK;
+ tmp |= TMIN(0);
+ WREG32(CG_FDO_CTRL2, tmp);
+
+ tmp = RREG32(CG_FDO_CTRL2) & ~FDO_PWM_MODE_MASK;
+ tmp |= FDO_PWM_MODE(mode);
+ WREG32(CG_FDO_CTRL2, tmp);
+}
+
+static int si_thermal_setup_fan_table(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ PP_SIslands_FanTable fan_table = { FDO_MODE_HARDWARE };
+ u32 duty100;
+ u32 t_diff1, t_diff2, pwm_diff1, pwm_diff2;
+ u16 fdo_min, slope1, slope2;
+ u32 reference_clock, tmp;
+ int ret;
+ u64 tmp64;
+
+ if (!si_pi->fan_table_start) {
+ adev->pm.dpm.fan.ucode_fan_control = false;
+ return 0;
+ }
+
+ duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+
+ if (duty100 == 0) {
+ adev->pm.dpm.fan.ucode_fan_control = false;
+ return 0;
+ }
+
+ tmp64 = (u64)adev->pm.dpm.fan.pwm_min * duty100;
+ do_div(tmp64, 10000);
+ fdo_min = (u16)tmp64;
+
+ t_diff1 = adev->pm.dpm.fan.t_med - adev->pm.dpm.fan.t_min;
+ t_diff2 = adev->pm.dpm.fan.t_high - adev->pm.dpm.fan.t_med;
+
+ pwm_diff1 = adev->pm.dpm.fan.pwm_med - adev->pm.dpm.fan.pwm_min;
+ pwm_diff2 = adev->pm.dpm.fan.pwm_high - adev->pm.dpm.fan.pwm_med;
+
+ slope1 = (u16)((50 + ((16 * duty100 * pwm_diff1) / t_diff1)) / 100);
+ slope2 = (u16)((50 + ((16 * duty100 * pwm_diff2) / t_diff2)) / 100);
+
+ fan_table.temp_min = cpu_to_be16((50 + adev->pm.dpm.fan.t_min) / 100);
+ fan_table.temp_med = cpu_to_be16((50 + adev->pm.dpm.fan.t_med) / 100);
+ fan_table.temp_max = cpu_to_be16((50 + adev->pm.dpm.fan.t_max) / 100);
+ fan_table.slope1 = cpu_to_be16(slope1);
+ fan_table.slope2 = cpu_to_be16(slope2);
+ fan_table.fdo_min = cpu_to_be16(fdo_min);
+ fan_table.hys_down = cpu_to_be16(adev->pm.dpm.fan.t_hyst);
+ fan_table.hys_up = cpu_to_be16(1);
+ fan_table.hys_slope = cpu_to_be16(1);
+ fan_table.temp_resp_lim = cpu_to_be16(5);
+ reference_clock = amdgpu_asic_get_xclk(adev);
+
+ fan_table.refresh_period = cpu_to_be32((adev->pm.dpm.fan.cycle_delay *
+ reference_clock) / 1600);
+ fan_table.fdo_max = cpu_to_be16((u16)duty100);
+
+ tmp = (RREG32(CG_MULT_THERMAL_CTRL) & TEMP_SEL_MASK) >> TEMP_SEL_SHIFT;
+ fan_table.temp_src = (uint8_t)tmp;
+
+ ret = amdgpu_si_copy_bytes_to_smc(adev,
+ si_pi->fan_table_start,
+ (u8 *)(&fan_table),
+ sizeof(fan_table),
+ si_pi->sram_end);
+
+ if (ret) {
+ DRM_ERROR("Failed to load fan table to the SMC.");
+ adev->pm.dpm.fan.ucode_fan_control = false;
+ }
+
+ return ret;
+}
+
+static int si_fan_ctrl_start_smc_fan_control(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ PPSMC_Result ret;
+
+ ret = amdgpu_si_send_msg_to_smc(adev, PPSMC_StartFanControl);
+ if (ret == PPSMC_Result_OK) {
+ si_pi->fan_is_controlled_by_smc = true;
+ return 0;
+ } else {
+ return -EINVAL;
+ }
+}
+
+static int si_fan_ctrl_stop_smc_fan_control(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ PPSMC_Result ret;
+
+ ret = amdgpu_si_send_msg_to_smc(adev, PPSMC_StopFanControl);
+
+ if (ret == PPSMC_Result_OK) {
+ si_pi->fan_is_controlled_by_smc = false;
+ return 0;
+ } else {
+ return -EINVAL;
+ }
+}
+
+static int si_dpm_get_fan_speed_percent(struct amdgpu_device *adev,
+ u32 *speed)
+{
+ u32 duty, duty100;
+ u64 tmp64;
+
+ if (adev->pm.no_fan)
+ return -ENOENT;
+
+ duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+ duty = (RREG32(CG_THERMAL_STATUS) & FDO_PWM_DUTY_MASK) >> FDO_PWM_DUTY_SHIFT;
+
+ if (duty100 == 0)
+ return -EINVAL;
+
+ tmp64 = (u64)duty * 100;
+ do_div(tmp64, duty100);
+ *speed = (u32)tmp64;
+
+ if (*speed > 100)
+ *speed = 100;
+
+ return 0;
+}
+
+static int si_dpm_set_fan_speed_percent(struct amdgpu_device *adev,
+ u32 speed)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 tmp;
+ u32 duty, duty100;
+ u64 tmp64;
+
+ if (adev->pm.no_fan)
+ return -ENOENT;
+
+ if (si_pi->fan_is_controlled_by_smc)
+ return -EINVAL;
+
+ if (speed > 100)
+ return -EINVAL;
+
+ duty100 = (RREG32(CG_FDO_CTRL1) & FMAX_DUTY100_MASK) >> FMAX_DUTY100_SHIFT;
+
+ if (duty100 == 0)
+ return -EINVAL;
+
+ tmp64 = (u64)speed * duty100;
+ do_div(tmp64, 100);
+ duty = (u32)tmp64;
+
+ tmp = RREG32(CG_FDO_CTRL0) & ~FDO_STATIC_DUTY_MASK;
+ tmp |= FDO_STATIC_DUTY(duty);
+ WREG32(CG_FDO_CTRL0, tmp);
+
+ return 0;
+}
+
+static void si_dpm_set_fan_control_mode(struct amdgpu_device *adev, u32 mode)
+{
+ if (mode) {
+ /* stop auto-manage */
+ if (adev->pm.dpm.fan.ucode_fan_control)
+ si_fan_ctrl_stop_smc_fan_control(adev);
+ si_fan_ctrl_set_static_mode(adev, mode);
+ } else {
+ /* restart auto-manage */
+ if (adev->pm.dpm.fan.ucode_fan_control)
+ si_thermal_start_smc_fan_control(adev);
+ else
+ si_fan_ctrl_set_default_mode(adev);
+ }
+}
+
+static u32 si_dpm_get_fan_control_mode(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 tmp;
+
+ if (si_pi->fan_is_controlled_by_smc)
+ return 0;
+
+ tmp = RREG32(CG_FDO_CTRL2) & FDO_PWM_MODE_MASK;
+ return (tmp >> FDO_PWM_MODE_SHIFT);
+}
+
+#if 0
+static int si_fan_ctrl_get_fan_speed_rpm(struct amdgpu_device *adev,
+ u32 *speed)
+{
+ u32 tach_period;
+ u32 xclk = amdgpu_asic_get_xclk(adev);
+
+ if (adev->pm.no_fan)
+ return -ENOENT;
+
+ if (adev->pm.fan_pulses_per_revolution == 0)
+ return -ENOENT;
+
+ tach_period = (RREG32(CG_TACH_STATUS) & TACH_PERIOD_MASK) >> TACH_PERIOD_SHIFT;
+ if (tach_period == 0)
+ return -ENOENT;
+
+ *speed = 60 * xclk * 10000 / tach_period;
+
+ return 0;
+}
+
+static int si_fan_ctrl_set_fan_speed_rpm(struct amdgpu_device *adev,
+ u32 speed)
+{
+ u32 tach_period, tmp;
+ u32 xclk = amdgpu_asic_get_xclk(adev);
+
+ if (adev->pm.no_fan)
+ return -ENOENT;
+
+ if (adev->pm.fan_pulses_per_revolution == 0)
+ return -ENOENT;
+
+ if ((speed < adev->pm.fan_min_rpm) ||
+ (speed > adev->pm.fan_max_rpm))
+ return -EINVAL;
+
+ if (adev->pm.dpm.fan.ucode_fan_control)
+ si_fan_ctrl_stop_smc_fan_control(adev);
+
+ tach_period = 60 * xclk * 10000 / (8 * speed);
+ tmp = RREG32(CG_TACH_CTRL) & ~TARGET_PERIOD_MASK;
+ tmp |= TARGET_PERIOD(tach_period);
+ WREG32(CG_TACH_CTRL, tmp);
+
+ si_fan_ctrl_set_static_mode(adev, FDO_PWM_MODE_STATIC_RPM);
+
+ return 0;
+}
+#endif
+
+static void si_fan_ctrl_set_default_mode(struct amdgpu_device *adev)
+{
+ struct si_power_info *si_pi = si_get_pi(adev);
+ u32 tmp;
+
+ if (!si_pi->fan_ctrl_is_in_default_mode) {
+ tmp = RREG32(CG_FDO_CTRL2) & ~FDO_PWM_MODE_MASK;
+ tmp |= FDO_PWM_MODE(si_pi->fan_ctrl_default_mode);
+ WREG32(CG_FDO_CTRL2, tmp);
+
+ tmp = RREG32(CG_FDO_CTRL2) & ~TMIN_MASK;
+ tmp |= TMIN(si_pi->t_min);
+ WREG32(CG_FDO_CTRL2, tmp);
+ si_pi->fan_ctrl_is_in_default_mode = true;
+ }
+}
+
+static void si_thermal_start_smc_fan_control(struct amdgpu_device *adev)
+{
+ if (adev->pm.dpm.fan.ucode_fan_control) {
+ si_fan_ctrl_start_smc_fan_control(adev);
+ si_fan_ctrl_set_static_mode(adev, FDO_PWM_MODE_STATIC);
+ }
+}
+
+static void si_thermal_initialize(struct amdgpu_device *adev)
+{
+ u32 tmp;
+
+ if (adev->pm.fan_pulses_per_revolution) {
+ tmp = RREG32(CG_TACH_CTRL) & ~EDGE_PER_REV_MASK;
+ tmp |= EDGE_PER_REV(adev->pm.fan_pulses_per_revolution -1);
+ WREG32(CG_TACH_CTRL, tmp);
+ }
+
+ tmp = RREG32(CG_FDO_CTRL2) & ~TACH_PWM_RESP_RATE_MASK;
+ tmp |= TACH_PWM_RESP_RATE(0x28);
+ WREG32(CG_FDO_CTRL2, tmp);
+}
+
+static int si_thermal_start_thermal_controller(struct amdgpu_device *adev)
+{
+ int ret;
+
+ si_thermal_initialize(adev);
+ ret = si_thermal_set_temperature_range(adev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
+ if (ret)
+ return ret;
+ ret = si_thermal_enable_alert(adev, true);
+ if (ret)
+ return ret;
+ if (adev->pm.dpm.fan.ucode_fan_control) {
+ ret = si_halt_smc(adev);
+ if (ret)
+ return ret;
+ ret = si_thermal_setup_fan_table(adev);
+ if (ret)
+ return ret;
+ ret = si_resume_smc(adev);
+ if (ret)
+ return ret;
+ si_thermal_start_smc_fan_control(adev);
+ }
+
+ return 0;
+}
+
+static void si_thermal_stop_thermal_controller(struct amdgpu_device *adev)
+{
+ if (!adev->pm.no_fan) {
+ si_fan_ctrl_set_default_mode(adev);
+ si_fan_ctrl_stop_smc_fan_control(adev);
+ }
+}
+
+static int si_dpm_enable(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct amdgpu_ps *boot_ps = adev->pm.dpm.boot_ps;
+ int ret;
+
+ if (amdgpu_si_is_smc_running(adev))
+ return -EINVAL;
+ if (pi->voltage_control || si_pi->voltage_control_svi2)
+ si_enable_voltage_control(adev, true);
+ if (pi->mvdd_control)
+ si_get_mvdd_configuration(adev);
+ if (pi->voltage_control || si_pi->voltage_control_svi2) {
+ ret = si_construct_voltage_tables(adev);
+ if (ret) {
+ DRM_ERROR("si_construct_voltage_tables failed\n");
+ return ret;
+ }
+ }
+ if (eg_pi->dynamic_ac_timing) {
+ ret = si_initialize_mc_reg_table(adev);
+ if (ret)
+ eg_pi->dynamic_ac_timing = false;
+ }
+ if (pi->dynamic_ss)
+ si_enable_spread_spectrum(adev, true);
+ if (pi->thermal_protection)
+ si_enable_thermal_protection(adev, true);
+ si_setup_bsp(adev);
+ si_program_git(adev);
+ si_program_tp(adev);
+ si_program_tpp(adev);
+ si_program_sstp(adev);
+ si_enable_display_gap(adev);
+ si_program_vc(adev);
+ ret = si_upload_firmware(adev);
+ if (ret) {
+ DRM_ERROR("si_upload_firmware failed\n");
+ return ret;
+ }
+ ret = si_process_firmware_header(adev);
+ if (ret) {
+ DRM_ERROR("si_process_firmware_header failed\n");
+ return ret;
+ }
+ ret = si_initial_switch_from_arb_f0_to_f1(adev);
+ if (ret) {
+ DRM_ERROR("si_initial_switch_from_arb_f0_to_f1 failed\n");
+ return ret;
+ }
+ ret = si_init_smc_table(adev);
+ if (ret) {
+ DRM_ERROR("si_init_smc_table failed\n");
+ return ret;
+ }
+ ret = si_init_smc_spll_table(adev);
+ if (ret) {
+ DRM_ERROR("si_init_smc_spll_table failed\n");
+ return ret;
+ }
+ ret = si_init_arb_table_index(adev);
+ if (ret) {
+ DRM_ERROR("si_init_arb_table_index failed\n");
+ return ret;
+ }
+ if (eg_pi->dynamic_ac_timing) {
+ ret = si_populate_mc_reg_table(adev, boot_ps);
+ if (ret) {
+ DRM_ERROR("si_populate_mc_reg_table failed\n");
+ return ret;
+ }
+ }
+ ret = si_initialize_smc_cac_tables(adev);
+ if (ret) {
+ DRM_ERROR("si_initialize_smc_cac_tables failed\n");
+ return ret;
+ }
+ ret = si_initialize_hardware_cac_manager(adev);
+ if (ret) {
+ DRM_ERROR("si_initialize_hardware_cac_manager failed\n");
+ return ret;
+ }
+ ret = si_initialize_smc_dte_tables(adev);
+ if (ret) {
+ DRM_ERROR("si_initialize_smc_dte_tables failed\n");
+ return ret;
+ }
+ ret = si_populate_smc_tdp_limits(adev, boot_ps);
+ if (ret) {
+ DRM_ERROR("si_populate_smc_tdp_limits failed\n");
+ return ret;
+ }
+ ret = si_populate_smc_tdp_limits_2(adev, boot_ps);
+ if (ret) {
+ DRM_ERROR("si_populate_smc_tdp_limits_2 failed\n");
+ return ret;
+ }
+ si_program_response_times(adev);
+ si_program_ds_registers(adev);
+ si_dpm_start_smc(adev);
+ ret = si_notify_smc_display_change(adev, false);
+ if (ret) {
+ DRM_ERROR("si_notify_smc_display_change failed\n");
+ return ret;
+ }
+ si_enable_sclk_control(adev, true);
+ si_start_dpm(adev);
+
+ si_enable_auto_throttle_source(adev, AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL, true);
+ si_thermal_start_thermal_controller(adev);
+ ni_update_current_ps(adev, boot_ps);
+
+ return 0;
+}
+
+static int si_set_temperature_range(struct amdgpu_device *adev)
+{
+ int ret;
+
+ ret = si_thermal_enable_alert(adev, false);
+ if (ret)
+ return ret;
+ ret = si_thermal_set_temperature_range(adev, R600_TEMP_RANGE_MIN, R600_TEMP_RANGE_MAX);
+ if (ret)
+ return ret;
+ ret = si_thermal_enable_alert(adev, true);
+ if (ret)
+ return ret;
+
+ return ret;
+}
+
+static void si_dpm_disable(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct amdgpu_ps *boot_ps = adev->pm.dpm.boot_ps;
+
+ if (!amdgpu_si_is_smc_running(adev))
+ return;
+ si_thermal_stop_thermal_controller(adev);
+ si_disable_ulv(adev);
+ si_clear_vc(adev);
+ if (pi->thermal_protection)
+ si_enable_thermal_protection(adev, false);
+ si_enable_power_containment(adev, boot_ps, false);
+ si_enable_smc_cac(adev, boot_ps, false);
+ si_enable_spread_spectrum(adev, false);
+ si_enable_auto_throttle_source(adev, AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL, false);
+ si_stop_dpm(adev);
+ si_reset_to_default(adev);
+ si_dpm_stop_smc(adev);
+ si_force_switch_to_arb_f0(adev);
+
+ ni_update_current_ps(adev, boot_ps);
+}
+
+static int si_dpm_pre_set_power_state(struct amdgpu_device *adev)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct amdgpu_ps requested_ps = *adev->pm.dpm.requested_ps;
+ struct amdgpu_ps *new_ps = &requested_ps;
+
+ ni_update_requested_ps(adev, new_ps);
+ si_apply_state_adjust_rules(adev, &eg_pi->requested_rps);
+
+ return 0;
+}
+
+static int si_power_control_set_level(struct amdgpu_device *adev)
+{
+ struct amdgpu_ps *new_ps = adev->pm.dpm.requested_ps;
+ int ret;
+
+ ret = si_restrict_performance_levels_before_switch(adev);
+ if (ret)
+ return ret;
+ ret = si_halt_smc(adev);
+ if (ret)
+ return ret;
+ ret = si_populate_smc_tdp_limits(adev, new_ps);
+ if (ret)
+ return ret;
+ ret = si_populate_smc_tdp_limits_2(adev, new_ps);
+ if (ret)
+ return ret;
+ ret = si_resume_smc(adev);
+ if (ret)
+ return ret;
+ ret = si_set_sw_state(adev);
+ if (ret)
+ return ret;
+ return 0;
+}
+
+static int si_dpm_set_power_state(struct amdgpu_device *adev)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct amdgpu_ps *new_ps = &eg_pi->requested_rps;
+ struct amdgpu_ps *old_ps = &eg_pi->current_rps;
+ int ret;
+
+ ret = si_disable_ulv(adev);
+ if (ret) {
+ DRM_ERROR("si_disable_ulv failed\n");
+ return ret;
+ }
+ ret = si_restrict_performance_levels_before_switch(adev);
+ if (ret) {
+ DRM_ERROR("si_restrict_performance_levels_before_switch failed\n");
+ return ret;
+ }
+ if (eg_pi->pcie_performance_request)
+ si_request_link_speed_change_before_state_change(adev, new_ps, old_ps);
+ ni_set_uvd_clock_before_set_eng_clock(adev, new_ps, old_ps);
+ ret = si_enable_power_containment(adev, new_ps, false);
+ if (ret) {
+ DRM_ERROR("si_enable_power_containment failed\n");
+ return ret;
+ }
+ ret = si_enable_smc_cac(adev, new_ps, false);
+ if (ret) {
+ DRM_ERROR("si_enable_smc_cac failed\n");
+ return ret;
+ }
+ ret = si_halt_smc(adev);
+ if (ret) {
+ DRM_ERROR("si_halt_smc failed\n");
+ return ret;
+ }
+ ret = si_upload_sw_state(adev, new_ps);
+ if (ret) {
+ DRM_ERROR("si_upload_sw_state failed\n");
+ return ret;
+ }
+ ret = si_upload_smc_data(adev);
+ if (ret) {
+ DRM_ERROR("si_upload_smc_data failed\n");
+ return ret;
+ }
+ ret = si_upload_ulv_state(adev);
+ if (ret) {
+ DRM_ERROR("si_upload_ulv_state failed\n");
+ return ret;
+ }
+ if (eg_pi->dynamic_ac_timing) {
+ ret = si_upload_mc_reg_table(adev, new_ps);
+ if (ret) {
+ DRM_ERROR("si_upload_mc_reg_table failed\n");
+ return ret;
+ }
+ }
+ ret = si_program_memory_timing_parameters(adev, new_ps);
+ if (ret) {
+ DRM_ERROR("si_program_memory_timing_parameters failed\n");
+ return ret;
+ }
+ si_set_pcie_lane_width_in_smc(adev, new_ps, old_ps);
+
+ ret = si_resume_smc(adev);
+ if (ret) {
+ DRM_ERROR("si_resume_smc failed\n");
+ return ret;
+ }
+ ret = si_set_sw_state(adev);
+ if (ret) {
+ DRM_ERROR("si_set_sw_state failed\n");
+ return ret;
+ }
+ ni_set_uvd_clock_after_set_eng_clock(adev, new_ps, old_ps);
+ if (eg_pi->pcie_performance_request)
+ si_notify_link_speed_change_after_state_change(adev, new_ps, old_ps);
+ ret = si_set_power_state_conditionally_enable_ulv(adev, new_ps);
+ if (ret) {
+ DRM_ERROR("si_set_power_state_conditionally_enable_ulv failed\n");
+ return ret;
+ }
+ ret = si_enable_smc_cac(adev, new_ps, true);
+ if (ret) {
+ DRM_ERROR("si_enable_smc_cac failed\n");
+ return ret;
+ }
+ ret = si_enable_power_containment(adev, new_ps, true);
+ if (ret) {
+ DRM_ERROR("si_enable_power_containment failed\n");
+ return ret;
+ }
+
+ ret = si_power_control_set_level(adev);
+ if (ret) {
+ DRM_ERROR("si_power_control_set_level failed\n");
+ return ret;
+ }
+
+ return 0;
+}
+
+static void si_dpm_post_set_power_state(struct amdgpu_device *adev)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct amdgpu_ps *new_ps = &eg_pi->requested_rps;
+
+ ni_update_current_ps(adev, new_ps);
+}
+
+#if 0
+void si_dpm_reset_asic(struct amdgpu_device *adev)
+{
+ si_restrict_performance_levels_before_switch(adev);
+ si_disable_ulv(adev);
+ si_set_boot_state(adev);
+}
+#endif
+
+static void si_dpm_display_configuration_changed(struct amdgpu_device *adev)
+{
+ si_program_display_gap(adev);
+}
+
+
+static void si_parse_pplib_non_clock_info(struct amdgpu_device *adev,
+ struct amdgpu_ps *rps,
+ struct _ATOM_PPLIB_NONCLOCK_INFO *non_clock_info,
+ u8 table_rev)
+{
+ rps->caps = le32_to_cpu(non_clock_info->ulCapsAndSettings);
+ rps->class = le16_to_cpu(non_clock_info->usClassification);
+ rps->class2 = le16_to_cpu(non_clock_info->usClassification2);
+
+ if (ATOM_PPLIB_NONCLOCKINFO_VER1 < table_rev) {
+ rps->vclk = le32_to_cpu(non_clock_info->ulVCLK);
+ rps->dclk = le32_to_cpu(non_clock_info->ulDCLK);
+ } else if (r600_is_uvd_state(rps->class, rps->class2)) {
+ rps->vclk = RV770_DEFAULT_VCLK_FREQ;
+ rps->dclk = RV770_DEFAULT_DCLK_FREQ;
+ } else {
+ rps->vclk = 0;
+ rps->dclk = 0;
+ }
+
+ if (rps->class & ATOM_PPLIB_CLASSIFICATION_BOOT)
+ adev->pm.dpm.boot_ps = rps;
+ if (rps->class & ATOM_PPLIB_CLASSIFICATION_UVDSTATE)
+ adev->pm.dpm.uvd_ps = rps;
+}
+
+static void si_parse_pplib_clock_info(struct amdgpu_device *adev,
+ struct amdgpu_ps *rps, int index,
+ union pplib_clock_info *clock_info)
+{
+ struct rv7xx_power_info *pi = rv770_get_pi(adev);
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_power_info *si_pi = si_get_pi(adev);
+ struct si_ps *ps = si_get_ps(rps);
+ u16 leakage_voltage;
+ struct rv7xx_pl *pl = &ps->performance_levels[index];
+ int ret;
+
+ ps->performance_level_count = index + 1;
+
+ pl->sclk = le16_to_cpu(clock_info->si.usEngineClockLow);
+ pl->sclk |= clock_info->si.ucEngineClockHigh << 16;
+ pl->mclk = le16_to_cpu(clock_info->si.usMemoryClockLow);
+ pl->mclk |= clock_info->si.ucMemoryClockHigh << 16;
+
+ pl->vddc = le16_to_cpu(clock_info->si.usVDDC);
+ pl->vddci = le16_to_cpu(clock_info->si.usVDDCI);
+ pl->flags = le32_to_cpu(clock_info->si.ulFlags);
+ pl->pcie_gen = r600_get_pcie_gen_support(adev,
+ si_pi->sys_pcie_mask,
+ si_pi->boot_pcie_gen,
+ clock_info->si.ucPCIEGen);
+
+ /* patch up vddc if necessary */
+ ret = si_get_leakage_voltage_from_leakage_index(adev, pl->vddc,
+ &leakage_voltage);
+ if (ret == 0)
+ pl->vddc = leakage_voltage;
+
+ if (rps->class & ATOM_PPLIB_CLASSIFICATION_ACPI) {
+ pi->acpi_vddc = pl->vddc;
+ eg_pi->acpi_vddci = pl->vddci;
+ si_pi->acpi_pcie_gen = pl->pcie_gen;
+ }
+
+ if ((rps->class2 & ATOM_PPLIB_CLASSIFICATION2_ULV) &&
+ index == 0) {
+ /* XXX disable for A0 tahiti */
+ si_pi->ulv.supported = false;
+ si_pi->ulv.pl = *pl;
+ si_pi->ulv.one_pcie_lane_in_ulv = false;
+ si_pi->ulv.volt_change_delay = SISLANDS_ULVVOLTAGECHANGEDELAY_DFLT;
+ si_pi->ulv.cg_ulv_parameter = SISLANDS_CGULVPARAMETER_DFLT;
+ si_pi->ulv.cg_ulv_control = SISLANDS_CGULVCONTROL_DFLT;
+ }
+
+ if (pi->min_vddc_in_table > pl->vddc)
+ pi->min_vddc_in_table = pl->vddc;
+
+ if (pi->max_vddc_in_table < pl->vddc)
+ pi->max_vddc_in_table = pl->vddc;
+
+ /* patch up boot state */
+ if (rps->class & ATOM_PPLIB_CLASSIFICATION_BOOT) {
+ u16 vddc, vddci, mvdd;
+ amdgpu_atombios_get_default_voltages(adev, &vddc, &vddci, &mvdd);
+ pl->mclk = adev->clock.default_mclk;
+ pl->sclk = adev->clock.default_sclk;
+ pl->vddc = vddc;
+ pl->vddci = vddci;
+ si_pi->mvdd_bootup_value = mvdd;
+ }
+
+ if ((rps->class & ATOM_PPLIB_CLASSIFICATION_UI_MASK) ==
+ ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE) {
+ adev->pm.dpm.dyn_state.max_clock_voltage_on_ac.sclk = pl->sclk;
+ adev->pm.dpm.dyn_state.max_clock_voltage_on_ac.mclk = pl->mclk;
+ adev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddc = pl->vddc;
+ adev->pm.dpm.dyn_state.max_clock_voltage_on_ac.vddci = pl->vddci;
+ }
+}
+
+union pplib_power_state {
+ struct _ATOM_PPLIB_STATE v1;
+ struct _ATOM_PPLIB_STATE_V2 v2;
+};
+
+static int si_parse_power_table(struct amdgpu_device *adev)
+{
+ struct amdgpu_mode_info *mode_info = &adev->mode_info;
+ struct _ATOM_PPLIB_NONCLOCK_INFO *non_clock_info;
+ union pplib_power_state *power_state;
+ int i, j, k, non_clock_array_index, clock_array_index;
+ union pplib_clock_info *clock_info;
+ struct _StateArray *state_array;
+ struct _ClockInfoArray *clock_info_array;
+ struct _NonClockInfoArray *non_clock_info_array;
+ union power_info *power_info;
+ int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
+ u16 data_offset;
+ u8 frev, crev;
+ u8 *power_state_offset;
+ struct si_ps *ps;
+
+ if (!amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL,
+ &frev, &crev, &data_offset))
+ return -EINVAL;
+ power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
+
+ amdgpu_add_thermal_controller(adev);
+
+ state_array = (struct _StateArray *)
+ (mode_info->atom_context->bios + data_offset +
+ le16_to_cpu(power_info->pplib.usStateArrayOffset));
+ clock_info_array = (struct _ClockInfoArray *)
+ (mode_info->atom_context->bios + data_offset +
+ le16_to_cpu(power_info->pplib.usClockInfoArrayOffset));
+ non_clock_info_array = (struct _NonClockInfoArray *)
+ (mode_info->atom_context->bios + data_offset +
+ le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
+
+ adev->pm.dpm.ps = kzalloc(sizeof(struct amdgpu_ps) *
+ state_array->ucNumEntries, GFP_KERNEL);
+ if (!adev->pm.dpm.ps)
+ return -ENOMEM;
+ power_state_offset = (u8 *)state_array->states;
+ for (i = 0; i < state_array->ucNumEntries; i++) {
+ u8 *idx;
+ power_state = (union pplib_power_state *)power_state_offset;
+ non_clock_array_index = power_state->v2.nonClockInfoIndex;
+ non_clock_info = (struct _ATOM_PPLIB_NONCLOCK_INFO *)
+ &non_clock_info_array->nonClockInfo[non_clock_array_index];
+ ps = kzalloc(sizeof(struct si_ps), GFP_KERNEL);
+ if (ps == NULL) {
+ kfree(adev->pm.dpm.ps);
+ return -ENOMEM;
+ }
+ adev->pm.dpm.ps[i].ps_priv = ps;
+ si_parse_pplib_non_clock_info(adev, &adev->pm.dpm.ps[i],
+ non_clock_info,
+ non_clock_info_array->ucEntrySize);
+ k = 0;
+ idx = (u8 *)&power_state->v2.clockInfoIndex[0];
+ for (j = 0; j < power_state->v2.ucNumDPMLevels; j++) {
+ clock_array_index = idx[j];
+ if (clock_array_index >= clock_info_array->ucNumEntries)
+ continue;
+ if (k >= SISLANDS_MAX_HARDWARE_POWERLEVELS)
+ break;
+ clock_info = (union pplib_clock_info *)
+ ((u8 *)&clock_info_array->clockInfo[0] +
+ (clock_array_index * clock_info_array->ucEntrySize));
+ si_parse_pplib_clock_info(adev,
+ &adev->pm.dpm.ps[i], k,
+ clock_info);
+ k++;
+ }
+ power_state_offset += 2 + power_state->v2.ucNumDPMLevels;
+ }
+ adev->pm.dpm.num_ps = state_array->ucNumEntries;
+
+ /* fill in the vce power states */
+ for (i = 0; i < AMDGPU_MAX_VCE_LEVELS; i++) {
+ u32 sclk, mclk;
+ clock_array_index = adev->pm.dpm.vce_states[i].clk_idx;
+ clock_info = (union pplib_clock_info *)
+ &clock_info_array->clockInfo[clock_array_index * clock_info_array->ucEntrySize];
+ sclk = le16_to_cpu(clock_info->si.usEngineClockLow);
+ sclk |= clock_info->si.ucEngineClockHigh << 16;
+ mclk = le16_to_cpu(clock_info->si.usMemoryClockLow);
+ mclk |= clock_info->si.ucMemoryClockHigh << 16;
+ adev->pm.dpm.vce_states[i].sclk = sclk;
+ adev->pm.dpm.vce_states[i].mclk = mclk;
+ }
+
+ return 0;
+}
+
+static int si_dpm_init(struct amdgpu_device *adev)
+{
+ struct rv7xx_power_info *pi;
+ struct evergreen_power_info *eg_pi;
+ struct ni_power_info *ni_pi;
+ struct si_power_info *si_pi;
+ struct atom_clock_dividers dividers;
+ int ret;
+ u32 mask;
+
+ si_pi = kzalloc(sizeof(struct si_power_info), GFP_KERNEL);
+ if (si_pi == NULL)
+ return -ENOMEM;
+ adev->pm.dpm.priv = si_pi;
+ ni_pi = &si_pi->ni;
+ eg_pi = &ni_pi->eg;
+ pi = &eg_pi->rv7xx;
+
+ ret = drm_pcie_get_speed_cap_mask(adev->ddev, &mask);
+ if (ret)
+ si_pi->sys_pcie_mask = 0;
+ else
+ si_pi->sys_pcie_mask = mask;
+ si_pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID;
+ si_pi->boot_pcie_gen = si_get_current_pcie_speed(adev);
+
+ si_set_max_cu_value(adev);
+
+ rv770_get_max_vddc(adev);
+ si_get_leakage_vddc(adev);
+ si_patch_dependency_tables_based_on_leakage(adev);
+
+ pi->acpi_vddc = 0;
+ eg_pi->acpi_vddci = 0;
+ pi->min_vddc_in_table = 0;
+ pi->max_vddc_in_table = 0;
+
+ ret = amdgpu_get_platform_caps(adev);
+ if (ret)
+ return ret;
+
+ ret = amdgpu_parse_extended_power_table(adev);
+ if (ret)
+ return ret;
+
+ ret = si_parse_power_table(adev);
+ if (ret)
+ return ret;
+
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
+ kzalloc(4 * sizeof(struct amdgpu_clock_voltage_dependency_entry), GFP_KERNEL);
+ if (!adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
+ amdgpu_free_extended_power_table(adev);
+ return -ENOMEM;
+ }
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.count = 4;
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[0].clk = 0;
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[0].v = 0;
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[1].clk = 36000;
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[1].v = 720;
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[2].clk = 54000;
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[2].v = 810;
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[3].clk = 72000;
+ adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries[3].v = 900;
+
+ if (adev->pm.dpm.voltage_response_time == 0)
+ adev->pm.dpm.voltage_response_time = R600_VOLTAGERESPONSETIME_DFLT;
+ if (adev->pm.dpm.backbias_response_time == 0)
+ adev->pm.dpm.backbias_response_time = R600_BACKBIASRESPONSETIME_DFLT;
+
+ ret = amdgpu_atombios_get_clock_dividers(adev, COMPUTE_ENGINE_PLL_PARAM,
+ 0, false, ÷rs);
+ if (ret)
+ pi->ref_div = dividers.ref_div + 1;
+ else
+ pi->ref_div = R600_REFERENCEDIVIDER_DFLT;
+
+ eg_pi->smu_uvd_hs = false;
+
+ pi->mclk_strobe_mode_threshold = 40000;
+ if (si_is_special_1gb_platform(adev))
+ pi->mclk_stutter_mode_threshold = 0;
+ else
+ pi->mclk_stutter_mode_threshold = pi->mclk_strobe_mode_threshold;
+ pi->mclk_edc_enable_threshold = 40000;
+ eg_pi->mclk_edc_wr_enable_threshold = 40000;
+
+ ni_pi->mclk_rtt_mode_threshold = eg_pi->mclk_edc_wr_enable_threshold;
+
+ pi->voltage_control =
+ amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDC,
+ VOLTAGE_OBJ_GPIO_LUT);
+ if (!pi->voltage_control) {
+ si_pi->voltage_control_svi2 =
+ amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDC,
+ VOLTAGE_OBJ_SVID2);
+ if (si_pi->voltage_control_svi2)
+ amdgpu_atombios_get_svi2_info(adev, SET_VOLTAGE_TYPE_ASIC_VDDC,
+ &si_pi->svd_gpio_id, &si_pi->svc_gpio_id);
+ }
+
+ pi->mvdd_control =
+ amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_MVDDC,
+ VOLTAGE_OBJ_GPIO_LUT);
+
+ eg_pi->vddci_control =
+ amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDCI,
+ VOLTAGE_OBJ_GPIO_LUT);
+ if (!eg_pi->vddci_control)
+ si_pi->vddci_control_svi2 =
+ amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDCI,
+ VOLTAGE_OBJ_SVID2);
+
+ si_pi->vddc_phase_shed_control =
+ amdgpu_atombios_is_voltage_gpio(adev, SET_VOLTAGE_TYPE_ASIC_VDDC,
+ VOLTAGE_OBJ_PHASE_LUT);
+
+ rv770_get_engine_memory_ss(adev);
+
+ pi->asi = RV770_ASI_DFLT;
+ pi->pasi = CYPRESS_HASI_DFLT;
+ pi->vrc = SISLANDS_VRC_DFLT;
+
+ pi->gfx_clock_gating = true;
+
+ eg_pi->sclk_deep_sleep = true;
+ si_pi->sclk_deep_sleep_above_low = false;
+
+ if (adev->pm.int_thermal_type != THERMAL_TYPE_NONE)
+ pi->thermal_protection = true;
+ else
+ pi->thermal_protection = false;
+
+ eg_pi->dynamic_ac_timing = true;
+
+ eg_pi->light_sleep = true;
+#if defined(CONFIG_ACPI)
+ eg_pi->pcie_performance_request =
+ amdgpu_acpi_is_pcie_performance_request_supported(adev);
+#else
+ eg_pi->pcie_performance_request = false;
+#endif
+
+ si_pi->sram_end = SMC_RAM_END;
+
+ adev->pm.dpm.dyn_state.mclk_sclk_ratio = 4;
+ adev->pm.dpm.dyn_state.sclk_mclk_delta = 15000;
+ adev->pm.dpm.dyn_state.vddc_vddci_delta = 200;
+ adev->pm.dpm.dyn_state.valid_sclk_values.count = 0;
+ adev->pm.dpm.dyn_state.valid_sclk_values.values = NULL;
+ adev->pm.dpm.dyn_state.valid_mclk_values.count = 0;
+ adev->pm.dpm.dyn_state.valid_mclk_values.values = NULL;
+
+ si_initialize_powertune_defaults(adev);
+
+ /* make sure dc limits are valid */
+ if ((adev->pm.dpm.dyn_state.max_clock_voltage_on_dc.sclk == 0) ||
+ (adev->pm.dpm.dyn_state.max_clock_voltage_on_dc.mclk == 0))
+ adev->pm.dpm.dyn_state.max_clock_voltage_on_dc =
+ adev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
+
+ si_pi->fan_ctrl_is_in_default_mode = true;
+
+ return 0;
+}
+
+static void si_dpm_fini(struct amdgpu_device *adev)
+{
+ int i;
+
+ if (adev->pm.dpm.ps)
+ for (i = 0; i < adev->pm.dpm.num_ps; i++)
+ kfree(adev->pm.dpm.ps[i].ps_priv);
+ kfree(adev->pm.dpm.ps);
+ kfree(adev->pm.dpm.priv);
+ kfree(adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries);
+ amdgpu_free_extended_power_table(adev);
+}
+
+static void si_dpm_debugfs_print_current_performance_level(struct amdgpu_device *adev,
+ struct seq_file *m)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct amdgpu_ps *rps = &eg_pi->current_rps;
+ struct si_ps *ps = si_get_ps(rps);
+ struct rv7xx_pl *pl;
+ u32 current_index =
+ (RREG32(TARGET_AND_CURRENT_PROFILE_INDEX) & CURRENT_STATE_INDEX_MASK) >>
+ CURRENT_STATE_INDEX_SHIFT;
+
+ if (current_index >= ps->performance_level_count) {
+ seq_printf(m, "invalid dpm profile %d\n", current_index);
+ } else {
+ pl = &ps->performance_levels[current_index];
+ seq_printf(m, "uvd vclk: %d dclk: %d\n", rps->vclk, rps->dclk);
+ seq_printf(m, "power level %d sclk: %u mclk: %u vddc: %u vddci: %u pcie gen: %u\n",
+ current_index, pl->sclk, pl->mclk, pl->vddc, pl->vddci, pl->pcie_gen + 1);
+ }
+}
+
+static int si_dpm_set_interrupt_state(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ unsigned type,
+ enum amdgpu_interrupt_state state)
+{
+ u32 cg_thermal_int;
+
+ switch (type) {
+ case AMDGPU_THERMAL_IRQ_LOW_TO_HIGH:
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ cg_thermal_int = RREG32_SMC(CG_THERMAL_INT);
+ cg_thermal_int |= THERM_INT_MASK_HIGH;
+ WREG32_SMC(CG_THERMAL_INT, cg_thermal_int);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ cg_thermal_int = RREG32_SMC(CG_THERMAL_INT);
+ cg_thermal_int &= ~THERM_INT_MASK_HIGH;
+ WREG32_SMC(CG_THERMAL_INT, cg_thermal_int);
+ break;
+ default:
+ break;
+ }
+ break;
+
+ case AMDGPU_THERMAL_IRQ_HIGH_TO_LOW:
+ switch (state) {
+ case AMDGPU_IRQ_STATE_DISABLE:
+ cg_thermal_int = RREG32_SMC(CG_THERMAL_INT);
+ cg_thermal_int |= THERM_INT_MASK_LOW;
+ WREG32_SMC(CG_THERMAL_INT, cg_thermal_int);
+ break;
+ case AMDGPU_IRQ_STATE_ENABLE:
+ cg_thermal_int = RREG32_SMC(CG_THERMAL_INT);
+ cg_thermal_int &= ~THERM_INT_MASK_LOW;
+ WREG32_SMC(CG_THERMAL_INT, cg_thermal_int);
+ break;
+ default:
+ break;
+ }
+ break;
+
+ default:
+ break;
+ }
+ return 0;
+}
+
+static int si_dpm_process_interrupt(struct amdgpu_device *adev,
+ struct amdgpu_irq_src *source,
+ struct amdgpu_iv_entry *entry)
+{
+ bool queue_thermal = false;
+
+ if (entry == NULL)
+ return -EINVAL;
+
+ switch (entry->src_id) {
+ case 230: /* thermal low to high */
+ DRM_DEBUG("IH: thermal low to high\n");
+ adev->pm.dpm.thermal.high_to_low = false;
+ queue_thermal = true;
+ break;
+ case 231: /* thermal high to low */
+ DRM_DEBUG("IH: thermal high to low\n");
+ adev->pm.dpm.thermal.high_to_low = true;
+ queue_thermal = true;
+ break;
+ default:
+ break;
+ }
+
+ if (queue_thermal)
+ schedule_work(&adev->pm.dpm.thermal.work);
+
+ return 0;
+}
+
+static int si_dpm_late_init(void *handle)
+{
+ int ret;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!amdgpu_dpm)
+ return 0;
+
+ /* init the sysfs and debugfs files late */
+ ret = amdgpu_pm_sysfs_init(adev);
+ if (ret)
+ return ret;
+
+ ret = si_set_temperature_range(adev);
+ if (ret)
+ return ret;
+#if 0 //TODO ?
+ si_dpm_powergate_uvd(adev, true);
+#endif
+ return 0;
+}
+
+/**
+ * si_dpm_init_microcode - load ucode images from disk
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Use the firmware interface to load the ucode images into
+ * the driver (not loaded into hw).
+ * Returns 0 on success, error on failure.
+ */
+static int si_dpm_init_microcode(struct amdgpu_device *adev)
+{
+ const char *chip_name;
+ char fw_name[30];
+ int err;
+
+ DRM_DEBUG("\n");
+ switch (adev->asic_type) {
+ case CHIP_TAHITI:
+ chip_name = "tahiti";
+ break;
+ case CHIP_PITCAIRN:
+ if ((adev->pdev->revision == 0x81) ||
+ (adev->pdev->device == 0x6810) ||
+ (adev->pdev->device == 0x6811) ||
+ (adev->pdev->device == 0x6816) ||
+ (adev->pdev->device == 0x6817) ||
+ (adev->pdev->device == 0x6806))
+ chip_name = "pitcairn_k";
+ else
+ chip_name = "pitcairn";
+ break;
+ case CHIP_VERDE:
+ if ((adev->pdev->revision == 0x81) ||
+ (adev->pdev->revision == 0x83) ||
+ (adev->pdev->revision == 0x87) ||
+ (adev->pdev->device == 0x6820) ||
+ (adev->pdev->device == 0x6821) ||
+ (adev->pdev->device == 0x6822) ||
+ (adev->pdev->device == 0x6823) ||
+ (adev->pdev->device == 0x682A) ||
+ (adev->pdev->device == 0x682B))
+ chip_name = "verde_k";
+ else
+ chip_name = "verde";
+ break;
+ case CHIP_OLAND:
+ if ((adev->pdev->revision == 0xC7) ||
+ (adev->pdev->revision == 0x80) ||
+ (adev->pdev->revision == 0x81) ||
+ (adev->pdev->revision == 0x83) ||
+ (adev->pdev->device == 0x6604) ||
+ (adev->pdev->device == 0x6605))
+ chip_name = "oland_k";
+ else
+ chip_name = "oland";
+ break;
+ case CHIP_HAINAN:
+ if ((adev->pdev->revision == 0x81) ||
+ (adev->pdev->revision == 0x83) ||
+ (adev->pdev->revision == 0xC3) ||
+ (adev->pdev->device == 0x6664) ||
+ (adev->pdev->device == 0x6665) ||
+ (adev->pdev->device == 0x6667))
+ chip_name = "hainan_k";
+ else
+ chip_name = "hainan";
+ break;
+ default: BUG();
+ }
+
+ snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
+ err = request_firmware(&adev->pm.fw, fw_name, adev->dev);
+ if (err)
+ goto out;
+ err = amdgpu_ucode_validate(adev->pm.fw);
+
+out:
+ if (err) {
+ DRM_ERROR("si_smc: Failed to load firmware. err = %d\"%s\"\n",
+ err, fw_name);
+ release_firmware(adev->pm.fw);
+ adev->pm.fw = NULL;
+ }
+ return err;
+
+}
+
+static int si_dpm_sw_init(void *handle)
+{
+ int ret;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ ret = amdgpu_irq_add_id(adev, 230, &adev->pm.dpm.thermal.irq);
+ if (ret)
+ return ret;
+
+ ret = amdgpu_irq_add_id(adev, 231, &adev->pm.dpm.thermal.irq);
+ if (ret)
+ return ret;
+
+ /* default to balanced state */
+ adev->pm.dpm.state = POWER_STATE_TYPE_BALANCED;
+ adev->pm.dpm.user_state = POWER_STATE_TYPE_BALANCED;
+ adev->pm.dpm.forced_level = AMDGPU_DPM_FORCED_LEVEL_AUTO;
+ adev->pm.default_sclk = adev->clock.default_sclk;
+ adev->pm.default_mclk = adev->clock.default_mclk;
+ adev->pm.current_sclk = adev->clock.default_sclk;
+ adev->pm.current_mclk = adev->clock.default_mclk;
+ adev->pm.int_thermal_type = THERMAL_TYPE_NONE;
+
+ if (amdgpu_dpm == 0)
+ return 0;
+
+ ret = si_dpm_init_microcode(adev);
+ if (ret)
+ return ret;
+
+ INIT_WORK(&adev->pm.dpm.thermal.work, amdgpu_dpm_thermal_work_handler);
+ mutex_lock(&adev->pm.mutex);
+ ret = si_dpm_init(adev);
+ if (ret)
+ goto dpm_failed;
+ adev->pm.dpm.current_ps = adev->pm.dpm.requested_ps = adev->pm.dpm.boot_ps;
+ if (amdgpu_dpm == 1)
+ amdgpu_pm_print_power_states(adev);
+ mutex_unlock(&adev->pm.mutex);
+ DRM_INFO("amdgpu: dpm initialized\n");
+
+ return 0;
+
+dpm_failed:
+ si_dpm_fini(adev);
+ mutex_unlock(&adev->pm.mutex);
+ DRM_ERROR("amdgpu: dpm initialization failed\n");
+ return ret;
+}
+
+static int si_dpm_sw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ mutex_lock(&adev->pm.mutex);
+ amdgpu_pm_sysfs_fini(adev);
+ si_dpm_fini(adev);
+ mutex_unlock(&adev->pm.mutex);
+
+ return 0;
+}
+
+static int si_dpm_hw_init(void *handle)
+{
+ int ret;
+
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!amdgpu_dpm)
+ return 0;
+
+ mutex_lock(&adev->pm.mutex);
+ si_dpm_setup_asic(adev);
+ ret = si_dpm_enable(adev);
+ if (ret)
+ adev->pm.dpm_enabled = false;
+ else
+ adev->pm.dpm_enabled = true;
+ mutex_unlock(&adev->pm.mutex);
+
+ return ret;
+}
+
+static int si_dpm_hw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (adev->pm.dpm_enabled) {
+ mutex_lock(&adev->pm.mutex);
+ si_dpm_disable(adev);
+ mutex_unlock(&adev->pm.mutex);
+ }
+
+ return 0;
+}
+
+static int si_dpm_suspend(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (adev->pm.dpm_enabled) {
+ mutex_lock(&adev->pm.mutex);
+ /* disable dpm */
+ si_dpm_disable(adev);
+ /* reset the power state */
+ adev->pm.dpm.current_ps = adev->pm.dpm.requested_ps = adev->pm.dpm.boot_ps;
+ mutex_unlock(&adev->pm.mutex);
+ }
+ return 0;
+}
+
+static int si_dpm_resume(void *handle)
+{
+ int ret;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (adev->pm.dpm_enabled) {
+ /* asic init will reset to the boot state */
+ mutex_lock(&adev->pm.mutex);
+ si_dpm_setup_asic(adev);
+ ret = si_dpm_enable(adev);
+ if (ret)
+ adev->pm.dpm_enabled = false;
+ else
+ adev->pm.dpm_enabled = true;
+ mutex_unlock(&adev->pm.mutex);
+ if (adev->pm.dpm_enabled)
+ amdgpu_pm_compute_clocks(adev);
+ }
+ return 0;
+}
+
+static bool si_dpm_is_idle(void *handle)
+{
+ /* XXX */
+ return true;
+}
+
+static int si_dpm_wait_for_idle(void *handle)
+{
+ /* XXX */
+ return 0;
+}
+
+static int si_dpm_soft_reset(void *handle)
+{
+ return 0;
+}
+
+static int si_dpm_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ return 0;
+}
+
+static int si_dpm_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ return 0;
+}
+
+/* get temperature in millidegrees */
+static int si_dpm_get_temp(struct amdgpu_device *adev)
+{
+ u32 temp;
+ int actual_temp = 0;
+
+ temp = (RREG32(CG_MULT_THERMAL_STATUS) & CTF_TEMP_MASK) >>
+ CTF_TEMP_SHIFT;
+
+ if (temp & 0x200)
+ actual_temp = 255;
+ else
+ actual_temp = temp & 0x1ff;
+
+ actual_temp = (actual_temp * 1000);
+
+ return actual_temp;
+}
+
+static u32 si_dpm_get_sclk(struct amdgpu_device *adev, bool low)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_ps *requested_state = si_get_ps(&eg_pi->requested_rps);
+
+ if (low)
+ return requested_state->performance_levels[0].sclk;
+ else
+ return requested_state->performance_levels[requested_state->performance_level_count - 1].sclk;
+}
+
+static u32 si_dpm_get_mclk(struct amdgpu_device *adev, bool low)
+{
+ struct evergreen_power_info *eg_pi = evergreen_get_pi(adev);
+ struct si_ps *requested_state = si_get_ps(&eg_pi->requested_rps);
+
+ if (low)
+ return requested_state->performance_levels[0].mclk;
+ else
+ return requested_state->performance_levels[requested_state->performance_level_count - 1].mclk;
+}
+
+static void si_dpm_print_power_state(struct amdgpu_device *adev,
+ struct amdgpu_ps *rps)
+{
+ struct si_ps *ps = si_get_ps(rps);
+ struct rv7xx_pl *pl;
+ int i;
+
+ amdgpu_dpm_print_class_info(rps->class, rps->class2);
+ amdgpu_dpm_print_cap_info(rps->caps);
+ DRM_INFO("\tuvd vclk: %d dclk: %d\n", rps->vclk, rps->dclk);
+ for (i = 0; i < ps->performance_level_count; i++) {
+ pl = &ps->performance_levels[i];
+ if (adev->asic_type >= CHIP_TAHITI)
+ DRM_INFO("\t\tpower level %d sclk: %u mclk: %u vddc: %u vddci: %u pcie gen: %u\n",
+ i, pl->sclk, pl->mclk, pl->vddc, pl->vddci, pl->pcie_gen + 1);
+ else
+ DRM_INFO("\t\tpower level %d sclk: %u mclk: %u vddc: %u vddci: %u\n",
+ i, pl->sclk, pl->mclk, pl->vddc, pl->vddci);
+ }
+ amdgpu_dpm_print_ps_status(adev, rps);
+}
+
+static int si_dpm_early_init(void *handle)
+{
+
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ si_dpm_set_dpm_funcs(adev);
+ si_dpm_set_irq_funcs(adev);
+ return 0;
+}
+
+
+const struct amd_ip_funcs si_dpm_ip_funcs = {
+ .name = "si_dpm",
+ .early_init = si_dpm_early_init,
+ .late_init = si_dpm_late_init,
+ .sw_init = si_dpm_sw_init,
+ .sw_fini = si_dpm_sw_fini,
+ .hw_init = si_dpm_hw_init,
+ .hw_fini = si_dpm_hw_fini,
+ .suspend = si_dpm_suspend,
+ .resume = si_dpm_resume,
+ .is_idle = si_dpm_is_idle,
+ .wait_for_idle = si_dpm_wait_for_idle,
+ .soft_reset = si_dpm_soft_reset,
+ .set_clockgating_state = si_dpm_set_clockgating_state,
+ .set_powergating_state = si_dpm_set_powergating_state,
+};
+
+static const struct amdgpu_dpm_funcs si_dpm_funcs = {
+ .get_temperature = &si_dpm_get_temp,
+ .pre_set_power_state = &si_dpm_pre_set_power_state,
+ .set_power_state = &si_dpm_set_power_state,
+ .post_set_power_state = &si_dpm_post_set_power_state,
+ .display_configuration_changed = &si_dpm_display_configuration_changed,
+ .get_sclk = &si_dpm_get_sclk,
+ .get_mclk = &si_dpm_get_mclk,
+ .print_power_state = &si_dpm_print_power_state,
+ .debugfs_print_current_performance_level = &si_dpm_debugfs_print_current_performance_level,
+ .force_performance_level = &si_dpm_force_performance_level,
+ .vblank_too_short = &si_dpm_vblank_too_short,
+ .set_fan_control_mode = &si_dpm_set_fan_control_mode,
+ .get_fan_control_mode = &si_dpm_get_fan_control_mode,
+ .set_fan_speed_percent = &si_dpm_set_fan_speed_percent,
+ .get_fan_speed_percent = &si_dpm_get_fan_speed_percent,
+};
+
+static void si_dpm_set_dpm_funcs(struct amdgpu_device *adev)
+{
+ if (adev->pm.funcs == NULL)
+ adev->pm.funcs = &si_dpm_funcs;
+}
+
+static const struct amdgpu_irq_src_funcs si_dpm_irq_funcs = {
+ .set = si_dpm_set_interrupt_state,
+ .process = si_dpm_process_interrupt,
+};
+
+static void si_dpm_set_irq_funcs(struct amdgpu_device *adev)
+{
+ adev->pm.dpm.thermal.irq.num_types = AMDGPU_THERMAL_IRQ_LAST;
+ adev->pm.dpm.thermal.irq.funcs = &si_dpm_irq_funcs;
+}
+
--- /dev/null
+/*
+ * Copyright 2012 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __SI_DPM_H__
+#define __SI_DPM_H__
+
+#include "amdgpu_atombios.h"
+#include "sislands_smc.h"
+
+#define MC_CG_CONFIG 0x96f
+#define MC_ARB_CG 0x9fa
+#define CG_ARB_REQ(x) ((x) << 0)
+#define CG_ARB_REQ_MASK (0xff << 0)
+
+#define MC_ARB_DRAM_TIMING_1 0x9fc
+#define MC_ARB_DRAM_TIMING_2 0x9fd
+#define MC_ARB_DRAM_TIMING_3 0x9fe
+#define MC_ARB_DRAM_TIMING2_1 0x9ff
+#define MC_ARB_DRAM_TIMING2_2 0xa00
+#define MC_ARB_DRAM_TIMING2_3 0xa01
+
+#define MAX_NO_OF_MVDD_VALUES 2
+#define MAX_NO_VREG_STEPS 32
+#define NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 16
+#define SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE 32
+#define SMC_NISLANDS_MC_REGISTER_ARRAY_SET_COUNT 20
+#define RV770_ASI_DFLT 1000
+#define CYPRESS_HASI_DFLT 400000
+#define PCIE_PERF_REQ_PECI_GEN1 2
+#define PCIE_PERF_REQ_PECI_GEN2 3
+#define PCIE_PERF_REQ_PECI_GEN3 4
+#define RV770_DEFAULT_VCLK_FREQ 53300 /* 10 khz */
+#define RV770_DEFAULT_DCLK_FREQ 40000 /* 10 khz */
+
+#define SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE 16
+
+#define RV770_SMC_TABLE_ADDRESS 0xB000
+#define RV770_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 3
+
+#define SMC_STROBE_RATIO 0x0F
+#define SMC_STROBE_ENABLE 0x10
+
+#define SMC_MC_EDC_RD_FLAG 0x01
+#define SMC_MC_EDC_WR_FLAG 0x02
+#define SMC_MC_RTT_ENABLE 0x04
+#define SMC_MC_STUTTER_EN 0x08
+
+#define RV770_SMC_VOLTAGEMASK_VDDC 0
+#define RV770_SMC_VOLTAGEMASK_MVDD 1
+#define RV770_SMC_VOLTAGEMASK_VDDCI 2
+#define RV770_SMC_VOLTAGEMASK_MAX 4
+
+#define NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 16
+#define NISLANDS_SMC_STROBE_RATIO 0x0F
+#define NISLANDS_SMC_STROBE_ENABLE 0x10
+
+#define NISLANDS_SMC_MC_EDC_RD_FLAG 0x01
+#define NISLANDS_SMC_MC_EDC_WR_FLAG 0x02
+#define NISLANDS_SMC_MC_RTT_ENABLE 0x04
+#define NISLANDS_SMC_MC_STUTTER_EN 0x08
+
+#define MAX_NO_VREG_STEPS 32
+
+#define NISLANDS_SMC_VOLTAGEMASK_VDDC 0
+#define NISLANDS_SMC_VOLTAGEMASK_MVDD 1
+#define NISLANDS_SMC_VOLTAGEMASK_VDDCI 2
+#define NISLANDS_SMC_VOLTAGEMASK_MAX 4
+
+#define SISLANDS_MCREGISTERTABLE_INITIAL_SLOT 0
+#define SISLANDS_MCREGISTERTABLE_ACPI_SLOT 1
+#define SISLANDS_MCREGISTERTABLE_ULV_SLOT 2
+#define SISLANDS_MCREGISTERTABLE_FIRST_DRIVERSTATE_SLOT 3
+
+#define SISLANDS_LEAKAGE_INDEX0 0xff01
+#define SISLANDS_MAX_LEAKAGE_COUNT 4
+
+#define SISLANDS_MAX_HARDWARE_POWERLEVELS 5
+#define SISLANDS_INITIAL_STATE_ARB_INDEX 0
+#define SISLANDS_ACPI_STATE_ARB_INDEX 1
+#define SISLANDS_ULV_STATE_ARB_INDEX 2
+#define SISLANDS_DRIVER_STATE_ARB_INDEX 3
+
+#define SISLANDS_DPM2_MAX_PULSE_SKIP 256
+
+#define SISLANDS_DPM2_NEAR_TDP_DEC 10
+#define SISLANDS_DPM2_ABOVE_SAFE_INC 5
+#define SISLANDS_DPM2_BELOW_SAFE_INC 20
+
+#define SISLANDS_DPM2_TDP_SAFE_LIMIT_PERCENT 80
+
+#define SISLANDS_DPM2_MAXPS_PERCENT_H 99
+#define SISLANDS_DPM2_MAXPS_PERCENT_M 99
+
+#define SISLANDS_DPM2_SQ_RAMP_MAX_POWER 0x3FFF
+#define SISLANDS_DPM2_SQ_RAMP_MIN_POWER 0x12
+#define SISLANDS_DPM2_SQ_RAMP_MAX_POWER_DELTA 0x15
+#define SISLANDS_DPM2_SQ_RAMP_STI_SIZE 0x1E
+#define SISLANDS_DPM2_SQ_RAMP_LTI_RATIO 0xF
+
+#define SISLANDS_DPM2_PWREFFICIENCYRATIO_MARGIN 10
+
+#define SISLANDS_VRC_DFLT 0xC000B3
+#define SISLANDS_ULVVOLTAGECHANGEDELAY_DFLT 1687
+#define SISLANDS_CGULVPARAMETER_DFLT 0x00040035
+#define SISLANDS_CGULVCONTROL_DFLT 0x1f007550
+
+#define SI_ASI_DFLT 10000
+#define SI_BSP_DFLT 0x41EB
+#define SI_BSU_DFLT 0x2
+#define SI_AH_DFLT 5
+#define SI_RLP_DFLT 25
+#define SI_RMP_DFLT 65
+#define SI_LHP_DFLT 40
+#define SI_LMP_DFLT 15
+#define SI_TD_DFLT 0
+#define SI_UTC_DFLT_00 0x24
+#define SI_UTC_DFLT_01 0x22
+#define SI_UTC_DFLT_02 0x22
+#define SI_UTC_DFLT_03 0x22
+#define SI_UTC_DFLT_04 0x22
+#define SI_UTC_DFLT_05 0x22
+#define SI_UTC_DFLT_06 0x22
+#define SI_UTC_DFLT_07 0x22
+#define SI_UTC_DFLT_08 0x22
+#define SI_UTC_DFLT_09 0x22
+#define SI_UTC_DFLT_10 0x22
+#define SI_UTC_DFLT_11 0x22
+#define SI_UTC_DFLT_12 0x22
+#define SI_UTC_DFLT_13 0x22
+#define SI_UTC_DFLT_14 0x22
+#define SI_DTC_DFLT_00 0x24
+#define SI_DTC_DFLT_01 0x22
+#define SI_DTC_DFLT_02 0x22
+#define SI_DTC_DFLT_03 0x22
+#define SI_DTC_DFLT_04 0x22
+#define SI_DTC_DFLT_05 0x22
+#define SI_DTC_DFLT_06 0x22
+#define SI_DTC_DFLT_07 0x22
+#define SI_DTC_DFLT_08 0x22
+#define SI_DTC_DFLT_09 0x22
+#define SI_DTC_DFLT_10 0x22
+#define SI_DTC_DFLT_11 0x22
+#define SI_DTC_DFLT_12 0x22
+#define SI_DTC_DFLT_13 0x22
+#define SI_DTC_DFLT_14 0x22
+#define SI_VRC_DFLT 0x0000C003
+#define SI_VOLTAGERESPONSETIME_DFLT 1000
+#define SI_BACKBIASRESPONSETIME_DFLT 1000
+#define SI_VRU_DFLT 0x3
+#define SI_SPLLSTEPTIME_DFLT 0x1000
+#define SI_SPLLSTEPUNIT_DFLT 0x3
+#define SI_TPU_DFLT 0
+#define SI_TPC_DFLT 0x200
+#define SI_SSTU_DFLT 0
+#define SI_SST_DFLT 0x00C8
+#define SI_GICST_DFLT 0x200
+#define SI_FCT_DFLT 0x0400
+#define SI_FCTU_DFLT 0
+#define SI_CTXCGTT3DRPHC_DFLT 0x20
+#define SI_CTXCGTT3DRSDC_DFLT 0x40
+#define SI_VDDC3DOORPHC_DFLT 0x100
+#define SI_VDDC3DOORSDC_DFLT 0x7
+#define SI_VDDC3DOORSU_DFLT 0
+#define SI_MPLLLOCKTIME_DFLT 100
+#define SI_MPLLRESETTIME_DFLT 150
+#define SI_VCOSTEPPCT_DFLT 20
+#define SI_ENDINGVCOSTEPPCT_DFLT 5
+#define SI_REFERENCEDIVIDER_DFLT 4
+
+#define SI_PM_NUMBER_OF_TC 15
+#define SI_PM_NUMBER_OF_SCLKS 20
+#define SI_PM_NUMBER_OF_MCLKS 4
+#define SI_PM_NUMBER_OF_VOLTAGE_LEVELS 4
+#define SI_PM_NUMBER_OF_ACTIVITY_LEVELS 3
+
+/* XXX are these ok? */
+#define SI_TEMP_RANGE_MIN (90 * 1000)
+#define SI_TEMP_RANGE_MAX (120 * 1000)
+
+#define FDO_PWM_MODE_STATIC 1
+#define FDO_PWM_MODE_STATIC_RPM 5
+
+enum ni_dc_cac_level
+{
+ NISLANDS_DCCAC_LEVEL_0 = 0,
+ NISLANDS_DCCAC_LEVEL_1,
+ NISLANDS_DCCAC_LEVEL_2,
+ NISLANDS_DCCAC_LEVEL_3,
+ NISLANDS_DCCAC_LEVEL_4,
+ NISLANDS_DCCAC_LEVEL_5,
+ NISLANDS_DCCAC_LEVEL_6,
+ NISLANDS_DCCAC_LEVEL_7,
+ NISLANDS_DCCAC_MAX_LEVELS
+};
+
+enum si_cac_config_reg_type
+{
+ SISLANDS_CACCONFIG_MMR = 0,
+ SISLANDS_CACCONFIG_CGIND,
+ SISLANDS_CACCONFIG_MAX
+};
+
+enum si_power_level {
+ SI_POWER_LEVEL_LOW = 0,
+ SI_POWER_LEVEL_MEDIUM = 1,
+ SI_POWER_LEVEL_HIGH = 2,
+ SI_POWER_LEVEL_CTXSW = 3,
+};
+
+enum si_td {
+ SI_TD_AUTO,
+ SI_TD_UP,
+ SI_TD_DOWN,
+};
+
+enum si_display_watermark {
+ SI_DISPLAY_WATERMARK_LOW = 0,
+ SI_DISPLAY_WATERMARK_HIGH = 1,
+};
+
+enum si_display_gap
+{
+ SI_PM_DISPLAY_GAP_VBLANK_OR_WM = 0,
+ SI_PM_DISPLAY_GAP_VBLANK = 1,
+ SI_PM_DISPLAY_GAP_WATERMARK = 2,
+ SI_PM_DISPLAY_GAP_IGNORE = 3,
+};
+
+extern const struct amd_ip_funcs si_dpm_ip_funcs;
+
+struct ni_leakage_coeffients
+{
+ u32 at;
+ u32 bt;
+ u32 av;
+ u32 bv;
+ s32 t_slope;
+ s32 t_intercept;
+ u32 t_ref;
+};
+
+struct SMC_Evergreen_MCRegisterAddress
+{
+ uint16_t s0;
+ uint16_t s1;
+};
+
+typedef struct SMC_Evergreen_MCRegisterAddress SMC_Evergreen_MCRegisterAddress;
+
+struct evergreen_mc_reg_entry {
+ u32 mclk_max;
+ u32 mc_data[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE];
+};
+
+struct evergreen_mc_reg_table {
+ u8 last;
+ u8 num_entries;
+ u16 valid_flag;
+ struct evergreen_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES];
+ SMC_Evergreen_MCRegisterAddress mc_reg_address[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE];
+};
+
+struct SMC_Evergreen_MCRegisterSet
+{
+ uint32_t value[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE];
+};
+
+typedef struct SMC_Evergreen_MCRegisterSet SMC_Evergreen_MCRegisterSet;
+
+struct SMC_Evergreen_MCRegisters
+{
+ uint8_t last;
+ uint8_t reserved[3];
+ SMC_Evergreen_MCRegisterAddress address[SMC_EVERGREEN_MC_REGISTER_ARRAY_SIZE];
+ SMC_Evergreen_MCRegisterSet data[5];
+};
+
+typedef struct SMC_Evergreen_MCRegisters SMC_Evergreen_MCRegisters;
+
+struct SMC_NIslands_MCRegisterSet
+{
+ uint32_t value[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE];
+};
+
+typedef struct SMC_NIslands_MCRegisterSet SMC_NIslands_MCRegisterSet;
+
+struct ni_mc_reg_entry {
+ u32 mclk_max;
+ u32 mc_data[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE];
+};
+
+struct SMC_NIslands_MCRegisterAddress
+{
+ uint16_t s0;
+ uint16_t s1;
+};
+
+typedef struct SMC_NIslands_MCRegisterAddress SMC_NIslands_MCRegisterAddress;
+
+struct SMC_NIslands_MCRegisters
+{
+ uint8_t last;
+ uint8_t reserved[3];
+ SMC_NIslands_MCRegisterAddress address[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE];
+ SMC_NIslands_MCRegisterSet data[SMC_NISLANDS_MC_REGISTER_ARRAY_SET_COUNT];
+};
+
+typedef struct SMC_NIslands_MCRegisters SMC_NIslands_MCRegisters;
+
+struct evergreen_ulv_param {
+ bool supported;
+ struct rv7xx_pl *pl;
+};
+
+struct evergreen_arb_registers {
+ u32 mc_arb_dram_timing;
+ u32 mc_arb_dram_timing2;
+ u32 mc_arb_rfsh_rate;
+ u32 mc_arb_burst_time;
+};
+
+struct at {
+ u32 rlp;
+ u32 rmp;
+ u32 lhp;
+ u32 lmp;
+};
+
+struct ni_clock_registers {
+ u32 cg_spll_func_cntl;
+ u32 cg_spll_func_cntl_2;
+ u32 cg_spll_func_cntl_3;
+ u32 cg_spll_func_cntl_4;
+ u32 cg_spll_spread_spectrum;
+ u32 cg_spll_spread_spectrum_2;
+ u32 mclk_pwrmgt_cntl;
+ u32 dll_cntl;
+ u32 mpll_ad_func_cntl;
+ u32 mpll_ad_func_cntl_2;
+ u32 mpll_dq_func_cntl;
+ u32 mpll_dq_func_cntl_2;
+ u32 mpll_ss1;
+ u32 mpll_ss2;
+};
+
+struct RV770_SMC_SCLK_VALUE
+{
+ uint32_t vCG_SPLL_FUNC_CNTL;
+ uint32_t vCG_SPLL_FUNC_CNTL_2;
+ uint32_t vCG_SPLL_FUNC_CNTL_3;
+ uint32_t vCG_SPLL_SPREAD_SPECTRUM;
+ uint32_t vCG_SPLL_SPREAD_SPECTRUM_2;
+ uint32_t sclk_value;
+};
+
+typedef struct RV770_SMC_SCLK_VALUE RV770_SMC_SCLK_VALUE;
+
+struct RV770_SMC_MCLK_VALUE
+{
+ uint32_t vMPLL_AD_FUNC_CNTL;
+ uint32_t vMPLL_AD_FUNC_CNTL_2;
+ uint32_t vMPLL_DQ_FUNC_CNTL;
+ uint32_t vMPLL_DQ_FUNC_CNTL_2;
+ uint32_t vMCLK_PWRMGT_CNTL;
+ uint32_t vDLL_CNTL;
+ uint32_t vMPLL_SS;
+ uint32_t vMPLL_SS2;
+ uint32_t mclk_value;
+};
+
+typedef struct RV770_SMC_MCLK_VALUE RV770_SMC_MCLK_VALUE;
+
+
+struct RV730_SMC_MCLK_VALUE
+{
+ uint32_t vMCLK_PWRMGT_CNTL;
+ uint32_t vDLL_CNTL;
+ uint32_t vMPLL_FUNC_CNTL;
+ uint32_t vMPLL_FUNC_CNTL2;
+ uint32_t vMPLL_FUNC_CNTL3;
+ uint32_t vMPLL_SS;
+ uint32_t vMPLL_SS2;
+ uint32_t mclk_value;
+};
+
+typedef struct RV730_SMC_MCLK_VALUE RV730_SMC_MCLK_VALUE;
+
+struct RV770_SMC_VOLTAGE_VALUE
+{
+ uint16_t value;
+ uint8_t index;
+ uint8_t padding;
+};
+
+typedef struct RV770_SMC_VOLTAGE_VALUE RV770_SMC_VOLTAGE_VALUE;
+
+union RV7XX_SMC_MCLK_VALUE
+{
+ RV770_SMC_MCLK_VALUE mclk770;
+ RV730_SMC_MCLK_VALUE mclk730;
+};
+
+typedef union RV7XX_SMC_MCLK_VALUE RV7XX_SMC_MCLK_VALUE, *LPRV7XX_SMC_MCLK_VALUE;
+
+struct RV770_SMC_HW_PERFORMANCE_LEVEL
+{
+ uint8_t arbValue;
+ union{
+ uint8_t seqValue;
+ uint8_t ACIndex;
+ };
+ uint8_t displayWatermark;
+ uint8_t gen2PCIE;
+ uint8_t gen2XSP;
+ uint8_t backbias;
+ uint8_t strobeMode;
+ uint8_t mcFlags;
+ uint32_t aT;
+ uint32_t bSP;
+ RV770_SMC_SCLK_VALUE sclk;
+ RV7XX_SMC_MCLK_VALUE mclk;
+ RV770_SMC_VOLTAGE_VALUE vddc;
+ RV770_SMC_VOLTAGE_VALUE mvdd;
+ RV770_SMC_VOLTAGE_VALUE vddci;
+ uint8_t reserved1;
+ uint8_t reserved2;
+ uint8_t stateFlags;
+ uint8_t padding;
+};
+
+typedef struct RV770_SMC_HW_PERFORMANCE_LEVEL RV770_SMC_HW_PERFORMANCE_LEVEL;
+
+struct RV770_SMC_SWSTATE
+{
+ uint8_t flags;
+ uint8_t padding1;
+ uint8_t padding2;
+ uint8_t padding3;
+ RV770_SMC_HW_PERFORMANCE_LEVEL levels[RV770_SMC_PERFORMANCE_LEVELS_PER_SWSTATE];
+};
+
+typedef struct RV770_SMC_SWSTATE RV770_SMC_SWSTATE;
+
+struct RV770_SMC_VOLTAGEMASKTABLE
+{
+ uint8_t highMask[RV770_SMC_VOLTAGEMASK_MAX];
+ uint32_t lowMask[RV770_SMC_VOLTAGEMASK_MAX];
+};
+
+typedef struct RV770_SMC_VOLTAGEMASKTABLE RV770_SMC_VOLTAGEMASKTABLE;
+
+struct RV770_SMC_STATETABLE
+{
+ uint8_t thermalProtectType;
+ uint8_t systemFlags;
+ uint8_t maxVDDCIndexInPPTable;
+ uint8_t extraFlags;
+ uint8_t highSMIO[MAX_NO_VREG_STEPS];
+ uint32_t lowSMIO[MAX_NO_VREG_STEPS];
+ RV770_SMC_VOLTAGEMASKTABLE voltageMaskTable;
+ RV770_SMC_SWSTATE initialState;
+ RV770_SMC_SWSTATE ACPIState;
+ RV770_SMC_SWSTATE driverState;
+ RV770_SMC_SWSTATE ULVState;
+};
+
+typedef struct RV770_SMC_STATETABLE RV770_SMC_STATETABLE;
+
+struct vddc_table_entry {
+ u16 vddc;
+ u8 vddc_index;
+ u8 high_smio;
+ u32 low_smio;
+};
+
+struct rv770_clock_registers {
+ u32 cg_spll_func_cntl;
+ u32 cg_spll_func_cntl_2;
+ u32 cg_spll_func_cntl_3;
+ u32 cg_spll_spread_spectrum;
+ u32 cg_spll_spread_spectrum_2;
+ u32 mpll_ad_func_cntl;
+ u32 mpll_ad_func_cntl_2;
+ u32 mpll_dq_func_cntl;
+ u32 mpll_dq_func_cntl_2;
+ u32 mclk_pwrmgt_cntl;
+ u32 dll_cntl;
+ u32 mpll_ss1;
+ u32 mpll_ss2;
+};
+
+struct rv730_clock_registers {
+ u32 cg_spll_func_cntl;
+ u32 cg_spll_func_cntl_2;
+ u32 cg_spll_func_cntl_3;
+ u32 cg_spll_spread_spectrum;
+ u32 cg_spll_spread_spectrum_2;
+ u32 mclk_pwrmgt_cntl;
+ u32 dll_cntl;
+ u32 mpll_func_cntl;
+ u32 mpll_func_cntl2;
+ u32 mpll_func_cntl3;
+ u32 mpll_ss;
+ u32 mpll_ss2;
+};
+
+union r7xx_clock_registers {
+ struct rv770_clock_registers rv770;
+ struct rv730_clock_registers rv730;
+};
+
+struct rv7xx_power_info {
+ /* flags */
+ bool mem_gddr5;
+ bool pcie_gen2;
+ bool dynamic_pcie_gen2;
+ bool acpi_pcie_gen2;
+ bool boot_in_gen2;
+ bool voltage_control; /* vddc */
+ bool mvdd_control;
+ bool sclk_ss;
+ bool mclk_ss;
+ bool dynamic_ss;
+ bool gfx_clock_gating;
+ bool mg_clock_gating;
+ bool mgcgtssm;
+ bool power_gating;
+ bool thermal_protection;
+ bool display_gap;
+ bool dcodt;
+ bool ulps;
+ /* registers */
+ union r7xx_clock_registers clk_regs;
+ u32 s0_vid_lower_smio_cntl;
+ /* voltage */
+ u32 vddc_mask_low;
+ u32 mvdd_mask_low;
+ u32 mvdd_split_frequency;
+ u32 mvdd_low_smio[MAX_NO_OF_MVDD_VALUES];
+ u16 max_vddc;
+ u16 max_vddc_in_table;
+ u16 min_vddc_in_table;
+ struct vddc_table_entry vddc_table[MAX_NO_VREG_STEPS];
+ u8 valid_vddc_entries;
+ /* dc odt */
+ u32 mclk_odt_threshold;
+ u8 odt_value_0[2];
+ u8 odt_value_1[2];
+ /* stored values */
+ u32 boot_sclk;
+ u16 acpi_vddc;
+ u32 ref_div;
+ u32 active_auto_throttle_sources;
+ u32 mclk_stutter_mode_threshold;
+ u32 mclk_strobe_mode_threshold;
+ u32 mclk_edc_enable_threshold;
+ u32 bsp;
+ u32 bsu;
+ u32 pbsp;
+ u32 pbsu;
+ u32 dsp;
+ u32 psp;
+ u32 asi;
+ u32 pasi;
+ u32 vrc;
+ u32 restricted_levels;
+ u32 rlp;
+ u32 rmp;
+ u32 lhp;
+ u32 lmp;
+ /* smc offsets */
+ u16 state_table_start;
+ u16 soft_regs_start;
+ u16 sram_end;
+ /* scratch structs */
+ RV770_SMC_STATETABLE smc_statetable;
+};
+
+struct rv7xx_pl {
+ u32 sclk;
+ u32 mclk;
+ u16 vddc;
+ u16 vddci; /* eg+ only */
+ u32 flags;
+ enum amdgpu_pcie_gen pcie_gen; /* si+ only */
+};
+
+struct rv7xx_ps {
+ struct rv7xx_pl high;
+ struct rv7xx_pl medium;
+ struct rv7xx_pl low;
+ bool dc_compatible;
+};
+
+struct si_ps {
+ u16 performance_level_count;
+ bool dc_compatible;
+ struct rv7xx_pl performance_levels[NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE];
+};
+
+struct ni_mc_reg_table {
+ u8 last;
+ u8 num_entries;
+ u16 valid_flag;
+ struct ni_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES];
+ SMC_NIslands_MCRegisterAddress mc_reg_address[SMC_NISLANDS_MC_REGISTER_ARRAY_SIZE];
+};
+
+struct ni_cac_data
+{
+ struct ni_leakage_coeffients leakage_coefficients;
+ u32 i_leakage;
+ s32 leakage_minimum_temperature;
+ u32 pwr_const;
+ u32 dc_cac_value;
+ u32 bif_cac_value;
+ u32 lkge_pwr;
+ u8 mc_wr_weight;
+ u8 mc_rd_weight;
+ u8 allow_ovrflw;
+ u8 num_win_tdp;
+ u8 l2num_win_tdp;
+ u8 lts_truncate_n;
+};
+
+struct evergreen_power_info {
+ /* must be first! */
+ struct rv7xx_power_info rv7xx;
+ /* flags */
+ bool vddci_control;
+ bool dynamic_ac_timing;
+ bool abm;
+ bool mcls;
+ bool light_sleep;
+ bool memory_transition;
+ bool pcie_performance_request;
+ bool pcie_performance_request_registered;
+ bool sclk_deep_sleep;
+ bool dll_default_on;
+ bool ls_clock_gating;
+ bool smu_uvd_hs;
+ bool uvd_enabled;
+ /* stored values */
+ u16 acpi_vddci;
+ u8 mvdd_high_index;
+ u8 mvdd_low_index;
+ u32 mclk_edc_wr_enable_threshold;
+ struct evergreen_mc_reg_table mc_reg_table;
+ struct atom_voltage_table vddc_voltage_table;
+ struct atom_voltage_table vddci_voltage_table;
+ struct evergreen_arb_registers bootup_arb_registers;
+ struct evergreen_ulv_param ulv;
+ struct at ats[2];
+ /* smc offsets */
+ u16 mc_reg_table_start;
+ struct amdgpu_ps current_rps;
+ struct rv7xx_ps current_ps;
+ struct amdgpu_ps requested_rps;
+ struct rv7xx_ps requested_ps;
+};
+
+struct PP_NIslands_Dpm2PerfLevel
+{
+ uint8_t MaxPS;
+ uint8_t TgtAct;
+ uint8_t MaxPS_StepInc;
+ uint8_t MaxPS_StepDec;
+ uint8_t PSST;
+ uint8_t NearTDPDec;
+ uint8_t AboveSafeInc;
+ uint8_t BelowSafeInc;
+ uint8_t PSDeltaLimit;
+ uint8_t PSDeltaWin;
+ uint8_t Reserved[6];
+};
+
+typedef struct PP_NIslands_Dpm2PerfLevel PP_NIslands_Dpm2PerfLevel;
+
+struct PP_NIslands_DPM2Parameters
+{
+ uint32_t TDPLimit;
+ uint32_t NearTDPLimit;
+ uint32_t SafePowerLimit;
+ uint32_t PowerBoostLimit;
+};
+typedef struct PP_NIslands_DPM2Parameters PP_NIslands_DPM2Parameters;
+
+struct NISLANDS_SMC_SCLK_VALUE
+{
+ uint32_t vCG_SPLL_FUNC_CNTL;
+ uint32_t vCG_SPLL_FUNC_CNTL_2;
+ uint32_t vCG_SPLL_FUNC_CNTL_3;
+ uint32_t vCG_SPLL_FUNC_CNTL_4;
+ uint32_t vCG_SPLL_SPREAD_SPECTRUM;
+ uint32_t vCG_SPLL_SPREAD_SPECTRUM_2;
+ uint32_t sclk_value;
+};
+
+typedef struct NISLANDS_SMC_SCLK_VALUE NISLANDS_SMC_SCLK_VALUE;
+
+struct NISLANDS_SMC_MCLK_VALUE
+{
+ uint32_t vMPLL_FUNC_CNTL;
+ uint32_t vMPLL_FUNC_CNTL_1;
+ uint32_t vMPLL_FUNC_CNTL_2;
+ uint32_t vMPLL_AD_FUNC_CNTL;
+ uint32_t vMPLL_AD_FUNC_CNTL_2;
+ uint32_t vMPLL_DQ_FUNC_CNTL;
+ uint32_t vMPLL_DQ_FUNC_CNTL_2;
+ uint32_t vMCLK_PWRMGT_CNTL;
+ uint32_t vDLL_CNTL;
+ uint32_t vMPLL_SS;
+ uint32_t vMPLL_SS2;
+ uint32_t mclk_value;
+};
+
+typedef struct NISLANDS_SMC_MCLK_VALUE NISLANDS_SMC_MCLK_VALUE;
+
+struct NISLANDS_SMC_VOLTAGE_VALUE
+{
+ uint16_t value;
+ uint8_t index;
+ uint8_t padding;
+};
+
+typedef struct NISLANDS_SMC_VOLTAGE_VALUE NISLANDS_SMC_VOLTAGE_VALUE;
+
+struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL
+{
+ uint8_t arbValue;
+ uint8_t ACIndex;
+ uint8_t displayWatermark;
+ uint8_t gen2PCIE;
+ uint8_t reserved1;
+ uint8_t reserved2;
+ uint8_t strobeMode;
+ uint8_t mcFlags;
+ uint32_t aT;
+ uint32_t bSP;
+ NISLANDS_SMC_SCLK_VALUE sclk;
+ NISLANDS_SMC_MCLK_VALUE mclk;
+ NISLANDS_SMC_VOLTAGE_VALUE vddc;
+ NISLANDS_SMC_VOLTAGE_VALUE mvdd;
+ NISLANDS_SMC_VOLTAGE_VALUE vddci;
+ NISLANDS_SMC_VOLTAGE_VALUE std_vddc;
+ uint32_t powergate_en;
+ uint8_t hUp;
+ uint8_t hDown;
+ uint8_t stateFlags;
+ uint8_t arbRefreshState;
+ uint32_t SQPowerThrottle;
+ uint32_t SQPowerThrottle_2;
+ uint32_t reserved[2];
+ PP_NIslands_Dpm2PerfLevel dpm2;
+};
+
+typedef struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL NISLANDS_SMC_HW_PERFORMANCE_LEVEL;
+
+struct NISLANDS_SMC_SWSTATE
+{
+ uint8_t flags;
+ uint8_t levelCount;
+ uint8_t padding2;
+ uint8_t padding3;
+ NISLANDS_SMC_HW_PERFORMANCE_LEVEL levels[1];
+};
+
+typedef struct NISLANDS_SMC_SWSTATE NISLANDS_SMC_SWSTATE;
+
+struct NISLANDS_SMC_VOLTAGEMASKTABLE
+{
+ uint8_t highMask[NISLANDS_SMC_VOLTAGEMASK_MAX];
+ uint32_t lowMask[NISLANDS_SMC_VOLTAGEMASK_MAX];
+};
+
+typedef struct NISLANDS_SMC_VOLTAGEMASKTABLE NISLANDS_SMC_VOLTAGEMASKTABLE;
+
+#define NISLANDS_MAX_NO_VREG_STEPS 32
+
+struct NISLANDS_SMC_STATETABLE
+{
+ uint8_t thermalProtectType;
+ uint8_t systemFlags;
+ uint8_t maxVDDCIndexInPPTable;
+ uint8_t extraFlags;
+ uint8_t highSMIO[NISLANDS_MAX_NO_VREG_STEPS];
+ uint32_t lowSMIO[NISLANDS_MAX_NO_VREG_STEPS];
+ NISLANDS_SMC_VOLTAGEMASKTABLE voltageMaskTable;
+ PP_NIslands_DPM2Parameters dpm2Params;
+ NISLANDS_SMC_SWSTATE initialState;
+ NISLANDS_SMC_SWSTATE ACPIState;
+ NISLANDS_SMC_SWSTATE ULVState;
+ NISLANDS_SMC_SWSTATE driverState;
+ NISLANDS_SMC_HW_PERFORMANCE_LEVEL dpmLevels[NISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE - 1];
+};
+
+typedef struct NISLANDS_SMC_STATETABLE NISLANDS_SMC_STATETABLE;
+
+struct ni_power_info {
+ /* must be first! */
+ struct evergreen_power_info eg;
+ struct ni_clock_registers clock_registers;
+ struct ni_mc_reg_table mc_reg_table;
+ u32 mclk_rtt_mode_threshold;
+ /* flags */
+ bool use_power_boost_limit;
+ bool support_cac_long_term_average;
+ bool cac_enabled;
+ bool cac_configuration_required;
+ bool driver_calculate_cac_leakage;
+ bool pc_enabled;
+ bool enable_power_containment;
+ bool enable_cac;
+ bool enable_sq_ramping;
+ /* smc offsets */
+ u16 arb_table_start;
+ u16 fan_table_start;
+ u16 cac_table_start;
+ u16 spll_table_start;
+ /* CAC stuff */
+ struct ni_cac_data cac_data;
+ u32 dc_cac_table[NISLANDS_DCCAC_MAX_LEVELS];
+ const struct ni_cac_weights *cac_weights;
+ u8 lta_window_size;
+ u8 lts_truncate;
+ struct si_ps current_ps;
+ struct si_ps requested_ps;
+ /* scratch structs */
+ SMC_NIslands_MCRegisters smc_mc_reg_table;
+ NISLANDS_SMC_STATETABLE smc_statetable;
+};
+
+struct si_cac_config_reg
+{
+ u32 offset;
+ u32 mask;
+ u32 shift;
+ u32 value;
+ enum si_cac_config_reg_type type;
+};
+
+struct si_powertune_data
+{
+ u32 cac_window;
+ u32 l2_lta_window_size_default;
+ u8 lts_truncate_default;
+ u8 shift_n_default;
+ u8 operating_temp;
+ struct ni_leakage_coeffients leakage_coefficients;
+ u32 fixed_kt;
+ u32 lkge_lut_v0_percent;
+ u8 dc_cac[NISLANDS_DCCAC_MAX_LEVELS];
+ bool enable_powertune_by_default;
+};
+
+struct si_dyn_powertune_data
+{
+ u32 cac_leakage;
+ s32 leakage_minimum_temperature;
+ u32 wintime;
+ u32 l2_lta_window_size;
+ u8 lts_truncate;
+ u8 shift_n;
+ u8 dc_pwr_value;
+ bool disable_uvd_powertune;
+};
+
+struct si_dte_data
+{
+ u32 tau[SMC_SISLANDS_DTE_MAX_FILTER_STAGES];
+ u32 r[SMC_SISLANDS_DTE_MAX_FILTER_STAGES];
+ u32 k;
+ u32 t0;
+ u32 max_t;
+ u8 window_size;
+ u8 temp_select;
+ u8 dte_mode;
+ u8 tdep_count;
+ u8 t_limits[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE];
+ u32 tdep_tau[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE];
+ u32 tdep_r[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE];
+ u32 t_threshold;
+ bool enable_dte_by_default;
+};
+
+struct si_clock_registers {
+ u32 cg_spll_func_cntl;
+ u32 cg_spll_func_cntl_2;
+ u32 cg_spll_func_cntl_3;
+ u32 cg_spll_func_cntl_4;
+ u32 cg_spll_spread_spectrum;
+ u32 cg_spll_spread_spectrum_2;
+ u32 dll_cntl;
+ u32 mclk_pwrmgt_cntl;
+ u32 mpll_ad_func_cntl;
+ u32 mpll_dq_func_cntl;
+ u32 mpll_func_cntl;
+ u32 mpll_func_cntl_1;
+ u32 mpll_func_cntl_2;
+ u32 mpll_ss1;
+ u32 mpll_ss2;
+};
+
+struct si_mc_reg_entry {
+ u32 mclk_max;
+ u32 mc_data[SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE];
+};
+
+struct si_mc_reg_table {
+ u8 last;
+ u8 num_entries;
+ u16 valid_flag;
+ struct si_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES];
+ SMC_NIslands_MCRegisterAddress mc_reg_address[SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE];
+};
+
+struct si_leakage_voltage_entry
+{
+ u16 voltage;
+ u16 leakage_index;
+};
+
+struct si_leakage_voltage
+{
+ u16 count;
+ struct si_leakage_voltage_entry entries[SISLANDS_MAX_LEAKAGE_COUNT];
+};
+
+
+struct si_ulv_param {
+ bool supported;
+ u32 cg_ulv_control;
+ u32 cg_ulv_parameter;
+ u32 volt_change_delay;
+ struct rv7xx_pl pl;
+ bool one_pcie_lane_in_ulv;
+};
+
+struct si_power_info {
+ /* must be first! */
+ struct ni_power_info ni;
+ struct si_clock_registers clock_registers;
+ struct si_mc_reg_table mc_reg_table;
+ struct atom_voltage_table mvdd_voltage_table;
+ struct atom_voltage_table vddc_phase_shed_table;
+ struct si_leakage_voltage leakage_voltage;
+ u16 mvdd_bootup_value;
+ struct si_ulv_param ulv;
+ u32 max_cu;
+ /* pcie gen */
+ enum amdgpu_pcie_gen force_pcie_gen;
+ enum amdgpu_pcie_gen boot_pcie_gen;
+ enum amdgpu_pcie_gen acpi_pcie_gen;
+ u32 sys_pcie_mask;
+ /* flags */
+ bool enable_dte;
+ bool enable_ppm;
+ bool vddc_phase_shed_control;
+ bool pspp_notify_required;
+ bool sclk_deep_sleep_above_low;
+ bool voltage_control_svi2;
+ bool vddci_control_svi2;
+ /* smc offsets */
+ u32 sram_end;
+ u32 state_table_start;
+ u32 soft_regs_start;
+ u32 mc_reg_table_start;
+ u32 arb_table_start;
+ u32 cac_table_start;
+ u32 dte_table_start;
+ u32 spll_table_start;
+ u32 papm_cfg_table_start;
+ u32 fan_table_start;
+ /* CAC stuff */
+ const struct si_cac_config_reg *cac_weights;
+ const struct si_cac_config_reg *lcac_config;
+ const struct si_cac_config_reg *cac_override;
+ const struct si_powertune_data *powertune_data;
+ struct si_dyn_powertune_data dyn_powertune_data;
+ /* DTE stuff */
+ struct si_dte_data dte_data;
+ /* scratch structs */
+ SMC_SIslands_MCRegisters smc_mc_reg_table;
+ SISLANDS_SMC_STATETABLE smc_statetable;
+ PP_SIslands_PAPMParameters papm_parm;
+ /* SVI2 */
+ u8 svd_gpio_id;
+ u8 svc_gpio_id;
+ /* fan control */
+ bool fan_ctrl_is_in_default_mode;
+ u32 t_min;
+ u32 fan_ctrl_default_mode;
+ bool fan_is_controlled_by_smc;
+};
+
+#endif
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "drmP.h"
+#include "amdgpu.h"
+#include "amdgpu_ih.h"
+#include "si/sid.h"
+#include "si_ih.h"
+
+static void si_ih_set_interrupt_funcs(struct amdgpu_device *adev);
+
+static void si_ih_enable_interrupts(struct amdgpu_device *adev)
+{
+ u32 ih_cntl = RREG32(IH_CNTL);
+ u32 ih_rb_cntl = RREG32(IH_RB_CNTL);
+
+ ih_cntl |= ENABLE_INTR;
+ ih_rb_cntl |= IH_RB_ENABLE;
+ WREG32(IH_CNTL, ih_cntl);
+ WREG32(IH_RB_CNTL, ih_rb_cntl);
+ adev->irq.ih.enabled = true;
+}
+
+static void si_ih_disable_interrupts(struct amdgpu_device *adev)
+{
+ u32 ih_rb_cntl = RREG32(IH_RB_CNTL);
+ u32 ih_cntl = RREG32(IH_CNTL);
+
+ ih_rb_cntl &= ~IH_RB_ENABLE;
+ ih_cntl &= ~ENABLE_INTR;
+ WREG32(IH_RB_CNTL, ih_rb_cntl);
+ WREG32(IH_CNTL, ih_cntl);
+ WREG32(IH_RB_RPTR, 0);
+ WREG32(IH_RB_WPTR, 0);
+ adev->irq.ih.enabled = false;
+ adev->irq.ih.rptr = 0;
+}
+
+static int si_ih_irq_init(struct amdgpu_device *adev)
+{
+ int rb_bufsz;
+ u32 interrupt_cntl, ih_cntl, ih_rb_cntl;
+ u64 wptr_off;
+
+ si_ih_disable_interrupts(adev);
+ WREG32(INTERRUPT_CNTL2, adev->irq.ih.gpu_addr >> 8);
+ interrupt_cntl = RREG32(INTERRUPT_CNTL);
+ interrupt_cntl &= ~IH_DUMMY_RD_OVERRIDE;
+ interrupt_cntl &= ~IH_REQ_NONSNOOP_EN;
+ WREG32(INTERRUPT_CNTL, interrupt_cntl);
+
+ WREG32(IH_RB_BASE, adev->irq.ih.gpu_addr >> 8);
+ rb_bufsz = order_base_2(adev->irq.ih.ring_size / 4);
+
+ ih_rb_cntl = IH_WPTR_OVERFLOW_ENABLE |
+ IH_WPTR_OVERFLOW_CLEAR |
+ (rb_bufsz << 1) |
+ IH_WPTR_WRITEBACK_ENABLE;
+
+ wptr_off = adev->wb.gpu_addr + (adev->irq.ih.wptr_offs * 4);
+ WREG32(IH_RB_WPTR_ADDR_LO, lower_32_bits(wptr_off));
+ WREG32(IH_RB_WPTR_ADDR_HI, upper_32_bits(wptr_off) & 0xFF);
+ WREG32(IH_RB_CNTL, ih_rb_cntl);
+ WREG32(IH_RB_RPTR, 0);
+ WREG32(IH_RB_WPTR, 0);
+
+ ih_cntl = MC_WRREQ_CREDIT(0x10) | MC_WR_CLEAN_CNT(0x10) | MC_VMID(0);
+ if (adev->irq.msi_enabled)
+ ih_cntl |= RPTR_REARM;
+ WREG32(IH_CNTL, ih_cntl);
+
+ pci_set_master(adev->pdev);
+ si_ih_enable_interrupts(adev);
+
+ return 0;
+}
+
+static void si_ih_irq_disable(struct amdgpu_device *adev)
+{
+ si_ih_disable_interrupts(adev);
+ mdelay(1);
+}
+
+static u32 si_ih_get_wptr(struct amdgpu_device *adev)
+{
+ u32 wptr, tmp;
+
+ wptr = le32_to_cpu(adev->wb.wb[adev->irq.ih.wptr_offs]);
+
+ if (wptr & IH_RB_WPTR__RB_OVERFLOW_MASK) {
+ wptr &= ~IH_RB_WPTR__RB_OVERFLOW_MASK;
+ dev_warn(adev->dev, "IH ring buffer overflow (0x%08X, 0x%08X, 0x%08X)\n",
+ wptr, adev->irq.ih.rptr, (wptr + 16) & adev->irq.ih.ptr_mask);
+ adev->irq.ih.rptr = (wptr + 16) & adev->irq.ih.ptr_mask;
+ tmp = RREG32(IH_RB_CNTL);
+ tmp |= IH_RB_CNTL__WPTR_OVERFLOW_CLEAR_MASK;
+ WREG32(IH_RB_CNTL, tmp);
+ }
+ return (wptr & adev->irq.ih.ptr_mask);
+}
+
+static void si_ih_decode_iv(struct amdgpu_device *adev,
+ struct amdgpu_iv_entry *entry)
+{
+ u32 ring_index = adev->irq.ih.rptr >> 2;
+ uint32_t dw[4];
+
+ dw[0] = le32_to_cpu(adev->irq.ih.ring[ring_index + 0]);
+ dw[1] = le32_to_cpu(adev->irq.ih.ring[ring_index + 1]);
+ dw[2] = le32_to_cpu(adev->irq.ih.ring[ring_index + 2]);
+ dw[3] = le32_to_cpu(adev->irq.ih.ring[ring_index + 3]);
+
+ entry->src_id = dw[0] & 0xff;
+ entry->src_data = dw[1] & 0xfffffff;
+ entry->ring_id = dw[2] & 0xff;
+ entry->vm_id = (dw[2] >> 8) & 0xff;
+
+ adev->irq.ih.rptr += 16;
+}
+
+static void si_ih_set_rptr(struct amdgpu_device *adev)
+{
+ WREG32(IH_RB_RPTR, adev->irq.ih.rptr);
+}
+
+static int si_ih_early_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ si_ih_set_interrupt_funcs(adev);
+
+ return 0;
+}
+
+static int si_ih_sw_init(void *handle)
+{
+ int r;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ r = amdgpu_ih_ring_init(adev, 64 * 1024, false);
+ if (r)
+ return r;
+
+ return amdgpu_irq_init(adev);
+}
+
+static int si_ih_sw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ amdgpu_irq_fini(adev);
+ amdgpu_ih_ring_fini(adev);
+
+ return 0;
+}
+
+static int si_ih_hw_init(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return si_ih_irq_init(adev);
+}
+
+static int si_ih_hw_fini(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ si_ih_irq_disable(adev);
+
+ return 0;
+}
+
+static int si_ih_suspend(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return si_ih_hw_fini(adev);
+}
+
+static int si_ih_resume(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ return si_ih_hw_init(adev);
+}
+
+static bool si_ih_is_idle(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 tmp = RREG32(SRBM_STATUS);
+
+ if (tmp & SRBM_STATUS__IH_BUSY_MASK)
+ return false;
+
+ return true;
+}
+
+static int si_ih_wait_for_idle(void *handle)
+{
+ unsigned i;
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ if (si_ih_is_idle(handle))
+ return 0;
+ udelay(1);
+ }
+ return -ETIMEDOUT;
+}
+
+static int si_ih_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ u32 srbm_soft_reset = 0;
+ u32 tmp = RREG32(SRBM_STATUS);
+
+ if (tmp & SRBM_STATUS__IH_BUSY_MASK)
+ srbm_soft_reset |= SRBM_SOFT_RESET__SOFT_RESET_IH_MASK;
+
+ if (srbm_soft_reset) {
+ tmp = RREG32(SRBM_SOFT_RESET);
+ tmp |= srbm_soft_reset;
+ dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp);
+ WREG32(SRBM_SOFT_RESET, tmp);
+ tmp = RREG32(SRBM_SOFT_RESET);
+
+ udelay(50);
+
+ tmp &= ~srbm_soft_reset;
+ WREG32(SRBM_SOFT_RESET, tmp);
+ tmp = RREG32(SRBM_SOFT_RESET);
+
+ udelay(50);
+ }
+
+ return 0;
+}
+
+static int si_ih_set_clockgating_state(void *handle,
+ enum amd_clockgating_state state)
+{
+ return 0;
+}
+
+static int si_ih_set_powergating_state(void *handle,
+ enum amd_powergating_state state)
+{
+ return 0;
+}
+
+const struct amd_ip_funcs si_ih_ip_funcs = {
+ .name = "si_ih",
+ .early_init = si_ih_early_init,
+ .late_init = NULL,
+ .sw_init = si_ih_sw_init,
+ .sw_fini = si_ih_sw_fini,
+ .hw_init = si_ih_hw_init,
+ .hw_fini = si_ih_hw_fini,
+ .suspend = si_ih_suspend,
+ .resume = si_ih_resume,
+ .is_idle = si_ih_is_idle,
+ .wait_for_idle = si_ih_wait_for_idle,
+ .soft_reset = si_ih_soft_reset,
+ .set_clockgating_state = si_ih_set_clockgating_state,
+ .set_powergating_state = si_ih_set_powergating_state,
+};
+
+static const struct amdgpu_ih_funcs si_ih_funcs = {
+ .get_wptr = si_ih_get_wptr,
+ .decode_iv = si_ih_decode_iv,
+ .set_rptr = si_ih_set_rptr
+};
+
+static void si_ih_set_interrupt_funcs(struct amdgpu_device *adev)
+{
+ if (adev->irq.ih_funcs == NULL)
+ adev->irq.ih_funcs = &si_ih_funcs;
+}
+
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __SI_IH_H__
+#define __SI_IH_H__
+
+extern const struct amd_ip_funcs si_ih_ip_funcs;
+
+#endif
--- /dev/null
+/*
+ * Copyright 2011 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Alex Deucher
+ */
+
+#include <linux/firmware.h>
+#include "drmP.h"
+#include "amdgpu.h"
+#include "si/sid.h"
+#include "ppsmc.h"
+#include "amdgpu_ucode.h"
+#include "sislands_smc.h"
+
+static int si_set_smc_sram_address(struct amdgpu_device *adev,
+ u32 smc_address, u32 limit)
+{
+ if (smc_address & 3)
+ return -EINVAL;
+ if ((smc_address + 3) > limit)
+ return -EINVAL;
+
+ WREG32(SMC_IND_INDEX_0, smc_address);
+ WREG32_P(SMC_IND_ACCESS_CNTL, 0, ~AUTO_INCREMENT_IND_0);
+
+ return 0;
+}
+
+int amdgpu_si_copy_bytes_to_smc(struct amdgpu_device *adev,
+ u32 smc_start_address,
+ const u8 *src, u32 byte_count, u32 limit)
+{
+ unsigned long flags;
+ int ret = 0;
+ u32 data, original_data, addr, extra_shift;
+
+ if (smc_start_address & 3)
+ return -EINVAL;
+ if ((smc_start_address + byte_count) > limit)
+ return -EINVAL;
+
+ addr = smc_start_address;
+
+ spin_lock_irqsave(&adev->smc_idx_lock, flags);
+ while (byte_count >= 4) {
+ /* SMC address space is BE */
+ data = (src[0] << 24) | (src[1] << 16) | (src[2] << 8) | src[3];
+
+ ret = si_set_smc_sram_address(adev, addr, limit);
+ if (ret)
+ goto done;
+
+ WREG32(SMC_IND_DATA_0, data);
+
+ src += 4;
+ byte_count -= 4;
+ addr += 4;
+ }
+
+ /* RMW for the final bytes */
+ if (byte_count > 0) {
+ data = 0;
+
+ ret = si_set_smc_sram_address(adev, addr, limit);
+ if (ret)
+ goto done;
+
+ original_data = RREG32(SMC_IND_DATA_0);
+ extra_shift = 8 * (4 - byte_count);
+
+ while (byte_count > 0) {
+ /* SMC address space is BE */
+ data = (data << 8) + *src++;
+ byte_count--;
+ }
+
+ data <<= extra_shift;
+ data |= (original_data & ~((~0UL) << extra_shift));
+
+ ret = si_set_smc_sram_address(adev, addr, limit);
+ if (ret)
+ goto done;
+
+ WREG32(SMC_IND_DATA_0, data);
+ }
+
+done:
+ spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
+
+ return ret;
+}
+
+void amdgpu_si_start_smc(struct amdgpu_device *adev)
+{
+ u32 tmp = RREG32_SMC(SMC_SYSCON_RESET_CNTL);
+
+ tmp &= ~RST_REG;
+
+ WREG32_SMC(SMC_SYSCON_RESET_CNTL, tmp);
+}
+
+void amdgpu_si_reset_smc(struct amdgpu_device *adev)
+{
+ u32 tmp;
+
+ RREG32(CB_CGTT_SCLK_CTRL);
+ RREG32(CB_CGTT_SCLK_CTRL);
+ RREG32(CB_CGTT_SCLK_CTRL);
+ RREG32(CB_CGTT_SCLK_CTRL);
+
+ tmp = RREG32_SMC(SMC_SYSCON_RESET_CNTL) |
+ RST_REG;
+ WREG32_SMC(SMC_SYSCON_RESET_CNTL, tmp);
+}
+
+int amdgpu_si_program_jump_on_start(struct amdgpu_device *adev)
+{
+ static const u8 data[] = { 0x0E, 0x00, 0x40, 0x40 };
+
+ return amdgpu_si_copy_bytes_to_smc(adev, 0x0, data, 4, sizeof(data)+1);
+}
+
+void amdgpu_si_smc_clock(struct amdgpu_device *adev, bool enable)
+{
+ u32 tmp = RREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0);
+
+ if (enable)
+ tmp &= ~CK_DISABLE;
+ else
+ tmp |= CK_DISABLE;
+
+ WREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0, tmp);
+}
+
+bool amdgpu_si_is_smc_running(struct amdgpu_device *adev)
+{
+ u32 rst = RREG32_SMC(SMC_SYSCON_RESET_CNTL);
+ u32 clk = RREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0);
+
+ if (!(rst & RST_REG) && !(clk & CK_DISABLE))
+ return true;
+
+ return false;
+}
+
+PPSMC_Result amdgpu_si_send_msg_to_smc(struct amdgpu_device *adev,
+ PPSMC_Msg msg)
+{
+ u32 tmp;
+ int i;
+
+ if (!amdgpu_si_is_smc_running(adev))
+ return PPSMC_Result_Failed;
+
+ WREG32(SMC_MESSAGE_0, msg);
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ tmp = RREG32(SMC_RESP_0);
+ if (tmp != 0)
+ break;
+ udelay(1);
+ }
+
+ return (PPSMC_Result)RREG32(SMC_RESP_0);
+}
+
+PPSMC_Result amdgpu_si_wait_for_smc_inactive(struct amdgpu_device *adev)
+{
+ u32 tmp;
+ int i;
+
+ if (!amdgpu_si_is_smc_running(adev))
+ return PPSMC_Result_OK;
+
+ for (i = 0; i < adev->usec_timeout; i++) {
+ tmp = RREG32_SMC(SMC_SYSCON_CLOCK_CNTL_0);
+ if ((tmp & CKEN) == 0)
+ break;
+ udelay(1);
+ }
+
+ return PPSMC_Result_OK;
+}
+
+int amdgpu_si_load_smc_ucode(struct amdgpu_device *adev, u32 limit)
+{
+ const struct smc_firmware_header_v1_0 *hdr;
+ unsigned long flags;
+ u32 ucode_start_address;
+ u32 ucode_size;
+ const u8 *src;
+ u32 data;
+
+ if (!adev->pm.fw)
+ return -EINVAL;
+
+ hdr = (const struct smc_firmware_header_v1_0 *)adev->pm.fw->data;
+
+ amdgpu_ucode_print_smc_hdr(&hdr->header);
+
+ adev->pm.fw_version = le32_to_cpu(hdr->header.ucode_version);
+ ucode_start_address = le32_to_cpu(hdr->ucode_start_addr);
+ ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes);
+ src = (const u8 *)
+ (adev->pm.fw->data + le32_to_cpu(hdr->header.ucode_array_offset_bytes));
+ if (ucode_size & 3)
+ return -EINVAL;
+
+ spin_lock_irqsave(&adev->smc_idx_lock, flags);
+ WREG32(SMC_IND_INDEX_0, ucode_start_address);
+ WREG32_P(SMC_IND_ACCESS_CNTL, AUTO_INCREMENT_IND_0, ~AUTO_INCREMENT_IND_0);
+ while (ucode_size >= 4) {
+ /* SMC address space is BE */
+ data = (src[0] << 24) | (src[1] << 16) | (src[2] << 8) | src[3];
+
+ WREG32(SMC_IND_DATA_0, data);
+
+ src += 4;
+ ucode_size -= 4;
+ }
+ WREG32_P(SMC_IND_ACCESS_CNTL, 0, ~AUTO_INCREMENT_IND_0);
+ spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
+
+ return 0;
+}
+
+int amdgpu_si_read_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address,
+ u32 *value, u32 limit)
+{
+ unsigned long flags;
+ int ret;
+
+ spin_lock_irqsave(&adev->smc_idx_lock, flags);
+ ret = si_set_smc_sram_address(adev, smc_address, limit);
+ if (ret == 0)
+ *value = RREG32(SMC_IND_DATA_0);
+ spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
+
+ return ret;
+}
+
+int amdgpu_si_write_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address,
+ u32 value, u32 limit)
+{
+ unsigned long flags;
+ int ret;
+
+ spin_lock_irqsave(&adev->smc_idx_lock, flags);
+ ret = si_set_smc_sram_address(adev, smc_address, limit);
+ if (ret == 0)
+ WREG32(SMC_IND_DATA_0, value);
+ spin_unlock_irqrestore(&adev->smc_idx_lock, flags);
+
+ return ret;
+}
--- /dev/null
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef PP_SISLANDS_SMC_H
+#define PP_SISLANDS_SMC_H
+
+#include "ppsmc.h"
+
+#pragma pack(push, 1)
+
+#define SISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE 16
+
+struct PP_SIslands_Dpm2PerfLevel
+{
+ uint8_t MaxPS;
+ uint8_t TgtAct;
+ uint8_t MaxPS_StepInc;
+ uint8_t MaxPS_StepDec;
+ uint8_t PSSamplingTime;
+ uint8_t NearTDPDec;
+ uint8_t AboveSafeInc;
+ uint8_t BelowSafeInc;
+ uint8_t PSDeltaLimit;
+ uint8_t PSDeltaWin;
+ uint16_t PwrEfficiencyRatio;
+ uint8_t Reserved[4];
+};
+
+typedef struct PP_SIslands_Dpm2PerfLevel PP_SIslands_Dpm2PerfLevel;
+
+struct PP_SIslands_DPM2Status
+{
+ uint32_t dpm2Flags;
+ uint8_t CurrPSkip;
+ uint8_t CurrPSkipPowerShift;
+ uint8_t CurrPSkipTDP;
+ uint8_t CurrPSkipOCP;
+ uint8_t MaxSPLLIndex;
+ uint8_t MinSPLLIndex;
+ uint8_t CurrSPLLIndex;
+ uint8_t InfSweepMode;
+ uint8_t InfSweepDir;
+ uint8_t TDPexceeded;
+ uint8_t reserved;
+ uint8_t SwitchDownThreshold;
+ uint32_t SwitchDownCounter;
+ uint32_t SysScalingFactor;
+};
+
+typedef struct PP_SIslands_DPM2Status PP_SIslands_DPM2Status;
+
+struct PP_SIslands_DPM2Parameters
+{
+ uint32_t TDPLimit;
+ uint32_t NearTDPLimit;
+ uint32_t SafePowerLimit;
+ uint32_t PowerBoostLimit;
+ uint32_t MinLimitDelta;
+};
+typedef struct PP_SIslands_DPM2Parameters PP_SIslands_DPM2Parameters;
+
+struct PP_SIslands_PAPMStatus
+{
+ uint32_t EstimatedDGPU_T;
+ uint32_t EstimatedDGPU_P;
+ uint32_t EstimatedAPU_T;
+ uint32_t EstimatedAPU_P;
+ uint8_t dGPU_T_Limit_Exceeded;
+ uint8_t reserved[3];
+};
+typedef struct PP_SIslands_PAPMStatus PP_SIslands_PAPMStatus;
+
+struct PP_SIslands_PAPMParameters
+{
+ uint32_t NearTDPLimitTherm;
+ uint32_t NearTDPLimitPAPM;
+ uint32_t PlatformPowerLimit;
+ uint32_t dGPU_T_Limit;
+ uint32_t dGPU_T_Warning;
+ uint32_t dGPU_T_Hysteresis;
+};
+typedef struct PP_SIslands_PAPMParameters PP_SIslands_PAPMParameters;
+
+struct SISLANDS_SMC_SCLK_VALUE
+{
+ uint32_t vCG_SPLL_FUNC_CNTL;
+ uint32_t vCG_SPLL_FUNC_CNTL_2;
+ uint32_t vCG_SPLL_FUNC_CNTL_3;
+ uint32_t vCG_SPLL_FUNC_CNTL_4;
+ uint32_t vCG_SPLL_SPREAD_SPECTRUM;
+ uint32_t vCG_SPLL_SPREAD_SPECTRUM_2;
+ uint32_t sclk_value;
+};
+
+typedef struct SISLANDS_SMC_SCLK_VALUE SISLANDS_SMC_SCLK_VALUE;
+
+struct SISLANDS_SMC_MCLK_VALUE
+{
+ uint32_t vMPLL_FUNC_CNTL;
+ uint32_t vMPLL_FUNC_CNTL_1;
+ uint32_t vMPLL_FUNC_CNTL_2;
+ uint32_t vMPLL_AD_FUNC_CNTL;
+ uint32_t vMPLL_DQ_FUNC_CNTL;
+ uint32_t vMCLK_PWRMGT_CNTL;
+ uint32_t vDLL_CNTL;
+ uint32_t vMPLL_SS;
+ uint32_t vMPLL_SS2;
+ uint32_t mclk_value;
+};
+
+typedef struct SISLANDS_SMC_MCLK_VALUE SISLANDS_SMC_MCLK_VALUE;
+
+struct SISLANDS_SMC_VOLTAGE_VALUE
+{
+ uint16_t value;
+ uint8_t index;
+ uint8_t phase_settings;
+};
+
+typedef struct SISLANDS_SMC_VOLTAGE_VALUE SISLANDS_SMC_VOLTAGE_VALUE;
+
+struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL
+{
+ uint8_t ACIndex;
+ uint8_t displayWatermark;
+ uint8_t gen2PCIE;
+ uint8_t UVDWatermark;
+ uint8_t VCEWatermark;
+ uint8_t strobeMode;
+ uint8_t mcFlags;
+ uint8_t padding;
+ uint32_t aT;
+ uint32_t bSP;
+ SISLANDS_SMC_SCLK_VALUE sclk;
+ SISLANDS_SMC_MCLK_VALUE mclk;
+ SISLANDS_SMC_VOLTAGE_VALUE vddc;
+ SISLANDS_SMC_VOLTAGE_VALUE mvdd;
+ SISLANDS_SMC_VOLTAGE_VALUE vddci;
+ SISLANDS_SMC_VOLTAGE_VALUE std_vddc;
+ uint8_t hysteresisUp;
+ uint8_t hysteresisDown;
+ uint8_t stateFlags;
+ uint8_t arbRefreshState;
+ uint32_t SQPowerThrottle;
+ uint32_t SQPowerThrottle_2;
+ uint32_t MaxPoweredUpCU;
+ SISLANDS_SMC_VOLTAGE_VALUE high_temp_vddc;
+ SISLANDS_SMC_VOLTAGE_VALUE low_temp_vddc;
+ uint32_t reserved[2];
+ PP_SIslands_Dpm2PerfLevel dpm2;
+};
+
+#define SISLANDS_SMC_STROBE_RATIO 0x0F
+#define SISLANDS_SMC_STROBE_ENABLE 0x10
+
+#define SISLANDS_SMC_MC_EDC_RD_FLAG 0x01
+#define SISLANDS_SMC_MC_EDC_WR_FLAG 0x02
+#define SISLANDS_SMC_MC_RTT_ENABLE 0x04
+#define SISLANDS_SMC_MC_STUTTER_EN 0x08
+#define SISLANDS_SMC_MC_PG_EN 0x10
+
+typedef struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL SISLANDS_SMC_HW_PERFORMANCE_LEVEL;
+
+struct SISLANDS_SMC_SWSTATE
+{
+ uint8_t flags;
+ uint8_t levelCount;
+ uint8_t padding2;
+ uint8_t padding3;
+ SISLANDS_SMC_HW_PERFORMANCE_LEVEL levels[1];
+};
+
+typedef struct SISLANDS_SMC_SWSTATE SISLANDS_SMC_SWSTATE;
+
+#define SISLANDS_SMC_VOLTAGEMASK_VDDC 0
+#define SISLANDS_SMC_VOLTAGEMASK_MVDD 1
+#define SISLANDS_SMC_VOLTAGEMASK_VDDCI 2
+#define SISLANDS_SMC_VOLTAGEMASK_MAX 4
+
+struct SISLANDS_SMC_VOLTAGEMASKTABLE
+{
+ uint32_t lowMask[SISLANDS_SMC_VOLTAGEMASK_MAX];
+};
+
+typedef struct SISLANDS_SMC_VOLTAGEMASKTABLE SISLANDS_SMC_VOLTAGEMASKTABLE;
+
+#define SISLANDS_MAX_NO_VREG_STEPS 32
+
+struct SISLANDS_SMC_STATETABLE
+{
+ uint8_t thermalProtectType;
+ uint8_t systemFlags;
+ uint8_t maxVDDCIndexInPPTable;
+ uint8_t extraFlags;
+ uint32_t lowSMIO[SISLANDS_MAX_NO_VREG_STEPS];
+ SISLANDS_SMC_VOLTAGEMASKTABLE voltageMaskTable;
+ SISLANDS_SMC_VOLTAGEMASKTABLE phaseMaskTable;
+ PP_SIslands_DPM2Parameters dpm2Params;
+ SISLANDS_SMC_SWSTATE initialState;
+ SISLANDS_SMC_SWSTATE ACPIState;
+ SISLANDS_SMC_SWSTATE ULVState;
+ SISLANDS_SMC_SWSTATE driverState;
+ SISLANDS_SMC_HW_PERFORMANCE_LEVEL dpmLevels[SISLANDS_MAX_SMC_PERFORMANCE_LEVELS_PER_SWSTATE - 1];
+};
+
+typedef struct SISLANDS_SMC_STATETABLE SISLANDS_SMC_STATETABLE;
+
+#define SI_SMC_SOFT_REGISTER_mclk_chg_timeout 0x0
+#define SI_SMC_SOFT_REGISTER_delay_vreg 0xC
+#define SI_SMC_SOFT_REGISTER_delay_acpi 0x28
+#define SI_SMC_SOFT_REGISTER_seq_index 0x5C
+#define SI_SMC_SOFT_REGISTER_mvdd_chg_time 0x60
+#define SI_SMC_SOFT_REGISTER_mclk_switch_lim 0x70
+#define SI_SMC_SOFT_REGISTER_watermark_threshold 0x78
+#define SI_SMC_SOFT_REGISTER_phase_shedding_delay 0x88
+#define SI_SMC_SOFT_REGISTER_ulv_volt_change_delay 0x8C
+#define SI_SMC_SOFT_REGISTER_mc_block_delay 0x98
+#define SI_SMC_SOFT_REGISTER_ticks_per_us 0xA8
+#define SI_SMC_SOFT_REGISTER_crtc_index 0xC4
+#define SI_SMC_SOFT_REGISTER_mclk_change_block_cp_min 0xC8
+#define SI_SMC_SOFT_REGISTER_mclk_change_block_cp_max 0xCC
+#define SI_SMC_SOFT_REGISTER_non_ulv_pcie_link_width 0xF4
+#define SI_SMC_SOFT_REGISTER_tdr_is_about_to_happen 0xFC
+#define SI_SMC_SOFT_REGISTER_vr_hot_gpio 0x100
+#define SI_SMC_SOFT_REGISTER_svi_rework_plat_type 0x118
+#define SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svd 0x11c
+#define SI_SMC_SOFT_REGISTER_svi_rework_gpio_id_svc 0x120
+
+struct PP_SIslands_FanTable
+{
+ uint8_t fdo_mode;
+ uint8_t padding;
+ int16_t temp_min;
+ int16_t temp_med;
+ int16_t temp_max;
+ int16_t slope1;
+ int16_t slope2;
+ int16_t fdo_min;
+ int16_t hys_up;
+ int16_t hys_down;
+ int16_t hys_slope;
+ int16_t temp_resp_lim;
+ int16_t temp_curr;
+ int16_t slope_curr;
+ int16_t pwm_curr;
+ uint32_t refresh_period;
+ int16_t fdo_max;
+ uint8_t temp_src;
+ int8_t padding2;
+};
+
+typedef struct PP_SIslands_FanTable PP_SIslands_FanTable;
+
+#define SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES 16
+#define SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES 32
+
+#define SMC_SISLANDS_SCALE_I 7
+#define SMC_SISLANDS_SCALE_R 12
+
+struct PP_SIslands_CacConfig
+{
+ uint16_t cac_lkge_lut[SMC_SISLANDS_LKGE_LUT_NUM_OF_TEMP_ENTRIES][SMC_SISLANDS_LKGE_LUT_NUM_OF_VOLT_ENTRIES];
+ uint32_t lkge_lut_V0;
+ uint32_t lkge_lut_Vstep;
+ uint32_t WinTime;
+ uint32_t R_LL;
+ uint32_t calculation_repeats;
+ uint32_t l2numWin_TDP;
+ uint32_t dc_cac;
+ uint8_t lts_truncate_n;
+ uint8_t SHIFT_N;
+ uint8_t log2_PG_LKG_SCALE;
+ uint8_t cac_temp;
+ uint32_t lkge_lut_T0;
+ uint32_t lkge_lut_Tstep;
+};
+
+typedef struct PP_SIslands_CacConfig PP_SIslands_CacConfig;
+
+#define SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE 16
+#define SMC_SISLANDS_MC_REGISTER_ARRAY_SET_COUNT 20
+
+struct SMC_SIslands_MCRegisterAddress
+{
+ uint16_t s0;
+ uint16_t s1;
+};
+
+typedef struct SMC_SIslands_MCRegisterAddress SMC_SIslands_MCRegisterAddress;
+
+struct SMC_SIslands_MCRegisterSet
+{
+ uint32_t value[SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE];
+};
+
+typedef struct SMC_SIslands_MCRegisterSet SMC_SIslands_MCRegisterSet;
+
+struct SMC_SIslands_MCRegisters
+{
+ uint8_t last;
+ uint8_t reserved[3];
+ SMC_SIslands_MCRegisterAddress address[SMC_SISLANDS_MC_REGISTER_ARRAY_SIZE];
+ SMC_SIslands_MCRegisterSet data[SMC_SISLANDS_MC_REGISTER_ARRAY_SET_COUNT];
+};
+
+typedef struct SMC_SIslands_MCRegisters SMC_SIslands_MCRegisters;
+
+struct SMC_SIslands_MCArbDramTimingRegisterSet
+{
+ uint32_t mc_arb_dram_timing;
+ uint32_t mc_arb_dram_timing2;
+ uint8_t mc_arb_rfsh_rate;
+ uint8_t mc_arb_burst_time;
+ uint8_t padding[2];
+};
+
+typedef struct SMC_SIslands_MCArbDramTimingRegisterSet SMC_SIslands_MCArbDramTimingRegisterSet;
+
+struct SMC_SIslands_MCArbDramTimingRegisters
+{
+ uint8_t arb_current;
+ uint8_t reserved[3];
+ SMC_SIslands_MCArbDramTimingRegisterSet data[16];
+};
+
+typedef struct SMC_SIslands_MCArbDramTimingRegisters SMC_SIslands_MCArbDramTimingRegisters;
+
+struct SMC_SISLANDS_SPLL_DIV_TABLE
+{
+ uint32_t freq[256];
+ uint32_t ss[256];
+};
+
+#define SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_MASK 0x01ffffff
+#define SMC_SISLANDS_SPLL_DIV_TABLE_FBDIV_SHIFT 0
+#define SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_MASK 0xfe000000
+#define SMC_SISLANDS_SPLL_DIV_TABLE_PDIV_SHIFT 25
+#define SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_MASK 0x000fffff
+#define SMC_SISLANDS_SPLL_DIV_TABLE_CLKV_SHIFT 0
+#define SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_MASK 0xfff00000
+#define SMC_SISLANDS_SPLL_DIV_TABLE_CLKS_SHIFT 20
+
+typedef struct SMC_SISLANDS_SPLL_DIV_TABLE SMC_SISLANDS_SPLL_DIV_TABLE;
+
+#define SMC_SISLANDS_DTE_MAX_FILTER_STAGES 5
+
+#define SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE 16
+
+struct Smc_SIslands_DTE_Configuration
+{
+ uint32_t tau[SMC_SISLANDS_DTE_MAX_FILTER_STAGES];
+ uint32_t R[SMC_SISLANDS_DTE_MAX_FILTER_STAGES];
+ uint32_t K;
+ uint32_t T0;
+ uint32_t MaxT;
+ uint8_t WindowSize;
+ uint8_t Tdep_count;
+ uint8_t temp_select;
+ uint8_t DTE_mode;
+ uint8_t T_limits[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE];
+ uint32_t Tdep_tau[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE];
+ uint32_t Tdep_R[SMC_SISLANDS_DTE_MAX_TEMPERATURE_DEPENDENT_ARRAY_SIZE];
+ uint32_t Tthreshold;
+};
+
+typedef struct Smc_SIslands_DTE_Configuration Smc_SIslands_DTE_Configuration;
+
+#define SMC_SISLANDS_DTE_STATUS_FLAG_DTE_ON 1
+
+#define SISLANDS_SMC_FIRMWARE_HEADER_LOCATION 0x10000
+
+#define SISLANDS_SMC_FIRMWARE_HEADER_version 0x0
+#define SISLANDS_SMC_FIRMWARE_HEADER_flags 0x4
+#define SISLANDS_SMC_FIRMWARE_HEADER_softRegisters 0xC
+#define SISLANDS_SMC_FIRMWARE_HEADER_stateTable 0x10
+#define SISLANDS_SMC_FIRMWARE_HEADER_fanTable 0x14
+#define SISLANDS_SMC_FIRMWARE_HEADER_CacConfigTable 0x18
+#define SISLANDS_SMC_FIRMWARE_HEADER_mcRegisterTable 0x24
+#define SISLANDS_SMC_FIRMWARE_HEADER_mcArbDramAutoRefreshTable 0x30
+#define SISLANDS_SMC_FIRMWARE_HEADER_spllTable 0x38
+#define SISLANDS_SMC_FIRMWARE_HEADER_DteConfiguration 0x40
+#define SISLANDS_SMC_FIRMWARE_HEADER_PAPMParameters 0x48
+
+#pragma pack(pop)
+
+int amdgpu_si_copy_bytes_to_smc(struct amdgpu_device *adev,
+ u32 smc_start_address,
+ const u8 *src, u32 byte_count, u32 limit);
+void amdgpu_si_start_smc(struct amdgpu_device *adev);
+void amdgpu_si_reset_smc(struct amdgpu_device *adev);
+int amdgpu_si_program_jump_on_start(struct amdgpu_device *adev);
+void amdgpu_si_smc_clock(struct amdgpu_device *adev, bool enable);
+bool amdgpu_si_is_smc_running(struct amdgpu_device *adev);
+PPSMC_Result amdgpu_si_send_msg_to_smc(struct amdgpu_device *adev, PPSMC_Msg msg);
+PPSMC_Result amdgpu_si_wait_for_smc_inactive(struct amdgpu_device *adev);
+int amdgpu_si_load_smc_ucode(struct amdgpu_device *adev, u32 limit);
+int amdgpu_si_read_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address,
+ u32 *value, u32 limit);
+int amdgpu_si_write_smc_sram_dword(struct amdgpu_device *adev, u32 smc_address,
+ u32 value, u32 limit);
+
+#endif
+
return -ETIMEDOUT;
}
-static int tonga_ih_soft_reset(void *handle)
+static int tonga_ih_check_soft_reset(void *handle)
{
- u32 srbm_soft_reset = 0;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset = 0;
u32 tmp = RREG32(mmSRBM_STATUS);
if (tmp & SRBM_STATUS__IH_BUSY_MASK)
SOFT_RESET_IH, 1);
if (srbm_soft_reset) {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_IH].hang = true;
+ adev->irq.srbm_soft_reset = srbm_soft_reset;
+ } else {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_IH].hang = false;
+ adev->irq.srbm_soft_reset = 0;
+ }
+
+ return 0;
+}
+
+static int tonga_ih_pre_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_IH].hang)
+ return 0;
+
+ return tonga_ih_hw_fini(adev);
+}
+
+static int tonga_ih_post_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_IH].hang)
+ return 0;
+
+ return tonga_ih_hw_init(adev);
+}
+
+static int tonga_ih_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_IH].hang)
+ return 0;
+ srbm_soft_reset = adev->irq.srbm_soft_reset;
+
+ if (srbm_soft_reset) {
+ u32 tmp;
+
tmp = RREG32(mmSRBM_SOFT_RESET);
tmp |= srbm_soft_reset;
dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp);
.resume = tonga_ih_resume,
.is_idle = tonga_ih_is_idle,
.wait_for_idle = tonga_ih_wait_for_idle,
+ .check_soft_reset = tonga_ih_check_soft_reset,
+ .pre_soft_reset = tonga_ih_pre_soft_reset,
.soft_reset = tonga_ih_soft_reset,
+ .post_soft_reset = tonga_ih_post_soft_reset,
.set_clockgating_state = tonga_ih_set_clockgating_state,
.set_powergating_state = tonga_ih_set_powergating_state,
};
ring = &adev->uvd.ring;
sprintf(ring->name, "uvd");
- r = amdgpu_ring_init(adev, ring, 512, CP_PACKET2, 0xf,
+ r = amdgpu_ring_init(adev, ring, 512, PACKET0(mmUVD_NO_OP, 0), 0xf,
&adev->uvd.irq, 0, AMDGPU_RING_TYPE_UVD);
return r;
amdgpu_ring_write(ring, ib->length_dw);
}
+static unsigned uvd_v4_2_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 4; /* uvd_v4_2_ring_emit_ib */
+}
+
+static unsigned uvd_v4_2_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 2 + /* uvd_v4_2_ring_emit_hdp_flush */
+ 2 + /* uvd_v4_2_ring_emit_hdp_invalidate */
+ 14; /* uvd_v4_2_ring_emit_fence x1 no user fence */
+}
+
/**
* uvd_v4_2_mc_resume - memory controller programming
*
.pad_ib = amdgpu_ring_generic_pad_ib,
.begin_use = amdgpu_uvd_ring_begin_use,
.end_use = amdgpu_uvd_ring_end_use,
+ .get_emit_ib_size = uvd_v4_2_ring_get_emit_ib_size,
+ .get_dma_frame_size = uvd_v4_2_ring_get_dma_frame_size,
};
static void uvd_v4_2_set_ring_funcs(struct amdgpu_device *adev)
ring = &adev->uvd.ring;
sprintf(ring->name, "uvd");
- r = amdgpu_ring_init(adev, ring, 512, CP_PACKET2, 0xf,
+ r = amdgpu_ring_init(adev, ring, 512, PACKET0(mmUVD_NO_OP, 0), 0xf,
&adev->uvd.irq, 0, AMDGPU_RING_TYPE_UVD);
return r;
amdgpu_ring_write(ring, ib->length_dw);
}
+static unsigned uvd_v5_0_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 6; /* uvd_v5_0_ring_emit_ib */
+}
+
+static unsigned uvd_v5_0_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 2 + /* uvd_v5_0_ring_emit_hdp_flush */
+ 2 + /* uvd_v5_0_ring_emit_hdp_invalidate */
+ 14; /* uvd_v5_0_ring_emit_fence x1 no user fence */
+}
+
static bool uvd_v5_0_is_idle(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
.pad_ib = amdgpu_ring_generic_pad_ib,
.begin_use = amdgpu_uvd_ring_begin_use,
.end_use = amdgpu_uvd_ring_end_use,
+ .get_emit_ib_size = uvd_v5_0_ring_get_emit_ib_size,
+ .get_dma_frame_size = uvd_v5_0_ring_get_dma_frame_size,
};
static void uvd_v5_0_set_ring_funcs(struct amdgpu_device *adev)
ring = &adev->uvd.ring;
sprintf(ring->name, "uvd");
- r = amdgpu_ring_init(adev, ring, 512, CP_PACKET2, 0xf,
+ r = amdgpu_ring_init(adev, ring, 512, PACKET0(mmUVD_NO_OP, 0), 0xf,
&adev->uvd.irq, 0, AMDGPU_RING_TYPE_UVD);
return r;
uvd_v6_0_mc_resume(adev);
- /* Set dynamic clock gating in S/W control mode */
- if (adev->cg_flags & AMD_CG_SUPPORT_UVD_MGCG) {
- uvd_v6_0_set_sw_clock_gating(adev);
- } else {
- /* disable clock gating */
- uint32_t data = RREG32(mmUVD_CGC_CTRL);
- data &= ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK;
- WREG32(mmUVD_CGC_CTRL, data);
- }
+ /* disable clock gating */
+ WREG32_FIELD(UVD_CGC_CTRL, DYN_CLOCK_MODE, 0);
/* disable interupt */
- WREG32_P(mmUVD_MASTINT_EN, 0, ~UVD_MASTINT_EN__VCPU_EN_MASK);
+ WREG32_FIELD(UVD_MASTINT_EN, VCPU_EN, 0);
/* stall UMC and register bus before resetting VCPU */
- WREG32_P(mmUVD_LMI_CTRL2, UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+ WREG32_FIELD(UVD_LMI_CTRL2, STALL_ARB_UMC, 1);
mdelay(1);
/* put LMI, VCPU, RBC etc... into reset */
mdelay(5);
/* take UVD block out of reset */
- WREG32_P(mmSRBM_SOFT_RESET, 0, ~SRBM_SOFT_RESET__SOFT_RESET_UVD_MASK);
+ WREG32_FIELD(SRBM_SOFT_RESET, SOFT_RESET_UVD, 0);
mdelay(5);
/* initialize UVD memory controller */
WREG32(mmUVD_VCPU_CNTL, UVD_VCPU_CNTL__CLK_EN_MASK);
/* enable UMC */
- WREG32_P(mmUVD_LMI_CTRL2, 0, ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+ WREG32_FIELD(UVD_LMI_CTRL2, STALL_ARB_UMC, 0);
/* boot up the VCPU */
WREG32(mmUVD_SOFT_RESET, 0);
break;
DRM_ERROR("UVD not responding, trying to reset the VCPU!!!\n");
- WREG32_P(mmUVD_SOFT_RESET, UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK,
- ~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+ WREG32_FIELD(UVD_SOFT_RESET, VCPU_SOFT_RESET, 1);
mdelay(10);
- WREG32_P(mmUVD_SOFT_RESET, 0,
- ~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+ WREG32_FIELD(UVD_SOFT_RESET, VCPU_SOFT_RESET, 0);
mdelay(10);
r = -1;
}
/* clear the bit 4 of UVD_STATUS */
WREG32_P(mmUVD_STATUS, 0, ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT));
+ /* force RBC into idle state */
rb_bufsz = order_base_2(ring->ring_size);
- tmp = 0;
- tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BUFSZ, rb_bufsz);
+ tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, rb_bufsz);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BLKSZ, 1);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_WPTR_POLL_EN, 0);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1);
tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1);
- /* force RBC into idle state */
WREG32(mmUVD_RBC_RB_CNTL, tmp);
/* set the write pointer delay */
ring->wptr = RREG32(mmUVD_RBC_RB_RPTR);
WREG32(mmUVD_RBC_RB_WPTR, ring->wptr);
- WREG32_P(mmUVD_RBC_RB_CNTL, 0, ~UVD_RBC_RB_CNTL__RB_NO_FETCH_MASK);
+ WREG32_FIELD(UVD_RBC_RB_CNTL, RB_NO_FETCH, 0);
return 0;
}
amdgpu_ring_write(ring, 0xE);
}
+static unsigned uvd_v6_0_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 8; /* uvd_v6_0_ring_emit_ib */
+}
+
+static unsigned uvd_v6_0_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 2 + /* uvd_v6_0_ring_emit_hdp_flush */
+ 2 + /* uvd_v6_0_ring_emit_hdp_invalidate */
+ 10 + /* uvd_v6_0_ring_emit_pipeline_sync */
+ 14; /* uvd_v6_0_ring_emit_fence x1 no user fence */
+}
+
+static unsigned uvd_v6_0_ring_get_dma_frame_size_vm(struct amdgpu_ring *ring)
+{
+ return
+ 2 + /* uvd_v6_0_ring_emit_hdp_flush */
+ 2 + /* uvd_v6_0_ring_emit_hdp_invalidate */
+ 10 + /* uvd_v6_0_ring_emit_pipeline_sync */
+ 20 + /* uvd_v6_0_ring_emit_vm_flush */
+ 14 + 14; /* uvd_v6_0_ring_emit_fence x2 vm fence */
+}
+
static bool uvd_v6_0_is_idle(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
for (i = 0; i < adev->usec_timeout; i++) {
- if (!(RREG32(mmSRBM_STATUS) & SRBM_STATUS__UVD_BUSY_MASK))
+ if (uvd_v6_0_is_idle(handle))
return 0;
}
return -ETIMEDOUT;
}
-static int uvd_v6_0_soft_reset(void *handle)
+#define AMDGPU_UVD_STATUS_BUSY_MASK 0xfd
+static int uvd_v6_0_check_soft_reset(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset = 0;
+ u32 tmp = RREG32(mmSRBM_STATUS);
+
+ if (REG_GET_FIELD(tmp, SRBM_STATUS, UVD_RQ_PENDING) ||
+ REG_GET_FIELD(tmp, SRBM_STATUS, UVD_BUSY) ||
+ (RREG32(mmUVD_STATUS) & AMDGPU_UVD_STATUS_BUSY_MASK))
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_UVD, 1);
+
+ if (srbm_soft_reset) {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_UVD].hang = true;
+ adev->uvd.srbm_soft_reset = srbm_soft_reset;
+ } else {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_UVD].hang = false;
+ adev->uvd.srbm_soft_reset = 0;
+ }
+ return 0;
+}
+static int uvd_v6_0_pre_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_UVD].hang)
+ return 0;
uvd_v6_0_stop(adev);
+ return 0;
+}
+
+static int uvd_v6_0_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_UVD].hang)
+ return 0;
+ srbm_soft_reset = adev->uvd.srbm_soft_reset;
+
+ if (srbm_soft_reset) {
+ u32 tmp;
+
+ tmp = RREG32(mmSRBM_SOFT_RESET);
+ tmp |= srbm_soft_reset;
+ dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp);
+ WREG32(mmSRBM_SOFT_RESET, tmp);
+ tmp = RREG32(mmSRBM_SOFT_RESET);
+
+ udelay(50);
+
+ tmp &= ~srbm_soft_reset;
+ WREG32(mmSRBM_SOFT_RESET, tmp);
+ tmp = RREG32(mmSRBM_SOFT_RESET);
+
+ /* Wait a little for things to settle down */
+ udelay(50);
+ }
+
+ return 0;
+}
+
+static int uvd_v6_0_post_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_UVD].hang)
+ return 0;
- WREG32_P(mmSRBM_SOFT_RESET, SRBM_SOFT_RESET__SOFT_RESET_UVD_MASK,
- ~SRBM_SOFT_RESET__SOFT_RESET_UVD_MASK);
mdelay(5);
return uvd_v6_0_start(adev);
enum amd_clockgating_state state)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
- bool enable = (state == AMD_CG_STATE_GATE) ? true : false;
- static int curstate = -1;
if (adev->asic_type == CHIP_FIJI ||
- adev->asic_type == CHIP_POLARIS10)
- uvd_v6_set_bypass_mode(adev, enable);
+ adev->asic_type == CHIP_POLARIS10)
+ uvd_v6_set_bypass_mode(adev, state == AMD_CG_STATE_GATE ? true : false);
if (!(adev->cg_flags & AMD_CG_SUPPORT_UVD_MGCG))
return 0;
- if (curstate == state)
- return 0;
-
- curstate = state;
- if (enable) {
+ if (state == AMD_CG_STATE_GATE) {
/* disable HW gating and enable Sw gating */
uvd_v6_0_set_sw_clock_gating(adev);
} else {
if (!(adev->pg_flags & AMD_PG_SUPPORT_UVD))
return 0;
+ WREG32(mmUVD_POWER_STATUS, UVD_POWER_STATUS__UVD_PG_EN_MASK);
+
if (state == AMD_PG_STATE_GATE) {
uvd_v6_0_stop(adev);
return 0;
.resume = uvd_v6_0_resume,
.is_idle = uvd_v6_0_is_idle,
.wait_for_idle = uvd_v6_0_wait_for_idle,
+ .check_soft_reset = uvd_v6_0_check_soft_reset,
+ .pre_soft_reset = uvd_v6_0_pre_soft_reset,
.soft_reset = uvd_v6_0_soft_reset,
+ .post_soft_reset = uvd_v6_0_post_soft_reset,
.set_clockgating_state = uvd_v6_0_set_clockgating_state,
.set_powergating_state = uvd_v6_0_set_powergating_state,
};
.pad_ib = amdgpu_ring_generic_pad_ib,
.begin_use = amdgpu_uvd_ring_begin_use,
.end_use = amdgpu_uvd_ring_end_use,
+ .get_emit_ib_size = uvd_v6_0_ring_get_emit_ib_size,
+ .get_dma_frame_size = uvd_v6_0_ring_get_dma_frame_size,
};
static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = {
.pad_ib = amdgpu_ring_generic_pad_ib,
.begin_use = amdgpu_uvd_ring_begin_use,
.end_use = amdgpu_uvd_ring_end_use,
+ .get_emit_ib_size = uvd_v6_0_ring_get_emit_ib_size,
+ .get_dma_frame_size = uvd_v6_0_ring_get_dma_frame_size_vm,
};
static void uvd_v6_0_set_ring_funcs(struct amdgpu_device *adev)
#include "amdgpu.h"
#include "amdgpu_vce.h"
#include "cikd.h"
-
#include "vce/vce_2_0_d.h"
#include "vce/vce_2_0_sh_mask.h"
-
+#include "smu/smu_7_0_1_d.h"
+#include "smu/smu_7_0_1_sh_mask.h"
#include "oss/oss_2_0_d.h"
#include "oss/oss_2_0_sh_mask.h"
#define VCE_V2_0_FW_SIZE (256 * 1024)
#define VCE_V2_0_STACK_SIZE (64 * 1024)
#define VCE_V2_0_DATA_SIZE (23552 * AMDGPU_MAX_VCE_HANDLES)
+#define VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK 0x02
static void vce_v2_0_mc_resume(struct amdgpu_device *adev);
static void vce_v2_0_set_ring_funcs(struct amdgpu_device *adev);
WREG32(mmVCE_RB_WPTR2, ring->wptr);
}
+static int vce_v2_0_lmi_clean(struct amdgpu_device *adev)
+{
+ int i, j;
+
+ for (i = 0; i < 10; ++i) {
+ for (j = 0; j < 100; ++j) {
+ uint32_t status = RREG32(mmVCE_LMI_STATUS);
+
+ if (status & 0x337f)
+ return 0;
+ mdelay(10);
+ }
+ }
+
+ return -ETIMEDOUT;
+}
+
+static int vce_v2_0_firmware_loaded(struct amdgpu_device *adev)
+{
+ int i, j;
+
+ for (i = 0; i < 10; ++i) {
+ for (j = 0; j < 100; ++j) {
+ uint32_t status = RREG32(mmVCE_STATUS);
+
+ if (status & VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK)
+ return 0;
+ mdelay(10);
+ }
+
+ DRM_ERROR("VCE not responding, trying to reset the ECPU!!!\n");
+ WREG32_P(mmVCE_SOFT_RESET,
+ VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
+ ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+ mdelay(10);
+ WREG32_P(mmVCE_SOFT_RESET, 0,
+ ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+ mdelay(10);
+ }
+
+ return -ETIMEDOUT;
+}
+
/**
* vce_v2_0_start - start VCE block
*
static int vce_v2_0_start(struct amdgpu_device *adev)
{
struct amdgpu_ring *ring;
- int i, j, r;
+ int r;
vce_v2_0_mc_resume(adev);
WREG32(mmVCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
WREG32(mmVCE_RB_SIZE2, ring->ring_size / 4);
- WREG32_P(mmVCE_VCPU_CNTL, VCE_VCPU_CNTL__CLK_EN_MASK, ~VCE_VCPU_CNTL__CLK_EN_MASK);
-
- WREG32_P(mmVCE_SOFT_RESET,
- VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
- ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
-
+ WREG32_FIELD(VCE_VCPU_CNTL, CLK_EN, 1);
+ WREG32_FIELD(VCE_SOFT_RESET, ECPU_SOFT_RESET, 1);
mdelay(100);
+ WREG32_FIELD(VCE_SOFT_RESET, ECPU_SOFT_RESET, 0);
- WREG32_P(mmVCE_SOFT_RESET, 0, ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
-
- for (i = 0; i < 10; ++i) {
- uint32_t status;
- for (j = 0; j < 100; ++j) {
- status = RREG32(mmVCE_STATUS);
- if (status & 2)
- break;
- mdelay(10);
- }
- r = 0;
- if (status & 2)
- break;
-
- DRM_ERROR("VCE not responding, trying to reset the ECPU!!!\n");
- WREG32_P(mmVCE_SOFT_RESET, VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
- ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
- mdelay(10);
- WREG32_P(mmVCE_SOFT_RESET, 0, ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
- mdelay(10);
- r = -1;
- }
+ r = vce_v2_0_firmware_loaded(adev);
/* clear BUSY flag */
WREG32_P(mmVCE_STATUS, 0, ~1);
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ adev->vce.num_rings = 2;
+
vce_v2_0_set_ring_funcs(adev);
vce_v2_0_set_irq_funcs(adev);
static int vce_v2_0_sw_init(void *handle)
{
struct amdgpu_ring *ring;
- int r;
+ int r, i;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
/* VCE */
if (r)
return r;
- ring = &adev->vce.ring[0];
- sprintf(ring->name, "vce0");
- r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf,
- &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE);
- if (r)
- return r;
-
- ring = &adev->vce.ring[1];
- sprintf(ring->name, "vce1");
- r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf,
- &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE);
- if (r)
- return r;
+ for (i = 0; i < adev->vce.num_rings; i++) {
+ ring = &adev->vce.ring[i];
+ sprintf(ring->name, "vce%d", i);
+ r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf,
+ &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE);
+ if (r)
+ return r;
+ }
return r;
}
static int vce_v2_0_hw_init(void *handle)
{
- struct amdgpu_ring *ring;
- int r;
+ int r, i;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
r = vce_v2_0_start(adev);
+ /* this error mean vcpu not in running state, so just skip ring test, not stop driver initialize */
if (r)
-/* this error mean vcpu not in running state, so just skip ring test, not stop driver initialize */
return 0;
- ring = &adev->vce.ring[0];
- ring->ready = true;
- r = amdgpu_ring_test_ring(ring);
- if (r) {
- ring->ready = false;
- return r;
- }
+ for (i = 0; i < adev->vce.num_rings; i++)
+ adev->vce.ring[i].ready = false;
- ring = &adev->vce.ring[1];
- ring->ready = true;
- r = amdgpu_ring_test_ring(ring);
- if (r) {
- ring->ready = false;
- return r;
+ for (i = 0; i < adev->vce.num_rings; i++) {
+ r = amdgpu_ring_test_ring(&adev->vce.ring[i]);
+ if (r)
+ return r;
+ else
+ adev->vce.ring[i].ready = true;
}
DRM_INFO("VCE initialized successfully.\n");
static void vce_v2_0_set_dyn_cg(struct amdgpu_device *adev, bool gated)
{
- u32 orig, tmp;
+ if (vce_v2_0_wait_for_idle(adev)) {
+ DRM_INFO("VCE is busy, Can't set clock gateing");
+ return;
+ }
- if (gated) {
- if (vce_v2_0_wait_for_idle(adev)) {
- DRM_INFO("VCE is busy, Can't set clock gateing");
- return;
- }
- WREG32_P(mmVCE_VCPU_CNTL, 0, ~VCE_VCPU_CNTL__CLK_EN_MASK);
- WREG32_P(mmVCE_SOFT_RESET, VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK, ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
- mdelay(100);
- WREG32(mmVCE_STATUS, 0);
- } else {
- WREG32_P(mmVCE_VCPU_CNTL, VCE_VCPU_CNTL__CLK_EN_MASK, ~VCE_VCPU_CNTL__CLK_EN_MASK);
- WREG32_P(mmVCE_SOFT_RESET, VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK, ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
- mdelay(100);
+ WREG32_P(mmVCE_LMI_CTRL2, 0x100, ~0x100);
+
+ if (vce_v2_0_lmi_clean(adev)) {
+ DRM_INFO("LMI is busy, Can't set clock gateing");
+ return;
}
- tmp = RREG32(mmVCE_CLOCK_GATING_B);
- tmp &= ~0x00060006;
+ WREG32_P(mmVCE_VCPU_CNTL, 0, ~VCE_VCPU_CNTL__CLK_EN_MASK);
+ WREG32_P(mmVCE_SOFT_RESET,
+ VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
+ ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+ WREG32(mmVCE_STATUS, 0);
+
+ if (gated)
+ WREG32(mmVCE_CGTT_CLK_OVERRIDE, 0);
+ /* LMI_MC/LMI_UMC always set in dynamic, set {CGC_*_GATE_MODE, CGC_*_SW_GATE} = {0, 0} */
if (gated) {
- tmp |= 0xe10000;
+ /* Force CLOCK OFF , set {CGC_*_GATE_MODE, CGC_*_SW_GATE} = {*, 1} */
+ WREG32(mmVCE_CLOCK_GATING_B, 0xe90010);
} else {
- tmp |= 0xe1;
- tmp &= ~0xe10000;
+ /* Force CLOCK ON, set {CGC_*_GATE_MODE, CGC_*_SW_GATE} = {1, 0} */
+ WREG32(mmVCE_CLOCK_GATING_B, 0x800f1);
}
- WREG32(mmVCE_CLOCK_GATING_B, tmp);
- orig = tmp = RREG32(mmVCE_UENC_CLOCK_GATING);
- tmp &= ~0x1fe000;
- tmp &= ~0xff000000;
- if (tmp != orig)
- WREG32(mmVCE_UENC_CLOCK_GATING, tmp);
+ /* Set VCE_UENC_CLOCK_GATING always in dynamic mode {*_FORCE_ON, *_FORCE_OFF} = {0, 0}*/;
+ WREG32(mmVCE_UENC_CLOCK_GATING, 0x40);
- orig = tmp = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
- tmp &= ~0x3fc;
- if (tmp != orig)
- WREG32(mmVCE_UENC_REG_CLOCK_GATING, tmp);
+ /* set VCE_UENC_REG_CLOCK_GATING always in dynamic mode */
+ WREG32(mmVCE_UENC_REG_CLOCK_GATING, 0x00);
- if (gated)
- WREG32(mmVCE_CGTT_CLK_OVERRIDE, 0);
- WREG32_P(mmVCE_SOFT_RESET, 0, ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+ WREG32_P(mmVCE_LMI_CTRL2, 0, ~0x100);
+ if(!gated) {
+ WREG32_P(mmVCE_VCPU_CNTL, VCE_VCPU_CNTL__CLK_EN_MASK, ~VCE_VCPU_CNTL__CLK_EN_MASK);
+ mdelay(100);
+ WREG32_P(mmVCE_SOFT_RESET, 0, ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+
+ vce_v2_0_firmware_loaded(adev);
+ WREG32_P(mmVCE_STATUS, 0, ~VCE_STATUS__JOB_BUSY_MASK);
+ }
}
static void vce_v2_0_disable_cg(struct amdgpu_device *adev)
WREG32(mmVCE_VCPU_CACHE_SIZE2, size);
WREG32_P(mmVCE_LMI_CTRL2, 0x0, ~0x100);
-
- WREG32_P(mmVCE_SYS_INT_EN, VCE_SYS_INT_EN__VCE_SYS_INT_TRAP_INTERRUPT_EN_MASK,
- ~VCE_SYS_INT_EN__VCE_SYS_INT_TRAP_INTERRUPT_EN_MASK);
+ WREG32_FIELD(VCE_SYS_INT_EN, VCE_SYS_INT_TRAP_INTERRUPT_EN, 1);
vce_v2_0_init_cg(adev);
}
static int vce_v2_0_wait_for_idle(void *handle)
{
- unsigned i;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ unsigned i;
for (i = 0; i < adev->usec_timeout; i++) {
- if (!(RREG32(mmSRBM_STATUS2) & SRBM_STATUS2__VCE_BUSY_MASK))
+ if (vce_v2_0_is_idle(handle))
return 0;
}
return -ETIMEDOUT;
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
- WREG32_P(mmSRBM_SOFT_RESET, SRBM_SOFT_RESET__SOFT_RESET_VCE_MASK,
- ~SRBM_SOFT_RESET__SOFT_RESET_VCE_MASK);
+ WREG32_FIELD(SRBM_SOFT_RESET, SOFT_RESET_VCE, 1);
mdelay(5);
return vce_v2_0_start(adev);
DRM_DEBUG("IH: VCE\n");
switch (entry->src_data) {
case 0:
- amdgpu_fence_process(&adev->vce.ring[0]);
- break;
case 1:
- amdgpu_fence_process(&adev->vce.ring[1]);
+ amdgpu_fence_process(&adev->vce.ring[entry->src_data]);
break;
default:
DRM_ERROR("Unhandled interrupt: %d %d\n",
return 0;
}
+static void vce_v2_0_set_bypass_mode(struct amdgpu_device *adev, bool enable)
+{
+ u32 tmp = RREG32_SMC(ixGCK_DFS_BYPASS_CNTL);
+
+ if (enable)
+ tmp |= GCK_DFS_BYPASS_CNTL__BYPASSECLK_MASK;
+ else
+ tmp &= ~GCK_DFS_BYPASS_CNTL__BYPASSECLK_MASK;
+
+ WREG32_SMC(ixGCK_DFS_BYPASS_CNTL, tmp);
+}
+
+
static int vce_v2_0_set_clockgating_state(void *handle,
enum amd_clockgating_state state)
{
bool gate = false;
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ bool enable = (state == AMD_CG_STATE_GATE) ? true : false;
+
+
+ vce_v2_0_set_bypass_mode(adev, enable);
if (state == AMD_CG_STATE_GATE)
gate = true;
.pad_ib = amdgpu_ring_generic_pad_ib,
.begin_use = amdgpu_vce_ring_begin_use,
.end_use = amdgpu_vce_ring_end_use,
+ .get_emit_ib_size = amdgpu_vce_ring_get_emit_ib_size,
+ .get_dma_frame_size = amdgpu_vce_ring_get_dma_frame_size,
};
static void vce_v2_0_set_ring_funcs(struct amdgpu_device *adev)
{
- adev->vce.ring[0].funcs = &vce_v2_0_ring_funcs;
- adev->vce.ring[1].funcs = &vce_v2_0_ring_funcs;
+ int i;
+
+ for (i = 0; i < adev->vce.num_rings; i++)
+ adev->vce.ring[i].funcs = &vce_v2_0_ring_funcs;
}
static const struct amdgpu_irq_src_funcs vce_v2_0_irq_funcs = {
#include "gca/gfx_8_0_d.h"
#include "smu/smu_7_1_2_d.h"
#include "smu/smu_7_1_2_sh_mask.h"
+#include "gca/gfx_8_0_d.h"
+#include "gca/gfx_8_0_sh_mask.h"
+
#define GRBM_GFX_INDEX__VCE_INSTANCE__SHIFT 0x04
#define GRBM_GFX_INDEX__VCE_INSTANCE_MASK 0x10
if (ring == &adev->vce.ring[0])
return RREG32(mmVCE_RB_RPTR);
- else
+ else if (ring == &adev->vce.ring[1])
return RREG32(mmVCE_RB_RPTR2);
+ else
+ return RREG32(mmVCE_RB_RPTR3);
}
/**
if (ring == &adev->vce.ring[0])
return RREG32(mmVCE_RB_WPTR);
- else
+ else if (ring == &adev->vce.ring[1])
return RREG32(mmVCE_RB_WPTR2);
+ else
+ return RREG32(mmVCE_RB_WPTR3);
}
/**
if (ring == &adev->vce.ring[0])
WREG32(mmVCE_RB_WPTR, ring->wptr);
- else
+ else if (ring == &adev->vce.ring[1])
WREG32(mmVCE_RB_WPTR2, ring->wptr);
+ else
+ WREG32(mmVCE_RB_WPTR3, ring->wptr);
}
static void vce_v3_0_override_vce_clock_gating(struct amdgpu_device *adev, bool override)
{
- u32 tmp, data;
-
- tmp = data = RREG32(mmVCE_RB_ARB_CTRL);
- if (override)
- data |= VCE_RB_ARB_CTRL__VCE_CGTT_OVERRIDE_MASK;
- else
- data &= ~VCE_RB_ARB_CTRL__VCE_CGTT_OVERRIDE_MASK;
-
- if (tmp != data)
- WREG32(mmVCE_RB_ARB_CTRL, data);
+ WREG32_FIELD(VCE_RB_ARB_CTRL, VCE_CGTT_OVERRIDE, override ? 1 : 0);
}
static void vce_v3_0_set_vce_sw_clock_gating(struct amdgpu_device *adev,
bool gated)
{
- u32 tmp, data;
+ u32 data;
+
/* Set Override to disable Clock Gating */
vce_v3_0_override_vce_clock_gating(adev, true);
- if (!gated) {
- /* Force CLOCK ON for VCE_CLOCK_GATING_B,
- * {*_FORCE_ON, *_FORCE_OFF} = {1, 0}
- * VREG can be FORCE ON or set to Dynamic, but can't be OFF
- */
- tmp = data = RREG32(mmVCE_CLOCK_GATING_B);
+ /* This function enables MGCG which is controlled by firmware.
+ With the clocks in the gated state the core is still
+ accessible but the firmware will throttle the clocks on the
+ fly as necessary.
+ */
+ if (gated) {
+ data = RREG32(mmVCE_CLOCK_GATING_B);
data |= 0x1ff;
data &= ~0xef0000;
- if (tmp != data)
- WREG32(mmVCE_CLOCK_GATING_B, data);
+ WREG32(mmVCE_CLOCK_GATING_B, data);
- /* Force CLOCK ON for VCE_UENC_CLOCK_GATING,
- * {*_FORCE_ON, *_FORCE_OFF} = {1, 0}
- */
- tmp = data = RREG32(mmVCE_UENC_CLOCK_GATING);
+ data = RREG32(mmVCE_UENC_CLOCK_GATING);
data |= 0x3ff000;
data &= ~0xffc00000;
- if (tmp != data)
- WREG32(mmVCE_UENC_CLOCK_GATING, data);
+ WREG32(mmVCE_UENC_CLOCK_GATING, data);
- /* set VCE_UENC_CLOCK_GATING_2 */
- tmp = data = RREG32(mmVCE_UENC_CLOCK_GATING_2);
+ data = RREG32(mmVCE_UENC_CLOCK_GATING_2);
data |= 0x2;
- data &= ~0x2;
- if (tmp != data)
- WREG32(mmVCE_UENC_CLOCK_GATING_2, data);
+ data &= ~0x00010000;
+ WREG32(mmVCE_UENC_CLOCK_GATING_2, data);
- /* Force CLOCK ON for VCE_UENC_REG_CLOCK_GATING */
- tmp = data = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
+ data = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
data |= 0x37f;
- if (tmp != data)
- WREG32(mmVCE_UENC_REG_CLOCK_GATING, data);
+ WREG32(mmVCE_UENC_REG_CLOCK_GATING, data);
- /* Force VCE_UENC_DMA_DCLK_CTRL Clock ON */
- tmp = data = RREG32(mmVCE_UENC_DMA_DCLK_CTRL);
+ data = RREG32(mmVCE_UENC_DMA_DCLK_CTRL);
data |= VCE_UENC_DMA_DCLK_CTRL__WRDMCLK_FORCEON_MASK |
- VCE_UENC_DMA_DCLK_CTRL__RDDMCLK_FORCEON_MASK |
- VCE_UENC_DMA_DCLK_CTRL__REGCLK_FORCEON_MASK |
- 0x8;
- if (tmp != data)
- WREG32(mmVCE_UENC_DMA_DCLK_CTRL, data);
+ VCE_UENC_DMA_DCLK_CTRL__RDDMCLK_FORCEON_MASK |
+ VCE_UENC_DMA_DCLK_CTRL__REGCLK_FORCEON_MASK |
+ 0x8;
+ WREG32(mmVCE_UENC_DMA_DCLK_CTRL, data);
} else {
- /* Force CLOCK OFF for VCE_CLOCK_GATING_B,
- * {*, *_FORCE_OFF} = {*, 1}
- * set VREG to Dynamic, as it can't be OFF
- */
- tmp = data = RREG32(mmVCE_CLOCK_GATING_B);
+ data = RREG32(mmVCE_CLOCK_GATING_B);
data &= ~0x80010;
data |= 0xe70008;
- if (tmp != data)
- WREG32(mmVCE_CLOCK_GATING_B, data);
- /* Force CLOCK OFF for VCE_UENC_CLOCK_GATING,
- * Force ClOCK OFF takes precedent over Force CLOCK ON setting.
- * {*_FORCE_ON, *_FORCE_OFF} = {*, 1}
- */
- tmp = data = RREG32(mmVCE_UENC_CLOCK_GATING);
+ WREG32(mmVCE_CLOCK_GATING_B, data);
+
+ data = RREG32(mmVCE_UENC_CLOCK_GATING);
data |= 0xffc00000;
- if (tmp != data)
- WREG32(mmVCE_UENC_CLOCK_GATING, data);
- /* Set VCE_UENC_CLOCK_GATING_2 */
- tmp = data = RREG32(mmVCE_UENC_CLOCK_GATING_2);
+ WREG32(mmVCE_UENC_CLOCK_GATING, data);
+
+ data = RREG32(mmVCE_UENC_CLOCK_GATING_2);
data |= 0x10000;
- if (tmp != data)
- WREG32(mmVCE_UENC_CLOCK_GATING_2, data);
- /* Set VCE_UENC_REG_CLOCK_GATING to dynamic */
- tmp = data = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
+ WREG32(mmVCE_UENC_CLOCK_GATING_2, data);
+
+ data = RREG32(mmVCE_UENC_REG_CLOCK_GATING);
data &= ~0xffc00000;
- if (tmp != data)
- WREG32(mmVCE_UENC_REG_CLOCK_GATING, data);
- /* Set VCE_UENC_DMA_DCLK_CTRL CG always in dynamic mode */
- tmp = data = RREG32(mmVCE_UENC_DMA_DCLK_CTRL);
+ WREG32(mmVCE_UENC_REG_CLOCK_GATING, data);
+
+ data = RREG32(mmVCE_UENC_DMA_DCLK_CTRL);
data &= ~(VCE_UENC_DMA_DCLK_CTRL__WRDMCLK_FORCEON_MASK |
- VCE_UENC_DMA_DCLK_CTRL__RDDMCLK_FORCEON_MASK |
- VCE_UENC_DMA_DCLK_CTRL__REGCLK_FORCEON_MASK |
- 0x8);
- if (tmp != data)
- WREG32(mmVCE_UENC_DMA_DCLK_CTRL, data);
+ VCE_UENC_DMA_DCLK_CTRL__RDDMCLK_FORCEON_MASK |
+ VCE_UENC_DMA_DCLK_CTRL__REGCLK_FORCEON_MASK |
+ 0x8);
+ WREG32(mmVCE_UENC_DMA_DCLK_CTRL, data);
}
vce_v3_0_override_vce_clock_gating(adev, false);
}
}
DRM_ERROR("VCE not responding, trying to reset the ECPU!!!\n");
- WREG32_P(mmVCE_SOFT_RESET,
- VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
- ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+ WREG32_FIELD(VCE_SOFT_RESET, ECPU_SOFT_RESET, 1);
mdelay(10);
- WREG32_P(mmVCE_SOFT_RESET, 0,
- ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+ WREG32_FIELD(VCE_SOFT_RESET, ECPU_SOFT_RESET, 0);
mdelay(10);
}
WREG32(mmVCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
WREG32(mmVCE_RB_SIZE2, ring->ring_size / 4);
+ ring = &adev->vce.ring[2];
+ WREG32(mmVCE_RB_RPTR3, ring->wptr);
+ WREG32(mmVCE_RB_WPTR3, ring->wptr);
+ WREG32(mmVCE_RB_BASE_LO3, ring->gpu_addr);
+ WREG32(mmVCE_RB_BASE_HI3, upper_32_bits(ring->gpu_addr));
+ WREG32(mmVCE_RB_SIZE3, ring->ring_size / 4);
+
mutex_lock(&adev->grbm_idx_mutex);
for (idx = 0; idx < 2; ++idx) {
if (adev->vce.harvest_config & (1 << idx))
continue;
- if (idx == 0)
- WREG32_P(mmGRBM_GFX_INDEX, 0,
- ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
- else
- WREG32_P(mmGRBM_GFX_INDEX,
- GRBM_GFX_INDEX__VCE_INSTANCE_MASK,
- ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
-
+ WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, idx);
vce_v3_0_mc_resume(adev, idx);
-
- WREG32_P(mmVCE_STATUS, VCE_STATUS__JOB_BUSY_MASK,
- ~VCE_STATUS__JOB_BUSY_MASK);
+ WREG32_FIELD(VCE_STATUS, JOB_BUSY, 1);
if (adev->asic_type >= CHIP_STONEY)
WREG32_P(mmVCE_VCPU_CNTL, 1, ~0x200001);
else
- WREG32_P(mmVCE_VCPU_CNTL, VCE_VCPU_CNTL__CLK_EN_MASK,
- ~VCE_VCPU_CNTL__CLK_EN_MASK);
-
- WREG32_P(mmVCE_SOFT_RESET, 0,
- ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+ WREG32_FIELD(VCE_VCPU_CNTL, CLK_EN, 1);
+ WREG32_FIELD(VCE_SOFT_RESET, ECPU_SOFT_RESET, 0);
mdelay(100);
r = vce_v3_0_firmware_loaded(adev);
/* clear BUSY flag */
- WREG32_P(mmVCE_STATUS, 0, ~VCE_STATUS__JOB_BUSY_MASK);
-
- /* Set Clock-Gating off */
- if (adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG)
- vce_v3_0_set_vce_sw_clock_gating(adev, false);
+ WREG32_FIELD(VCE_STATUS, JOB_BUSY, 0);
if (r) {
DRM_ERROR("VCE not responding, giving up!!!\n");
}
}
- WREG32_P(mmGRBM_GFX_INDEX, 0, ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
+ WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, 0);
mutex_unlock(&adev->grbm_idx_mutex);
return 0;
if (adev->vce.harvest_config & (1 << idx))
continue;
- if (idx == 0)
- WREG32_P(mmGRBM_GFX_INDEX, 0,
- ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
- else
- WREG32_P(mmGRBM_GFX_INDEX,
- GRBM_GFX_INDEX__VCE_INSTANCE_MASK,
- ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
+ WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, idx);
if (adev->asic_type >= CHIP_STONEY)
WREG32_P(mmVCE_VCPU_CNTL, 0, ~0x200001);
else
- WREG32_P(mmVCE_VCPU_CNTL, 0,
- ~VCE_VCPU_CNTL__CLK_EN_MASK);
+ WREG32_FIELD(VCE_VCPU_CNTL, CLK_EN, 0);
+
/* hold on ECPU */
- WREG32_P(mmVCE_SOFT_RESET,
- VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK,
- ~VCE_SOFT_RESET__ECPU_SOFT_RESET_MASK);
+ WREG32_FIELD(VCE_SOFT_RESET, ECPU_SOFT_RESET, 1);
/* clear BUSY flag */
- WREG32_P(mmVCE_STATUS, 0, ~VCE_STATUS__JOB_BUSY_MASK);
+ WREG32_FIELD(VCE_STATUS, JOB_BUSY, 0);
/* Set Clock-Gating off */
if (adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG)
vce_v3_0_set_vce_sw_clock_gating(adev, false);
}
- WREG32_P(mmGRBM_GFX_INDEX, 0, ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
+ WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, 0);
mutex_unlock(&adev->grbm_idx_mutex);
return 0;
(AMDGPU_VCE_HARVEST_VCE0 | AMDGPU_VCE_HARVEST_VCE1))
return -ENOENT;
+ adev->vce.num_rings = 3;
+
vce_v3_0_set_ring_funcs(adev);
vce_v3_0_set_irq_funcs(adev);
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
struct amdgpu_ring *ring;
- int r;
+ int r, i;
/* VCE */
r = amdgpu_irq_add_id(adev, 167, &adev->vce.irq);
if (r)
return r;
- ring = &adev->vce.ring[0];
- sprintf(ring->name, "vce0");
- r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf,
- &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE);
- if (r)
- return r;
-
- ring = &adev->vce.ring[1];
- sprintf(ring->name, "vce1");
- r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf,
- &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE);
- if (r)
- return r;
+ for (i = 0; i < adev->vce.num_rings; i++) {
+ ring = &adev->vce.ring[i];
+ sprintf(ring->name, "vce%d", i);
+ r = amdgpu_ring_init(adev, ring, 512, VCE_CMD_NO_OP, 0xf,
+ &adev->vce.irq, 0, AMDGPU_RING_TYPE_VCE);
+ if (r)
+ return r;
+ }
return r;
}
if (r)
return r;
- adev->vce.ring[0].ready = false;
- adev->vce.ring[1].ready = false;
+ for (i = 0; i < adev->vce.num_rings; i++)
+ adev->vce.ring[i].ready = false;
- for (i = 0; i < 2; i++) {
+ for (i = 0; i < adev->vce.num_rings; i++) {
r = amdgpu_ring_test_ring(&adev->vce.ring[i]);
if (r)
return r;
WREG32_P(mmVCE_CLOCK_GATING_A, 0, ~(1 << 16));
WREG32_P(mmVCE_UENC_CLOCK_GATING, 0x1FF000, ~0xFF9FF000);
WREG32_P(mmVCE_UENC_REG_CLOCK_GATING, 0x3F, ~0x3F);
- WREG32(mmVCE_CLOCK_GATING_B, 0xf7);
+ WREG32(mmVCE_CLOCK_GATING_B, 0x1FF);
WREG32(mmVCE_LMI_CTRL, 0x00398000);
WREG32_P(mmVCE_LMI_CACHE_CTRL, 0x0, ~0x1);
}
WREG32_P(mmVCE_LMI_CTRL2, 0x0, ~0x100);
-
- WREG32_P(mmVCE_SYS_INT_EN, VCE_SYS_INT_EN__VCE_SYS_INT_TRAP_INTERRUPT_EN_MASK,
- ~VCE_SYS_INT_EN__VCE_SYS_INT_TRAP_INTERRUPT_EN_MASK);
+ WREG32_FIELD(VCE_SYS_INT_EN, VCE_SYS_INT_TRAP_INTERRUPT_EN, 1);
}
static bool vce_v3_0_is_idle(void *handle)
return -ETIMEDOUT;
}
+#define VCE_STATUS_VCPU_REPORT_AUTO_BUSY_MASK 0x00000008L /* AUTO_BUSY */
+#define VCE_STATUS_VCPU_REPORT_RB0_BUSY_MASK 0x00000010L /* RB0_BUSY */
+#define VCE_STATUS_VCPU_REPORT_RB1_BUSY_MASK 0x00000020L /* RB1_BUSY */
+#define AMDGPU_VCE_STATUS_BUSY_MASK (VCE_STATUS_VCPU_REPORT_AUTO_BUSY_MASK | \
+ VCE_STATUS_VCPU_REPORT_RB0_BUSY_MASK)
+
+static int vce_v3_0_check_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+ u32 srbm_soft_reset = 0;
+
+ /* According to VCE team , we should use VCE_STATUS instead
+ * SRBM_STATUS.VCE_BUSY bit for busy status checking.
+ * GRBM_GFX_INDEX.INSTANCE_INDEX is used to specify which VCE
+ * instance's registers are accessed
+ * (0 for 1st instance, 10 for 2nd instance).
+ *
+ *VCE_STATUS
+ *|UENC|ACPI|AUTO ACTIVE|RB1 |RB0 |RB2 | |FW_LOADED|JOB |
+ *|----+----+-----------+----+----+----+----------+---------+----|
+ *|bit8|bit7| bit6 |bit5|bit4|bit3| bit2 | bit1 |bit0|
+ *
+ * VCE team suggest use bit 3--bit 6 for busy status check
+ */
+ mutex_lock(&adev->grbm_idx_mutex);
+ WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0);
+ if (RREG32(mmVCE_STATUS) & AMDGPU_VCE_STATUS_BUSY_MASK) {
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE0, 1);
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE1, 1);
+ }
+ WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0x10);
+ if (RREG32(mmVCE_STATUS) & AMDGPU_VCE_STATUS_BUSY_MASK) {
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE0, 1);
+ srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_VCE1, 1);
+ }
+ WREG32_FIELD(GRBM_GFX_INDEX, INSTANCE_INDEX, 0);
+
+ if (srbm_soft_reset) {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_VCE].hang = true;
+ adev->vce.srbm_soft_reset = srbm_soft_reset;
+ } else {
+ adev->ip_block_status[AMD_IP_BLOCK_TYPE_VCE].hang = false;
+ adev->vce.srbm_soft_reset = 0;
+ }
+ mutex_unlock(&adev->grbm_idx_mutex);
+ return 0;
+}
+
static int vce_v3_0_soft_reset(void *handle)
{
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
- u32 mask = 0;
+ u32 srbm_soft_reset;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_VCE].hang)
+ return 0;
+ srbm_soft_reset = adev->vce.srbm_soft_reset;
+
+ if (srbm_soft_reset) {
+ u32 tmp;
- mask |= (adev->vce.harvest_config & AMDGPU_VCE_HARVEST_VCE0) ? 0 : SRBM_SOFT_RESET__SOFT_RESET_VCE0_MASK;
- mask |= (adev->vce.harvest_config & AMDGPU_VCE_HARVEST_VCE1) ? 0 : SRBM_SOFT_RESET__SOFT_RESET_VCE1_MASK;
+ tmp = RREG32(mmSRBM_SOFT_RESET);
+ tmp |= srbm_soft_reset;
+ dev_info(adev->dev, "SRBM_SOFT_RESET=0x%08X\n", tmp);
+ WREG32(mmSRBM_SOFT_RESET, tmp);
+ tmp = RREG32(mmSRBM_SOFT_RESET);
+
+ udelay(50);
+
+ tmp &= ~srbm_soft_reset;
+ WREG32(mmSRBM_SOFT_RESET, tmp);
+ tmp = RREG32(mmSRBM_SOFT_RESET);
+
+ /* Wait a little for things to settle down */
+ udelay(50);
+ }
+
+ return 0;
+}
+
+static int vce_v3_0_pre_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_VCE].hang)
+ return 0;
- WREG32_P(mmSRBM_SOFT_RESET, mask,
- ~(SRBM_SOFT_RESET__SOFT_RESET_VCE0_MASK |
- SRBM_SOFT_RESET__SOFT_RESET_VCE1_MASK));
mdelay(5);
- return vce_v3_0_start(adev);
+ return vce_v3_0_suspend(adev);
+}
+
+
+static int vce_v3_0_post_soft_reset(void *handle)
+{
+ struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+ if (!adev->ip_block_status[AMD_IP_BLOCK_TYPE_VCE].hang)
+ return 0;
+
+ mdelay(5);
+
+ return vce_v3_0_resume(adev);
}
static int vce_v3_0_set_interrupt_state(struct amdgpu_device *adev,
{
DRM_DEBUG("IH: VCE\n");
- WREG32_P(mmVCE_SYS_INT_STATUS,
- VCE_SYS_INT_STATUS__VCE_SYS_INT_TRAP_INTERRUPT_INT_MASK,
- ~VCE_SYS_INT_STATUS__VCE_SYS_INT_TRAP_INTERRUPT_INT_MASK);
+ WREG32_FIELD(VCE_SYS_INT_STATUS, VCE_SYS_INT_TRAP_INTERRUPT_INT, 1);
switch (entry->src_data) {
case 0:
case 1:
+ case 2:
amdgpu_fence_process(&adev->vce.ring[entry->src_data]);
break;
default:
return 0;
}
-static void vce_v3_set_bypass_mode(struct amdgpu_device *adev, bool enable)
+static void vce_v3_0_set_bypass_mode(struct amdgpu_device *adev, bool enable)
{
u32 tmp = RREG32_SMC(ixGCK_DFS_BYPASS_CNTL);
bool enable = (state == AMD_CG_STATE_GATE) ? true : false;
int i;
- if (adev->asic_type == CHIP_POLARIS10)
- vce_v3_set_bypass_mode(adev, enable);
+ if ((adev->asic_type == CHIP_POLARIS10) ||
+ (adev->asic_type == CHIP_TONGA))
+ vce_v3_0_set_bypass_mode(adev, enable);
if (!(adev->cg_flags & AMD_CG_SUPPORT_VCE_MGCG))
return 0;
if (adev->vce.harvest_config & (1 << i))
continue;
- if (i == 0)
- WREG32_P(mmGRBM_GFX_INDEX, 0,
- ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
- else
- WREG32_P(mmGRBM_GFX_INDEX,
- GRBM_GFX_INDEX__VCE_INSTANCE_MASK,
- ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
+ WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, i);
if (enable) {
/* initialize VCE_CLOCK_GATING_A: Clock ON/OFF delay */
vce_v3_0_set_vce_sw_clock_gating(adev, enable);
}
- WREG32_P(mmGRBM_GFX_INDEX, 0, ~GRBM_GFX_INDEX__VCE_INSTANCE_MASK);
+ WREG32_FIELD(GRBM_GFX_INDEX, VCE_INSTANCE, 0);
mutex_unlock(&adev->grbm_idx_mutex);
return 0;
return vce_v3_0_start(adev);
}
+static void vce_v3_0_ring_emit_ib(struct amdgpu_ring *ring,
+ struct amdgpu_ib *ib, unsigned int vm_id, bool ctx_switch)
+{
+ amdgpu_ring_write(ring, VCE_CMD_IB_VM);
+ amdgpu_ring_write(ring, vm_id);
+ amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
+ amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr));
+ amdgpu_ring_write(ring, ib->length_dw);
+}
+
+static void vce_v3_0_emit_vm_flush(struct amdgpu_ring *ring,
+ unsigned int vm_id, uint64_t pd_addr)
+{
+ amdgpu_ring_write(ring, VCE_CMD_UPDATE_PTB);
+ amdgpu_ring_write(ring, vm_id);
+ amdgpu_ring_write(ring, pd_addr >> 12);
+
+ amdgpu_ring_write(ring, VCE_CMD_FLUSH_TLB);
+ amdgpu_ring_write(ring, vm_id);
+ amdgpu_ring_write(ring, VCE_CMD_END);
+}
+
+static void vce_v3_0_emit_pipeline_sync(struct amdgpu_ring *ring)
+{
+ uint32_t seq = ring->fence_drv.sync_seq;
+ uint64_t addr = ring->fence_drv.gpu_addr;
+
+ amdgpu_ring_write(ring, VCE_CMD_WAIT_GE);
+ amdgpu_ring_write(ring, lower_32_bits(addr));
+ amdgpu_ring_write(ring, upper_32_bits(addr));
+ amdgpu_ring_write(ring, seq);
+}
+
+static unsigned vce_v3_0_ring_get_emit_ib_size(struct amdgpu_ring *ring)
+{
+ return
+ 5; /* vce_v3_0_ring_emit_ib */
+}
+
+static unsigned vce_v3_0_ring_get_dma_frame_size(struct amdgpu_ring *ring)
+{
+ return
+ 4 + /* vce_v3_0_emit_pipeline_sync */
+ 6; /* amdgpu_vce_ring_emit_fence x1 no user fence */
+}
+
+static unsigned vce_v3_0_ring_get_dma_frame_size_vm(struct amdgpu_ring *ring)
+{
+ return
+ 6 + /* vce_v3_0_emit_vm_flush */
+ 4 + /* vce_v3_0_emit_pipeline_sync */
+ 6 + 6; /* amdgpu_vce_ring_emit_fence x2 vm fence */
+}
+
const struct amd_ip_funcs vce_v3_0_ip_funcs = {
.name = "vce_v3_0",
.early_init = vce_v3_0_early_init,
.resume = vce_v3_0_resume,
.is_idle = vce_v3_0_is_idle,
.wait_for_idle = vce_v3_0_wait_for_idle,
+ .check_soft_reset = vce_v3_0_check_soft_reset,
+ .pre_soft_reset = vce_v3_0_pre_soft_reset,
.soft_reset = vce_v3_0_soft_reset,
+ .post_soft_reset = vce_v3_0_post_soft_reset,
.set_clockgating_state = vce_v3_0_set_clockgating_state,
.set_powergating_state = vce_v3_0_set_powergating_state,
};
-static const struct amdgpu_ring_funcs vce_v3_0_ring_funcs = {
+static const struct amdgpu_ring_funcs vce_v3_0_ring_phys_funcs = {
.get_rptr = vce_v3_0_ring_get_rptr,
.get_wptr = vce_v3_0_ring_get_wptr,
.set_wptr = vce_v3_0_ring_set_wptr,
.pad_ib = amdgpu_ring_generic_pad_ib,
.begin_use = amdgpu_vce_ring_begin_use,
.end_use = amdgpu_vce_ring_end_use,
+ .get_emit_ib_size = vce_v3_0_ring_get_emit_ib_size,
+ .get_dma_frame_size = vce_v3_0_ring_get_dma_frame_size,
+};
+
+static const struct amdgpu_ring_funcs vce_v3_0_ring_vm_funcs = {
+ .get_rptr = vce_v3_0_ring_get_rptr,
+ .get_wptr = vce_v3_0_ring_get_wptr,
+ .set_wptr = vce_v3_0_ring_set_wptr,
+ .parse_cs = NULL,
+ .emit_ib = vce_v3_0_ring_emit_ib,
+ .emit_vm_flush = vce_v3_0_emit_vm_flush,
+ .emit_pipeline_sync = vce_v3_0_emit_pipeline_sync,
+ .emit_fence = amdgpu_vce_ring_emit_fence,
+ .test_ring = amdgpu_vce_ring_test_ring,
+ .test_ib = amdgpu_vce_ring_test_ib,
+ .insert_nop = amdgpu_ring_insert_nop,
+ .pad_ib = amdgpu_ring_generic_pad_ib,
+ .begin_use = amdgpu_vce_ring_begin_use,
+ .end_use = amdgpu_vce_ring_end_use,
+ .get_emit_ib_size = vce_v3_0_ring_get_emit_ib_size,
+ .get_dma_frame_size = vce_v3_0_ring_get_dma_frame_size_vm,
};
static void vce_v3_0_set_ring_funcs(struct amdgpu_device *adev)
{
- adev->vce.ring[0].funcs = &vce_v3_0_ring_funcs;
- adev->vce.ring[1].funcs = &vce_v3_0_ring_funcs;
+ int i;
+
+ if (adev->asic_type >= CHIP_STONEY) {
+ for (i = 0; i < adev->vce.num_rings; i++)
+ adev->vce.ring[i].funcs = &vce_v3_0_ring_vm_funcs;
+ DRM_INFO("VCE enabled in VM mode\n");
+ } else {
+ for (i = 0; i < adev->vce.num_rings; i++)
+ adev->vce.ring[i].funcs = &vce_v3_0_ring_phys_funcs;
+ DRM_INFO("VCE enabled in physical mode\n");
+ }
}
static const struct amdgpu_irq_src_funcs vce_v3_0_irq_funcs = {
#if defined(CONFIG_DRM_AMD_ACP)
#include "amdgpu_acp.h"
#endif
+#include "dce_virtual.h"
MODULE_FIRMWARE("amdgpu/polaris10_smc.bin");
MODULE_FIRMWARE("amdgpu/polaris10_smc_sk.bin");
},
};
+static const struct amdgpu_ip_block_version topaz_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vi_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 7,
+ .minor = 4,
+ .rev = 0,
+ .funcs = &gmc_v7_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 2,
+ .minor = 4,
+ .rev = 0,
+ .funcs = &iceland_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 1,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 8,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gfx_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 2,
+ .minor = 4,
+ .rev = 0,
+ .funcs = &sdma_v2_4_ip_funcs,
+ },
+};
+
static const struct amdgpu_ip_block_version tonga_ip_blocks[] =
{
/* ORDER MATTERS! */
},
};
+static const struct amdgpu_ip_block_version tonga_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vi_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 8,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &tonga_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 10,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 8,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gfx_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &sdma_v3_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 5,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &uvd_v5_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vce_v3_0_ip_funcs,
+ },
+};
+
static const struct amdgpu_ip_block_version fiji_ip_blocks[] =
{
/* ORDER MATTERS! */
},
};
+static const struct amdgpu_ip_block_version fiji_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vi_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 8,
+ .minor = 5,
+ .rev = 0,
+ .funcs = &gmc_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &tonga_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 10,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 8,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gfx_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &sdma_v3_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &uvd_v6_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vce_v3_0_ip_funcs,
+ },
+};
+
static const struct amdgpu_ip_block_version polaris11_ip_blocks[] =
{
/* ORDER MATTERS! */
},
};
+static const struct amdgpu_ip_block_version polaris11_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vi_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 8,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &gmc_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 3,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &tonga_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 7,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 11,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 8,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gfx_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 3,
+ .minor = 1,
+ .rev = 0,
+ .funcs = &sdma_v3_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 6,
+ .minor = 3,
+ .rev = 0,
+ .funcs = &uvd_v6_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 3,
+ .minor = 4,
+ .rev = 0,
+ .funcs = &vce_v3_0_ip_funcs,
+ },
+};
+
static const struct amdgpu_ip_block_version cz_ip_blocks[] =
{
/* ORDER MATTERS! */
#endif
};
+static const struct amdgpu_ip_block_version cz_ip_blocks_vd[] =
+{
+ /* ORDER MATTERS! */
+ {
+ .type = AMD_IP_BLOCK_TYPE_COMMON,
+ .major = 2,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vi_common_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GMC,
+ .major = 8,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gmc_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_IH,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &cz_ih_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SMC,
+ .major = 8,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &amdgpu_pp_ip_funcs
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_DCE,
+ .major = 11,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &dce_virtual_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_GFX,
+ .major = 8,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &gfx_v8_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_SDMA,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &sdma_v3_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_UVD,
+ .major = 6,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &uvd_v6_0_ip_funcs,
+ },
+ {
+ .type = AMD_IP_BLOCK_TYPE_VCE,
+ .major = 3,
+ .minor = 0,
+ .rev = 0,
+ .funcs = &vce_v3_0_ip_funcs,
+ },
+#if defined(CONFIG_DRM_AMD_ACP)
+ {
+ .type = AMD_IP_BLOCK_TYPE_ACP,
+ .major = 2,
+ .minor = 2,
+ .rev = 0,
+ .funcs = &acp_ip_funcs,
+ },
+#endif
+};
+
int vi_set_ip_blocks(struct amdgpu_device *adev)
{
- switch (adev->asic_type) {
- case CHIP_TOPAZ:
- adev->ip_blocks = topaz_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(topaz_ip_blocks);
- break;
- case CHIP_FIJI:
- adev->ip_blocks = fiji_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(fiji_ip_blocks);
- break;
- case CHIP_TONGA:
- adev->ip_blocks = tonga_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(tonga_ip_blocks);
- break;
- case CHIP_POLARIS11:
- case CHIP_POLARIS10:
- adev->ip_blocks = polaris11_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(polaris11_ip_blocks);
- break;
- case CHIP_CARRIZO:
- case CHIP_STONEY:
- adev->ip_blocks = cz_ip_blocks;
- adev->num_ip_blocks = ARRAY_SIZE(cz_ip_blocks);
- break;
- default:
- /* FIXME: not supported yet */
- return -EINVAL;
+ if (adev->enable_virtual_display) {
+ switch (adev->asic_type) {
+ case CHIP_TOPAZ:
+ adev->ip_blocks = topaz_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(topaz_ip_blocks_vd);
+ break;
+ case CHIP_FIJI:
+ adev->ip_blocks = fiji_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(fiji_ip_blocks_vd);
+ break;
+ case CHIP_TONGA:
+ adev->ip_blocks = tonga_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(tonga_ip_blocks_vd);
+ break;
+ case CHIP_POLARIS11:
+ case CHIP_POLARIS10:
+ adev->ip_blocks = polaris11_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(polaris11_ip_blocks_vd);
+ break;
+
+ case CHIP_CARRIZO:
+ case CHIP_STONEY:
+ adev->ip_blocks = cz_ip_blocks_vd;
+ adev->num_ip_blocks = ARRAY_SIZE(cz_ip_blocks_vd);
+ break;
+ default:
+ /* FIXME: not supported yet */
+ return -EINVAL;
+ }
+ } else {
+ switch (adev->asic_type) {
+ case CHIP_TOPAZ:
+ adev->ip_blocks = topaz_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(topaz_ip_blocks);
+ break;
+ case CHIP_FIJI:
+ adev->ip_blocks = fiji_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(fiji_ip_blocks);
+ break;
+ case CHIP_TONGA:
+ adev->ip_blocks = tonga_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(tonga_ip_blocks);
+ break;
+ case CHIP_POLARIS11:
+ case CHIP_POLARIS10:
+ adev->ip_blocks = polaris11_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(polaris11_ip_blocks);
+ break;
+ case CHIP_CARRIZO:
+ case CHIP_STONEY:
+ adev->ip_blocks = cz_ip_blocks;
+ adev->num_ip_blocks = ARRAY_SIZE(cz_ip_blocks);
+ break;
+ default:
+ /* FIXME: not supported yet */
+ return -EINVAL;
+ }
}
return 0;
AMD_CG_SUPPORT_HDP_MGCG |
AMD_CG_SUPPORT_HDP_LS |
AMD_CG_SUPPORT_SDMA_MGCG |
- AMD_CG_SUPPORT_SDMA_LS;
+ AMD_CG_SUPPORT_SDMA_LS |
+ AMD_CG_SUPPORT_VCE_MGCG;
+ /* rev0 hardware requires workarounds to support PG */
adev->pg_flags = 0;
+ if (adev->rev_id != 0x00) {
+ adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG |
+ AMD_PG_SUPPORT_GFX_SMG |
+ AMD_PG_SUPPORT_GFX_PIPELINE |
+ AMD_PG_SUPPORT_UVD |
+ AMD_PG_SUPPORT_VCE;
+ }
adev->external_rev_id = adev->rev_id + 0x1;
break;
case CHIP_STONEY:
AMD_CG_SUPPORT_HDP_MGCG |
AMD_CG_SUPPORT_HDP_LS |
AMD_CG_SUPPORT_SDMA_MGCG |
- AMD_CG_SUPPORT_SDMA_LS;
- adev->external_rev_id = adev->rev_id + 0x1;
+ AMD_CG_SUPPORT_SDMA_LS |
+ AMD_CG_SUPPORT_VCE_MGCG;
+ adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG |
+ AMD_PG_SUPPORT_GFX_SMG |
+ AMD_PG_SUPPORT_GFX_PIPELINE |
+ AMD_PG_SUPPORT_UVD |
+ AMD_PG_SUPPORT_VCE;
+ adev->external_rev_id = adev->rev_id + 0x61;
break;
default:
/* FIXME: not supported yet */
#define VCE_CMD_IB_AUTO 0x00000005
#define VCE_CMD_SEMAPHORE 0x00000006
+#define VCE_CMD_IB_VM 0x00000102
+#define VCE_CMD_WAIT_GE 0x00000106
+#define VCE_CMD_UPDATE_PTB 0x00000107
+#define VCE_CMD_FLUSH_TLB 0x00000108
#endif
* Supported ASIC types
*/
enum amd_asic_type {
- CHIP_BONAIRE = 0,
+ CHIP_TAHITI = 0,
+ CHIP_PITCAIRN,
+ CHIP_VERDE,
+ CHIP_OLAND,
+ CHIP_HAINAN,
+ CHIP_BONAIRE,
CHIP_KAVERI,
CHIP_KABINI,
CHIP_HAWAII,
bool (*is_idle)(void *handle);
/* poll for idle */
int (*wait_for_idle)(void *handle);
+ /* check soft reset the IP block */
+ int (*check_soft_reset)(void *handle);
+ /* pre soft reset the IP block */
+ int (*pre_soft_reset)(void *handle);
/* soft reset the IP block */
int (*soft_reset)(void *handle);
+ /* post soft reset the IP block */
+ int (*post_soft_reset)(void *handle);
/* enable/disable cg for the IP block */
int (*set_clockgating_state)(void *handle,
enum amd_clockgating_state state);
--- /dev/null
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+static const u32 si_SECT_CONTEXT_def_1[] =
+{
+ 0x00000000, // DB_RENDER_CONTROL
+ 0x00000000, // DB_COUNT_CONTROL
+ 0x00000000, // DB_DEPTH_VIEW
+ 0x00000000, // DB_RENDER_OVERRIDE
+ 0x00000000, // DB_RENDER_OVERRIDE2
+ 0x00000000, // DB_HTILE_DATA_BASE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // DB_DEPTH_BOUNDS_MIN
+ 0x00000000, // DB_DEPTH_BOUNDS_MAX
+ 0x00000000, // DB_STENCIL_CLEAR
+ 0x00000000, // DB_DEPTH_CLEAR
+ 0x00000000, // PA_SC_SCREEN_SCISSOR_TL
+ 0x40004000, // PA_SC_SCREEN_SCISSOR_BR
+ 0, // HOLE
+ 0x00000000, // DB_DEPTH_INFO
+ 0x00000000, // DB_Z_INFO
+ 0x00000000, // DB_STENCIL_INFO
+ 0x00000000, // DB_Z_READ_BASE
+ 0x00000000, // DB_STENCIL_READ_BASE
+ 0x00000000, // DB_Z_WRITE_BASE
+ 0x00000000, // DB_STENCIL_WRITE_BASE
+ 0x00000000, // DB_DEPTH_SIZE
+ 0x00000000, // DB_DEPTH_SLICE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // TA_BC_BASE_ADDR
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // COHER_DEST_BASE_2
+ 0x00000000, // COHER_DEST_BASE_3
+ 0x00000000, // PA_SC_WINDOW_OFFSET
+ 0x80000000, // PA_SC_WINDOW_SCISSOR_TL
+ 0x40004000, // PA_SC_WINDOW_SCISSOR_BR
+ 0x0000ffff, // PA_SC_CLIPRECT_RULE
+ 0x00000000, // PA_SC_CLIPRECT_0_TL
+ 0x40004000, // PA_SC_CLIPRECT_0_BR
+ 0x00000000, // PA_SC_CLIPRECT_1_TL
+ 0x40004000, // PA_SC_CLIPRECT_1_BR
+ 0x00000000, // PA_SC_CLIPRECT_2_TL
+ 0x40004000, // PA_SC_CLIPRECT_2_BR
+ 0x00000000, // PA_SC_CLIPRECT_3_TL
+ 0x40004000, // PA_SC_CLIPRECT_3_BR
+ 0xaa99aaaa, // PA_SC_EDGERULE
+ 0x00000000, // PA_SU_HARDWARE_SCREEN_OFFSET
+ 0xffffffff, // CB_TARGET_MASK
+ 0xffffffff, // CB_SHADER_MASK
+ 0x80000000, // PA_SC_GENERIC_SCISSOR_TL
+ 0x40004000, // PA_SC_GENERIC_SCISSOR_BR
+ 0x00000000, // COHER_DEST_BASE_0
+ 0x00000000, // COHER_DEST_BASE_1
+ 0x80000000, // PA_SC_VPORT_SCISSOR_0_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_0_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_1_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_1_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_2_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_2_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_3_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_3_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_4_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_4_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_5_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_5_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_6_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_6_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_7_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_7_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_8_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_8_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_9_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_9_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_10_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_10_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_11_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_11_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_12_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_12_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_13_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_13_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_14_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_14_BR
+ 0x80000000, // PA_SC_VPORT_SCISSOR_15_TL
+ 0x40004000, // PA_SC_VPORT_SCISSOR_15_BR
+ 0x00000000, // PA_SC_VPORT_ZMIN_0
+ 0x3f800000, // PA_SC_VPORT_ZMAX_0
+ 0x00000000, // PA_SC_VPORT_ZMIN_1
+ 0x3f800000, // PA_SC_VPORT_ZMAX_1
+ 0x00000000, // PA_SC_VPORT_ZMIN_2
+ 0x3f800000, // PA_SC_VPORT_ZMAX_2
+ 0x00000000, // PA_SC_VPORT_ZMIN_3
+ 0x3f800000, // PA_SC_VPORT_ZMAX_3
+ 0x00000000, // PA_SC_VPORT_ZMIN_4
+ 0x3f800000, // PA_SC_VPORT_ZMAX_4
+ 0x00000000, // PA_SC_VPORT_ZMIN_5
+ 0x3f800000, // PA_SC_VPORT_ZMAX_5
+ 0x00000000, // PA_SC_VPORT_ZMIN_6
+ 0x3f800000, // PA_SC_VPORT_ZMAX_6
+ 0x00000000, // PA_SC_VPORT_ZMIN_7
+ 0x3f800000, // PA_SC_VPORT_ZMAX_7
+ 0x00000000, // PA_SC_VPORT_ZMIN_8
+ 0x3f800000, // PA_SC_VPORT_ZMAX_8
+ 0x00000000, // PA_SC_VPORT_ZMIN_9
+ 0x3f800000, // PA_SC_VPORT_ZMAX_9
+ 0x00000000, // PA_SC_VPORT_ZMIN_10
+ 0x3f800000, // PA_SC_VPORT_ZMAX_10
+ 0x00000000, // PA_SC_VPORT_ZMIN_11
+ 0x3f800000, // PA_SC_VPORT_ZMAX_11
+ 0x00000000, // PA_SC_VPORT_ZMIN_12
+ 0x3f800000, // PA_SC_VPORT_ZMAX_12
+ 0x00000000, // PA_SC_VPORT_ZMIN_13
+ 0x3f800000, // PA_SC_VPORT_ZMAX_13
+ 0x00000000, // PA_SC_VPORT_ZMIN_14
+ 0x3f800000, // PA_SC_VPORT_ZMAX_14
+ 0x00000000, // PA_SC_VPORT_ZMIN_15
+ 0x3f800000, // PA_SC_VPORT_ZMAX_15
+};
+static const u32 si_SECT_CONTEXT_def_2[] =
+{
+ 0x00000000, // CP_PERFMON_CNTX_CNTL
+ 0x00000000, // CP_RINGID
+ 0x00000000, // CP_VMID
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0xffffffff, // VGT_MAX_VTX_INDX
+ 0x00000000, // VGT_MIN_VTX_INDX
+ 0x00000000, // VGT_INDX_OFFSET
+ 0x00000000, // VGT_MULTI_PRIM_IB_RESET_INDX
+ 0, // HOLE
+ 0x00000000, // CB_BLEND_RED
+ 0x00000000, // CB_BLEND_GREEN
+ 0x00000000, // CB_BLEND_BLUE
+ 0x00000000, // CB_BLEND_ALPHA
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // DB_STENCIL_CONTROL
+ 0x00000000, // DB_STENCILREFMASK
+ 0x00000000, // DB_STENCILREFMASK_BF
+ 0, // HOLE
+ 0x00000000, // PA_CL_VPORT_XSCALE
+ 0x00000000, // PA_CL_VPORT_XOFFSET
+ 0x00000000, // PA_CL_VPORT_YSCALE
+ 0x00000000, // PA_CL_VPORT_YOFFSET
+ 0x00000000, // PA_CL_VPORT_ZSCALE
+ 0x00000000, // PA_CL_VPORT_ZOFFSET
+ 0x00000000, // PA_CL_VPORT_XSCALE_1
+ 0x00000000, // PA_CL_VPORT_XOFFSET_1
+ 0x00000000, // PA_CL_VPORT_YSCALE_1
+ 0x00000000, // PA_CL_VPORT_YOFFSET_1
+ 0x00000000, // PA_CL_VPORT_ZSCALE_1
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_1
+ 0x00000000, // PA_CL_VPORT_XSCALE_2
+ 0x00000000, // PA_CL_VPORT_XOFFSET_2
+ 0x00000000, // PA_CL_VPORT_YSCALE_2
+ 0x00000000, // PA_CL_VPORT_YOFFSET_2
+ 0x00000000, // PA_CL_VPORT_ZSCALE_2
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_2
+ 0x00000000, // PA_CL_VPORT_XSCALE_3
+ 0x00000000, // PA_CL_VPORT_XOFFSET_3
+ 0x00000000, // PA_CL_VPORT_YSCALE_3
+ 0x00000000, // PA_CL_VPORT_YOFFSET_3
+ 0x00000000, // PA_CL_VPORT_ZSCALE_3
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_3
+ 0x00000000, // PA_CL_VPORT_XSCALE_4
+ 0x00000000, // PA_CL_VPORT_XOFFSET_4
+ 0x00000000, // PA_CL_VPORT_YSCALE_4
+ 0x00000000, // PA_CL_VPORT_YOFFSET_4
+ 0x00000000, // PA_CL_VPORT_ZSCALE_4
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_4
+ 0x00000000, // PA_CL_VPORT_XSCALE_5
+ 0x00000000, // PA_CL_VPORT_XOFFSET_5
+ 0x00000000, // PA_CL_VPORT_YSCALE_5
+ 0x00000000, // PA_CL_VPORT_YOFFSET_5
+ 0x00000000, // PA_CL_VPORT_ZSCALE_5
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_5
+ 0x00000000, // PA_CL_VPORT_XSCALE_6
+ 0x00000000, // PA_CL_VPORT_XOFFSET_6
+ 0x00000000, // PA_CL_VPORT_YSCALE_6
+ 0x00000000, // PA_CL_VPORT_YOFFSET_6
+ 0x00000000, // PA_CL_VPORT_ZSCALE_6
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_6
+ 0x00000000, // PA_CL_VPORT_XSCALE_7
+ 0x00000000, // PA_CL_VPORT_XOFFSET_7
+ 0x00000000, // PA_CL_VPORT_YSCALE_7
+ 0x00000000, // PA_CL_VPORT_YOFFSET_7
+ 0x00000000, // PA_CL_VPORT_ZSCALE_7
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_7
+ 0x00000000, // PA_CL_VPORT_XSCALE_8
+ 0x00000000, // PA_CL_VPORT_XOFFSET_8
+ 0x00000000, // PA_CL_VPORT_YSCALE_8
+ 0x00000000, // PA_CL_VPORT_YOFFSET_8
+ 0x00000000, // PA_CL_VPORT_ZSCALE_8
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_8
+ 0x00000000, // PA_CL_VPORT_XSCALE_9
+ 0x00000000, // PA_CL_VPORT_XOFFSET_9
+ 0x00000000, // PA_CL_VPORT_YSCALE_9
+ 0x00000000, // PA_CL_VPORT_YOFFSET_9
+ 0x00000000, // PA_CL_VPORT_ZSCALE_9
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_9
+ 0x00000000, // PA_CL_VPORT_XSCALE_10
+ 0x00000000, // PA_CL_VPORT_XOFFSET_10
+ 0x00000000, // PA_CL_VPORT_YSCALE_10
+ 0x00000000, // PA_CL_VPORT_YOFFSET_10
+ 0x00000000, // PA_CL_VPORT_ZSCALE_10
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_10
+ 0x00000000, // PA_CL_VPORT_XSCALE_11
+ 0x00000000, // PA_CL_VPORT_XOFFSET_11
+ 0x00000000, // PA_CL_VPORT_YSCALE_11
+ 0x00000000, // PA_CL_VPORT_YOFFSET_11
+ 0x00000000, // PA_CL_VPORT_ZSCALE_11
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_11
+ 0x00000000, // PA_CL_VPORT_XSCALE_12
+ 0x00000000, // PA_CL_VPORT_XOFFSET_12
+ 0x00000000, // PA_CL_VPORT_YSCALE_12
+ 0x00000000, // PA_CL_VPORT_YOFFSET_12
+ 0x00000000, // PA_CL_VPORT_ZSCALE_12
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_12
+ 0x00000000, // PA_CL_VPORT_XSCALE_13
+ 0x00000000, // PA_CL_VPORT_XOFFSET_13
+ 0x00000000, // PA_CL_VPORT_YSCALE_13
+ 0x00000000, // PA_CL_VPORT_YOFFSET_13
+ 0x00000000, // PA_CL_VPORT_ZSCALE_13
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_13
+ 0x00000000, // PA_CL_VPORT_XSCALE_14
+ 0x00000000, // PA_CL_VPORT_XOFFSET_14
+ 0x00000000, // PA_CL_VPORT_YSCALE_14
+ 0x00000000, // PA_CL_VPORT_YOFFSET_14
+ 0x00000000, // PA_CL_VPORT_ZSCALE_14
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_14
+ 0x00000000, // PA_CL_VPORT_XSCALE_15
+ 0x00000000, // PA_CL_VPORT_XOFFSET_15
+ 0x00000000, // PA_CL_VPORT_YSCALE_15
+ 0x00000000, // PA_CL_VPORT_YOFFSET_15
+ 0x00000000, // PA_CL_VPORT_ZSCALE_15
+ 0x00000000, // PA_CL_VPORT_ZOFFSET_15
+ 0x00000000, // PA_CL_UCP_0_X
+ 0x00000000, // PA_CL_UCP_0_Y
+ 0x00000000, // PA_CL_UCP_0_Z
+ 0x00000000, // PA_CL_UCP_0_W
+ 0x00000000, // PA_CL_UCP_1_X
+ 0x00000000, // PA_CL_UCP_1_Y
+ 0x00000000, // PA_CL_UCP_1_Z
+ 0x00000000, // PA_CL_UCP_1_W
+ 0x00000000, // PA_CL_UCP_2_X
+ 0x00000000, // PA_CL_UCP_2_Y
+ 0x00000000, // PA_CL_UCP_2_Z
+ 0x00000000, // PA_CL_UCP_2_W
+ 0x00000000, // PA_CL_UCP_3_X
+ 0x00000000, // PA_CL_UCP_3_Y
+ 0x00000000, // PA_CL_UCP_3_Z
+ 0x00000000, // PA_CL_UCP_3_W
+ 0x00000000, // PA_CL_UCP_4_X
+ 0x00000000, // PA_CL_UCP_4_Y
+ 0x00000000, // PA_CL_UCP_4_Z
+ 0x00000000, // PA_CL_UCP_4_W
+ 0x00000000, // PA_CL_UCP_5_X
+ 0x00000000, // PA_CL_UCP_5_Y
+ 0x00000000, // PA_CL_UCP_5_Z
+ 0x00000000, // PA_CL_UCP_5_W
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // SPI_PS_INPUT_CNTL_0
+ 0x00000000, // SPI_PS_INPUT_CNTL_1
+ 0x00000000, // SPI_PS_INPUT_CNTL_2
+ 0x00000000, // SPI_PS_INPUT_CNTL_3
+ 0x00000000, // SPI_PS_INPUT_CNTL_4
+ 0x00000000, // SPI_PS_INPUT_CNTL_5
+ 0x00000000, // SPI_PS_INPUT_CNTL_6
+ 0x00000000, // SPI_PS_INPUT_CNTL_7
+ 0x00000000, // SPI_PS_INPUT_CNTL_8
+ 0x00000000, // SPI_PS_INPUT_CNTL_9
+ 0x00000000, // SPI_PS_INPUT_CNTL_10
+ 0x00000000, // SPI_PS_INPUT_CNTL_11
+ 0x00000000, // SPI_PS_INPUT_CNTL_12
+ 0x00000000, // SPI_PS_INPUT_CNTL_13
+ 0x00000000, // SPI_PS_INPUT_CNTL_14
+ 0x00000000, // SPI_PS_INPUT_CNTL_15
+ 0x00000000, // SPI_PS_INPUT_CNTL_16
+ 0x00000000, // SPI_PS_INPUT_CNTL_17
+ 0x00000000, // SPI_PS_INPUT_CNTL_18
+ 0x00000000, // SPI_PS_INPUT_CNTL_19
+ 0x00000000, // SPI_PS_INPUT_CNTL_20
+ 0x00000000, // SPI_PS_INPUT_CNTL_21
+ 0x00000000, // SPI_PS_INPUT_CNTL_22
+ 0x00000000, // SPI_PS_INPUT_CNTL_23
+ 0x00000000, // SPI_PS_INPUT_CNTL_24
+ 0x00000000, // SPI_PS_INPUT_CNTL_25
+ 0x00000000, // SPI_PS_INPUT_CNTL_26
+ 0x00000000, // SPI_PS_INPUT_CNTL_27
+ 0x00000000, // SPI_PS_INPUT_CNTL_28
+ 0x00000000, // SPI_PS_INPUT_CNTL_29
+ 0x00000000, // SPI_PS_INPUT_CNTL_30
+ 0x00000000, // SPI_PS_INPUT_CNTL_31
+ 0x00000000, // SPI_VS_OUT_CONFIG
+ 0, // HOLE
+ 0x00000000, // SPI_PS_INPUT_ENA
+ 0x00000000, // SPI_PS_INPUT_ADDR
+ 0x00000000, // SPI_INTERP_CONTROL_0
+ 0x00000002, // SPI_PS_IN_CONTROL
+ 0, // HOLE
+ 0x00000000, // SPI_BARYC_CNTL
+ 0, // HOLE
+ 0x00000000, // SPI_TMPRING_SIZE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // SPI_WAVE_MGMT_1
+ 0x00000000, // SPI_WAVE_MGMT_2
+ 0x00000000, // SPI_SHADER_POS_FORMAT
+ 0x00000000, // SPI_SHADER_Z_FORMAT
+ 0x00000000, // SPI_SHADER_COL_FORMAT
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // CB_BLEND0_CONTROL
+ 0x00000000, // CB_BLEND1_CONTROL
+ 0x00000000, // CB_BLEND2_CONTROL
+ 0x00000000, // CB_BLEND3_CONTROL
+ 0x00000000, // CB_BLEND4_CONTROL
+ 0x00000000, // CB_BLEND5_CONTROL
+ 0x00000000, // CB_BLEND6_CONTROL
+ 0x00000000, // CB_BLEND7_CONTROL
+};
+static const u32 si_SECT_CONTEXT_def_3[] =
+{
+ 0x00000000, // PA_CL_POINT_X_RAD
+ 0x00000000, // PA_CL_POINT_Y_RAD
+ 0x00000000, // PA_CL_POINT_SIZE
+ 0x00000000, // PA_CL_POINT_CULL_RAD
+ 0x00000000, // VGT_DMA_BASE_HI
+ 0x00000000, // VGT_DMA_BASE
+};
+static const u32 si_SECT_CONTEXT_def_4[] =
+{
+ 0x00000000, // DB_DEPTH_CONTROL
+ 0x00000000, // DB_EQAA
+ 0x00000000, // CB_COLOR_CONTROL
+ 0x00000000, // DB_SHADER_CONTROL
+ 0x00090000, // PA_CL_CLIP_CNTL
+ 0x00000004, // PA_SU_SC_MODE_CNTL
+ 0x00000000, // PA_CL_VTE_CNTL
+ 0x00000000, // PA_CL_VS_OUT_CNTL
+ 0x00000000, // PA_CL_NANINF_CNTL
+ 0x00000000, // PA_SU_LINE_STIPPLE_CNTL
+ 0x00000000, // PA_SU_LINE_STIPPLE_SCALE
+ 0x00000000, // PA_SU_PRIM_FILTER_CNTL
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // PA_SU_POINT_SIZE
+ 0x00000000, // PA_SU_POINT_MINMAX
+ 0x00000000, // PA_SU_LINE_CNTL
+ 0x00000000, // PA_SC_LINE_STIPPLE
+ 0x00000000, // VGT_OUTPUT_PATH_CNTL
+ 0x00000000, // VGT_HOS_CNTL
+ 0x00000000, // VGT_HOS_MAX_TESS_LEVEL
+ 0x00000000, // VGT_HOS_MIN_TESS_LEVEL
+ 0x00000000, // VGT_HOS_REUSE_DEPTH
+ 0x00000000, // VGT_GROUP_PRIM_TYPE
+ 0x00000000, // VGT_GROUP_FIRST_DECR
+ 0x00000000, // VGT_GROUP_DECR
+ 0x00000000, // VGT_GROUP_VECT_0_CNTL
+ 0x00000000, // VGT_GROUP_VECT_1_CNTL
+ 0x00000000, // VGT_GROUP_VECT_0_FMT_CNTL
+ 0x00000000, // VGT_GROUP_VECT_1_FMT_CNTL
+ 0x00000000, // VGT_GS_MODE
+ 0, // HOLE
+ 0x00000000, // PA_SC_MODE_CNTL_0
+ 0x00000000, // PA_SC_MODE_CNTL_1
+ 0x00000000, // VGT_ENHANCE
+ 0x00000100, // VGT_GS_PER_ES
+ 0x00000080, // VGT_ES_PER_GS
+ 0x00000002, // VGT_GS_PER_VS
+ 0x00000000, // VGT_GSVS_RING_OFFSET_1
+ 0x00000000, // VGT_GSVS_RING_OFFSET_2
+ 0x00000000, // VGT_GSVS_RING_OFFSET_3
+ 0x00000000, // VGT_GS_OUT_PRIM_TYPE
+ 0x00000000, // IA_ENHANCE
+};
+static const u32 si_SECT_CONTEXT_def_5[] =
+{
+ 0x00000000, // VGT_PRIMITIVEID_EN
+};
+static const u32 si_SECT_CONTEXT_def_6[] =
+{
+ 0x00000000, // VGT_PRIMITIVEID_RESET
+};
+static const u32 si_SECT_CONTEXT_def_7[] =
+{
+ 0x00000000, // VGT_MULTI_PRIM_IB_RESET_EN
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // VGT_INSTANCE_STEP_RATE_0
+ 0x00000000, // VGT_INSTANCE_STEP_RATE_1
+ 0x000000ff, // IA_MULTI_VGT_PARAM
+ 0x00000000, // VGT_ESGS_RING_ITEMSIZE
+ 0x00000000, // VGT_GSVS_RING_ITEMSIZE
+ 0x00000000, // VGT_REUSE_OFF
+ 0x00000000, // VGT_VTX_CNT_EN
+ 0x00000000, // DB_HTILE_SURFACE
+ 0x00000000, // DB_SRESULTS_COMPARE_STATE0
+ 0x00000000, // DB_SRESULTS_COMPARE_STATE1
+ 0x00000000, // DB_PRELOAD_CONTROL
+ 0, // HOLE
+ 0x00000000, // VGT_STRMOUT_BUFFER_SIZE_0
+ 0x00000000, // VGT_STRMOUT_VTX_STRIDE_0
+ 0, // HOLE
+ 0x00000000, // VGT_STRMOUT_BUFFER_OFFSET_0
+ 0x00000000, // VGT_STRMOUT_BUFFER_SIZE_1
+ 0x00000000, // VGT_STRMOUT_VTX_STRIDE_1
+ 0, // HOLE
+ 0x00000000, // VGT_STRMOUT_BUFFER_OFFSET_1
+ 0x00000000, // VGT_STRMOUT_BUFFER_SIZE_2
+ 0x00000000, // VGT_STRMOUT_VTX_STRIDE_2
+ 0, // HOLE
+ 0x00000000, // VGT_STRMOUT_BUFFER_OFFSET_2
+ 0x00000000, // VGT_STRMOUT_BUFFER_SIZE_3
+ 0x00000000, // VGT_STRMOUT_VTX_STRIDE_3
+ 0, // HOLE
+ 0x00000000, // VGT_STRMOUT_BUFFER_OFFSET_3
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // VGT_STRMOUT_DRAW_OPAQUE_OFFSET
+ 0x00000000, // VGT_STRMOUT_DRAW_OPAQUE_BUFFER_FILLED_SIZE
+ 0x00000000, // VGT_STRMOUT_DRAW_OPAQUE_VERTEX_STRIDE
+ 0, // HOLE
+ 0x00000000, // VGT_GS_MAX_VERT_OUT
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // VGT_SHADER_STAGES_EN
+ 0x00000000, // VGT_LS_HS_CONFIG
+ 0x00000000, // VGT_GS_VERT_ITEMSIZE
+ 0x00000000, // VGT_GS_VERT_ITEMSIZE_1
+ 0x00000000, // VGT_GS_VERT_ITEMSIZE_2
+ 0x00000000, // VGT_GS_VERT_ITEMSIZE_3
+ 0x00000000, // VGT_TF_PARAM
+ 0x00000000, // DB_ALPHA_TO_MASK
+ 0, // HOLE
+ 0x00000000, // PA_SU_POLY_OFFSET_DB_FMT_CNTL
+ 0x00000000, // PA_SU_POLY_OFFSET_CLAMP
+ 0x00000000, // PA_SU_POLY_OFFSET_FRONT_SCALE
+ 0x00000000, // PA_SU_POLY_OFFSET_FRONT_OFFSET
+ 0x00000000, // PA_SU_POLY_OFFSET_BACK_SCALE
+ 0x00000000, // PA_SU_POLY_OFFSET_BACK_OFFSET
+ 0x00000000, // VGT_GS_INSTANCE_CNT
+ 0x00000000, // VGT_STRMOUT_CONFIG
+ 0x00000000, // VGT_STRMOUT_BUFFER_CONFIG
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // PA_SC_CENTROID_PRIORITY_0
+ 0x00000000, // PA_SC_CENTROID_PRIORITY_1
+ 0x00001000, // PA_SC_LINE_CNTL
+ 0x00000000, // PA_SC_AA_CONFIG
+ 0x00000005, // PA_SU_VTX_CNTL
+ 0x3f800000, // PA_CL_GB_VERT_CLIP_ADJ
+ 0x3f800000, // PA_CL_GB_VERT_DISC_ADJ
+ 0x3f800000, // PA_CL_GB_HORZ_CLIP_ADJ
+ 0x3f800000, // PA_CL_GB_HORZ_DISC_ADJ
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_0
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_1
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_2
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y0_3
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_0
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_1
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_2
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y0_3
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_0
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_1
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_2
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X0Y1_3
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_0
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_1
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_2
+ 0x00000000, // PA_SC_AA_SAMPLE_LOCS_PIXEL_X1Y1_3
+ 0xffffffff, // PA_SC_AA_MASK_X0Y0_X1Y0
+ 0xffffffff, // PA_SC_AA_MASK_X0Y1_X1Y1
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0, // HOLE
+ 0x0000000e, // VGT_VERTEX_REUSE_BLOCK_CNTL
+ 0x00000010, // VGT_OUT_DEALLOC_CNTL
+ 0x00000000, // CB_COLOR0_BASE
+ 0x00000000, // CB_COLOR0_PITCH
+ 0x00000000, // CB_COLOR0_SLICE
+ 0x00000000, // CB_COLOR0_VIEW
+ 0x00000000, // CB_COLOR0_INFO
+ 0x00000000, // CB_COLOR0_ATTRIB
+ 0, // HOLE
+ 0x00000000, // CB_COLOR0_CMASK
+ 0x00000000, // CB_COLOR0_CMASK_SLICE
+ 0x00000000, // CB_COLOR0_FMASK
+ 0x00000000, // CB_COLOR0_FMASK_SLICE
+ 0x00000000, // CB_COLOR0_CLEAR_WORD0
+ 0x00000000, // CB_COLOR0_CLEAR_WORD1
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // CB_COLOR1_BASE
+ 0x00000000, // CB_COLOR1_PITCH
+ 0x00000000, // CB_COLOR1_SLICE
+ 0x00000000, // CB_COLOR1_VIEW
+ 0x00000000, // CB_COLOR1_INFO
+ 0x00000000, // CB_COLOR1_ATTRIB
+ 0, // HOLE
+ 0x00000000, // CB_COLOR1_CMASK
+ 0x00000000, // CB_COLOR1_CMASK_SLICE
+ 0x00000000, // CB_COLOR1_FMASK
+ 0x00000000, // CB_COLOR1_FMASK_SLICE
+ 0x00000000, // CB_COLOR1_CLEAR_WORD0
+ 0x00000000, // CB_COLOR1_CLEAR_WORD1
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // CB_COLOR2_BASE
+ 0x00000000, // CB_COLOR2_PITCH
+ 0x00000000, // CB_COLOR2_SLICE
+ 0x00000000, // CB_COLOR2_VIEW
+ 0x00000000, // CB_COLOR2_INFO
+ 0x00000000, // CB_COLOR2_ATTRIB
+ 0, // HOLE
+ 0x00000000, // CB_COLOR2_CMASK
+ 0x00000000, // CB_COLOR2_CMASK_SLICE
+ 0x00000000, // CB_COLOR2_FMASK
+ 0x00000000, // CB_COLOR2_FMASK_SLICE
+ 0x00000000, // CB_COLOR2_CLEAR_WORD0
+ 0x00000000, // CB_COLOR2_CLEAR_WORD1
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // CB_COLOR3_BASE
+ 0x00000000, // CB_COLOR3_PITCH
+ 0x00000000, // CB_COLOR3_SLICE
+ 0x00000000, // CB_COLOR3_VIEW
+ 0x00000000, // CB_COLOR3_INFO
+ 0x00000000, // CB_COLOR3_ATTRIB
+ 0, // HOLE
+ 0x00000000, // CB_COLOR3_CMASK
+ 0x00000000, // CB_COLOR3_CMASK_SLICE
+ 0x00000000, // CB_COLOR3_FMASK
+ 0x00000000, // CB_COLOR3_FMASK_SLICE
+ 0x00000000, // CB_COLOR3_CLEAR_WORD0
+ 0x00000000, // CB_COLOR3_CLEAR_WORD1
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // CB_COLOR4_BASE
+ 0x00000000, // CB_COLOR4_PITCH
+ 0x00000000, // CB_COLOR4_SLICE
+ 0x00000000, // CB_COLOR4_VIEW
+ 0x00000000, // CB_COLOR4_INFO
+ 0x00000000, // CB_COLOR4_ATTRIB
+ 0, // HOLE
+ 0x00000000, // CB_COLOR4_CMASK
+ 0x00000000, // CB_COLOR4_CMASK_SLICE
+ 0x00000000, // CB_COLOR4_FMASK
+ 0x00000000, // CB_COLOR4_FMASK_SLICE
+ 0x00000000, // CB_COLOR4_CLEAR_WORD0
+ 0x00000000, // CB_COLOR4_CLEAR_WORD1
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // CB_COLOR5_BASE
+ 0x00000000, // CB_COLOR5_PITCH
+ 0x00000000, // CB_COLOR5_SLICE
+ 0x00000000, // CB_COLOR5_VIEW
+ 0x00000000, // CB_COLOR5_INFO
+ 0x00000000, // CB_COLOR5_ATTRIB
+ 0, // HOLE
+ 0x00000000, // CB_COLOR5_CMASK
+ 0x00000000, // CB_COLOR5_CMASK_SLICE
+ 0x00000000, // CB_COLOR5_FMASK
+ 0x00000000, // CB_COLOR5_FMASK_SLICE
+ 0x00000000, // CB_COLOR5_CLEAR_WORD0
+ 0x00000000, // CB_COLOR5_CLEAR_WORD1
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // CB_COLOR6_BASE
+ 0x00000000, // CB_COLOR6_PITCH
+ 0x00000000, // CB_COLOR6_SLICE
+ 0x00000000, // CB_COLOR6_VIEW
+ 0x00000000, // CB_COLOR6_INFO
+ 0x00000000, // CB_COLOR6_ATTRIB
+ 0, // HOLE
+ 0x00000000, // CB_COLOR6_CMASK
+ 0x00000000, // CB_COLOR6_CMASK_SLICE
+ 0x00000000, // CB_COLOR6_FMASK
+ 0x00000000, // CB_COLOR6_FMASK_SLICE
+ 0x00000000, // CB_COLOR6_CLEAR_WORD0
+ 0x00000000, // CB_COLOR6_CLEAR_WORD1
+ 0, // HOLE
+ 0, // HOLE
+ 0x00000000, // CB_COLOR7_BASE
+ 0x00000000, // CB_COLOR7_PITCH
+ 0x00000000, // CB_COLOR7_SLICE
+ 0x00000000, // CB_COLOR7_VIEW
+ 0x00000000, // CB_COLOR7_INFO
+ 0x00000000, // CB_COLOR7_ATTRIB
+ 0, // HOLE
+ 0x00000000, // CB_COLOR7_CMASK
+ 0x00000000, // CB_COLOR7_CMASK_SLICE
+ 0x00000000, // CB_COLOR7_FMASK
+ 0x00000000, // CB_COLOR7_FMASK_SLICE
+ 0x00000000, // CB_COLOR7_CLEAR_WORD0
+ 0x00000000, // CB_COLOR7_CLEAR_WORD1
+};
+static const struct cs_extent_def si_SECT_CONTEXT_defs[] =
+{
+ {si_SECT_CONTEXT_def_1, 0x0000a000, 212 },
+ {si_SECT_CONTEXT_def_2, 0x0000a0d8, 272 },
+ {si_SECT_CONTEXT_def_3, 0x0000a1f5, 6 },
+ {si_SECT_CONTEXT_def_4, 0x0000a200, 157 },
+ {si_SECT_CONTEXT_def_5, 0x0000a2a1, 1 },
+ {si_SECT_CONTEXT_def_6, 0x0000a2a3, 1 },
+ {si_SECT_CONTEXT_def_7, 0x0000a2a5, 233 },
+ { NULL, 0, 0 }
+};
+static const struct cs_section_def si_cs_data[] = {
+ { si_SECT_CONTEXT_defs, SECT_CONTEXT },
+ { NULL, SECT_NONE }
+};
--- /dev/null
+/*
+ * Copyright 2010 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Alex Deucher
+ */
+#ifndef __SI_REG_H__
+#define __SI_REG_H__
+
+/* SI */
+#define SI_DC_GPIO_HPD_MASK 0x196c
+#define SI_DC_GPIO_HPD_A 0x196d
+#define SI_DC_GPIO_HPD_EN 0x196e
+#define SI_DC_GPIO_HPD_Y 0x196f
+
+#define SI_GRPH_CONTROL 0x1a01
+# define SI_GRPH_DEPTH(x) (((x) & 0x3) << 0)
+# define SI_GRPH_DEPTH_8BPP 0
+# define SI_GRPH_DEPTH_16BPP 1
+# define SI_GRPH_DEPTH_32BPP 2
+# define SI_GRPH_NUM_BANKS(x) (((x) & 0x3) << 2)
+# define SI_ADDR_SURF_2_BANK 0
+# define SI_ADDR_SURF_4_BANK 1
+# define SI_ADDR_SURF_8_BANK 2
+# define SI_ADDR_SURF_16_BANK 3
+# define SI_GRPH_Z(x) (((x) & 0x3) << 4)
+# define SI_GRPH_BANK_WIDTH(x) (((x) & 0x3) << 6)
+# define SI_ADDR_SURF_BANK_WIDTH_1 0
+# define SI_ADDR_SURF_BANK_WIDTH_2 1
+# define SI_ADDR_SURF_BANK_WIDTH_4 2
+# define SI_ADDR_SURF_BANK_WIDTH_8 3
+# define SI_GRPH_FORMAT(x) (((x) & 0x7) << 8)
+/* 8 BPP */
+# define SI_GRPH_FORMAT_INDEXED 0
+/* 16 BPP */
+# define SI_GRPH_FORMAT_ARGB1555 0
+# define SI_GRPH_FORMAT_ARGB565 1
+# define SI_GRPH_FORMAT_ARGB4444 2
+# define SI_GRPH_FORMAT_AI88 3
+# define SI_GRPH_FORMAT_MONO16 4
+# define SI_GRPH_FORMAT_BGRA5551 5
+/* 32 BPP */
+# define SI_GRPH_FORMAT_ARGB8888 0
+# define SI_GRPH_FORMAT_ARGB2101010 1
+# define SI_GRPH_FORMAT_32BPP_DIG 2
+# define SI_GRPH_FORMAT_8B_ARGB2101010 3
+# define SI_GRPH_FORMAT_BGRA1010102 4
+# define SI_GRPH_FORMAT_8B_BGRA1010102 5
+# define SI_GRPH_FORMAT_RGB111110 6
+# define SI_GRPH_FORMAT_BGR101111 7
+# define SI_GRPH_BANK_HEIGHT(x) (((x) & 0x3) << 11)
+# define SI_ADDR_SURF_BANK_HEIGHT_1 0
+# define SI_ADDR_SURF_BANK_HEIGHT_2 1
+# define SI_ADDR_SURF_BANK_HEIGHT_4 2
+# define SI_ADDR_SURF_BANK_HEIGHT_8 3
+# define SI_GRPH_TILE_SPLIT(x) (((x) & 0x7) << 13)
+# define SI_ADDR_SURF_TILE_SPLIT_64B 0
+# define SI_ADDR_SURF_TILE_SPLIT_128B 1
+# define SI_ADDR_SURF_TILE_SPLIT_256B 2
+# define SI_ADDR_SURF_TILE_SPLIT_512B 3
+# define SI_ADDR_SURF_TILE_SPLIT_1KB 4
+# define SI_ADDR_SURF_TILE_SPLIT_2KB 5
+# define SI_ADDR_SURF_TILE_SPLIT_4KB 6
+# define SI_GRPH_MACRO_TILE_ASPECT(x) (((x) & 0x3) << 18)
+# define SI_ADDR_SURF_MACRO_TILE_ASPECT_1 0
+# define SI_ADDR_SURF_MACRO_TILE_ASPECT_2 1
+# define SI_ADDR_SURF_MACRO_TILE_ASPECT_4 2
+# define SI_ADDR_SURF_MACRO_TILE_ASPECT_8 3
+# define SI_GRPH_ARRAY_MODE(x) (((x) & 0x7) << 20)
+# define SI_GRPH_ARRAY_LINEAR_GENERAL 0
+# define SI_GRPH_ARRAY_LINEAR_ALIGNED 1
+# define SI_GRPH_ARRAY_1D_TILED_THIN1 2
+# define SI_GRPH_ARRAY_2D_TILED_THIN1 4
+# define SI_GRPH_PIPE_CONFIG(x) (((x) & 0x1f) << 24)
+# define SI_ADDR_SURF_P2 0
+# define SI_ADDR_SURF_P4_8x16 4
+# define SI_ADDR_SURF_P4_16x16 5
+# define SI_ADDR_SURF_P4_16x32 6
+# define SI_ADDR_SURF_P4_32x32 7
+# define SI_ADDR_SURF_P8_16x16_8x16 8
+# define SI_ADDR_SURF_P8_16x32_8x16 9
+# define SI_ADDR_SURF_P8_32x32_8x16 10
+# define SI_ADDR_SURF_P8_16x32_16x16 11
+# define SI_ADDR_SURF_P8_32x32_16x16 12
+# define SI_ADDR_SURF_P8_32x32_16x32 13
+# define SI_ADDR_SURF_P8_32x64_32x32 14
+
+#endif
--- /dev/null
+/*
+ * Copyright 2011 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Alex Deucher
+ */
+#ifndef SI_H
+#define SI_H
+
+#define TAHITI_RB_BITMAP_WIDTH_PER_SH 2
+
+#define TAHITI_GB_ADDR_CONFIG_GOLDEN 0x12011003
+#define VERDE_GB_ADDR_CONFIG_GOLDEN 0x12010002
+#define HAINAN_GB_ADDR_CONFIG_GOLDEN 0x02010001
+
+#define SI_MAX_SH_GPRS 256
+#define SI_MAX_TEMP_GPRS 16
+#define SI_MAX_SH_THREADS 256
+#define SI_MAX_SH_STACK_ENTRIES 4096
+#define SI_MAX_FRC_EOV_CNT 16384
+#define SI_MAX_BACKENDS 8
+#define SI_MAX_BACKENDS_MASK 0xFF
+#define SI_MAX_BACKENDS_PER_SE_MASK 0x0F
+#define SI_MAX_SIMDS 12
+#define SI_MAX_SIMDS_MASK 0x0FFF
+#define SI_MAX_SIMDS_PER_SE_MASK 0x00FF
+#define SI_MAX_PIPES 8
+#define SI_MAX_PIPES_MASK 0xFF
+#define SI_MAX_PIPES_PER_SIMD_MASK 0x3F
+#define SI_MAX_LDS_NUM 0xFFFF
+#define SI_MAX_TCC 16
+#define SI_MAX_TCC_MASK 0xFFFF
+
+#define AMDGPU_NUM_OF_VMIDS 8
+
+/* SMC IND accessor regs */
+#define SMC_IND_INDEX_0 0x80
+#define SMC_IND_DATA_0 0x81
+
+#define SMC_IND_ACCESS_CNTL 0x8A
+# define AUTO_INCREMENT_IND_0 (1 << 0)
+#define SMC_MESSAGE_0 0x8B
+#define SMC_RESP_0 0x8C
+
+/* CG IND registers are accessed via SMC indirect space + SMC_CG_IND_START */
+#define SMC_CG_IND_START 0xc0030000
+#define SMC_CG_IND_END 0xc0040000
+
+#define CG_CGTT_LOCAL_0 0x400
+#define CG_CGTT_LOCAL_1 0x401
+
+/* SMC IND registers */
+#define SMC_SYSCON_RESET_CNTL 0x80000000
+# define RST_REG (1 << 0)
+#define SMC_SYSCON_CLOCK_CNTL_0 0x80000004
+# define CK_DISABLE (1 << 0)
+# define CKEN (1 << 24)
+
+#define VGA_HDP_CONTROL 0xCA
+#define VGA_MEMORY_DISABLE (1 << 4)
+
+#define DCCG_DISP_SLOW_SELECT_REG 0x13F
+#define DCCG_DISP1_SLOW_SELECT(x) ((x) << 0)
+#define DCCG_DISP1_SLOW_SELECT_MASK (7 << 0)
+#define DCCG_DISP1_SLOW_SELECT_SHIFT 0
+#define DCCG_DISP2_SLOW_SELECT(x) ((x) << 4)
+#define DCCG_DISP2_SLOW_SELECT_MASK (7 << 4)
+#define DCCG_DISP2_SLOW_SELECT_SHIFT 4
+
+#define CG_SPLL_FUNC_CNTL 0x180
+#define SPLL_RESET (1 << 0)
+#define SPLL_SLEEP (1 << 1)
+#define SPLL_BYPASS_EN (1 << 3)
+#define SPLL_REF_DIV(x) ((x) << 4)
+#define SPLL_REF_DIV_MASK (0x3f << 4)
+#define SPLL_PDIV_A(x) ((x) << 20)
+#define SPLL_PDIV_A_MASK (0x7f << 20)
+#define SPLL_PDIV_A_SHIFT 20
+#define CG_SPLL_FUNC_CNTL_2 0x181
+#define SCLK_MUX_SEL(x) ((x) << 0)
+#define SCLK_MUX_SEL_MASK (0x1ff << 0)
+#define SPLL_CTLREQ_CHG (1 << 23)
+#define SCLK_MUX_UPDATE (1 << 26)
+#define CG_SPLL_FUNC_CNTL_3 0x182
+#define SPLL_FB_DIV(x) ((x) << 0)
+#define SPLL_FB_DIV_MASK (0x3ffffff << 0)
+#define SPLL_FB_DIV_SHIFT 0
+#define SPLL_DITHEN (1 << 28)
+#define CG_SPLL_FUNC_CNTL_4 0x183
+
+#define SPLL_STATUS 0x185
+#define SPLL_CHG_STATUS (1 << 1)
+#define SPLL_CNTL_MODE 0x186
+#define SPLL_SW_DIR_CONTROL (1 << 0)
+# define SPLL_REFCLK_SEL(x) ((x) << 26)
+# define SPLL_REFCLK_SEL_MASK (3 << 26)
+
+#define CG_SPLL_SPREAD_SPECTRUM 0x188
+#define SSEN (1 << 0)
+#define CLK_S(x) ((x) << 4)
+#define CLK_S_MASK (0xfff << 4)
+#define CLK_S_SHIFT 4
+#define CG_SPLL_SPREAD_SPECTRUM_2 0x189
+#define CLK_V(x) ((x) << 0)
+#define CLK_V_MASK (0x3ffffff << 0)
+#define CLK_V_SHIFT 0
+
+#define CG_SPLL_AUTOSCALE_CNTL 0x18b
+# define AUTOSCALE_ON_SS_CLEAR (1 << 9)
+
+/* discrete uvd clocks */
+#define CG_UPLL_FUNC_CNTL 0x18d
+# define UPLL_RESET_MASK 0x00000001
+# define UPLL_SLEEP_MASK 0x00000002
+# define UPLL_BYPASS_EN_MASK 0x00000004
+# define UPLL_CTLREQ_MASK 0x00000008
+# define UPLL_VCO_MODE_MASK 0x00000600
+# define UPLL_REF_DIV_MASK 0x003F0000
+# define UPLL_CTLACK_MASK 0x40000000
+# define UPLL_CTLACK2_MASK 0x80000000
+#define CG_UPLL_FUNC_CNTL_2 0x18e
+# define UPLL_PDIV_A(x) ((x) << 0)
+# define UPLL_PDIV_A_MASK 0x0000007F
+# define UPLL_PDIV_B(x) ((x) << 8)
+# define UPLL_PDIV_B_MASK 0x00007F00
+# define VCLK_SRC_SEL(x) ((x) << 20)
+# define VCLK_SRC_SEL_MASK 0x01F00000
+# define DCLK_SRC_SEL(x) ((x) << 25)
+# define DCLK_SRC_SEL_MASK 0x3E000000
+#define CG_UPLL_FUNC_CNTL_3 0x18f
+# define UPLL_FB_DIV(x) ((x) << 0)
+# define UPLL_FB_DIV_MASK 0x01FFFFFF
+#define CG_UPLL_FUNC_CNTL_4 0x191
+# define UPLL_SPARE_ISPARE9 0x00020000
+#define CG_UPLL_FUNC_CNTL_5 0x192
+# define RESET_ANTI_MUX_MASK 0x00000200
+#define CG_UPLL_SPREAD_SPECTRUM 0x194
+# define SSEN_MASK 0x00000001
+
+#define MPLL_BYPASSCLK_SEL 0x197
+# define MPLL_CLKOUT_SEL(x) ((x) << 8)
+# define MPLL_CLKOUT_SEL_MASK 0xFF00
+
+#define CG_CLKPIN_CNTL 0x198
+# define XTALIN_DIVIDE (1 << 1)
+# define BCLK_AS_XCLK (1 << 2)
+#define CG_CLKPIN_CNTL_2 0x199
+# define FORCE_BIF_REFCLK_EN (1 << 3)
+# define MUX_TCLK_TO_XCLK (1 << 8)
+
+#define THM_CLK_CNTL 0x19b
+# define CMON_CLK_SEL(x) ((x) << 0)
+# define CMON_CLK_SEL_MASK 0xFF
+# define TMON_CLK_SEL(x) ((x) << 8)
+# define TMON_CLK_SEL_MASK 0xFF00
+#define MISC_CLK_CNTL 0x19c
+# define DEEP_SLEEP_CLK_SEL(x) ((x) << 0)
+# define DEEP_SLEEP_CLK_SEL_MASK 0xFF
+# define ZCLK_SEL(x) ((x) << 8)
+# define ZCLK_SEL_MASK 0xFF00
+
+#define CG_THERMAL_CTRL 0x1c0
+#define DPM_EVENT_SRC(x) ((x) << 0)
+#define DPM_EVENT_SRC_MASK (7 << 0)
+#define DIG_THERM_DPM(x) ((x) << 14)
+#define DIG_THERM_DPM_MASK 0x003FC000
+#define DIG_THERM_DPM_SHIFT 14
+#define CG_THERMAL_STATUS 0x1c1
+#define FDO_PWM_DUTY(x) ((x) << 9)
+#define FDO_PWM_DUTY_MASK (0xff << 9)
+#define FDO_PWM_DUTY_SHIFT 9
+#define CG_THERMAL_INT 0x1c2
+#define DIG_THERM_INTH(x) ((x) << 8)
+#define DIG_THERM_INTH_MASK 0x0000FF00
+#define DIG_THERM_INTH_SHIFT 8
+#define DIG_THERM_INTL(x) ((x) << 16)
+#define DIG_THERM_INTL_MASK 0x00FF0000
+#define DIG_THERM_INTL_SHIFT 16
+#define THERM_INT_MASK_HIGH (1 << 24)
+#define THERM_INT_MASK_LOW (1 << 25)
+
+#define CG_MULT_THERMAL_CTRL 0x1c4
+#define TEMP_SEL(x) ((x) << 20)
+#define TEMP_SEL_MASK (0xff << 20)
+#define TEMP_SEL_SHIFT 20
+#define CG_MULT_THERMAL_STATUS 0x1c5
+#define ASIC_MAX_TEMP(x) ((x) << 0)
+#define ASIC_MAX_TEMP_MASK 0x000001ff
+#define ASIC_MAX_TEMP_SHIFT 0
+#define CTF_TEMP(x) ((x) << 9)
+#define CTF_TEMP_MASK 0x0003fe00
+#define CTF_TEMP_SHIFT 9
+
+#define CG_FDO_CTRL0 0x1d5
+#define FDO_STATIC_DUTY(x) ((x) << 0)
+#define FDO_STATIC_DUTY_MASK 0x000000FF
+#define FDO_STATIC_DUTY_SHIFT 0
+#define CG_FDO_CTRL1 0x1d6
+#define FMAX_DUTY100(x) ((x) << 0)
+#define FMAX_DUTY100_MASK 0x000000FF
+#define FMAX_DUTY100_SHIFT 0
+#define CG_FDO_CTRL2 0x1d7
+#define TMIN(x) ((x) << 0)
+#define TMIN_MASK 0x000000FF
+#define TMIN_SHIFT 0
+#define FDO_PWM_MODE(x) ((x) << 11)
+#define FDO_PWM_MODE_MASK (7 << 11)
+#define FDO_PWM_MODE_SHIFT 11
+#define TACH_PWM_RESP_RATE(x) ((x) << 25)
+#define TACH_PWM_RESP_RATE_MASK (0x7f << 25)
+#define TACH_PWM_RESP_RATE_SHIFT 25
+
+#define CG_TACH_CTRL 0x1dc
+# define EDGE_PER_REV(x) ((x) << 0)
+# define EDGE_PER_REV_MASK (0x7 << 0)
+# define EDGE_PER_REV_SHIFT 0
+# define TARGET_PERIOD(x) ((x) << 3)
+# define TARGET_PERIOD_MASK 0xfffffff8
+# define TARGET_PERIOD_SHIFT 3
+#define CG_TACH_STATUS 0x1dd
+# define TACH_PERIOD(x) ((x) << 0)
+# define TACH_PERIOD_MASK 0xffffffff
+# define TACH_PERIOD_SHIFT 0
+
+#define GENERAL_PWRMGT 0x1e0
+# define GLOBAL_PWRMGT_EN (1 << 0)
+# define STATIC_PM_EN (1 << 1)
+# define THERMAL_PROTECTION_DIS (1 << 2)
+# define THERMAL_PROTECTION_TYPE (1 << 3)
+# define SW_SMIO_INDEX(x) ((x) << 6)
+# define SW_SMIO_INDEX_MASK (1 << 6)
+# define SW_SMIO_INDEX_SHIFT 6
+# define VOLT_PWRMGT_EN (1 << 10)
+# define DYN_SPREAD_SPECTRUM_EN (1 << 23)
+#define CG_TPC 0x1e1
+#define SCLK_PWRMGT_CNTL 0x1e2
+# define SCLK_PWRMGT_OFF (1 << 0)
+# define SCLK_LOW_D1 (1 << 1)
+# define FIR_RESET (1 << 4)
+# define FIR_FORCE_TREND_SEL (1 << 5)
+# define FIR_TREND_MODE (1 << 6)
+# define DYN_GFX_CLK_OFF_EN (1 << 7)
+# define GFX_CLK_FORCE_ON (1 << 8)
+# define GFX_CLK_REQUEST_OFF (1 << 9)
+# define GFX_CLK_FORCE_OFF (1 << 10)
+# define GFX_CLK_OFF_ACPI_D1 (1 << 11)
+# define GFX_CLK_OFF_ACPI_D2 (1 << 12)
+# define GFX_CLK_OFF_ACPI_D3 (1 << 13)
+# define DYN_LIGHT_SLEEP_EN (1 << 14)
+
+#define TARGET_AND_CURRENT_PROFILE_INDEX 0x1e6
+# define CURRENT_STATE_INDEX_MASK (0xf << 4)
+# define CURRENT_STATE_INDEX_SHIFT 4
+
+#define CG_FTV 0x1ef
+
+#define CG_FFCT_0 0x1f0
+# define UTC_0(x) ((x) << 0)
+# define UTC_0_MASK (0x3ff << 0)
+# define DTC_0(x) ((x) << 10)
+# define DTC_0_MASK (0x3ff << 10)
+
+#define CG_BSP 0x1ff
+# define BSP(x) ((x) << 0)
+# define BSP_MASK (0xffff << 0)
+# define BSU(x) ((x) << 16)
+# define BSU_MASK (0xf << 16)
+#define CG_AT 0x200
+# define CG_R(x) ((x) << 0)
+# define CG_R_MASK (0xffff << 0)
+# define CG_L(x) ((x) << 16)
+# define CG_L_MASK (0xffff << 16)
+
+#define CG_GIT 0x201
+# define CG_GICST(x) ((x) << 0)
+# define CG_GICST_MASK (0xffff << 0)
+# define CG_GIPOT(x) ((x) << 16)
+# define CG_GIPOT_MASK (0xffff << 16)
+
+#define CG_SSP 0x203
+# define SST(x) ((x) << 0)
+# define SST_MASK (0xffff << 0)
+# define SSTU(x) ((x) << 16)
+# define SSTU_MASK (0xf << 16)
+
+#define CG_DISPLAY_GAP_CNTL 0x20a
+# define DISP1_GAP(x) ((x) << 0)
+# define DISP1_GAP_MASK (3 << 0)
+# define DISP2_GAP(x) ((x) << 2)
+# define DISP2_GAP_MASK (3 << 2)
+# define VBI_TIMER_COUNT(x) ((x) << 4)
+# define VBI_TIMER_COUNT_MASK (0x3fff << 4)
+# define VBI_TIMER_UNIT(x) ((x) << 20)
+# define VBI_TIMER_UNIT_MASK (7 << 20)
+# define DISP1_GAP_MCHG(x) ((x) << 24)
+# define DISP1_GAP_MCHG_MASK (3 << 24)
+# define DISP2_GAP_MCHG(x) ((x) << 26)
+# define DISP2_GAP_MCHG_MASK (3 << 26)
+
+#define CG_ULV_CONTROL 0x21e
+#define CG_ULV_PARAMETER 0x21f
+
+#define SMC_SCRATCH0 0x221
+
+#define CG_CAC_CTRL 0x22e
+# define CAC_WINDOW(x) ((x) << 0)
+# define CAC_WINDOW_MASK 0x00ffffff
+
+#define DMIF_ADDR_CONFIG 0x2F5
+
+#define DMIF_ADDR_CALC 0x300
+
+#define PIPE0_DMIF_BUFFER_CONTROL 0x0328
+# define DMIF_BUFFERS_ALLOCATED(x) ((x) << 0)
+# define DMIF_BUFFERS_ALLOCATED_COMPLETED (1 << 4)
+
+#define SRBM_STATUS 0x394
+#define GRBM_RQ_PENDING (1 << 5)
+#define VMC_BUSY (1 << 8)
+#define MCB_BUSY (1 << 9)
+#define MCB_NON_DISPLAY_BUSY (1 << 10)
+#define MCC_BUSY (1 << 11)
+#define MCD_BUSY (1 << 12)
+#define SEM_BUSY (1 << 14)
+#define IH_BUSY (1 << 17)
+
+#define SRBM_SOFT_RESET 0x398
+#define SOFT_RESET_BIF (1 << 1)
+#define SOFT_RESET_DC (1 << 5)
+#define SOFT_RESET_DMA1 (1 << 6)
+#define SOFT_RESET_GRBM (1 << 8)
+#define SOFT_RESET_HDP (1 << 9)
+#define SOFT_RESET_IH (1 << 10)
+#define SOFT_RESET_MC (1 << 11)
+#define SOFT_RESET_ROM (1 << 14)
+#define SOFT_RESET_SEM (1 << 15)
+#define SOFT_RESET_VMC (1 << 17)
+#define SOFT_RESET_DMA (1 << 20)
+#define SOFT_RESET_TST (1 << 21)
+#define SOFT_RESET_REGBB (1 << 22)
+#define SOFT_RESET_ORB (1 << 23)
+
+#define CC_SYS_RB_BACKEND_DISABLE 0x3A0
+#define GC_USER_SYS_RB_BACKEND_DISABLE 0x3A1
+
+#define SRBM_READ_ERROR 0x3A6
+#define SRBM_INT_CNTL 0x3A8
+#define SRBM_INT_ACK 0x3AA
+
+#define SRBM_STATUS2 0x3B1
+#define DMA_BUSY (1 << 5)
+#define DMA1_BUSY (1 << 6)
+
+#define VM_L2_CNTL 0x500
+#define ENABLE_L2_CACHE (1 << 0)
+#define ENABLE_L2_FRAGMENT_PROCESSING (1 << 1)
+#define L2_CACHE_PTE_ENDIAN_SWAP_MODE(x) ((x) << 2)
+#define L2_CACHE_PDE_ENDIAN_SWAP_MODE(x) ((x) << 4)
+#define ENABLE_L2_PTE_CACHE_LRU_UPDATE_BY_WRITE (1 << 9)
+#define ENABLE_L2_PDE0_CACHE_LRU_UPDATE_BY_WRITE (1 << 10)
+#define EFFECTIVE_L2_QUEUE_SIZE(x) (((x) & 7) << 15)
+#define CONTEXT1_IDENTITY_ACCESS_MODE(x) (((x) & 3) << 19)
+#define VM_L2_CNTL2 0x501
+#define INVALIDATE_ALL_L1_TLBS (1 << 0)
+#define INVALIDATE_L2_CACHE (1 << 1)
+#define INVALIDATE_CACHE_MODE(x) ((x) << 26)
+#define INVALIDATE_PTE_AND_PDE_CACHES 0
+#define INVALIDATE_ONLY_PTE_CACHES 1
+#define INVALIDATE_ONLY_PDE_CACHES 2
+#define VM_L2_CNTL3 0x502
+#define BANK_SELECT(x) ((x) << 0)
+#define L2_CACHE_UPDATE_MODE(x) ((x) << 6)
+#define L2_CACHE_BIGK_FRAGMENT_SIZE(x) ((x) << 15)
+#define L2_CACHE_BIGK_ASSOCIATIVITY (1 << 20)
+#define VM_L2_STATUS 0x503
+#define L2_BUSY (1 << 0)
+#define VM_CONTEXT0_CNTL 0x504
+#define ENABLE_CONTEXT (1 << 0)
+#define PAGE_TABLE_DEPTH(x) (((x) & 3) << 1)
+#define RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 3)
+#define RANGE_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 4)
+#define DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 6)
+#define DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 7)
+#define PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 9)
+#define PDE0_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 10)
+#define VALID_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 12)
+#define VALID_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 13)
+#define READ_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 15)
+#define READ_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 16)
+#define WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT (1 << 18)
+#define WRITE_PROTECTION_FAULT_ENABLE_DEFAULT (1 << 19)
+#define PAGE_TABLE_BLOCK_SIZE(x) (((x) & 0xF) << 24)
+#define VM_CONTEXT1_CNTL 0x505
+#define VM_CONTEXT0_CNTL2 0x50C
+#define VM_CONTEXT1_CNTL2 0x50D
+#define VM_CONTEXT8_PAGE_TABLE_BASE_ADDR 0x50E
+#define VM_CONTEXT9_PAGE_TABLE_BASE_ADDR 0x50F
+#define VM_CONTEXT10_PAGE_TABLE_BASE_ADDR 0x510
+#define VM_CONTEXT11_PAGE_TABLE_BASE_ADDR 0x511
+#define VM_CONTEXT12_PAGE_TABLE_BASE_ADDR 0x512
+#define VM_CONTEXT13_PAGE_TABLE_BASE_ADDR 0x513
+#define VM_CONTEXT14_PAGE_TABLE_BASE_ADDR 0x514
+#define VM_CONTEXT15_PAGE_TABLE_BASE_ADDR 0x515
+
+#define VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x53f
+#define VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x537
+#define PROTECTIONS_MASK (0xf << 0)
+#define PROTECTIONS_SHIFT 0
+ /* bit 0: range
+ * bit 1: pde0
+ * bit 2: valid
+ * bit 3: read
+ * bit 4: write
+ */
+#define MEMORY_CLIENT_ID_MASK (0xff << 12)
+#define MEMORY_CLIENT_ID_SHIFT 12
+#define MEMORY_CLIENT_RW_MASK (1 << 24)
+#define MEMORY_CLIENT_RW_SHIFT 24
+#define FAULT_VMID_MASK (0xf << 25)
+#define FAULT_VMID_SHIFT 25
+
+#define VM_INVALIDATE_REQUEST 0x51E
+#define VM_INVALIDATE_RESPONSE 0x51F
+
+#define VM_CONTEXT0_PROTECTION_FAULT_DEFAULT_ADDR 0x546
+#define VM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR 0x547
+
+#define VM_CONTEXT0_PAGE_TABLE_BASE_ADDR 0x54F
+#define VM_CONTEXT1_PAGE_TABLE_BASE_ADDR 0x550
+#define VM_CONTEXT2_PAGE_TABLE_BASE_ADDR 0x551
+#define VM_CONTEXT3_PAGE_TABLE_BASE_ADDR 0x552
+#define VM_CONTEXT4_PAGE_TABLE_BASE_ADDR 0x553
+#define VM_CONTEXT5_PAGE_TABLE_BASE_ADDR 0x554
+#define VM_CONTEXT6_PAGE_TABLE_BASE_ADDR 0x555
+#define VM_CONTEXT7_PAGE_TABLE_BASE_ADDR 0x556
+#define VM_CONTEXT0_PAGE_TABLE_START_ADDR 0x557
+#define VM_CONTEXT1_PAGE_TABLE_START_ADDR 0x558
+
+#define VM_CONTEXT0_PAGE_TABLE_END_ADDR 0x55F
+#define VM_CONTEXT1_PAGE_TABLE_END_ADDR 0x560
+
+#define VM_L2_CG 0x570
+#define MC_CG_ENABLE (1 << 18)
+#define MC_LS_ENABLE (1 << 19)
+
+#define MC_SHARED_CHMAP 0x801
+#define NOOFCHAN_SHIFT 12
+#define NOOFCHAN_MASK 0x0000f000
+#define MC_SHARED_CHREMAP 0x802
+
+#define MC_VM_FB_LOCATION 0x809
+#define MC_VM_AGP_TOP 0x80A
+#define MC_VM_AGP_BOT 0x80B
+#define MC_VM_AGP_BASE 0x80C
+#define MC_VM_SYSTEM_APERTURE_LOW_ADDR 0x80D
+#define MC_VM_SYSTEM_APERTURE_HIGH_ADDR 0x80E
+#define MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR 0x80F
+
+#define MC_VM_MX_L1_TLB_CNTL 0x819
+#define ENABLE_L1_TLB (1 << 0)
+#define ENABLE_L1_FRAGMENT_PROCESSING (1 << 1)
+#define SYSTEM_ACCESS_MODE_PA_ONLY (0 << 3)
+#define SYSTEM_ACCESS_MODE_USE_SYS_MAP (1 << 3)
+#define SYSTEM_ACCESS_MODE_IN_SYS (2 << 3)
+#define SYSTEM_ACCESS_MODE_NOT_IN_SYS (3 << 3)
+#define SYSTEM_APERTURE_UNMAPPED_ACCESS_PASS_THRU (0 << 5)
+#define ENABLE_ADVANCED_DRIVER_MODEL (1 << 6)
+
+#define MC_SHARED_BLACKOUT_CNTL 0x82B
+
+#define MC_HUB_MISC_HUB_CG 0x82E
+#define MC_HUB_MISC_VM_CG 0x82F
+
+#define MC_HUB_MISC_SIP_CG 0x830
+
+#define MC_XPB_CLK_GAT 0x91E
+
+#define MC_CITF_MISC_RD_CG 0x992
+#define MC_CITF_MISC_WR_CG 0x993
+#define MC_CITF_MISC_VM_CG 0x994
+
+#define MC_ARB_RAMCFG 0x9D8
+#define NOOFBANK_SHIFT 0
+#define NOOFBANK_MASK 0x00000003
+#define NOOFRANK_SHIFT 2
+#define NOOFRANK_MASK 0x00000004
+#define NOOFROWS_SHIFT 3
+#define NOOFROWS_MASK 0x00000038
+#define NOOFCOLS_SHIFT 6
+#define NOOFCOLS_MASK 0x000000C0
+#define CHANSIZE_SHIFT 8
+#define CHANSIZE_MASK 0x00000100
+#define CHANSIZE_OVERRIDE (1 << 11)
+#define NOOFGROUPS_SHIFT 12
+#define NOOFGROUPS_MASK 0x00001000
+
+#define MC_ARB_DRAM_TIMING 0x9DD
+#define MC_ARB_DRAM_TIMING2 0x9DE
+
+#define MC_ARB_BURST_TIME 0xA02
+#define STATE0(x) ((x) << 0)
+#define STATE0_MASK (0x1f << 0)
+#define STATE0_SHIFT 0
+#define STATE1(x) ((x) << 5)
+#define STATE1_MASK (0x1f << 5)
+#define STATE1_SHIFT 5
+#define STATE2(x) ((x) << 10)
+#define STATE2_MASK (0x1f << 10)
+#define STATE2_SHIFT 10
+#define STATE3(x) ((x) << 15)
+#define STATE3_MASK (0x1f << 15)
+#define STATE3_SHIFT 15
+
+#define MC_SEQ_TRAIN_WAKEUP_CNTL 0xA3A
+#define TRAIN_DONE_D0 (1 << 30)
+#define TRAIN_DONE_D1 (1 << 31)
+
+#define MC_SEQ_SUP_CNTL 0xA32
+#define RUN_MASK (1 << 0)
+#define MC_SEQ_SUP_PGM 0xA33
+#define MC_PMG_AUTO_CMD 0xA34
+
+#define MC_IO_PAD_CNTL_D0 0xA74
+#define MEM_FALL_OUT_CMD (1 << 8)
+
+#define MC_SEQ_RAS_TIMING 0xA28
+#define MC_SEQ_CAS_TIMING 0xA29
+#define MC_SEQ_MISC_TIMING 0xA2A
+#define MC_SEQ_MISC_TIMING2 0xA2B
+#define MC_SEQ_PMG_TIMING 0xA2C
+#define MC_SEQ_RD_CTL_D0 0xA2D
+#define MC_SEQ_RD_CTL_D1 0xA2E
+#define MC_SEQ_WR_CTL_D0 0xA2F
+#define MC_SEQ_WR_CTL_D1 0xA30
+
+#define MC_SEQ_MISC0 0xA80
+#define MC_SEQ_MISC0_VEN_ID_SHIFT 8
+#define MC_SEQ_MISC0_VEN_ID_MASK 0x00000f00
+#define MC_SEQ_MISC0_VEN_ID_VALUE 3
+#define MC_SEQ_MISC0_REV_ID_SHIFT 12
+#define MC_SEQ_MISC0_REV_ID_MASK 0x0000f000
+#define MC_SEQ_MISC0_REV_ID_VALUE 1
+#define MC_SEQ_MISC0_GDDR5_SHIFT 28
+#define MC_SEQ_MISC0_GDDR5_MASK 0xf0000000
+#define MC_SEQ_MISC0_GDDR5_VALUE 5
+#define MC_SEQ_MISC1 0xA81
+#define MC_SEQ_RESERVE_M 0xA82
+#define MC_PMG_CMD_EMRS 0xA83
+
+#define MC_SEQ_IO_DEBUG_INDEX 0xA91
+#define MC_SEQ_IO_DEBUG_DATA 0xA92
+
+#define MC_SEQ_MISC5 0xA95
+#define MC_SEQ_MISC6 0xA96
+
+#define MC_SEQ_MISC7 0xA99
+
+#define MC_SEQ_RAS_TIMING_LP 0xA9B
+#define MC_SEQ_CAS_TIMING_LP 0xA9C
+#define MC_SEQ_MISC_TIMING_LP 0xA9D
+#define MC_SEQ_MISC_TIMING2_LP 0xA9E
+#define MC_SEQ_WR_CTL_D0_LP 0xA9F
+#define MC_SEQ_WR_CTL_D1_LP 0xAA0
+#define MC_SEQ_PMG_CMD_EMRS_LP 0xAA1
+#define MC_SEQ_PMG_CMD_MRS_LP 0xAA2
+
+#define MC_PMG_CMD_MRS 0xAAB
+
+#define MC_SEQ_RD_CTL_D0_LP 0xAC7
+#define MC_SEQ_RD_CTL_D1_LP 0xAC8
+
+#define MC_PMG_CMD_MRS1 0xAD1
+#define MC_SEQ_PMG_CMD_MRS1_LP 0xAD2
+#define MC_SEQ_PMG_TIMING_LP 0xAD3
+
+#define MC_SEQ_WR_CTL_2 0xAD5
+#define MC_SEQ_WR_CTL_2_LP 0xAD6
+#define MC_PMG_CMD_MRS2 0xAD7
+#define MC_SEQ_PMG_CMD_MRS2_LP 0xAD8
+
+#define MCLK_PWRMGT_CNTL 0xAE8
+# define DLL_SPEED(x) ((x) << 0)
+# define DLL_SPEED_MASK (0x1f << 0)
+# define DLL_READY (1 << 6)
+# define MC_INT_CNTL (1 << 7)
+# define MRDCK0_PDNB (1 << 8)
+# define MRDCK1_PDNB (1 << 9)
+# define MRDCK0_RESET (1 << 16)
+# define MRDCK1_RESET (1 << 17)
+# define DLL_READY_READ (1 << 24)
+#define DLL_CNTL 0xAE9
+# define MRDCK0_BYPASS (1 << 24)
+# define MRDCK1_BYPASS (1 << 25)
+
+#define MPLL_CNTL_MODE 0xAEC
+# define MPLL_MCLK_SEL (1 << 11)
+#define MPLL_FUNC_CNTL 0xAED
+#define BWCTRL(x) ((x) << 20)
+#define BWCTRL_MASK (0xff << 20)
+#define MPLL_FUNC_CNTL_1 0xAEE
+#define VCO_MODE(x) ((x) << 0)
+#define VCO_MODE_MASK (3 << 0)
+#define CLKFRAC(x) ((x) << 4)
+#define CLKFRAC_MASK (0xfff << 4)
+#define CLKF(x) ((x) << 16)
+#define CLKF_MASK (0xfff << 16)
+#define MPLL_FUNC_CNTL_2 0xAEF
+#define MPLL_AD_FUNC_CNTL 0xAF0
+#define YCLK_POST_DIV(x) ((x) << 0)
+#define YCLK_POST_DIV_MASK (7 << 0)
+#define MPLL_DQ_FUNC_CNTL 0xAF1
+#define YCLK_SEL(x) ((x) << 4)
+#define YCLK_SEL_MASK (1 << 4)
+
+#define MPLL_SS1 0xAF3
+#define CLKV(x) ((x) << 0)
+#define CLKV_MASK (0x3ffffff << 0)
+#define MPLL_SS2 0xAF4
+#define CLKS(x) ((x) << 0)
+#define CLKS_MASK (0xfff << 0)
+
+#define HDP_HOST_PATH_CNTL 0xB00
+#define CLOCK_GATING_DIS (1 << 23)
+#define HDP_NONSURFACE_BASE 0xB01
+#define HDP_NONSURFACE_INFO 0xB02
+#define HDP_NONSURFACE_SIZE 0xB03
+
+#define HDP_DEBUG0 0xBCC
+
+#define HDP_ADDR_CONFIG 0xBD2
+#define HDP_MISC_CNTL 0xBD3
+#define HDP_FLUSH_INVALIDATE_CACHE (1 << 0)
+#define HDP_MEM_POWER_LS 0xBD4
+#define HDP_LS_ENABLE (1 << 0)
+
+#define ATC_MISC_CG 0xCD4
+
+#define IH_RB_CNTL 0xF80
+# define IH_RB_ENABLE (1 << 0)
+# define IH_IB_SIZE(x) ((x) << 1) /* log2 */
+# define IH_RB_FULL_DRAIN_ENABLE (1 << 6)
+# define IH_WPTR_WRITEBACK_ENABLE (1 << 8)
+# define IH_WPTR_WRITEBACK_TIMER(x) ((x) << 9) /* log2 */
+# define IH_WPTR_OVERFLOW_ENABLE (1 << 16)
+# define IH_WPTR_OVERFLOW_CLEAR (1 << 31)
+#define IH_RB_BASE 0xF81
+#define IH_RB_RPTR 0xF82
+#define IH_RB_WPTR 0xF83
+# define RB_OVERFLOW (1 << 0)
+# define WPTR_OFFSET_MASK 0x3fffc
+#define IH_RB_WPTR_ADDR_HI 0xF84
+#define IH_RB_WPTR_ADDR_LO 0xF85
+#define IH_CNTL 0xF86
+# define ENABLE_INTR (1 << 0)
+# define IH_MC_SWAP(x) ((x) << 1)
+# define IH_MC_SWAP_NONE 0
+# define IH_MC_SWAP_16BIT 1
+# define IH_MC_SWAP_32BIT 2
+# define IH_MC_SWAP_64BIT 3
+# define RPTR_REARM (1 << 4)
+# define MC_WRREQ_CREDIT(x) ((x) << 15)
+# define MC_WR_CLEAN_CNT(x) ((x) << 20)
+# define MC_VMID(x) ((x) << 25)
+
+#define CONFIG_MEMSIZE 0x150A
+
+#define INTERRUPT_CNTL 0x151A
+# define IH_DUMMY_RD_OVERRIDE (1 << 0)
+# define IH_DUMMY_RD_EN (1 << 1)
+# define IH_REQ_NONSNOOP_EN (1 << 3)
+# define GEN_IH_INT_EN (1 << 8)
+#define INTERRUPT_CNTL2 0x151B
+
+#define HDP_MEM_COHERENCY_FLUSH_CNTL 0x1520
+
+#define BIF_FB_EN 0x1524
+#define FB_READ_EN (1 << 0)
+#define FB_WRITE_EN (1 << 1)
+
+#define HDP_REG_COHERENCY_FLUSH_CNTL 0x1528
+
+/* DCE6 ELD audio interface */
+#define AZ_F0_CODEC_ENDPOINT_INDEX 0x1780
+# define AZ_ENDPOINT_REG_INDEX(x) (((x) & 0xff) << 0)
+# define AZ_ENDPOINT_REG_WRITE_EN (1 << 8)
+#define AZ_F0_CODEC_ENDPOINT_DATA 0x1781
+
+#define AZ_F0_CODEC_PIN_CONTROL_CHANNEL_SPEAKER 0x25
+#define SPEAKER_ALLOCATION(x) (((x) & 0x7f) << 0)
+#define SPEAKER_ALLOCATION_MASK (0x7f << 0)
+#define SPEAKER_ALLOCATION_SHIFT 0
+#define HDMI_CONNECTION (1 << 16)
+#define DP_CONNECTION (1 << 17)
+
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR0 0x28 /* LPCM */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR1 0x29 /* AC3 */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR2 0x2A /* MPEG1 */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR3 0x2B /* MP3 */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR4 0x2C /* MPEG2 */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR5 0x2D /* AAC */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR6 0x2E /* DTS */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR7 0x2F /* ATRAC */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR8 0x30 /* one bit audio - leave at 0 (default) */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR9 0x31 /* Dolby Digital */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR10 0x32 /* DTS-HD */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR11 0x33 /* MAT-MLP */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR12 0x34 /* DTS */
+#define AZ_F0_CODEC_PIN_CONTROL_AUDIO_DESCRIPTOR13 0x35 /* WMA Pro */
+# define MAX_CHANNELS(x) (((x) & 0x7) << 0)
+/* max channels minus one. 7 = 8 channels */
+# define SUPPORTED_FREQUENCIES(x) (((x) & 0xff) << 8)
+# define DESCRIPTOR_BYTE_2(x) (((x) & 0xff) << 16)
+# define SUPPORTED_FREQUENCIES_STEREO(x) (((x) & 0xff) << 24) /* LPCM only */
+/* SUPPORTED_FREQUENCIES, SUPPORTED_FREQUENCIES_STEREO
+ * bit0 = 32 kHz
+ * bit1 = 44.1 kHz
+ * bit2 = 48 kHz
+ * bit3 = 88.2 kHz
+ * bit4 = 96 kHz
+ * bit5 = 176.4 kHz
+ * bit6 = 192 kHz
+ */
+
+#define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_LIPSYNC 0x37
+# define VIDEO_LIPSYNC(x) (((x) & 0xff) << 0)
+# define AUDIO_LIPSYNC(x) (((x) & 0xff) << 8)
+/* VIDEO_LIPSYNC, AUDIO_LIPSYNC
+ * 0 = invalid
+ * x = legal delay value
+ * 255 = sync not supported
+ */
+#define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_HBR 0x38
+# define HBR_CAPABLE (1 << 0) /* enabled by default */
+
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO0 0x3a
+# define MANUFACTURER_ID(x) (((x) & 0xffff) << 0)
+# define PRODUCT_ID(x) (((x) & 0xffff) << 16)
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO1 0x3b
+# define SINK_DESCRIPTION_LEN(x) (((x) & 0xff) << 0)
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO2 0x3c
+# define PORT_ID0(x) (((x) & 0xffffffff) << 0)
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO3 0x3d
+# define PORT_ID1(x) (((x) & 0xffffffff) << 0)
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO4 0x3e
+# define DESCRIPTION0(x) (((x) & 0xff) << 0)
+# define DESCRIPTION1(x) (((x) & 0xff) << 8)
+# define DESCRIPTION2(x) (((x) & 0xff) << 16)
+# define DESCRIPTION3(x) (((x) & 0xff) << 24)
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO5 0x3f
+# define DESCRIPTION4(x) (((x) & 0xff) << 0)
+# define DESCRIPTION5(x) (((x) & 0xff) << 8)
+# define DESCRIPTION6(x) (((x) & 0xff) << 16)
+# define DESCRIPTION7(x) (((x) & 0xff) << 24)
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO6 0x40
+# define DESCRIPTION8(x) (((x) & 0xff) << 0)
+# define DESCRIPTION9(x) (((x) & 0xff) << 8)
+# define DESCRIPTION10(x) (((x) & 0xff) << 16)
+# define DESCRIPTION11(x) (((x) & 0xff) << 24)
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO7 0x41
+# define DESCRIPTION12(x) (((x) & 0xff) << 0)
+# define DESCRIPTION13(x) (((x) & 0xff) << 8)
+# define DESCRIPTION14(x) (((x) & 0xff) << 16)
+# define DESCRIPTION15(x) (((x) & 0xff) << 24)
+#define AZ_F0_CODEC_PIN_CONTROL_SINK_INFO8 0x42
+# define DESCRIPTION16(x) (((x) & 0xff) << 0)
+# define DESCRIPTION17(x) (((x) & 0xff) << 8)
+
+#define AZ_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL 0x54
+# define AUDIO_ENABLED (1 << 31)
+
+#define AZ_F0_CODEC_PIN_CONTROL_RESPONSE_CONFIGURATION_DEFAULT 0x56
+#define PORT_CONNECTIVITY_MASK (3 << 30)
+#define PORT_CONNECTIVITY_SHIFT 30
+
+#define DC_LB_MEMORY_SPLIT 0x1AC3
+#define DC_LB_MEMORY_CONFIG(x) ((x) << 20)
+
+#define PRIORITY_A_CNT 0x1AC6
+#define PRIORITY_MARK_MASK 0x7fff
+#define PRIORITY_OFF (1 << 16)
+#define PRIORITY_ALWAYS_ON (1 << 20)
+#define PRIORITY_B_CNT 0x1AC7
+
+#define DPG_PIPE_ARBITRATION_CONTROL3 0x1B32
+# define LATENCY_WATERMARK_MASK(x) ((x) << 16)
+#define DPG_PIPE_LATENCY_CONTROL 0x1B33
+# define LATENCY_LOW_WATERMARK(x) ((x) << 0)
+# define LATENCY_HIGH_WATERMARK(x) ((x) << 16)
+
+/* 0x6bb8, 0x77b8, 0x103b8, 0x10fb8, 0x11bb8, 0x127b8 */
+#define VLINE_STATUS 0x1AEE
+# define VLINE_OCCURRED (1 << 0)
+# define VLINE_ACK (1 << 4)
+# define VLINE_STAT (1 << 12)
+# define VLINE_INTERRUPT (1 << 16)
+# define VLINE_INTERRUPT_TYPE (1 << 17)
+/* 0x6bbc, 0x77bc, 0x103bc, 0x10fbc, 0x11bbc, 0x127bc */
+#define VBLANK_STATUS 0x1AEF
+# define VBLANK_OCCURRED (1 << 0)
+# define VBLANK_ACK (1 << 4)
+# define VBLANK_STAT (1 << 12)
+# define VBLANK_INTERRUPT (1 << 16)
+# define VBLANK_INTERRUPT_TYPE (1 << 17)
+
+/* 0x6b40, 0x7740, 0x10340, 0x10f40, 0x11b40, 0x12740 */
+#define INT_MASK 0x1AD0
+# define VBLANK_INT_MASK (1 << 0)
+# define VLINE_INT_MASK (1 << 4)
+
+#define DISP_INTERRUPT_STATUS 0x183D
+# define LB_D1_VLINE_INTERRUPT (1 << 2)
+# define LB_D1_VBLANK_INTERRUPT (1 << 3)
+# define DC_HPD1_INTERRUPT (1 << 17)
+# define DC_HPD1_RX_INTERRUPT (1 << 18)
+# define DACA_AUTODETECT_INTERRUPT (1 << 22)
+# define DACB_AUTODETECT_INTERRUPT (1 << 23)
+# define DC_I2C_SW_DONE_INTERRUPT (1 << 24)
+# define DC_I2C_HW_DONE_INTERRUPT (1 << 25)
+#define DISP_INTERRUPT_STATUS_CONTINUE 0x183E
+# define LB_D2_VLINE_INTERRUPT (1 << 2)
+# define LB_D2_VBLANK_INTERRUPT (1 << 3)
+# define DC_HPD2_INTERRUPT (1 << 17)
+# define DC_HPD2_RX_INTERRUPT (1 << 18)
+# define DISP_TIMER_INTERRUPT (1 << 24)
+#define DISP_INTERRUPT_STATUS_CONTINUE2 0x183F
+# define LB_D3_VLINE_INTERRUPT (1 << 2)
+# define LB_D3_VBLANK_INTERRUPT (1 << 3)
+# define DC_HPD3_INTERRUPT (1 << 17)
+# define DC_HPD3_RX_INTERRUPT (1 << 18)
+#define DISP_INTERRUPT_STATUS_CONTINUE3 0x1840
+# define LB_D4_VLINE_INTERRUPT (1 << 2)
+# define LB_D4_VBLANK_INTERRUPT (1 << 3)
+# define DC_HPD4_INTERRUPT (1 << 17)
+# define DC_HPD4_RX_INTERRUPT (1 << 18)
+#define DISP_INTERRUPT_STATUS_CONTINUE4 0x1853
+# define LB_D5_VLINE_INTERRUPT (1 << 2)
+# define LB_D5_VBLANK_INTERRUPT (1 << 3)
+# define DC_HPD5_INTERRUPT (1 << 17)
+# define DC_HPD5_RX_INTERRUPT (1 << 18)
+#define DISP_INTERRUPT_STATUS_CONTINUE5 0x1854
+# define LB_D6_VLINE_INTERRUPT (1 << 2)
+# define LB_D6_VBLANK_INTERRUPT (1 << 3)
+# define DC_HPD6_INTERRUPT (1 << 17)
+# define DC_HPD6_RX_INTERRUPT (1 << 18)
+
+/* 0x6858, 0x7458, 0x10058, 0x10c58, 0x11858, 0x12458 */
+#define GRPH_INT_STATUS 0x1A16
+# define GRPH_PFLIP_INT_OCCURRED (1 << 0)
+# define GRPH_PFLIP_INT_CLEAR (1 << 8)
+/* 0x685c, 0x745c, 0x1005c, 0x10c5c, 0x1185c, 0x1245c */
+#define GRPH_INT_CONTROL 0x1A17
+# define GRPH_PFLIP_INT_MASK (1 << 0)
+# define GRPH_PFLIP_INT_TYPE (1 << 8)
+
+#define DAC_AUTODETECT_INT_CONTROL 0x19F2
+
+#define DC_HPD1_INT_STATUS 0x1807
+#define DC_HPD2_INT_STATUS 0x180A
+#define DC_HPD3_INT_STATUS 0x180D
+#define DC_HPD4_INT_STATUS 0x1810
+#define DC_HPD5_INT_STATUS 0x1813
+#define DC_HPD6_INT_STATUS 0x1816
+# define DC_HPDx_INT_STATUS (1 << 0)
+# define DC_HPDx_SENSE (1 << 1)
+# define DC_HPDx_RX_INT_STATUS (1 << 8)
+
+#define DC_HPD1_INT_CONTROL 0x1808
+#define DC_HPD2_INT_CONTROL 0x180B
+#define DC_HPD3_INT_CONTROL 0x180E
+#define DC_HPD4_INT_CONTROL 0x1811
+#define DC_HPD5_INT_CONTROL 0x1814
+#define DC_HPD6_INT_CONTROL 0x1817
+# define DC_HPDx_INT_ACK (1 << 0)
+# define DC_HPDx_INT_POLARITY (1 << 8)
+# define DC_HPDx_INT_EN (1 << 16)
+# define DC_HPDx_RX_INT_ACK (1 << 20)
+# define DC_HPDx_RX_INT_EN (1 << 24)
+
+#define DC_HPD1_CONTROL 0x1809
+#define DC_HPD2_CONTROL 0x180C
+#define DC_HPD3_CONTROL 0x180F
+#define DC_HPD4_CONTROL 0x1812
+#define DC_HPD5_CONTROL 0x1815
+#define DC_HPD6_CONTROL 0x1818
+# define DC_HPDx_CONNECTION_TIMER(x) ((x) << 0)
+# define DC_HPDx_RX_INT_TIMER(x) ((x) << 16)
+# define DC_HPDx_EN (1 << 28)
+
+#define DPG_PIPE_STUTTER_CONTROL 0x1B35
+# define STUTTER_ENABLE (1 << 0)
+
+/* 0x6e98, 0x7a98, 0x10698, 0x11298, 0x11e98, 0x12a98 */
+#define CRTC_STATUS_FRAME_COUNT 0x1BA6
+
+/* Audio clocks */
+#define DCCG_AUDIO_DTO_SOURCE 0x05ac
+# define DCCG_AUDIO_DTO0_SOURCE_SEL(x) ((x) << 0) /* crtc0 - crtc5 */
+# define DCCG_AUDIO_DTO_SEL (1 << 4) /* 0=dto0 1=dto1 */
+
+#define DCCG_AUDIO_DTO0_PHASE 0x05b0
+#define DCCG_AUDIO_DTO0_MODULE 0x05b4
+#define DCCG_AUDIO_DTO1_PHASE 0x05c0
+#define DCCG_AUDIO_DTO1_MODULE 0x05c4
+
+#define AFMT_AUDIO_SRC_CONTROL 0x1c4f
+#define AFMT_AUDIO_SRC_SELECT(x) (((x) & 7) << 0)
+/* AFMT_AUDIO_SRC_SELECT
+ * 0 = stream0
+ * 1 = stream1
+ * 2 = stream2
+ * 3 = stream3
+ * 4 = stream4
+ * 5 = stream5
+ */
+
+#define GRBM_CNTL 0x2000
+#define GRBM_READ_TIMEOUT(x) ((x) << 0)
+
+#define GRBM_STATUS2 0x2002
+#define RLC_RQ_PENDING (1 << 0)
+#define RLC_BUSY (1 << 8)
+#define TC_BUSY (1 << 9)
+
+#define GRBM_STATUS 0x2004
+#define CMDFIFO_AVAIL_MASK 0x0000000F
+#define RING2_RQ_PENDING (1 << 4)
+#define SRBM_RQ_PENDING (1 << 5)
+#define RING1_RQ_PENDING (1 << 6)
+#define CF_RQ_PENDING (1 << 7)
+#define PF_RQ_PENDING (1 << 8)
+#define GDS_DMA_RQ_PENDING (1 << 9)
+#define GRBM_EE_BUSY (1 << 10)
+#define DB_CLEAN (1 << 12)
+#define CB_CLEAN (1 << 13)
+#define TA_BUSY (1 << 14)
+#define GDS_BUSY (1 << 15)
+#define VGT_BUSY (1 << 17)
+#define IA_BUSY_NO_DMA (1 << 18)
+#define IA_BUSY (1 << 19)
+#define SX_BUSY (1 << 20)
+#define SPI_BUSY (1 << 22)
+#define BCI_BUSY (1 << 23)
+#define SC_BUSY (1 << 24)
+#define PA_BUSY (1 << 25)
+#define DB_BUSY (1 << 26)
+#define CP_COHERENCY_BUSY (1 << 28)
+#define CP_BUSY (1 << 29)
+#define CB_BUSY (1 << 30)
+#define GUI_ACTIVE (1 << 31)
+#define GRBM_STATUS_SE0 0x2005
+#define GRBM_STATUS_SE1 0x2006
+#define SE_DB_CLEAN (1 << 1)
+#define SE_CB_CLEAN (1 << 2)
+#define SE_BCI_BUSY (1 << 22)
+#define SE_VGT_BUSY (1 << 23)
+#define SE_PA_BUSY (1 << 24)
+#define SE_TA_BUSY (1 << 25)
+#define SE_SX_BUSY (1 << 26)
+#define SE_SPI_BUSY (1 << 27)
+#define SE_SC_BUSY (1 << 29)
+#define SE_DB_BUSY (1 << 30)
+#define SE_CB_BUSY (1 << 31)
+
+#define GRBM_SOFT_RESET 0x2008
+#define SOFT_RESET_CP (1 << 0)
+#define SOFT_RESET_CB (1 << 1)
+#define SOFT_RESET_RLC (1 << 2)
+#define SOFT_RESET_DB (1 << 3)
+#define SOFT_RESET_GDS (1 << 4)
+#define SOFT_RESET_PA (1 << 5)
+#define SOFT_RESET_SC (1 << 6)
+#define SOFT_RESET_BCI (1 << 7)
+#define SOFT_RESET_SPI (1 << 8)
+#define SOFT_RESET_SX (1 << 10)
+#define SOFT_RESET_TC (1 << 11)
+#define SOFT_RESET_TA (1 << 12)
+#define SOFT_RESET_VGT (1 << 14)
+#define SOFT_RESET_IA (1 << 15)
+
+#define GRBM_GFX_INDEX 0x200B
+#define INSTANCE_INDEX(x) ((x) << 0)
+#define SH_INDEX(x) ((x) << 8)
+#define SE_INDEX(x) ((x) << 16)
+#define SH_BROADCAST_WRITES (1 << 29)
+#define INSTANCE_BROADCAST_WRITES (1 << 30)
+#define SE_BROADCAST_WRITES (1 << 31)
+
+#define GRBM_INT_CNTL 0x2018
+# define RDERR_INT_ENABLE (1 << 0)
+# define GUI_IDLE_INT_ENABLE (1 << 19)
+
+#define CP_STRMOUT_CNTL 0x213F
+#define SCRATCH_REG0 0x2140
+#define SCRATCH_REG1 0x2141
+#define SCRATCH_REG2 0x2142
+#define SCRATCH_REG3 0x2143
+#define SCRATCH_REG4 0x2144
+#define SCRATCH_REG5 0x2145
+#define SCRATCH_REG6 0x2146
+#define SCRATCH_REG7 0x2147
+
+#define SCRATCH_UMSK 0x2150
+#define SCRATCH_ADDR 0x2151
+
+#define CP_SEM_WAIT_TIMER 0x216F
+
+#define CP_SEM_INCOMPLETE_TIMER_CNTL 0x2172
+
+#define CP_ME_CNTL 0x21B6
+#define CP_CE_HALT (1 << 24)
+#define CP_PFP_HALT (1 << 26)
+#define CP_ME_HALT (1 << 28)
+
+#define CP_COHER_CNTL2 0x217A
+
+#define CP_RB2_RPTR 0x21BE
+#define CP_RB1_RPTR 0x21BF
+#define CP_RB0_RPTR 0x21C0
+#define CP_RB_WPTR_DELAY 0x21C1
+
+#define CP_QUEUE_THRESHOLDS 0x21D8
+#define ROQ_IB1_START(x) ((x) << 0)
+#define ROQ_IB2_START(x) ((x) << 8)
+#define CP_MEQ_THRESHOLDS 0x21D9
+#define MEQ1_START(x) ((x) << 0)
+#define MEQ2_START(x) ((x) << 8)
+
+#define CP_PERFMON_CNTL 0x21FF
+
+#define VGT_VTX_VECT_EJECT_REG 0x222C
+
+#define VGT_CACHE_INVALIDATION 0x2231
+#define CACHE_INVALIDATION(x) ((x) << 0)
+#define VC_ONLY 0
+#define TC_ONLY 1
+#define VC_AND_TC 2
+#define AUTO_INVLD_EN(x) ((x) << 6)
+#define NO_AUTO 0
+#define ES_AUTO 1
+#define GS_AUTO 2
+#define ES_AND_GS_AUTO 3
+#define VGT_ESGS_RING_SIZE 0x2232
+#define VGT_GSVS_RING_SIZE 0x2233
+
+#define VGT_GS_VERTEX_REUSE 0x2235
+
+#define VGT_PRIMITIVE_TYPE 0x2256
+#define VGT_INDEX_TYPE 0x2257
+
+#define VGT_NUM_INDICES 0x225C
+#define VGT_NUM_INSTANCES 0x225D
+
+#define VGT_TF_RING_SIZE 0x2262
+
+#define VGT_HS_OFFCHIP_PARAM 0x226C
+
+#define VGT_TF_MEMORY_BASE 0x226E
+
+#define CC_GC_SHADER_ARRAY_CONFIG 0x226F
+#define INACTIVE_CUS_MASK 0xFFFF0000
+#define INACTIVE_CUS_SHIFT 16
+#define GC_USER_SHADER_ARRAY_CONFIG 0x2270
+
+#define PA_CL_ENHANCE 0x2285
+#define CLIP_VTX_REORDER_ENA (1 << 0)
+#define NUM_CLIP_SEQ(x) ((x) << 1)
+
+#define PA_SU_LINE_STIPPLE_VALUE 0x2298
+
+#define PA_SC_LINE_STIPPLE_STATE 0x22C4
+
+#define PA_SC_FORCE_EOV_MAX_CNTS 0x22C9
+#define FORCE_EOV_MAX_CLK_CNT(x) ((x) << 0)
+#define FORCE_EOV_MAX_REZ_CNT(x) ((x) << 16)
+
+#define PA_SC_FIFO_SIZE 0x22F3
+#define SC_FRONTEND_PRIM_FIFO_SIZE(x) ((x) << 0)
+#define SC_BACKEND_PRIM_FIFO_SIZE(x) ((x) << 6)
+#define SC_HIZ_TILE_FIFO_SIZE(x) ((x) << 15)
+#define SC_EARLYZ_TILE_FIFO_SIZE(x) ((x) << 23)
+
+#define PA_SC_ENHANCE 0x22FC
+
+#define SQ_CONFIG 0x2300
+
+#define SQC_CACHES 0x2302
+
+#define SQ_POWER_THROTTLE 0x2396
+#define MIN_POWER(x) ((x) << 0)
+#define MIN_POWER_MASK (0x3fff << 0)
+#define MIN_POWER_SHIFT 0
+#define MAX_POWER(x) ((x) << 16)
+#define MAX_POWER_MASK (0x3fff << 16)
+#define MAX_POWER_SHIFT 0
+#define SQ_POWER_THROTTLE2 0x2397
+#define MAX_POWER_DELTA(x) ((x) << 0)
+#define MAX_POWER_DELTA_MASK (0x3fff << 0)
+#define MAX_POWER_DELTA_SHIFT 0
+#define STI_SIZE(x) ((x) << 16)
+#define STI_SIZE_MASK (0x3ff << 16)
+#define STI_SIZE_SHIFT 16
+#define LTI_RATIO(x) ((x) << 27)
+#define LTI_RATIO_MASK (0xf << 27)
+#define LTI_RATIO_SHIFT 27
+
+#define SX_DEBUG_1 0x2418
+
+#define SPI_STATIC_THREAD_MGMT_1 0x2438
+#define SPI_STATIC_THREAD_MGMT_2 0x2439
+#define SPI_STATIC_THREAD_MGMT_3 0x243A
+#define SPI_PS_MAX_WAVE_ID 0x243B
+
+#define SPI_CONFIG_CNTL 0x2440
+
+#define SPI_CONFIG_CNTL_1 0x244F
+#define VTX_DONE_DELAY(x) ((x) << 0)
+#define INTERP_ONE_PRIM_PER_ROW (1 << 4)
+
+#define CGTS_TCC_DISABLE 0x2452
+#define CGTS_USER_TCC_DISABLE 0x2453
+#define TCC_DISABLE_MASK 0xFFFF0000
+#define TCC_DISABLE_SHIFT 16
+#define CGTS_SM_CTRL_REG 0x2454
+#define OVERRIDE (1 << 21)
+#define LS_OVERRIDE (1 << 22)
+
+#define SPI_LB_CU_MASK 0x24D5
+
+#define TA_CNTL_AUX 0x2542
+
+#define CC_RB_BACKEND_DISABLE 0x263D
+#define BACKEND_DISABLE(x) ((x) << 16)
+#define GB_ADDR_CONFIG 0x263E
+#define NUM_PIPES(x) ((x) << 0)
+#define NUM_PIPES_MASK 0x00000007
+#define NUM_PIPES_SHIFT 0
+#define PIPE_INTERLEAVE_SIZE(x) ((x) << 4)
+#define PIPE_INTERLEAVE_SIZE_MASK 0x00000070
+#define PIPE_INTERLEAVE_SIZE_SHIFT 4
+#define NUM_SHADER_ENGINES(x) ((x) << 12)
+#define NUM_SHADER_ENGINES_MASK 0x00003000
+#define NUM_SHADER_ENGINES_SHIFT 12
+#define SHADER_ENGINE_TILE_SIZE(x) ((x) << 16)
+#define SHADER_ENGINE_TILE_SIZE_MASK 0x00070000
+#define SHADER_ENGINE_TILE_SIZE_SHIFT 16
+#define NUM_GPUS(x) ((x) << 20)
+#define NUM_GPUS_MASK 0x00700000
+#define NUM_GPUS_SHIFT 20
+#define MULTI_GPU_TILE_SIZE(x) ((x) << 24)
+#define MULTI_GPU_TILE_SIZE_MASK 0x03000000
+#define MULTI_GPU_TILE_SIZE_SHIFT 24
+#define ROW_SIZE(x) ((x) << 28)
+#define ROW_SIZE_MASK 0x30000000
+#define ROW_SIZE_SHIFT 28
+
+#define GB_TILE_MODE0 0x2644
+# define MICRO_TILE_MODE(x) ((x) << 0)
+# define ADDR_SURF_DISPLAY_MICRO_TILING 0
+# define ADDR_SURF_THIN_MICRO_TILING 1
+# define ADDR_SURF_DEPTH_MICRO_TILING 2
+# define ARRAY_MODE(x) ((x) << 2)
+# define ARRAY_LINEAR_GENERAL 0
+# define ARRAY_LINEAR_ALIGNED 1
+# define ARRAY_1D_TILED_THIN1 2
+# define ARRAY_2D_TILED_THIN1 4
+# define PIPE_CONFIG(x) ((x) << 6)
+# define ADDR_SURF_P2 0
+# define ADDR_SURF_P4_8x16 4
+# define ADDR_SURF_P4_16x16 5
+# define ADDR_SURF_P4_16x32 6
+# define ADDR_SURF_P4_32x32 7
+# define ADDR_SURF_P8_16x16_8x16 8
+# define ADDR_SURF_P8_16x32_8x16 9
+# define ADDR_SURF_P8_32x32_8x16 10
+# define ADDR_SURF_P8_16x32_16x16 11
+# define ADDR_SURF_P8_32x32_16x16 12
+# define ADDR_SURF_P8_32x32_16x32 13
+# define ADDR_SURF_P8_32x64_32x32 14
+# define TILE_SPLIT(x) ((x) << 11)
+# define ADDR_SURF_TILE_SPLIT_64B 0
+# define ADDR_SURF_TILE_SPLIT_128B 1
+# define ADDR_SURF_TILE_SPLIT_256B 2
+# define ADDR_SURF_TILE_SPLIT_512B 3
+# define ADDR_SURF_TILE_SPLIT_1KB 4
+# define ADDR_SURF_TILE_SPLIT_2KB 5
+# define ADDR_SURF_TILE_SPLIT_4KB 6
+# define BANK_WIDTH(x) ((x) << 14)
+# define ADDR_SURF_BANK_WIDTH_1 0
+# define ADDR_SURF_BANK_WIDTH_2 1
+# define ADDR_SURF_BANK_WIDTH_4 2
+# define ADDR_SURF_BANK_WIDTH_8 3
+# define BANK_HEIGHT(x) ((x) << 16)
+# define ADDR_SURF_BANK_HEIGHT_1 0
+# define ADDR_SURF_BANK_HEIGHT_2 1
+# define ADDR_SURF_BANK_HEIGHT_4 2
+# define ADDR_SURF_BANK_HEIGHT_8 3
+# define MACRO_TILE_ASPECT(x) ((x) << 18)
+# define ADDR_SURF_MACRO_ASPECT_1 0
+# define ADDR_SURF_MACRO_ASPECT_2 1
+# define ADDR_SURF_MACRO_ASPECT_4 2
+# define ADDR_SURF_MACRO_ASPECT_8 3
+# define NUM_BANKS(x) ((x) << 20)
+# define ADDR_SURF_2_BANK 0
+# define ADDR_SURF_4_BANK 1
+# define ADDR_SURF_8_BANK 2
+# define ADDR_SURF_16_BANK 3
+#define GB_TILE_MODE1 0x2645
+#define GB_TILE_MODE2 0x2646
+#define GB_TILE_MODE3 0x2647
+#define GB_TILE_MODE4 0x2648
+#define GB_TILE_MODE5 0x2649
+#define GB_TILE_MODE6 0x264a
+#define GB_TILE_MODE7 0x264b
+#define GB_TILE_MODE8 0x264c
+#define GB_TILE_MODE9 0x264d
+#define GB_TILE_MODE10 0x264e
+#define GB_TILE_MODE11 0x264f
+#define GB_TILE_MODE12 0x2650
+#define GB_TILE_MODE13 0x2651
+#define GB_TILE_MODE14 0x2652
+#define GB_TILE_MODE15 0x2653
+#define GB_TILE_MODE16 0x2654
+#define GB_TILE_MODE17 0x2655
+#define GB_TILE_MODE18 0x2656
+#define GB_TILE_MODE19 0x2657
+#define GB_TILE_MODE20 0x2658
+#define GB_TILE_MODE21 0x2659
+#define GB_TILE_MODE22 0x265a
+#define GB_TILE_MODE23 0x265b
+#define GB_TILE_MODE24 0x265c
+#define GB_TILE_MODE25 0x265d
+#define GB_TILE_MODE26 0x265e
+#define GB_TILE_MODE27 0x265f
+#define GB_TILE_MODE28 0x2660
+#define GB_TILE_MODE29 0x2661
+#define GB_TILE_MODE30 0x2662
+#define GB_TILE_MODE31 0x2663
+
+#define CB_PERFCOUNTER0_SELECT0 0x2688
+#define CB_PERFCOUNTER0_SELECT1 0x2689
+#define CB_PERFCOUNTER1_SELECT0 0x268A
+#define CB_PERFCOUNTER1_SELECT1 0x268B
+#define CB_PERFCOUNTER2_SELECT0 0x268C
+#define CB_PERFCOUNTER2_SELECT1 0x268D
+#define CB_PERFCOUNTER3_SELECT0 0x268E
+#define CB_PERFCOUNTER3_SELECT1 0x268F
+
+#define CB_CGTT_SCLK_CTRL 0x2698
+
+#define GC_USER_RB_BACKEND_DISABLE 0x26DF
+#define BACKEND_DISABLE_MASK 0x00FF0000
+#define BACKEND_DISABLE_SHIFT 16
+
+#define TCP_CHAN_STEER_LO 0x2B03
+#define TCP_CHAN_STEER_HI 0x2B94
+
+#define CP_RB0_BASE 0x3040
+#define CP_RB0_CNTL 0x3041
+#define RB_BUFSZ(x) ((x) << 0)
+#define RB_BLKSZ(x) ((x) << 8)
+#define BUF_SWAP_32BIT (2 << 16)
+#define RB_NO_UPDATE (1 << 27)
+#define RB_RPTR_WR_ENA (1 << 31)
+
+#define CP_RB0_RPTR_ADDR 0x3043
+#define CP_RB0_RPTR_ADDR_HI 0x3044
+#define CP_RB0_WPTR 0x3045
+
+#define CP_PFP_UCODE_ADDR 0x3054
+#define CP_PFP_UCODE_DATA 0x3055
+#define CP_ME_RAM_RADDR 0x3056
+#define CP_ME_RAM_WADDR 0x3057
+#define CP_ME_RAM_DATA 0x3058
+
+#define CP_CE_UCODE_ADDR 0x305A
+#define CP_CE_UCODE_DATA 0x305B
+
+#define CP_RB1_BASE 0x3060
+#define CP_RB1_CNTL 0x3061
+#define CP_RB1_RPTR_ADDR 0x3062
+#define CP_RB1_RPTR_ADDR_HI 0x3063
+#define CP_RB1_WPTR 0x3064
+#define CP_RB2_BASE 0x3065
+#define CP_RB2_CNTL 0x3066
+#define CP_RB2_RPTR_ADDR 0x3067
+#define CP_RB2_RPTR_ADDR_HI 0x3068
+#define CP_RB2_WPTR 0x3069
+#define CP_INT_CNTL_RING0 0x306A
+#define CP_INT_CNTL_RING1 0x306B
+#define CP_INT_CNTL_RING2 0x306C
+# define CNTX_BUSY_INT_ENABLE (1 << 19)
+# define CNTX_EMPTY_INT_ENABLE (1 << 20)
+# define WAIT_MEM_SEM_INT_ENABLE (1 << 21)
+# define TIME_STAMP_INT_ENABLE (1 << 26)
+# define CP_RINGID2_INT_ENABLE (1 << 29)
+# define CP_RINGID1_INT_ENABLE (1 << 30)
+# define CP_RINGID0_INT_ENABLE (1 << 31)
+#define CP_INT_STATUS_RING0 0x306D
+#define CP_INT_STATUS_RING1 0x306E
+#define CP_INT_STATUS_RING2 0x306F
+# define WAIT_MEM_SEM_INT_STAT (1 << 21)
+# define TIME_STAMP_INT_STAT (1 << 26)
+# define CP_RINGID2_INT_STAT (1 << 29)
+# define CP_RINGID1_INT_STAT (1 << 30)
+# define CP_RINGID0_INT_STAT (1 << 31)
+
+#define CP_MEM_SLP_CNTL 0x3079
+# define CP_MEM_LS_EN (1 << 0)
+
+#define CP_DEBUG 0x307F
+
+#define RLC_CNTL 0x30C0
+# define RLC_ENABLE (1 << 0)
+#define RLC_RL_BASE 0x30C1
+#define RLC_RL_SIZE 0x30C2
+#define RLC_LB_CNTL 0x30C3
+# define LOAD_BALANCE_ENABLE (1 << 0)
+#define RLC_SAVE_AND_RESTORE_BASE 0x30C4
+#define RLC_LB_CNTR_MAX 0x30C5
+#define RLC_LB_CNTR_INIT 0x30C6
+
+#define RLC_CLEAR_STATE_RESTORE_BASE 0x30C8
+
+#define RLC_UCODE_ADDR 0x30CB
+#define RLC_UCODE_DATA 0x30CC
+
+#define RLC_GPU_CLOCK_COUNT_LSB 0x30CE
+#define RLC_GPU_CLOCK_COUNT_MSB 0x30CF
+#define RLC_CAPTURE_GPU_CLOCK_COUNT 0x30D0
+#define RLC_MC_CNTL 0x30D1
+#define RLC_UCODE_CNTL 0x30D2
+#define RLC_STAT 0x30D3
+# define RLC_BUSY_STATUS (1 << 0)
+# define GFX_POWER_STATUS (1 << 1)
+# define GFX_CLOCK_STATUS (1 << 2)
+# define GFX_LS_STATUS (1 << 3)
+
+#define RLC_PG_CNTL 0x30D7
+# define GFX_PG_ENABLE (1 << 0)
+# define GFX_PG_SRC (1 << 1)
+
+#define RLC_CGTT_MGCG_OVERRIDE 0x3100
+#define RLC_CGCG_CGLS_CTRL 0x3101
+# define CGCG_EN (1 << 0)
+# define CGLS_EN (1 << 1)
+
+#define RLC_TTOP_D 0x3105
+# define RLC_PUD(x) ((x) << 0)
+# define RLC_PUD_MASK (0xff << 0)
+# define RLC_PDD(x) ((x) << 8)
+# define RLC_PDD_MASK (0xff << 8)
+# define RLC_TTPD(x) ((x) << 16)
+# define RLC_TTPD_MASK (0xff << 16)
+# define RLC_MSD(x) ((x) << 24)
+# define RLC_MSD_MASK (0xff << 24)
+
+#define RLC_LB_INIT_CU_MASK 0x3107
+
+#define RLC_PG_AO_CU_MASK 0x310B
+#define RLC_MAX_PG_CU 0x310C
+# define MAX_PU_CU(x) ((x) << 0)
+# define MAX_PU_CU_MASK (0xff << 0)
+#define RLC_AUTO_PG_CTRL 0x310C
+# define AUTO_PG_EN (1 << 0)
+# define GRBM_REG_SGIT(x) ((x) << 3)
+# define GRBM_REG_SGIT_MASK (0xffff << 3)
+# define PG_AFTER_GRBM_REG_ST(x) ((x) << 19)
+# define PG_AFTER_GRBM_REG_ST_MASK (0x1fff << 19)
+
+#define RLC_SERDES_WR_MASTER_MASK_0 0x3115
+#define RLC_SERDES_WR_MASTER_MASK_1 0x3116
+#define RLC_SERDES_WR_CTRL 0x3117
+
+#define RLC_SERDES_MASTER_BUSY_0 0x3119
+#define RLC_SERDES_MASTER_BUSY_1 0x311A
+
+#define RLC_GCPM_GENERAL_3 0x311E
+
+#define DB_RENDER_CONTROL 0xA000
+
+#define DB_DEPTH_INFO 0xA00F
+
+#define PA_SC_RASTER_CONFIG 0xA0D4
+# define RASTER_CONFIG_RB_MAP_0 0
+# define RASTER_CONFIG_RB_MAP_1 1
+# define RASTER_CONFIG_RB_MAP_2 2
+# define RASTER_CONFIG_RB_MAP_3 3
+
+#define VGT_EVENT_INITIATOR 0xA2A4
+# define SAMPLE_STREAMOUTSTATS1 (1 << 0)
+# define SAMPLE_STREAMOUTSTATS2 (2 << 0)
+# define SAMPLE_STREAMOUTSTATS3 (3 << 0)
+# define CACHE_FLUSH_TS (4 << 0)
+# define CACHE_FLUSH (6 << 0)
+# define CS_PARTIAL_FLUSH (7 << 0)
+# define VGT_STREAMOUT_RESET (10 << 0)
+# define END_OF_PIPE_INCR_DE (11 << 0)
+# define END_OF_PIPE_IB_END (12 << 0)
+# define RST_PIX_CNT (13 << 0)
+# define VS_PARTIAL_FLUSH (15 << 0)
+# define PS_PARTIAL_FLUSH (16 << 0)
+# define CACHE_FLUSH_AND_INV_TS_EVENT (20 << 0)
+# define ZPASS_DONE (21 << 0)
+# define CACHE_FLUSH_AND_INV_EVENT (22 << 0)
+# define PERFCOUNTER_START (23 << 0)
+# define PERFCOUNTER_STOP (24 << 0)
+# define PIPELINESTAT_START (25 << 0)
+# define PIPELINESTAT_STOP (26 << 0)
+# define PERFCOUNTER_SAMPLE (27 << 0)
+# define SAMPLE_PIPELINESTAT (30 << 0)
+# define SAMPLE_STREAMOUTSTATS (32 << 0)
+# define RESET_VTX_CNT (33 << 0)
+# define VGT_FLUSH (36 << 0)
+# define BOTTOM_OF_PIPE_TS (40 << 0)
+# define DB_CACHE_FLUSH_AND_INV (42 << 0)
+# define FLUSH_AND_INV_DB_DATA_TS (43 << 0)
+# define FLUSH_AND_INV_DB_META (44 << 0)
+# define FLUSH_AND_INV_CB_DATA_TS (45 << 0)
+# define FLUSH_AND_INV_CB_META (46 << 0)
+# define CS_DONE (47 << 0)
+# define PS_DONE (48 << 0)
+# define FLUSH_AND_INV_CB_PIXEL_DATA (49 << 0)
+# define THREAD_TRACE_START (51 << 0)
+# define THREAD_TRACE_STOP (52 << 0)
+# define THREAD_TRACE_FLUSH (54 << 0)
+# define THREAD_TRACE_FINISH (55 << 0)
+
+/* PIF PHY0 registers idx/data 0x8/0xc */
+#define PB0_PIF_CNTL 0x10
+# define LS2_EXIT_TIME(x) ((x) << 17)
+# define LS2_EXIT_TIME_MASK (0x7 << 17)
+# define LS2_EXIT_TIME_SHIFT 17
+#define PB0_PIF_PAIRING 0x11
+# define MULTI_PIF (1 << 25)
+#define PB0_PIF_PWRDOWN_0 0x12
+# define PLL_POWER_STATE_IN_TXS2_0(x) ((x) << 7)
+# define PLL_POWER_STATE_IN_TXS2_0_MASK (0x7 << 7)
+# define PLL_POWER_STATE_IN_TXS2_0_SHIFT 7
+# define PLL_POWER_STATE_IN_OFF_0(x) ((x) << 10)
+# define PLL_POWER_STATE_IN_OFF_0_MASK (0x7 << 10)
+# define PLL_POWER_STATE_IN_OFF_0_SHIFT 10
+# define PLL_RAMP_UP_TIME_0(x) ((x) << 24)
+# define PLL_RAMP_UP_TIME_0_MASK (0x7 << 24)
+# define PLL_RAMP_UP_TIME_0_SHIFT 24
+#define PB0_PIF_PWRDOWN_1 0x13
+# define PLL_POWER_STATE_IN_TXS2_1(x) ((x) << 7)
+# define PLL_POWER_STATE_IN_TXS2_1_MASK (0x7 << 7)
+# define PLL_POWER_STATE_IN_TXS2_1_SHIFT 7
+# define PLL_POWER_STATE_IN_OFF_1(x) ((x) << 10)
+# define PLL_POWER_STATE_IN_OFF_1_MASK (0x7 << 10)
+# define PLL_POWER_STATE_IN_OFF_1_SHIFT 10
+# define PLL_RAMP_UP_TIME_1(x) ((x) << 24)
+# define PLL_RAMP_UP_TIME_1_MASK (0x7 << 24)
+# define PLL_RAMP_UP_TIME_1_SHIFT 24
+
+#define PB0_PIF_PWRDOWN_2 0x17
+# define PLL_POWER_STATE_IN_TXS2_2(x) ((x) << 7)
+# define PLL_POWER_STATE_IN_TXS2_2_MASK (0x7 << 7)
+# define PLL_POWER_STATE_IN_TXS2_2_SHIFT 7
+# define PLL_POWER_STATE_IN_OFF_2(x) ((x) << 10)
+# define PLL_POWER_STATE_IN_OFF_2_MASK (0x7 << 10)
+# define PLL_POWER_STATE_IN_OFF_2_SHIFT 10
+# define PLL_RAMP_UP_TIME_2(x) ((x) << 24)
+# define PLL_RAMP_UP_TIME_2_MASK (0x7 << 24)
+# define PLL_RAMP_UP_TIME_2_SHIFT 24
+#define PB0_PIF_PWRDOWN_3 0x18
+# define PLL_POWER_STATE_IN_TXS2_3(x) ((x) << 7)
+# define PLL_POWER_STATE_IN_TXS2_3_MASK (0x7 << 7)
+# define PLL_POWER_STATE_IN_TXS2_3_SHIFT 7
+# define PLL_POWER_STATE_IN_OFF_3(x) ((x) << 10)
+# define PLL_POWER_STATE_IN_OFF_3_MASK (0x7 << 10)
+# define PLL_POWER_STATE_IN_OFF_3_SHIFT 10
+# define PLL_RAMP_UP_TIME_3(x) ((x) << 24)
+# define PLL_RAMP_UP_TIME_3_MASK (0x7 << 24)
+# define PLL_RAMP_UP_TIME_3_SHIFT 24
+/* PIF PHY1 registers idx/data 0x10/0x14 */
+#define PB1_PIF_CNTL 0x10
+#define PB1_PIF_PAIRING 0x11
+#define PB1_PIF_PWRDOWN_0 0x12
+#define PB1_PIF_PWRDOWN_1 0x13
+
+#define PB1_PIF_PWRDOWN_2 0x17
+#define PB1_PIF_PWRDOWN_3 0x18
+/* PCIE registers idx/data 0x30/0x34 */
+#define PCIE_CNTL2 0x1c /* PCIE */
+# define SLV_MEM_LS_EN (1 << 16)
+# define SLV_MEM_AGGRESSIVE_LS_EN (1 << 17)
+# define MST_MEM_LS_EN (1 << 18)
+# define REPLAY_MEM_LS_EN (1 << 19)
+#define PCIE_LC_STATUS1 0x28 /* PCIE */
+# define LC_REVERSE_RCVR (1 << 0)
+# define LC_REVERSE_XMIT (1 << 1)
+# define LC_OPERATING_LINK_WIDTH_MASK (0x7 << 2)
+# define LC_OPERATING_LINK_WIDTH_SHIFT 2
+# define LC_DETECTED_LINK_WIDTH_MASK (0x7 << 5)
+# define LC_DETECTED_LINK_WIDTH_SHIFT 5
+
+#define PCIE_P_CNTL 0x40 /* PCIE */
+# define P_IGNORE_EDB_ERR (1 << 6)
+
+/* PCIE PORT registers idx/data 0x38/0x3c */
+#define PCIE_LC_CNTL 0xa0
+# define LC_L0S_INACTIVITY(x) ((x) << 8)
+# define LC_L0S_INACTIVITY_MASK (0xf << 8)
+# define LC_L0S_INACTIVITY_SHIFT 8
+# define LC_L1_INACTIVITY(x) ((x) << 12)
+# define LC_L1_INACTIVITY_MASK (0xf << 12)
+# define LC_L1_INACTIVITY_SHIFT 12
+# define LC_PMI_TO_L1_DIS (1 << 16)
+# define LC_ASPM_TO_L1_DIS (1 << 24)
+#define PCIE_LC_LINK_WIDTH_CNTL 0xa2 /* PCIE_P */
+# define LC_LINK_WIDTH_SHIFT 0
+# define LC_LINK_WIDTH_MASK 0x7
+# define LC_LINK_WIDTH_X0 0
+# define LC_LINK_WIDTH_X1 1
+# define LC_LINK_WIDTH_X2 2
+# define LC_LINK_WIDTH_X4 3
+# define LC_LINK_WIDTH_X8 4
+# define LC_LINK_WIDTH_X16 6
+# define LC_LINK_WIDTH_RD_SHIFT 4
+# define LC_LINK_WIDTH_RD_MASK 0x70
+# define LC_RECONFIG_ARC_MISSING_ESCAPE (1 << 7)
+# define LC_RECONFIG_NOW (1 << 8)
+# define LC_RENEGOTIATION_SUPPORT (1 << 9)
+# define LC_RENEGOTIATE_EN (1 << 10)
+# define LC_SHORT_RECONFIG_EN (1 << 11)
+# define LC_UPCONFIGURE_SUPPORT (1 << 12)
+# define LC_UPCONFIGURE_DIS (1 << 13)
+# define LC_DYN_LANES_PWR_STATE(x) ((x) << 21)
+# define LC_DYN_LANES_PWR_STATE_MASK (0x3 << 21)
+# define LC_DYN_LANES_PWR_STATE_SHIFT 21
+#define PCIE_LC_N_FTS_CNTL 0xa3 /* PCIE_P */
+# define LC_XMIT_N_FTS(x) ((x) << 0)
+# define LC_XMIT_N_FTS_MASK (0xff << 0)
+# define LC_XMIT_N_FTS_SHIFT 0
+# define LC_XMIT_N_FTS_OVERRIDE_EN (1 << 8)
+# define LC_N_FTS_MASK (0xff << 24)
+#define PCIE_LC_SPEED_CNTL 0xa4 /* PCIE_P */
+# define LC_GEN2_EN_STRAP (1 << 0)
+# define LC_GEN3_EN_STRAP (1 << 1)
+# define LC_TARGET_LINK_SPEED_OVERRIDE_EN (1 << 2)
+# define LC_TARGET_LINK_SPEED_OVERRIDE_MASK (0x3 << 3)
+# define LC_TARGET_LINK_SPEED_OVERRIDE_SHIFT 3
+# define LC_FORCE_EN_SW_SPEED_CHANGE (1 << 5)
+# define LC_FORCE_DIS_SW_SPEED_CHANGE (1 << 6)
+# define LC_FORCE_EN_HW_SPEED_CHANGE (1 << 7)
+# define LC_FORCE_DIS_HW_SPEED_CHANGE (1 << 8)
+# define LC_INITIATE_LINK_SPEED_CHANGE (1 << 9)
+# define LC_SPEED_CHANGE_ATTEMPTS_ALLOWED_MASK (0x3 << 10)
+# define LC_SPEED_CHANGE_ATTEMPTS_ALLOWED_SHIFT 10
+# define LC_CURRENT_DATA_RATE_MASK (0x3 << 13) /* 0/1/2 = gen1/2/3 */
+# define LC_CURRENT_DATA_RATE_SHIFT 13
+# define LC_CLR_FAILED_SPD_CHANGE_CNT (1 << 16)
+# define LC_OTHER_SIDE_EVER_SENT_GEN2 (1 << 18)
+# define LC_OTHER_SIDE_SUPPORTS_GEN2 (1 << 19)
+# define LC_OTHER_SIDE_EVER_SENT_GEN3 (1 << 20)
+# define LC_OTHER_SIDE_SUPPORTS_GEN3 (1 << 21)
+
+#define PCIE_LC_CNTL2 0xb1
+# define LC_ALLOW_PDWN_IN_L1 (1 << 17)
+# define LC_ALLOW_PDWN_IN_L23 (1 << 18)
+
+#define PCIE_LC_CNTL3 0xb5 /* PCIE_P */
+# define LC_GO_TO_RECOVERY (1 << 30)
+#define PCIE_LC_CNTL4 0xb6 /* PCIE_P */
+# define LC_REDO_EQ (1 << 5)
+# define LC_SET_QUIESCE (1 << 13)
+
+/*
+ * UVD
+ */
+#define UVD_UDEC_ADDR_CONFIG 0x3bd3
+#define UVD_UDEC_DB_ADDR_CONFIG 0x3bd4
+#define UVD_UDEC_DBW_ADDR_CONFIG 0x3bd5
+#define UVD_RBC_RB_RPTR 0x3da4
+#define UVD_RBC_RB_WPTR 0x3da5
+#define UVD_STATUS 0x3daf
+
+#define UVD_CGC_CTRL 0x3dc2
+# define DCM (1 << 0)
+# define CG_DT(x) ((x) << 2)
+# define CG_DT_MASK (0xf << 2)
+# define CLK_OD(x) ((x) << 6)
+# define CLK_OD_MASK (0x1f << 6)
+
+ /* UVD CTX indirect */
+#define UVD_CGC_MEM_CTRL 0xC0
+#define UVD_CGC_CTRL2 0xC1
+# define DYN_OR_EN (1 << 0)
+# define DYN_RR_EN (1 << 1)
+# define G_DIV_ID(x) ((x) << 2)
+# define G_DIV_ID_MASK (0x7 << 2)
+
+/*
+ * PM4
+ */
+#define PACKET0(reg, n) ((RADEON_PACKET_TYPE0 << 30) | \
+ (((reg) >> 2) & 0xFFFF) | \
+ ((n) & 0x3FFF) << 16)
+#define CP_PACKET2 0x80000000
+#define PACKET2_PAD_SHIFT 0
+#define PACKET2_PAD_MASK (0x3fffffff << 0)
+
+#define PACKET2(v) (CP_PACKET2 | REG_SET(PACKET2_PAD, (v)))
+#define RADEON_PACKET_TYPE3 3
+#define PACKET3(op, n) ((RADEON_PACKET_TYPE3 << 30) | \
+ (((op) & 0xFF) << 8) | \
+ ((n) & 0x3FFF) << 16)
+
+#define PACKET3_COMPUTE(op, n) (PACKET3(op, n) | 1 << 1)
+
+/* Packet 3 types */
+#define PACKET3_NOP 0x10
+#define PACKET3_SET_BASE 0x11
+#define PACKET3_BASE_INDEX(x) ((x) << 0)
+#define GDS_PARTITION_BASE 2
+#define CE_PARTITION_BASE 3
+#define PACKET3_CLEAR_STATE 0x12
+#define PACKET3_INDEX_BUFFER_SIZE 0x13
+#define PACKET3_DISPATCH_DIRECT 0x15
+#define PACKET3_DISPATCH_INDIRECT 0x16
+#define PACKET3_ALLOC_GDS 0x1B
+#define PACKET3_WRITE_GDS_RAM 0x1C
+#define PACKET3_ATOMIC_GDS 0x1D
+#define PACKET3_ATOMIC 0x1E
+#define PACKET3_OCCLUSION_QUERY 0x1F
+#define PACKET3_SET_PREDICATION 0x20
+#define PACKET3_REG_RMW 0x21
+#define PACKET3_COND_EXEC 0x22
+#define PACKET3_PRED_EXEC 0x23
+#define PACKET3_DRAW_INDIRECT 0x24
+#define PACKET3_DRAW_INDEX_INDIRECT 0x25
+#define PACKET3_INDEX_BASE 0x26
+#define PACKET3_DRAW_INDEX_2 0x27
+#define PACKET3_CONTEXT_CONTROL 0x28
+#define PACKET3_INDEX_TYPE 0x2A
+#define PACKET3_DRAW_INDIRECT_MULTI 0x2C
+#define PACKET3_DRAW_INDEX_AUTO 0x2D
+#define PACKET3_DRAW_INDEX_IMMD 0x2E
+#define PACKET3_NUM_INSTANCES 0x2F
+#define PACKET3_DRAW_INDEX_MULTI_AUTO 0x30
+#define PACKET3_INDIRECT_BUFFER_CONST 0x31
+#define PACKET3_INDIRECT_BUFFER 0x3F
+#define PACKET3_STRMOUT_BUFFER_UPDATE 0x34
+#define PACKET3_DRAW_INDEX_OFFSET_2 0x35
+#define PACKET3_DRAW_INDEX_MULTI_ELEMENT 0x36
+#define PACKET3_WRITE_DATA 0x37
+#define WRITE_DATA_DST_SEL(x) ((x) << 8)
+ /* 0 - register
+ * 1 - memory (sync - via GRBM)
+ * 2 - tc/l2
+ * 3 - gds
+ * 4 - reserved
+ * 5 - memory (async - direct)
+ */
+#define WR_ONE_ADDR (1 << 16)
+#define WR_CONFIRM (1 << 20)
+#define WRITE_DATA_ENGINE_SEL(x) ((x) << 30)
+ /* 0 - me
+ * 1 - pfp
+ * 2 - ce
+ */
+#define PACKET3_DRAW_INDEX_INDIRECT_MULTI 0x38
+#define PACKET3_MEM_SEMAPHORE 0x39
+#define PACKET3_MPEG_INDEX 0x3A
+#define PACKET3_COPY_DW 0x3B
+#define PACKET3_WAIT_REG_MEM 0x3C
+#define WAIT_REG_MEM_FUNCTION(x) ((x) << 0)
+ /* 0 - always
+ * 1 - <
+ * 2 - <=
+ * 3 - ==
+ * 4 - !=
+ * 5 - >=
+ * 6 - >
+ */
+#define WAIT_REG_MEM_MEM_SPACE(x) ((x) << 4)
+ /* 0 - reg
+ * 1 - mem
+ */
+#define WAIT_REG_MEM_ENGINE(x) ((x) << 8)
+ /* 0 - me
+ * 1 - pfp
+ */
+#define PACKET3_MEM_WRITE 0x3D
+#define PACKET3_COPY_DATA 0x40
+#define PACKET3_CP_DMA 0x41
+/* 1. header
+ * 2. SRC_ADDR_LO or DATA [31:0]
+ * 3. CP_SYNC [31] | SRC_SEL [30:29] | ENGINE [27] | DST_SEL [21:20] |
+ * SRC_ADDR_HI [7:0]
+ * 4. DST_ADDR_LO [31:0]
+ * 5. DST_ADDR_HI [7:0]
+ * 6. COMMAND [30:21] | BYTE_COUNT [20:0]
+ */
+# define PACKET3_CP_DMA_DST_SEL(x) ((x) << 20)
+ /* 0 - DST_ADDR
+ * 1 - GDS
+ */
+# define PACKET3_CP_DMA_ENGINE(x) ((x) << 27)
+ /* 0 - ME
+ * 1 - PFP
+ */
+# define PACKET3_CP_DMA_SRC_SEL(x) ((x) << 29)
+ /* 0 - SRC_ADDR
+ * 1 - GDS
+ * 2 - DATA
+ */
+# define PACKET3_CP_DMA_CP_SYNC (1 << 31)
+/* COMMAND */
+# define PACKET3_CP_DMA_DIS_WC (1 << 21)
+# define PACKET3_CP_DMA_CMD_SRC_SWAP(x) ((x) << 22)
+ /* 0 - none
+ * 1 - 8 in 16
+ * 2 - 8 in 32
+ * 3 - 8 in 64
+ */
+# define PACKET3_CP_DMA_CMD_DST_SWAP(x) ((x) << 24)
+ /* 0 - none
+ * 1 - 8 in 16
+ * 2 - 8 in 32
+ * 3 - 8 in 64
+ */
+# define PACKET3_CP_DMA_CMD_SAS (1 << 26)
+ /* 0 - memory
+ * 1 - register
+ */
+# define PACKET3_CP_DMA_CMD_DAS (1 << 27)
+ /* 0 - memory
+ * 1 - register
+ */
+# define PACKET3_CP_DMA_CMD_SAIC (1 << 28)
+# define PACKET3_CP_DMA_CMD_DAIC (1 << 29)
+# define PACKET3_CP_DMA_CMD_RAW_WAIT (1 << 30)
+#define PACKET3_PFP_SYNC_ME 0x42
+#define PACKET3_SURFACE_SYNC 0x43
+# define PACKET3_DEST_BASE_0_ENA (1 << 0)
+# define PACKET3_DEST_BASE_1_ENA (1 << 1)
+# define PACKET3_CB0_DEST_BASE_ENA (1 << 6)
+# define PACKET3_CB1_DEST_BASE_ENA (1 << 7)
+# define PACKET3_CB2_DEST_BASE_ENA (1 << 8)
+# define PACKET3_CB3_DEST_BASE_ENA (1 << 9)
+# define PACKET3_CB4_DEST_BASE_ENA (1 << 10)
+# define PACKET3_CB5_DEST_BASE_ENA (1 << 11)
+# define PACKET3_CB6_DEST_BASE_ENA (1 << 12)
+# define PACKET3_CB7_DEST_BASE_ENA (1 << 13)
+# define PACKET3_DB_DEST_BASE_ENA (1 << 14)
+# define PACKET3_DEST_BASE_2_ENA (1 << 19)
+# define PACKET3_DEST_BASE_3_ENA (1 << 21)
+# define PACKET3_TCL1_ACTION_ENA (1 << 22)
+# define PACKET3_TC_ACTION_ENA (1 << 23)
+# define PACKET3_CB_ACTION_ENA (1 << 25)
+# define PACKET3_DB_ACTION_ENA (1 << 26)
+# define PACKET3_SH_KCACHE_ACTION_ENA (1 << 27)
+# define PACKET3_SH_ICACHE_ACTION_ENA (1 << 29)
+#define PACKET3_ME_INITIALIZE 0x44
+#define PACKET3_ME_INITIALIZE_DEVICE_ID(x) ((x) << 16)
+#define PACKET3_COND_WRITE 0x45
+#define PACKET3_EVENT_WRITE 0x46
+#define EVENT_TYPE(x) ((x) << 0)
+#define EVENT_INDEX(x) ((x) << 8)
+ /* 0 - any non-TS event
+ * 1 - ZPASS_DONE
+ * 2 - SAMPLE_PIPELINESTAT
+ * 3 - SAMPLE_STREAMOUTSTAT*
+ * 4 - *S_PARTIAL_FLUSH
+ * 5 - EOP events
+ * 6 - EOS events
+ * 7 - CACHE_FLUSH, CACHE_FLUSH_AND_INV_EVENT
+ */
+#define INV_L2 (1 << 20)
+ /* INV TC L2 cache when EVENT_INDEX = 7 */
+#define PACKET3_EVENT_WRITE_EOP 0x47
+#define DATA_SEL(x) ((x) << 29)
+ /* 0 - discard
+ * 1 - send low 32bit data
+ * 2 - send 64bit data
+ * 3 - send 64bit counter value
+ */
+#define INT_SEL(x) ((x) << 24)
+ /* 0 - none
+ * 1 - interrupt only (DATA_SEL = 0)
+ * 2 - interrupt when data write is confirmed
+ */
+#define PACKET3_EVENT_WRITE_EOS 0x48
+#define PACKET3_PREAMBLE_CNTL 0x4A
+# define PACKET3_PREAMBLE_BEGIN_CLEAR_STATE (2 << 28)
+# define PACKET3_PREAMBLE_END_CLEAR_STATE (3 << 28)
+#define PACKET3_ONE_REG_WRITE 0x57
+#define PACKET3_LOAD_CONFIG_REG 0x5F
+#define PACKET3_LOAD_CONTEXT_REG 0x60
+#define PACKET3_LOAD_SH_REG 0x61
+#define PACKET3_SET_CONFIG_REG 0x68
+#define PACKET3_SET_CONFIG_REG_START 0x00002000
+#define PACKET3_SET_CONFIG_REG_END 0x00002c00
+#define PACKET3_SET_CONTEXT_REG 0x69
+#define PACKET3_SET_CONTEXT_REG_START 0x000a000
+#define PACKET3_SET_CONTEXT_REG_END 0x000a400
+#define PACKET3_SET_CONTEXT_REG_INDIRECT 0x73
+#define PACKET3_SET_RESOURCE_INDIRECT 0x74
+#define PACKET3_SET_SH_REG 0x76
+#define PACKET3_SET_SH_REG_START 0x00002c00
+#define PACKET3_SET_SH_REG_END 0x00003000
+#define PACKET3_SET_SH_REG_OFFSET 0x77
+#define PACKET3_ME_WRITE 0x7A
+#define PACKET3_SCRATCH_RAM_WRITE 0x7D
+#define PACKET3_SCRATCH_RAM_READ 0x7E
+#define PACKET3_CE_WRITE 0x7F
+#define PACKET3_LOAD_CONST_RAM 0x80
+#define PACKET3_WRITE_CONST_RAM 0x81
+#define PACKET3_WRITE_CONST_RAM_OFFSET 0x82
+#define PACKET3_DUMP_CONST_RAM 0x83
+#define PACKET3_INCREMENT_CE_COUNTER 0x84
+#define PACKET3_INCREMENT_DE_COUNTER 0x85
+#define PACKET3_WAIT_ON_CE_COUNTER 0x86
+#define PACKET3_WAIT_ON_DE_COUNTER 0x87
+#define PACKET3_WAIT_ON_DE_COUNTER_DIFF 0x88
+#define PACKET3_SET_CE_DE_COUNTERS 0x89
+#define PACKET3_WAIT_ON_AVAIL_BUFFER 0x8A
+#define PACKET3_SWITCH_BUFFER 0x8B
+
+/* ASYNC DMA - first instance at 0xd000, second at 0xd800 */
+#define DMA0_REGISTER_OFFSET 0x0 /* not a register */
+#define DMA1_REGISTER_OFFSET 0x200 /* not a register */
+
+#define DMA_RB_CNTL 0x3400
+# define DMA_RB_ENABLE (1 << 0)
+# define DMA_RB_SIZE(x) ((x) << 1) /* log2 */
+# define DMA_RB_SWAP_ENABLE (1 << 9) /* 8IN32 */
+# define DMA_RPTR_WRITEBACK_ENABLE (1 << 12)
+# define DMA_RPTR_WRITEBACK_SWAP_ENABLE (1 << 13) /* 8IN32 */
+# define DMA_RPTR_WRITEBACK_TIMER(x) ((x) << 16) /* log2 */
+#define DMA_RB_BASE 0x3401
+#define DMA_RB_RPTR 0x3402
+#define DMA_RB_WPTR 0x3403
+
+#define DMA_RB_RPTR_ADDR_HI 0x3407
+#define DMA_RB_RPTR_ADDR_LO 0x3408
+
+#define DMA_IB_CNTL 0x3409
+# define DMA_IB_ENABLE (1 << 0)
+# define DMA_IB_SWAP_ENABLE (1 << 4)
+# define CMD_VMID_FORCE (1 << 31)
+#define DMA_IB_RPTR 0x340a
+#define DMA_CNTL 0x340b
+# define TRAP_ENABLE (1 << 0)
+# define SEM_INCOMPLETE_INT_ENABLE (1 << 1)
+# define SEM_WAIT_INT_ENABLE (1 << 2)
+# define DATA_SWAP_ENABLE (1 << 3)
+# define FENCE_SWAP_ENABLE (1 << 4)
+# define CTXEMPTY_INT_ENABLE (1 << 28)
+#define DMA_STATUS_REG 0x340d
+# define DMA_IDLE (1 << 0)
+#define DMA_TILING_CONFIG 0x342e
+
+#define DMA_POWER_CNTL 0x342f
+# define MEM_POWER_OVERRIDE (1 << 8)
+#define DMA_CLK_CTRL 0x3430
+
+#define DMA_PG 0x3435
+# define PG_CNTL_ENABLE (1 << 0)
+#define DMA_PGFSM_CONFIG 0x3436
+#define DMA_PGFSM_WRITE 0x3437
+
+#define DMA_PACKET(cmd, b, t, s, n) ((((cmd) & 0xF) << 28) | \
+ (((b) & 0x1) << 26) | \
+ (((t) & 0x1) << 23) | \
+ (((s) & 0x1) << 22) | \
+ (((n) & 0xFFFFF) << 0))
+
+#define DMA_IB_PACKET(cmd, vmid, n) ((((cmd) & 0xF) << 28) | \
+ (((vmid) & 0xF) << 20) | \
+ (((n) & 0xFFFFF) << 0))
+
+#define DMA_PTE_PDE_PACKET(n) ((2 << 28) | \
+ (1 << 26) | \
+ (1 << 21) | \
+ (((n) & 0xFFFFF) << 0))
+
+/* async DMA Packet types */
+#define DMA_PACKET_WRITE 0x2
+#define DMA_PACKET_COPY 0x3
+#define DMA_PACKET_INDIRECT_BUFFER 0x4
+#define DMA_PACKET_SEMAPHORE 0x5
+#define DMA_PACKET_FENCE 0x6
+#define DMA_PACKET_TRAP 0x7
+#define DMA_PACKET_SRBM_WRITE 0x9
+#define DMA_PACKET_CONSTANT_FILL 0xd
+#define DMA_PACKET_POLL_REG_MEM 0xe
+#define DMA_PACKET_NOP 0xf
+
+#define VCE_STATUS 0x20004
+#define VCE_VCPU_CNTL 0x20014
+#define VCE_CLK_EN (1 << 0)
+#define VCE_VCPU_CACHE_OFFSET0 0x20024
+#define VCE_VCPU_CACHE_SIZE0 0x20028
+#define VCE_VCPU_CACHE_OFFSET1 0x2002c
+#define VCE_VCPU_CACHE_SIZE1 0x20030
+#define VCE_VCPU_CACHE_OFFSET2 0x20034
+#define VCE_VCPU_CACHE_SIZE2 0x20038
+#define VCE_SOFT_RESET 0x20120
+#define VCE_ECPU_SOFT_RESET (1 << 0)
+#define VCE_FME_SOFT_RESET (1 << 2)
+#define VCE_RB_BASE_LO2 0x2016c
+#define VCE_RB_BASE_HI2 0x20170
+#define VCE_RB_SIZE2 0x20174
+#define VCE_RB_RPTR2 0x20178
+#define VCE_RB_WPTR2 0x2017c
+#define VCE_RB_BASE_LO 0x20180
+#define VCE_RB_BASE_HI 0x20184
+#define VCE_RB_SIZE 0x20188
+#define VCE_RB_RPTR 0x2018c
+#define VCE_RB_WPTR 0x20190
+#define VCE_CLOCK_GATING_A 0x202f8
+#define VCE_CLOCK_GATING_B 0x202fc
+#define VCE_UENC_CLOCK_GATING 0x205bc
+#define VCE_UENC_REG_CLOCK_GATING 0x205c0
+#define VCE_FW_REG_STATUS 0x20e10
+# define VCE_FW_REG_STATUS_BUSY (1 << 0)
+# define VCE_FW_REG_STATUS_PASS (1 << 3)
+# define VCE_FW_REG_STATUS_DONE (1 << 11)
+#define VCE_LMI_FW_START_KEYSEL 0x20e18
+#define VCE_LMI_FW_PERIODIC_CTRL 0x20e20
+#define VCE_LMI_CTRL2 0x20e74
+#define VCE_LMI_CTRL 0x20e98
+#define VCE_LMI_VM_CTRL 0x20ea0
+#define VCE_LMI_SWAP_CNTL 0x20eb4
+#define VCE_LMI_SWAP_CNTL1 0x20eb8
+#define VCE_LMI_CACHE_CTRL 0x20ef4
+
+#define VCE_CMD_NO_OP 0x00000000
+#define VCE_CMD_END 0x00000001
+#define VCE_CMD_IB 0x00000002
+#define VCE_CMD_FENCE 0x00000003
+#define VCE_CMD_TRAP 0x00000004
+#define VCE_CMD_IB_AUTO 0x00000005
+#define VCE_CMD_SEMAPHORE 0x00000006
+
+
+//#dce stupp
+/* display controller offsets used for crtc/cur/lut/grph/viewport/etc. */
+#define SI_CRTC0_REGISTER_OFFSET 0 //(0x6df0 - 0x6df0)/4
+#define SI_CRTC1_REGISTER_OFFSET 0x300 //(0x79f0 - 0x6df0)/4
+#define SI_CRTC2_REGISTER_OFFSET 0x2600 //(0x105f0 - 0x6df0)/4
+#define SI_CRTC3_REGISTER_OFFSET 0x2900 //(0x111f0 - 0x6df0)/4
+#define SI_CRTC4_REGISTER_OFFSET 0x2c00 //(0x11df0 - 0x6df0)/4
+#define SI_CRTC5_REGISTER_OFFSET 0x2f00 //(0x129f0 - 0x6df0)/4
+
+#define CURSOR_WIDTH 64
+#define CURSOR_HEIGHT 64
+#define AMDGPU_MM_INDEX 0x0000
+#define AMDGPU_MM_DATA 0x0001
+
+#define VERDE_NUM_CRTC 6
+#define BLACKOUT_MODE_MASK 0x00000007
+#define VGA_RENDER_CONTROL 0xC0
+#define R_000300_VGA_RENDER_CONTROL 0xC0
+#define C_000300_VGA_VSTATUS_CNTL 0xFFFCFFFF
+#define EVERGREEN_CRTC_STATUS 0x1BA3
+#define EVERGREEN_CRTC_V_BLANK (1 << 0)
+#define EVERGREEN_CRTC_STATUS_POSITION 0x1BA4
+/* CRTC blocks at 0x6df0, 0x79f0, 0x105f0, 0x111f0, 0x11df0, 0x129f0 */
+#define EVERGREEN_CRTC_V_BLANK_START_END 0x1b8d
+#define EVERGREEN_CRTC_CONTROL 0x1b9c
+#define EVERGREEN_CRTC_MASTER_EN (1 << 0)
+#define EVERGREEN_CRTC_DISP_READ_REQUEST_DISABLE (1 << 24)
+#define EVERGREEN_CRTC_BLANK_CONTROL 0x1b9d
+#define EVERGREEN_CRTC_BLANK_DATA_EN (1 << 8)
+#define EVERGREEN_CRTC_V_BLANK (1 << 0)
+#define EVERGREEN_CRTC_STATUS_HV_COUNT 0x1ba8
+#define EVERGREEN_CRTC_UPDATE_LOCK 0x1bb5
+#define EVERGREEN_MASTER_UPDATE_LOCK 0x1bbd
+#define EVERGREEN_MASTER_UPDATE_MODE 0x1bbe
+#define EVERGREEN_GRPH_UPDATE_LOCK (1 << 16)
+#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1a07
+#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1a08
+#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS 0x1a04
+#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS 0x1a05
+#define EVERGREEN_GRPH_UPDATE 0x1a11
+#define EVERGREEN_VGA_MEMORY_BASE_ADDRESS 0xc4
+#define EVERGREEN_VGA_MEMORY_BASE_ADDRESS_HIGH 0xc9
+#define EVERGREEN_GRPH_SURFACE_UPDATE_PENDING (1 << 2)
+
+#define EVERGREEN_DATA_FORMAT 0x1ac0
+# define EVERGREEN_INTERLEAVE_EN (1 << 0)
+
+#define MC_SHARED_CHMAP__NOOFCHAN_MASK 0xf000
+#define MC_SHARED_CHMAP__NOOFCHAN__SHIFT 0xc
+
+#define R600_D1GRPH_ARRAY_MODE_LINEAR_GENERAL (0 << 20)
+#define R600_D1GRPH_ARRAY_MODE_LINEAR_ALIGNED (1 << 20)
+#define R600_D1GRPH_ARRAY_MODE_1D_TILED_THIN1 (2 << 20)
+#define R600_D1GRPH_ARRAY_MODE_2D_TILED_THIN1 (4 << 20)
+
+#define R700_D1GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1a45
+#define R700_D2GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1845
+
+#define R700_D2GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1847
+#define R700_D1GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1a47
+
+#define DISP_INTERRUPT_STATUS__LB_D1_VBLANK_INTERRUPT_MASK 0x8
+#define DISP_INTERRUPT_STATUS_CONTINUE__LB_D2_VBLANK_INTERRUPT_MASK 0x8
+#define DISP_INTERRUPT_STATUS_CONTINUE2__LB_D3_VBLANK_INTERRUPT_MASK 0x8
+#define DISP_INTERRUPT_STATUS_CONTINUE3__LB_D4_VBLANK_INTERRUPT_MASK 0x8
+#define DISP_INTERRUPT_STATUS_CONTINUE4__LB_D5_VBLANK_INTERRUPT_MASK 0x8
+#define DISP_INTERRUPT_STATUS_CONTINUE5__LB_D6_VBLANK_INTERRUPT_MASK 0x8
+
+#define DISP_INTERRUPT_STATUS__LB_D1_VLINE_INTERRUPT_MASK 0x4
+#define DISP_INTERRUPT_STATUS_CONTINUE__LB_D2_VLINE_INTERRUPT_MASK 0x4
+#define DISP_INTERRUPT_STATUS_CONTINUE2__LB_D3_VLINE_INTERRUPT_MASK 0x4
+#define DISP_INTERRUPT_STATUS_CONTINUE3__LB_D4_VLINE_INTERRUPT_MASK 0x4
+#define DISP_INTERRUPT_STATUS_CONTINUE4__LB_D5_VLINE_INTERRUPT_MASK 0x4
+#define DISP_INTERRUPT_STATUS_CONTINUE5__LB_D6_VLINE_INTERRUPT_MASK 0x4
+
+#define DISP_INTERRUPT_STATUS__DC_HPD1_INTERRUPT_MASK 0x20000
+#define DISP_INTERRUPT_STATUS_CONTINUE__DC_HPD2_INTERRUPT_MASK 0x20000
+#define DISP_INTERRUPT_STATUS_CONTINUE2__DC_HPD3_INTERRUPT_MASK 0x20000
+#define DISP_INTERRUPT_STATUS_CONTINUE3__DC_HPD4_INTERRUPT_MASK 0x20000
+#define DISP_INTERRUPT_STATUS_CONTINUE4__DC_HPD5_INTERRUPT_MASK 0x20000
+#define DISP_INTERRUPT_STATUS_CONTINUE5__DC_HPD6_INTERRUPT_MASK 0x20000
+
+#define GRPH_INTERRUPT_STATUS__GRPH_PFLIP_INT_OCCURRED_MASK 0x1
+#define GRPH_INTERRUPT_STATUS__GRPH_PFLIP_INT_CLEAR_MASK 0x100
+
+#define DC_HPD1_INT_CONTROL__DC_HPD1_INT_ACK_MASK 0x1
+
+#define R600_D1GRPH_SWAP_CONTROL 0x1843
+#define R600_D1GRPH_SWAP_ENDIAN_NONE (0 << 0)
+#define R600_D1GRPH_SWAP_ENDIAN_16BIT (1 << 0)
+#define R600_D1GRPH_SWAP_ENDIAN_32BIT (2 << 0)
+#define R600_D1GRPH_SWAP_ENDIAN_64BIT (3 << 0)
+
+#define AVIVO_D1VGA_CONTROL 0x00cc
+# define AVIVO_DVGA_CONTROL_MODE_ENABLE (1 << 0)
+# define AVIVO_DVGA_CONTROL_TIMING_SELECT (1 << 8)
+# define AVIVO_DVGA_CONTROL_SYNC_POLARITY_SELECT (1 << 9)
+# define AVIVO_DVGA_CONTROL_OVERSCAN_TIMING_SELECT (1 << 10)
+# define AVIVO_DVGA_CONTROL_OVERSCAN_COLOR_EN (1 << 16)
+# define AVIVO_DVGA_CONTROL_ROTATE (1 << 24)
+#define AVIVO_D2VGA_CONTROL 0x00ce
+
+#define R600_BUS_CNTL 0x1508
+# define R600_BIOS_ROM_DIS (1 << 1)
+
+#define R600_ROM_CNTL 0x580
+# define R600_SCK_OVERWRITE (1 << 1)
+# define R600_SCK_PRESCALE_CRYSTAL_CLK_SHIFT 28
+# define R600_SCK_PRESCALE_CRYSTAL_CLK_MASK (0xf << 28)
+
+#define GRPH_INTERRUPT_CONTROL__GRPH_PFLIP_INT_MASK_MASK 0x1
+
+#define FMT_BIT_DEPTH_CONTROL 0x1bf2
+#define FMT_TRUNCATE_EN (1 << 0)
+#define FMT_TRUNCATE_DEPTH (1 << 4)
+#define FMT_SPATIAL_DITHER_EN (1 << 8)
+#define FMT_SPATIAL_DITHER_MODE(x) ((x) << 9)
+#define FMT_SPATIAL_DITHER_DEPTH (1 << 12)
+#define FMT_FRAME_RANDOM_ENABLE (1 << 13)
+#define FMT_RGB_RANDOM_ENABLE (1 << 14)
+#define FMT_HIGHPASS_RANDOM_ENABLE (1 << 15)
+#define FMT_TEMPORAL_DITHER_EN (1 << 16)
+#define FMT_TEMPORAL_DITHER_DEPTH (1 << 20)
+#define FMT_TEMPORAL_DITHER_OFFSET(x) ((x) << 21)
+#define FMT_TEMPORAL_LEVEL (1 << 24)
+#define FMT_TEMPORAL_DITHER_RESET (1 << 25)
+#define FMT_25FRC_SEL(x) ((x) << 26)
+#define FMT_50FRC_SEL(x) ((x) << 28)
+#define FMT_75FRC_SEL(x) ((x) << 30)
+
+#define EVERGREEN_DC_LUT_CONTROL 0x1a80
+#define EVERGREEN_DC_LUT_BLACK_OFFSET_BLUE 0x1a81
+#define EVERGREEN_DC_LUT_BLACK_OFFSET_GREEN 0x1a82
+#define EVERGREEN_DC_LUT_BLACK_OFFSET_RED 0x1a83
+#define EVERGREEN_DC_LUT_WHITE_OFFSET_BLUE 0x1a84
+#define EVERGREEN_DC_LUT_WHITE_OFFSET_GREEN 0x1a85
+#define EVERGREEN_DC_LUT_WHITE_OFFSET_RED 0x1a86
+#define EVERGREEN_DC_LUT_30_COLOR 0x1a7c
+#define EVERGREEN_DC_LUT_RW_INDEX 0x1a79
+#define EVERGREEN_DC_LUT_WRITE_EN_MASK 0x1a7e
+#define EVERGREEN_DC_LUT_RW_MODE 0x1a78
+
+#define EVERGREEN_GRPH_ENABLE 0x1a00
+#define EVERGREEN_GRPH_CONTROL 0x1a01
+#define EVERGREEN_GRPH_DEPTH(x) (((x) & 0x3) << 0)
+#define EVERGREEN_GRPH_DEPTH_8BPP 0
+#define EVERGREEN_GRPH_DEPTH_16BPP 1
+#define EVERGREEN_GRPH_DEPTH_32BPP 2
+#define EVERGREEN_GRPH_NUM_BANKS(x) (((x) & 0x3) << 2)
+#define EVERGREEN_ADDR_SURF_2_BANK 0
+#define EVERGREEN_ADDR_SURF_4_BANK 1
+#define EVERGREEN_ADDR_SURF_8_BANK 2
+#define EVERGREEN_ADDR_SURF_16_BANK 3
+#define EVERGREEN_GRPH_Z(x) (((x) & 0x3) << 4)
+#define EVERGREEN_GRPH_BANK_WIDTH(x) (((x) & 0x3) << 6)
+#define EVERGREEN_ADDR_SURF_BANK_WIDTH_1 0
+#define EVERGREEN_ADDR_SURF_BANK_WIDTH_2 1
+#define EVERGREEN_ADDR_SURF_BANK_WIDTH_4 2
+#define EVERGREEN_ADDR_SURF_BANK_WIDTH_8 3
+#define EVERGREEN_GRPH_FORMAT(x) (((x) & 0x7) << 8)
+
+#define EVERGREEN_GRPH_FORMAT_INDEXED 0
+#define EVERGREEN_GRPH_FORMAT_ARGB1555 0
+#define EVERGREEN_GRPH_FORMAT_ARGB565 1
+#define EVERGREEN_GRPH_FORMAT_ARGB4444 2
+#define EVERGREEN_GRPH_FORMAT_AI88 3
+#define EVERGREEN_GRPH_FORMAT_MONO16 4
+#define EVERGREEN_GRPH_FORMAT_BGRA5551 5
+
+/* 32 BPP */
+#define EVERGREEN_GRPH_FORMAT_ARGB8888 0
+#define EVERGREEN_GRPH_FORMAT_ARGB2101010 1
+#define EVERGREEN_GRPH_FORMAT_32BPP_DIG 2
+#define EVERGREEN_GRPH_FORMAT_8B_ARGB2101010 3
+#define EVERGREEN_GRPH_FORMAT_BGRA1010102 4
+#define EVERGREEN_GRPH_FORMAT_8B_BGRA1010102 5
+#define EVERGREEN_GRPH_FORMAT_RGB111110 6
+#define EVERGREEN_GRPH_FORMAT_BGR101111 7
+#define EVERGREEN_GRPH_BANK_HEIGHT(x) (((x) & 0x3) << 11)
+#define EVERGREEN_ADDR_SURF_BANK_HEIGHT_1 0
+#define EVERGREEN_ADDR_SURF_BANK_HEIGHT_2 1
+#define EVERGREEN_ADDR_SURF_BANK_HEIGHT_4 2
+#define EVERGREEN_ADDR_SURF_BANK_HEIGHT_8 3
+#define EVERGREEN_GRPH_TILE_SPLIT(x) (((x) & 0x7) << 13)
+#define EVERGREEN_ADDR_SURF_TILE_SPLIT_64B 0
+#define EVERGREEN_ADDR_SURF_TILE_SPLIT_128B 1
+#define EVERGREEN_ADDR_SURF_TILE_SPLIT_256B 2
+#define EVERGREEN_ADDR_SURF_TILE_SPLIT_512B 3
+#define EVERGREEN_ADDR_SURF_TILE_SPLIT_1KB 4
+#define EVERGREEN_ADDR_SURF_TILE_SPLIT_2KB 5
+#define EVERGREEN_ADDR_SURF_TILE_SPLIT_4KB 6
+#define EVERGREEN_GRPH_MACRO_TILE_ASPECT(x) (((x) & 0x3) << 18)
+#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_1 0
+#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_2 1
+#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_4 2
+#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_8 3
+#define EVERGREEN_GRPH_ARRAY_MODE(x) (((x) & 0x7) << 20)
+#define EVERGREEN_GRPH_ARRAY_LINEAR_GENERAL 0
+#define EVERGREEN_GRPH_ARRAY_LINEAR_ALIGNED 1
+#define EVERGREEN_GRPH_ARRAY_1D_TILED_THIN1 2
+#define EVERGREEN_GRPH_ARRAY_2D_TILED_THIN1 4
+#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_1 0
+#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_2 1
+#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_4 2
+#define EVERGREEN_ADDR_SURF_MACRO_TILE_ASPECT_8 3
+
+#define EVERGREEN_GRPH_SWAP_CONTROL 0x1a03
+#define EVERGREEN_GRPH_ENDIAN_SWAP(x) (((x) & 0x3) << 0)
+# define EVERGREEN_GRPH_ENDIAN_NONE 0
+# define EVERGREEN_GRPH_ENDIAN_8IN16 1
+# define EVERGREEN_GRPH_ENDIAN_8IN32 2
+# define EVERGREEN_GRPH_ENDIAN_8IN64 3
+
+#define EVERGREEN_D3VGA_CONTROL 0xf8
+#define EVERGREEN_D4VGA_CONTROL 0xf9
+#define EVERGREEN_D5VGA_CONTROL 0xfa
+#define EVERGREEN_D6VGA_CONTROL 0xfb
+
+#define EVERGREEN_GRPH_SURFACE_ADDRESS_MASK 0xffffff00
+
+#define EVERGREEN_GRPH_LUT_10BIT_BYPASS_CONTROL 0x1a02
+#define EVERGREEN_LUT_10BIT_BYPASS_EN (1 << 8)
+
+#define EVERGREEN_GRPH_PITCH 0x1a06
+#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1a07
+#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1a08
+#define EVERGREEN_GRPH_SURFACE_OFFSET_X 0x1a09
+#define EVERGREEN_GRPH_SURFACE_OFFSET_Y 0x1a0a
+#define EVERGREEN_GRPH_X_START 0x1a0b
+#define EVERGREEN_GRPH_Y_START 0x1a0c
+#define EVERGREEN_GRPH_X_END 0x1a0d
+#define EVERGREEN_GRPH_Y_END 0x1a0e
+#define EVERGREEN_GRPH_UPDATE 0x1a11
+#define EVERGREEN_GRPH_SURFACE_UPDATE_PENDING (1 << 2)
+#define EVERGREEN_GRPH_UPDATE_LOCK (1 << 16)
+#define EVERGREEN_GRPH_FLIP_CONTROL 0x1a12
+#define EVERGREEN_GRPH_SURFACE_UPDATE_H_RETRACE_EN (1 << 0)
+
+#define EVERGREEN_VIEWPORT_START 0x1b5c
+#define EVERGREEN_VIEWPORT_SIZE 0x1b5d
+#define EVERGREEN_DESKTOP_HEIGHT 0x1ac1
+
+/* CUR blocks at 0x6998, 0x7598, 0x10198, 0x10d98, 0x11998, 0x12598 */
+#define EVERGREEN_CUR_CONTROL 0x1a66
+# define EVERGREEN_CURSOR_EN (1 << 0)
+# define EVERGREEN_CURSOR_MODE(x) (((x) & 0x3) << 8)
+# define EVERGREEN_CURSOR_MONO 0
+# define EVERGREEN_CURSOR_24_1 1
+# define EVERGREEN_CURSOR_24_8_PRE_MULT 2
+# define EVERGREEN_CURSOR_24_8_UNPRE_MULT 3
+# define EVERGREEN_CURSOR_2X_MAGNIFY (1 << 16)
+# define EVERGREEN_CURSOR_FORCE_MC_ON (1 << 20)
+# define EVERGREEN_CURSOR_URGENT_CONTROL(x) (((x) & 0x7) << 24)
+# define EVERGREEN_CURSOR_URGENT_ALWAYS 0
+# define EVERGREEN_CURSOR_URGENT_1_8 1
+# define EVERGREEN_CURSOR_URGENT_1_4 2
+# define EVERGREEN_CURSOR_URGENT_3_8 3
+# define EVERGREEN_CURSOR_URGENT_1_2 4
+#define EVERGREEN_CUR_SURFACE_ADDRESS 0x1a67
+# define EVERGREEN_CUR_SURFACE_ADDRESS_MASK 0xfffff000
+#define EVERGREEN_CUR_SIZE 0x1a68
+#define EVERGREEN_CUR_SURFACE_ADDRESS_HIGH 0x1a69
+#define EVERGREEN_CUR_POSITION 0x1a6a
+#define EVERGREEN_CUR_HOT_SPOT 0x1a6b
+#define EVERGREEN_CUR_COLOR1 0x1a6c
+#define EVERGREEN_CUR_COLOR2 0x1a6d
+#define EVERGREEN_CUR_UPDATE 0x1a6e
+# define EVERGREEN_CURSOR_UPDATE_PENDING (1 << 0)
+# define EVERGREEN_CURSOR_UPDATE_TAKEN (1 << 1)
+# define EVERGREEN_CURSOR_UPDATE_LOCK (1 << 16)
+# define EVERGREEN_CURSOR_DISABLE_MULTIPLE_UPDATE (1 << 24)
+
+
+#define NI_INPUT_CSC_CONTROL 0x1a35
+# define NI_INPUT_CSC_GRPH_MODE(x) (((x) & 0x3) << 0)
+# define NI_INPUT_CSC_BYPASS 0
+# define NI_INPUT_CSC_PROG_COEFF 1
+# define NI_INPUT_CSC_PROG_SHARED_MATRIXA 2
+# define NI_INPUT_CSC_OVL_MODE(x) (((x) & 0x3) << 4)
+
+#define NI_OUTPUT_CSC_CONTROL 0x1a3c
+# define NI_OUTPUT_CSC_GRPH_MODE(x) (((x) & 0x7) << 0)
+# define NI_OUTPUT_CSC_BYPASS 0
+# define NI_OUTPUT_CSC_TV_RGB 1
+# define NI_OUTPUT_CSC_YCBCR_601 2
+# define NI_OUTPUT_CSC_YCBCR_709 3
+# define NI_OUTPUT_CSC_PROG_COEFF 4
+# define NI_OUTPUT_CSC_PROG_SHARED_MATRIXB 5
+# define NI_OUTPUT_CSC_OVL_MODE(x) (((x) & 0x7) << 4)
+
+#define NI_DEGAMMA_CONTROL 0x1a58
+# define NI_GRPH_DEGAMMA_MODE(x) (((x) & 0x3) << 0)
+# define NI_DEGAMMA_BYPASS 0
+# define NI_DEGAMMA_SRGB_24 1
+# define NI_DEGAMMA_XVYCC_222 2
+# define NI_OVL_DEGAMMA_MODE(x) (((x) & 0x3) << 4)
+# define NI_ICON_DEGAMMA_MODE(x) (((x) & 0x3) << 8)
+# define NI_CURSOR_DEGAMMA_MODE(x) (((x) & 0x3) << 12)
+
+#define NI_GAMUT_REMAP_CONTROL 0x1a59
+# define NI_GRPH_GAMUT_REMAP_MODE(x) (((x) & 0x3) << 0)
+# define NI_GAMUT_REMAP_BYPASS 0
+# define NI_GAMUT_REMAP_PROG_COEFF 1
+# define NI_GAMUT_REMAP_PROG_SHARED_MATRIXA 2
+# define NI_GAMUT_REMAP_PROG_SHARED_MATRIXB 3
+# define NI_OVL_GAMUT_REMAP_MODE(x) (((x) & 0x3) << 4)
+
+#define NI_REGAMMA_CONTROL 0x1aa0
+# define NI_GRPH_REGAMMA_MODE(x) (((x) & 0x7) << 0)
+# define NI_REGAMMA_BYPASS 0
+# define NI_REGAMMA_SRGB_24 1
+# define NI_REGAMMA_XVYCC_222 2
+# define NI_REGAMMA_PROG_A 3
+# define NI_REGAMMA_PROG_B 4
+# define NI_OVL_REGAMMA_MODE(x) (((x) & 0x7) << 4)
+
+
+#define NI_PRESCALE_GRPH_CONTROL 0x1a2d
+# define NI_GRPH_PRESCALE_BYPASS (1 << 4)
+
+#define NI_PRESCALE_OVL_CONTROL 0x1a31
+# define NI_OVL_PRESCALE_BYPASS (1 << 4)
+
+#define NI_INPUT_GAMMA_CONTROL 0x1a10
+# define NI_GRPH_INPUT_GAMMA_MODE(x) (((x) & 0x3) << 0)
+# define NI_INPUT_GAMMA_USE_LUT 0
+# define NI_INPUT_GAMMA_BYPASS 1
+# define NI_INPUT_GAMMA_SRGB_24 2
+# define NI_INPUT_GAMMA_XVYCC_222 3
+# define NI_OVL_INPUT_GAMMA_MODE(x) (((x) & 0x3) << 4)
+
+#define IH_RB_WPTR__RB_OVERFLOW_MASK 0x1
+#define IH_RB_CNTL__WPTR_OVERFLOW_CLEAR_MASK 0x80000000
+#define SRBM_STATUS__IH_BUSY_MASK 0x20000
+#define SRBM_SOFT_RESET__SOFT_RESET_IH_MASK 0x400
+
+#define BLACKOUT_MODE_MASK 0x00000007
+#define VGA_RENDER_CONTROL 0xC0
+#define R_000300_VGA_RENDER_CONTROL 0xC0
+#define C_000300_VGA_VSTATUS_CNTL 0xFFFCFFFF
+#define EVERGREEN_CRTC_STATUS 0x1BA3
+#define EVERGREEN_CRTC_V_BLANK (1 << 0)
+#define EVERGREEN_CRTC_STATUS_POSITION 0x1BA4
+/* CRTC blocks at 0x6df0, 0x79f0, 0x105f0, 0x111f0, 0x11df0, 0x129f0 */
+#define EVERGREEN_CRTC_V_BLANK_START_END 0x1b8d
+#define EVERGREEN_CRTC_CONTROL 0x1b9c
+# define EVERGREEN_CRTC_MASTER_EN (1 << 0)
+# define EVERGREEN_CRTC_DISP_READ_REQUEST_DISABLE (1 << 24)
+#define EVERGREEN_CRTC_BLANK_CONTROL 0x1b9d
+# define EVERGREEN_CRTC_BLANK_DATA_EN (1 << 8)
+# define EVERGREEN_CRTC_V_BLANK (1 << 0)
+#define EVERGREEN_CRTC_STATUS_HV_COUNT 0x1ba8
+#define EVERGREEN_CRTC_UPDATE_LOCK 0x1bb5
+#define EVERGREEN_MASTER_UPDATE_LOCK 0x1bbd
+#define EVERGREEN_MASTER_UPDATE_MODE 0x1bbe
+#define EVERGREEN_GRPH_UPDATE_LOCK (1 << 16)
+#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS_HIGH 0x1a07
+#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS_HIGH 0x1a08
+#define EVERGREEN_GRPH_PRIMARY_SURFACE_ADDRESS 0x1a04
+#define EVERGREEN_GRPH_SECONDARY_SURFACE_ADDRESS 0x1a05
+#define EVERGREEN_GRPH_UPDATE 0x1a11
+#define EVERGREEN_VGA_MEMORY_BASE_ADDRESS 0xc4
+#define EVERGREEN_VGA_MEMORY_BASE_ADDRESS_HIGH 0xc9
+#define EVERGREEN_GRPH_SURFACE_UPDATE_PENDING (1 << 2)
+
+#define mmVM_CONTEXT1_CNTL__xxRANGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x10
+#define mmVM_CONTEXT1_CNTL__xxRANGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x4
+#define mmVM_CONTEXT1_CNTL__xxDUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x80
+#define mmVM_CONTEXT1_CNTL__xxDUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x7
+#define mmVM_CONTEXT1_CNTL__xxPDE0_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x400
+#define mmVM_CONTEXT1_CNTL__xxPDE0_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xa
+#define mmVM_CONTEXT1_CNTL__xxVALID_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x2000
+#define mmVM_CONTEXT1_CNTL__xxVALID_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0xd
+#define mmVM_CONTEXT1_CNTL__xxREAD_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x10000
+#define mmVM_CONTEXT1_CNTL__xxREAD_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x10
+#define mmVM_CONTEXT1_CNTL__xxWRITE_PROTECTION_FAULT_ENABLE_DEFAULT_MASK 0x80000
+#define mmVM_CONTEXT1_CNTL__xxWRITE_PROTECTION_FAULT_ENABLE_DEFAULT__SHIFT 0x13
+
+#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxVMID_MASK 0x1e000000
+#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxVMID__SHIFT 0x19
+#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxPROTECTIONS_MASK 0xff
+#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxPROTECTIONS__SHIFT 0x0
+#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxMEMORY_CLIENT_ID_MASK 0xff000
+#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxMEMORY_CLIENT_ID__SHIFT 0xc
+#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxMEMORY_CLIENT_RW_MASK 0x1000000
+#define mmVM_CONTEXT1_PROTECTION_FAULT_STATUS__xxMEMORY_CLIENT_RW__SHIFT 0x18
+
+#define mmMC_SHARED_BLACKOUT_CNTL__xxBLACKOUT_MODE_MASK 0x7
+#define mmMC_SHARED_BLACKOUT_CNTL__xxBLACKOUT_MODE__SHIFT 0x0
+
+#define mmBIF_FB_EN__xxFB_READ_EN_MASK 0x1
+#define mmBIF_FB_EN__xxFB_READ_EN__SHIFT 0x0
+#define mmBIF_FB_EN__xxFB_WRITE_EN_MASK 0x2
+#define mmBIF_FB_EN__xxFB_WRITE_EN__SHIFT 0x1
+
+#define mmSRBM_SOFT_RESET__xxSOFT_RESET_VMC_MASK 0x20000
+#define mmSRBM_SOFT_RESET__xxSOFT_RESET_VMC__SHIFT 0x11
+#define mmSRBM_SOFT_RESET__xxSOFT_RESET_MC_MASK 0x800
+#define mmSRBM_SOFT_RESET__xxSOFT_RESET_MC__SHIFT 0xb
+
+#define VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x8
+#define VM_CONTEXT1_CNTL__RANGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x3
+#define VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x40
+#define VM_CONTEXT1_CNTL__DUMMY_PAGE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x6
+#define VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x200
+#define VM_CONTEXT1_CNTL__PDE0_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x9
+#define VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x1000
+#define VM_CONTEXT1_CNTL__VALID_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xc
+#define VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x8000
+#define VM_CONTEXT1_CNTL__READ_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0xf
+#define VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT_MASK 0x40000
+#define VM_CONTEXT1_CNTL__WRITE_PROTECTION_FAULT_ENABLE_INTERRUPT__SHIFT 0x12
+
+#define MC_SEQ_MISC0__MT__MASK 0xf0000000
+#define MC_SEQ_MISC0__MT__GDDR1 0x10000000
+#define MC_SEQ_MISC0__MT__DDR2 0x20000000
+#define MC_SEQ_MISC0__MT__GDDR3 0x30000000
+#define MC_SEQ_MISC0__MT__GDDR4 0x40000000
+#define MC_SEQ_MISC0__MT__GDDR5 0x50000000
+#define MC_SEQ_MISC0__MT__HBM 0x60000000
+#define MC_SEQ_MISC0__MT__DDR3 0xB0000000
+
+#define SRBM_STATUS__MCB_BUSY_MASK 0x200
+#define SRBM_STATUS__MCB_BUSY__SHIFT 0x9
+#define SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK 0x400
+#define SRBM_STATUS__MCB_NON_DISPLAY_BUSY__SHIFT 0xa
+#define SRBM_STATUS__MCC_BUSY_MASK 0x800
+#define SRBM_STATUS__MCC_BUSY__SHIFT 0xb
+#define SRBM_STATUS__MCD_BUSY_MASK 0x1000
+#define SRBM_STATUS__MCD_BUSY__SHIFT 0xc
+#define SRBM_STATUS__VMC_BUSY_MASK 0x100
+#define SRBM_STATUS__VMC_BUSY__SHIFT 0x8
+
+
+#define GRBM_STATUS__GUI_ACTIVE_MASK 0x80000000
+#define CP_INT_CNTL_RING__TIME_STAMP_INT_ENABLE_MASK 0x4000000
+#define CP_INT_CNTL_RING0__PRIV_REG_INT_ENABLE_MASK 0x800000
+#define CP_INT_CNTL_RING0__PRIV_INSTR_INT_ENABLE_MASK 0x400000
+#define PACKET3_SEM_WAIT_ON_SIGNAL (0x1 << 12)
+#define PACKET3_SEM_SEL_SIGNAL (0x6 << 29)
+#define PACKET3_SEM_SEL_WAIT (0x7 << 29)
+
+#define CONFIG_CNTL 0x1509
+#define CC_DRM_ID_STRAPS 0X1559
+#define AMDGPU_PCIE_INDEX 0xc
+#define AMDGPU_PCIE_DATA 0xd
+
+#define DMA_SEM_INCOMPLETE_TIMER_CNTL 0x3411
+#define DMA_SEM_WAIT_FAIL_TIMER_CNTL 0x3412
+#define DMA_MODE 0x342f
+#define DMA_RB_RPTR_ADDR_HI 0x3407
+#define DMA_RB_RPTR_ADDR_LO 0x3408
+#define DMA_BUSY_MASK 0x20
+#define DMA1_BUSY_MASK 0X40
+#define SDMA_MAX_INSTANCE 2
+
+#define PCIE_BUS_CLK 10000
+#define TCLK (PCIE_BUS_CLK / 10)
+#define CC_DRM_ID_STRAPS__ATI_REV_ID_MASK 0xf0000000
+#define CC_DRM_ID_STRAPS__ATI_REV_ID__SHIFT 0x1c
+#define PCIE_PORT_INDEX 0xe
+#define PCIE_PORT_DATA 0xf
+#define EVERGREEN_PIF_PHY0_INDEX 0x8
+#define EVERGREEN_PIF_PHY0_DATA 0xc
+#define EVERGREEN_PIF_PHY1_INDEX 0x10
+#define EVERGREEN_PIF_PHY1_DATA 0x14
+
+#define MC_VM_FB_OFFSET 0x81a
+
+#endif
#define mmUVD_UDEC_ADDR_CONFIG 0x3bd3
#define mmUVD_UDEC_DB_ADDR_CONFIG 0x3bd4
#define mmUVD_UDEC_DBW_ADDR_CONFIG 0x3bd5
+#define mmUVD_NO_OP 0x3bff
#define mmUVD_SEMA_CNTL 0x3d00
#define mmUVD_LMI_EXT40_ADDR 0x3d26
#define mmUVD_CTX_INDEX 0x3d28
#define mmUVD_UDEC_ADDR_CONFIG 0x3bd3
#define mmUVD_UDEC_DB_ADDR_CONFIG 0x3bd4
#define mmUVD_UDEC_DBW_ADDR_CONFIG 0x3bd5
+#define mmUVD_NO_OP 0x3bff
#define mmUVD_LMI_RBC_RB_64BIT_BAR_LOW 0x3c69
#define mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH 0x3c68
#define mmUVD_LMI_RBC_IB_64BIT_BAR_LOW 0x3c67
#define mmUVD_UDEC_DB_ADDR_CONFIG 0x3bd4
#define mmUVD_UDEC_DBW_ADDR_CONFIG 0x3bd5
#define mmUVD_POWER_STATUS_U 0x3bfd
+#define mmUVD_NO_OP 0x3bff
#define mmUVD_LMI_RBC_RB_64BIT_BAR_LOW 0x3c69
#define mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH 0x3c68
#define mmUVD_LMI_RBC_IB_64BIT_BAR_LOW 0x3c67
union
{
ATOM_COMPUTE_CLOCK_FREQ ulClock; //Input Parameter
+ ULONG ulClockParams; //ULONG access for BE
ATOM_S_MPLL_FB_DIVIDER ulFbDiv; //Output Parameter
};
UCHAR ucRefDiv; //Output Parameter
union
{
ATOM_COMPUTE_CLOCK_FREQ ulClock; //Input Parameter
+ ULONG ulClockParams; //ULONG access for BE
ATOM_S_MPLL_FB_DIVIDER ulFbDiv; //Output Parameter
};
UCHAR ucRefDiv; //Output Parameter
CGS_SYSTEM_INFO_PG_FLAGS,
CGS_SYSTEM_INFO_GFX_CU_INFO,
CGS_SYSTEM_INFO_GFX_SE_INFO,
+ CGS_SYSTEM_INFO_PCIE_SUB_SYS_ID,
+ CGS_SYSTEM_INFO_PCIE_SUB_SYS_VENDOR_ID,
CGS_SYSTEM_INFO_ID_MAXIMUM,
};
#include "eventmanager.h"
#include "pp_debug.h"
+
#define PP_CHECK(handle) \
do { \
if ((handle) == NULL || (handle)->pp_valid != PP_VALID) \
pp_handle = (struct pp_instance *)handle;
eventmgr = pp_handle->eventmgr;
- if (eventmgr != NULL || eventmgr->pp_eventmgr_fini != NULL)
+ if (eventmgr != NULL && eventmgr->pp_eventmgr_fini != NULL)
eventmgr->pp_eventmgr_fini(eventmgr);
smumgr = pp_handle->smu_mgr;
- if (smumgr != NULL || smumgr->smumgr_funcs != NULL ||
+ if (smumgr != NULL && smumgr->smumgr_funcs != NULL &&
smumgr->smumgr_funcs->smu_fini != NULL)
smumgr->smumgr_funcs->smu_fini(smumgr);
ret = pem_handle_event(pp_handle->eventmgr, event_id, &data);
break;
case AMD_PP_EVENT_READJUST_POWER_STATE:
- pp_handle->hwmgr->current_ps = pp_handle->hwmgr->boot_ps;
ret = pem_handle_event(pp_handle->eventmgr, event_id, &data);
break;
default:
PP_CHECK_HW(hwmgr);
if (!hwmgr->hardcode_pp_table) {
- hwmgr->hardcode_pp_table =
- kzalloc(hwmgr->soft_pp_table_size, GFP_KERNEL);
+ hwmgr->hardcode_pp_table = kmemdup(hwmgr->soft_pp_table,
+ hwmgr->soft_pp_table_size,
+ GFP_KERNEL);
if (!hwmgr->hardcode_pp_table)
return -ENOMEM;
-
- /* to avoid powerplay crash when hardcode pptable is empty */
- memcpy(hwmgr->hardcode_pp_table, hwmgr->soft_pp_table,
- hwmgr->soft_pp_table_size);
}
memcpy(hwmgr->hardcode_pp_table, buf, size);
int i;
table_entries = hwmgr->num_ps;
+
state = hwmgr->ps;
for (i = 0; i < table_entries; i++) {
if (state->id == *state_id) {
- hwmgr->request_ps = state;
+ memcpy(hwmgr->request_ps, state, hwmgr->ps_size);
return 0;
}
state = (struct pp_power_state *)((unsigned long)state + hwmgr->ps_size);
if (!equal || phm_check_smc_update_required_for_display_configuration(hwmgr)) {
phm_apply_state_adjust_rules(hwmgr, requested, pcurrent);
phm_set_power_state(hwmgr, &pcurrent->hardware, &requested->hardware);
- hwmgr->current_ps = requested;
+ memcpy(hwmgr->current_ps, hwmgr->request_ps, hwmgr->ps_size);
}
return 0;
}
HARDWARE_MGR = hwmgr.o processpptables.o functiontables.o \
hardwaremanager.o pp_acpi.o cz_hwmgr.o \
- cz_clockpowergating.o \
- tonga_processpptables.o ppatomctrl.o \
+ cz_clockpowergating.o tonga_powertune.o\
+ process_pptables_v1_0.o ppatomctrl.o \
tonga_hwmgr.o pppcielanes.o tonga_thermal.o\
fiji_powertune.o fiji_hwmgr.o tonga_clockpowergating.o \
fiji_clockpowergating.o fiji_thermal.o \
polaris10_hwmgr.o polaris10_powertune.o polaris10_thermal.o \
- polaris10_clockpowergating.o
+ polaris10_clockpowergating.o iceland_hwmgr.o \
+ iceland_clockpowergating.o iceland_thermal.o \
+ iceland_powertune.o
AMD_PP_HWMGR = $(addprefix $(AMD_PP_PATH)/hwmgr/,$(HARDWARE_MGR))
int result;
cz_hwmgr->gfx_ramp_step = 256*25/100;
-
cz_hwmgr->gfx_ramp_delay = 1; /* by default, we delay 1us */
for (i = 0; i < CZ_MAX_HARDWARE_POWERLEVELS; i++)
cz_hwmgr->mgcg_cgtt_local0 = 0x00000000;
cz_hwmgr->mgcg_cgtt_local1 = 0x00000000;
-
cz_hwmgr->clock_slow_down_freq = 25000;
-
cz_hwmgr->skip_clock_slow_down = 1;
-
cz_hwmgr->enable_nb_ps_policy = 1; /* disable until UNB is ready, Enabled */
-
cz_hwmgr->voltage_drop_in_dce_power_gating = 0; /* disable until fully verified */
-
cz_hwmgr->voting_rights_clients = 0x00C00033;
-
cz_hwmgr->static_screen_threshold = 8;
-
cz_hwmgr->ddi_power_gating_disabled = 0;
-
cz_hwmgr->bapm_enabled = 1;
-
cz_hwmgr->voltage_drop_threshold = 0;
-
cz_hwmgr->gfx_power_gating_threshold = 500;
-
cz_hwmgr->vce_slow_sclk_threshold = 20000;
-
cz_hwmgr->dce_slow_sclk_threshold = 30000;
-
cz_hwmgr->disable_driver_thermal_policy = 1;
-
cz_hwmgr->disable_nb_ps3_in_battery = 0;
phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
phm_cap_set(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_NonABMSupportInPPLib);
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_SclkDeepSleep);
-
phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_DynamicM3Arbiter);
PHM_PlatformCaps_DynamicPatchPowerState);
cz_hwmgr->thermal_auto_throttling_treshold = 0;
-
cz_hwmgr->tdr_clock = 0;
-
cz_hwmgr->disable_gfx_power_gating_in_uvd = 0;
phm_cap_set(hwmgr->platform_descriptor.platformCaps,
(uint8_t)cz_hwmgr->sys_info.bootup_nb_voltage_index;
cz_hwmgr->boot_power_level.dsDividerIndex = 0;
-
cz_hwmgr->boot_power_level.ssDividerIndex = 0;
-
cz_hwmgr->boot_power_level.allowGnbSlow = 1;
-
cz_hwmgr->boot_power_level.forceNBPstate = 0;
-
cz_hwmgr->boot_power_level.hysteresis_up = 0;
-
cz_hwmgr->boot_power_level.numSIMDToPowerDown = 0;
-
cz_hwmgr->boot_power_level.display_wm = 0;
-
cz_hwmgr->boot_power_level.vce_wm = 0;
return 0;
cz_hwmgr->sclk_dpm.soft_max_clk = table->entries[table->count - 1].clk;
clock = hwmgr->display_config.min_core_set_clock;
-;
if (clock == 0)
printk(KERN_INFO "[ powerplay ] min_core_set_clock not set\n");
smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
PPSMC_MSG_SetWatermarkFrequency,
- cz_hwmgr->sclk_dpm.soft_max_clk);
+ cz_hwmgr->sclk_dpm.soft_max_clk);
return 0;
}
PP_DBG_LOG("enabling ALL SMU features.\n");
dpm_features |= NB_DPM_MASK;
ret = smum_send_msg_to_smc_with_parameter(
- hwmgr->smumgr,
- PPSMC_MSG_EnableAllSmuFeatures,
- dpm_features);
+ hwmgr->smumgr,
+ PPSMC_MSG_EnableAllSmuFeatures,
+ dpm_features);
if (ret == 0)
cz_hwmgr->is_nb_dpm_enabled = true;
}
static int cz_hwmgr_backend_fini(struct pp_hwmgr *hwmgr)
{
- if (hwmgr != NULL || hwmgr->backend != NULL) {
+ if (hwmgr != NULL && hwmgr->backend != NULL) {
kfree(hwmgr->backend);
kfree(hwmgr);
}
PPSMC_MSG_SetUvdHardMin));
cz_enable_disable_uvd_dpm(hwmgr, true);
- } else
+ } else {
cz_enable_disable_uvd_dpm(hwmgr, true);
- } else
+ }
+ } else {
cz_enable_disable_uvd_dpm(hwmgr, false);
+ }
return 0;
}
struct cz_hwmgr *hw_data = (struct cz_hwmgr *)(hwmgr->backend);
if (separation_time !=
- hw_data->cc6_settings.cpu_pstate_separation_time
- || cc6_disable !=
- hw_data->cc6_settings.cpu_cc6_disable
- || pstate_disable !=
- hw_data->cc6_settings.cpu_pstate_disable
- || pstate_switch_disable !=
- hw_data->cc6_settings.nb_pstate_switch_disable) {
+ hw_data->cc6_settings.cpu_pstate_separation_time ||
+ cc6_disable != hw_data->cc6_settings.cpu_cc6_disable ||
+ pstate_disable != hw_data->cc6_settings.cpu_pstate_disable ||
+ pstate_switch_disable != hw_data->cc6_settings.nb_pstate_switch_disable) {
hw_data->cc6_settings.cc6_setting_changed = true;
ps = cast_const_PhwCzPowerState(state);
level_index = index > ps->level - 1 ? ps->level - 1 : index;
-
- level->coreClock = ps->levels[level_index].engineClock;
+ level->coreClock = ps->levels[level_index].engineClock;
if (designation == PHM_PerformanceLevelDesignation_PowerContainment) {
for (i = 1; i < ps->level; i++) {
#include "dce/dce_10_0_sh_mask.h"
#include "pppcielanes.h"
#include "fiji_hwmgr.h"
-#include "tonga_processpptables.h"
-#include "tonga_pptable.h"
+#include "process_pptables_v1_0.h"
+#include "pptable_v1_0.h"
#include "pp_debug.h"
#include "pp_acpi.h"
#include "amd_pcie_helpers.h"
static const unsigned long PhwFiji_Magic = (unsigned long)(PHM_VIslands_Magic);
-struct fiji_power_state *cast_phw_fiji_power_state(
+static struct fiji_power_state *cast_phw_fiji_power_state(
struct pp_hw_power_state *hw_ps)
{
PP_ASSERT_WITH_CODE((PhwFiji_Magic == hw_ps->magic),
return (struct fiji_power_state *)hw_ps;
}
-const struct fiji_power_state *cast_const_phw_fiji_power_state(
+static const struct
+fiji_power_state *cast_const_phw_fiji_power_state(
const struct pp_hw_power_state *hw_ps)
{
PP_ASSERT_WITH_CODE((PhwFiji_Magic == hw_ps->magic),
phm_cap_set(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_TablelessHardwareInterface);
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_SclkDeepSleep);
-
data->gpio_debug = 0;
phm_cap_set(hwmgr->platform_descriptor.platformCaps,
* @param voltage - voltage to look for
* @return 0 on success
*/
-uint8_t fiji_get_voltage_index(
+static uint8_t fiji_get_voltage_index(
struct phm_ppt_v1_voltage_lookup_table *lookup_table, uint16_t voltage)
{
uint8_t count = (uint8_t) (lookup_table->count);
* @return always 0
*/
-int fiji_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr,
+static int fiji_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr,
struct SMU73_Discrete_DpmTable *table)
{
int result;
* @param mclk the MCLK value to be used in the decision if MVDD should be high or low.
* @param voltage the SMC VOLTAGE structure to be populated
*/
-int fiji_populate_mvdd_value(struct pp_hwmgr *hwmgr,
+static int fiji_populate_mvdd_value(struct pp_hwmgr *hwmgr,
uint32_t mclk, SMIO_Pattern *smio_pat)
{
const struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend);
ps = (struct fiji_power_state *)(&state->hardware);
- result = tonga_get_powerplay_table_entry(hwmgr, entry_index, state,
+ result = get_powerplay_table_entry_v1_0(hwmgr, entry_index, state,
fiji_get_pp_table_entry_callback_func);
/* This is the earliest time we have all the dependency table and the VBIOS boot state
return 0;
}
-int fiji_enable_disable_uvd_dpm(struct pp_hwmgr *hwmgr, bool enable)
+static int fiji_enable_disable_uvd_dpm(struct pp_hwmgr *hwmgr, bool enable)
{
return smum_send_msg_to_smc(hwmgr->smumgr, enable ?
(PPSMC_Msg)PPSMC_MSG_UVDDPM_Enable :
PPSMC_MSG_VCEDPM_Disable);
}
-int fiji_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable)
+static int fiji_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable)
{
return smum_send_msg_to_smc(hwmgr->smumgr, enable?
PPSMC_MSG_SAMUDPM_Enable :
PPSMC_MSG_SAMUDPM_Disable);
}
-int fiji_enable_disable_acp_dpm(struct pp_hwmgr *hwmgr, bool enable)
+static int fiji_enable_disable_acp_dpm(struct pp_hwmgr *hwmgr, bool enable)
{
return smum_send_msg_to_smc(hwmgr->smumgr, enable?
PPSMC_MSG_ACPDPM_Enable :
return;
}
-int fiji_upload_dpm_level_enable_mask(struct pp_hwmgr *hwmgr)
+static int fiji_upload_dpm_level_enable_mask(struct pp_hwmgr *hwmgr)
{
int result;
struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend);
return 0;
}
-int fiji_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
+static int fiji_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
{
return fiji_program_display_gap(hwmgr);
}
PPSMC_MSG_SetFanRpmMax, us_max_fan_rpm);
}
-int fiji_dpm_set_interrupt_state(void *private_data,
+static int fiji_dpm_set_interrupt_state(void *private_data,
unsigned src_id, unsigned type,
int enabled)
{
return 0;
}
-int fiji_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr,
+static int fiji_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr,
const void *thermal_interrupt_info)
{
int result;
(pl1->pcie_lane == pl2->pcie_lane));
}
-int fiji_check_states_equal(struct pp_hwmgr *hwmgr, const struct pp_hw_power_state *pstate1, const struct pp_hw_power_state *pstate2, bool *equal)
+static int
+fiji_check_states_equal(struct pp_hwmgr *hwmgr,
+ const struct pp_hw_power_state *pstate1,
+ const struct pp_hw_power_state *pstate2, bool *equal)
{
const struct fiji_power_state *psa = cast_const_phw_fiji_power_state(pstate1);
const struct fiji_power_state *psb = cast_const_phw_fiji_power_state(pstate2);
return 0;
}
-bool fiji_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr)
+static bool
+fiji_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr)
{
struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend);
bool is_update_required = false;
.dynamic_state_management_enable = &fiji_enable_dpm_tasks,
.dynamic_state_management_disable = &fiji_disable_dpm_tasks,
.force_dpm_level = &fiji_dpm_force_dpm_level,
- .get_num_of_pp_table_entries = &tonga_get_number_of_powerplay_table_entries,
+ .get_num_of_pp_table_entries = &get_number_of_powerplay_table_entries_v1_0,
.get_power_state_size = &fiji_get_power_state_size,
.get_pp_table_entry = &fiji_get_pp_table_entry,
.patch_boot_state = &fiji_patch_boot_state,
int fiji_hwmgr_init(struct pp_hwmgr *hwmgr)
{
hwmgr->hwmgr_func = &fiji_hwmgr_funcs;
- hwmgr->pptable_func = &tonga_pptable_funcs;
+ hwmgr->pptable_func = &pptable_v1_0_funcs;
pp_fiji_thermal_initialize(hwmgr);
return 0;
}
fiji_hwmgr->power_tune_defaults = &fiji_power_tune_data_set_array[0];
/* Assume disabled */
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_PowerContainment);
phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_CAC);
phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
fiji_hwmgr->fast_watermark_threshold = 100;
- if (hwmgr->powercontainment_enabled) {
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_PowerContainment);
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
tmp = 1;
fiji_hwmgr->enable_dte_feature = tmp ? false : true;
fiji_hwmgr->enable_tdc_limit_feature = tmp ? true : false;
return 0;
}
-int fiji_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr)
+static int fiji_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr)
{
int result;
* @param Result the last failure code
* @return result from set temperature range routine
*/
-int tf_fiji_thermal_setup_fan_table(struct pp_hwmgr *hwmgr,
+static int tf_fiji_thermal_setup_fan_table(struct pp_hwmgr *hwmgr,
void *input, void *output, void *storage, int result)
{
struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend);
* @param Result the last failure code
* @return result from set temperature range routine
*/
-int tf_fiji_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr,
+static int tf_fiji_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr,
void *input, void *output, void *storage, int result)
{
/* If the fantable setup has failed we could have disabled
#include "hwmgr.h"
#include "hardwaremanager.h"
#include "power_state.h"
-#include "pp_acpi.h"
-#include "amd_acpi.h"
#include "pp_debug.h"
#define PHM_FUNC_CHECK(hw) \
return -EINVAL; \
} while (0)
-void phm_init_dynamic_caps(struct pp_hwmgr *hwmgr)
-{
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableVoltageTransition);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableEngineTransition);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMemoryTransition);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMGClockGating);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMGCGTSSM);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableLSClockGating);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_Force3DClockSupport);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableLightSleep);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMCLS);
- phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisablePowerGating);
-
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableDPM);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableSMUUVDHandshake);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_ThermalAutoThrottling);
-
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest);
-
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_NoOD5Support);
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_UserMaxClockForMultiDisplays);
-
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_VpuRecoveryInProgress);
-
- phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_UVDDPM);
- phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_VCEDPM);
-
- if (acpi_atcs_functions_supported(hwmgr->device, ATCS_FUNCTION_PCIE_PERFORMANCE_REQUEST) &&
- acpi_atcs_functions_supported(hwmgr->device, ATCS_FUNCTION_PCIE_DEVICE_READY_NOTIFICATION))
- phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest);
-}
-
bool phm_is_hw_access_blocked(struct pp_hwmgr *hwmgr)
{
return hwmgr->block_hw_access;
#include "pp_debug.h"
#include "ppatomctrl.h"
#include "ppsmc.h"
-
-#define VOLTAGE_SCALE 4
+#include "pp_acpi.h"
+#include "amd_acpi.h"
extern int cz_hwmgr_init(struct pp_hwmgr *hwmgr);
extern int tonga_hwmgr_init(struct pp_hwmgr *hwmgr);
extern int fiji_hwmgr_init(struct pp_hwmgr *hwmgr);
extern int polaris10_hwmgr_init(struct pp_hwmgr *hwmgr);
+extern int iceland_hwmgr_init(struct pp_hwmgr *hwmgr);
+
+static void hwmgr_init_default_caps(struct pp_hwmgr *hwmgr);
+static int hwmgr_set_user_specify_caps(struct pp_hwmgr *hwmgr);
+
+uint8_t convert_to_vid(uint16_t vddc)
+{
+ return (uint8_t) ((6200 - (vddc * VOLTAGE_SCALE)) / 25);
+}
int hwmgr_init(struct amd_pp_init *pp_init, struct pp_instance *handle)
{
hwmgr->device = pp_init->device;
hwmgr->chip_family = pp_init->chip_family;
hwmgr->chip_id = pp_init->chip_id;
- hwmgr->hw_revision = pp_init->rev_id;
hwmgr->usec_timeout = AMD_MAX_USEC_TIMEOUT;
hwmgr->power_source = PP_PowerSource_AC;
- hwmgr->powercontainment_enabled = pp_init->powercontainment_enabled;
+ hwmgr->pp_table_version = PP_TABLE_V1;
+
+ hwmgr_init_default_caps(hwmgr);
+ hwmgr_set_user_specify_caps(hwmgr);
switch (hwmgr->chip_family) {
case AMDGPU_FAMILY_CZ:
break;
case AMDGPU_FAMILY_VI:
switch (hwmgr->chip_id) {
+ case CHIP_TOPAZ:
+ iceland_hwmgr_init(hwmgr);
+ break;
case CHIP_TONGA:
tonga_hwmgr_init(hwmgr);
break;
return -EINVAL;
}
- phm_init_dynamic_caps(hwmgr);
-
return 0;
}
kfree(hwmgr->set_temperature_range.function_list);
kfree(hwmgr->ps);
+ kfree(hwmgr->current_ps);
+ kfree(hwmgr->request_ps);
kfree(hwmgr);
return 0;
}
sizeof(struct pp_power_state);
hwmgr->ps = kzalloc(size * table_entries, GFP_KERNEL);
-
if (hwmgr->ps == NULL)
return -ENOMEM;
+ hwmgr->request_ps = kzalloc(size, GFP_KERNEL);
+ if (hwmgr->request_ps == NULL)
+ return -ENOMEM;
+
+ hwmgr->current_ps = kzalloc(size, GFP_KERNEL);
+ if (hwmgr->current_ps == NULL)
+ return -ENOMEM;
+
state = hwmgr->ps;
for (i = 0; i < table_entries; i++) {
if (state->classification.flags & PP_StateClassificationFlag_Boot) {
hwmgr->boot_ps = state;
- hwmgr->current_ps = hwmgr->request_ps = state;
+ memcpy(hwmgr->current_ps, state, size);
+ memcpy(hwmgr->request_ps, state, size);
}
state->id = i + 1; /* assigned unique num for every power state id */
state = (struct pp_power_state *)((unsigned long)state + size);
}
+
return 0;
}
return 0;
}
-int phm_wait_for_register_unequal(struct pp_hwmgr *hwmgr,
- uint32_t index, uint32_t value, uint32_t mask)
-{
- uint32_t i;
- uint32_t cur_value;
-
- if (hwmgr == NULL || hwmgr->device == NULL) {
- printk(KERN_ERR "[ powerplay ] Invalid Hardware Manager!");
- return -EINVAL;
- }
-
- for (i = 0; i < hwmgr->usec_timeout; i++) {
- cur_value = cgs_read_register(hwmgr->device, index);
- if ((cur_value & mask) != (value & mask))
- break;
- udelay(1);
- }
-
- /* timeout means wrong logic*/
- if (i == hwmgr->usec_timeout)
- return -1;
- return 0;
-}
-
/**
* Returns once the part of the register indicated by the mask has
phm_wait_on_register(hwmgr, indirect_port + 1, mask, value);
}
-void phm_wait_for_indirect_register_unequal(struct pp_hwmgr *hwmgr,
- uint32_t indirect_port,
- uint32_t index,
- uint32_t value,
- uint32_t mask)
-{
- if (hwmgr == NULL || hwmgr->device == NULL) {
- printk(KERN_ERR "[ powerplay ] Invalid Hardware Manager!");
- return;
- }
- cgs_write_register(hwmgr->device, indirect_port, index);
- phm_wait_for_register_unequal(hwmgr, indirect_port + 1,
- value, mask);
-}
bool phm_cf_want_uvd_power_gating(struct pp_hwmgr *hwmgr)
{
return i - 1;
}
+uint8_t phm_get_voltage_id(pp_atomctrl_voltage_table *voltage_table,
+ uint32_t voltage)
+{
+ uint8_t count = (uint8_t) (voltage_table->count);
+ uint8_t i = 0;
+
+ PP_ASSERT_WITH_CODE((NULL != voltage_table),
+ "Voltage Table empty.", return 0;);
+ PP_ASSERT_WITH_CODE((0 != count),
+ "Voltage Table empty.", return 0;);
+
+ for (i = 0; i < count; i++) {
+ /* find first voltage bigger than requested */
+ if (voltage_table->entries[i].value >= voltage)
+ return i;
+ }
+
+ /* voltage is bigger than max voltage in the table */
+ return i - 1;
+}
+
uint16_t phm_find_closest_vddci(struct pp_atomctrl_voltage_table *vddci_table, uint16_t vddci)
{
uint32_t i;
table_clk_vlt->entries[2].v = 810;
table_clk_vlt->entries[3].clk = PP_DAL_POWERLEVEL_PERFORMANCE;
table_clk_vlt->entries[3].v = 900;
- pptable_info->vddc_dep_on_dal_pwrl = table_clk_vlt;
+ if (pptable_info != NULL)
+ pptable_info->vddc_dep_on_dal_pwrl = table_clk_vlt;
hwmgr->dyn_state.vddc_dep_on_dal_pwrl = table_clk_vlt;
}
printk(KERN_ERR "DAL requested level can not"
" found a available voltage in VDDC DPM Table \n");
}
+
+void hwmgr_init_default_caps(struct pp_hwmgr *hwmgr)
+{
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableVoltageTransition);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableEngineTransition);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMemoryTransition);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMGClockGating);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMGCGTSSM);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableLSClockGating);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_Force3DClockSupport);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableLightSleep);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableMCLS);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisablePowerGating);
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableDPM);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_DisableSMUUVDHandshake);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_ThermalAutoThrottling);
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest);
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_NoOD5Support);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_UserMaxClockForMultiDisplays);
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_VpuRecoveryInProgress);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_UVDDPM);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_VCEDPM);
+
+ if (acpi_atcs_functions_supported(hwmgr->device, ATCS_FUNCTION_PCIE_PERFORMANCE_REQUEST) &&
+ acpi_atcs_functions_supported(hwmgr->device, ATCS_FUNCTION_PCIE_DEVICE_READY_NOTIFICATION))
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DynamicPatchPowerState);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EnableSMU7ThermalManagement);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DynamicPowerManagement);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SMC);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DynamicUVDState);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_FanSpeedInTableIsRPM);
+
+ return;
+}
+
+int hwmgr_set_user_specify_caps(struct pp_hwmgr *hwmgr)
+{
+ if (amdgpu_sclk_deep_sleep_en)
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SclkDeepSleep);
+ else
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SclkDeepSleep);
+
+ if (amdgpu_powercontainment)
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment);
+ else
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment);
+
+ hwmgr->feature_mask = amdgpu_pp_feature_mask;
+
+ return 0;
+}
+
+int phm_get_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
+ uint32_t sclk, uint16_t id, uint16_t *voltage)
+{
+ uint32_t vol;
+ int ret = 0;
+
+ if (hwmgr->chip_id < CHIP_POLARIS10) {
+ atomctrl_get_voltage_evv_on_sclk(hwmgr, voltage_type, sclk, id, voltage);
+ if (*voltage >= 2000 || *voltage == 0)
+ *voltage = 1150;
+ } else {
+ ret = atomctrl_get_voltage_evv_on_sclk_ai(hwmgr, voltage_type, sclk, id, &vol);
+ *voltage = (uint16_t)vol/100;
+ }
+ return ret;
+}
+
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+
+#include "hwmgr.h"
+#include "iceland_clockpowergating.h"
+#include "ppsmc.h"
+#include "iceland_hwmgr.h"
+
+int iceland_phm_powerdown_uvd(struct pp_hwmgr *hwmgr)
+{
+ /* iceland does not have MM hardware block */
+ return 0;
+}
+
+static int iceland_phm_powerup_uvd(struct pp_hwmgr *hwmgr)
+{
+ /* iceland does not have MM hardware block */
+ return 0;
+}
+
+static int iceland_phm_powerdown_vce(struct pp_hwmgr *hwmgr)
+{
+ /* iceland does not have MM hardware block */
+ return 0;
+}
+
+static int iceland_phm_powerup_vce(struct pp_hwmgr *hwmgr)
+{
+ /* iceland does not have MM hardware block */
+ return 0;
+}
+
+int iceland_phm_set_asic_block_gating(struct pp_hwmgr *hwmgr, enum
+ PHM_AsicBlock block, enum PHM_ClockGateSetting gating)
+{
+ int ret = 0;
+
+ switch (block) {
+ case PHM_AsicBlock_UVD_MVC:
+ case PHM_AsicBlock_UVD:
+ case PHM_AsicBlock_UVD_HD:
+ case PHM_AsicBlock_UVD_SD:
+ if (gating == PHM_ClockGateSetting_StaticOff)
+ ret = iceland_phm_powerdown_uvd(hwmgr);
+ else
+ ret = iceland_phm_powerup_uvd(hwmgr);
+ break;
+ case PHM_AsicBlock_GFX:
+ default:
+ break;
+ }
+
+ return ret;
+}
+
+int iceland_phm_disable_clock_power_gating(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ data->uvd_power_gated = false;
+ data->vce_power_gated = false;
+
+ iceland_phm_powerup_uvd(hwmgr);
+ iceland_phm_powerup_vce(hwmgr);
+
+ return 0;
+}
+
+int iceland_phm_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate)
+{
+ if (bgate) {
+ iceland_update_uvd_dpm(hwmgr, true);
+ iceland_phm_powerdown_uvd(hwmgr);
+ } else {
+ iceland_phm_powerup_uvd(hwmgr);
+ iceland_update_uvd_dpm(hwmgr, false);
+ }
+
+ return 0;
+}
+
+int iceland_phm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
+{
+ if (bgate)
+ return iceland_phm_powerdown_vce(hwmgr);
+ else
+ return iceland_phm_powerup_vce(hwmgr);
+
+ return 0;
+}
+
+int iceland_phm_update_clock_gatings(struct pp_hwmgr *hwmgr,
+ const uint32_t *msg_id)
+{
+ /* iceland does not have MM hardware block */
+ return 0;
+}
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+
+#ifndef _ICELAND_CLOCK_POWER_GATING_H_
+#define _ICELAND_CLOCK_POWER_GATING_H_
+
+#include "iceland_hwmgr.h"
+#include "pp_asicblocks.h"
+
+extern int iceland_phm_set_asic_block_gating(struct pp_hwmgr *hwmgr, enum PHM_AsicBlock block, enum PHM_ClockGateSetting gating);
+extern int iceland_phm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate);
+extern int iceland_phm_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate);
+extern int iceland_phm_powerdown_uvd(struct pp_hwmgr *hwmgr);
+extern int iceland_phm_disable_clock_power_gating(struct pp_hwmgr *hwmgr);
+extern int iceland_phm_update_clock_gatings(struct pp_hwmgr *hwmgr, const uint32_t *msg_id);
+#endif /* _ICELAND_CLOCK_POWER_GATING_H_ */
--- /dev/null
+#ifndef ICELAND_DYN_DEFAULTS_H
+#define ICELAND_DYN_DEFAULTS_H
+
+enum ICELANDdpm_TrendDetection
+{
+ ICELANDdpm_TrendDetection_AUTO,
+ ICELANDdpm_TrendDetection_UP,
+ ICELANDdpm_TrendDetection_DOWN
+};
+typedef enum ICELANDdpm_TrendDetection ICELANDdpm_TrendDetection;
+
+
+#define PPICELAND_VOTINGRIGHTSCLIENTS_DFLT0 0x3FFFC102
+#define PPICELAND_VOTINGRIGHTSCLIENTS_DFLT1 0x000400
+#define PPICELAND_VOTINGRIGHTSCLIENTS_DFLT2 0xC00080
+#define PPICELAND_VOTINGRIGHTSCLIENTS_DFLT3 0xC00200
+#define PPICELAND_VOTINGRIGHTSCLIENTS_DFLT4 0xC01680
+#define PPICELAND_VOTINGRIGHTSCLIENTS_DFLT5 0xC00033
+#define PPICELAND_VOTINGRIGHTSCLIENTS_DFLT6 0xC00033
+#define PPICELAND_VOTINGRIGHTSCLIENTS_DFLT7 0x3FFFC000
+
+
+#define PPICELAND_THERMALPROTECTCOUNTER_DFLT 0x200
+
+#define PPICELAND_STATICSCREENTHRESHOLDUNIT_DFLT 0
+
+#define PPICELAND_STATICSCREENTHRESHOLD_DFLT 0x00C8
+
+#define PPICELAND_GFXIDLECLOCKSTOPTHRESHOLD_DFLT 0x200
+
+#define PPICELAND_REFERENCEDIVIDER_DFLT 4
+
+#define PPICELAND_ULVVOLTAGECHANGEDELAY_DFLT 1687
+
+#define PPICELAND_CGULVPARAMETER_DFLT 0x00040035
+#define PPICELAND_CGULVCONTROL_DFLT 0x00007450
+#define PPICELAND_TARGETACTIVITY_DFLT 30
+#define PPICELAND_MCLK_TARGETACTIVITY_DFLT 10
+
+#endif
+
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/fb.h>
+#include "linux/delay.h"
+#include "pp_acpi.h"
+#include "hwmgr.h"
+#include <atombios.h>
+#include "iceland_hwmgr.h"
+#include "pptable.h"
+#include "processpptables.h"
+#include "pp_debug.h"
+#include "ppsmc.h"
+#include "cgs_common.h"
+#include "pppcielanes.h"
+#include "iceland_dyn_defaults.h"
+#include "smumgr.h"
+#include "iceland_smumgr.h"
+#include "iceland_clockpowergating.h"
+#include "iceland_thermal.h"
+#include "iceland_powertune.h"
+
+#include "gmc/gmc_8_1_d.h"
+#include "gmc/gmc_8_1_sh_mask.h"
+
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+
+#include "smu/smu_7_1_1_d.h"
+#include "smu/smu_7_1_1_sh_mask.h"
+
+#include "cgs_linux.h"
+#include "eventmgr.h"
+#include "amd_pcie_helpers.h"
+
+#define MC_CG_ARB_FREQ_F0 0x0a
+#define MC_CG_ARB_FREQ_F1 0x0b
+#define MC_CG_ARB_FREQ_F2 0x0c
+#define MC_CG_ARB_FREQ_F3 0x0d
+
+#define MC_CG_SEQ_DRAMCONF_S0 0x05
+#define MC_CG_SEQ_DRAMCONF_S1 0x06
+#define MC_CG_SEQ_YCLK_SUSPEND 0x04
+#define MC_CG_SEQ_YCLK_RESUME 0x0a
+
+#define PCIE_BUS_CLK 10000
+#define TCLK (PCIE_BUS_CLK / 10)
+
+#define SMC_RAM_END 0x40000
+#define SMC_CG_IND_START 0xc0030000
+#define SMC_CG_IND_END 0xc0040000 /* First byte after SMC_CG_IND*/
+
+#define VOLTAGE_SCALE 4
+#define VOLTAGE_VID_OFFSET_SCALE1 625
+#define VOLTAGE_VID_OFFSET_SCALE2 100
+
+const uint32_t iceland_magic = (uint32_t)(PHM_VIslands_Magic);
+
+#define MC_SEQ_MISC0_GDDR5_SHIFT 28
+#define MC_SEQ_MISC0_GDDR5_MASK 0xf0000000
+#define MC_SEQ_MISC0_GDDR5_VALUE 5
+
+/** Values for the CG_THERMAL_CTRL::DPM_EVENT_SRC field. */
+enum DPM_EVENT_SRC {
+ DPM_EVENT_SRC_ANALOG = 0, /* Internal analog trip point */
+ DPM_EVENT_SRC_EXTERNAL = 1, /* External (GPIO 17) signal */
+ DPM_EVENT_SRC_DIGITAL = 2, /* Internal digital trip point (DIG_THERM_DPM) */
+ DPM_EVENT_SRC_ANALOG_OR_EXTERNAL = 3, /* Internal analog or external */
+ DPM_EVENT_SRC_DIGITAL_OR_EXTERNAL = 4 /* Internal digital or external */
+};
+
+static int iceland_read_clock_registers(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ data->clock_registers.vCG_SPLL_FUNC_CNTL =
+ cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_SPLL_FUNC_CNTL);
+ data->clock_registers.vCG_SPLL_FUNC_CNTL_2 =
+ cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_SPLL_FUNC_CNTL_2);
+ data->clock_registers.vCG_SPLL_FUNC_CNTL_3 =
+ cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_SPLL_FUNC_CNTL_3);
+ data->clock_registers.vCG_SPLL_FUNC_CNTL_4 =
+ cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_SPLL_FUNC_CNTL_4);
+ data->clock_registers.vCG_SPLL_SPREAD_SPECTRUM =
+ cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_SPLL_SPREAD_SPECTRUM);
+ data->clock_registers.vCG_SPLL_SPREAD_SPECTRUM_2 =
+ cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_SPLL_SPREAD_SPECTRUM_2);
+ data->clock_registers.vDLL_CNTL =
+ cgs_read_register(hwmgr->device, mmDLL_CNTL);
+ data->clock_registers.vMCLK_PWRMGT_CNTL =
+ cgs_read_register(hwmgr->device, mmMCLK_PWRMGT_CNTL);
+ data->clock_registers.vMPLL_AD_FUNC_CNTL =
+ cgs_read_register(hwmgr->device, mmMPLL_AD_FUNC_CNTL);
+ data->clock_registers.vMPLL_DQ_FUNC_CNTL =
+ cgs_read_register(hwmgr->device, mmMPLL_DQ_FUNC_CNTL);
+ data->clock_registers.vMPLL_FUNC_CNTL =
+ cgs_read_register(hwmgr->device, mmMPLL_FUNC_CNTL);
+ data->clock_registers.vMPLL_FUNC_CNTL_1 =
+ cgs_read_register(hwmgr->device, mmMPLL_FUNC_CNTL_1);
+ data->clock_registers.vMPLL_FUNC_CNTL_2 =
+ cgs_read_register(hwmgr->device, mmMPLL_FUNC_CNTL_2);
+ data->clock_registers.vMPLL_SS1 =
+ cgs_read_register(hwmgr->device, mmMPLL_SS1);
+ data->clock_registers.vMPLL_SS2 =
+ cgs_read_register(hwmgr->device, mmMPLL_SS2);
+
+ return 0;
+}
+
+/**
+ * Find out if memory is GDDR5.
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+int iceland_get_memory_type(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ uint32_t temp;
+
+ temp = cgs_read_register(hwmgr->device, mmMC_SEQ_MISC0);
+
+ data->is_memory_GDDR5 = (MC_SEQ_MISC0_GDDR5_VALUE ==
+ ((temp & MC_SEQ_MISC0_GDDR5_MASK) >>
+ MC_SEQ_MISC0_GDDR5_SHIFT));
+
+ return 0;
+}
+
+int iceland_update_uvd_dpm(struct pp_hwmgr *hwmgr, bool bgate)
+{
+ /* iceland does not have MM hardware blocks */
+ return 0;
+}
+
+/**
+ * Enables Dynamic Power Management by SMC
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+int iceland_enable_acpi_power_management(struct pp_hwmgr *hwmgr)
+{
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT, STATIC_PM_EN, 1);
+
+ return 0;
+}
+
+/**
+ * Find the MC microcode version and store it in the HwMgr struct
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+int iceland_get_mc_microcode_version(struct pp_hwmgr *hwmgr)
+{
+ cgs_write_register(hwmgr->device, mmMC_SEQ_IO_DEBUG_INDEX, 0x9F);
+
+ hwmgr->microcode_version_info.MC = cgs_read_register(hwmgr->device, mmMC_SEQ_IO_DEBUG_DATA);
+
+ return 0;
+}
+
+static int iceland_init_sclk_threshold(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ data->low_sclk_interrupt_threshold = 0;
+
+ return 0;
+}
+
+
+static int iceland_setup_asic_task(struct pp_hwmgr *hwmgr)
+{
+ int tmp_result, result = 0;
+
+ tmp_result = iceland_read_clock_registers(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to read clock registers!", result = tmp_result);
+
+ tmp_result = iceland_get_memory_type(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to get memory type!", result = tmp_result);
+
+ tmp_result = iceland_enable_acpi_power_management(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to enable ACPI power management!", result = tmp_result);
+
+ tmp_result = iceland_get_mc_microcode_version(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to get MC microcode version!", result = tmp_result);
+
+ tmp_result = iceland_init_sclk_threshold(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to init sclk threshold!", result = tmp_result);
+
+ return result;
+}
+
+static bool cf_iceland_voltage_control(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ return ICELAND_VOLTAGE_CONTROL_NONE != data->voltage_control;
+}
+
+/*
+ * -------------- Voltage Tables ----------------------
+ * If the voltage table would be bigger than what will fit into the
+ * state table on the SMC keep only the higher entries.
+ */
+
+static void iceland_trim_voltage_table_to_fit_state_table(
+ struct pp_hwmgr *hwmgr,
+ uint32_t max_voltage_steps,
+ pp_atomctrl_voltage_table *voltage_table)
+{
+ unsigned int i, diff;
+
+ if (voltage_table->count <= max_voltage_steps) {
+ return;
+ }
+
+ diff = voltage_table->count - max_voltage_steps;
+
+ for (i = 0; i < max_voltage_steps; i++) {
+ voltage_table->entries[i] = voltage_table->entries[i + diff];
+ }
+
+ voltage_table->count = max_voltage_steps;
+
+ return;
+}
+
+/**
+ * Enable voltage control
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+int iceland_enable_voltage_control(struct pp_hwmgr *hwmgr)
+{
+ /* enable voltage control */
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT, VOLT_PWRMGT_EN, 1);
+
+ return 0;
+}
+
+static int iceland_get_svi2_voltage_table(struct pp_hwmgr *hwmgr,
+ struct phm_clock_voltage_dependency_table *voltage_dependency_table,
+ pp_atomctrl_voltage_table *voltage_table)
+{
+ uint32_t i;
+
+ PP_ASSERT_WITH_CODE((NULL != voltage_table),
+ "Voltage Dependency Table empty.", return -EINVAL;);
+
+ voltage_table->mask_low = 0;
+ voltage_table->phase_delay = 0;
+ voltage_table->count = voltage_dependency_table->count;
+
+ for (i = 0; i < voltage_dependency_table->count; i++) {
+ voltage_table->entries[i].value =
+ voltage_dependency_table->entries[i].v;
+ voltage_table->entries[i].smio_low = 0;
+ }
+
+ return 0;
+}
+
+/**
+ * Create Voltage Tables.
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+int iceland_construct_voltage_tables(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ int result;
+
+ /* GPIO voltage */
+ if (ICELAND_VOLTAGE_CONTROL_BY_GPIO == data->voltage_control) {
+ result = atomctrl_get_voltage_table_v3(hwmgr,
+ VOLTAGE_TYPE_VDDC, VOLTAGE_OBJ_GPIO_LUT,
+ &data->vddc_voltage_table);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Failed to retrieve VDDC table.", return result;);
+ } else if (ICELAND_VOLTAGE_CONTROL_BY_SVID2 == data->voltage_control) {
+ /* SVI2 VDDC voltage */
+ result = iceland_get_svi2_voltage_table(hwmgr,
+ hwmgr->dyn_state.vddc_dependency_on_mclk,
+ &data->vddc_voltage_table);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Failed to retrieve SVI2 VDDC table from dependancy table.", return result;);
+ }
+
+ PP_ASSERT_WITH_CODE(
+ (data->vddc_voltage_table.count <= (SMU71_MAX_LEVELS_VDDC)),
+ "Too many voltage values for VDDC. Trimming to fit state table.",
+ iceland_trim_voltage_table_to_fit_state_table(hwmgr,
+ SMU71_MAX_LEVELS_VDDC, &(data->vddc_voltage_table));
+ );
+
+ /* GPIO */
+ if (ICELAND_VOLTAGE_CONTROL_BY_GPIO == data->vdd_ci_control) {
+ result = atomctrl_get_voltage_table_v3(hwmgr,
+ VOLTAGE_TYPE_VDDCI, VOLTAGE_OBJ_GPIO_LUT, &(data->vddci_voltage_table));
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Failed to retrieve VDDCI table.", return result;);
+ }
+
+ /* SVI2 VDDCI voltage */
+ if (ICELAND_VOLTAGE_CONTROL_BY_SVID2 == data->vdd_ci_control) {
+ result = iceland_get_svi2_voltage_table(hwmgr,
+ hwmgr->dyn_state.vddci_dependency_on_mclk,
+ &data->vddci_voltage_table);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Failed to retrieve SVI2 VDDCI table from dependancy table.", return result;);
+ }
+
+ PP_ASSERT_WITH_CODE(
+ (data->vddci_voltage_table.count <= (SMU71_MAX_LEVELS_VDDCI)),
+ "Too many voltage values for VDDCI. Trimming to fit state table.",
+ iceland_trim_voltage_table_to_fit_state_table(hwmgr,
+ SMU71_MAX_LEVELS_VDDCI, &(data->vddci_voltage_table));
+ );
+
+
+ /* GPIO */
+ if (ICELAND_VOLTAGE_CONTROL_BY_GPIO == data->mvdd_control) {
+ result = atomctrl_get_voltage_table_v3(hwmgr,
+ VOLTAGE_TYPE_MVDDC, VOLTAGE_OBJ_GPIO_LUT, &(data->mvdd_voltage_table));
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Failed to retrieve table.", return result;);
+ }
+
+ /* SVI2 voltage control */
+ if (ICELAND_VOLTAGE_CONTROL_BY_SVID2 == data->mvdd_control) {
+ result = iceland_get_svi2_voltage_table(hwmgr,
+ hwmgr->dyn_state.mvdd_dependency_on_mclk,
+ &data->mvdd_voltage_table);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Failed to retrieve SVI2 MVDD table from dependancy table.", return result;);
+ }
+
+ PP_ASSERT_WITH_CODE(
+ (data->mvdd_voltage_table.count <= (SMU71_MAX_LEVELS_MVDD)),
+ "Too many voltage values for MVDD. Trimming to fit state table.",
+ iceland_trim_voltage_table_to_fit_state_table(hwmgr,
+ SMU71_MAX_LEVELS_MVDD, &(data->mvdd_voltage_table));
+ );
+
+ return 0;
+}
+
+/*---------------------------MC----------------------------*/
+
+uint8_t iceland_get_memory_module_index(struct pp_hwmgr *hwmgr)
+{
+ return (uint8_t) (0xFF & (cgs_read_register(hwmgr->device, mmBIOS_SCRATCH_4) >> 16));
+}
+
+bool iceland_check_s0_mc_reg_index(uint16_t inReg, uint16_t *outReg)
+{
+ bool result = true;
+
+ switch (inReg) {
+ case mmMC_SEQ_RAS_TIMING:
+ *outReg = mmMC_SEQ_RAS_TIMING_LP;
+ break;
+
+ case mmMC_SEQ_DLL_STBY:
+ *outReg = mmMC_SEQ_DLL_STBY_LP;
+ break;
+
+ case mmMC_SEQ_G5PDX_CMD0:
+ *outReg = mmMC_SEQ_G5PDX_CMD0_LP;
+ break;
+
+ case mmMC_SEQ_G5PDX_CMD1:
+ *outReg = mmMC_SEQ_G5PDX_CMD1_LP;
+ break;
+
+ case mmMC_SEQ_G5PDX_CTRL:
+ *outReg = mmMC_SEQ_G5PDX_CTRL_LP;
+ break;
+
+ case mmMC_SEQ_CAS_TIMING:
+ *outReg = mmMC_SEQ_CAS_TIMING_LP;
+ break;
+
+ case mmMC_SEQ_MISC_TIMING:
+ *outReg = mmMC_SEQ_MISC_TIMING_LP;
+ break;
+
+ case mmMC_SEQ_MISC_TIMING2:
+ *outReg = mmMC_SEQ_MISC_TIMING2_LP;
+ break;
+
+ case mmMC_SEQ_PMG_DVS_CMD:
+ *outReg = mmMC_SEQ_PMG_DVS_CMD_LP;
+ break;
+
+ case mmMC_SEQ_PMG_DVS_CTL:
+ *outReg = mmMC_SEQ_PMG_DVS_CTL_LP;
+ break;
+
+ case mmMC_SEQ_RD_CTL_D0:
+ *outReg = mmMC_SEQ_RD_CTL_D0_LP;
+ break;
+
+ case mmMC_SEQ_RD_CTL_D1:
+ *outReg = mmMC_SEQ_RD_CTL_D1_LP;
+ break;
+
+ case mmMC_SEQ_WR_CTL_D0:
+ *outReg = mmMC_SEQ_WR_CTL_D0_LP;
+ break;
+
+ case mmMC_SEQ_WR_CTL_D1:
+ *outReg = mmMC_SEQ_WR_CTL_D1_LP;
+ break;
+
+ case mmMC_PMG_CMD_EMRS:
+ *outReg = mmMC_SEQ_PMG_CMD_EMRS_LP;
+ break;
+
+ case mmMC_PMG_CMD_MRS:
+ *outReg = mmMC_SEQ_PMG_CMD_MRS_LP;
+ break;
+
+ case mmMC_PMG_CMD_MRS1:
+ *outReg = mmMC_SEQ_PMG_CMD_MRS1_LP;
+ break;
+
+ case mmMC_SEQ_PMG_TIMING:
+ *outReg = mmMC_SEQ_PMG_TIMING_LP;
+ break;
+
+ case mmMC_PMG_CMD_MRS2:
+ *outReg = mmMC_SEQ_PMG_CMD_MRS2_LP;
+ break;
+
+ case mmMC_SEQ_WR_CTL_2:
+ *outReg = mmMC_SEQ_WR_CTL_2_LP;
+ break;
+
+ default:
+ result = false;
+ break;
+ }
+
+ return result;
+}
+
+int iceland_set_s0_mc_reg_index(phw_iceland_mc_reg_table *table)
+{
+ uint32_t i;
+ uint16_t address;
+
+ for (i = 0; i < table->last; i++) {
+ table->mc_reg_address[i].s0 =
+ iceland_check_s0_mc_reg_index(table->mc_reg_address[i].s1, &address)
+ ? address : table->mc_reg_address[i].s1;
+ }
+ return 0;
+}
+
+int iceland_copy_vbios_smc_reg_table(const pp_atomctrl_mc_reg_table *table, phw_iceland_mc_reg_table *ni_table)
+{
+ uint8_t i, j;
+
+ PP_ASSERT_WITH_CODE((table->last <= SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE),
+ "Invalid VramInfo table.", return -1);
+ PP_ASSERT_WITH_CODE((table->num_entries <= MAX_AC_TIMING_ENTRIES),
+ "Invalid VramInfo table.", return -1);
+
+ for (i = 0; i < table->last; i++) {
+ ni_table->mc_reg_address[i].s1 = table->mc_reg_address[i].s1;
+ }
+ ni_table->last = table->last;
+
+ for (i = 0; i < table->num_entries; i++) {
+ ni_table->mc_reg_table_entry[i].mclk_max =
+ table->mc_reg_table_entry[i].mclk_max;
+ for (j = 0; j < table->last; j++) {
+ ni_table->mc_reg_table_entry[i].mc_data[j] =
+ table->mc_reg_table_entry[i].mc_data[j];
+ }
+ }
+
+ ni_table->num_entries = table->num_entries;
+
+ return 0;
+}
+
+/**
+ * VBIOS omits some information to reduce size, we need to recover them here.
+ * 1. when we see mmMC_SEQ_MISC1, bit[31:16] EMRS1, need to be write to mmMC_PMG_CMD_EMRS /_LP[15:0].
+ * Bit[15:0] MRS, need to be update mmMC_PMG_CMD_MRS/_LP[15:0]
+ * 2. when we see mmMC_SEQ_RESERVE_M, bit[15:0] EMRS2, need to be write to mmMC_PMG_CMD_MRS1/_LP[15:0].
+ * 3. need to set these data for each clock range
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @param table the address of MCRegTable
+ * @return always 0
+ */
+static int iceland_set_mc_special_registers(struct pp_hwmgr *hwmgr, phw_iceland_mc_reg_table *table)
+{
+ uint8_t i, j, k;
+ uint32_t temp_reg;
+ const iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ for (i = 0, j = table->last; i < table->last; i++) {
+ PP_ASSERT_WITH_CODE((j < SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE),
+ "Invalid VramInfo table.", return -1);
+ switch (table->mc_reg_address[i].s1) {
+ /*
+ * mmMC_SEQ_MISC1, bit[31:16] EMRS1, need to be write
+ * to mmMC_PMG_CMD_EMRS/_LP[15:0]. Bit[15:0] MRS, need
+ * to be update mmMC_PMG_CMD_MRS/_LP[15:0]
+ */
+ case mmMC_SEQ_MISC1:
+ temp_reg = cgs_read_register(hwmgr->device, mmMC_PMG_CMD_EMRS);
+ table->mc_reg_address[j].s1 = mmMC_PMG_CMD_EMRS;
+ table->mc_reg_address[j].s0 = mmMC_SEQ_PMG_CMD_EMRS_LP;
+ for (k = 0; k < table->num_entries; k++) {
+ table->mc_reg_table_entry[k].mc_data[j] =
+ ((temp_reg & 0xffff0000)) |
+ ((table->mc_reg_table_entry[k].mc_data[i] & 0xffff0000) >> 16);
+ }
+ j++;
+ PP_ASSERT_WITH_CODE((j < SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE),
+ "Invalid VramInfo table.", return -1);
+
+ temp_reg = cgs_read_register(hwmgr->device, mmMC_PMG_CMD_MRS);
+ table->mc_reg_address[j].s1 = mmMC_PMG_CMD_MRS;
+ table->mc_reg_address[j].s0 = mmMC_SEQ_PMG_CMD_MRS_LP;
+ for (k = 0; k < table->num_entries; k++) {
+ table->mc_reg_table_entry[k].mc_data[j] =
+ (temp_reg & 0xffff0000) |
+ (table->mc_reg_table_entry[k].mc_data[i] & 0x0000ffff);
+
+ if (!data->is_memory_GDDR5) {
+ table->mc_reg_table_entry[k].mc_data[j] |= 0x100;
+ }
+ }
+ j++;
+ PP_ASSERT_WITH_CODE((j <= SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE),
+ "Invalid VramInfo table.", return -1);
+
+ if (!data->is_memory_GDDR5) {
+ table->mc_reg_address[j].s1 = mmMC_PMG_AUTO_CMD;
+ table->mc_reg_address[j].s0 = mmMC_PMG_AUTO_CMD;
+ for (k = 0; k < table->num_entries; k++) {
+ table->mc_reg_table_entry[k].mc_data[j] =
+ (table->mc_reg_table_entry[k].mc_data[i] & 0xffff0000) >> 16;
+ }
+ j++;
+ PP_ASSERT_WITH_CODE((j <= SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE),
+ "Invalid VramInfo table.", return -1);
+ }
+
+ break;
+
+ case mmMC_SEQ_RESERVE_M:
+ temp_reg = cgs_read_register(hwmgr->device, mmMC_PMG_CMD_MRS1);
+ table->mc_reg_address[j].s1 = mmMC_PMG_CMD_MRS1;
+ table->mc_reg_address[j].s0 = mmMC_SEQ_PMG_CMD_MRS1_LP;
+ for (k = 0; k < table->num_entries; k++) {
+ table->mc_reg_table_entry[k].mc_data[j] =
+ (temp_reg & 0xffff0000) |
+ (table->mc_reg_table_entry[k].mc_data[i] & 0x0000ffff);
+ }
+ j++;
+ PP_ASSERT_WITH_CODE((j <= SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE),
+ "Invalid VramInfo table.", return -1);
+ break;
+
+ default:
+ break;
+ }
+
+ }
+
+ table->last = j;
+
+ return 0;
+}
+
+
+static int iceland_set_valid_flag(phw_iceland_mc_reg_table *table)
+{
+ uint8_t i, j;
+ for (i = 0; i < table->last; i++) {
+ for (j = 1; j < table->num_entries; j++) {
+ if (table->mc_reg_table_entry[j-1].mc_data[i] !=
+ table->mc_reg_table_entry[j].mc_data[i]) {
+ table->validflag |= (1<<i);
+ break;
+ }
+ }
+ }
+
+ return 0;
+}
+
+static int iceland_initialize_mc_reg_table(struct pp_hwmgr *hwmgr)
+{
+ int result;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ pp_atomctrl_mc_reg_table *table;
+ phw_iceland_mc_reg_table *ni_table = &data->iceland_mc_reg_table;
+ uint8_t module_index = iceland_get_memory_module_index(hwmgr);
+
+ table = kzalloc(sizeof(pp_atomctrl_mc_reg_table), GFP_KERNEL);
+
+ if (NULL == table)
+ return -ENOMEM;
+
+ /* Program additional LP registers that are no longer programmed by VBIOS */
+ cgs_write_register(hwmgr->device, mmMC_SEQ_RAS_TIMING_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_RAS_TIMING));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_CAS_TIMING_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_CAS_TIMING));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_DLL_STBY_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_DLL_STBY));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_G5PDX_CMD0_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_G5PDX_CMD0));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_G5PDX_CMD1_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_G5PDX_CMD1));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_G5PDX_CTRL_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_G5PDX_CTRL));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_DVS_CMD_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_PMG_DVS_CMD));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_DVS_CTL_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_PMG_DVS_CTL));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_MISC_TIMING_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_MISC_TIMING));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_MISC_TIMING2_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_MISC_TIMING2));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_CMD_EMRS_LP, cgs_read_register(hwmgr->device, mmMC_PMG_CMD_EMRS));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_CMD_MRS_LP, cgs_read_register(hwmgr->device, mmMC_PMG_CMD_MRS));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_CMD_MRS1_LP, cgs_read_register(hwmgr->device, mmMC_PMG_CMD_MRS1));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_WR_CTL_D0_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_WR_CTL_D0));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_WR_CTL_D1_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_WR_CTL_D1));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_RD_CTL_D0_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_RD_CTL_D0));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_RD_CTL_D1_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_RD_CTL_D1));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_TIMING_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_PMG_TIMING));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_CMD_MRS2_LP, cgs_read_register(hwmgr->device, mmMC_PMG_CMD_MRS2));
+ cgs_write_register(hwmgr->device, mmMC_SEQ_WR_CTL_2_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_WR_CTL_2));
+
+ memset(table, 0x00, sizeof(pp_atomctrl_mc_reg_table));
+
+ result = atomctrl_initialize_mc_reg_table(hwmgr, module_index, table);
+
+ if (0 == result)
+ result = iceland_copy_vbios_smc_reg_table(table, ni_table);
+
+ if (0 == result) {
+ iceland_set_s0_mc_reg_index(ni_table);
+ result = iceland_set_mc_special_registers(hwmgr, ni_table);
+ }
+
+ if (0 == result)
+ iceland_set_valid_flag(ni_table);
+
+ kfree(table);
+ return result;
+}
+
+/**
+ * Programs static screed detection parameters
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+int iceland_program_static_screen_threshold_parameters(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ /* Set static screen threshold unit*/
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device,
+ CGS_IND_REG__SMC, CG_STATIC_SCREEN_PARAMETER, STATIC_SCREEN_THRESHOLD_UNIT,
+ data->static_screen_threshold_unit);
+ /* Set static screen threshold*/
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device,
+ CGS_IND_REG__SMC, CG_STATIC_SCREEN_PARAMETER, STATIC_SCREEN_THRESHOLD,
+ data->static_screen_threshold);
+
+ return 0;
+}
+
+/**
+ * Setup display gap for glitch free memory clock switching.
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+int iceland_enable_display_gap(struct pp_hwmgr *hwmgr)
+{
+ uint32_t display_gap = cgs_read_ind_register(hwmgr->device,
+ CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL);
+
+ display_gap = PHM_SET_FIELD(display_gap,
+ CG_DISPLAY_GAP_CNTL, DISP_GAP, DISPLAY_GAP_IGNORE);
+
+ display_gap = PHM_SET_FIELD(display_gap,
+ CG_DISPLAY_GAP_CNTL, DISP_GAP_MCHG, DISPLAY_GAP_VBLANK);
+
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_DISPLAY_GAP_CNTL, display_gap);
+
+ return 0;
+}
+
+/**
+ * Programs activity state transition voting clients
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+int iceland_program_voting_clients(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ /* Clear reset for voting clients before enabling DPM */
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+ SCLK_PWRMGT_CNTL, RESET_SCLK_CNT, 0);
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+ SCLK_PWRMGT_CNTL, RESET_BUSY_CNT, 0);
+
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_FREQ_TRAN_VOTING_0, data->voting_rights_clients0);
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_FREQ_TRAN_VOTING_1, data->voting_rights_clients1);
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_FREQ_TRAN_VOTING_2, data->voting_rights_clients2);
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_FREQ_TRAN_VOTING_3, data->voting_rights_clients3);
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_FREQ_TRAN_VOTING_4, data->voting_rights_clients4);
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_FREQ_TRAN_VOTING_5, data->voting_rights_clients5);
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_FREQ_TRAN_VOTING_6, data->voting_rights_clients6);
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_FREQ_TRAN_VOTING_7, data->voting_rights_clients7);
+
+ return 0;
+}
+
+static int iceland_upload_firmware(struct pp_hwmgr *hwmgr)
+{
+ int ret = 0;
+
+ if (!iceland_is_smc_ram_running(hwmgr->smumgr))
+ ret = iceland_smu_upload_firmware_image(hwmgr->smumgr);
+
+ return ret;
+}
+
+/**
+ * Get the location of various tables inside the FW image.
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ */
+static int iceland_process_firmware_header(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ uint32_t tmp;
+ int result;
+ bool error = 0;
+
+ result = iceland_read_smc_sram_dword(hwmgr->smumgr,
+ SMU71_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU71_Firmware_Header, DpmTable),
+ &tmp, data->sram_end);
+
+ if (0 == result) {
+ data->dpm_table_start = tmp;
+ }
+
+ error |= (0 != result);
+
+ result = iceland_read_smc_sram_dword(hwmgr->smumgr,
+ SMU71_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU71_Firmware_Header, SoftRegisters),
+ &tmp, data->sram_end);
+
+ if (0 == result) {
+ data->soft_regs_start = tmp;
+ }
+
+ error |= (0 != result);
+
+
+ result = iceland_read_smc_sram_dword(hwmgr->smumgr,
+ SMU71_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU71_Firmware_Header, mcRegisterTable),
+ &tmp, data->sram_end);
+
+ if (0 == result) {
+ data->mc_reg_table_start = tmp;
+ }
+
+ result = iceland_read_smc_sram_dword(hwmgr->smumgr,
+ SMU71_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU71_Firmware_Header, FanTable),
+ &tmp, data->sram_end);
+
+ if (0 == result) {
+ data->fan_table_start = tmp;
+ }
+
+ error |= (0 != result);
+
+ result = iceland_read_smc_sram_dword(hwmgr->smumgr,
+ SMU71_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU71_Firmware_Header, mcArbDramTimingTable),
+ &tmp, data->sram_end);
+
+ if (0 == result) {
+ data->arb_table_start = tmp;
+ }
+
+ error |= (0 != result);
+
+
+ result = iceland_read_smc_sram_dword(hwmgr->smumgr,
+ SMU71_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU71_Firmware_Header, Version),
+ &tmp, data->sram_end);
+
+ if (0 == result) {
+ hwmgr->microcode_version_info.SMC = tmp;
+ }
+
+ error |= (0 != result);
+
+ result = iceland_read_smc_sram_dword(hwmgr->smumgr,
+ SMU71_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU71_Firmware_Header, UlvSettings),
+ &tmp, data->sram_end);
+
+ if (0 == result) {
+ data->ulv_settings_start = tmp;
+ }
+
+ error |= (0 != result);
+
+ return error ? 1 : 0;
+}
+
+/*
+* Copy one arb setting to another and then switch the active set.
+* arbFreqSrc and arbFreqDest is one of the MC_CG_ARB_FREQ_Fx constants.
+*/
+int iceland_copy_and_switch_arb_sets(struct pp_hwmgr *hwmgr,
+ uint32_t arbFreqSrc, uint32_t arbFreqDest)
+{
+ uint32_t mc_arb_dram_timing;
+ uint32_t mc_arb_dram_timing2;
+ uint32_t burst_time;
+ uint32_t mc_cg_config;
+
+ switch (arbFreqSrc) {
+ case MC_CG_ARB_FREQ_F0:
+ mc_arb_dram_timing = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING);
+ mc_arb_dram_timing2 = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING2);
+ burst_time = PHM_READ_FIELD(hwmgr->device, MC_ARB_BURST_TIME, STATE0);
+ break;
+
+ case MC_CG_ARB_FREQ_F1:
+ mc_arb_dram_timing = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING_1);
+ mc_arb_dram_timing2 = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING2_1);
+ burst_time = PHM_READ_FIELD(hwmgr->device, MC_ARB_BURST_TIME, STATE1);
+ break;
+
+ default:
+ return -1;
+ }
+
+ switch (arbFreqDest) {
+ case MC_CG_ARB_FREQ_F0:
+ cgs_write_register(hwmgr->device, mmMC_ARB_DRAM_TIMING, mc_arb_dram_timing);
+ cgs_write_register(hwmgr->device, mmMC_ARB_DRAM_TIMING2, mc_arb_dram_timing2);
+ PHM_WRITE_FIELD(hwmgr->device, MC_ARB_BURST_TIME, STATE0, burst_time);
+ break;
+
+ case MC_CG_ARB_FREQ_F1:
+ cgs_write_register(hwmgr->device, mmMC_ARB_DRAM_TIMING_1, mc_arb_dram_timing);
+ cgs_write_register(hwmgr->device, mmMC_ARB_DRAM_TIMING2_1, mc_arb_dram_timing2);
+ PHM_WRITE_FIELD(hwmgr->device, MC_ARB_BURST_TIME, STATE1, burst_time);
+ break;
+
+ default:
+ return -1;
+ }
+
+ mc_cg_config = cgs_read_register(hwmgr->device, mmMC_CG_CONFIG);
+ mc_cg_config |= 0x0000000F;
+ cgs_write_register(hwmgr->device, mmMC_CG_CONFIG, mc_cg_config);
+ PHM_WRITE_FIELD(hwmgr->device, MC_ARB_CG, CG_ARB_REQ, arbFreqDest);
+
+ return 0;
+}
+
+/**
+ * Initial switch from ARB F0->F1
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ * This function is to be called from the SetPowerState table.
+ */
+int iceland_initial_switch_from_arb_f0_to_f1(struct pp_hwmgr *hwmgr)
+{
+ return iceland_copy_and_switch_arb_sets(hwmgr, MC_CG_ARB_FREQ_F0, MC_CG_ARB_FREQ_F1);
+}
+
+/* ---------------------------------------- ULV related functions ----------------------------------------------------*/
+
+
+static int iceland_reset_single_dpm_table(
+ struct pp_hwmgr *hwmgr,
+ struct iceland_single_dpm_table *dpm_table,
+ uint32_t count)
+{
+ uint32_t i;
+ if (!(count <= MAX_REGULAR_DPM_NUMBER))
+ printk(KERN_ERR "[ powerplay ] Fatal error, can not set up single DPM \
+ table entries to exceed max number! \n");
+
+ dpm_table->count = count;
+ for (i = 0; i < MAX_REGULAR_DPM_NUMBER; i++) {
+ dpm_table->dpm_levels[i].enabled = 0;
+ }
+
+ return 0;
+}
+
+static void iceland_setup_pcie_table_entry(
+ struct iceland_single_dpm_table *dpm_table,
+ uint32_t index, uint32_t pcie_gen,
+ uint32_t pcie_lanes)
+{
+ dpm_table->dpm_levels[index].value = pcie_gen;
+ dpm_table->dpm_levels[index].param1 = pcie_lanes;
+ dpm_table->dpm_levels[index].enabled = 1;
+}
+
+/*
+ * Set up the PCIe DPM table as follows:
+ *
+ * A = Performance State, Max, Gen Speed
+ * C = Performance State, Min, Gen Speed
+ * 1 = Performance State, Max, Lane #
+ * 3 = Performance State, Min, Lane #
+ *
+ * B = Power Saving State, Max, Gen Speed
+ * D = Power Saving State, Min, Gen Speed
+ * 2 = Power Saving State, Max, Lane #
+ * 4 = Power Saving State, Min, Lane #
+ *
+ *
+ * DPM Index Gen Speed Lane #
+ * 5 A 1
+ * 4 B 2
+ * 3 C 1
+ * 2 D 2
+ * 1 C 3
+ * 0 D 4
+ *
+ */
+static int iceland_setup_default_pcie_tables(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ PP_ASSERT_WITH_CODE((data->use_pcie_performance_levels ||
+ data->use_pcie_power_saving_levels),
+ "No pcie performance levels!", return -EINVAL);
+
+ if (data->use_pcie_performance_levels && !data->use_pcie_power_saving_levels) {
+ data->pcie_gen_power_saving = data->pcie_gen_performance;
+ data->pcie_lane_power_saving = data->pcie_lane_performance;
+ } else if (!data->use_pcie_performance_levels && data->use_pcie_power_saving_levels) {
+ data->pcie_gen_performance = data->pcie_gen_power_saving;
+ data->pcie_lane_performance = data->pcie_lane_power_saving;
+ }
+
+ iceland_reset_single_dpm_table(hwmgr, &data->dpm_table.pcie_speed_table, SMU71_MAX_LEVELS_LINK);
+
+ /* Hardcode Pcie Table */
+ iceland_setup_pcie_table_entry(&data->dpm_table.pcie_speed_table, 0,
+ get_pcie_gen_support(data->pcie_gen_cap, PP_Min_PCIEGen),
+ get_pcie_lane_support(data->pcie_lane_cap, PP_Max_PCIELane));
+ iceland_setup_pcie_table_entry(&data->dpm_table.pcie_speed_table, 1,
+ get_pcie_gen_support(data->pcie_gen_cap, PP_Min_PCIEGen),
+ get_pcie_lane_support(data->pcie_lane_cap, PP_Max_PCIELane));
+ iceland_setup_pcie_table_entry(&data->dpm_table.pcie_speed_table, 2,
+ get_pcie_gen_support(data->pcie_gen_cap, PP_Max_PCIEGen),
+ get_pcie_lane_support(data->pcie_lane_cap, PP_Max_PCIELane));
+ iceland_setup_pcie_table_entry(&data->dpm_table.pcie_speed_table, 3,
+ get_pcie_gen_support(data->pcie_gen_cap, PP_Max_PCIEGen),
+ get_pcie_lane_support(data->pcie_lane_cap, PP_Max_PCIELane));
+ iceland_setup_pcie_table_entry(&data->dpm_table.pcie_speed_table, 4,
+ get_pcie_gen_support(data->pcie_gen_cap, PP_Max_PCIEGen),
+ get_pcie_lane_support(data->pcie_lane_cap, PP_Max_PCIELane));
+ iceland_setup_pcie_table_entry(&data->dpm_table.pcie_speed_table, 5,
+ get_pcie_gen_support(data->pcie_gen_cap, PP_Max_PCIEGen),
+ get_pcie_lane_support(data->pcie_lane_cap, PP_Max_PCIELane));
+ data->dpm_table.pcie_speed_table.count = 6;
+
+ return 0;
+
+}
+
+
+/*
+ * This function is to initalize all DPM state tables for SMU7 based on the dependency table.
+ * Dynamic state patching function will then trim these state tables to the allowed range based
+ * on the power policy or external client requests, such as UVD request, etc.
+ */
+static int iceland_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ uint32_t i;
+
+ struct phm_clock_voltage_dependency_table *allowed_vdd_sclk_table =
+ hwmgr->dyn_state.vddc_dependency_on_sclk;
+ struct phm_clock_voltage_dependency_table *allowed_vdd_mclk_table =
+ hwmgr->dyn_state.vddc_dependency_on_mclk;
+ struct phm_cac_leakage_table *std_voltage_table =
+ hwmgr->dyn_state.cac_leakage_table;
+
+ PP_ASSERT_WITH_CODE(allowed_vdd_sclk_table != NULL,
+ "SCLK dependency table is missing. This table is mandatory", return -1);
+ PP_ASSERT_WITH_CODE(allowed_vdd_sclk_table->count >= 1,
+ "SCLK dependency table has to have is missing. This table is mandatory", return -1);
+
+ PP_ASSERT_WITH_CODE(allowed_vdd_mclk_table != NULL,
+ "MCLK dependency table is missing. This table is mandatory", return -1);
+ PP_ASSERT_WITH_CODE(allowed_vdd_mclk_table->count >= 1,
+ "VMCLK dependency table has to have is missing. This table is mandatory", return -1);
+
+ /* clear the state table to reset everything to default */
+ memset(&(data->dpm_table), 0x00, sizeof(data->dpm_table));
+ iceland_reset_single_dpm_table(hwmgr, &data->dpm_table.sclk_table, SMU71_MAX_LEVELS_GRAPHICS);
+ iceland_reset_single_dpm_table(hwmgr, &data->dpm_table.mclk_table, SMU71_MAX_LEVELS_MEMORY);
+ iceland_reset_single_dpm_table(hwmgr, &data->dpm_table.vddc_table, SMU71_MAX_LEVELS_VDDC);
+ iceland_reset_single_dpm_table(hwmgr, &data->dpm_table.vdd_ci_table, SMU71_MAX_LEVELS_VDDCI);
+ iceland_reset_single_dpm_table(hwmgr, &data->dpm_table.mvdd_table, SMU71_MAX_LEVELS_MVDD);
+
+ PP_ASSERT_WITH_CODE(allowed_vdd_sclk_table != NULL,
+ "SCLK dependency table is missing. This table is mandatory", return -1);
+ /* Initialize Sclk DPM table based on allow Sclk values*/
+ data->dpm_table.sclk_table.count = 0;
+
+ for (i = 0; i < allowed_vdd_sclk_table->count; i++) {
+ if (i == 0 || data->dpm_table.sclk_table.dpm_levels[data->dpm_table.sclk_table.count-1].value !=
+ allowed_vdd_sclk_table->entries[i].clk) {
+ data->dpm_table.sclk_table.dpm_levels[data->dpm_table.sclk_table.count].value =
+ allowed_vdd_sclk_table->entries[i].clk;
+ data->dpm_table.sclk_table.dpm_levels[data->dpm_table.sclk_table.count].enabled = 1; /*(i==0) ? 1 : 0; to do */
+ data->dpm_table.sclk_table.count++;
+ }
+ }
+
+ PP_ASSERT_WITH_CODE(allowed_vdd_mclk_table != NULL,
+ "MCLK dependency table is missing. This table is mandatory", return -1);
+ /* Initialize Mclk DPM table based on allow Mclk values */
+ data->dpm_table.mclk_table.count = 0;
+ for (i = 0; i < allowed_vdd_mclk_table->count; i++) {
+ if (i == 0 || data->dpm_table.mclk_table.dpm_levels[data->dpm_table.mclk_table.count-1].value !=
+ allowed_vdd_mclk_table->entries[i].clk) {
+ data->dpm_table.mclk_table.dpm_levels[data->dpm_table.mclk_table.count].value =
+ allowed_vdd_mclk_table->entries[i].clk;
+ data->dpm_table.mclk_table.dpm_levels[data->dpm_table.mclk_table.count].enabled = 1; /*(i==0) ? 1 : 0; */
+ data->dpm_table.mclk_table.count++;
+ }
+ }
+
+ /* Initialize Vddc DPM table based on allow Vddc values. And populate corresponding std values. */
+ for (i = 0; i < allowed_vdd_sclk_table->count; i++) {
+ data->dpm_table.vddc_table.dpm_levels[i].value = allowed_vdd_mclk_table->entries[i].v;
+ data->dpm_table.vddc_table.dpm_levels[i].param1 = std_voltage_table->entries[i].Leakage;
+ /* param1 is for corresponding std voltage */
+ data->dpm_table.vddc_table.dpm_levels[i].enabled = 1;
+ }
+
+ data->dpm_table.vddc_table.count = allowed_vdd_sclk_table->count;
+ allowed_vdd_mclk_table = hwmgr->dyn_state.vddci_dependency_on_mclk;
+
+ if (NULL != allowed_vdd_mclk_table) {
+ /* Initialize Vddci DPM table based on allow Mclk values */
+ for (i = 0; i < allowed_vdd_mclk_table->count; i++) {
+ data->dpm_table.vdd_ci_table.dpm_levels[i].value = allowed_vdd_mclk_table->entries[i].v;
+ data->dpm_table.vdd_ci_table.dpm_levels[i].enabled = 1;
+ }
+ data->dpm_table.vdd_ci_table.count = allowed_vdd_mclk_table->count;
+ }
+
+ allowed_vdd_mclk_table = hwmgr->dyn_state.mvdd_dependency_on_mclk;
+
+ if (NULL != allowed_vdd_mclk_table) {
+ /*
+ * Initialize MVDD DPM table based on allow Mclk
+ * values
+ */
+ for (i = 0; i < allowed_vdd_mclk_table->count; i++) {
+ data->dpm_table.mvdd_table.dpm_levels[i].value = allowed_vdd_mclk_table->entries[i].v;
+ data->dpm_table.mvdd_table.dpm_levels[i].enabled = 1;
+ }
+ data->dpm_table.mvdd_table.count = allowed_vdd_mclk_table->count;
+ }
+
+ /* setup PCIE gen speed levels*/
+ iceland_setup_default_pcie_tables(hwmgr);
+
+ /* save a copy of the default DPM table*/
+ memcpy(&(data->golden_dpm_table), &(data->dpm_table), sizeof(struct iceland_dpm_table));
+
+ return 0;
+}
+
+/**
+ * @brief PhwIceland_GetVoltageOrder
+ * Returns index of requested voltage record in lookup(table)
+ * @param hwmgr - pointer to hardware manager
+ * @param lookutab - lookup list to search in
+ * @param voltage - voltage to look for
+ * @return 0 on success
+ */
+uint8_t iceland_get_voltage_index(phm_ppt_v1_voltage_lookup_table *look_up_table,
+ uint16_t voltage)
+{
+ uint8_t count = (uint8_t) (look_up_table->count);
+ uint8_t i;
+
+ PP_ASSERT_WITH_CODE((NULL != look_up_table), "Lookup Table empty.", return 0;);
+ PP_ASSERT_WITH_CODE((0 != count), "Lookup Table empty.", return 0;);
+
+ for (i = 0; i < count; i++) {
+ /* find first voltage equal or bigger than requested */
+ if (look_up_table->entries[i].us_vdd >= voltage)
+ return i;
+ }
+
+ /* voltage is bigger than max voltage in the table */
+ return i-1;
+}
+
+
+static int iceland_get_std_voltage_value_sidd(struct pp_hwmgr *hwmgr,
+ pp_atomctrl_voltage_table_entry *tab, uint16_t *hi,
+ uint16_t *lo)
+{
+ uint16_t v_index;
+ bool vol_found = false;
+ *hi = tab->value * VOLTAGE_SCALE;
+ *lo = tab->value * VOLTAGE_SCALE;
+
+ /* SCLK/VDDC Dependency Table has to exist. */
+ PP_ASSERT_WITH_CODE(NULL != hwmgr->dyn_state.vddc_dependency_on_sclk,
+ "The SCLK/VDDC Dependency Table does not exist.\n",
+ return -EINVAL);
+
+ if (NULL == hwmgr->dyn_state.cac_leakage_table) {
+ pr_warning("CAC Leakage Table does not exist, using vddc.\n");
+ return 0;
+ }
+
+ /*
+ * Since voltage in the sclk/vddc dependency table is not
+ * necessarily in ascending order because of ELB voltage
+ * patching, loop through entire list to find exact voltage.
+ */
+ for (v_index = 0; (uint32_t)v_index < hwmgr->dyn_state.vddc_dependency_on_sclk->count; v_index++) {
+ if (tab->value == hwmgr->dyn_state.vddc_dependency_on_sclk->entries[v_index].v) {
+ vol_found = true;
+ if ((uint32_t)v_index < hwmgr->dyn_state.cac_leakage_table->count) {
+ *lo = hwmgr->dyn_state.cac_leakage_table->entries[v_index].Vddc * VOLTAGE_SCALE;
+ *hi = (uint16_t)(hwmgr->dyn_state.cac_leakage_table->entries[v_index].Leakage * VOLTAGE_SCALE);
+ } else {
+ pr_warning("Index from SCLK/VDDC Dependency Table exceeds the CAC Leakage Table index, using maximum index from CAC table.\n");
+ *lo = hwmgr->dyn_state.cac_leakage_table->entries[hwmgr->dyn_state.cac_leakage_table->count - 1].Vddc * VOLTAGE_SCALE;
+ *hi = (uint16_t)(hwmgr->dyn_state.cac_leakage_table->entries[hwmgr->dyn_state.cac_leakage_table->count - 1].Leakage * VOLTAGE_SCALE);
+ }
+ break;
+ }
+ }
+
+ /*
+ * If voltage is not found in the first pass, loop again to
+ * find the best match, equal or higher value.
+ */
+ if (!vol_found) {
+ for (v_index = 0; (uint32_t)v_index < hwmgr->dyn_state.vddc_dependency_on_sclk->count; v_index++) {
+ if (tab->value <= hwmgr->dyn_state.vddc_dependency_on_sclk->entries[v_index].v) {
+ vol_found = true;
+ if ((uint32_t)v_index < hwmgr->dyn_state.cac_leakage_table->count) {
+ *lo = hwmgr->dyn_state.cac_leakage_table->entries[v_index].Vddc * VOLTAGE_SCALE;
+ *hi = (uint16_t)(hwmgr->dyn_state.cac_leakage_table->entries[v_index].Leakage) * VOLTAGE_SCALE;
+ } else {
+ pr_warning("Index from SCLK/VDDC Dependency Table exceeds the CAC Leakage Table index in second look up, using maximum index from CAC table.");
+ *lo = hwmgr->dyn_state.cac_leakage_table->entries[hwmgr->dyn_state.cac_leakage_table->count - 1].Vddc * VOLTAGE_SCALE;
+ *hi = (uint16_t)(hwmgr->dyn_state.cac_leakage_table->entries[hwmgr->dyn_state.cac_leakage_table->count - 1].Leakage * VOLTAGE_SCALE);
+ }
+ break;
+ }
+ }
+
+ if (!vol_found)
+ pr_warning("Unable to get std_vddc from SCLK/VDDC Dependency Table, using vddc.\n");
+ }
+
+ return 0;
+}
+
+static int iceland_populate_smc_voltage_table(struct pp_hwmgr *hwmgr,
+ pp_atomctrl_voltage_table_entry *tab,
+ SMU71_Discrete_VoltageLevel *smc_voltage_tab) {
+ int result;
+
+
+ result = iceland_get_std_voltage_value_sidd(hwmgr, tab,
+ &smc_voltage_tab->StdVoltageHiSidd,
+ &smc_voltage_tab->StdVoltageLoSidd);
+ if (0 != result) {
+ smc_voltage_tab->StdVoltageHiSidd = tab->value * VOLTAGE_SCALE;
+ smc_voltage_tab->StdVoltageLoSidd = tab->value * VOLTAGE_SCALE;
+ }
+
+ smc_voltage_tab->Voltage = PP_HOST_TO_SMC_US(tab->value * VOLTAGE_SCALE);
+ CONVERT_FROM_HOST_TO_SMC_US(smc_voltage_tab->StdVoltageHiSidd);
+ CONVERT_FROM_HOST_TO_SMC_US(smc_voltage_tab->StdVoltageHiSidd);
+
+ return 0;
+}
+
+/**
+ * Vddc table preparation for SMC.
+ *
+ * @param hwmgr the address of the hardware manager
+ * @param table the SMC DPM table structure to be populated
+ * @return always 0
+ */
+static int iceland_populate_smc_vddc_table(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ unsigned int count;
+ int result;
+
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ table->VddcLevelCount = data->vddc_voltage_table.count;
+ for (count = 0; count < table->VddcLevelCount; count++) {
+ result = iceland_populate_smc_voltage_table(hwmgr,
+ &data->vddc_voltage_table.entries[count],
+ &table->VddcLevel[count]);
+ PP_ASSERT_WITH_CODE(0 == result, "do not populate SMC VDDC voltage table", return -EINVAL);
+
+ /* GPIO voltage control */
+ if (ICELAND_VOLTAGE_CONTROL_BY_GPIO == data->voltage_control)
+ table->VddcLevel[count].Smio |= data->vddc_voltage_table.entries[count].smio_low;
+ else if (ICELAND_VOLTAGE_CONTROL_BY_SVID2 == data->voltage_control)
+ table->VddcLevel[count].Smio = 0;
+ }
+
+ CONVERT_FROM_HOST_TO_SMC_UL(table->VddcLevelCount);
+
+ return 0;
+}
+
+/**
+ * Vddci table preparation for SMC.
+ *
+ * @param *hwmgr The address of the hardware manager.
+ * @param *table The SMC DPM table structure to be populated.
+ * @return 0
+ */
+static int iceland_populate_smc_vdd_ci_table(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ int result;
+ uint32_t count;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ table->VddciLevelCount = data->vddci_voltage_table.count;
+ for (count = 0; count < table->VddciLevelCount; count++) {
+ result = iceland_populate_smc_voltage_table(hwmgr,
+ &data->vddci_voltage_table.entries[count],
+ &table->VddciLevel[count]);
+ PP_ASSERT_WITH_CODE(0 == result, "do not populate SMC VDDCI voltage table", return -EINVAL);
+
+ /* GPIO voltage control */
+ if (ICELAND_VOLTAGE_CONTROL_BY_GPIO == data->vdd_ci_control)
+ table->VddciLevel[count].Smio |= data->vddci_voltage_table.entries[count].smio_low;
+ else
+ table->VddciLevel[count].Smio = 0;
+ }
+
+ CONVERT_FROM_HOST_TO_SMC_UL(table->VddcLevelCount);
+
+ return 0;
+}
+
+/**
+ * Mvdd table preparation for SMC.
+ *
+ * @param *hwmgr The address of the hardware manager.
+ * @param *table The SMC DPM table structure to be populated.
+ * @return 0
+ */
+static int iceland_populate_smc_mvdd_table(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ int result;
+ uint32_t count;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ table->MvddLevelCount = data->mvdd_voltage_table.count;
+ for (count = 0; count < table->MvddLevelCount; count++) {
+ result = iceland_populate_smc_voltage_table(hwmgr,
+ &data->mvdd_voltage_table.entries[count],
+ &table->MvddLevel[count]);
+ PP_ASSERT_WITH_CODE(0 == result, "do not populate SMC VDDCI voltage table", return -EINVAL);
+
+ /* GPIO voltage control */
+ if (ICELAND_VOLTAGE_CONTROL_BY_GPIO == data->mvdd_control)
+ table->MvddLevel[count].Smio |= data->mvdd_voltage_table.entries[count].smio_low;
+ else
+ table->MvddLevel[count].Smio = 0;
+ }
+
+ CONVERT_FROM_HOST_TO_SMC_UL(table->MvddLevelCount);
+
+ return 0;
+}
+
+int iceland_populate_bapm_vddc_vid_sidd(struct pp_hwmgr *hwmgr)
+{
+ int i;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint8_t * hi_vid = data->power_tune_table.BapmVddCVidHiSidd;
+ uint8_t * lo_vid = data->power_tune_table.BapmVddCVidLoSidd;
+
+ PP_ASSERT_WITH_CODE(NULL != hwmgr->dyn_state.cac_leakage_table,
+ "The CAC Leakage table does not exist!", return -EINVAL);
+ PP_ASSERT_WITH_CODE(hwmgr->dyn_state.cac_leakage_table->count <= 8,
+ "There should never be more than 8 entries for BapmVddcVid!!!", return -EINVAL);
+ PP_ASSERT_WITH_CODE(hwmgr->dyn_state.cac_leakage_table->count == hwmgr->dyn_state.vddc_dependency_on_sclk->count,
+ "CACLeakageTable->count and VddcDependencyOnSCLk->count not equal", return -EINVAL);
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_EVV)) {
+ for (i = 0; (uint32_t) i < hwmgr->dyn_state.cac_leakage_table->count; i++) {
+ lo_vid[i] = convert_to_vid(hwmgr->dyn_state.cac_leakage_table->entries[i].Vddc1);
+ hi_vid[i] = convert_to_vid(hwmgr->dyn_state.cac_leakage_table->entries[i].Vddc2);
+ }
+ } else {
+ PP_ASSERT_WITH_CODE(false, "Iceland should always support EVV", return -EINVAL);
+ }
+
+ return 0;
+}
+
+int iceland_populate_vddc_vid(struct pp_hwmgr *hwmgr)
+{
+ int i;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint8_t *vid = data->power_tune_table.VddCVid;
+
+ PP_ASSERT_WITH_CODE(data->vddc_voltage_table.count <= 8,
+ "There should never be more than 8 entries for VddcVid!!!",
+ return -EINVAL);
+
+ for (i = 0; i < (int)data->vddc_voltage_table.count; i++) {
+ vid[i] = convert_to_vid(data->vddc_voltage_table.entries[i].value);
+ }
+
+ return 0;
+}
+
+/**
+ * Preparation of voltage tables for SMC.
+ *
+ * @param hwmgr the address of the hardware manager
+ * @param table the SMC DPM table structure to be populated
+ * @return always 0
+ */
+
+int iceland_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ int result;
+
+ result = iceland_populate_smc_vddc_table(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "can not populate VDDC voltage table to SMC", return -1);
+
+ result = iceland_populate_smc_vdd_ci_table(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "can not populate VDDCI voltage table to SMC", return -1);
+
+ result = iceland_populate_smc_mvdd_table(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "can not populate MVDD voltage table to SMC", return -1);
+
+ return 0;
+}
+
+
+/**
+ * Re-generate the DPM level mask value
+ * @param hwmgr the address of the hardware manager
+ */
+static uint32_t iceland_get_dpm_level_enable_mask_value(
+ struct iceland_single_dpm_table * dpm_table)
+{
+ uint32_t i;
+ uint32_t mask_value = 0;
+
+ for (i = dpm_table->count; i > 0; i--) {
+ mask_value = mask_value << 1;
+
+ if (dpm_table->dpm_levels[i-1].enabled)
+ mask_value |= 0x1;
+ else
+ mask_value &= 0xFFFFFFFE;
+ }
+ return mask_value;
+}
+
+int iceland_populate_memory_timing_parameters(
+ struct pp_hwmgr *hwmgr,
+ uint32_t engine_clock,
+ uint32_t memory_clock,
+ struct SMU71_Discrete_MCArbDramTimingTableEntry *arb_regs
+ )
+{
+ uint32_t dramTiming;
+ uint32_t dramTiming2;
+ uint32_t burstTime;
+ int result;
+
+ result = atomctrl_set_engine_dram_timings_rv770(hwmgr,
+ engine_clock, memory_clock);
+
+ PP_ASSERT_WITH_CODE(result == 0,
+ "Error calling VBIOS to set DRAM_TIMING.", return result);
+
+ dramTiming = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING);
+ dramTiming2 = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING2);
+ burstTime = PHM_READ_FIELD(hwmgr->device, MC_ARB_BURST_TIME, STATE0);
+
+ arb_regs->McArbDramTiming = PP_HOST_TO_SMC_UL(dramTiming);
+ arb_regs->McArbDramTiming2 = PP_HOST_TO_SMC_UL(dramTiming2);
+ arb_regs->McArbBurstTime = (uint8_t)burstTime;
+
+ return 0;
+}
+
+/**
+ * Setup parameters for the MC ARB.
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return always 0
+ * This function is to be called from the SetPowerState table.
+ */
+int iceland_program_memory_timing_parameters(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ int result = 0;
+ SMU71_Discrete_MCArbDramTimingTable arb_regs;
+ uint32_t i, j;
+
+ memset(&arb_regs, 0x00, sizeof(SMU71_Discrete_MCArbDramTimingTable));
+
+ for (i = 0; i < data->dpm_table.sclk_table.count; i++) {
+ for (j = 0; j < data->dpm_table.mclk_table.count; j++) {
+ result = iceland_populate_memory_timing_parameters
+ (hwmgr, data->dpm_table.sclk_table.dpm_levels[i].value,
+ data->dpm_table.mclk_table.dpm_levels[j].value,
+ &arb_regs.entries[i][j]);
+
+ if (0 != result) {
+ break;
+ }
+ }
+ }
+
+ if (0 == result) {
+ result = iceland_copy_bytes_to_smc(
+ hwmgr->smumgr,
+ data->arb_table_start,
+ (uint8_t *)&arb_regs,
+ sizeof(SMU71_Discrete_MCArbDramTimingTable),
+ data->sram_end
+ );
+ }
+
+ return result;
+}
+
+static int iceland_populate_smc_link_level(struct pp_hwmgr *hwmgr, SMU71_Discrete_DpmTable *table)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_dpm_table *dpm_table = &data->dpm_table;
+ uint32_t i;
+
+ /* Index (dpm_table->pcie_speed_table.count) is reserved for PCIE boot level. */
+ for (i = 0; i <= dpm_table->pcie_speed_table.count; i++) {
+ table->LinkLevel[i].PcieGenSpeed =
+ (uint8_t)dpm_table->pcie_speed_table.dpm_levels[i].value;
+ table->LinkLevel[i].PcieLaneCount =
+ (uint8_t)encode_pcie_lane_width(dpm_table->pcie_speed_table.dpm_levels[i].param1);
+ table->LinkLevel[i].EnabledForActivity =
+ 1;
+ table->LinkLevel[i].SPC =
+ (uint8_t)(data->pcie_spc_cap & 0xff);
+ table->LinkLevel[i].DownThreshold =
+ PP_HOST_TO_SMC_UL(5);
+ table->LinkLevel[i].UpThreshold =
+ PP_HOST_TO_SMC_UL(30);
+ }
+
+ data->smc_state_table.LinkLevelCount =
+ (uint8_t)dpm_table->pcie_speed_table.count;
+ data->dpm_level_enable_mask.pcie_dpm_enable_mask =
+ iceland_get_dpm_level_enable_mask_value(&dpm_table->pcie_speed_table);
+
+ return 0;
+}
+
+static int iceland_populate_smc_uvd_level(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ return 0;
+}
+
+uint8_t iceland_get_voltage_id(pp_atomctrl_voltage_table *voltage_table,
+ uint32_t voltage)
+{
+ uint8_t count = (uint8_t) (voltage_table->count);
+ uint8_t i = 0;
+
+ PP_ASSERT_WITH_CODE((NULL != voltage_table),
+ "Voltage Table empty.", return 0;);
+ PP_ASSERT_WITH_CODE((0 != count),
+ "Voltage Table empty.", return 0;);
+
+ for (i = 0; i < count; i++) {
+ /* find first voltage bigger than requested */
+ if (voltage_table->entries[i].value >= voltage)
+ return i;
+ }
+
+ /* voltage is bigger than max voltage in the table */
+ return i - 1;
+}
+
+static int iceland_populate_smc_vce_level(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ return 0;
+}
+
+static int iceland_populate_smc_acp_level(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ return 0;
+}
+
+static int iceland_populate_smc_samu_level(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ return 0;
+}
+
+
+static int iceland_populate_smc_svi2_config(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *tab)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ if(ICELAND_VOLTAGE_CONTROL_BY_SVID2 == data->voltage_control)
+ tab->SVI2Enable |= VDDC_ON_SVI2;
+
+ if(ICELAND_VOLTAGE_CONTROL_BY_SVID2 == data->vdd_ci_control)
+ tab->SVI2Enable |= VDDCI_ON_SVI2;
+ else
+ tab->MergedVddci = 1;
+
+ if(ICELAND_VOLTAGE_CONTROL_BY_SVID2 == data->mvdd_control)
+ tab->SVI2Enable |= MVDD_ON_SVI2;
+
+ PP_ASSERT_WITH_CODE( tab->SVI2Enable != (VDDC_ON_SVI2 | VDDCI_ON_SVI2 | MVDD_ON_SVI2) &&
+ (tab->SVI2Enable & VDDC_ON_SVI2), "SVI2 domain configuration is incorrect!", return -EINVAL);
+
+ return 0;
+}
+
+static int iceland_get_dependecy_volt_by_clk(struct pp_hwmgr *hwmgr,
+ struct phm_clock_voltage_dependency_table *allowed_clock_voltage_table,
+ uint32_t clock, uint32_t *vol)
+{
+ uint32_t i = 0;
+
+ /* clock - voltage dependency table is empty table */
+ if (allowed_clock_voltage_table->count == 0)
+ return -EINVAL;
+
+ for (i = 0; i < allowed_clock_voltage_table->count; i++) {
+ /* find first sclk bigger than request */
+ if (allowed_clock_voltage_table->entries[i].clk >= clock) {
+ *vol = allowed_clock_voltage_table->entries[i].v;
+ return 0;
+ }
+ }
+
+ /* sclk is bigger than max sclk in the dependence table */
+ *vol = allowed_clock_voltage_table->entries[i - 1].v;
+
+ return 0;
+}
+
+static uint8_t iceland_get_mclk_frequency_ratio(uint32_t memory_clock,
+ bool strobe_mode)
+{
+ uint8_t mc_para_index;
+
+ if (strobe_mode) {
+ if (memory_clock < 12500) {
+ mc_para_index = 0x00;
+ } else if (memory_clock > 47500) {
+ mc_para_index = 0x0f;
+ } else {
+ mc_para_index = (uint8_t)((memory_clock - 10000) / 2500);
+ }
+ } else {
+ if (memory_clock < 65000) {
+ mc_para_index = 0x00;
+ } else if (memory_clock > 135000) {
+ mc_para_index = 0x0f;
+ } else {
+ mc_para_index = (uint8_t)((memory_clock - 60000) / 5000);
+ }
+ }
+
+ return mc_para_index;
+}
+
+static uint8_t iceland_get_ddr3_mclk_frequency_ratio(uint32_t memory_clock)
+{
+ uint8_t mc_para_index;
+
+ if (memory_clock < 10000) {
+ mc_para_index = 0;
+ } else if (memory_clock >= 80000) {
+ mc_para_index = 0x0f;
+ } else {
+ mc_para_index = (uint8_t)((memory_clock - 10000) / 5000 + 1);
+ }
+
+ return mc_para_index;
+}
+
+static int iceland_populate_phase_value_based_on_sclk(struct pp_hwmgr *hwmgr, const struct phm_phase_shedding_limits_table *pl,
+ uint32_t sclk, uint32_t *p_shed)
+{
+ unsigned int i;
+
+ /* use the minimum phase shedding */
+ *p_shed = 1;
+
+ /*
+ * PPGen ensures the phase shedding limits table is sorted
+ * from lowest voltage/sclk/mclk to highest voltage/sclk/mclk.
+ * VBIOS ensures the phase shedding masks table is sorted from
+ * least phases enabled (phase shedding on) to most phases
+ * enabled (phase shedding off).
+ */
+ for (i = 0; i < pl->count; i++) {
+ if (sclk < pl->entries[i].Sclk) {
+ /* Enable phase shedding */
+ *p_shed = i;
+ break;
+ }
+ }
+
+ return 0;
+}
+
+static int iceland_populate_phase_value_based_on_mclk(struct pp_hwmgr *hwmgr, const struct phm_phase_shedding_limits_table *pl,
+ uint32_t memory_clock, uint32_t *p_shed)
+{
+ unsigned int i;
+
+ /* use the minimum phase shedding */
+ *p_shed = 1;
+
+ /*
+ * PPGen ensures the phase shedding limits table is sorted
+ * from lowest voltage/sclk/mclk to highest voltage/sclk/mclk.
+ * VBIOS ensures the phase shedding masks table is sorted from
+ * least phases enabled (phase shedding on) to most phases
+ * enabled (phase shedding off).
+ */
+ for (i = 0; i < pl->count; i++) {
+ if (memory_clock < pl->entries[i].Mclk) {
+ /* Enable phase shedding */
+ *p_shed = i;
+ break;
+ }
+ }
+
+ return 0;
+}
+
+/**
+ * Populates the SMC MCLK structure using the provided memory clock
+ *
+ * @param hwmgr the address of the hardware manager
+ * @param memory_clock the memory clock to use to populate the structure
+ * @param sclk the SMC SCLK structure to be populated
+ */
+static int iceland_calculate_mclk_params(
+ struct pp_hwmgr *hwmgr,
+ uint32_t memory_clock,
+ SMU71_Discrete_MemoryLevel *mclk,
+ bool strobe_mode,
+ bool dllStateOn
+ )
+{
+ const iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ uint32_t dll_cntl = data->clock_registers.vDLL_CNTL;
+ uint32_t mclk_pwrmgt_cntl = data->clock_registers.vMCLK_PWRMGT_CNTL;
+ uint32_t mpll_ad_func_cntl = data->clock_registers.vMPLL_AD_FUNC_CNTL;
+ uint32_t mpll_dq_func_cntl = data->clock_registers.vMPLL_DQ_FUNC_CNTL;
+ uint32_t mpll_func_cntl = data->clock_registers.vMPLL_FUNC_CNTL;
+ uint32_t mpll_func_cntl_1 = data->clock_registers.vMPLL_FUNC_CNTL_1;
+ uint32_t mpll_func_cntl_2 = data->clock_registers.vMPLL_FUNC_CNTL_2;
+ uint32_t mpll_ss1 = data->clock_registers.vMPLL_SS1;
+ uint32_t mpll_ss2 = data->clock_registers.vMPLL_SS2;
+
+ pp_atomctrl_memory_clock_param mpll_param;
+ int result;
+
+ result = atomctrl_get_memory_pll_dividers_si(hwmgr,
+ memory_clock, &mpll_param, strobe_mode);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Error retrieving Memory Clock Parameters from VBIOS.", return result);
+
+ /* MPLL_FUNC_CNTL setup*/
+ mpll_func_cntl = PHM_SET_FIELD(mpll_func_cntl, MPLL_FUNC_CNTL, BWCTRL, mpll_param.bw_ctrl);
+
+ /* MPLL_FUNC_CNTL_1 setup*/
+ mpll_func_cntl_1 = PHM_SET_FIELD(mpll_func_cntl_1,
+ MPLL_FUNC_CNTL_1, CLKF, mpll_param.mpll_fb_divider.cl_kf);
+ mpll_func_cntl_1 = PHM_SET_FIELD(mpll_func_cntl_1,
+ MPLL_FUNC_CNTL_1, CLKFRAC, mpll_param.mpll_fb_divider.clk_frac);
+ mpll_func_cntl_1 = PHM_SET_FIELD(mpll_func_cntl_1,
+ MPLL_FUNC_CNTL_1, VCO_MODE, mpll_param.vco_mode);
+
+ /* MPLL_AD_FUNC_CNTL setup*/
+ mpll_ad_func_cntl = PHM_SET_FIELD(mpll_ad_func_cntl,
+ MPLL_AD_FUNC_CNTL, YCLK_POST_DIV, mpll_param.mpll_post_divider);
+
+ if (data->is_memory_GDDR5) {
+ /* MPLL_DQ_FUNC_CNTL setup*/
+ mpll_dq_func_cntl = PHM_SET_FIELD(mpll_dq_func_cntl,
+ MPLL_DQ_FUNC_CNTL, YCLK_SEL, mpll_param.yclk_sel);
+ mpll_dq_func_cntl = PHM_SET_FIELD(mpll_dq_func_cntl,
+ MPLL_DQ_FUNC_CNTL, YCLK_POST_DIV, mpll_param.mpll_post_divider);
+ }
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_MemorySpreadSpectrumSupport)) {
+ /*
+ ************************************
+ Fref = Reference Frequency
+ NF = Feedback divider ratio
+ NR = Reference divider ratio
+ Fnom = Nominal VCO output frequency = Fref * NF / NR
+ Fs = Spreading Rate
+ D = Percentage down-spread / 2
+ Fint = Reference input frequency to PFD = Fref / NR
+ NS = Spreading rate divider ratio = int(Fint / (2 * Fs))
+ CLKS = NS - 1 = ISS_STEP_NUM[11:0]
+ NV = D * Fs / Fnom * 4 * ((Fnom/Fref * NR) ^ 2)
+ CLKV = 65536 * NV = ISS_STEP_SIZE[25:0]
+ *************************************
+ */
+ pp_atomctrl_internal_ss_info ss_info;
+ uint32_t freq_nom;
+ uint32_t tmp;
+ uint32_t reference_clock = atomctrl_get_mpll_reference_clock(hwmgr);
+
+ /* for GDDR5 for all modes and DDR3 */
+ if (1 == mpll_param.qdr)
+ freq_nom = memory_clock * 4 * (1 << mpll_param.mpll_post_divider);
+ else
+ freq_nom = memory_clock * 2 * (1 << mpll_param.mpll_post_divider);
+
+ /* tmp = (freq_nom / reference_clock * reference_divider) ^ 2 Note: S.I. reference_divider = 1*/
+ tmp = (freq_nom / reference_clock);
+ tmp = tmp * tmp;
+
+ if (0 == atomctrl_get_memory_clock_spread_spectrum(hwmgr, freq_nom, &ss_info)) {
+ /* ss_info.speed_spectrum_percentage -- in unit of 0.01% */
+ /* ss.Info.speed_spectrum_rate -- in unit of khz */
+ /* CLKS = reference_clock / (2 * speed_spectrum_rate * reference_divider) * 10 */
+ /* = reference_clock * 5 / speed_spectrum_rate */
+ uint32_t clks = reference_clock * 5 / ss_info.speed_spectrum_rate;
+
+ /* CLKV = 65536 * speed_spectrum_percentage / 2 * spreadSpecrumRate / freq_nom * 4 / 100000 * ((freq_nom / reference_clock) ^ 2) */
+ /* = 131 * speed_spectrum_percentage * speed_spectrum_rate / 100 * ((freq_nom / reference_clock) ^ 2) / freq_nom */
+ uint32_t clkv =
+ (uint32_t)((((131 * ss_info.speed_spectrum_percentage *
+ ss_info.speed_spectrum_rate) / 100) * tmp) / freq_nom);
+
+ mpll_ss1 = PHM_SET_FIELD(mpll_ss1, MPLL_SS1, CLKV, clkv);
+ mpll_ss2 = PHM_SET_FIELD(mpll_ss2, MPLL_SS2, CLKS, clks);
+ }
+ }
+
+ /* MCLK_PWRMGT_CNTL setup */
+ mclk_pwrmgt_cntl = PHM_SET_FIELD(mclk_pwrmgt_cntl,
+ MCLK_PWRMGT_CNTL, DLL_SPEED, mpll_param.dll_speed);
+ mclk_pwrmgt_cntl = PHM_SET_FIELD(mclk_pwrmgt_cntl,
+ MCLK_PWRMGT_CNTL, MRDCK0_PDNB, dllStateOn);
+ mclk_pwrmgt_cntl = PHM_SET_FIELD(mclk_pwrmgt_cntl,
+ MCLK_PWRMGT_CNTL, MRDCK1_PDNB, dllStateOn);
+
+
+ /* Save the result data to outpupt memory level structure */
+ mclk->MclkFrequency = memory_clock;
+ mclk->MpllFuncCntl = mpll_func_cntl;
+ mclk->MpllFuncCntl_1 = mpll_func_cntl_1;
+ mclk->MpllFuncCntl_2 = mpll_func_cntl_2;
+ mclk->MpllAdFuncCntl = mpll_ad_func_cntl;
+ mclk->MpllDqFuncCntl = mpll_dq_func_cntl;
+ mclk->MclkPwrmgtCntl = mclk_pwrmgt_cntl;
+ mclk->DllCntl = dll_cntl;
+ mclk->MpllSs1 = mpll_ss1;
+ mclk->MpllSs2 = mpll_ss2;
+
+ return 0;
+}
+
+static int iceland_populate_single_memory_level(
+ struct pp_hwmgr *hwmgr,
+ uint32_t memory_clock,
+ SMU71_Discrete_MemoryLevel *memory_level
+ )
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ int result = 0;
+ bool dllStateOn;
+ struct cgs_display_info info = {0};
+
+
+ if (NULL != hwmgr->dyn_state.vddc_dependency_on_mclk) {
+ result = iceland_get_dependecy_volt_by_clk(hwmgr,
+ hwmgr->dyn_state.vddc_dependency_on_mclk, memory_clock, &memory_level->MinVddc);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "can not find MinVddc voltage value from memory VDDC voltage dependency table", return result);
+ }
+
+ if (data->vdd_ci_control == ICELAND_VOLTAGE_CONTROL_NONE) {
+ memory_level->MinVddci = memory_level->MinVddc;
+ } else if (NULL != hwmgr->dyn_state.vddci_dependency_on_mclk) {
+ result = iceland_get_dependecy_volt_by_clk(hwmgr,
+ hwmgr->dyn_state.vddci_dependency_on_mclk,
+ memory_clock,
+ &memory_level->MinVddci);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "can not find MinVddci voltage value from memory VDDCI voltage dependency table", return result);
+ }
+
+ if (NULL != hwmgr->dyn_state.mvdd_dependency_on_mclk) {
+ result = iceland_get_dependecy_volt_by_clk(hwmgr,
+ hwmgr->dyn_state.mvdd_dependency_on_mclk, memory_clock, &memory_level->MinMvdd);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "can not find MinMVDD voltage value from memory MVDD voltage dependency table", return result);
+ }
+
+ memory_level->MinVddcPhases = 1;
+
+ if (data->vddc_phase_shed_control) {
+ iceland_populate_phase_value_based_on_mclk(hwmgr, hwmgr->dyn_state.vddc_phase_shed_limits_table,
+ memory_clock, &memory_level->MinVddcPhases);
+ }
+
+ memory_level->EnabledForThrottle = 1;
+ memory_level->EnabledForActivity = 1;
+ memory_level->UpHyst = 0;
+ memory_level->DownHyst = 100;
+ memory_level->VoltageDownHyst = 0;
+
+ /* Indicates maximum activity level for this performance level.*/
+ memory_level->ActivityLevel = (uint16_t)data->mclk_activity_target;
+ memory_level->StutterEnable = 0;
+ memory_level->StrobeEnable = 0;
+ memory_level->EdcReadEnable = 0;
+ memory_level->EdcWriteEnable = 0;
+ memory_level->RttEnable = 0;
+
+ /* default set to low watermark. Highest level will be set to high later.*/
+ memory_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
+
+ cgs_get_active_displays_info(hwmgr->device, &info);
+ data->display_timing.num_existing_displays = info.display_count;
+
+ //if ((data->mclk_stutter_mode_threshold != 0) &&
+ // (memory_clock <= data->mclk_stutter_mode_threshold) &&
+ // (data->is_uvd_enabled == 0)
+ // && (PHM_READ_FIELD(hwmgr->device, DPG_PIPE_STUTTER_CONTROL, STUTTER_ENABLE) & 0x1)
+ // && (data->display_timing.num_existing_displays <= 2)
+ // && (data->display_timing.num_existing_displays != 0))
+ // memory_level->StutterEnable = 1;
+
+ /* decide strobe mode*/
+ memory_level->StrobeEnable = (data->mclk_strobe_mode_threshold != 0) &&
+ (memory_clock <= data->mclk_strobe_mode_threshold);
+
+ /* decide EDC mode and memory clock ratio*/
+ if (data->is_memory_GDDR5) {
+ memory_level->StrobeRatio = iceland_get_mclk_frequency_ratio(memory_clock,
+ memory_level->StrobeEnable);
+
+ if ((data->mclk_edc_enable_threshold != 0) &&
+ (memory_clock > data->mclk_edc_enable_threshold)) {
+ memory_level->EdcReadEnable = 1;
+ }
+
+ if ((data->mclk_edc_wr_enable_threshold != 0) &&
+ (memory_clock > data->mclk_edc_wr_enable_threshold)) {
+ memory_level->EdcWriteEnable = 1;
+ }
+
+ if (memory_level->StrobeEnable) {
+ if (iceland_get_mclk_frequency_ratio(memory_clock, 1) >=
+ ((cgs_read_register(hwmgr->device, mmMC_SEQ_MISC7) >> 16) & 0xf)) {
+ dllStateOn = ((cgs_read_register(hwmgr->device, mmMC_SEQ_MISC5) >> 1) & 0x1) ? 1 : 0;
+ } else {
+ dllStateOn = ((cgs_read_register(hwmgr->device, mmMC_SEQ_MISC6) >> 1) & 0x1) ? 1 : 0;
+ }
+
+ } else {
+ dllStateOn = data->dll_defaule_on;
+ }
+ } else {
+ memory_level->StrobeRatio =
+ iceland_get_ddr3_mclk_frequency_ratio(memory_clock);
+ dllStateOn = ((cgs_read_register(hwmgr->device, mmMC_SEQ_MISC5) >> 1) & 0x1) ? 1 : 0;
+ }
+
+ result = iceland_calculate_mclk_params(hwmgr,
+ memory_clock, memory_level, memory_level->StrobeEnable, dllStateOn);
+
+ if (0 == result) {
+ memory_level->MinVddc = PP_HOST_TO_SMC_UL(memory_level->MinVddc * VOLTAGE_SCALE);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MinVddcPhases);
+ memory_level->MinVddci = PP_HOST_TO_SMC_UL(memory_level->MinVddci * VOLTAGE_SCALE);
+ memory_level->MinMvdd = PP_HOST_TO_SMC_UL(memory_level->MinMvdd * VOLTAGE_SCALE);
+ /* MCLK frequency in units of 10KHz*/
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MclkFrequency);
+ /* Indicates maximum activity level for this performance level.*/
+ CONVERT_FROM_HOST_TO_SMC_US(memory_level->ActivityLevel);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MpllFuncCntl);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MpllFuncCntl_1);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MpllFuncCntl_2);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MpllAdFuncCntl);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MpllDqFuncCntl);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MclkPwrmgtCntl);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->DllCntl);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MpllSs1);
+ CONVERT_FROM_HOST_TO_SMC_UL(memory_level->MpllSs2);
+ }
+
+ return result;
+}
+
+/**
+ * Populates the SMC MVDD structure using the provided memory clock.
+ *
+ * @param hwmgr the address of the hardware manager
+ * @param mclk the MCLK value to be used in the decision if MVDD should be high or low.
+ * @param voltage the SMC VOLTAGE structure to be populated
+ */
+int iceland_populate_mvdd_value(struct pp_hwmgr *hwmgr, uint32_t mclk, SMU71_Discrete_VoltageLevel *voltage)
+{
+ const iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ uint32_t i = 0;
+
+ if (ICELAND_VOLTAGE_CONTROL_NONE != data->mvdd_control) {
+ /* find mvdd value which clock is more than request */
+ for (i = 0; i < hwmgr->dyn_state.mvdd_dependency_on_mclk->count; i++) {
+ if (mclk <= hwmgr->dyn_state.mvdd_dependency_on_mclk->entries[i].clk) {
+ /* Always round to higher voltage. */
+ voltage->Voltage = data->mvdd_voltage_table.entries[i].value;
+ break;
+ }
+ }
+
+ PP_ASSERT_WITH_CODE(i < hwmgr->dyn_state.mvdd_dependency_on_mclk->count,
+ "MVDD Voltage is outside the supported range.", return -1);
+
+ } else {
+ return -1;
+ }
+
+ return 0;
+}
+
+
+static int iceland_populate_smc_acpi_level(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_DpmTable *table)
+{
+ int result = 0;
+ const iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ pp_atomctrl_clock_dividers_vi dividers;
+ SMU71_Discrete_VoltageLevel voltage_level;
+ uint32_t spll_func_cntl = data->clock_registers.vCG_SPLL_FUNC_CNTL;
+ uint32_t spll_func_cntl_2 = data->clock_registers.vCG_SPLL_FUNC_CNTL_2;
+ uint32_t dll_cntl = data->clock_registers.vDLL_CNTL;
+ uint32_t mclk_pwrmgt_cntl = data->clock_registers.vMCLK_PWRMGT_CNTL;
+
+ /* The ACPI state should not do DPM on DC (or ever).*/
+ table->ACPILevel.Flags &= ~PPSMC_SWSTATE_FLAG_DC;
+
+ if (data->acpi_vddc)
+ table->ACPILevel.MinVddc = PP_HOST_TO_SMC_UL(data->acpi_vddc * VOLTAGE_SCALE);
+ else
+ table->ACPILevel.MinVddc = PP_HOST_TO_SMC_UL(data->min_vddc_in_pp_table * VOLTAGE_SCALE);
+
+ table->ACPILevel.MinVddcPhases = (data->vddc_phase_shed_control) ? 0 : 1;
+
+ /* assign zero for now*/
+ table->ACPILevel.SclkFrequency = atomctrl_get_reference_clock(hwmgr);
+
+ /* get the engine clock dividers for this clock value*/
+ result = atomctrl_get_engine_pll_dividers_vi(hwmgr,
+ table->ACPILevel.SclkFrequency, ÷rs);
+
+ PP_ASSERT_WITH_CODE(result == 0,
+ "Error retrieving Engine Clock dividers from VBIOS.", return result);
+
+ /* divider ID for required SCLK*/
+ table->ACPILevel.SclkDid = (uint8_t)dividers.pll_post_divider;
+ table->ACPILevel.DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
+ table->ACPILevel.DeepSleepDivId = 0;
+
+ spll_func_cntl = PHM_SET_FIELD(spll_func_cntl,
+ CG_SPLL_FUNC_CNTL, SPLL_PWRON, 0);
+ spll_func_cntl = PHM_SET_FIELD(spll_func_cntl,
+ CG_SPLL_FUNC_CNTL, SPLL_RESET, 1);
+ spll_func_cntl_2 = PHM_SET_FIELD(spll_func_cntl_2,
+ CG_SPLL_FUNC_CNTL_2, SCLK_MUX_SEL, 4);
+
+ table->ACPILevel.CgSpllFuncCntl = spll_func_cntl;
+ table->ACPILevel.CgSpllFuncCntl2 = spll_func_cntl_2;
+ table->ACPILevel.CgSpllFuncCntl3 = data->clock_registers.vCG_SPLL_FUNC_CNTL_3;
+ table->ACPILevel.CgSpllFuncCntl4 = data->clock_registers.vCG_SPLL_FUNC_CNTL_4;
+ table->ACPILevel.SpllSpreadSpectrum = data->clock_registers.vCG_SPLL_SPREAD_SPECTRUM;
+ table->ACPILevel.SpllSpreadSpectrum2 = data->clock_registers.vCG_SPLL_SPREAD_SPECTRUM_2;
+ table->ACPILevel.CcPwrDynRm = 0;
+ table->ACPILevel.CcPwrDynRm1 = 0;
+
+
+ /* For various features to be enabled/disabled while this level is active.*/
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.Flags);
+ /* SCLK frequency in units of 10KHz*/
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.SclkFrequency);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CgSpllFuncCntl);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CgSpllFuncCntl2);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CgSpllFuncCntl3);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CgSpllFuncCntl4);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.SpllSpreadSpectrum);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.SpllSpreadSpectrum2);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CcPwrDynRm);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CcPwrDynRm1);
+
+ table->MemoryACPILevel.MinVddc = table->ACPILevel.MinVddc;
+ table->MemoryACPILevel.MinVddcPhases = table->ACPILevel.MinVddcPhases;
+
+ /* CONVERT_FROM_HOST_TO_SMC_UL(table->MemoryACPILevel.MinVoltage);*/
+
+ if (0 == iceland_populate_mvdd_value(hwmgr, 0, &voltage_level))
+ table->MemoryACPILevel.MinMvdd =
+ PP_HOST_TO_SMC_UL(voltage_level.Voltage * VOLTAGE_SCALE);
+ else
+ table->MemoryACPILevel.MinMvdd = 0;
+
+ /* Force reset on DLL*/
+ mclk_pwrmgt_cntl = PHM_SET_FIELD(mclk_pwrmgt_cntl,
+ MCLK_PWRMGT_CNTL, MRDCK0_RESET, 0x1);
+ mclk_pwrmgt_cntl = PHM_SET_FIELD(mclk_pwrmgt_cntl,
+ MCLK_PWRMGT_CNTL, MRDCK1_RESET, 0x1);
+
+ /* Disable DLL in ACPIState*/
+ mclk_pwrmgt_cntl = PHM_SET_FIELD(mclk_pwrmgt_cntl,
+ MCLK_PWRMGT_CNTL, MRDCK0_PDNB, 0);
+ mclk_pwrmgt_cntl = PHM_SET_FIELD(mclk_pwrmgt_cntl,
+ MCLK_PWRMGT_CNTL, MRDCK1_PDNB, 0);
+
+ /* Enable DLL bypass signal*/
+ dll_cntl = PHM_SET_FIELD(dll_cntl,
+ DLL_CNTL, MRDCK0_BYPASS, 0);
+ dll_cntl = PHM_SET_FIELD(dll_cntl,
+ DLL_CNTL, MRDCK1_BYPASS, 0);
+
+ table->MemoryACPILevel.DllCntl =
+ PP_HOST_TO_SMC_UL(dll_cntl);
+ table->MemoryACPILevel.MclkPwrmgtCntl =
+ PP_HOST_TO_SMC_UL(mclk_pwrmgt_cntl);
+ table->MemoryACPILevel.MpllAdFuncCntl =
+ PP_HOST_TO_SMC_UL(data->clock_registers.vMPLL_AD_FUNC_CNTL);
+ table->MemoryACPILevel.MpllDqFuncCntl =
+ PP_HOST_TO_SMC_UL(data->clock_registers.vMPLL_DQ_FUNC_CNTL);
+ table->MemoryACPILevel.MpllFuncCntl =
+ PP_HOST_TO_SMC_UL(data->clock_registers.vMPLL_FUNC_CNTL);
+ table->MemoryACPILevel.MpllFuncCntl_1 =
+ PP_HOST_TO_SMC_UL(data->clock_registers.vMPLL_FUNC_CNTL_1);
+ table->MemoryACPILevel.MpllFuncCntl_2 =
+ PP_HOST_TO_SMC_UL(data->clock_registers.vMPLL_FUNC_CNTL_2);
+ table->MemoryACPILevel.MpllSs1 =
+ PP_HOST_TO_SMC_UL(data->clock_registers.vMPLL_SS1);
+ table->MemoryACPILevel.MpllSs2 =
+ PP_HOST_TO_SMC_UL(data->clock_registers.vMPLL_SS2);
+
+ table->MemoryACPILevel.EnabledForThrottle = 0;
+ table->MemoryACPILevel.EnabledForActivity = 0;
+ table->MemoryACPILevel.UpHyst = 0;
+ table->MemoryACPILevel.DownHyst = 100;
+ table->MemoryACPILevel.VoltageDownHyst = 0;
+ /* Indicates maximum activity level for this performance level.*/
+ table->MemoryACPILevel.ActivityLevel = PP_HOST_TO_SMC_US((uint16_t)data->mclk_activity_target);
+
+ table->MemoryACPILevel.StutterEnable = 0;
+ table->MemoryACPILevel.StrobeEnable = 0;
+ table->MemoryACPILevel.EdcReadEnable = 0;
+ table->MemoryACPILevel.EdcWriteEnable = 0;
+ table->MemoryACPILevel.RttEnable = 0;
+
+ return result;
+}
+
+static int iceland_find_boot_level(struct iceland_single_dpm_table *table, uint32_t value, uint32_t *boot_level)
+{
+ int result = 0;
+ uint32_t i;
+
+ for (i = 0; i < table->count; i++) {
+ if (value == table->dpm_levels[i].value) {
+ *boot_level = i;
+ result = 0;
+ }
+ }
+ return result;
+}
+
+/**
+ * Calculates the SCLK dividers using the provided engine clock
+ *
+ * @param hwmgr the address of the hardware manager
+ * @param engine_clock the engine clock to use to populate the structure
+ * @param sclk the SMC SCLK structure to be populated
+ */
+int iceland_calculate_sclk_params(struct pp_hwmgr *hwmgr,
+ uint32_t engine_clock, SMU71_Discrete_GraphicsLevel *sclk)
+{
+ const iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ pp_atomctrl_clock_dividers_vi dividers;
+ uint32_t spll_func_cntl = data->clock_registers.vCG_SPLL_FUNC_CNTL;
+ uint32_t spll_func_cntl_3 = data->clock_registers.vCG_SPLL_FUNC_CNTL_3;
+ uint32_t spll_func_cntl_4 = data->clock_registers.vCG_SPLL_FUNC_CNTL_4;
+ uint32_t cg_spll_spread_spectrum = data->clock_registers.vCG_SPLL_SPREAD_SPECTRUM;
+ uint32_t cg_spll_spread_spectrum_2 = data->clock_registers.vCG_SPLL_SPREAD_SPECTRUM_2;
+ uint32_t reference_clock;
+ uint32_t reference_divider;
+ uint32_t fbdiv;
+ int result;
+
+ /* get the engine clock dividers for this clock value*/
+ result = atomctrl_get_engine_pll_dividers_vi(hwmgr, engine_clock, ÷rs);
+
+ PP_ASSERT_WITH_CODE(result == 0,
+ "Error retrieving Engine Clock dividers from VBIOS.", return result);
+
+ /* To get FBDIV we need to multiply this by 16384 and divide it by Fref.*/
+ reference_clock = atomctrl_get_reference_clock(hwmgr);
+
+ reference_divider = 1 + dividers.uc_pll_ref_div;
+
+ /* low 14 bits is fraction and high 12 bits is divider*/
+ fbdiv = dividers.ul_fb_div.ul_fb_divider & 0x3FFFFFF;
+
+ /* SPLL_FUNC_CNTL setup*/
+ spll_func_cntl = PHM_SET_FIELD(spll_func_cntl,
+ CG_SPLL_FUNC_CNTL, SPLL_REF_DIV, dividers.uc_pll_ref_div);
+ spll_func_cntl = PHM_SET_FIELD(spll_func_cntl,
+ CG_SPLL_FUNC_CNTL, SPLL_PDIV_A, dividers.uc_pll_post_div);
+
+ /* SPLL_FUNC_CNTL_3 setup*/
+ spll_func_cntl_3 = PHM_SET_FIELD(spll_func_cntl_3,
+ CG_SPLL_FUNC_CNTL_3, SPLL_FB_DIV, fbdiv);
+
+ /* set to use fractional accumulation*/
+ spll_func_cntl_3 = PHM_SET_FIELD(spll_func_cntl_3,
+ CG_SPLL_FUNC_CNTL_3, SPLL_DITHEN, 1);
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EngineSpreadSpectrumSupport)) {
+ pp_atomctrl_internal_ss_info ss_info;
+
+ uint32_t vcoFreq = engine_clock * dividers.uc_pll_post_div;
+ if (0 == atomctrl_get_engine_clock_spread_spectrum(hwmgr, vcoFreq, &ss_info)) {
+ /*
+ * ss_info.speed_spectrum_percentage -- in unit of 0.01%
+ * ss_info.speed_spectrum_rate -- in unit of khz
+ */
+ /* clks = reference_clock * 10 / (REFDIV + 1) / speed_spectrum_rate / 2 */
+ uint32_t clkS = reference_clock * 5 / (reference_divider * ss_info.speed_spectrum_rate);
+
+ /* clkv = 2 * D * fbdiv / NS */
+ uint32_t clkV = 4 * ss_info.speed_spectrum_percentage * fbdiv / (clkS * 10000);
+
+ cg_spll_spread_spectrum =
+ PHM_SET_FIELD(cg_spll_spread_spectrum, CG_SPLL_SPREAD_SPECTRUM, CLKS, clkS);
+ cg_spll_spread_spectrum =
+ PHM_SET_FIELD(cg_spll_spread_spectrum, CG_SPLL_SPREAD_SPECTRUM, SSEN, 1);
+ cg_spll_spread_spectrum_2 =
+ PHM_SET_FIELD(cg_spll_spread_spectrum_2, CG_SPLL_SPREAD_SPECTRUM_2, CLKV, clkV);
+ }
+ }
+
+ sclk->SclkFrequency = engine_clock;
+ sclk->CgSpllFuncCntl3 = spll_func_cntl_3;
+ sclk->CgSpllFuncCntl4 = spll_func_cntl_4;
+ sclk->SpllSpreadSpectrum = cg_spll_spread_spectrum;
+ sclk->SpllSpreadSpectrum2 = cg_spll_spread_spectrum_2;
+ sclk->SclkDid = (uint8_t)dividers.pll_post_divider;
+
+ return 0;
+}
+
+static uint8_t iceland_get_sleep_divider_id_from_clock(struct pp_hwmgr *hwmgr,
+ uint32_t engine_clock, uint32_t min_engine_clock_in_sr)
+{
+ uint32_t i, temp;
+ uint32_t min = (min_engine_clock_in_sr > ICELAND_MINIMUM_ENGINE_CLOCK) ?
+ min_engine_clock_in_sr : ICELAND_MINIMUM_ENGINE_CLOCK;
+
+ PP_ASSERT_WITH_CODE((engine_clock >= min),
+ "Engine clock can't satisfy stutter requirement!", return 0);
+
+ for (i = ICELAND_MAX_DEEPSLEEP_DIVIDER_ID;; i--) {
+ temp = engine_clock / (1 << i);
+
+ if(temp >= min || i == 0)
+ break;
+ }
+ return (uint8_t)i;
+}
+
+/**
+ * Populates single SMC SCLK structure using the provided engine clock
+ *
+ * @param hwmgr the address of the hardware manager
+ * @param engine_clock the engine clock to use to populate the structure
+ * @param sclk the SMC SCLK structure to be populated
+ */
+static int iceland_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
+ uint32_t engine_clock, uint16_t sclk_activity_level_threshold,
+ SMU71_Discrete_GraphicsLevel *graphic_level)
+{
+ int result;
+ uint32_t threshold;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ result = iceland_calculate_sclk_params(hwmgr, engine_clock, graphic_level);
+
+
+ /* populate graphics levels*/
+ result = iceland_get_dependecy_volt_by_clk(hwmgr,
+ hwmgr->dyn_state.vddc_dependency_on_sclk, engine_clock, &graphic_level->MinVddc);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "can not find VDDC voltage value for VDDC engine clock dependency table", return result);
+
+ /* SCLK frequency in units of 10KHz*/
+ graphic_level->SclkFrequency = engine_clock;
+
+ /*
+ * Minimum VDDC phases required to support this level, it
+ * should get from dependence table.
+ */
+ graphic_level->MinVddcPhases = 1;
+
+ if (data->vddc_phase_shed_control) {
+ iceland_populate_phase_value_based_on_sclk(hwmgr,
+ hwmgr->dyn_state.vddc_phase_shed_limits_table,
+ engine_clock,
+ &graphic_level->MinVddcPhases);
+ }
+
+ /* Indicates maximum activity level for this performance level. 50% for now*/
+ graphic_level->ActivityLevel = sclk_activity_level_threshold;
+
+ graphic_level->CcPwrDynRm = 0;
+ graphic_level->CcPwrDynRm1 = 0;
+ /* this level can be used if activity is high enough.*/
+ graphic_level->EnabledForActivity = 1;
+ /* this level can be used for throttling.*/
+ graphic_level->EnabledForThrottle = 1;
+ graphic_level->UpHyst = 0;
+ graphic_level->DownHyst = 100;
+ graphic_level->VoltageDownHyst = 0;
+ graphic_level->PowerThrottle = 0;
+
+ threshold = engine_clock * data->fast_watermark_threshold / 100;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SclkDeepSleep)) {
+ graphic_level->DeepSleepDivId =
+ iceland_get_sleep_divider_id_from_clock(hwmgr, engine_clock,
+ data->display_timing.min_clock_insr);
+ }
+
+ /* Default to slow, highest DPM level will be set to PPSMC_DISPLAY_WATERMARK_LOW later.*/
+ graphic_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
+
+ if (0 == result) {
+ graphic_level->MinVddc = PP_HOST_TO_SMC_UL(graphic_level->MinVddc * VOLTAGE_SCALE);
+ /* CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->MinVoltage);*/
+ CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->MinVddcPhases);
+ CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->SclkFrequency);
+ CONVERT_FROM_HOST_TO_SMC_US(graphic_level->ActivityLevel);
+ CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->CgSpllFuncCntl3);
+ CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->CgSpllFuncCntl4);
+ CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->SpllSpreadSpectrum);
+ CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->SpllSpreadSpectrum2);
+ CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->CcPwrDynRm);
+ CONVERT_FROM_HOST_TO_SMC_UL(graphic_level->CcPwrDynRm1);
+ }
+
+ return result;
+}
+
+/**
+ * Populates all SMC SCLK levels' structure based on the trimmed allowed dpm engine clock states
+ *
+ * @param hwmgr the address of the hardware manager
+ */
+static int iceland_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_dpm_table *dpm_table = &data->dpm_table;
+ int result = 0;
+ uint32_t level_array_adress = data->dpm_table_start +
+ offsetof(SMU71_Discrete_DpmTable, GraphicsLevel);
+
+ uint32_t level_array_size = sizeof(SMU71_Discrete_GraphicsLevel) * SMU71_MAX_LEVELS_GRAPHICS;
+ SMU71_Discrete_GraphicsLevel *levels = data->smc_state_table.GraphicsLevel;
+ uint32_t i;
+ uint8_t highest_pcie_level_enabled = 0, lowest_pcie_level_enabled = 0, mid_pcie_level_enabled = 0, count = 0;
+ memset(levels, 0x00, level_array_size);
+
+ for (i = 0; i < dpm_table->sclk_table.count; i++) {
+ result = iceland_populate_single_graphic_level(hwmgr,
+ dpm_table->sclk_table.dpm_levels[i].value,
+ (uint16_t)data->activity_target[i],
+ &(data->smc_state_table.GraphicsLevel[i]));
+ if (0 != result)
+ return result;
+
+ /* Making sure only DPM level 0-1 have Deep Sleep Div ID populated. */
+ if (i > 1)
+ data->smc_state_table.GraphicsLevel[i].DeepSleepDivId = 0;
+ }
+
+ /* set highest level watermark to high */
+ if (dpm_table->sclk_table.count > 1)
+ data->smc_state_table.GraphicsLevel[dpm_table->sclk_table.count-1].DisplayWatermark =
+ PPSMC_DISPLAY_WATERMARK_HIGH;
+
+ data->smc_state_table.GraphicsDpmLevelCount =
+ (uint8_t)dpm_table->sclk_table.count;
+ data->dpm_level_enable_mask.sclk_dpm_enable_mask =
+ iceland_get_dpm_level_enable_mask_value(&dpm_table->sclk_table);
+
+ while ((data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+ (1 << (highest_pcie_level_enabled + 1))) != 0) {
+ highest_pcie_level_enabled++;
+ }
+
+ while ((data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+ (1 << lowest_pcie_level_enabled)) == 0) {
+ lowest_pcie_level_enabled++;
+ }
+
+ while ((count < highest_pcie_level_enabled) &&
+ ((data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+ (1 << (lowest_pcie_level_enabled + 1 + count))) == 0)) {
+ count++;
+ }
+
+ mid_pcie_level_enabled = (lowest_pcie_level_enabled+1+count) < highest_pcie_level_enabled ?
+ (lowest_pcie_level_enabled + 1 + count) : highest_pcie_level_enabled;
+
+ /* set pcieDpmLevel to highest_pcie_level_enabled*/
+ for (i = 2; i < dpm_table->sclk_table.count; i++) {
+ data->smc_state_table.GraphicsLevel[i].pcieDpmLevel = highest_pcie_level_enabled;
+ }
+
+ /* set pcieDpmLevel to lowest_pcie_level_enabled*/
+ data->smc_state_table.GraphicsLevel[0].pcieDpmLevel = lowest_pcie_level_enabled;
+
+ /* set pcieDpmLevel to mid_pcie_level_enabled*/
+ data->smc_state_table.GraphicsLevel[1].pcieDpmLevel = mid_pcie_level_enabled;
+
+ /* level count will send to smc once at init smc table and never change*/
+ result = iceland_copy_bytes_to_smc(hwmgr->smumgr, level_array_adress, (uint8_t *)levels, (uint32_t)level_array_size, data->sram_end);
+
+ if (0 != result)
+ return result;
+
+ return 0;
+}
+
+/**
+ * Populates all SMC MCLK levels' structure based on the trimmed allowed dpm memory clock states
+ *
+ * @param hwmgr the address of the hardware manager
+ */
+
+static int iceland_populate_all_memory_levels(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_dpm_table *dpm_table = &data->dpm_table;
+ int result;
+ /* populate MCLK dpm table to SMU7 */
+ uint32_t level_array_adress = data->dpm_table_start + offsetof(SMU71_Discrete_DpmTable, MemoryLevel);
+ uint32_t level_array_size = sizeof(SMU71_Discrete_MemoryLevel) * SMU71_MAX_LEVELS_MEMORY;
+ SMU71_Discrete_MemoryLevel *levels = data->smc_state_table.MemoryLevel;
+ uint32_t i;
+
+ memset(levels, 0x00, level_array_size);
+
+ for (i = 0; i < dpm_table->mclk_table.count; i++) {
+ PP_ASSERT_WITH_CODE((0 != dpm_table->mclk_table.dpm_levels[i].value),
+ "can not populate memory level as memory clock is zero", return -1);
+ result = iceland_populate_single_memory_level(hwmgr, dpm_table->mclk_table.dpm_levels[i].value,
+ &(data->smc_state_table.MemoryLevel[i]));
+ if (0 != result) {
+ return result;
+ }
+ }
+
+ /* Only enable level 0 for now.*/
+ data->smc_state_table.MemoryLevel[0].EnabledForActivity = 1;
+
+ /*
+ * in order to prevent MC activity from stutter mode to push DPM up.
+ * the UVD change complements this by putting the MCLK in a higher state
+ * by default such that we are not effected by up threshold or and MCLK DPM latency.
+ */
+ data->smc_state_table.MemoryLevel[0].ActivityLevel = 0x1F;
+ CONVERT_FROM_HOST_TO_SMC_US(data->smc_state_table.MemoryLevel[0].ActivityLevel);
+
+ data->smc_state_table.MemoryDpmLevelCount = (uint8_t)dpm_table->mclk_table.count;
+ data->dpm_level_enable_mask.mclk_dpm_enable_mask = iceland_get_dpm_level_enable_mask_value(&dpm_table->mclk_table);
+ /* set highest level watermark to high*/
+ data->smc_state_table.MemoryLevel[dpm_table->mclk_table.count-1].DisplayWatermark = PPSMC_DISPLAY_WATERMARK_HIGH;
+
+ /* level count will send to smc once at init smc table and never change*/
+ result = iceland_copy_bytes_to_smc(hwmgr->smumgr,
+ level_array_adress, (uint8_t *)levels, (uint32_t)level_array_size, data->sram_end);
+
+ if (0 != result) {
+ return result;
+ }
+
+ return 0;
+}
+
+struct ICELAND_DLL_SPEED_SETTING
+{
+ uint16_t Min; /* Minimum Data Rate*/
+ uint16_t Max; /* Maximum Data Rate*/
+ uint32_t dll_speed; /* The desired DLL_SPEED setting*/
+};
+
+static int iceland_populate_ulv_level(struct pp_hwmgr *hwmgr, SMU71_Discrete_Ulv *pstate)
+{
+ int result = 0;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ uint32_t voltage_response_time, ulv_voltage;
+
+ pstate->CcPwrDynRm = 0;
+ pstate->CcPwrDynRm1 = 0;
+
+ //backbiasResponseTime is use for ULV state voltage value.
+ result = pp_tables_get_response_times(hwmgr, &voltage_response_time, &ulv_voltage);
+ PP_ASSERT_WITH_CODE((0 == result), "can not get ULV voltage value", return result;);
+
+ if(!ulv_voltage) {
+ data->ulv.ulv_supported = false;
+ return 0;
+ }
+
+ if (ICELAND_VOLTAGE_CONTROL_BY_SVID2 != data->voltage_control) {
+ /* use minimum voltage if ulv voltage in pptable is bigger than minimum voltage */
+ if (ulv_voltage > hwmgr->dyn_state.vddc_dependency_on_sclk->entries[0].v) {
+ pstate->VddcOffset = 0;
+ }
+ else {
+ /* used in SMIO Mode. not implemented for now. this is backup only for CI. */
+ pstate->VddcOffset = (uint16_t)(hwmgr->dyn_state.vddc_dependency_on_sclk->entries[0].v - ulv_voltage);
+ }
+ } else {
+ /* use minimum voltage if ulv voltage in pptable is bigger than minimum voltage */
+ if(ulv_voltage > hwmgr->dyn_state.vddc_dependency_on_sclk->entries[0].v) {
+ pstate->VddcOffsetVid = 0;
+ } else {
+ /* used in SVI2 Mode */
+ pstate->VddcOffsetVid = (uint8_t)((hwmgr->dyn_state.vddc_dependency_on_sclk->entries[0].v - ulv_voltage) * VOLTAGE_VID_OFFSET_SCALE2 / VOLTAGE_VID_OFFSET_SCALE1);
+ }
+ }
+
+ /* used in SVI2 Mode to shed phase */
+ pstate->VddcPhase = (data->vddc_phase_shed_control) ? 0 : 1;
+
+ if (0 == result) {
+ CONVERT_FROM_HOST_TO_SMC_UL(pstate->CcPwrDynRm);
+ CONVERT_FROM_HOST_TO_SMC_UL(pstate->CcPwrDynRm1);
+ CONVERT_FROM_HOST_TO_SMC_US(pstate->VddcOffset);
+ }
+
+ return result;
+}
+
+static int iceland_populate_ulv_state(struct pp_hwmgr *hwmgr, SMU71_Discrete_Ulv *ulv)
+{
+ return iceland_populate_ulv_level(hwmgr, ulv);
+}
+
+static int iceland_populate_smc_initial_state(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ uint8_t count, level;
+
+ count = (uint8_t)(hwmgr->dyn_state.vddc_dependency_on_sclk->count);
+
+ for (level = 0; level < count; level++) {
+ if (hwmgr->dyn_state.vddc_dependency_on_sclk->entries[level].clk
+ >= data->vbios_boot_state.sclk_bootup_value) {
+ data->smc_state_table.GraphicsBootLevel = level;
+ break;
+ }
+ }
+
+ count = (uint8_t)(hwmgr->dyn_state.vddc_dependency_on_mclk->count);
+
+ for (level = 0; level < count; level++) {
+ if (hwmgr->dyn_state.vddc_dependency_on_mclk->entries[level].clk
+ >= data->vbios_boot_state.mclk_bootup_value) {
+ data->smc_state_table.MemoryBootLevel = level;
+ break;
+ }
+ }
+
+ return 0;
+}
+
+/**
+ * Initializes the SMC table and uploads it
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @param pInput the pointer to input data (PowerState)
+ * @return always 0
+ */
+static int iceland_init_smc_table(struct pp_hwmgr *hwmgr)
+{
+ int result;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ SMU71_Discrete_DpmTable *table = &(data->smc_state_table);
+ const struct phw_iceland_ulv_parm *ulv = &(data->ulv);
+
+ result = iceland_setup_default_dpm_tables(hwmgr);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to setup default DPM tables!", return result;);
+ memset(&(data->smc_state_table), 0x00, sizeof(data->smc_state_table));
+
+ if (ICELAND_VOLTAGE_CONTROL_NONE != data->voltage_control) {
+ iceland_populate_smc_voltage_tables(hwmgr, table);
+ }
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_AutomaticDCTransition)) {
+ table->SystemFlags |= PPSMC_SYSTEMFLAG_GPIO_DC;
+ }
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_StepVddc)) {
+ table->SystemFlags |= PPSMC_SYSTEMFLAG_STEPVDDC;
+ }
+
+ if (data->is_memory_GDDR5) {
+ table->SystemFlags |= PPSMC_SYSTEMFLAG_GDDR5;
+ }
+
+ if (ulv->ulv_supported) {
+ result = iceland_populate_ulv_state(hwmgr, &data->ulv_setting);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize ULV state!", return result;);
+
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCG_ULV_PARAMETER, ulv->ch_ulv_parameter);
+ }
+
+ result = iceland_populate_smc_link_level(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize Link Level!", return result;);
+
+ result = iceland_populate_all_graphic_levels(hwmgr);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize Graphics Level!", return result;);
+
+ result = iceland_populate_all_memory_levels(hwmgr);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize Memory Level!", return result;);
+
+ result = iceland_populate_smc_acpi_level(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize ACPI Level!", return result;);
+
+ result = iceland_populate_smc_vce_level(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize VCE Level!", return result;);
+
+ result = iceland_populate_smc_acp_level(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize ACP Level!", return result;);
+
+ result = iceland_populate_smc_samu_level(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize SAMU Level!", return result;);
+
+ /*
+ * Since only the initial state is completely set up at this
+ * point (the other states are just copies of the boot state)
+ * we only need to populate the ARB settings for the initial
+ * state.
+ */
+ result = iceland_program_memory_timing_parameters(hwmgr);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to Write ARB settings for the initial state.", return result;);
+
+ result = iceland_populate_smc_uvd_level(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize UVD Level!", return result;);
+
+ table->GraphicsBootLevel = 0;
+ table->MemoryBootLevel = 0;
+
+ /* find boot level from dpm table */
+ result = iceland_find_boot_level(&(data->dpm_table.sclk_table),
+ data->vbios_boot_state.sclk_bootup_value,
+ (uint32_t *)&(data->smc_state_table.GraphicsBootLevel));
+
+ if (result)
+ pr_warning("VBIOS did not find boot engine clock value in dependency table.\n");
+
+ result = iceland_find_boot_level(&(data->dpm_table.mclk_table),
+ data->vbios_boot_state.mclk_bootup_value,
+ (uint32_t *)&(data->smc_state_table.MemoryBootLevel));
+
+ if (result)
+ pr_warning("VBIOS did not find boot memory clock value in dependency table.\n");
+
+ table->BootVddc = data->vbios_boot_state.vddc_bootup_value;
+ if (ICELAND_VOLTAGE_CONTROL_NONE == data->vdd_ci_control) {
+ table->BootVddci = table->BootVddc;
+ }
+ else {
+ table->BootVddci = data->vbios_boot_state.vddci_bootup_value;
+ }
+ table->BootMVdd = data->vbios_boot_state.mvdd_bootup_value;
+
+ result = iceland_populate_smc_initial_state(hwmgr);
+ PP_ASSERT_WITH_CODE(0 == result, "Failed to initialize Boot State!", return result);
+
+ result = iceland_populate_bapm_parameters_in_dpm_table(hwmgr);
+ PP_ASSERT_WITH_CODE(0 == result, "Failed to populate BAPM Parameters!", return result);
+
+ table->GraphicsVoltageChangeEnable = 1;
+ table->GraphicsThermThrottleEnable = 1;
+ table->GraphicsInterval = 1;
+ table->VoltageInterval = 1;
+ table->ThermalInterval = 1;
+ table->TemperatureLimitHigh =
+ (data->thermal_temp_setting.temperature_high *
+ ICELAND_Q88_FORMAT_CONVERSION_UNIT) / PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+ table->TemperatureLimitLow =
+ (data->thermal_temp_setting.temperature_low *
+ ICELAND_Q88_FORMAT_CONVERSION_UNIT) / PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+ table->MemoryVoltageChangeEnable = 1;
+ table->MemoryInterval = 1;
+ table->VoltageResponseTime = 0;
+ table->PhaseResponseTime = 0;
+ table->MemoryThermThrottleEnable = 1;
+ table->PCIeBootLinkLevel = 0;
+ table->PCIeGenInterval = 1;
+
+ result = iceland_populate_smc_svi2_config(hwmgr, table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to populate SVI2 setting!", return result);
+
+ table->ThermGpio = 17;
+ table->SclkStepSize = 0x4000;
+
+ CONVERT_FROM_HOST_TO_SMC_UL(table->SystemFlags);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->SmioMaskVddcVid);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->SmioMaskVddcPhase);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->SmioMaskVddciVid);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->SmioMaskMvddVid);
+ CONVERT_FROM_HOST_TO_SMC_UL(table->SclkStepSize);
+ CONVERT_FROM_HOST_TO_SMC_US(table->TemperatureLimitHigh);
+ CONVERT_FROM_HOST_TO_SMC_US(table->TemperatureLimitLow);
+ CONVERT_FROM_HOST_TO_SMC_US(table->VoltageResponseTime);
+ CONVERT_FROM_HOST_TO_SMC_US(table->PhaseResponseTime);
+
+ table->BootVddc = PP_HOST_TO_SMC_US(table->BootVddc * VOLTAGE_SCALE);
+ table->BootVddci = PP_HOST_TO_SMC_US(table->BootVddci * VOLTAGE_SCALE);
+ table->BootMVdd = PP_HOST_TO_SMC_US(table->BootMVdd * VOLTAGE_SCALE);
+
+ /* Upload all dpm data to SMC memory.(dpm level, dpm level count etc) */
+ result = iceland_copy_bytes_to_smc(hwmgr->smumgr, data->dpm_table_start +
+ offsetof(SMU71_Discrete_DpmTable, SystemFlags),
+ (uint8_t *)&(table->SystemFlags),
+ sizeof(SMU71_Discrete_DpmTable) - 3 * sizeof(SMU71_PIDController),
+ data->sram_end);
+
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to upload dpm data to SMC memory!", return result);
+
+ /* Upload all ulv setting to SMC memory.(dpm level, dpm level count etc) */
+ result = iceland_copy_bytes_to_smc(hwmgr->smumgr,
+ data->ulv_settings_start,
+ (uint8_t *)&(data->ulv_setting),
+ sizeof(SMU71_Discrete_Ulv),
+ data->sram_end);
+
+#if 0
+ /* Notify SMC to follow new GPIO scheme */
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_AutomaticDCTransition)) {
+ if (0 == iceland_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_UseNewGPIOScheme))
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SMCtoPPLIBAcdcGpioScheme);
+ }
+#endif
+
+ return result;
+}
+
+int iceland_populate_mc_reg_address(struct pp_hwmgr *hwmgr, SMU71_Discrete_MCRegisters *mc_reg_table)
+{
+ const struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ uint32_t i, j;
+
+ for (i = 0, j = 0; j < data->iceland_mc_reg_table.last; j++) {
+ if (data->iceland_mc_reg_table.validflag & 1<<j) {
+ PP_ASSERT_WITH_CODE(i < SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE,
+ "Index of mc_reg_table->address[] array out of boundary", return -1);
+ mc_reg_table->address[i].s0 =
+ PP_HOST_TO_SMC_US(data->iceland_mc_reg_table.mc_reg_address[j].s0);
+ mc_reg_table->address[i].s1 =
+ PP_HOST_TO_SMC_US(data->iceland_mc_reg_table.mc_reg_address[j].s1);
+ i++;
+ }
+ }
+
+ mc_reg_table->last = (uint8_t)i;
+
+ return 0;
+}
+
+/* convert register values from driver to SMC format */
+void iceland_convert_mc_registers(
+ const phw_iceland_mc_reg_entry * pEntry,
+ SMU71_Discrete_MCRegisterSet *pData,
+ uint32_t numEntries, uint32_t validflag)
+{
+ uint32_t i, j;
+
+ for (i = 0, j = 0; j < numEntries; j++) {
+ if (validflag & 1<<j) {
+ pData->value[i] = PP_HOST_TO_SMC_UL(pEntry->mc_data[j]);
+ i++;
+ }
+ }
+}
+
+/* find the entry in the memory range table, then populate the value to SMC's iceland_mc_reg_table */
+int iceland_convert_mc_reg_table_entry_to_smc(
+ struct pp_hwmgr *hwmgr,
+ const uint32_t memory_clock,
+ SMU71_Discrete_MCRegisterSet *mc_reg_table_data
+ )
+{
+ const iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint32_t i = 0;
+
+ for (i = 0; i < data->iceland_mc_reg_table.num_entries; i++) {
+ if (memory_clock <=
+ data->iceland_mc_reg_table.mc_reg_table_entry[i].mclk_max) {
+ break;
+ }
+ }
+
+ if ((i == data->iceland_mc_reg_table.num_entries) && (i > 0))
+ --i;
+
+ iceland_convert_mc_registers(&data->iceland_mc_reg_table.mc_reg_table_entry[i],
+ mc_reg_table_data, data->iceland_mc_reg_table.last, data->iceland_mc_reg_table.validflag);
+
+ return 0;
+}
+
+int iceland_convert_mc_reg_table_to_smc(struct pp_hwmgr *hwmgr,
+ SMU71_Discrete_MCRegisters *mc_reg_table)
+{
+ int result = 0;
+ iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ int res;
+ uint32_t i;
+
+ for (i = 0; i < data->dpm_table.mclk_table.count; i++) {
+ res = iceland_convert_mc_reg_table_entry_to_smc(
+ hwmgr,
+ data->dpm_table.mclk_table.dpm_levels[i].value,
+ &mc_reg_table->data[i]
+ );
+
+ if (0 != res)
+ result = res;
+ }
+
+ return result;
+}
+
+int iceland_populate_initial_mc_reg_table(struct pp_hwmgr *hwmgr)
+{
+ int result;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ memset(&data->mc_reg_table, 0x00, sizeof(SMU71_Discrete_MCRegisters));
+ result = iceland_populate_mc_reg_address(hwmgr, &(data->mc_reg_table));
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize MCRegTable for the MC register addresses!", return result;);
+
+ result = iceland_convert_mc_reg_table_to_smc(hwmgr, &data->mc_reg_table);
+ PP_ASSERT_WITH_CODE(0 == result,
+ "Failed to initialize MCRegTable for driver state!", return result;);
+
+ return iceland_copy_bytes_to_smc(hwmgr->smumgr, data->mc_reg_table_start,
+ (uint8_t *)&data->mc_reg_table, sizeof(SMU71_Discrete_MCRegisters), data->sram_end);
+}
+
+int iceland_notify_smc_display_change(struct pp_hwmgr *hwmgr, bool has_display)
+{
+ PPSMC_Msg msg = has_display? (PPSMC_Msg)PPSMC_HasDisplay : (PPSMC_Msg)PPSMC_NoDisplay;
+
+ return (smum_send_msg_to_smc(hwmgr->smumgr, msg) == 0) ? 0 : -1;
+}
+
+int iceland_enable_sclk_control(struct pp_hwmgr *hwmgr)
+{
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SCLK_PWRMGT_CNTL, SCLK_PWRMGT_OFF, 0);
+
+ return 0;
+}
+
+int iceland_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ /* enable SCLK dpm */
+ if (0 == data->sclk_dpm_key_disabled) {
+ PP_ASSERT_WITH_CODE(
+ (0 == smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_DPM_Enable)),
+ "Failed to enable SCLK DPM during DPM Start Function!",
+ return -1);
+ }
+
+ /* enable MCLK dpm */
+ if (0 == data->mclk_dpm_key_disabled) {
+ PP_ASSERT_WITH_CODE(
+ (0 == smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_MCLKDPM_Enable)),
+ "Failed to enable MCLK DPM during DPM Start Function!",
+ return -1);
+
+ PHM_WRITE_FIELD(hwmgr->device, MC_SEQ_CNTL_3, CAC_EN, 0x1);
+
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixLCAC_MC0_CNTL, 0x05);/* CH0,1 read */
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixLCAC_MC1_CNTL, 0x05);/* CH2,3 read */
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixLCAC_CPL_CNTL, 0x100005);/*Read */
+
+ udelay(10);
+
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixLCAC_MC0_CNTL, 0x400005);/* CH0,1 write */
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixLCAC_MC1_CNTL, 0x400005);/* CH2,3 write */
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixLCAC_CPL_CNTL, 0x500005);/* write */
+
+ }
+
+ return 0;
+}
+
+int iceland_start_dpm(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ /* enable general power management */
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT, GLOBAL_PWRMGT_EN, 1);
+ /* enable sclk deep sleep */
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SCLK_PWRMGT_CNTL, DYNAMIC_PM_EN, 1);
+
+ /* prepare for PCIE DPM */
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SOFT_REGISTERS_TABLE_12, VoltageChangeTimeout, 0x1000);
+
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__PCIE, SWRST_COMMAND_1, RESETLC, 0x0);
+
+ PP_ASSERT_WITH_CODE(
+ (0 == smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_Voltage_Cntl_Enable)),
+ "Failed to enable voltage DPM during DPM Start Function!",
+ return -1);
+
+ if (0 != iceland_enable_sclk_mclk_dpm(hwmgr)) {
+ PP_ASSERT_WITH_CODE(0, "Failed to enable Sclk DPM and Mclk DPM!", return -1);
+ }
+
+ /* enable PCIE dpm */
+ if (0 == data->pcie_dpm_key_disabled) {
+ PP_ASSERT_WITH_CODE(
+ (0 == smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_PCIeDPM_Enable)),
+ "Failed to enable pcie DPM during DPM Start Function!",
+ return -1
+ );
+ }
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_Falcon_QuickTransition)) {
+ smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_EnableACDCGPIOInterrupt);
+ }
+
+ return 0;
+}
+
+static void iceland_set_dpm_event_sources(struct pp_hwmgr *hwmgr,
+ uint32_t sources)
+{
+ bool protection;
+ enum DPM_EVENT_SRC src;
+
+ switch (sources) {
+ default:
+ printk(KERN_ERR "Unknown throttling event sources.");
+ /* fall through */
+ case 0:
+ protection = false;
+ /* src is unused */
+ break;
+ case (1 << PHM_AutoThrottleSource_Thermal):
+ protection = true;
+ src = DPM_EVENT_SRC_DIGITAL;
+ break;
+ case (1 << PHM_AutoThrottleSource_External):
+ protection = true;
+ src = DPM_EVENT_SRC_EXTERNAL;
+ break;
+ case (1 << PHM_AutoThrottleSource_External) |
+ (1 << PHM_AutoThrottleSource_Thermal):
+ protection = true;
+ src = DPM_EVENT_SRC_DIGITAL_OR_EXTERNAL;
+ break;
+ }
+ /* Order matters - don't enable thermal protection for the wrong source. */
+ if (protection) {
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_CTRL,
+ DPM_EVENT_SRC, src);
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT,
+ THERMAL_PROTECTION_DIS,
+ !phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_ThermalController));
+ } else
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT,
+ THERMAL_PROTECTION_DIS, 1);
+}
+
+static int iceland_enable_auto_throttle_source(struct pp_hwmgr *hwmgr,
+ PHM_AutoThrottleSource source)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ if (!(data->active_auto_throttle_sources & (1 << source))) {
+ data->active_auto_throttle_sources |= 1 << source;
+ iceland_set_dpm_event_sources(hwmgr, data->active_auto_throttle_sources);
+ }
+ return 0;
+}
+
+static int iceland_enable_thermal_auto_throttle(struct pp_hwmgr *hwmgr)
+{
+ return iceland_enable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal);
+}
+
+static int iceland_tf_start_smc(struct pp_hwmgr *hwmgr)
+{
+ int ret = 0;
+
+ if (!iceland_is_smc_ram_running(hwmgr->smumgr))
+ ret = iceland_smu_start_smc(hwmgr->smumgr);
+
+ return ret;
+}
+
+/**
+* Programs the Deep Sleep registers
+*
+* @param pHwMgr the address of the powerplay hardware manager.
+* @param pInput the pointer to input data (PhwEvergreen_DisplayConfiguration)
+* @param pOutput the pointer to output data (unused)
+* @param pStorage the pointer to temporary storage (unused)
+* @param Result the last failure code (unused)
+* @return always 0
+*/
+static int iceland_enable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr)
+{
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SclkDeepSleep)) {
+ if (smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_MASTER_DeepSleep_ON) != 0)
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to enable Master Deep Sleep switch failed!",
+ return -EINVAL);
+ } else {
+ if (smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_MASTER_DeepSleep_OFF) != 0)
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to disable Master Deep Sleep switch failed!",
+ return -EINVAL);
+ }
+
+ return 0;
+}
+
+static int iceland_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
+{
+ int tmp_result, result = 0;
+
+ if (cf_iceland_voltage_control(hwmgr)) {
+ tmp_result = iceland_enable_voltage_control(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to enable voltage control!", return tmp_result);
+
+ tmp_result = iceland_construct_voltage_tables(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to contruct voltage tables!", return tmp_result);
+ }
+
+ tmp_result = iceland_initialize_mc_reg_table(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to initialize MC reg table!", return tmp_result);
+
+ tmp_result = iceland_program_static_screen_threshold_parameters(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to program static screen threshold parameters!", return tmp_result);
+
+ tmp_result = iceland_enable_display_gap(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to enable display gap!", return tmp_result);
+
+ tmp_result = iceland_program_voting_clients(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to program voting clients!", return tmp_result);
+
+ tmp_result = iceland_upload_firmware(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to upload firmware header!", return tmp_result);
+
+ tmp_result = iceland_process_firmware_header(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to process firmware header!", return tmp_result);
+
+ tmp_result = iceland_initial_switch_from_arb_f0_to_f1(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to initialize switch from ArbF0 to F1!", return tmp_result);
+
+ tmp_result = iceland_init_smc_table(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to initialize SMC table!", return tmp_result);
+
+ tmp_result = iceland_populate_initial_mc_reg_table(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to populate initialize MC Reg table!", return tmp_result);
+
+ tmp_result = iceland_populate_pm_fuses(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to populate PM fuses!", return tmp_result);
+
+ /* start SMC */
+ tmp_result = iceland_tf_start_smc(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to start SMC!", return tmp_result);
+
+ /* enable SCLK control */
+ tmp_result = iceland_enable_sclk_control(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to enable SCLK control!", return tmp_result);
+
+ tmp_result = iceland_enable_deep_sleep_master_switch(hwmgr);
+ PP_ASSERT_WITH_CODE((tmp_result == 0),
+ "Failed to enable deep sleep!", return tmp_result);
+
+ /* enable DPM */
+ tmp_result = iceland_start_dpm(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to start DPM!", return tmp_result);
+
+ tmp_result = iceland_enable_smc_cac(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to enable SMC CAC!", return tmp_result);
+
+ tmp_result = iceland_enable_power_containment(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to enable power containment!", return tmp_result);
+
+ tmp_result = iceland_power_control_set_level(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to power control set level!", result = tmp_result);
+
+ tmp_result = iceland_enable_thermal_auto_throttle(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to enable thermal auto throttle!", result = tmp_result);
+
+ return result;
+}
+
+static int iceland_hwmgr_backend_fini(struct pp_hwmgr *hwmgr)
+{
+ return phm_hwmgr_backend_fini(hwmgr);
+}
+
+static void iceland_initialize_dpm_defaults(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct phw_iceland_ulv_parm *ulv;
+
+ ulv = &data->ulv;
+ ulv->ch_ulv_parameter = PPICELAND_CGULVPARAMETER_DFLT;
+ data->voting_rights_clients0 = PPICELAND_VOTINGRIGHTSCLIENTS_DFLT0;
+ data->voting_rights_clients1 = PPICELAND_VOTINGRIGHTSCLIENTS_DFLT1;
+ data->voting_rights_clients2 = PPICELAND_VOTINGRIGHTSCLIENTS_DFLT2;
+ data->voting_rights_clients3 = PPICELAND_VOTINGRIGHTSCLIENTS_DFLT3;
+ data->voting_rights_clients4 = PPICELAND_VOTINGRIGHTSCLIENTS_DFLT4;
+ data->voting_rights_clients5 = PPICELAND_VOTINGRIGHTSCLIENTS_DFLT5;
+ data->voting_rights_clients6 = PPICELAND_VOTINGRIGHTSCLIENTS_DFLT6;
+ data->voting_rights_clients7 = PPICELAND_VOTINGRIGHTSCLIENTS_DFLT7;
+
+ data->static_screen_threshold_unit = PPICELAND_STATICSCREENTHRESHOLDUNIT_DFLT;
+ data->static_screen_threshold = PPICELAND_STATICSCREENTHRESHOLD_DFLT;
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_ABM);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_NonABMSupportInPPLib);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DynamicACTiming);
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DisableMemoryTransition);
+
+ iceland_initialize_power_tune_defaults(hwmgr);
+
+ data->mclk_strobe_mode_threshold = 40000;
+ data->mclk_stutter_mode_threshold = 30000;
+ data->mclk_edc_enable_threshold = 40000;
+ data->mclk_edc_wr_enable_threshold = 40000;
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DisableMCLS);
+
+ data->pcie_gen_performance.max = PP_PCIEGen1;
+ data->pcie_gen_performance.min = PP_PCIEGen3;
+ data->pcie_gen_power_saving.max = PP_PCIEGen1;
+ data->pcie_gen_power_saving.min = PP_PCIEGen3;
+
+ data->pcie_lane_performance.max = 0;
+ data->pcie_lane_performance.min = 16;
+ data->pcie_lane_power_saving.max = 0;
+ data->pcie_lane_power_saving.min = 16;
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SclkThrottleLowNotification);
+}
+
+static int iceland_get_evv_voltage(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+ uint16_t virtual_voltage_id;
+ uint16_t vddc = 0;
+ uint16_t i;
+
+ /* the count indicates actual number of entries */
+ data->vddc_leakage.count = 0;
+ data->vddci_leakage.count = 0;
+
+ if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_EVV)) {
+ pr_err("Iceland should always support EVV\n");
+ return -EINVAL;
+ }
+
+ /* retrieve voltage for leakage ID (0xff01 + i) */
+ for (i = 0; i < ICELAND_MAX_LEAKAGE_COUNT; i++) {
+ virtual_voltage_id = ATOM_VIRTUAL_VOLTAGE_ID0 + i;
+
+ PP_ASSERT_WITH_CODE((0 == atomctrl_get_voltage_evv(hwmgr, virtual_voltage_id, &vddc)),
+ "Error retrieving EVV voltage value!\n", continue);
+
+ if (vddc >= 2000)
+ pr_warning("Invalid VDDC value!\n");
+
+ if (vddc != 0 && vddc != virtual_voltage_id) {
+ data->vddc_leakage.actual_voltage[data->vddc_leakage.count] = vddc;
+ data->vddc_leakage.leakage_id[data->vddc_leakage.count] = virtual_voltage_id;
+ data->vddc_leakage.count++;
+ }
+ }
+
+ return 0;
+}
+
+static void iceland_patch_with_vddc_leakage(struct pp_hwmgr *hwmgr,
+ uint32_t *vddc)
+{
+ iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint32_t leakage_index;
+ struct phw_iceland_leakage_voltage *leakage_table = &data->vddc_leakage;
+
+ /* search for leakage voltage ID 0xff01 ~ 0xff08 */
+ for (leakage_index = 0; leakage_index < leakage_table->count; leakage_index++) {
+ /*
+ * If this voltage matches a leakage voltage ID, patch
+ * with actual leakage voltage.
+ */
+ if (leakage_table->leakage_id[leakage_index] == *vddc) {
+ /*
+ * Need to make sure vddc is less than 2v or
+ * else, it could burn the ASIC.
+ */
+ if (leakage_table->actual_voltage[leakage_index] >= 2000)
+ pr_warning("Invalid VDDC value!\n");
+ *vddc = leakage_table->actual_voltage[leakage_index];
+ /* we found leakage voltage */
+ break;
+ }
+ }
+
+ if (*vddc >= ATOM_VIRTUAL_VOLTAGE_ID0)
+ pr_warning("Voltage value looks like a Leakage ID but it's not patched\n");
+}
+
+static void iceland_patch_with_vddci_leakage(struct pp_hwmgr *hwmgr,
+ uint32_t *vddci)
+{
+ iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint32_t leakage_index;
+ struct phw_iceland_leakage_voltage *leakage_table = &data->vddci_leakage;
+
+ /* search for leakage voltage ID 0xff01 ~ 0xff08 */
+ for (leakage_index = 0; leakage_index < leakage_table->count; leakage_index++) {
+ /*
+ * If this voltage matches a leakage voltage ID, patch
+ * with actual leakage voltage.
+ */
+ if (leakage_table->leakage_id[leakage_index] == *vddci) {
+ *vddci = leakage_table->actual_voltage[leakage_index];
+ /* we found leakage voltage */
+ break;
+ }
+ }
+
+ if (*vddci >= ATOM_VIRTUAL_VOLTAGE_ID0)
+ pr_warning("Voltage value looks like a Leakage ID but it's not patched\n");
+}
+
+static int iceland_patch_vddc(struct pp_hwmgr *hwmgr,
+ struct phm_clock_voltage_dependency_table *tab)
+{
+ uint16_t i;
+
+ if (tab)
+ for (i = 0; i < tab->count; i++)
+ iceland_patch_with_vddc_leakage(hwmgr, &tab->entries[i].v);
+
+ return 0;
+}
+
+static int iceland_patch_vddci(struct pp_hwmgr *hwmgr,
+ struct phm_clock_voltage_dependency_table *tab)
+{
+ uint16_t i;
+
+ if (tab)
+ for (i = 0; i < tab->count; i++)
+ iceland_patch_with_vddci_leakage(hwmgr, &tab->entries[i].v);
+
+ return 0;
+}
+
+static int iceland_patch_vce_vddc(struct pp_hwmgr *hwmgr,
+ struct phm_vce_clock_voltage_dependency_table *tab)
+{
+ uint16_t i;
+
+ if (tab)
+ for (i = 0; i < tab->count; i++)
+ iceland_patch_with_vddc_leakage(hwmgr, &tab->entries[i].v);
+
+ return 0;
+}
+
+
+static int iceland_patch_uvd_vddc(struct pp_hwmgr *hwmgr,
+ struct phm_uvd_clock_voltage_dependency_table *tab)
+{
+ uint16_t i;
+
+ if (tab)
+ for (i = 0; i < tab->count; i++)
+ iceland_patch_with_vddc_leakage(hwmgr, &tab->entries[i].v);
+
+ return 0;
+}
+
+static int iceland_patch_vddc_shed_limit(struct pp_hwmgr *hwmgr,
+ struct phm_phase_shedding_limits_table *tab)
+{
+ uint16_t i;
+
+ if (tab)
+ for (i = 0; i < tab->count; i++)
+ iceland_patch_with_vddc_leakage(hwmgr, &tab->entries[i].Voltage);
+
+ return 0;
+}
+
+static int iceland_patch_samu_vddc(struct pp_hwmgr *hwmgr,
+ struct phm_samu_clock_voltage_dependency_table *tab)
+{
+ uint16_t i;
+
+ if (tab)
+ for (i = 0; i < tab->count; i++)
+ iceland_patch_with_vddc_leakage(hwmgr, &tab->entries[i].v);
+
+ return 0;
+}
+
+static int iceland_patch_acp_vddc(struct pp_hwmgr *hwmgr,
+ struct phm_acp_clock_voltage_dependency_table *tab)
+{
+ uint16_t i;
+
+ if (tab)
+ for (i = 0; i < tab->count; i++)
+ iceland_patch_with_vddc_leakage(hwmgr, &tab->entries[i].v);
+
+ return 0;
+}
+
+static int iceland_patch_limits_vddc(struct pp_hwmgr *hwmgr,
+ struct phm_clock_and_voltage_limits *tab)
+{
+ if (tab) {
+ iceland_patch_with_vddc_leakage(hwmgr, (uint32_t *)&tab->vddc);
+ iceland_patch_with_vddci_leakage(hwmgr, (uint32_t *)&tab->vddci);
+ }
+
+ return 0;
+}
+
+static int iceland_patch_cac_vddc(struct pp_hwmgr *hwmgr, struct phm_cac_leakage_table *tab)
+{
+ uint32_t i;
+ uint32_t vddc;
+
+ if (tab) {
+ for (i = 0; i < tab->count; i++) {
+ vddc = (uint32_t)(tab->entries[i].Vddc);
+ iceland_patch_with_vddc_leakage(hwmgr, &vddc);
+ tab->entries[i].Vddc = (uint16_t)vddc;
+ }
+ }
+
+ return 0;
+}
+
+static int iceland_patch_dependency_tables_with_leakage(struct pp_hwmgr *hwmgr)
+{
+ int tmp;
+
+ tmp = iceland_patch_vddc(hwmgr, hwmgr->dyn_state.vddc_dependency_on_sclk);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_vddc(hwmgr, hwmgr->dyn_state.vddc_dependency_on_mclk);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_vddc(hwmgr, hwmgr->dyn_state.vddc_dep_on_dal_pwrl);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_vddci(hwmgr, hwmgr->dyn_state.vddci_dependency_on_mclk);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_vce_vddc(hwmgr, hwmgr->dyn_state.vce_clock_voltage_dependency_table);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_uvd_vddc(hwmgr, hwmgr->dyn_state.uvd_clock_voltage_dependency_table);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_samu_vddc(hwmgr, hwmgr->dyn_state.samu_clock_voltage_dependency_table);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_acp_vddc(hwmgr, hwmgr->dyn_state.acp_clock_voltage_dependency_table);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_vddc_shed_limit(hwmgr, hwmgr->dyn_state.vddc_phase_shed_limits_table);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_limits_vddc(hwmgr, &hwmgr->dyn_state.max_clock_voltage_on_ac);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_limits_vddc(hwmgr, &hwmgr->dyn_state.max_clock_voltage_on_dc);
+ if(tmp)
+ return -EINVAL;
+
+ tmp = iceland_patch_cac_vddc(hwmgr, hwmgr->dyn_state.cac_leakage_table);
+ if(tmp)
+ return -EINVAL;
+
+ return 0;
+}
+
+static int iceland_set_private_var_based_on_pptale(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ struct phm_clock_voltage_dependency_table *allowed_sclk_vddc_table = hwmgr->dyn_state.vddc_dependency_on_sclk;
+ struct phm_clock_voltage_dependency_table *allowed_mclk_vddc_table = hwmgr->dyn_state.vddc_dependency_on_mclk;
+ struct phm_clock_voltage_dependency_table *allowed_mclk_vddci_table = hwmgr->dyn_state.vddci_dependency_on_mclk;
+
+ PP_ASSERT_WITH_CODE(allowed_sclk_vddc_table != NULL,
+ "VDDC dependency on SCLK table is missing. This table is mandatory\n", return -EINVAL);
+ PP_ASSERT_WITH_CODE(allowed_sclk_vddc_table->count >= 1,
+ "VDDC dependency on SCLK table has to have is missing. This table is mandatory\n", return -EINVAL);
+
+ PP_ASSERT_WITH_CODE(allowed_mclk_vddc_table != NULL,
+ "VDDC dependency on MCLK table is missing. This table is mandatory\n", return -EINVAL);
+ PP_ASSERT_WITH_CODE(allowed_mclk_vddc_table->count >= 1,
+ "VDD dependency on MCLK table has to have is missing. This table is mandatory\n", return -EINVAL);
+
+ data->min_vddc_in_pp_table = (uint16_t)allowed_sclk_vddc_table->entries[0].v;
+ data->max_vddc_in_pp_table = (uint16_t)allowed_sclk_vddc_table->entries[allowed_sclk_vddc_table->count - 1].v;
+
+ hwmgr->dyn_state.max_clock_voltage_on_ac.sclk =
+ allowed_sclk_vddc_table->entries[allowed_sclk_vddc_table->count - 1].clk;
+ hwmgr->dyn_state.max_clock_voltage_on_ac.mclk =
+ allowed_mclk_vddc_table->entries[allowed_mclk_vddc_table->count - 1].clk;
+ hwmgr->dyn_state.max_clock_voltage_on_ac.vddc =
+ allowed_sclk_vddc_table->entries[allowed_sclk_vddc_table->count - 1].v;
+
+ if (allowed_mclk_vddci_table != NULL && allowed_mclk_vddci_table->count >= 1) {
+ data->min_vddci_in_pp_table = (uint16_t)allowed_mclk_vddci_table->entries[0].v;
+ data->max_vddci_in_pp_table = (uint16_t)allowed_mclk_vddci_table->entries[allowed_mclk_vddci_table->count - 1].v;
+ }
+
+ if (hwmgr->dyn_state.vddci_dependency_on_mclk != NULL && hwmgr->dyn_state.vddci_dependency_on_mclk->count > 1)
+ hwmgr->dyn_state.max_clock_voltage_on_ac.vddci = hwmgr->dyn_state.vddci_dependency_on_mclk->entries[hwmgr->dyn_state.vddci_dependency_on_mclk->count - 1].v;
+
+ return 0;
+}
+
+static int iceland_initializa_dynamic_state_adjustment_rule_settings(struct pp_hwmgr *hwmgr)
+{
+ uint32_t table_size;
+ struct phm_clock_voltage_dependency_table *table_clk_vlt;
+
+ hwmgr->dyn_state.mclk_sclk_ratio = 4;
+ hwmgr->dyn_state.sclk_mclk_delta = 15000; /* 150 MHz */
+ hwmgr->dyn_state.vddc_vddci_delta = 200; /* 200mV */
+
+ /* initialize vddc_dep_on_dal_pwrl table */
+ table_size = sizeof(uint32_t) + 4 * sizeof(struct phm_clock_voltage_dependency_record);
+ table_clk_vlt = (struct phm_clock_voltage_dependency_table *)kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == table_clk_vlt) {
+ pr_err("[ powerplay ] Can not allocate space for vddc_dep_on_dal_pwrl! \n");
+ return -ENOMEM;
+ } else {
+ table_clk_vlt->count = 4;
+ table_clk_vlt->entries[0].clk = PP_DAL_POWERLEVEL_ULTRALOW;
+ table_clk_vlt->entries[0].v = 0;
+ table_clk_vlt->entries[1].clk = PP_DAL_POWERLEVEL_LOW;
+ table_clk_vlt->entries[1].v = 720;
+ table_clk_vlt->entries[2].clk = PP_DAL_POWERLEVEL_NOMINAL;
+ table_clk_vlt->entries[2].v = 810;
+ table_clk_vlt->entries[3].clk = PP_DAL_POWERLEVEL_PERFORMANCE;
+ table_clk_vlt->entries[3].v = 900;
+ hwmgr->dyn_state.vddc_dep_on_dal_pwrl = table_clk_vlt;
+ }
+
+ return 0;
+}
+
+/**
+ * Initializes the Volcanic Islands Hardware Manager
+ *
+ * @param hwmgr the address of the powerplay hardware manager.
+ * @return 1 if success; otherwise appropriate error code.
+ */
+static int iceland_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
+{
+ int result = 0;
+ SMU71_Discrete_DpmTable *table = NULL;
+ iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ pp_atomctrl_gpio_pin_assignment gpio_pin_assignment;
+ bool stay_in_boot;
+ struct phw_iceland_ulv_parm *ulv;
+ struct cgs_system_info sys_info = {0};
+
+ PP_ASSERT_WITH_CODE((NULL != hwmgr),
+ "Invalid Parameter!", return -EINVAL;);
+
+ data->dll_defaule_on = 0;
+ data->sram_end = SMC_RAM_END;
+
+ data->activity_target[0] = PPICELAND_TARGETACTIVITY_DFLT;
+ data->activity_target[1] = PPICELAND_TARGETACTIVITY_DFLT;
+ data->activity_target[2] = PPICELAND_TARGETACTIVITY_DFLT;
+ data->activity_target[3] = PPICELAND_TARGETACTIVITY_DFLT;
+ data->activity_target[4] = PPICELAND_TARGETACTIVITY_DFLT;
+ data->activity_target[5] = PPICELAND_TARGETACTIVITY_DFLT;
+ data->activity_target[6] = PPICELAND_TARGETACTIVITY_DFLT;
+ data->activity_target[7] = PPICELAND_TARGETACTIVITY_DFLT;
+
+ data->mclk_activity_target = PPICELAND_MCLK_TARGETACTIVITY_DFLT;
+
+ data->sclk_dpm_key_disabled = 0;
+ data->mclk_dpm_key_disabled = 0;
+ data->pcie_dpm_key_disabled = 0;
+ data->pcc_monitor_enabled = 0;
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_UnTabledHardwareInterface);
+
+ data->gpio_debug = 0;
+ data->engine_clock_data = 0;
+ data->memory_clock_data = 0;
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SclkDeepSleepAboveLow);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DynamicPatchPowerState);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_TablelessHardwareInterface);
+
+ /* Initializes DPM default values. */
+ iceland_initialize_dpm_defaults(hwmgr);
+
+ /* Enable Platform EVV support. */
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EVV);
+
+ /* Get leakage voltage based on leakage ID. */
+ result = iceland_get_evv_voltage(hwmgr);
+ if (result)
+ goto failed;
+
+ /**
+ * Patch our voltage dependency table with actual leakage
+ * voltage. We need to perform leakage translation before it's
+ * used by other functions such as
+ * iceland_set_hwmgr_variables_based_on_pptable.
+ */
+ result = iceland_patch_dependency_tables_with_leakage(hwmgr);
+ if (result)
+ goto failed;
+
+ /* Parse pptable data read from VBIOS. */
+ result = iceland_set_private_var_based_on_pptale(hwmgr);
+ if (result)
+ goto failed;
+
+ /* ULV support */
+ ulv = &(data->ulv);
+ ulv->ulv_supported = 1;
+
+ /* Initalize Dynamic State Adjustment Rule Settings*/
+ result = iceland_initializa_dynamic_state_adjustment_rule_settings(hwmgr);
+ if (result) {
+ pr_err("[ powerplay ] iceland_initializa_dynamic_state_adjustment_rule_settings failed!\n");
+ goto failed;
+ }
+
+ data->voltage_control = ICELAND_VOLTAGE_CONTROL_NONE;
+ data->vdd_ci_control = ICELAND_VOLTAGE_CONTROL_NONE;
+ data->mvdd_control = ICELAND_VOLTAGE_CONTROL_NONE;
+
+ /*
+ * Hardcode thermal temperature settings for now, these will
+ * be overwritten if a custom policy exists.
+ */
+ data->thermal_temp_setting.temperature_low = 99500;
+ data->thermal_temp_setting.temperature_high = 100000;
+ data->thermal_temp_setting.temperature_shutdown = 104000;
+ data->uvd_enabled = false;
+
+ table = &data->smc_state_table;
+
+ if (atomctrl_get_pp_assign_pin(hwmgr, VDDC_VRHOT_GPIO_PINID,
+ &gpio_pin_assignment)) {
+ table->VRHotGpio = gpio_pin_assignment.uc_gpio_pin_bit_shift;
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_RegulatorHot);
+ } else {
+ table->VRHotGpio = ICELAND_UNUSED_GPIO_PIN;
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_RegulatorHot);
+ }
+
+ if (atomctrl_get_pp_assign_pin(hwmgr, PP_AC_DC_SWITCH_GPIO_PINID,
+ &gpio_pin_assignment)) {
+ table->AcDcGpio = gpio_pin_assignment.uc_gpio_pin_bit_shift;
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_AutomaticDCTransition);
+ } else {
+ table->AcDcGpio = ICELAND_UNUSED_GPIO_PIN;
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_AutomaticDCTransition);
+ }
+
+ /*
+ * If ucGPIO_ID=VDDC_PCC_GPIO_PINID in GPIO_LUTable, Peak.
+ * Current Control feature is enabled and we should program
+ * PCC HW register
+ */
+ if (atomctrl_get_pp_assign_pin(hwmgr, VDDC_PCC_GPIO_PINID,
+ &gpio_pin_assignment)) {
+ uint32_t temp_reg = cgs_read_ind_register(hwmgr->device,
+ CGS_IND_REG__SMC,
+ ixCNB_PWRMGT_CNTL);
+
+ switch (gpio_pin_assignment.uc_gpio_pin_bit_shift) {
+ case 0:
+ temp_reg = PHM_SET_FIELD(temp_reg,
+ CNB_PWRMGT_CNTL, GNB_SLOW_MODE, 0x1);
+ break;
+ case 1:
+ temp_reg = PHM_SET_FIELD(temp_reg,
+ CNB_PWRMGT_CNTL, GNB_SLOW_MODE, 0x2);
+ break;
+ case 2:
+ temp_reg = PHM_SET_FIELD(temp_reg,
+ CNB_PWRMGT_CNTL, GNB_SLOW, 0x1);
+ break;
+ case 3:
+ temp_reg = PHM_SET_FIELD(temp_reg,
+ CNB_PWRMGT_CNTL, FORCE_NB_PS1, 0x1);
+ break;
+ case 4:
+ temp_reg = PHM_SET_FIELD(temp_reg,
+ CNB_PWRMGT_CNTL, DPM_ENABLED, 0x1);
+ break;
+ default:
+ pr_warning("[ powerplay ] Failed to setup PCC HW register! Wrong GPIO assigned for VDDC_PCC_GPIO_PINID!\n");
+ break;
+ }
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+ ixCNB_PWRMGT_CNTL, temp_reg);
+ }
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EnableSMU7ThermalManagement);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SMU7);
+
+ if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr,
+ VOLTAGE_TYPE_VDDC,
+ VOLTAGE_OBJ_GPIO_LUT))
+ data->voltage_control = ICELAND_VOLTAGE_CONTROL_BY_GPIO;
+ else if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr,
+ VOLTAGE_TYPE_VDDC,
+ VOLTAGE_OBJ_SVID2))
+ data->voltage_control = ICELAND_VOLTAGE_CONTROL_BY_SVID2;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_ControlVDDCI)) {
+ if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr,
+ VOLTAGE_TYPE_VDDCI,
+ VOLTAGE_OBJ_GPIO_LUT))
+ data->vdd_ci_control = ICELAND_VOLTAGE_CONTROL_BY_GPIO;
+ else if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr,
+ VOLTAGE_TYPE_VDDCI,
+ VOLTAGE_OBJ_SVID2))
+ data->vdd_ci_control = ICELAND_VOLTAGE_CONTROL_BY_SVID2;
+ }
+
+ if (data->vdd_ci_control == ICELAND_VOLTAGE_CONTROL_NONE)
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_ControlVDDCI);
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EnableMVDDControl)) {
+ if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr,
+ VOLTAGE_TYPE_MVDDC,
+ VOLTAGE_OBJ_GPIO_LUT))
+ data->mvdd_control = ICELAND_VOLTAGE_CONTROL_BY_GPIO;
+ else if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr,
+ VOLTAGE_TYPE_MVDDC,
+ VOLTAGE_OBJ_SVID2))
+ data->mvdd_control = ICELAND_VOLTAGE_CONTROL_BY_SVID2;
+ }
+
+ if (data->mvdd_control == ICELAND_VOLTAGE_CONTROL_NONE)
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EnableMVDDControl);
+
+ data->vddc_phase_shed_control = false;
+
+ stay_in_boot = phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_StayInBootState);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DynamicPowerManagement);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_ActivityReporting);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_GFXClockGatingSupport);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_MemorySpreadSpectrumSupport);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EngineSpreadSpectrumSupport);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DynamicPCIEGen2Support);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SMC);
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DisablePowerGating);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_BACO);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_ThermalAutoThrottling);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DisableLSClockGating);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SamuDPM);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_AcpDPM);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_OD6inACSupport);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EnablePlatformPowerManagement);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PauseMMSessions);
+
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_OD6PlusinACSupport);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PauseMMSessions);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_GFXClockGatingManagedInCAIL);
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_IcelandULPSSWWorkAround);
+
+
+ /* iceland doesn't support UVD and VCE */
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_UVDPowerGating);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_VCEPowerGating);
+
+ sys_info.size = sizeof(struct cgs_system_info);
+ sys_info.info_id = CGS_SYSTEM_INFO_PG_FLAGS;
+ result = cgs_query_system_info(hwmgr->device, &sys_info);
+ if (!result) {
+ if (sys_info.value & AMD_PG_SUPPORT_UVD)
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_UVDPowerGating);
+ if (sys_info.value & AMD_PG_SUPPORT_VCE)
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_VCEPowerGating);
+
+ data->is_tlu_enabled = false;
+ hwmgr->platform_descriptor.hardwareActivityPerformanceLevels =
+ ICELAND_MAX_HARDWARE_POWERLEVELS;
+ hwmgr->platform_descriptor.hardwarePerformanceLevels = 2;
+ hwmgr->platform_descriptor.minimumClocksReductionPercentage = 50;
+
+ sys_info.size = sizeof(struct cgs_system_info);
+ sys_info.info_id = CGS_SYSTEM_INFO_PCIE_GEN_INFO;
+ result = cgs_query_system_info(hwmgr->device, &sys_info);
+ if (result)
+ data->pcie_gen_cap = AMDGPU_DEFAULT_PCIE_GEN_MASK;
+ else
+ data->pcie_gen_cap = (uint32_t)sys_info.value;
+ if (data->pcie_gen_cap & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3)
+ data->pcie_spc_cap = 20;
+ sys_info.size = sizeof(struct cgs_system_info);
+ sys_info.info_id = CGS_SYSTEM_INFO_PCIE_MLW;
+ result = cgs_query_system_info(hwmgr->device, &sys_info);
+ if (result)
+ data->pcie_lane_cap = AMDGPU_DEFAULT_PCIE_MLW_MASK;
+ else
+ data->pcie_lane_cap = (uint32_t)sys_info.value;
+ } else {
+ /* Ignore return value in here, we are cleaning up a mess. */
+ iceland_hwmgr_backend_fini(hwmgr);
+ }
+
+ return 0;
+failed:
+ return result;
+}
+
+static int iceland_get_num_of_entries(struct pp_hwmgr *hwmgr)
+{
+ int result;
+ unsigned long ret = 0;
+
+ result = pp_tables_get_num_of_entries(hwmgr, &ret);
+
+ return result ? 0 : ret;
+}
+
+static const unsigned long PhwIceland_Magic = (unsigned long)(PHM_VIslands_Magic);
+
+struct iceland_power_state *cast_phw_iceland_power_state(
+ struct pp_hw_power_state *hw_ps)
+{
+ if (hw_ps == NULL)
+ return NULL;
+
+ PP_ASSERT_WITH_CODE((PhwIceland_Magic == hw_ps->magic),
+ "Invalid Powerstate Type!",
+ return NULL);
+
+ return (struct iceland_power_state *)hw_ps;
+}
+
+static int iceland_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
+ struct pp_power_state *prequest_ps,
+ const struct pp_power_state *pcurrent_ps)
+{
+ struct iceland_power_state *iceland_ps =
+ cast_phw_iceland_power_state(&prequest_ps->hardware);
+
+ uint32_t sclk;
+ uint32_t mclk;
+ struct PP_Clocks minimum_clocks = {0};
+ bool disable_mclk_switching;
+ bool disable_mclk_switching_for_frame_lock;
+ struct cgs_display_info info = {0};
+ const struct phm_clock_and_voltage_limits *max_limits;
+ uint32_t i;
+ iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ int32_t count;
+ int32_t stable_pstate_sclk = 0, stable_pstate_mclk = 0;
+
+ data->battery_state = (PP_StateUILabel_Battery == prequest_ps->classification.ui_label);
+
+ PP_ASSERT_WITH_CODE(iceland_ps->performance_level_count == 2,
+ "VI should always have 2 performance levels",
+ );
+
+ max_limits = (PP_PowerSource_AC == hwmgr->power_source) ?
+ &(hwmgr->dyn_state.max_clock_voltage_on_ac) :
+ &(hwmgr->dyn_state.max_clock_voltage_on_dc);
+
+ if (PP_PowerSource_DC == hwmgr->power_source) {
+ for (i = 0; i < iceland_ps->performance_level_count; i++) {
+ if (iceland_ps->performance_levels[i].memory_clock > max_limits->mclk)
+ iceland_ps->performance_levels[i].memory_clock = max_limits->mclk;
+ if (iceland_ps->performance_levels[i].engine_clock > max_limits->sclk)
+ iceland_ps->performance_levels[i].engine_clock = max_limits->sclk;
+ }
+ }
+
+ iceland_ps->vce_clocks.EVCLK = hwmgr->vce_arbiter.evclk;
+ iceland_ps->vce_clocks.ECCLK = hwmgr->vce_arbiter.ecclk;
+
+ cgs_get_active_displays_info(hwmgr->device, &info);
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_StablePState)) {
+
+ max_limits = &(hwmgr->dyn_state.max_clock_voltage_on_ac);
+ stable_pstate_sclk = (max_limits->sclk * 75) / 100;
+
+ for (count = hwmgr->dyn_state.vddc_dependency_on_sclk->count-1; count >= 0; count--) {
+ if (stable_pstate_sclk >= hwmgr->dyn_state.vddc_dependency_on_sclk->entries[count].clk) {
+ stable_pstate_sclk = hwmgr->dyn_state.vddc_dependency_on_sclk->entries[count].clk;
+ break;
+ }
+ }
+
+ if (count < 0)
+ stable_pstate_sclk = hwmgr->dyn_state.vddc_dependency_on_sclk->entries[0].clk;
+
+ stable_pstate_mclk = max_limits->mclk;
+
+ minimum_clocks.engineClock = stable_pstate_sclk;
+ minimum_clocks.memoryClock = stable_pstate_mclk;
+ }
+
+ if (minimum_clocks.engineClock < hwmgr->gfx_arbiter.sclk)
+ minimum_clocks.engineClock = hwmgr->gfx_arbiter.sclk;
+
+ if (minimum_clocks.memoryClock < hwmgr->gfx_arbiter.mclk)
+ minimum_clocks.memoryClock = hwmgr->gfx_arbiter.mclk;
+
+ iceland_ps->sclk_threshold = hwmgr->gfx_arbiter.sclk_threshold;
+
+ if (0 != hwmgr->gfx_arbiter.sclk_over_drive) {
+ PP_ASSERT_WITH_CODE((hwmgr->gfx_arbiter.sclk_over_drive <= hwmgr->platform_descriptor.overdriveLimit.engineClock),
+ "Overdrive sclk exceeds limit",
+ hwmgr->gfx_arbiter.sclk_over_drive = hwmgr->platform_descriptor.overdriveLimit.engineClock);
+
+ if (hwmgr->gfx_arbiter.sclk_over_drive >= hwmgr->gfx_arbiter.sclk)
+ iceland_ps->performance_levels[1].engine_clock = hwmgr->gfx_arbiter.sclk_over_drive;
+ }
+
+ if (0 != hwmgr->gfx_arbiter.mclk_over_drive) {
+ PP_ASSERT_WITH_CODE((hwmgr->gfx_arbiter.mclk_over_drive <= hwmgr->platform_descriptor.overdriveLimit.memoryClock),
+ "Overdrive mclk exceeds limit",
+ hwmgr->gfx_arbiter.mclk_over_drive = hwmgr->platform_descriptor.overdriveLimit.memoryClock);
+
+ if (hwmgr->gfx_arbiter.mclk_over_drive >= hwmgr->gfx_arbiter.mclk)
+ iceland_ps->performance_levels[1].memory_clock = hwmgr->gfx_arbiter.mclk_over_drive;
+ }
+
+ disable_mclk_switching_for_frame_lock = phm_cap_enabled(
+ hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DisableMclkSwitchingForFrameLock);
+
+ disable_mclk_switching = (1 < info.display_count) ||
+ disable_mclk_switching_for_frame_lock;
+
+ sclk = iceland_ps->performance_levels[0].engine_clock;
+ mclk = iceland_ps->performance_levels[0].memory_clock;
+
+ if (disable_mclk_switching)
+ mclk = iceland_ps->performance_levels[iceland_ps->performance_level_count - 1].memory_clock;
+
+ if (sclk < minimum_clocks.engineClock)
+ sclk = (minimum_clocks.engineClock > max_limits->sclk) ? max_limits->sclk : minimum_clocks.engineClock;
+
+ if (mclk < minimum_clocks.memoryClock)
+ mclk = (minimum_clocks.memoryClock > max_limits->mclk) ? max_limits->mclk : minimum_clocks.memoryClock;
+
+ iceland_ps->performance_levels[0].engine_clock = sclk;
+ iceland_ps->performance_levels[0].memory_clock = mclk;
+
+ iceland_ps->performance_levels[1].engine_clock =
+ (iceland_ps->performance_levels[1].engine_clock >= iceland_ps->performance_levels[0].engine_clock) ?
+ iceland_ps->performance_levels[1].engine_clock :
+ iceland_ps->performance_levels[0].engine_clock;
+
+ if (disable_mclk_switching) {
+ if (mclk < iceland_ps->performance_levels[1].memory_clock)
+ mclk = iceland_ps->performance_levels[1].memory_clock;
+
+ iceland_ps->performance_levels[0].memory_clock = mclk;
+ iceland_ps->performance_levels[1].memory_clock = mclk;
+ } else {
+ if (iceland_ps->performance_levels[1].memory_clock < iceland_ps->performance_levels[0].memory_clock)
+ iceland_ps->performance_levels[1].memory_clock = iceland_ps->performance_levels[0].memory_clock;
+ }
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_StablePState)) {
+ for (i=0; i < iceland_ps->performance_level_count; i++) {
+ iceland_ps->performance_levels[i].engine_clock = stable_pstate_sclk;
+ iceland_ps->performance_levels[i].memory_clock = stable_pstate_mclk;
+ iceland_ps->performance_levels[i].pcie_gen = data->pcie_gen_performance.max;
+ iceland_ps->performance_levels[i].pcie_lane = data->pcie_gen_performance.max;
+ }
+ }
+
+ return 0;
+}
+
+static bool iceland_is_dpm_running(struct pp_hwmgr *hwmgr)
+{
+ /*
+ * We return the status of Voltage Control instead of checking SCLK/MCLK DPM
+ * because we may have test scenarios that need us intentionly disable SCLK/MCLK DPM,
+ * whereas voltage control is a fundemental change that will not be disabled
+ */
+ return (0 == PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+ FEATURE_STATUS, VOLTAGE_CONTROLLER_ON) ? 1 : 0);
+}
+
+/**
+ * force DPM power State
+ *
+ * @param hwmgr: the address of the powerplay hardware manager.
+ * @param n : DPM level
+ * @return The response that came from the SMC.
+ */
+int iceland_dpm_force_state(struct pp_hwmgr *hwmgr, uint32_t n)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ /* Checking if DPM is running. If we discover hang because of this, we should skip this message. */
+ PP_ASSERT_WITH_CODE(0 == iceland_is_dpm_running(hwmgr),
+ "Trying to force SCLK when DPM is disabled", return -1;);
+ if (0 == data->sclk_dpm_key_disabled)
+ return (0 == smum_send_msg_to_smc_with_parameter(
+ hwmgr->smumgr,
+ PPSMC_MSG_DPM_ForceState,
+ n) ? 0 : 1);
+
+ return 0;
+}
+
+/**
+ * force DPM power State
+ *
+ * @param hwmgr: the address of the powerplay hardware manager.
+ * @param n : DPM level
+ * @return The response that came from the SMC.
+ */
+int iceland_dpm_force_state_mclk(struct pp_hwmgr *hwmgr, uint32_t n)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ /* Checking if DPM is running. If we discover hang because of this, we should skip this message. */
+ PP_ASSERT_WITH_CODE(0 == iceland_is_dpm_running(hwmgr),
+ "Trying to Force MCLK when DPM is disabled", return -1;);
+ if (0 == data->mclk_dpm_key_disabled)
+ return (0 == smum_send_msg_to_smc_with_parameter(
+ hwmgr->smumgr,
+ PPSMC_MSG_MCLKDPM_ForceState,
+ n) ? 0 : 1);
+
+ return 0;
+}
+
+/**
+ * force DPM power State
+ *
+ * @param hwmgr: the address of the powerplay hardware manager.
+ * @param n : DPM level
+ * @return The response that came from the SMC.
+ */
+int iceland_dpm_force_state_pcie(struct pp_hwmgr *hwmgr, uint32_t n)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ /* Checking if DPM is running. If we discover hang because of this, we should skip this message.*/
+ PP_ASSERT_WITH_CODE(0 == iceland_is_dpm_running(hwmgr),
+ "Trying to Force PCIE level when DPM is disabled", return -1;);
+ if (0 == data->pcie_dpm_key_disabled)
+ return (0 == smum_send_msg_to_smc_with_parameter(
+ hwmgr->smumgr,
+ PPSMC_MSG_PCIeDPM_ForceLevel,
+ n) ? 0 : 1);
+
+ return 0;
+}
+
+static int iceland_force_dpm_highest(struct pp_hwmgr *hwmgr)
+{
+ uint32_t level, tmp;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ if (0 == data->sclk_dpm_key_disabled) {
+ /* SCLK */
+ if (data->dpm_level_enable_mask.sclk_dpm_enable_mask != 0) {
+ level = 0;
+ tmp = data->dpm_level_enable_mask.sclk_dpm_enable_mask;
+ while (tmp >>= 1)
+ level++ ;
+
+ if (0 != level) {
+ PP_ASSERT_WITH_CODE((0 == iceland_dpm_force_state(hwmgr, level)),
+ "force highest sclk dpm state failed!", return -1);
+ PHM_WAIT_INDIRECT_FIELD(hwmgr->device,
+ SMC_IND, TARGET_AND_CURRENT_PROFILE_INDEX, CURR_SCLK_INDEX, level);
+ }
+ }
+ }
+
+ if (0 == data->mclk_dpm_key_disabled) {
+ /* MCLK */
+ if (data->dpm_level_enable_mask.mclk_dpm_enable_mask != 0) {
+ level = 0;
+ tmp = data->dpm_level_enable_mask.mclk_dpm_enable_mask;
+ while (tmp >>= 1)
+ level++ ;
+
+ if (0 != level) {
+ PP_ASSERT_WITH_CODE((0 == iceland_dpm_force_state_mclk(hwmgr, level)),
+ "force highest mclk dpm state failed!", return -1);
+ PHM_WAIT_INDIRECT_FIELD(hwmgr->device, SMC_IND,
+ TARGET_AND_CURRENT_PROFILE_INDEX, CURR_MCLK_INDEX, level);
+ }
+ }
+ }
+
+ if (0 == data->pcie_dpm_key_disabled) {
+ /* PCIE */
+ if (data->dpm_level_enable_mask.pcie_dpm_enable_mask != 0) {
+ level = 0;
+ tmp = data->dpm_level_enable_mask.pcie_dpm_enable_mask;
+ while (tmp >>= 1)
+ level++ ;
+
+ if (0 != level) {
+ PP_ASSERT_WITH_CODE((0 == iceland_dpm_force_state_pcie(hwmgr, level)),
+ "force highest pcie dpm state failed!", return -1);
+ }
+ }
+ }
+
+ return 0;
+}
+
+static uint32_t iceland_get_lowest_enable_level(struct pp_hwmgr *hwmgr,
+ uint32_t level_mask)
+{
+ uint32_t level = 0;
+
+ while (0 == (level_mask & (1 << level)))
+ level++;
+
+ return level;
+}
+
+static int iceland_force_dpm_lowest(struct pp_hwmgr *hwmgr)
+{
+ uint32_t level;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ /* for now force only sclk */
+ if (0 != data->dpm_level_enable_mask.sclk_dpm_enable_mask) {
+ level = iceland_get_lowest_enable_level(hwmgr,
+ data->dpm_level_enable_mask.sclk_dpm_enable_mask);
+
+ PP_ASSERT_WITH_CODE((0 == iceland_dpm_force_state(hwmgr, level)),
+ "force sclk dpm state failed!", return -1);
+
+ PHM_WAIT_INDIRECT_FIELD(hwmgr->device, SMC_IND,
+ TARGET_AND_CURRENT_PROFILE_INDEX,
+ CURR_SCLK_INDEX,
+ level);
+ }
+
+ return 0;
+}
+
+int iceland_unforce_dpm_levels(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ PP_ASSERT_WITH_CODE (0 == iceland_is_dpm_running(hwmgr),
+ "Trying to Unforce DPM when DPM is disabled. Returning without sending SMC message.",
+ return -1);
+
+ if (0 == data->sclk_dpm_key_disabled) {
+ PP_ASSERT_WITH_CODE((0 == smum_send_msg_to_smc(
+ hwmgr->smumgr,
+ PPSMC_MSG_NoForcedLevel)),
+ "unforce sclk dpm state failed!",
+ return -1);
+ }
+
+ if (0 == data->mclk_dpm_key_disabled) {
+ PP_ASSERT_WITH_CODE((0 == smum_send_msg_to_smc(
+ hwmgr->smumgr,
+ PPSMC_MSG_MCLKDPM_NoForcedLevel)),
+ "unforce mclk dpm state failed!",
+ return -1);
+ }
+
+ if (0 == data->pcie_dpm_key_disabled) {
+ PP_ASSERT_WITH_CODE((0 == smum_send_msg_to_smc(
+ hwmgr->smumgr,
+ PPSMC_MSG_PCIeDPM_UnForceLevel)),
+ "unforce pcie level failed!",
+ return -1);
+ }
+
+ return 0;
+}
+
+static int iceland_force_dpm_level(struct pp_hwmgr *hwmgr,
+ enum amd_dpm_forced_level level)
+{
+ int ret = 0;
+
+ switch (level) {
+ case AMD_DPM_FORCED_LEVEL_HIGH:
+ ret = iceland_force_dpm_highest(hwmgr);
+ if (ret)
+ return ret;
+ break;
+ case AMD_DPM_FORCED_LEVEL_LOW:
+ ret = iceland_force_dpm_lowest(hwmgr);
+ if (ret)
+ return ret;
+ break;
+ case AMD_DPM_FORCED_LEVEL_AUTO:
+ ret = iceland_unforce_dpm_levels(hwmgr);
+ if (ret)
+ return ret;
+ break;
+ default:
+ break;
+ }
+
+ hwmgr->dpm_level = level;
+ return ret;
+}
+
+const struct iceland_power_state *cast_const_phw_iceland_power_state(
+ const struct pp_hw_power_state *hw_ps)
+{
+ if (hw_ps == NULL)
+ return NULL;
+
+ PP_ASSERT_WITH_CODE((PhwIceland_Magic == hw_ps->magic),
+ "Invalid Powerstate Type!",
+ return NULL);
+
+ return (const struct iceland_power_state *)hw_ps;
+}
+
+static int iceland_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, const void *input)
+{
+ const struct phm_set_power_state_input *states = (const struct phm_set_power_state_input *)input;
+ const struct iceland_power_state *iceland_ps = cast_const_phw_iceland_power_state(states->pnew_state);
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_single_dpm_table *psclk_table = &(data->dpm_table.sclk_table);
+ uint32_t sclk = iceland_ps->performance_levels[iceland_ps->performance_level_count-1].engine_clock;
+ struct iceland_single_dpm_table *pmclk_table = &(data->dpm_table.mclk_table);
+ uint32_t mclk = iceland_ps->performance_levels[iceland_ps->performance_level_count-1].memory_clock;
+ struct PP_Clocks min_clocks = {0};
+ uint32_t i;
+ struct cgs_display_info info = {0};
+
+ data->need_update_smu7_dpm_table = 0;
+
+ for (i = 0; i < psclk_table->count; i++) {
+ if (sclk == psclk_table->dpm_levels[i].value)
+ break;
+ }
+
+ if (i >= psclk_table->count)
+ data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
+ else {
+ /*
+ * TODO: Check SCLK in DAL's minimum clocks in case DeepSleep
+ * divider update is required.
+ */
+ if(data->display_timing.min_clock_insr != min_clocks.engineClockInSR)
+ data->need_update_smu7_dpm_table |= DPMTABLE_UPDATE_SCLK;
+ }
+
+ for (i = 0; i < pmclk_table->count; i++) {
+ if (mclk == pmclk_table->dpm_levels[i].value)
+ break;
+ }
+
+ if (i >= pmclk_table->count)
+ data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
+
+ cgs_get_active_displays_info(hwmgr->device, &info);
+
+ if (data->display_timing.num_existing_displays != info.display_count)
+ data->need_update_smu7_dpm_table |= DPMTABLE_UPDATE_MCLK;
+
+ return 0;
+}
+
+static uint16_t iceland_get_maximum_link_speed(struct pp_hwmgr *hwmgr, const struct iceland_power_state *hw_ps)
+{
+ uint32_t i;
+ uint32_t pcie_speed, max_speed = 0;
+
+ for (i = 0; i < hw_ps->performance_level_count; i++) {
+ pcie_speed = hw_ps->performance_levels[i].pcie_gen;
+ if (max_speed < pcie_speed)
+ max_speed = pcie_speed;
+ }
+
+ return max_speed;
+}
+
+static uint16_t iceland_get_current_pcie_speed(struct pp_hwmgr *hwmgr)
+{
+ uint32_t speed_cntl = 0;
+
+ speed_cntl = cgs_read_ind_register(hwmgr->device,
+ CGS_IND_REG__PCIE,
+ ixPCIE_LC_SPEED_CNTL);
+ return((uint16_t)PHM_GET_FIELD(speed_cntl,
+ PCIE_LC_SPEED_CNTL, LC_CURRENT_DATA_RATE));
+}
+
+
+static int iceland_request_link_speed_change_before_state_change(struct pp_hwmgr *hwmgr, const void *input)
+{
+ const struct phm_set_power_state_input *states = (const struct phm_set_power_state_input *)input;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ const struct iceland_power_state *iceland_nps = cast_const_phw_iceland_power_state(states->pnew_state);
+ const struct iceland_power_state *iceland_cps = cast_const_phw_iceland_power_state(states->pcurrent_state);
+
+ uint16_t target_link_speed = iceland_get_maximum_link_speed(hwmgr, iceland_nps);
+ uint16_t current_link_speed;
+
+ if (data->force_pcie_gen == PP_PCIEGenInvalid)
+ current_link_speed = iceland_get_maximum_link_speed(hwmgr, iceland_cps);
+ else
+ current_link_speed = data->force_pcie_gen;
+
+ data->force_pcie_gen = PP_PCIEGenInvalid;
+ data->pspp_notify_required = false;
+ if (target_link_speed > current_link_speed) {
+ switch(target_link_speed) {
+ case PP_PCIEGen3:
+ if (0 == acpi_pcie_perf_request(hwmgr->device, PCIE_PERF_REQ_GEN3, false))
+ break;
+ data->force_pcie_gen = PP_PCIEGen2;
+ if (current_link_speed == PP_PCIEGen2)
+ break;
+ case PP_PCIEGen2:
+ if (0 == acpi_pcie_perf_request(hwmgr->device, PCIE_PERF_REQ_GEN2, false))
+ break;
+ default:
+ data->force_pcie_gen = iceland_get_current_pcie_speed(hwmgr);
+ break;
+ }
+ } else {
+ if (target_link_speed < current_link_speed)
+ data->pspp_notify_required = true;
+ }
+
+ return 0;
+}
+
+static int iceland_freeze_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ if (0 == data->need_update_smu7_dpm_table)
+ return 0;
+
+ if ((0 == data->sclk_dpm_key_disabled) &&
+ (data->need_update_smu7_dpm_table &
+ (DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_UPDATE_SCLK))) {
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_is_dpm_running(hwmgr),
+ "Trying to freeze SCLK DPM when DPM is disabled",
+ );
+ PP_ASSERT_WITH_CODE(
+ 0 == smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_SCLKDPM_FreezeLevel),
+ "Failed to freeze SCLK DPM during FreezeSclkMclkDPM Function!",
+ return -1);
+ }
+
+ if ((0 == data->mclk_dpm_key_disabled) &&
+ (data->need_update_smu7_dpm_table &
+ DPMTABLE_OD_UPDATE_MCLK)) {
+ PP_ASSERT_WITH_CODE(0 == iceland_is_dpm_running(hwmgr),
+ "Trying to freeze MCLK DPM when DPM is disabled",
+ );
+ PP_ASSERT_WITH_CODE(
+ 0 == smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_MCLKDPM_FreezeLevel),
+ "Failed to freeze MCLK DPM during FreezeSclkMclkDPM Function!",
+ return -1);
+ }
+
+ return 0;
+}
+
+static int iceland_populate_and_upload_sclk_mclk_dpm_levels(struct pp_hwmgr *hwmgr, const void *input)
+{
+ int result = 0;
+
+ const struct phm_set_power_state_input *states = (const struct phm_set_power_state_input *)input;
+ const struct iceland_power_state *iceland_ps = cast_const_phw_iceland_power_state(states->pnew_state);
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint32_t sclk = iceland_ps->performance_levels[iceland_ps->performance_level_count-1].engine_clock;
+ uint32_t mclk = iceland_ps->performance_levels[iceland_ps->performance_level_count-1].memory_clock;
+ struct iceland_dpm_table *pdpm_table = &data->dpm_table;
+
+ struct iceland_dpm_table *pgolden_dpm_table = &data->golden_dpm_table;
+ uint32_t dpm_count, clock_percent;
+ uint32_t i;
+
+ if (0 == data->need_update_smu7_dpm_table)
+ return 0;
+
+ if (data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_SCLK) {
+ pdpm_table->sclk_table.dpm_levels[pdpm_table->sclk_table.count-1].value = sclk;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_OD6PlusinACSupport) ||
+ phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_OD6PlusinDCSupport)) {
+ /*
+ * Need to do calculation based on the golden DPM table
+ * as the Heatmap GPU Clock axis is also based on the default values
+ */
+ PP_ASSERT_WITH_CODE(
+ (pgolden_dpm_table->sclk_table.dpm_levels[pgolden_dpm_table->sclk_table.count-1].value != 0),
+ "Divide by 0!",
+ return -1);
+ dpm_count = pdpm_table->sclk_table.count < 2 ? 0 : pdpm_table->sclk_table.count-2;
+ for (i = dpm_count; i > 1; i--) {
+ if (sclk > pgolden_dpm_table->sclk_table.dpm_levels[pgolden_dpm_table->sclk_table.count-1].value) {
+ clock_percent = ((sclk - pgolden_dpm_table->sclk_table.dpm_levels[pgolden_dpm_table->sclk_table.count-1].value)*100) /
+ pgolden_dpm_table->sclk_table.dpm_levels[pgolden_dpm_table->sclk_table.count-1].value;
+
+ pdpm_table->sclk_table.dpm_levels[i].value =
+ pgolden_dpm_table->sclk_table.dpm_levels[i].value +
+ (pgolden_dpm_table->sclk_table.dpm_levels[i].value * clock_percent)/100;
+
+ } else if (pgolden_dpm_table->sclk_table.dpm_levels[pdpm_table->sclk_table.count-1].value > sclk) {
+ clock_percent = ((pgolden_dpm_table->sclk_table.dpm_levels[pgolden_dpm_table->sclk_table.count-1].value - sclk)*100) /
+ pgolden_dpm_table->sclk_table.dpm_levels[pgolden_dpm_table->sclk_table.count-1].value;
+
+ pdpm_table->sclk_table.dpm_levels[i].value =
+ pgolden_dpm_table->sclk_table.dpm_levels[i].value -
+ (pgolden_dpm_table->sclk_table.dpm_levels[i].value * clock_percent)/100;
+ } else
+ pdpm_table->sclk_table.dpm_levels[i].value =
+ pgolden_dpm_table->sclk_table.dpm_levels[i].value;
+ }
+ }
+ }
+
+ if (data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_MCLK) {
+ pdpm_table->mclk_table.dpm_levels[pdpm_table->mclk_table.count-1].value = mclk;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_OD6PlusinACSupport) ||
+ phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_OD6PlusinDCSupport)) {
+
+ PP_ASSERT_WITH_CODE(
+ (pgolden_dpm_table->mclk_table.dpm_levels[pgolden_dpm_table->mclk_table.count-1].value != 0),
+ "Divide by 0!",
+ return -1);
+ dpm_count = pdpm_table->mclk_table.count < 2? 0 : pdpm_table->mclk_table.count-2;
+ for (i = dpm_count; i > 1; i--) {
+ if (mclk > pgolden_dpm_table->mclk_table.dpm_levels[pgolden_dpm_table->mclk_table.count-1].value) {
+ clock_percent = ((mclk - pgolden_dpm_table->mclk_table.dpm_levels[pgolden_dpm_table->mclk_table.count-1].value)*100) /
+ pgolden_dpm_table->mclk_table.dpm_levels[pgolden_dpm_table->mclk_table.count-1].value;
+
+ pdpm_table->mclk_table.dpm_levels[i].value =
+ pgolden_dpm_table->mclk_table.dpm_levels[i].value +
+ (pgolden_dpm_table->mclk_table.dpm_levels[i].value * clock_percent)/100;
+
+ } else if (pgolden_dpm_table->mclk_table.dpm_levels[pdpm_table->mclk_table.count-1].value > mclk) {
+ clock_percent = ((pgolden_dpm_table->mclk_table.dpm_levels[pgolden_dpm_table->mclk_table.count-1].value - mclk)*100) /
+ pgolden_dpm_table->mclk_table.dpm_levels[pgolden_dpm_table->mclk_table.count-1].value;
+
+ pdpm_table->mclk_table.dpm_levels[i].value =
+ pgolden_dpm_table->mclk_table.dpm_levels[i].value -
+ (pgolden_dpm_table->mclk_table.dpm_levels[i].value * clock_percent)/100;
+ } else
+ pdpm_table->mclk_table.dpm_levels[i].value = pgolden_dpm_table->mclk_table.dpm_levels[i].value;
+ }
+ }
+ }
+
+
+ if (data->need_update_smu7_dpm_table & (DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_UPDATE_SCLK)) {
+ result = iceland_populate_all_graphic_levels(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Failed to populate SCLK during PopulateNewDPMClocksStates Function!",
+ return result);
+ }
+
+ if (data->need_update_smu7_dpm_table & (DPMTABLE_OD_UPDATE_MCLK + DPMTABLE_UPDATE_MCLK)) {
+ /*populate MCLK dpm table to SMU7 */
+ result = iceland_populate_all_memory_levels(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Failed to populate MCLK during PopulateNewDPMClocksStates Function!",
+ return result);
+ }
+
+ return result;
+}
+
+static int iceland_trim_single_dpm_states(struct pp_hwmgr *hwmgr,
+ struct iceland_single_dpm_table *pdpm_table,
+ uint32_t low_limit, uint32_t high_limit)
+{
+ uint32_t i;
+
+ for (i = 0; i < pdpm_table->count; i++) {
+ if ((pdpm_table->dpm_levels[i].value < low_limit) ||
+ (pdpm_table->dpm_levels[i].value > high_limit))
+ pdpm_table->dpm_levels[i].enabled = false;
+ else
+ pdpm_table->dpm_levels[i].enabled = true;
+ }
+ return 0;
+}
+
+static int iceland_trim_dpm_states(struct pp_hwmgr *hwmgr, const struct iceland_power_state *hw_state)
+{
+ int result = 0;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint32_t high_limit_count;
+
+ PP_ASSERT_WITH_CODE((hw_state->performance_level_count >= 1),
+ "power state did not have any performance level",
+ return -1);
+
+ high_limit_count = (1 == hw_state->performance_level_count) ? 0: 1;
+
+ iceland_trim_single_dpm_states(hwmgr, &(data->dpm_table.sclk_table),
+ hw_state->performance_levels[0].engine_clock,
+ hw_state->performance_levels[high_limit_count].engine_clock);
+
+ iceland_trim_single_dpm_states(hwmgr, &(data->dpm_table.mclk_table),
+ hw_state->performance_levels[0].memory_clock,
+ hw_state->performance_levels[high_limit_count].memory_clock);
+
+ return result;
+}
+
+static int iceland_generate_dpm_level_enable_mask(struct pp_hwmgr *hwmgr, const void *input)
+{
+ int result;
+ const struct phm_set_power_state_input *states = (const struct phm_set_power_state_input *)input;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ const struct iceland_power_state *iceland_ps = cast_const_phw_iceland_power_state(states->pnew_state);
+
+ result = iceland_trim_dpm_states(hwmgr, iceland_ps);
+ if (0 != result)
+ return result;
+
+ data->dpm_level_enable_mask.sclk_dpm_enable_mask = iceland_get_dpm_level_enable_mask_value(&data->dpm_table.sclk_table);
+ data->dpm_level_enable_mask.mclk_dpm_enable_mask = iceland_get_dpm_level_enable_mask_value(&data->dpm_table.mclk_table);
+ data->last_mclk_dpm_enable_mask = data->dpm_level_enable_mask.mclk_dpm_enable_mask;
+ if (data->uvd_enabled && (data->dpm_level_enable_mask.mclk_dpm_enable_mask & 1))
+ data->dpm_level_enable_mask.mclk_dpm_enable_mask &= 0xFFFFFFFE;
+
+ data->dpm_level_enable_mask.pcie_dpm_enable_mask = iceland_get_dpm_level_enable_mask_value(&data->dpm_table.pcie_speed_table);
+
+ return 0;
+}
+
+static int iceland_update_vce_dpm(struct pp_hwmgr *hwmgr, const void *input)
+{
+ return 0;
+}
+
+static int iceland_update_sclk_threshold(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ int result = 0;
+ uint32_t low_sclk_interrupt_threshold = 0;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SclkThrottleLowNotification)
+ && (hwmgr->gfx_arbiter.sclk_threshold != data->low_sclk_interrupt_threshold)) {
+ data->low_sclk_interrupt_threshold = hwmgr->gfx_arbiter.sclk_threshold;
+ low_sclk_interrupt_threshold = data->low_sclk_interrupt_threshold;
+
+ CONVERT_FROM_HOST_TO_SMC_UL(low_sclk_interrupt_threshold);
+
+ result = iceland_copy_bytes_to_smc(
+ hwmgr->smumgr,
+ data->dpm_table_start + offsetof(SMU71_Discrete_DpmTable,
+ LowSclkInterruptThreshold),
+ (uint8_t *)&low_sclk_interrupt_threshold,
+ sizeof(uint32_t),
+ data->sram_end
+ );
+ }
+
+ return result;
+}
+
+static int iceland_update_and_upload_mc_reg_table(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ uint32_t address;
+ int32_t result;
+
+ if (0 == (data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_MCLK))
+ return 0;
+
+
+ memset(&data->mc_reg_table, 0, sizeof(SMU71_Discrete_MCRegisters));
+
+ result = iceland_convert_mc_reg_table_to_smc(hwmgr, &(data->mc_reg_table));
+
+ if(result != 0)
+ return result;
+
+
+ address = data->mc_reg_table_start + (uint32_t)offsetof(SMU71_Discrete_MCRegisters, data[0]);
+
+ return iceland_copy_bytes_to_smc(hwmgr->smumgr, address,
+ (uint8_t *)&data->mc_reg_table.data[0],
+ sizeof(SMU71_Discrete_MCRegisterSet) * data->dpm_table.mclk_table.count,
+ data->sram_end);
+}
+
+static int iceland_program_memory_timing_parameters_conditionally(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ if (data->need_update_smu7_dpm_table &
+ (DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_OD_UPDATE_MCLK))
+ return iceland_program_memory_timing_parameters(hwmgr);
+
+ return 0;
+}
+
+static int iceland_unfreeze_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ if (0 == data->need_update_smu7_dpm_table)
+ return 0;
+
+ if ((0 == data->sclk_dpm_key_disabled) &&
+ (data->need_update_smu7_dpm_table &
+ (DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_UPDATE_SCLK))) {
+
+ PP_ASSERT_WITH_CODE(0 == iceland_is_dpm_running(hwmgr),
+ "Trying to Unfreeze SCLK DPM when DPM is disabled",
+ );
+ PP_ASSERT_WITH_CODE(
+ 0 == smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_SCLKDPM_UnfreezeLevel),
+ "Failed to unfreeze SCLK DPM during UnFreezeSclkMclkDPM Function!",
+ return -1);
+ }
+
+ if ((0 == data->mclk_dpm_key_disabled) &&
+ (data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_MCLK)) {
+
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_is_dpm_running(hwmgr),
+ "Trying to Unfreeze MCLK DPM when DPM is disabled",
+ );
+ PP_ASSERT_WITH_CODE(
+ 0 == smum_send_msg_to_smc(hwmgr->smumgr,
+ PPSMC_MSG_MCLKDPM_UnfreezeLevel),
+ "Failed to unfreeze MCLK DPM during UnFreezeSclkMclkDPM Function!",
+ return -1);
+ }
+
+ data->need_update_smu7_dpm_table = 0;
+
+ return 0;
+}
+
+static int iceland_notify_link_speed_change_after_state_change(struct pp_hwmgr *hwmgr, const void *input)
+{
+ const struct phm_set_power_state_input *states = (const struct phm_set_power_state_input *)input;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ const struct iceland_power_state *iceland_ps = cast_const_phw_iceland_power_state(states->pnew_state);
+ uint16_t target_link_speed = iceland_get_maximum_link_speed(hwmgr, iceland_ps);
+ uint8_t request;
+
+ if (data->pspp_notify_required ||
+ data->pcie_performance_request) {
+ if (target_link_speed == PP_PCIEGen3)
+ request = PCIE_PERF_REQ_GEN3;
+ else if (target_link_speed == PP_PCIEGen2)
+ request = PCIE_PERF_REQ_GEN2;
+ else
+ request = PCIE_PERF_REQ_GEN1;
+
+ if(request == PCIE_PERF_REQ_GEN1 && iceland_get_current_pcie_speed(hwmgr) > 0) {
+ data->pcie_performance_request = false;
+ return 0;
+ }
+
+ if (0 != acpi_pcie_perf_request(hwmgr->device, request, false)) {
+ if (PP_PCIEGen2 == target_link_speed)
+ printk("PSPP request to switch to Gen2 from Gen3 Failed!");
+ else
+ printk("PSPP request to switch to Gen1 from Gen2 Failed!");
+ }
+ }
+
+ data->pcie_performance_request = false;
+ return 0;
+}
+
+int iceland_upload_dpm_level_enable_mask(struct pp_hwmgr *hwmgr)
+{
+ PPSMC_Result result;
+ iceland_hwmgr *data = (iceland_hwmgr *)(hwmgr->backend);
+
+ if (0 == data->sclk_dpm_key_disabled) {
+ /* Checking if DPM is running. If we discover hang because of this, we should skip this message.*/
+ if (0 != iceland_is_dpm_running(hwmgr))
+ printk(KERN_ERR "[ powerplay ] Trying to set Enable Sclk Mask when DPM is disabled \n");
+
+ if (0 != data->dpm_level_enable_mask.sclk_dpm_enable_mask) {
+ result = smum_send_msg_to_smc_with_parameter(
+ hwmgr->smumgr,
+ (PPSMC_Msg)PPSMC_MSG_SCLKDPM_SetEnabledMask,
+ data->dpm_level_enable_mask.sclk_dpm_enable_mask);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Set Sclk Dpm enable Mask failed", return -1);
+ }
+ }
+
+ if (0 == data->mclk_dpm_key_disabled) {
+ /* Checking if DPM is running. If we discover hang because of this, we should skip this message.*/
+ if (0 != iceland_is_dpm_running(hwmgr))
+ printk(KERN_ERR "[ powerplay ] Trying to set Enable Mclk Mask when DPM is disabled \n");
+
+ if (0 != data->dpm_level_enable_mask.mclk_dpm_enable_mask) {
+ result = smum_send_msg_to_smc_with_parameter(
+ hwmgr->smumgr,
+ (PPSMC_Msg)PPSMC_MSG_MCLKDPM_SetEnabledMask,
+ data->dpm_level_enable_mask.mclk_dpm_enable_mask);
+ PP_ASSERT_WITH_CODE((0 == result),
+ "Set Mclk Dpm enable Mask failed", return -1);
+ }
+ }
+
+ return 0;
+}
+
+static int iceland_set_power_state_tasks(struct pp_hwmgr *hwmgr, const void *input)
+{
+ int tmp_result, result = 0;
+
+ tmp_result = iceland_find_dpm_states_clocks_in_dpm_table(hwmgr, input);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to find DPM states clocks in DPM table!", result = tmp_result);
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest)) {
+ tmp_result = iceland_request_link_speed_change_before_state_change(hwmgr, input);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to request link speed change before state change!", result = tmp_result);
+ }
+
+ tmp_result = iceland_freeze_sclk_mclk_dpm(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to freeze SCLK MCLK DPM!", result = tmp_result);
+
+ tmp_result = iceland_populate_and_upload_sclk_mclk_dpm_levels(hwmgr, input);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to populate and upload SCLK MCLK DPM levels!", result = tmp_result);
+
+ tmp_result = iceland_generate_dpm_level_enable_mask(hwmgr, input);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to generate DPM level enabled mask!", result = tmp_result);
+
+ tmp_result = iceland_update_vce_dpm(hwmgr, input);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to update VCE DPM!", result = tmp_result);
+
+ tmp_result = iceland_update_sclk_threshold(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to update SCLK threshold!", result = tmp_result);
+
+ tmp_result = iceland_update_and_upload_mc_reg_table(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to upload MC reg table!", result = tmp_result);
+
+ tmp_result = iceland_program_memory_timing_parameters_conditionally(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to program memory timing parameters!", result = tmp_result);
+
+ tmp_result = iceland_unfreeze_sclk_mclk_dpm(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to unfreeze SCLK MCLK DPM!", result = tmp_result);
+
+ tmp_result = iceland_upload_dpm_level_enable_mask(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to upload DPM level enabled mask!", result = tmp_result);
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_PCIEPerformanceRequest)) {
+ tmp_result = iceland_notify_link_speed_change_after_state_change(hwmgr, input);
+ PP_ASSERT_WITH_CODE((0 == tmp_result), "Failed to notify link speed change after state change!", result = tmp_result);
+ }
+
+ return result;
+}
+
+static int iceland_get_power_state_size(struct pp_hwmgr *hwmgr)
+{
+ return sizeof(struct iceland_power_state);
+}
+
+static int iceland_dpm_get_mclk(struct pp_hwmgr *hwmgr, bool low)
+{
+ struct pp_power_state *ps;
+ struct iceland_power_state *iceland_ps;
+
+ if (hwmgr == NULL)
+ return -EINVAL;
+
+ ps = hwmgr->request_ps;
+
+ if (ps == NULL)
+ return -EINVAL;
+
+ iceland_ps = cast_phw_iceland_power_state(&ps->hardware);
+
+ if (low)
+ return iceland_ps->performance_levels[0].memory_clock;
+ else
+ return iceland_ps->performance_levels[iceland_ps->performance_level_count-1].memory_clock;
+}
+
+static int iceland_dpm_get_sclk(struct pp_hwmgr *hwmgr, bool low)
+{
+ struct pp_power_state *ps;
+ struct iceland_power_state *iceland_ps;
+
+ if (hwmgr == NULL)
+ return -EINVAL;
+
+ ps = hwmgr->request_ps;
+
+ if (ps == NULL)
+ return -EINVAL;
+
+ iceland_ps = cast_phw_iceland_power_state(&ps->hardware);
+
+ if (low)
+ return iceland_ps->performance_levels[0].engine_clock;
+ else
+ return iceland_ps->performance_levels[iceland_ps->performance_level_count-1].engine_clock;
+}
+
+static int iceland_get_current_pcie_lane_number(
+ struct pp_hwmgr *hwmgr)
+{
+ uint32_t link_width;
+
+ link_width = PHM_READ_INDIRECT_FIELD(hwmgr->device,
+ CGS_IND_REG__PCIE,
+ PCIE_LC_LINK_WIDTH_CNTL,
+ LC_LINK_WIDTH_RD);
+
+ PP_ASSERT_WITH_CODE((7 >= link_width),
+ "Invalid PCIe lane width!", return 0);
+
+ return decode_pcie_lane_width(link_width);
+}
+
+static int iceland_dpm_patch_boot_state(struct pp_hwmgr *hwmgr,
+ struct pp_hw_power_state *hw_ps)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_power_state *ps = (struct iceland_power_state *)hw_ps;
+ ATOM_FIRMWARE_INFO_V2_2 *fw_info;
+ uint16_t size;
+ uint8_t frev, crev;
+ int index = GetIndexIntoMasterTable(DATA, FirmwareInfo);
+
+ /* First retrieve the Boot clocks and VDDC from the firmware info table.
+ * We assume here that fw_info is unchanged if this call fails.
+ */
+ fw_info = (ATOM_FIRMWARE_INFO_V2_2 *)cgs_atom_get_data_table(
+ hwmgr->device, index,
+ &size, &frev, &crev);
+ if (!fw_info)
+ /* During a test, there is no firmware info table. */
+ return 0;
+
+ /* Patch the state. */
+ data->vbios_boot_state.sclk_bootup_value = le32_to_cpu(fw_info->ulDefaultEngineClock);
+ data->vbios_boot_state.mclk_bootup_value = le32_to_cpu(fw_info->ulDefaultMemoryClock);
+ data->vbios_boot_state.mvdd_bootup_value = le16_to_cpu(fw_info->usBootUpMVDDCVoltage);
+ data->vbios_boot_state.vddc_bootup_value = le16_to_cpu(fw_info->usBootUpVDDCVoltage);
+ data->vbios_boot_state.vddci_bootup_value = le16_to_cpu(fw_info->usBootUpVDDCIVoltage);
+ data->vbios_boot_state.pcie_gen_bootup_value = iceland_get_current_pcie_speed(hwmgr);
+ data->vbios_boot_state.pcie_lane_bootup_value =
+ (uint16_t)iceland_get_current_pcie_lane_number(hwmgr);
+
+ /* set boot power state */
+ ps->performance_levels[0].memory_clock = data->vbios_boot_state.mclk_bootup_value;
+ ps->performance_levels[0].engine_clock = data->vbios_boot_state.sclk_bootup_value;
+ ps->performance_levels[0].pcie_gen = data->vbios_boot_state.pcie_gen_bootup_value;
+ ps->performance_levels[0].pcie_lane = data->vbios_boot_state.pcie_lane_bootup_value;
+
+ return 0;
+}
+
+static int iceland_get_pp_table_entry_callback_func(struct pp_hwmgr *hwmgr,
+ struct pp_hw_power_state *power_state,
+ unsigned int index, const void *clock_info)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_power_state *iceland_power_state = cast_phw_iceland_power_state(power_state);
+ const ATOM_PPLIB_CI_CLOCK_INFO *visland_clk_info = clock_info;
+ struct iceland_performance_level *performance_level;
+ uint32_t engine_clock, memory_clock;
+ uint16_t pcie_gen_from_bios;
+
+ engine_clock = visland_clk_info->ucEngineClockHigh << 16 | visland_clk_info->usEngineClockLow;
+ memory_clock = visland_clk_info->ucMemoryClockHigh << 16 | visland_clk_info->usMemoryClockLow;
+
+ if (!(data->mc_micro_code_feature & DISABLE_MC_LOADMICROCODE) && memory_clock > data->highest_mclk)
+ data->highest_mclk = memory_clock;
+
+ performance_level = &(iceland_power_state->performance_levels
+ [iceland_power_state->performance_level_count++]);
+
+ PP_ASSERT_WITH_CODE(
+ (iceland_power_state->performance_level_count < SMU71_MAX_LEVELS_GRAPHICS),
+ "Performance levels exceeds SMC limit!",
+ return -1);
+
+ PP_ASSERT_WITH_CODE(
+ (iceland_power_state->performance_level_count <=
+ hwmgr->platform_descriptor.hardwareActivityPerformanceLevels),
+ "Performance levels exceeds Driver limit!",
+ return -1);
+
+ /* Performance levels are arranged from low to high. */
+ performance_level->memory_clock = memory_clock;
+ performance_level->engine_clock = engine_clock;
+
+ pcie_gen_from_bios = visland_clk_info->ucPCIEGen;
+
+ performance_level->pcie_gen = get_pcie_gen_support(data->pcie_gen_cap, pcie_gen_from_bios);
+ performance_level->pcie_lane = get_pcie_lane_support(data->pcie_lane_cap, visland_clk_info->usPCIELane);
+
+ return 0;
+}
+
+static int iceland_get_pp_table_entry(struct pp_hwmgr *hwmgr,
+ unsigned long entry_index, struct pp_power_state *state)
+{
+ int result;
+ struct iceland_power_state *ps;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct phm_clock_voltage_dependency_table *dep_mclk_table =
+ hwmgr->dyn_state.vddci_dependency_on_mclk;
+
+ memset(&state->hardware, 0x00, sizeof(struct pp_hw_power_state));
+
+ state->hardware.magic = PHM_VIslands_Magic;
+
+ ps = (struct iceland_power_state *)(&state->hardware);
+
+ result = pp_tables_get_entry(hwmgr, entry_index, state,
+ iceland_get_pp_table_entry_callback_func);
+
+ /*
+ * This is the earliest time we have all the dependency table
+ * and the VBIOS boot state as
+ * PP_Tables_GetPowerPlayTableEntry retrieves the VBIOS boot
+ * state if there is only one VDDCI/MCLK level, check if it's
+ * the same as VBIOS boot state
+ */
+ if (dep_mclk_table != NULL && dep_mclk_table->count == 1) {
+ if (dep_mclk_table->entries[0].clk !=
+ data->vbios_boot_state.mclk_bootup_value)
+ printk(KERN_ERR "Single MCLK entry VDDCI/MCLK dependency table "
+ "does not match VBIOS boot MCLK level");
+ if (dep_mclk_table->entries[0].v !=
+ data->vbios_boot_state.vddci_bootup_value)
+ printk(KERN_ERR "Single VDDCI entry VDDCI/MCLK dependency table "
+ "does not match VBIOS boot VDDCI level");
+ }
+
+ /* set DC compatible flag if this state supports DC */
+ if (!state->validation.disallowOnDC)
+ ps->dc_compatible = true;
+
+ if (state->classification.flags & PP_StateClassificationFlag_ACPI)
+ data->acpi_pcie_gen = ps->performance_levels[0].pcie_gen;
+ else if (0 != (state->classification.flags & PP_StateClassificationFlag_Boot)) {
+ if (data->bacos.best_match == 0xffff) {
+ /* For C.I. use boot state as base BACO state */
+ data->bacos.best_match = PP_StateClassificationFlag_Boot;
+ data->bacos.performance_level = ps->performance_levels[0];
+ }
+ }
+
+
+ ps->uvd_clocks.VCLK = state->uvd_clocks.VCLK;
+ ps->uvd_clocks.DCLK = state->uvd_clocks.DCLK;
+
+ if (!result) {
+ uint32_t i;
+
+ switch (state->classification.ui_label) {
+ case PP_StateUILabel_Performance:
+ data->use_pcie_performance_levels = true;
+
+ for (i = 0; i < ps->performance_level_count; i++) {
+ if (data->pcie_gen_performance.max <
+ ps->performance_levels[i].pcie_gen)
+ data->pcie_gen_performance.max =
+ ps->performance_levels[i].pcie_gen;
+
+ if (data->pcie_gen_performance.min >
+ ps->performance_levels[i].pcie_gen)
+ data->pcie_gen_performance.min =
+ ps->performance_levels[i].pcie_gen;
+
+ if (data->pcie_lane_performance.max <
+ ps->performance_levels[i].pcie_lane)
+ data->pcie_lane_performance.max =
+ ps->performance_levels[i].pcie_lane;
+
+ if (data->pcie_lane_performance.min >
+ ps->performance_levels[i].pcie_lane)
+ data->pcie_lane_performance.min =
+ ps->performance_levels[i].pcie_lane;
+ }
+ break;
+ case PP_StateUILabel_Battery:
+ data->use_pcie_power_saving_levels = true;
+
+ for (i = 0; i < ps->performance_level_count; i++) {
+ if (data->pcie_gen_power_saving.max <
+ ps->performance_levels[i].pcie_gen)
+ data->pcie_gen_power_saving.max =
+ ps->performance_levels[i].pcie_gen;
+
+ if (data->pcie_gen_power_saving.min >
+ ps->performance_levels[i].pcie_gen)
+ data->pcie_gen_power_saving.min =
+ ps->performance_levels[i].pcie_gen;
+
+ if (data->pcie_lane_power_saving.max <
+ ps->performance_levels[i].pcie_lane)
+ data->pcie_lane_power_saving.max =
+ ps->performance_levels[i].pcie_lane;
+
+ if (data->pcie_lane_power_saving.min >
+ ps->performance_levels[i].pcie_lane)
+ data->pcie_lane_power_saving.min =
+ ps->performance_levels[i].pcie_lane;
+ }
+ break;
+ default:
+ break;
+ }
+ }
+ return 0;
+}
+
+static void
+iceland_print_current_perforce_level(struct pp_hwmgr *hwmgr, struct seq_file *m)
+{
+ uint32_t sclk, mclk, activity_percent;
+ uint32_t offset;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ smum_send_msg_to_smc(hwmgr->smumgr, (PPSMC_Msg)(PPSMC_MSG_API_GetSclkFrequency));
+
+ sclk = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+
+ smum_send_msg_to_smc(hwmgr->smumgr, (PPSMC_Msg)(PPSMC_MSG_API_GetMclkFrequency));
+
+ mclk = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+ seq_printf(m, "\n [ mclk ]: %u MHz\n\n [ sclk ]: %u MHz\n", mclk/100, sclk/100);
+
+ offset = data->soft_regs_start + offsetof(SMU71_SoftRegisters, AverageGraphicsActivity);
+ activity_percent = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, offset);
+ activity_percent += 0x80;
+ activity_percent >>= 8;
+
+ seq_printf(m, "\n [GPU load]: %u%%\n\n", activity_percent > 100 ? 100 : activity_percent);
+
+ seq_printf(m, "uvd %sabled\n", data->uvd_power_gated ? "dis" : "en");
+
+ seq_printf(m, "vce %sabled\n", data->vce_power_gated ? "dis" : "en");
+}
+
+int iceland_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr)
+{
+ uint32_t num_active_displays = 0;
+ struct cgs_display_info info = {0};
+ info.mode_info = NULL;
+
+ cgs_get_active_displays_info(hwmgr->device, &info);
+
+ num_active_displays = info.display_count;
+
+ if (num_active_displays > 1) /* to do && (pHwMgr->pPECI->displayConfiguration.bMultiMonitorInSync != TRUE)) */
+ iceland_notify_smc_display_change(hwmgr, false);
+ else
+ iceland_notify_smc_display_change(hwmgr, true);
+
+ return 0;
+}
+
+/**
+* Programs the display gap
+*
+* @param hwmgr the address of the powerplay hardware manager.
+* @return always OK
+*/
+int iceland_program_display_gap(struct pp_hwmgr *hwmgr)
+{
+ uint32_t num_active_displays = 0;
+ uint32_t display_gap = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL);
+ uint32_t display_gap2;
+ uint32_t pre_vbi_time_in_us;
+ uint32_t frame_time_in_us;
+ uint32_t ref_clock;
+ uint32_t refresh_rate = 0;
+ struct cgs_display_info info = {0};
+ struct cgs_mode_info mode_info;
+
+ info.mode_info = &mode_info;
+
+ cgs_get_active_displays_info(hwmgr->device, &info);
+ num_active_displays = info.display_count;
+
+ display_gap = PHM_SET_FIELD(display_gap, CG_DISPLAY_GAP_CNTL, DISP_GAP, (num_active_displays > 0)? DISPLAY_GAP_VBLANK_OR_WM : DISPLAY_GAP_IGNORE);
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL, display_gap);
+
+ ref_clock = mode_info.ref_clock;
+ refresh_rate = mode_info.refresh_rate;
+
+ if(0 == refresh_rate)
+ refresh_rate = 60;
+
+ frame_time_in_us = 1000000 / refresh_rate;
+
+ pre_vbi_time_in_us = frame_time_in_us - 200 - mode_info.vblank_time_us;
+ display_gap2 = pre_vbi_time_in_us * (ref_clock / 100);
+
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL2, display_gap2);
+
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SOFT_REGISTERS_TABLE_4, PreVBlankGap, 0x64);
+
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SOFT_REGISTERS_TABLE_5, VBlankTimeout, (frame_time_in_us - pre_vbi_time_in_us));
+
+ if (num_active_displays == 1)
+ iceland_notify_smc_display_change(hwmgr, true);
+
+ return 0;
+}
+
+int iceland_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
+{
+ iceland_program_display_gap(hwmgr);
+
+ return 0;
+}
+
+/**
+* Set maximum target operating fan output PWM
+*
+* @param pHwMgr: the address of the powerplay hardware manager.
+* @param usMaxFanPwm: max operating fan PWM in percents
+* @return The response that came from the SMC.
+*/
+static int iceland_set_max_fan_pwm_output(struct pp_hwmgr *hwmgr, uint16_t us_max_fan_pwm)
+{
+ hwmgr->thermal_controller.advanceFanControlParameters.usMaxFanPWM = us_max_fan_pwm;
+
+ if (phm_is_hw_access_blocked(hwmgr))
+ return 0;
+
+ return (0 == smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, PPSMC_MSG_SetFanPwmMax, us_max_fan_pwm) ? 0 : -1);
+}
+
+/**
+* Set maximum target operating fan output RPM
+*
+* @param pHwMgr: the address of the powerplay hardware manager.
+* @param usMaxFanRpm: max operating fan RPM value.
+* @return The response that came from the SMC.
+*/
+static int iceland_set_max_fan_rpm_output(struct pp_hwmgr *hwmgr, uint16_t us_max_fan_pwm)
+{
+ hwmgr->thermal_controller.advanceFanControlParameters.usMaxFanRPM = us_max_fan_pwm;
+
+ if (phm_is_hw_access_blocked(hwmgr))
+ return 0;
+
+ return (0 == smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, PPSMC_MSG_SetFanRpmMax, us_max_fan_pwm) ? 0 : -1);
+}
+
+static int iceland_dpm_set_interrupt_state(void *private_data,
+ unsigned src_id, unsigned type,
+ int enabled)
+{
+ uint32_t cg_thermal_int;
+ struct pp_hwmgr *hwmgr = ((struct pp_eventmgr *)private_data)->hwmgr;
+
+ if (hwmgr == NULL)
+ return -EINVAL;
+
+ switch (type) {
+ case AMD_THERMAL_IRQ_LOW_TO_HIGH:
+ if (enabled) {
+ cg_thermal_int = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_THERMAL_INT);
+ cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK;
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_THERMAL_INT, cg_thermal_int);
+ } else {
+ cg_thermal_int = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_THERMAL_INT);
+ cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK;
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_THERMAL_INT, cg_thermal_int);
+ }
+ break;
+
+ case AMD_THERMAL_IRQ_HIGH_TO_LOW:
+ if (enabled) {
+ cg_thermal_int = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_THERMAL_INT);
+ cg_thermal_int |= CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK;
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_THERMAL_INT, cg_thermal_int);
+ } else {
+ cg_thermal_int = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_THERMAL_INT);
+ cg_thermal_int &= ~CG_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK;
+ cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_THERMAL_INT, cg_thermal_int);
+ }
+ break;
+ default:
+ break;
+ }
+ return 0;
+}
+
+static int iceland_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr,
+ const void *thermal_interrupt_info)
+{
+ int result;
+ const struct pp_interrupt_registration_info *info =
+ (const struct pp_interrupt_registration_info *)thermal_interrupt_info;
+
+ if (info == NULL)
+ return -EINVAL;
+
+ result = cgs_add_irq_source(hwmgr->device, 230, AMD_THERMAL_IRQ_LAST,
+ iceland_dpm_set_interrupt_state,
+ info->call_back, info->context);
+
+ if (result)
+ return -EINVAL;
+
+ result = cgs_add_irq_source(hwmgr->device, 231, AMD_THERMAL_IRQ_LAST,
+ iceland_dpm_set_interrupt_state,
+ info->call_back, info->context);
+
+ if (result)
+ return -EINVAL;
+
+ return 0;
+}
+
+
+static bool iceland_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ bool is_update_required = false;
+ struct cgs_display_info info = {0,0,NULL};
+
+ cgs_get_active_displays_info(hwmgr->device, &info);
+
+ if (data->display_timing.num_existing_displays != info.display_count)
+ is_update_required = true;
+/* TO DO NEED TO GET DEEP SLEEP CLOCK FROM DAL
+ if (phm_cap_enabled(hwmgr->hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep)) {
+ cgs_get_min_clock_settings(hwmgr->device, &min_clocks);
+ if(min_clocks.engineClockInSR != data->display_timing.minClockInSR)
+ is_update_required = true;
+*/
+ return is_update_required;
+}
+
+
+static inline bool iceland_are_power_levels_equal(const struct iceland_performance_level *pl1,
+ const struct iceland_performance_level *pl2)
+{
+ return ((pl1->memory_clock == pl2->memory_clock) &&
+ (pl1->engine_clock == pl2->engine_clock) &&
+ (pl1->pcie_gen == pl2->pcie_gen) &&
+ (pl1->pcie_lane == pl2->pcie_lane));
+}
+
+int iceland_check_states_equal(struct pp_hwmgr *hwmgr, const struct pp_hw_power_state *pstate1,
+ const struct pp_hw_power_state *pstate2, bool *equal)
+{
+ const struct iceland_power_state *psa = cast_const_phw_iceland_power_state(pstate1);
+ const struct iceland_power_state *psb = cast_const_phw_iceland_power_state(pstate2);
+ int i;
+
+ if (equal == NULL || psa == NULL || psb == NULL)
+ return -EINVAL;
+
+ /* If the two states don't even have the same number of performance levels they cannot be the same state. */
+ if (psa->performance_level_count != psb->performance_level_count) {
+ *equal = false;
+ return 0;
+ }
+
+ for (i = 0; i < psa->performance_level_count; i++) {
+ if (!iceland_are_power_levels_equal(&(psa->performance_levels[i]), &(psb->performance_levels[i]))) {
+ /* If we have found even one performance level pair that is different the states are different. */
+ *equal = false;
+ return 0;
+ }
+ }
+
+ /* If all performance levels are the same try to use the UVD clocks to break the tie.*/
+ *equal = ((psa->uvd_clocks.VCLK == psb->uvd_clocks.VCLK) && (psa->uvd_clocks.DCLK == psb->uvd_clocks.DCLK));
+ *equal &= ((psa->vce_clocks.EVCLK == psb->vce_clocks.EVCLK) && (psa->vce_clocks.ECCLK == psb->vce_clocks.ECCLK));
+ *equal &= (psa->sclk_threshold == psb->sclk_threshold);
+ *equal &= (psa->acp_clk == psb->acp_clk);
+
+ return 0;
+}
+
+static int iceland_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
+{
+ if (mode) {
+ /* stop auto-manage */
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_MicrocodeFanControl))
+ iceland_fan_ctrl_stop_smc_fan_control(hwmgr);
+ iceland_fan_ctrl_set_static_mode(hwmgr, mode);
+ } else
+ /* restart auto-manage */
+ iceland_fan_ctrl_reset_fan_speed_to_default(hwmgr);
+
+ return 0;
+}
+
+static int iceland_get_fan_control_mode(struct pp_hwmgr *hwmgr)
+{
+ if (hwmgr->fan_ctrl_is_in_default_mode)
+ return hwmgr->fan_ctrl_default_mode;
+ else
+ return PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+ CG_FDO_CTRL2, FDO_PWM_MODE);
+}
+
+static int iceland_force_clock_level(struct pp_hwmgr *hwmgr,
+ enum pp_clock_type type, uint32_t mask)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ if (hwmgr->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL)
+ return -EINVAL;
+
+ switch (type) {
+ case PP_SCLK:
+ if (!data->sclk_dpm_key_disabled)
+ smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+ PPSMC_MSG_SCLKDPM_SetEnabledMask,
+ data->dpm_level_enable_mask.sclk_dpm_enable_mask & mask);
+ break;
+ case PP_MCLK:
+ if (!data->mclk_dpm_key_disabled)
+ smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+ PPSMC_MSG_MCLKDPM_SetEnabledMask,
+ data->dpm_level_enable_mask.mclk_dpm_enable_mask & mask);
+ break;
+ case PP_PCIE:
+ {
+ uint32_t tmp = mask & data->dpm_level_enable_mask.pcie_dpm_enable_mask;
+ uint32_t level = 0;
+
+ while (tmp >>= 1)
+ level++;
+
+ if (!data->pcie_dpm_key_disabled)
+ smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+ PPSMC_MSG_PCIeDPM_ForceLevel,
+ level);
+ break;
+ }
+ default:
+ break;
+ }
+
+ return 0;
+}
+
+static int iceland_print_clock_levels(struct pp_hwmgr *hwmgr,
+ enum pp_clock_type type, char *buf)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_single_dpm_table *sclk_table = &(data->dpm_table.sclk_table);
+ struct iceland_single_dpm_table *mclk_table = &(data->dpm_table.mclk_table);
+ struct iceland_single_dpm_table *pcie_table = &(data->dpm_table.pcie_speed_table);
+ int i, now, size = 0;
+ uint32_t clock, pcie_speed;
+
+ switch (type) {
+ case PP_SCLK:
+ smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_API_GetSclkFrequency);
+ clock = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+
+ for (i = 0; i < sclk_table->count; i++) {
+ if (clock > sclk_table->dpm_levels[i].value)
+ continue;
+ break;
+ }
+ now = i;
+
+ for (i = 0; i < sclk_table->count; i++)
+ size += sprintf(buf + size, "%d: %uMhz %s\n",
+ i, sclk_table->dpm_levels[i].value / 100,
+ (i == now) ? "*" : "");
+ break;
+ case PP_MCLK:
+ smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_API_GetMclkFrequency);
+ clock = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+
+ for (i = 0; i < mclk_table->count; i++) {
+ if (clock > mclk_table->dpm_levels[i].value)
+ continue;
+ break;
+ }
+ now = i;
+
+ for (i = 0; i < mclk_table->count; i++)
+ size += sprintf(buf + size, "%d: %uMhz %s\n",
+ i, mclk_table->dpm_levels[i].value / 100,
+ (i == now) ? "*" : "");
+ break;
+ case PP_PCIE:
+ pcie_speed = iceland_get_current_pcie_speed(hwmgr);
+ for (i = 0; i < pcie_table->count; i++) {
+ if (pcie_speed != pcie_table->dpm_levels[i].value)
+ continue;
+ break;
+ }
+ now = i;
+
+ for (i = 0; i < pcie_table->count; i++)
+ size += sprintf(buf + size, "%d: %s %s\n", i,
+ (pcie_table->dpm_levels[i].value == 0) ? "2.5GB, x8" :
+ (pcie_table->dpm_levels[i].value == 1) ? "5.0GB, x16" :
+ (pcie_table->dpm_levels[i].value == 2) ? "8.0GB, x16" : "",
+ (i == now) ? "*" : "");
+ break;
+ default:
+ break;
+ }
+ return size;
+}
+
+static int iceland_get_sclk_od(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_single_dpm_table *sclk_table = &(data->dpm_table.sclk_table);
+ struct iceland_single_dpm_table *golden_sclk_table =
+ &(data->golden_dpm_table.sclk_table);
+ int value;
+
+ value = (sclk_table->dpm_levels[sclk_table->count - 1].value -
+ golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value) *
+ 100 /
+ golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value;
+
+ return value;
+}
+
+static int iceland_set_sclk_od(struct pp_hwmgr *hwmgr, uint32_t value)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_single_dpm_table *golden_sclk_table =
+ &(data->golden_dpm_table.sclk_table);
+ struct pp_power_state *ps;
+ struct iceland_power_state *iceland_ps;
+
+ if (value > 20)
+ value = 20;
+
+ ps = hwmgr->request_ps;
+
+ if (ps == NULL)
+ return -EINVAL;
+
+ iceland_ps = cast_phw_iceland_power_state(&ps->hardware);
+
+ iceland_ps->performance_levels[iceland_ps->performance_level_count - 1].engine_clock =
+ golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value *
+ value / 100 +
+ golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value;
+
+ return 0;
+}
+
+static int iceland_get_mclk_od(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_single_dpm_table *mclk_table = &(data->dpm_table.mclk_table);
+ struct iceland_single_dpm_table *golden_mclk_table =
+ &(data->golden_dpm_table.mclk_table);
+ int value;
+
+ value = (mclk_table->dpm_levels[mclk_table->count - 1].value -
+ golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value) *
+ 100 /
+ golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value;
+
+ return value;
+}
+
+uint32_t iceland_get_xclk(struct pp_hwmgr *hwmgr)
+{
+ uint32_t reference_clock;
+ uint32_t tc;
+ uint32_t divide;
+
+ ATOM_FIRMWARE_INFO *fw_info;
+ uint16_t size;
+ uint8_t frev, crev;
+ int index = GetIndexIntoMasterTable(DATA, FirmwareInfo);
+
+ tc = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_CLKPIN_CNTL_2, MUX_TCLK_TO_XCLK);
+
+ if (tc)
+ return TCLK;
+
+ fw_info = (ATOM_FIRMWARE_INFO *)cgs_atom_get_data_table(hwmgr->device, index,
+ &size, &frev, &crev);
+
+ if (!fw_info)
+ return 0;
+
+ reference_clock = le16_to_cpu(fw_info->usReferenceClock);
+
+ divide = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_CLKPIN_CNTL, XTALIN_DIVIDE);
+
+ if (0 != divide)
+ return reference_clock / 4;
+
+ return reference_clock;
+}
+
+static int iceland_set_mclk_od(struct pp_hwmgr *hwmgr, uint32_t value)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_single_dpm_table *golden_mclk_table =
+ &(data->golden_dpm_table.mclk_table);
+ struct pp_power_state *ps;
+ struct iceland_power_state *iceland_ps;
+
+ if (value > 20)
+ value = 20;
+
+ ps = hwmgr->request_ps;
+
+ if (ps == NULL)
+ return -EINVAL;
+
+ iceland_ps = cast_phw_iceland_power_state(&ps->hardware);
+
+ iceland_ps->performance_levels[iceland_ps->performance_level_count - 1].memory_clock =
+ golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value *
+ value / 100 +
+ golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value;
+
+ return 0;
+}
+
+static const struct pp_hwmgr_func iceland_hwmgr_funcs = {
+ .backend_init = &iceland_hwmgr_backend_init,
+ .backend_fini = &iceland_hwmgr_backend_fini,
+ .asic_setup = &iceland_setup_asic_task,
+ .dynamic_state_management_enable = &iceland_enable_dpm_tasks,
+ .apply_state_adjust_rules = iceland_apply_state_adjust_rules,
+ .force_dpm_level = &iceland_force_dpm_level,
+ .power_state_set = iceland_set_power_state_tasks,
+ .get_power_state_size = iceland_get_power_state_size,
+ .get_mclk = iceland_dpm_get_mclk,
+ .get_sclk = iceland_dpm_get_sclk,
+ .patch_boot_state = iceland_dpm_patch_boot_state,
+ .get_pp_table_entry = iceland_get_pp_table_entry,
+ .get_num_of_pp_table_entries = iceland_get_num_of_entries,
+ .print_current_perforce_level = iceland_print_current_perforce_level,
+ .powerdown_uvd = iceland_phm_powerdown_uvd,
+ .powergate_uvd = iceland_phm_powergate_uvd,
+ .powergate_vce = iceland_phm_powergate_vce,
+ .disable_clock_power_gating = iceland_phm_disable_clock_power_gating,
+ .update_clock_gatings = iceland_phm_update_clock_gatings,
+ .notify_smc_display_config_after_ps_adjustment = iceland_notify_smc_display_config_after_ps_adjustment,
+ .display_config_changed = iceland_display_configuration_changed_task,
+ .set_max_fan_pwm_output = iceland_set_max_fan_pwm_output,
+ .set_max_fan_rpm_output = iceland_set_max_fan_rpm_output,
+ .get_temperature = iceland_thermal_get_temperature,
+ .stop_thermal_controller = iceland_thermal_stop_thermal_controller,
+ .get_fan_speed_info = iceland_fan_ctrl_get_fan_speed_info,
+ .get_fan_speed_percent = iceland_fan_ctrl_get_fan_speed_percent,
+ .set_fan_speed_percent = iceland_fan_ctrl_set_fan_speed_percent,
+ .reset_fan_speed_to_default = iceland_fan_ctrl_reset_fan_speed_to_default,
+ .get_fan_speed_rpm = iceland_fan_ctrl_get_fan_speed_rpm,
+ .set_fan_speed_rpm = iceland_fan_ctrl_set_fan_speed_rpm,
+ .uninitialize_thermal_controller = iceland_thermal_ctrl_uninitialize_thermal_controller,
+ .register_internal_thermal_interrupt = iceland_register_internal_thermal_interrupt,
+ .check_smc_update_required_for_display_configuration = iceland_check_smc_update_required_for_display_configuration,
+ .check_states_equal = iceland_check_states_equal,
+ .set_fan_control_mode = iceland_set_fan_control_mode,
+ .get_fan_control_mode = iceland_get_fan_control_mode,
+ .force_clock_level = iceland_force_clock_level,
+ .print_clock_levels = iceland_print_clock_levels,
+ .get_sclk_od = iceland_get_sclk_od,
+ .set_sclk_od = iceland_set_sclk_od,
+ .get_mclk_od = iceland_get_mclk_od,
+ .set_mclk_od = iceland_set_mclk_od,
+};
+
+int iceland_hwmgr_init(struct pp_hwmgr *hwmgr)
+{
+ iceland_hwmgr *data;
+
+ data = kzalloc (sizeof(iceland_hwmgr), GFP_KERNEL);
+ if (data == NULL)
+ return -ENOMEM;
+ memset(data, 0x00, sizeof(iceland_hwmgr));
+
+ hwmgr->backend = data;
+ hwmgr->hwmgr_func = &iceland_hwmgr_funcs;
+ hwmgr->pptable_func = &pptable_funcs;
+
+ /* thermal */
+ pp_iceland_thermal_initialize(hwmgr);
+ return 0;
+}
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+#ifndef ICELAND_HWMGR_H
+#define ICELAND_HWMGR_H
+
+#include "hwmgr.h"
+#include "ppatomctrl.h"
+#include "ppinterrupt.h"
+#include "ppsmc.h"
+#include "iceland_powertune.h"
+#include "pp_endian.h"
+#include "smu71_discrete.h"
+
+#define ICELAND_MAX_HARDWARE_POWERLEVELS 2
+#define ICELAND_DYNCLK_NUMBER_OF_TREND_COEFFICIENTS 15
+
+struct iceland_performance_level {
+ uint32_t memory_clock;
+ uint32_t engine_clock;
+ uint16_t pcie_gen;
+ uint16_t pcie_lane;
+};
+
+struct _phw_iceland_bacos {
+ uint32_t best_match;
+ uint32_t baco_flags;
+ struct iceland_performance_level performance_level;
+};
+typedef struct _phw_iceland_bacos phw_iceland_bacos;
+
+struct _phw_iceland_uvd_clocks {
+ uint32_t VCLK;
+ uint32_t DCLK;
+};
+
+typedef struct _phw_iceland_uvd_clocks phw_iceland_uvd_clocks;
+
+struct _phw_iceland_vce_clocks {
+ uint32_t EVCLK;
+ uint32_t ECCLK;
+};
+
+typedef struct _phw_iceland_vce_clocks phw_iceland_vce_clocks;
+
+struct iceland_power_state {
+ uint32_t magic;
+ phw_iceland_uvd_clocks uvd_clocks;
+ phw_iceland_vce_clocks vce_clocks;
+ uint32_t sam_clk;
+ uint32_t acp_clk;
+ uint16_t performance_level_count;
+ bool dc_compatible;
+ uint32_t sclk_threshold;
+ struct iceland_performance_level performance_levels[ICELAND_MAX_HARDWARE_POWERLEVELS];
+};
+
+struct _phw_iceland_dpm_level {
+ bool enabled;
+ uint32_t value;
+ uint32_t param1;
+};
+typedef struct _phw_iceland_dpm_level phw_iceland_dpm_level;
+
+#define ICELAND_MAX_DEEPSLEEP_DIVIDER_ID 5
+#define MAX_REGULAR_DPM_NUMBER 8
+#define ICELAND_MINIMUM_ENGINE_CLOCK 5000
+
+struct iceland_single_dpm_table {
+ uint32_t count;
+ phw_iceland_dpm_level dpm_levels[MAX_REGULAR_DPM_NUMBER];
+};
+
+struct iceland_dpm_table {
+ struct iceland_single_dpm_table sclk_table;
+ struct iceland_single_dpm_table mclk_table;
+ struct iceland_single_dpm_table pcie_speed_table;
+ struct iceland_single_dpm_table vddc_table;
+ struct iceland_single_dpm_table vdd_gfx_table;
+ struct iceland_single_dpm_table vdd_ci_table;
+ struct iceland_single_dpm_table mvdd_table;
+};
+typedef struct _phw_iceland_dpm_table phw_iceland_dpm_table;
+
+
+struct _phw_iceland_clock_regisiters {
+ uint32_t vCG_SPLL_FUNC_CNTL;
+ uint32_t vCG_SPLL_FUNC_CNTL_2;
+ uint32_t vCG_SPLL_FUNC_CNTL_3;
+ uint32_t vCG_SPLL_FUNC_CNTL_4;
+ uint32_t vCG_SPLL_SPREAD_SPECTRUM;
+ uint32_t vCG_SPLL_SPREAD_SPECTRUM_2;
+ uint32_t vDLL_CNTL;
+ uint32_t vMCLK_PWRMGT_CNTL;
+ uint32_t vMPLL_AD_FUNC_CNTL;
+ uint32_t vMPLL_DQ_FUNC_CNTL;
+ uint32_t vMPLL_FUNC_CNTL;
+ uint32_t vMPLL_FUNC_CNTL_1;
+ uint32_t vMPLL_FUNC_CNTL_2;
+ uint32_t vMPLL_SS1;
+ uint32_t vMPLL_SS2;
+};
+typedef struct _phw_iceland_clock_regisiters phw_iceland_clock_registers;
+
+struct _phw_iceland_voltage_smio_registers {
+ uint32_t vs0_vid_lower_smio_cntl;
+};
+typedef struct _phw_iceland_voltage_smio_registers phw_iceland_voltage_smio_registers;
+
+
+struct _phw_iceland_mc_reg_entry {
+ uint32_t mclk_max;
+ uint32_t mc_data[SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE];
+};
+typedef struct _phw_iceland_mc_reg_entry phw_iceland_mc_reg_entry;
+
+struct _phw_iceland_mc_reg_table {
+ uint8_t last; /* number of registers*/
+ uint8_t num_entries; /* number of entries in mc_reg_table_entry used*/
+ uint16_t validflag; /* indicate the corresponding register is valid or not. 1: valid, 0: invalid. bit0->address[0], bit1->address[1], etc.*/
+ phw_iceland_mc_reg_entry mc_reg_table_entry[MAX_AC_TIMING_ENTRIES];
+ SMU71_Discrete_MCRegisterAddress mc_reg_address[SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE];
+};
+typedef struct _phw_iceland_mc_reg_table phw_iceland_mc_reg_table;
+
+#define DISABLE_MC_LOADMICROCODE 1
+#define DISABLE_MC_CFGPROGRAMMING 2
+
+
+/*Ultra Low Voltage parameter structure */
+struct phw_iceland_ulv_parm{
+ bool ulv_supported;
+ uint32_t ch_ulv_parameter;
+ uint32_t ulv_volt_change_delay;
+ struct iceland_performance_level ulv_power_level;
+};
+
+#define ICELAND_MAX_LEAKAGE_COUNT 8
+
+struct phw_iceland_leakage_voltage {
+ uint16_t count;
+ uint16_t leakage_id[ICELAND_MAX_LEAKAGE_COUNT];
+ uint16_t actual_voltage[ICELAND_MAX_LEAKAGE_COUNT];
+};
+
+struct _phw_iceland_display_timing {
+ uint32_t min_clock_insr;
+ uint32_t num_existing_displays;
+};
+typedef struct _phw_iceland_display_timing phw_iceland_display_timing;
+
+
+struct phw_iceland_thermal_temperature_setting
+{
+ long temperature_low;
+ long temperature_high;
+ long temperature_shutdown;
+};
+
+struct _phw_iceland_dpmlevel_enable_mask {
+ uint32_t uvd_dpm_enable_mask;
+ uint32_t vce_dpm_enable_mask;
+ uint32_t acp_dpm_enable_mask;
+ uint32_t samu_dpm_enable_mask;
+ uint32_t sclk_dpm_enable_mask;
+ uint32_t mclk_dpm_enable_mask;
+ uint32_t pcie_dpm_enable_mask;
+};
+typedef struct _phw_iceland_dpmlevel_enable_mask phw_iceland_dpmlevel_enable_mask;
+
+struct _phw_iceland_pcie_perf_range {
+ uint16_t max;
+ uint16_t min;
+};
+typedef struct _phw_iceland_pcie_perf_range phw_iceland_pcie_perf_range;
+
+struct _phw_iceland_vbios_boot_state {
+ uint16_t mvdd_bootup_value;
+ uint16_t vddc_bootup_value;
+ uint16_t vddci_bootup_value;
+ uint16_t vddgfx_bootup_value;
+ uint32_t sclk_bootup_value;
+ uint32_t mclk_bootup_value;
+ uint16_t pcie_gen_bootup_value;
+ uint16_t pcie_lane_bootup_value;
+};
+typedef struct _phw_iceland_vbios_boot_state phw_iceland_vbios_boot_state;
+
+#define DPMTABLE_OD_UPDATE_SCLK 0x00000001
+#define DPMTABLE_OD_UPDATE_MCLK 0x00000002
+#define DPMTABLE_UPDATE_SCLK 0x00000004
+#define DPMTABLE_UPDATE_MCLK 0x00000008
+
+/* We need to review which fields are needed. */
+/* This is mostly a copy of the RV7xx/Evergreen structure which is close, but not identical to the N.Islands one. */
+struct iceland_hwmgr {
+ struct iceland_dpm_table dpm_table;
+ struct iceland_dpm_table golden_dpm_table;
+
+ uint32_t voting_rights_clients0;
+ uint32_t voting_rights_clients1;
+ uint32_t voting_rights_clients2;
+ uint32_t voting_rights_clients3;
+ uint32_t voting_rights_clients4;
+ uint32_t voting_rights_clients5;
+ uint32_t voting_rights_clients6;
+ uint32_t voting_rights_clients7;
+ uint32_t static_screen_threshold_unit;
+ uint32_t static_screen_threshold;
+ uint32_t voltage_control;
+ uint32_t vdd_gfx_control;
+
+ uint32_t vddc_vddci_delta;
+ uint32_t vddc_vddgfx_delta;
+
+ struct pp_interrupt_registration_info internal_high_thermal_interrupt_info;
+ struct pp_interrupt_registration_info internal_low_thermal_interrupt_info;
+ struct pp_interrupt_registration_info smc_to_host_interrupt_info;
+ uint32_t active_auto_throttle_sources;
+
+ struct pp_interrupt_registration_info external_throttle_interrupt;
+ irq_handler_func_t external_throttle_callback;
+ void *external_throttle_context;
+
+ struct pp_interrupt_registration_info ctf_interrupt_info;
+ irq_handler_func_t ctf_callback;
+ void *ctf_context;
+
+ phw_iceland_clock_registers clock_registers;
+ phw_iceland_voltage_smio_registers voltage_smio_registers;
+
+ bool is_memory_GDDR5;
+ uint16_t acpi_vddc;
+ bool pspp_notify_required; /* Flag to indicate if PSPP notification to SBIOS is required */
+ uint16_t force_pcie_gen; /* The forced PCI-E speed if not 0xffff */
+ uint16_t acpi_pcie_gen; /* The PCI-E speed at ACPI time */
+ uint32_t pcie_gen_cap; /* The PCI-E speed capabilities bitmap from CAIL */
+ uint32_t pcie_lane_cap; /* The PCI-E lane capabilities bitmap from CAIL */
+ uint32_t pcie_spc_cap; /* Symbol Per Clock Capabilities from registry */
+ struct phw_iceland_leakage_voltage vddc_leakage; /* The Leakage VDDC supported (based on leakage ID).*/
+ struct phw_iceland_leakage_voltage vddcgfx_leakage; /* The Leakage VDDC supported (based on leakage ID). */
+ struct phw_iceland_leakage_voltage vddci_leakage; /* The Leakage VDDCI supported (based on leakage ID). */
+
+ uint32_t mvdd_control;
+ uint32_t vddc_mask_low;
+ uint32_t mvdd_mask_low;
+ uint16_t max_vddc_in_pp_table; /* the maximum VDDC value in the powerplay table*/
+ uint16_t min_vddc_in_pp_table;
+ uint16_t max_vddci_in_pp_table; /* the maximum VDDCI value in the powerplay table */
+ uint16_t min_vddci_in_pp_table;
+ uint32_t mclk_strobe_mode_threshold;
+ uint32_t mclk_stutter_mode_threshold;
+ uint32_t mclk_edc_enable_threshold;
+ uint32_t mclk_edc_wr_enable_threshold;
+ bool is_uvd_enabled;
+ bool is_xdma_enabled;
+ phw_iceland_vbios_boot_state vbios_boot_state;
+
+ bool battery_state;
+ bool is_tlu_enabled;
+ bool pcie_performance_request;
+
+ /* -------------- SMC SRAM Address of firmware header tables ----------------*/
+ uint32_t sram_end; /* The first address after the SMC SRAM. */
+ uint32_t dpm_table_start; /* The start of the dpm table in the SMC SRAM. */
+ uint32_t soft_regs_start; /* The start of the soft registers in the SMC SRAM. */
+ uint32_t mc_reg_table_start; /* The start of the mc register table in the SMC SRAM. */
+ uint32_t fan_table_start; /* The start of the fan table in the SMC SRAM. */
+ uint32_t arb_table_start; /* The start of the ARB setting table in the SMC SRAM. */
+ uint32_t ulv_settings_start;
+ SMU71_Discrete_DpmTable smc_state_table; /* The carbon copy of the SMC state table. */
+ SMU71_Discrete_MCRegisters mc_reg_table;
+ SMU71_Discrete_Ulv ulv_setting; /* The carbon copy of ULV setting. */
+
+ /* -------------- Stuff originally coming from Evergreen --------------------*/
+ phw_iceland_mc_reg_table iceland_mc_reg_table;
+ uint32_t vdd_ci_control;
+ pp_atomctrl_voltage_table vddc_voltage_table;
+ pp_atomctrl_voltage_table vddci_voltage_table;
+ pp_atomctrl_voltage_table vddgfx_voltage_table;
+ pp_atomctrl_voltage_table mvdd_voltage_table;
+
+ uint32_t mgcg_cgtt_local2;
+ uint32_t mgcg_cgtt_local3;
+ uint32_t gpio_debug;
+ uint32_t mc_micro_code_feature;
+ uint32_t highest_mclk;
+ uint16_t acpi_vdd_ci;
+ uint8_t mvdd_high_index;
+ uint8_t mvdd_low_index;
+ bool dll_defaule_on;
+ bool performance_request_registered;
+
+ /* ----------------- Low Power Features ---------------------*/
+ phw_iceland_bacos bacos;
+ struct phw_iceland_ulv_parm ulv;
+
+ /* ----------------- CAC Stuff ---------------------*/
+ uint32_t cac_table_start;
+ bool cac_configuration_required; /* TRUE if PP_CACConfigurationRequired == 1 */
+ bool driver_calculate_cac_leakage; /* TRUE if PP_DriverCalculateCACLeakage == 1 */
+ bool cac_enabled;
+
+ /* ----------------- DPM2 Parameters ---------------------*/
+ uint32_t power_containment_features;
+ bool enable_bapm_feature;
+ bool enable_dte_feature;
+ bool enable_tdc_limit_feature;
+ bool enable_pkg_pwr_tracking_feature;
+ bool disable_uvd_power_tune_feature;
+ struct iceland_pt_defaults *power_tune_defaults;
+ SMU71_Discrete_PmFuses power_tune_table;
+ uint32_t ul_dte_tj_offset; /* Fudge factor in DPM table to correct HW DTE errors */
+ uint32_t fast_watermark_threshold; /* use fast watermark if clock is equal or above this. In percentage of the target high sclk. */
+
+ /* ----------------- Phase Shedding ---------------------*/
+ bool vddc_phase_shed_control;
+
+ /* --------------------- DI/DT --------------------------*/
+ phw_iceland_display_timing display_timing;
+
+ /* --------- ReadRegistry data for memory and engine clock margins ---- */
+ uint32_t engine_clock_data;
+ uint32_t memory_clock_data;
+
+ /* -------- Thermal Temperature Setting --------------*/
+ struct phw_iceland_thermal_temperature_setting thermal_temp_setting;
+ phw_iceland_dpmlevel_enable_mask dpm_level_enable_mask;
+
+ uint32_t need_update_smu7_dpm_table;
+ uint32_t sclk_dpm_key_disabled;
+ uint32_t mclk_dpm_key_disabled;
+ uint32_t pcie_dpm_key_disabled;
+ /* used to store the previous dal min sclock */
+ uint32_t min_engine_clocks;
+ phw_iceland_pcie_perf_range pcie_gen_performance;
+ phw_iceland_pcie_perf_range pcie_lane_performance;
+ phw_iceland_pcie_perf_range pcie_gen_power_saving;
+ phw_iceland_pcie_perf_range pcie_lane_power_saving;
+ bool use_pcie_performance_levels;
+ bool use_pcie_power_saving_levels;
+ /* percentage value from 0-100, default 50 */
+ uint32_t activity_target[SMU71_MAX_LEVELS_GRAPHICS];
+ uint32_t mclk_activity_target;
+ uint32_t low_sclk_interrupt_threshold;
+ uint32_t last_mclk_dpm_enable_mask;
+ bool uvd_enabled;
+ uint32_t pcc_monitor_enabled;
+
+ /* --------- Power Gating States ------------*/
+ bool uvd_power_gated; /* 1: gated, 0:not gated */
+ bool vce_power_gated; /* 1: gated, 0:not gated */
+ bool samu_power_gated; /* 1: gated, 0:not gated */
+ bool acp_power_gated; /* 1: gated, 0:not gated */
+ bool pg_acp_init;
+
+ /* soft pptable for re-uploading into smu */
+ void *soft_pp_table;
+};
+
+typedef struct iceland_hwmgr iceland_hwmgr;
+
+int iceland_hwmgr_init(struct pp_hwmgr *hwmgr);
+int iceland_update_uvd_dpm(struct pp_hwmgr *hwmgr, bool bgate);
+uint32_t iceland_get_xclk(struct pp_hwmgr *hwmgr);
+int iceland_populate_bapm_vddc_vid_sidd(struct pp_hwmgr *hwmgr);
+int iceland_populate_vddc_vid(struct pp_hwmgr *hwmgr);
+
+#define ICELAND_DPM2_NEAR_TDP_DEC 10
+#define ICELAND_DPM2_ABOVE_SAFE_INC 5
+#define ICELAND_DPM2_BELOW_SAFE_INC 20
+
+/*
+ * Log2 of the LTA window size (l2numWin_TDP). Eg. If LTA windows size
+ * is 128, then this value should be Log2(128) = 7.
+ */
+#define ICELAND_DPM2_LTA_WINDOW_SIZE 7
+
+#define ICELAND_DPM2_LTS_TRUNCATE 0
+
+#define ICELAND_DPM2_TDP_SAFE_LIMIT_PERCENT 80 // Maximum 100
+
+#define ICELAND_DPM2_MAXPS_PERCENT_H 90 // Maximum 0xFF
+#define ICELAND_DPM2_MAXPS_PERCENT_M 90 // Maximum 0xFF
+
+#define ICELAND_DPM2_PWREFFICIENCYRATIO_MARGIN 50
+
+#define ICELAND_DPM2_SQ_RAMP_MAX_POWER 0x3FFF
+#define ICELAND_DPM2_SQ_RAMP_MIN_POWER 0x12
+#define ICELAND_DPM2_SQ_RAMP_MAX_POWER_DELTA 0x15
+#define ICELAND_DPM2_SQ_RAMP_SHORT_TERM_INTERVAL_SIZE 0x1E
+#define ICELAND_DPM2_SQ_RAMP_LONG_TERM_INTERVAL_RATIO 0xF
+
+#define ICELAND_VOLTAGE_CONTROL_NONE 0x0
+#define ICELAND_VOLTAGE_CONTROL_BY_GPIO 0x1
+#define ICELAND_VOLTAGE_CONTROL_BY_SVID2 0x2
+
+/* convert to Q8.8 format for firmware */
+#define ICELAND_Q88_FORMAT_CONVERSION_UNIT 256
+
+#define ICELAND_UNUSED_GPIO_PIN 0x7F
+
+#endif
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+
+#include "amdgpu.h"
+#include "hwmgr.h"
+#include "smumgr.h"
+#include "iceland_hwmgr.h"
+#include "iceland_powertune.h"
+#include "iceland_smumgr.h"
+#include "smu71_discrete.h"
+#include "smu71.h"
+#include "pp_debug.h"
+#include "cgs_common.h"
+#include "pp_endian.h"
+
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+
+#define VOLTAGE_SCALE 4
+#define POWERTUNE_DEFAULT_SET_MAX 1
+
+#define DEVICE_ID_VI_ICELAND_M_6900 0x6900
+#define DEVICE_ID_VI_ICELAND_M_6901 0x6901
+#define DEVICE_ID_VI_ICELAND_M_6902 0x6902
+#define DEVICE_ID_VI_ICELAND_M_6903 0x6903
+
+
+struct iceland_pt_defaults defaults_iceland =
+{
+ /*
+ * sviLoadLIneEn, SviLoadLineVddC, TDC_VDDC_ThrottleReleaseLimitPerc,
+ * TDC_MAWt, TdcWaterfallCtl, DTEAmbientTempBase, DisplayCac, BAPM_TEMP_GRADIENT
+ */
+ 1, 0xF, 0xFD, 0x19, 5, 45, 0, 0xB0000,
+ { 0x79, 0x253, 0x25D, 0xAE, 0x72, 0x80, 0x83, 0x86, 0x6F, 0xC8, 0xC9, 0xC9, 0x2F, 0x4D, 0x61 },
+ { 0x17C, 0x172, 0x180, 0x1BC, 0x1B3, 0x1BD, 0x206, 0x200, 0x203, 0x25D, 0x25A, 0x255, 0x2C3, 0x2C5, 0x2B4 }
+};
+
+/* 35W - XT, XTL */
+struct iceland_pt_defaults defaults_icelandxt =
+{
+ /*
+ * sviLoadLIneEn, SviLoadLineVddC,
+ * TDC_VDDC_ThrottleReleaseLimitPerc, TDC_MAWt,
+ * TdcWaterfallCtl, DTEAmbientTempBase, DisplayCac,
+ * BAPM_TEMP_GRADIENT
+ */
+ 1, 0xF, 0xFD, 0x19, 5, 45, 0, 0x0,
+ { 0xA7, 0x0, 0x0, 0xB5, 0x0, 0x0, 0x9F, 0x0, 0x0, 0xD6, 0x0, 0x0, 0xD7, 0x0, 0x0},
+ { 0x1EA, 0x0, 0x0, 0x224, 0x0, 0x0, 0x25E, 0x0, 0x0, 0x28E, 0x0, 0x0, 0x2AB, 0x0, 0x0}
+};
+
+/* 25W - PRO, LE */
+struct iceland_pt_defaults defaults_icelandpro =
+{
+ /*
+ * sviLoadLIneEn, SviLoadLineVddC,
+ * TDC_VDDC_ThrottleReleaseLimitPerc, TDC_MAWt,
+ * TdcWaterfallCtl, DTEAmbientTempBase, DisplayCac,
+ * BAPM_TEMP_GRADIENT
+ */
+ 1, 0xF, 0xFD, 0x19, 5, 45, 0, 0x0,
+ { 0xB7, 0x0, 0x0, 0xC3, 0x0, 0x0, 0xB5, 0x0, 0x0, 0xEA, 0x0, 0x0, 0xE6, 0x0, 0x0},
+ { 0x1EA, 0x0, 0x0, 0x224, 0x0, 0x0, 0x25E, 0x0, 0x0, 0x28E, 0x0, 0x0, 0x2AB, 0x0, 0x0}
+};
+
+void iceland_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint32_t tmp = 0;
+ struct cgs_system_info sys_info = {0};
+ uint32_t pdev_id;
+
+ sys_info.size = sizeof(struct cgs_system_info);
+ sys_info.info_id = CGS_SYSTEM_INFO_PCIE_DEV;
+ cgs_query_system_info(hwmgr->device, &sys_info);
+ pdev_id = (uint32_t)sys_info.value;
+
+ switch (pdev_id) {
+ case DEVICE_ID_VI_ICELAND_M_6900:
+ case DEVICE_ID_VI_ICELAND_M_6903:
+ data->power_tune_defaults = &defaults_icelandxt;
+ break;
+
+ case DEVICE_ID_VI_ICELAND_M_6901:
+ case DEVICE_ID_VI_ICELAND_M_6902:
+ data->power_tune_defaults = &defaults_icelandpro;
+ break;
+ default:
+ /* TODO: need to assign valid defaults */
+ data->power_tune_defaults = &defaults_iceland;
+ pr_warning("Unknown V.I. Device ID.\n");
+ break;
+ }
+
+ /* Assume disabled */
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_CAC);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SQRamping);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DBRamping);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_TDRamping);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_TCPRamping);
+
+ data->ul_dte_tj_offset = tmp;
+
+ if (!tmp) {
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_CAC);
+
+ data->fast_watermark_threshold = 100;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
+ tmp = 1;
+ data->enable_dte_feature = tmp ? false : true;
+ data->enable_tdc_limit_feature = tmp ? true : false;
+ data->enable_pkg_pwr_tracking_feature = tmp ? true : false;
+ }
+ }
+}
+
+int iceland_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ struct iceland_pt_defaults *defaults = data->power_tune_defaults;
+ SMU71_Discrete_DpmTable *dpm_table = &(data->smc_state_table);
+ struct phm_cac_tdp_table *cac_dtp_table = hwmgr->dyn_state.cac_dtp_table;
+ struct phm_ppm_table *ppm = hwmgr->dyn_state.ppm_parameter_table;
+ uint16_t *def1, *def2;
+ int i, j, k;
+
+ /*
+ * TDP number of fraction bits are changed from 8 to 7 for Iceland
+ * as requested by SMC team
+ */
+ dpm_table->DefaultTdp = PP_HOST_TO_SMC_US((uint16_t)(cac_dtp_table->usTDP * 256));
+ dpm_table->TargetTdp = PP_HOST_TO_SMC_US((uint16_t)(cac_dtp_table->usConfigurableTDP * 256));
+
+ dpm_table->DTETjOffset = (uint8_t)data->ul_dte_tj_offset;
+
+ dpm_table->GpuTjMax = (uint8_t)(data->thermal_temp_setting.temperature_high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES);
+ dpm_table->GpuTjHyst = 8;
+
+ dpm_table->DTEAmbientTempBase = defaults->dte_ambient_temp_base;
+
+ /* The following are for new Iceland Multi-input fan/thermal control */
+ if(NULL != ppm) {
+ dpm_table->PPM_PkgPwrLimit = (uint16_t)ppm->dgpu_tdp * 256 / 1000;
+ dpm_table->PPM_TemperatureLimit = (uint16_t)ppm->tj_max * 256;
+ } else {
+ dpm_table->PPM_PkgPwrLimit = 0;
+ dpm_table->PPM_TemperatureLimit = 0;
+ }
+
+ CONVERT_FROM_HOST_TO_SMC_US(dpm_table->PPM_PkgPwrLimit);
+ CONVERT_FROM_HOST_TO_SMC_US(dpm_table->PPM_TemperatureLimit);
+
+ dpm_table->BAPM_TEMP_GRADIENT = PP_HOST_TO_SMC_UL(defaults->bamp_temp_gradient);
+ def1 = defaults->bapmti_r;
+ def2 = defaults->bapmti_rc;
+
+ for (i = 0; i < SMU71_DTE_ITERATIONS; i++) {
+ for (j = 0; j < SMU71_DTE_SOURCES; j++) {
+ for (k = 0; k < SMU71_DTE_SINKS; k++) {
+ dpm_table->BAPMTI_R[i][j][k] = PP_HOST_TO_SMC_US(*def1);
+ dpm_table->BAPMTI_RC[i][j][k] = PP_HOST_TO_SMC_US(*def2);
+ def1++;
+ def2++;
+ }
+ }
+ }
+
+ return 0;
+}
+
+static int iceland_populate_svi_load_line(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ const struct iceland_pt_defaults *defaults = data->power_tune_defaults;
+
+ data->power_tune_table.SviLoadLineEn = defaults->svi_load_line_en;
+ data->power_tune_table.SviLoadLineVddC = defaults->svi_load_line_vddc;
+ data->power_tune_table.SviLoadLineTrimVddC = 3;
+ data->power_tune_table.SviLoadLineOffsetVddC = 0;
+
+ return 0;
+}
+
+static int iceland_populate_tdc_limit(struct pp_hwmgr *hwmgr)
+{
+ uint16_t tdc_limit;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ const struct iceland_pt_defaults *defaults = data->power_tune_defaults;
+
+ /* TDC number of fraction bits are changed from 8 to 7
+ * for Iceland as requested by SMC team
+ */
+ tdc_limit = (uint16_t)(hwmgr->dyn_state.cac_dtp_table->usTDC * 256);
+ data->power_tune_table.TDC_VDDC_PkgLimit =
+ CONVERT_FROM_HOST_TO_SMC_US(tdc_limit);
+ data->power_tune_table.TDC_VDDC_ThrottleReleaseLimitPerc =
+ defaults->tdc_vddc_throttle_release_limit_perc;
+ data->power_tune_table.TDC_MAWt = defaults->tdc_mawt;
+
+ return 0;
+}
+
+static int iceland_populate_dw8(struct pp_hwmgr *hwmgr, uint32_t fuse_table_offset)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ const struct iceland_pt_defaults *defaults = data->power_tune_defaults;
+ uint32_t temp;
+
+ if (iceland_read_smc_sram_dword(hwmgr->smumgr,
+ fuse_table_offset +
+ offsetof(SMU71_Discrete_PmFuses, TdcWaterfallCtl),
+ (uint32_t *)&temp, data->sram_end))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to read PmFuses.DW6 (SviLoadLineEn) from SMC Failed!",
+ return -EINVAL);
+ else
+ data->power_tune_table.TdcWaterfallCtl = defaults->tdc_waterfall_ctl;
+
+ return 0;
+}
+
+static int iceland_populate_temperature_scaler(struct pp_hwmgr *hwmgr)
+{
+ return 0;
+}
+
+static int iceland_populate_gnb_lpml(struct pp_hwmgr *hwmgr)
+{
+ int i;
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ /* Currently not used. Set all to zero. */
+ for (i = 0; i < 8; i++)
+ data->power_tune_table.GnbLPML[i] = 0;
+
+ return 0;
+}
+
+static int iceland_min_max_vgnb_lpml_id_from_bapm_vddc(struct pp_hwmgr *hwmgr)
+{
+ return 0;
+}
+
+static int iceland_populate_bapm_vddc_base_leakage_sidd(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint16_t HiSidd = data->power_tune_table.BapmVddCBaseLeakageHiSidd;
+ uint16_t LoSidd = data->power_tune_table.BapmVddCBaseLeakageLoSidd;
+ struct phm_cac_tdp_table *cac_table = hwmgr->dyn_state.cac_dtp_table;
+
+ HiSidd = (uint16_t)(cac_table->usHighCACLeakage / 100 * 256);
+ LoSidd = (uint16_t)(cac_table->usLowCACLeakage / 100 * 256);
+
+ data->power_tune_table.BapmVddCBaseLeakageHiSidd =
+ CONVERT_FROM_HOST_TO_SMC_US(HiSidd);
+ data->power_tune_table.BapmVddCBaseLeakageLoSidd =
+ CONVERT_FROM_HOST_TO_SMC_US(LoSidd);
+
+ return 0;
+}
+
+int iceland_populate_pm_fuses(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ uint32_t pm_fuse_table_offset;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
+ if (iceland_read_smc_sram_dword(hwmgr->smumgr,
+ SMU71_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU71_Firmware_Header, PmFuseTable),
+ &pm_fuse_table_offset, data->sram_end))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to get pm_fuse_table_offset Failed!",
+ return -EINVAL);
+
+ /* DW0 - DW3 */
+ if (iceland_populate_bapm_vddc_vid_sidd(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate bapm vddc vid Failed!",
+ return -EINVAL);
+
+ /* DW4 - DW5 */
+ if (iceland_populate_vddc_vid(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate vddc vid Failed!",
+ return -EINVAL);
+
+ /* DW6 */
+ if (iceland_populate_svi_load_line(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate SviLoadLine Failed!",
+ return -EINVAL);
+ /* DW7 */
+ if (iceland_populate_tdc_limit(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate TDCLimit Failed!", return -EINVAL);
+ /* DW8 */
+ if (iceland_populate_dw8(hwmgr, pm_fuse_table_offset))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate TdcWaterfallCtl, "
+ "LPMLTemperature Min and Max Failed!",
+ return -EINVAL);
+
+ /* DW9-DW12 */
+ if (0 != iceland_populate_temperature_scaler(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate LPMLTemperatureScaler Failed!",
+ return -EINVAL);
+
+ /* DW13-DW16 */
+ if (iceland_populate_gnb_lpml(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate GnbLPML Failed!",
+ return -EINVAL);
+
+ /* DW17 */
+ if (iceland_min_max_vgnb_lpml_id_from_bapm_vddc(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate GnbLPML Min and Max Vid Failed!",
+ return -EINVAL);
+
+ /* DW18 */
+ if (iceland_populate_bapm_vddc_base_leakage_sidd(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate BapmVddCBaseLeakage Hi and Lo Sidd Failed!",
+ return -EINVAL);
+
+ if (iceland_copy_bytes_to_smc(hwmgr->smumgr, pm_fuse_table_offset,
+ (uint8_t *)&data->power_tune_table,
+ sizeof(struct SMU71_Discrete_PmFuses), data->sram_end))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to download PmFuseTable Failed!",
+ return -EINVAL);
+ }
+ return 0;
+}
+
+int iceland_enable_smc_cac(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ int result = 0;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_CAC)) {
+ int smc_result;
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_EnableCac));
+ PP_ASSERT_WITH_CODE((0 == smc_result),
+ "Failed to enable CAC in SMC.", result = -1);
+
+ data->cac_enabled = (0 == smc_result) ? true : false;
+ }
+ return result;
+}
+
+static int iceland_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+
+ if(data->power_containment_features &
+ POWERCONTAINMENT_FEATURE_PkgPwrLimit)
+ return smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+ PPSMC_MSG_PkgPwrSetLimit, n);
+ return 0;
+}
+
+static int iceland_set_overdriver_target_tdp(struct pp_hwmgr *pHwMgr, uint32_t target_tdp)
+{
+ return smum_send_msg_to_smc_with_parameter(pHwMgr->smumgr,
+ PPSMC_MSG_OverDriveSetTargetTdp, target_tdp);
+}
+
+int iceland_enable_power_containment(struct pp_hwmgr *hwmgr)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ SMU71_Discrete_DpmTable *dpm_table = &data->smc_state_table;
+ int smc_result;
+ int result = 0;
+ uint32_t is_asic_kicker;
+
+ data->power_containment_features = 0;
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
+ is_asic_kicker = cgs_read_register(hwmgr->device, mmCC_BIF_BX_STRAP2);
+ is_asic_kicker = (is_asic_kicker >> 12) & 0x01;
+
+ if (data->enable_bapm_feature &&
+ (!is_asic_kicker ||
+ phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DisableUsingActualTemperatureForPowerCalc))) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_EnableDTE));
+ PP_ASSERT_WITH_CODE((0 == smc_result),
+ "Failed to enable BAPM in SMC.", result = -1;);
+ if (0 == smc_result)
+ data->power_containment_features |= POWERCONTAINMENT_FEATURE_BAPM;
+ }
+
+ if (is_asic_kicker && !phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DisableUsingActualTemperatureForPowerCalc))
+ dpm_table->DTEMode = 2;
+
+ if (data->enable_tdc_limit_feature) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_TDCLimitEnable));
+ PP_ASSERT_WITH_CODE((0 == smc_result),
+ "Failed to enable TDCLimit in SMC.", result = -1;);
+ if (0 == smc_result)
+ data->power_containment_features |=
+ POWERCONTAINMENT_FEATURE_TDCLimit;
+ }
+
+ if (data->enable_pkg_pwr_tracking_feature) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_PkgPwrLimitEnable));
+ PP_ASSERT_WITH_CODE((0 == smc_result),
+ "Failed to enable PkgPwrTracking in SMC.", result = -1;);
+ if (0 == smc_result) {
+ struct phm_cac_tdp_table *cac_table =
+ hwmgr->dyn_state.cac_dtp_table;
+ uint32_t default_limit =
+ (uint32_t)(cac_table->usMaximumPowerDeliveryLimit * 256);
+
+ data->power_containment_features |=
+ POWERCONTAINMENT_FEATURE_PkgPwrLimit;
+
+ if (iceland_set_power_limit(hwmgr, default_limit))
+ printk(KERN_ERR "Failed to set Default Power Limit in SMC!");
+ }
+ }
+ }
+ return result;
+}
+
+int iceland_power_control_set_level(struct pp_hwmgr *hwmgr)
+{
+ struct phm_cac_tdp_table *cac_table = hwmgr->dyn_state.cac_dtp_table;
+ int adjust_percent, target_tdp;
+ int result = 0;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
+ /* adjustment percentage has already been validated */
+ adjust_percent = hwmgr->platform_descriptor.TDPAdjustmentPolarity ?
+ hwmgr->platform_descriptor.TDPAdjustment :
+ (-1 * hwmgr->platform_descriptor.TDPAdjustment);
+ /*
+ * SMC requested that target_tdp to be 7 bit fraction in DPM table
+ * but message to be 8 bit fraction for messages
+ */
+ target_tdp = ((100 + adjust_percent) * (int)(cac_table->usTDP * 256)) / 100;
+ result = iceland_set_overdriver_target_tdp(hwmgr, (uint32_t)target_tdp);
+ }
+
+ return result;
+}
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+#ifndef ICELAND_POWERTUNE_H
+#define ICELAND_POWERTUNE_H
+
+#include "smu71.h"
+
+enum iceland_pt_config_reg_type {
+ ICELAND_CONFIGREG_MMR = 0,
+ ICELAND_CONFIGREG_SMC_IND,
+ ICELAND_CONFIGREG_DIDT_IND,
+ ICELAND_CONFIGREG_CACHE,
+ ICELAND_CONFIGREG_MAX
+};
+
+/* PowerContainment Features */
+#define POWERCONTAINMENT_FEATURE_DTE 0x00000001
+#define POWERCONTAINMENT_FEATURE_TDCLimit 0x00000002
+#define POWERCONTAINMENT_FEATURE_PkgPwrLimit 0x00000004
+#define POWERCONTAINMENT_FEATURE_BAPM 0x00000001
+
+struct iceland_pt_config_reg {
+ uint32_t offset;
+ uint32_t mask;
+ uint32_t shift;
+ uint32_t value;
+ enum iceland_pt_config_reg_type type;
+};
+
+struct iceland_pt_defaults
+{
+ uint8_t svi_load_line_en;
+ uint8_t svi_load_line_vddc;
+ uint8_t tdc_vddc_throttle_release_limit_perc;
+ uint8_t tdc_mawt;
+ uint8_t tdc_waterfall_ctl;
+ uint8_t dte_ambient_temp_base;
+ uint32_t display_cac;
+ uint32_t bamp_temp_gradient;
+ uint16_t bapmti_r[SMU71_DTE_ITERATIONS * SMU71_DTE_SOURCES * SMU71_DTE_SINKS];
+ uint16_t bapmti_rc[SMU71_DTE_ITERATIONS * SMU71_DTE_SOURCES * SMU71_DTE_SINKS];
+};
+
+void iceland_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr);
+int iceland_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr);
+int iceland_populate_pm_fuses(struct pp_hwmgr *hwmgr);
+int iceland_enable_smc_cac(struct pp_hwmgr *hwmgr);
+int iceland_enable_power_containment(struct pp_hwmgr *hwmgr);
+int iceland_power_control_set_level(struct pp_hwmgr *hwmgr);
+
+#endif /* ICELAND_POWERTUNE_H */
+
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+#include <asm/div64.h>
+#include "iceland_thermal.h"
+#include "iceland_hwmgr.h"
+#include "iceland_smumgr.h"
+#include "atombios.h"
+#include "ppsmc.h"
+
+#include "gmc/gmc_8_1_d.h"
+#include "gmc/gmc_8_1_sh_mask.h"
+
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+
+#include "smu/smu_7_1_1_d.h"
+#include "smu/smu_7_1_1_sh_mask.h"
+
+
+/**
+* Get Fan Speed Control Parameters.
+* @param hwmgr the address of the powerplay hardware manager.
+* @param pSpeed is the address of the structure where the result is to be placed.
+* @exception Always succeeds except if we cannot zero out the output structure.
+*/
+int iceland_fan_ctrl_get_fan_speed_info(struct pp_hwmgr *hwmgr,
+ struct phm_fan_speed_info *fan_speed_info)
+{
+
+ if (hwmgr->thermal_controller.fanInfo.bNoFan)
+ return 0;
+
+ fan_speed_info->supports_percent_read = true;
+ fan_speed_info->supports_percent_write = true;
+ fan_speed_info->min_percent = 0;
+ fan_speed_info->max_percent = 100;
+
+ if (0 != hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution) {
+ fan_speed_info->supports_rpm_read = true;
+ fan_speed_info->supports_rpm_write = true;
+ fan_speed_info->min_rpm = hwmgr->thermal_controller.fanInfo.ulMinRPM;
+ fan_speed_info->max_rpm = hwmgr->thermal_controller.fanInfo.ulMaxRPM;
+ } else {
+ fan_speed_info->min_rpm = 0;
+ fan_speed_info->max_rpm = 0;
+ }
+
+ return 0;
+}
+
+/**
+* Get Fan Speed in percent.
+* @param hwmgr the address of the powerplay hardware manager.
+* @param pSpeed is the address of the structure where the result is to be placed.
+* @exception Fails is the 100% setting appears to be 0.
+*/
+int iceland_fan_ctrl_get_fan_speed_percent(struct pp_hwmgr *hwmgr, uint32_t *speed)
+{
+ uint32_t duty100;
+ uint32_t duty;
+ uint64_t tmp64;
+
+ if (hwmgr->thermal_controller.fanInfo.bNoFan)
+ return 0;
+
+ duty100 = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL1, FMAX_DUTY100);
+ duty = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_STATUS, FDO_PWM_DUTY);
+
+ if (0 == duty100)
+ return -EINVAL;
+
+
+ tmp64 = (uint64_t)duty * 100;
+ do_div(tmp64, duty100);
+ *speed = (uint32_t)tmp64;
+
+ if (*speed > 100)
+ *speed = 100;
+
+ return 0;
+}
+
+/**
+* Get Fan Speed in RPM.
+* @param hwmgr the address of the powerplay hardware manager.
+* @param speed is the address of the structure where the result is to be placed.
+* @exception Returns not supported if no fan is found or if pulses per revolution are not set
+*/
+int iceland_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed)
+{
+ return 0;
+}
+
+/**
+* Set Fan Speed Control to static mode, so that the user can decide what speed to use.
+* @param hwmgr the address of the powerplay hardware manager.
+* mode the fan control mode, 0 default, 1 by percent, 5, by RPM
+* @exception Should always succeed.
+*/
+int iceland_fan_ctrl_set_static_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
+{
+
+ if (hwmgr->fan_ctrl_is_in_default_mode) {
+ hwmgr->fan_ctrl_default_mode = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL2, FDO_PWM_MODE);
+ hwmgr->tmin = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL2, TMIN);
+ hwmgr->fan_ctrl_is_in_default_mode = false;
+ }
+
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL2, TMIN, 0);
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL2, FDO_PWM_MODE, mode);
+
+ return 0;
+}
+
+/**
+* Reset Fan Speed Control to default mode.
+* @param hwmgr the address of the powerplay hardware manager.
+* @exception Should always succeed.
+*/
+static int iceland_fan_ctrl_set_default_mode(struct pp_hwmgr *hwmgr)
+{
+ if (!hwmgr->fan_ctrl_is_in_default_mode) {
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL2, FDO_PWM_MODE, hwmgr->fan_ctrl_default_mode);
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL2, TMIN, hwmgr->tmin);
+ hwmgr->fan_ctrl_is_in_default_mode = true;
+ }
+
+ return 0;
+}
+
+int iceland_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr)
+{
+ return (smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_StartFanControl) == 0) ? 0 : -EINVAL;
+}
+
+
+int iceland_fan_ctrl_stop_smc_fan_control(struct pp_hwmgr *hwmgr)
+{
+ return (smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_StopFanControl) == 0) ? 0 : -EINVAL;
+}
+
+/**
+* Set Fan Speed in percent.
+* @param hwmgr the address of the powerplay hardware manager.
+* @param speed is the percentage value (0% - 100%) to be set.
+* @exception Fails is the 100% setting appears to be 0.
+*/
+int iceland_fan_ctrl_set_fan_speed_percent(struct pp_hwmgr *hwmgr, uint32_t speed)
+{
+ uint32_t duty100;
+ uint32_t duty;
+ uint64_t tmp64;
+
+ if (hwmgr->thermal_controller.fanInfo.bNoFan)
+ return -EINVAL;
+
+ if (speed > 100) {
+ pr_warning("Cannot set more than 100%% duty cycle. Set it to 100.\n");
+ speed = 100;
+ }
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl))
+ iceland_fan_ctrl_stop_smc_fan_control(hwmgr);
+
+ duty100 = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL1, FMAX_DUTY100);
+
+ if (0 == duty100)
+ return -EINVAL;
+
+ tmp64 = (uint64_t)speed * duty100;
+ do_div(tmp64, 100);
+ duty = (uint32_t)tmp64;
+
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL0, FDO_STATIC_DUTY, duty);
+
+ return iceland_fan_ctrl_set_static_mode(hwmgr, FDO_PWM_MODE_STATIC);
+}
+
+/**
+* Reset Fan Speed to default.
+* @param hwmgr the address of the powerplay hardware manager.
+* @exception Always succeeds.
+*/
+int iceland_fan_ctrl_reset_fan_speed_to_default(struct pp_hwmgr *hwmgr)
+{
+ int result;
+
+ if (hwmgr->thermal_controller.fanInfo.bNoFan)
+ return 0;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl)) {
+ result = iceland_fan_ctrl_set_static_mode(hwmgr, FDO_PWM_MODE_STATIC);
+ if (0 == result)
+ result = iceland_fan_ctrl_start_smc_fan_control(hwmgr);
+ } else
+ result = iceland_fan_ctrl_set_default_mode(hwmgr);
+
+ return result;
+}
+
+/**
+* Set Fan Speed in RPM.
+* @param hwmgr the address of the powerplay hardware manager.
+* @param speed is the percentage value (min - max) to be set.
+* @exception Fails is the speed not lie between min and max.
+*/
+int iceland_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed)
+{
+ return 0;
+}
+
+/**
+* Reads the remote temperature from the SIslands thermal controller.
+*
+* @param hwmgr The address of the hardware manager.
+*/
+int iceland_thermal_get_temperature(struct pp_hwmgr *hwmgr)
+{
+ int temp;
+
+ temp = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_MULT_THERMAL_STATUS, CTF_TEMP);
+
+ /*
+ * Bit 9 means the reading is lower than the lowest usable
+ * value.
+ */
+ if (0 != (0x200 & temp))
+ temp = ICELAND_THERMAL_MAXIMUM_TEMP_READING;
+ else
+ temp = (temp & 0x1ff);
+
+ temp = temp * PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+
+ return temp;
+}
+
+/**
+* Set the requested temperature range for high and low alert signals
+*
+* @param hwmgr The address of the hardware manager.
+* @param range Temperature range to be programmed for high and low alert signals
+* @exception PP_Result_BadInput if the input data is not valid.
+*/
+static int iceland_thermal_set_temperature_range(struct pp_hwmgr *hwmgr, uint32_t low_temp, uint32_t high_temp)
+{
+ uint32_t low = ICELAND_THERMAL_MINIMUM_ALERT_TEMP * PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+ uint32_t high = ICELAND_THERMAL_MAXIMUM_ALERT_TEMP * PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
+
+ if (low < low_temp)
+ low = low_temp;
+ if (high > high_temp)
+ high = high_temp;
+
+ if (low > high)
+ return -EINVAL;
+
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_INT, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_INT, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_CTRL, DIG_THERM_DPM, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+
+ return 0;
+}
+
+/**
+* Programs thermal controller one-time setting registers
+*
+* @param hwmgr The address of the hardware manager.
+*/
+static int iceland_thermal_initialize(struct pp_hwmgr *hwmgr)
+{
+ if (0 != hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution)
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+ CG_TACH_CTRL, EDGE_PER_REV,
+ hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution - 1);
+
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL2, TACH_PWM_RESP_RATE, 0x28);
+
+ return 0;
+}
+
+/**
+* Enable thermal alerts on the RV770 thermal controller.
+*
+* @param hwmgr The address of the hardware manager.
+*/
+static int iceland_thermal_enable_alert(struct pp_hwmgr *hwmgr)
+{
+ uint32_t alert;
+
+ alert = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_INT, THERM_INT_MASK);
+ alert &= ~(ICELAND_THERMAL_HIGH_ALERT_MASK | ICELAND_THERMAL_LOW_ALERT_MASK);
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_INT, THERM_INT_MASK, alert);
+
+ /* send message to SMU to enable internal thermal interrupts */
+ return (smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_Thermal_Cntl_Enable) == 0) ? 0 : -1;
+}
+
+/**
+* Disable thermal alerts on the RV770 thermal controller.
+* @param hwmgr The address of the hardware manager.
+*/
+static int iceland_thermal_disable_alert(struct pp_hwmgr *hwmgr)
+{
+ uint32_t alert;
+
+ alert = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_INT, THERM_INT_MASK);
+ alert |= (ICELAND_THERMAL_HIGH_ALERT_MASK | ICELAND_THERMAL_LOW_ALERT_MASK);
+ PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_INT, THERM_INT_MASK, alert);
+
+ /* send message to SMU to disable internal thermal interrupts */
+ return (smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_Thermal_Cntl_Disable) == 0) ? 0 : -1;
+}
+
+/**
+* Uninitialize the thermal controller.
+* Currently just disables alerts.
+* @param hwmgr The address of the hardware manager.
+*/
+int iceland_thermal_stop_thermal_controller(struct pp_hwmgr *hwmgr)
+{
+ int result = iceland_thermal_disable_alert(hwmgr);
+
+ if (result)
+ pr_warning("Failed to disable thermal alerts!\n");
+
+ if (hwmgr->thermal_controller.fanInfo.bNoFan)
+ iceland_fan_ctrl_set_default_mode(hwmgr);
+
+ return result;
+}
+
+/**
+* Set up the fan table to control the fan using the SMC.
+* @param hwmgr the address of the powerplay hardware manager.
+* @param pInput the pointer to input data
+* @param pOutput the pointer to output data
+* @param pStorage the pointer to temporary storage
+* @param Result the last failure code
+* @return result from set temperature range routine
+*/
+int tf_iceland_thermal_setup_fan_table(struct pp_hwmgr *hwmgr, void *input, void *output, void *storage, int result)
+{
+ struct iceland_hwmgr *data = (struct iceland_hwmgr *)(hwmgr->backend);
+ SMU71_Discrete_FanTable fan_table = { FDO_MODE_HARDWARE };
+ uint32_t duty100;
+ uint32_t t_diff1, t_diff2, pwm_diff1, pwm_diff2;
+ uint16_t fdo_min, slope1, slope2;
+ uint32_t reference_clock;
+ int res;
+ uint64_t tmp64;
+
+ if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl))
+ return 0;
+
+ if (0 == data->fan_table_start) {
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl);
+ return 0;
+ }
+
+ duty100 = PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_FDO_CTRL1, FMAX_DUTY100);
+
+ if (0 == duty100) {
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl);
+ return 0;
+ }
+
+ tmp64 = hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin * duty100;
+ do_div(tmp64, 10000);
+ fdo_min = (uint16_t)tmp64;
+
+ t_diff1 = hwmgr->thermal_controller.advanceFanControlParameters.usTMed - hwmgr->thermal_controller.advanceFanControlParameters.usTMin;
+ t_diff2 = hwmgr->thermal_controller.advanceFanControlParameters.usTHigh - hwmgr->thermal_controller.advanceFanControlParameters.usTMed;
+
+ pwm_diff1 = hwmgr->thermal_controller.advanceFanControlParameters.usPWMMed - hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin;
+ pwm_diff2 = hwmgr->thermal_controller.advanceFanControlParameters.usPWMHigh - hwmgr->thermal_controller.advanceFanControlParameters.usPWMMed;
+
+ slope1 = (uint16_t)((50 + ((16 * duty100 * pwm_diff1) / t_diff1)) / 100);
+ slope2 = (uint16_t)((50 + ((16 * duty100 * pwm_diff2) / t_diff2)) / 100);
+
+ fan_table.TempMin = cpu_to_be16((50 + hwmgr->thermal_controller.advanceFanControlParameters.usTMin) / 100);
+ fan_table.TempMed = cpu_to_be16((50 + hwmgr->thermal_controller.advanceFanControlParameters.usTMed) / 100);
+ fan_table.TempMax = cpu_to_be16((50 + hwmgr->thermal_controller.advanceFanControlParameters.usTMax) / 100);
+
+ fan_table.Slope1 = cpu_to_be16(slope1);
+ fan_table.Slope2 = cpu_to_be16(slope2);
+
+ fan_table.FdoMin = cpu_to_be16(fdo_min);
+
+ fan_table.HystDown = cpu_to_be16(hwmgr->thermal_controller.advanceFanControlParameters.ucTHyst);
+
+ fan_table.HystUp = cpu_to_be16(1);
+
+ fan_table.HystSlope = cpu_to_be16(1);
+
+ fan_table.TempRespLim = cpu_to_be16(5);
+
+ reference_clock = iceland_get_xclk(hwmgr);
+
+ fan_table.RefreshPeriod = cpu_to_be32((hwmgr->thermal_controller.advanceFanControlParameters.ulCycleDelay * reference_clock) / 1600);
+
+ fan_table.FdoMax = cpu_to_be16((uint16_t)duty100);
+
+ fan_table.TempSrc = (uint8_t)PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_MULT_THERMAL_CTRL, TEMP_SEL);
+
+ //fan_table.FanControl_GL_Flag = 1;
+
+ res = iceland_copy_bytes_to_smc(hwmgr->smumgr, data->fan_table_start, (uint8_t *)&fan_table, (uint32_t)sizeof(fan_table), data->sram_end);
+/* TO DO FOR SOME DEVICE ID 0X692b, send this msg return invalid command.
+ if (res == 0 && hwmgr->thermal_controller.advanceFanControlParameters.ucMinimumPWMLimit != 0)
+ res = (0 == smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, PPSMC_MSG_SetFanMinPwm, \
+ hwmgr->thermal_controller.advanceFanControlParameters.ucMinimumPWMLimit) ? 0 : -1);
+
+ if (res == 0 && hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit != 0)
+ res = (0 == smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, PPSMC_MSG_SetFanSclkTarget, \
+ hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit) ? 0 : -1);
+
+ if (0 != res)
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl);
+*/
+ return 0;
+}
+
+/**
+* Start the fan control on the SMC.
+* @param hwmgr the address of the powerplay hardware manager.
+* @param pInput the pointer to input data
+* @param pOutput the pointer to output data
+* @param pStorage the pointer to temporary storage
+* @param Result the last failure code
+* @return result from set temperature range routine
+*/
+int tf_iceland_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr, void *input, void *output, void *storage, int result)
+{
+/* If the fantable setup has failed we could have disabled PHM_PlatformCaps_MicrocodeFanControl even after this function was included in the table.
+ * Make sure that we still think controlling the fan is OK.
+*/
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_MicrocodeFanControl)) {
+ iceland_fan_ctrl_start_smc_fan_control(hwmgr);
+ iceland_fan_ctrl_set_static_mode(hwmgr, FDO_PWM_MODE_STATIC);
+ }
+
+ return 0;
+}
+
+/**
+* Set temperature range for high and low alerts
+* @param hwmgr the address of the powerplay hardware manager.
+* @param pInput the pointer to input data
+* @param pOutput the pointer to output data
+* @param pStorage the pointer to temporary storage
+* @param Result the last failure code
+* @return result from set temperature range routine
+*/
+static int tf_iceland_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
+ void *input, void *output, void *storage, int result)
+{
+ struct PP_TemperatureRange *range = (struct PP_TemperatureRange *)input;
+
+ if (range == NULL)
+ return -EINVAL;
+
+ return iceland_thermal_set_temperature_range(hwmgr, range->min, range->max);
+}
+
+/**
+* Programs one-time setting registers
+* @param hwmgr the address of the powerplay hardware manager.
+* @param pInput the pointer to input data
+* @param pOutput the pointer to output data
+* @param pStorage the pointer to temporary storage
+* @param Result the last failure code
+* @return result from initialize thermal controller routine
+*/
+static int tf_iceland_thermal_initialize(struct pp_hwmgr *hwmgr, void *input,
+ void *output, void *storage, int result)
+{
+ return iceland_thermal_initialize(hwmgr);
+}
+
+/**
+* Enable high and low alerts
+* @param hwmgr the address of the powerplay hardware manager.
+* @param pInput the pointer to input data
+* @param pOutput the pointer to output data
+* @param pStorage the pointer to temporary storage
+* @param Result the last failure code
+* @return result from enable alert routine
+*/
+static int tf_iceland_thermal_enable_alert(struct pp_hwmgr *hwmgr,
+ void *input, void *output, void *storage, int result)
+{
+ return iceland_thermal_enable_alert(hwmgr);
+}
+
+/**
+* Disable high and low alerts
+* @param hwmgr the address of the powerplay hardware manager.
+* @param pInput the pointer to input data
+* @param pOutput the pointer to output data
+* @param pStorage the pointer to temporary storage
+* @param Result the last failure code
+* @return result from disable alert routine
+*/
+static int tf_iceland_thermal_disable_alert(struct pp_hwmgr *hwmgr, void *input, void *output, void *storage, int result)
+{
+ return iceland_thermal_disable_alert(hwmgr);
+}
+
+static const struct phm_master_table_item iceland_thermal_start_thermal_controller_master_list[] = {
+ { NULL, tf_iceland_thermal_initialize },
+ { NULL, tf_iceland_thermal_set_temperature_range },
+ { NULL, tf_iceland_thermal_enable_alert },
+ /*
+ * We should restrict performance levels to low before we halt
+ * the SMC. On the other hand we are still in boot state when
+ * we do this so it would be pointless. If this assumption
+ * changes we have to revisit this table.
+ */
+ { NULL, tf_iceland_thermal_setup_fan_table},
+ { NULL, tf_iceland_thermal_start_smc_fan_control},
+ { NULL, NULL }
+};
+
+static const struct phm_master_table_header iceland_thermal_start_thermal_controller_master = {
+ 0,
+ PHM_MasterTableFlag_None,
+ iceland_thermal_start_thermal_controller_master_list
+};
+
+static const struct phm_master_table_item iceland_thermal_set_temperature_range_master_list[] = {
+ { NULL, tf_iceland_thermal_disable_alert},
+ { NULL, tf_iceland_thermal_set_temperature_range},
+ { NULL, tf_iceland_thermal_enable_alert},
+ { NULL, NULL }
+};
+
+static const struct phm_master_table_header iceland_thermal_set_temperature_range_master = {
+ 0,
+ PHM_MasterTableFlag_None,
+ iceland_thermal_set_temperature_range_master_list
+};
+
+int iceland_thermal_ctrl_uninitialize_thermal_controller(struct pp_hwmgr *hwmgr)
+{
+ if (!hwmgr->thermal_controller.fanInfo.bNoFan)
+ iceland_fan_ctrl_set_default_mode(hwmgr);
+ return 0;
+}
+
+/**
+* Initializes the thermal controller related functions in the Hardware Manager structure.
+* @param hwmgr The address of the hardware manager.
+* @exception Any error code from the low-level communication.
+*/
+int pp_iceland_thermal_initialize(struct pp_hwmgr *hwmgr)
+{
+ int result;
+
+ result = phm_construct_table(hwmgr, &iceland_thermal_set_temperature_range_master, &(hwmgr->set_temperature_range));
+
+ if (0 == result) {
+ result = phm_construct_table(hwmgr,
+ &iceland_thermal_start_thermal_controller_master,
+ &(hwmgr->start_thermal_controller));
+ if (0 != result)
+ phm_destroy_table(hwmgr, &(hwmgr->set_temperature_range));
+ }
+
+ if (0 == result)
+ hwmgr->fan_ctrl_is_in_default_mode = true;
+ return result;
+}
+
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+
+#ifndef ICELAND_THERMAL_H
+#define ICELAND_THERMAL_H
+
+#include "hwmgr.h"
+
+#define ICELAND_THERMAL_HIGH_ALERT_MASK 0x1
+#define ICELAND_THERMAL_LOW_ALERT_MASK 0x2
+
+#define ICELAND_THERMAL_MINIMUM_TEMP_READING -256
+#define ICELAND_THERMAL_MAXIMUM_TEMP_READING 255
+
+#define ICELAND_THERMAL_MINIMUM_ALERT_TEMP 0
+#define ICELAND_THERMAL_MAXIMUM_ALERT_TEMP 255
+
+#define FDO_PWM_MODE_STATIC 1
+#define FDO_PWM_MODE_STATIC_RPM 5
+
+
+extern int iceland_thermal_get_temperature(struct pp_hwmgr *hwmgr);
+extern int iceland_thermal_stop_thermal_controller(struct pp_hwmgr *hwmgr);
+extern int iceland_fan_ctrl_get_fan_speed_info(struct pp_hwmgr *hwmgr, struct phm_fan_speed_info *fan_speed_info);
+extern int iceland_fan_ctrl_get_fan_speed_percent(struct pp_hwmgr *hwmgr, uint32_t *speed);
+extern int iceland_fan_ctrl_set_static_mode(struct pp_hwmgr *hwmgr, uint32_t mode);
+extern int iceland_fan_ctrl_set_fan_speed_percent(struct pp_hwmgr *hwmgr, uint32_t speed);
+extern int iceland_fan_ctrl_reset_fan_speed_to_default(struct pp_hwmgr *hwmgr);
+extern int pp_iceland_thermal_initialize(struct pp_hwmgr *hwmgr);
+extern int iceland_thermal_ctrl_uninitialize_thermal_controller(struct pp_hwmgr *hwmgr);
+extern int iceland_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed);
+extern int iceland_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed);
+extern int iceland_fan_ctrl_stop_smc_fan_control(struct pp_hwmgr *hwmgr);
+
+#endif
+
return 0;
}
-int polaris10_phm_powerup_uvd(struct pp_hwmgr *hwmgr)
+static int polaris10_phm_powerup_uvd(struct pp_hwmgr *hwmgr)
{
if (phm_cf_want_uvd_power_gating(hwmgr)) {
if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
return 0;
}
-int polaris10_phm_powerdown_vce(struct pp_hwmgr *hwmgr)
+static int polaris10_phm_powerdown_vce(struct pp_hwmgr *hwmgr)
{
if (phm_cf_want_vce_power_gating(hwmgr))
return smum_send_msg_to_smc(hwmgr->smumgr,
return 0;
}
-int polaris10_phm_powerup_vce(struct pp_hwmgr *hwmgr)
+static int polaris10_phm_powerup_vce(struct pp_hwmgr *hwmgr)
{
if (phm_cf_want_vce_power_gating(hwmgr))
return smum_send_msg_to_smc(hwmgr->smumgr,
return 0;
}
-int polaris10_phm_powerdown_samu(struct pp_hwmgr *hwmgr)
+static int polaris10_phm_powerdown_samu(struct pp_hwmgr *hwmgr)
{
if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_SamuPowerGating))
return 0;
}
-int polaris10_phm_powerup_samu(struct pp_hwmgr *hwmgr)
+static int polaris10_phm_powerup_samu(struct pp_hwmgr *hwmgr)
{
if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_SamuPowerGating))
#include "pp_debug.h"
#include "ppatomctrl.h"
#include "atombios.h"
-#include "tonga_pptable.h"
+#include "pptable_v1_0.h"
#include "pppcielanes.h"
#include "amd_pcie_helpers.h"
#include "hardwaremanager.h"
-#include "tonga_processpptables.h"
+#include "process_pptables_v1_0.h"
#include "cgs_common.h"
#include "smu74.h"
#include "smu_ucode_xfer_vi.h"
#define PCIE_BUS_CLK 10000
#define TCLK (PCIE_BUS_CLK / 10)
-
-static const uint16_t polaris10_clock_stretcher_lookup_table[2][4] =
-{ {600, 1050, 3, 0}, {600, 1050, 6, 1} };
-
-/* [FF, SS] type, [] 4 voltage ranges, and [Floor Freq, Boundary Freq, VID min , VID max] */
-static const uint32_t polaris10_clock_stretcher_ddt_table[2][4][4] =
-{ { {265, 529, 120, 128}, {325, 650, 96, 119}, {430, 860, 32, 95}, {0, 0, 0, 31} },
- { {275, 550, 104, 112}, {319, 638, 96, 103}, {360, 720, 64, 95}, {384, 768, 32, 63} } };
-
-/* [Use_For_Low_freq] value, [0%, 5%, 10%, 7.14%, 14.28%, 20%] (coming from PWR_CKS_CNTL.stretch_amount reg spec) */
-static const uint8_t polaris10_clock_stretch_amount_conversion[2][6] =
-{ {0, 1, 3, 2, 4, 5}, {0, 2, 4, 5, 6, 5} };
-
/** Values for the CG_THERMAL_CTRL::DPM_EVENT_SRC field. */
enum DPM_EVENT_SRC {
DPM_EVENT_SRC_ANALOG = 0,
static const unsigned long PhwPolaris10_Magic = (unsigned long)(PHM_VIslands_Magic);
-struct polaris10_power_state *cast_phw_polaris10_power_state(
+static struct polaris10_power_state *cast_phw_polaris10_power_state(
struct pp_hw_power_state *hw_ps)
{
PP_ASSERT_WITH_CODE((PhwPolaris10_Magic == hw_ps->magic),
return (struct polaris10_power_state *)hw_ps;
}
-const struct polaris10_power_state *cast_const_phw_polaris10_power_state(
+static const struct polaris10_power_state *
+cast_const_phw_polaris10_power_state(
const struct pp_hw_power_state *hw_ps)
{
PP_ASSERT_WITH_CODE((PhwPolaris10_Magic == hw_ps->magic),
* @param hwmgr the address of the powerplay hardware manager.
* @return always 0
*/
-int phm_get_mc_microcode_version (struct pp_hwmgr *hwmgr)
+static int phm_get_mc_microcode_version(struct pp_hwmgr *hwmgr)
{
cgs_write_register(hwmgr->device, mmMC_SEQ_IO_DEBUG_INDEX, 0x9F);
return 0;
}
-uint16_t phm_get_current_pcie_speed(struct pp_hwmgr *hwmgr)
+static uint16_t phm_get_current_pcie_speed(struct pp_hwmgr *hwmgr)
{
uint32_t speedCntl = 0;
PCIE_LC_SPEED_CNTL, LC_CURRENT_DATA_RATE));
}
-int phm_get_current_pcie_lane_number(struct pp_hwmgr *hwmgr)
+static int phm_get_current_pcie_lane_number(struct pp_hwmgr *hwmgr)
{
uint32_t link_width;
* @param pHwMgr the address of the powerplay hardware manager.
* @return always PP_Result_OK
*/
-int polaris10_enable_smc_voltage_controller(struct pp_hwmgr *hwmgr)
+static int polaris10_enable_smc_voltage_controller(struct pp_hwmgr *hwmgr)
{
PP_ASSERT_WITH_CODE(
(hwmgr->smumgr->smumgr_funcs->send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_Voltage_Cntl_Enable) == 0),
* on the power policy or external client requests,
* such as UVD request, etc.
*/
-int polaris10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
+static int polaris10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
{
struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
struct phm_ppt_v1_information *table_info =
return 0;
}
-uint8_t convert_to_vid(uint16_t vddc)
-{
- return (uint8_t) ((6200 - (vddc * VOLTAGE_SCALE)) / 25);
-}
-
/**
* Mvdd table preparation for SMC.
*
* @return always 0
*/
-int polaris10_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr,
+static int polaris10_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr,
struct SMU74_Discrete_DpmTable *table)
{
polaris10_populate_smc_vddci_table(hwmgr, table);
* @param mclk the MCLK value to be used in the decision if MVDD should be high or low.
* @param voltage the SMC VOLTAGE structure to be populated
*/
-int polaris10_populate_mvdd_value(struct pp_hwmgr *hwmgr,
+static int polaris10_populate_mvdd_value(struct pp_hwmgr *hwmgr,
uint32_t mclk, SMIO_Pattern *smio_pat)
{
const struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
}
-int polaris10_populate_avfs_parameters(struct pp_hwmgr *hwmgr)
+static int polaris10_populate_avfs_parameters(struct pp_hwmgr *hwmgr)
{
struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
SMU74_Discrete_DpmTable *table = &(data->smc_state_table);
return polaris10_disable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal);
}
-int polaris10_pcie_performance_request(struct pp_hwmgr *hwmgr)
+static int polaris10_pcie_performance_request(struct pp_hwmgr *hwmgr)
{
struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
data->pcie_performance_request = true;
return 0;
}
-int polaris10_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
+static int polaris10_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
{
int tmp_result, result = 0;
tmp_result = (!polaris10_is_dpm_running(hwmgr)) ? 0 : -1;
return 0;
}
-int polaris10_hwmgr_backend_fini(struct pp_hwmgr *hwmgr)
+static int polaris10_hwmgr_backend_fini(struct pp_hwmgr *hwmgr)
{
return phm_hwmgr_backend_fini(hwmgr);
}
-int polaris10_set_features_platform_caps(struct pp_hwmgr *hwmgr)
+static int polaris10_set_features_platform_caps(struct pp_hwmgr *hwmgr)
{
struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_SclkDeepSleep);
-
phm_cap_set(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_DynamicPatchPowerState);
phm_cap_set(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_TCPRamping);
- if (hwmgr->powercontainment_enabled)
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_PowerContainment);
- else
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_PowerContainment);
-
phm_cap_set(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_CAC);
continue;
}
- /* need to make sure vddc is less than 2v or else, it could burn the ASIC.
- * real voltage level in unit of 0.01mv */
+ /* need to make sure vddc is less than 2V or else, it could burn the ASIC.
+ * real voltage level in unit of 0.01mV */
PP_ASSERT_WITH_CODE((vddc < 200000 && vddc != 0),
"Invalid VDDC value", result = -EINVAL;);
return 0;
}
-int polaris10_patch_voltage_workaround(struct pp_hwmgr *hwmgr)
+static int polaris10_patch_voltage_workaround(struct pp_hwmgr *hwmgr)
{
struct phm_ppt_v1_information *table_info =
(struct phm_ppt_v1_information *)(hwmgr->pptable);
struct phm_ppt_v1_voltage_lookup_table *lookup_table =
table_info->vddc_lookup_table;
uint32_t i;
+ uint32_t hw_revision, sub_vendor_id, sub_sys_id;
+ struct cgs_system_info sys_info = {0};
+
+ sys_info.size = sizeof(struct cgs_system_info);
- if (hwmgr->chip_id == CHIP_POLARIS10 && hwmgr->hw_revision == 0xC7) {
+ sys_info.info_id = CGS_SYSTEM_INFO_PCIE_REV;
+ cgs_query_system_info(hwmgr->device, &sys_info);
+ hw_revision = (uint32_t)sys_info.value;
+
+ sys_info.info_id = CGS_SYSTEM_INFO_PCIE_SUB_SYS_ID;
+ cgs_query_system_info(hwmgr->device, &sys_info);
+ sub_sys_id = (uint32_t)sys_info.value;
+
+ sys_info.info_id = CGS_SYSTEM_INFO_PCIE_SUB_SYS_VENDOR_ID;
+ cgs_query_system_info(hwmgr->device, &sys_info);
+ sub_vendor_id = (uint32_t)sys_info.value;
+
+ if (hwmgr->chip_id == CHIP_POLARIS10 && hw_revision == 0xC7 &&
+ ((sub_sys_id == 0xb37 && sub_vendor_id == 0x1002) ||
+ (sub_sys_id == 0x4a8 && sub_vendor_id == 0x1043) ||
+ (sub_sys_id == 0x9480 && sub_vendor_id == 0x1682))) {
if (lookup_table->entries[dep_mclk_table->entries[dep_mclk_table->count-1].vddInd].us_vdd >= 1000)
return 0;
}
-int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
+static int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
{
struct polaris10_hwmgr *data;
struct pp_atomctrl_gpio_pin_assignment gpio_pin_assignment;
ps = (struct polaris10_power_state *)(&state->hardware);
- result = tonga_get_powerplay_table_entry(hwmgr, entry_index, state,
+ result = get_powerplay_table_entry_v1_0(hwmgr, entry_index, state,
polaris10_get_pp_table_entry_callback_func);
/* This is the earliest time we have all the dependency table and the VBIOS boot state
return 0;
}
-int polaris10_enable_disable_uvd_dpm(struct pp_hwmgr *hwmgr, bool enable)
+static int
+polaris10_enable_disable_uvd_dpm(struct pp_hwmgr *hwmgr, bool enable)
{
return smum_send_msg_to_smc(hwmgr->smumgr, enable ?
PPSMC_MSG_UVDDPM_Enable :
PPSMC_MSG_VCEDPM_Disable);
}
-int polaris10_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable)
+static int
+polaris10_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable)
{
return smum_send_msg_to_smc(hwmgr->smumgr, enable?
PPSMC_MSG_SAMUDPM_Enable :
}
-int polaris10_notify_smc_display_change(struct pp_hwmgr *hwmgr, bool has_display)
+static int
+polaris10_notify_smc_display_change(struct pp_hwmgr *hwmgr, bool has_display)
{
PPSMC_Msg msg = has_display ? (PPSMC_Msg)PPSMC_HasDisplay : (PPSMC_Msg)PPSMC_NoDisplay;
return (smum_send_msg_to_smc(hwmgr->smumgr, msg) == 0) ? 0 : -1;
}
-int polaris10_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr)
+static int
+polaris10_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr)
{
uint32_t num_active_displays = 0;
struct cgs_display_info info = {0};
* @param hwmgr the address of the powerplay hardware manager.
* @return always OK
*/
-int polaris10_program_display_gap(struct pp_hwmgr *hwmgr)
+static int polaris10_program_display_gap(struct pp_hwmgr *hwmgr)
{
struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
uint32_t num_active_displays = 0;
}
-int polaris10_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
+static int polaris10_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
{
return polaris10_program_display_gap(hwmgr);
}
PPSMC_MSG_SetFanRpmMax, us_max_fan_rpm);
}
-int polaris10_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr,
+static int
+polaris10_register_internal_thermal_interrupt(struct pp_hwmgr *hwmgr,
const void *thermal_interrupt_info)
{
return 0;
}
-bool polaris10_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr)
+static bool polaris10_check_smc_update_required_for_display_configuration(
+ struct pp_hwmgr *hwmgr)
{
struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
bool is_update_required = false;
(pl1->pcie_lane == pl2->pcie_lane));
}
-int polaris10_check_states_equal(struct pp_hwmgr *hwmgr, const struct pp_hw_power_state *pstate1, const struct pp_hw_power_state *pstate2, bool *equal)
+static int polaris10_check_states_equal(struct pp_hwmgr *hwmgr,
+ const struct pp_hw_power_state *pstate1,
+ const struct pp_hw_power_state *pstate2, bool *equal)
{
const struct polaris10_power_state *psa = cast_const_phw_polaris10_power_state(pstate1);
const struct polaris10_power_state *psb = cast_const_phw_polaris10_power_state(pstate2);
return 0;
}
-int polaris10_upload_mc_firmware(struct pp_hwmgr *hwmgr)
+static int polaris10_upload_mc_firmware(struct pp_hwmgr *hwmgr)
{
struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
return 0;
}
-int polaris10_setup_asic_task(struct pp_hwmgr *hwmgr)
+static int polaris10_setup_asic_task(struct pp_hwmgr *hwmgr)
{
int tmp_result, result = 0;
.get_sclk = polaris10_dpm_get_sclk,
.patch_boot_state = polaris10_dpm_patch_boot_state,
.get_pp_table_entry = polaris10_get_pp_table_entry,
- .get_num_of_pp_table_entries = tonga_get_number_of_powerplay_table_entries,
+ .get_num_of_pp_table_entries = get_number_of_powerplay_table_entries_v1_0,
.print_current_perforce_level = polaris10_print_current_perforce_level,
.powerdown_uvd = polaris10_phm_powerdown_uvd,
.powergate_uvd = polaris10_phm_powergate_uvd,
int polaris10_hwmgr_init(struct pp_hwmgr *hwmgr)
{
hwmgr->hwmgr_func = &polaris10_hwmgr_funcs;
- hwmgr->pptable_func = &tonga_pptable_funcs;
+ hwmgr->pptable_func = &pptable_v1_0_funcs;
pp_polaris10_thermal_initialize(hwmgr);
return 0;
#include "ppatomctrl.h"
#include "polaris10_ppsmc.h"
#include "polaris10_powertune.h"
+#include "polaris10_smumgr.h"
#define POLARIS10_MAX_HARDWARE_POWERLEVELS 2
uint16_t max;
uint16_t min;
};
-struct polaris10_range_table {
- uint32_t trans_lower_frequency; /* in 10khz */
- uint32_t trans_upper_frequency;
-};
struct polaris10_hwmgr {
struct polaris10_dpm_table dpm_table;
enum polaris10_pt_config_reg_type type;
};
-struct polaris10_pt_defaults {
- uint8_t SviLoadLineEn;
- uint8_t SviLoadLineVddC;
- uint8_t TDC_VDDC_ThrottleReleaseLimitPerc;
- uint8_t TDC_MAWt;
- uint8_t TdcWaterfallCtl;
- uint8_t DTEAmbientTempBase;
-
- uint32_t DisplayCac;
- uint32_t BAPM_TEMP_GRADIENT;
- uint16_t BAPMTI_R[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS];
- uint16_t BAPMTI_RC[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS];
-};
void polaris10_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr);
int polaris10_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr);
return 0;
}
-int polaris10_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr)
+static int polaris10_fan_ctrl_start_smc_fan_control(struct pp_hwmgr *hwmgr)
{
int result;
* @param Result the last failure code
* @return result from set temperature range routine
*/
-int tf_polaris10_thermal_setup_fan_table(struct pp_hwmgr *hwmgr,
+static int tf_polaris10_thermal_setup_fan_table(struct pp_hwmgr *hwmgr,
void *input, void *output, void *storage, int result)
{
struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend);
* @param Result the last failure code
* @return result from set temperature range routine
*/
-int tf_polaris10_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr,
+static int tf_polaris10_thermal_start_smc_fan_control(struct pp_hwmgr *hwmgr,
void *input, void *output, void *storage, int result)
{
/* If the fantable setup has failed we could have disabled
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef TONGA_PPTABLE_H
+#define TONGA_PPTABLE_H
+
+/** \file
+ * This is a PowerPlay table header file
+ */
+#pragma pack(push, 1)
+
+#include "hwmgr.h"
+
+#define ATOM_TONGA_PP_FANPARAMETERS_TACHOMETER_PULSES_PER_REVOLUTION_MASK 0x0f
+#define ATOM_TONGA_PP_FANPARAMETERS_NOFAN 0x80 /* No fan is connected to this controller. */
+
+#define ATOM_TONGA_PP_THERMALCONTROLLER_NONE 0
+#define ATOM_TONGA_PP_THERMALCONTROLLER_LM96163 17
+#define ATOM_TONGA_PP_THERMALCONTROLLER_TONGA 21
+#define ATOM_TONGA_PP_THERMALCONTROLLER_FIJI 22
+
+/*
+ * Thermal controller 'combo type' to use an external controller for Fan control and an internal controller for thermal.
+ * We probably should reserve the bit 0x80 for this use.
+ * To keep the number of these types low we should also use the same code for all ASICs (i.e. do not distinguish RV6xx and RV7xx Internal here).
+ * The driver can pick the correct internal controller based on the ASIC.
+ */
+
+#define ATOM_TONGA_PP_THERMALCONTROLLER_ADT7473_WITH_INTERNAL 0x89 /* ADT7473 Fan Control + Internal Thermal Controller */
+#define ATOM_TONGA_PP_THERMALCONTROLLER_EMC2103_WITH_INTERNAL 0x8D /* EMC2103 Fan Control + Internal Thermal Controller */
+
+/*/* ATOM_TONGA_POWERPLAYTABLE::ulPlatformCaps */
+#define ATOM_TONGA_PP_PLATFORM_CAP_VDDGFX_CONTROL 0x1 /* This cap indicates whether vddgfx will be a separated power rail. */
+#define ATOM_TONGA_PP_PLATFORM_CAP_POWERPLAY 0x2 /* This cap indicates whether this is a mobile part and CCC need to show Powerplay page. */
+#define ATOM_TONGA_PP_PLATFORM_CAP_SBIOSPOWERSOURCE 0x4 /* This cap indicates whether power source notificaiton is done by SBIOS directly. */
+#define ATOM_TONGA_PP_PLATFORM_CAP_DISABLE_VOLTAGE_ISLAND 0x8 /* Enable the option to overwrite voltage island feature to be disabled, regardless of VddGfx power rail support. */
+#define ____RETIRE16____ 0x10
+#define ATOM_TONGA_PP_PLATFORM_CAP_HARDWAREDC 0x20 /* This cap indicates whether power source notificaiton is done by GPIO directly. */
+#define ____RETIRE64____ 0x40
+#define ____RETIRE128____ 0x80
+#define ____RETIRE256____ 0x100
+#define ____RETIRE512____ 0x200
+#define ____RETIRE1024____ 0x400
+#define ____RETIRE2048____ 0x800
+#define ATOM_TONGA_PP_PLATFORM_CAP_MVDD_CONTROL 0x1000 /* This cap indicates dynamic MVDD is required. Uncheck to disable it. */
+#define ____RETIRE2000____ 0x2000
+#define ____RETIRE4000____ 0x4000
+#define ATOM_TONGA_PP_PLATFORM_CAP_VDDCI_CONTROL 0x8000 /* This cap indicates dynamic VDDCI is required. Uncheck to disable it. */
+#define ____RETIRE10000____ 0x10000
+#define ATOM_TONGA_PP_PLATFORM_CAP_BACO 0x20000 /* Enable to indicate the driver supports BACO state. */
+
+#define ATOM_TONGA_PP_PLATFORM_CAP_OUTPUT_THERMAL2GPIO17 0x100000 /* Enable to indicate the driver supports thermal2GPIO17. */
+#define ATOM_TONGA_PP_PLATFORM_COMBINE_PCC_WITH_THERMAL_SIGNAL 0x1000000 /* Enable to indicate if thermal and PCC are sharing the same GPIO */
+#define ATOM_TONGA_PLATFORM_LOAD_POST_PRODUCTION_FIRMWARE 0x2000000
+
+/* ATOM_PPLIB_NONCLOCK_INFO::usClassification */
+#define ATOM_PPLIB_CLASSIFICATION_UI_MASK 0x0007
+#define ATOM_PPLIB_CLASSIFICATION_UI_SHIFT 0
+#define ATOM_PPLIB_CLASSIFICATION_UI_NONE 0
+#define ATOM_PPLIB_CLASSIFICATION_UI_BATTERY 1
+#define ATOM_PPLIB_CLASSIFICATION_UI_BALANCED 3
+#define ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE 5
+/* 2, 4, 6, 7 are reserved */
+
+#define ATOM_PPLIB_CLASSIFICATION_BOOT 0x0008
+#define ATOM_PPLIB_CLASSIFICATION_THERMAL 0x0010
+#define ATOM_PPLIB_CLASSIFICATION_LIMITEDPOWERSOURCE 0x0020
+#define ATOM_PPLIB_CLASSIFICATION_REST 0x0040
+#define ATOM_PPLIB_CLASSIFICATION_FORCED 0x0080
+#define ATOM_PPLIB_CLASSIFICATION_ACPI 0x1000
+
+/* ATOM_PPLIB_NONCLOCK_INFO::usClassification2 */
+#define ATOM_PPLIB_CLASSIFICATION2_LIMITEDPOWERSOURCE_2 0x0001
+
+#define ATOM_Tonga_DISALLOW_ON_DC 0x00004000
+#define ATOM_Tonga_ENABLE_VARIBRIGHT 0x00008000
+
+#define ATOM_Tonga_TABLE_REVISION_TONGA 7
+
+typedef struct _ATOM_Tonga_POWERPLAYTABLE {
+ ATOM_COMMON_TABLE_HEADER sHeader;
+
+ UCHAR ucTableRevision;
+ USHORT usTableSize; /*the size of header structure */
+
+ ULONG ulGoldenPPID;
+ ULONG ulGoldenRevision;
+ USHORT usFormatID;
+
+ USHORT usVoltageTime; /*in microseconds */
+ ULONG ulPlatformCaps; /*See ATOM_Tonga_CAPS_* */
+
+ ULONG ulMaxODEngineClock; /*For Overdrive. */
+ ULONG ulMaxODMemoryClock; /*For Overdrive. */
+
+ USHORT usPowerControlLimit;
+ USHORT usUlvVoltageOffset; /*in mv units */
+
+ USHORT usStateArrayOffset; /*points to ATOM_Tonga_State_Array */
+ USHORT usFanTableOffset; /*points to ATOM_Tonga_Fan_Table */
+ USHORT usThermalControllerOffset; /*points to ATOM_Tonga_Thermal_Controller */
+ USHORT usReserv; /*CustomThermalPolicy removed for Tonga. Keep this filed as reserved. */
+
+ USHORT usMclkDependencyTableOffset; /*points to ATOM_Tonga_MCLK_Dependency_Table */
+ USHORT usSclkDependencyTableOffset; /*points to ATOM_Tonga_SCLK_Dependency_Table */
+ USHORT usVddcLookupTableOffset; /*points to ATOM_Tonga_Voltage_Lookup_Table */
+ USHORT usVddgfxLookupTableOffset; /*points to ATOM_Tonga_Voltage_Lookup_Table */
+
+ USHORT usMMDependencyTableOffset; /*points to ATOM_Tonga_MM_Dependency_Table */
+
+ USHORT usVCEStateTableOffset; /*points to ATOM_Tonga_VCE_State_Table; */
+
+ USHORT usPPMTableOffset; /*points to ATOM_Tonga_PPM_Table */
+ USHORT usPowerTuneTableOffset; /*points to ATOM_PowerTune_Table */
+
+ USHORT usHardLimitTableOffset; /*points to ATOM_Tonga_Hard_Limit_Table */
+
+ USHORT usPCIETableOffset; /*points to ATOM_Tonga_PCIE_Table */
+
+ USHORT usGPIOTableOffset; /*points to ATOM_Tonga_GPIO_Table */
+
+ USHORT usReserved[6]; /*TODO: modify reserved size to fit structure aligning */
+} ATOM_Tonga_POWERPLAYTABLE;
+
+typedef struct _ATOM_Tonga_State {
+ UCHAR ucEngineClockIndexHigh;
+ UCHAR ucEngineClockIndexLow;
+
+ UCHAR ucMemoryClockIndexHigh;
+ UCHAR ucMemoryClockIndexLow;
+
+ UCHAR ucPCIEGenLow;
+ UCHAR ucPCIEGenHigh;
+
+ UCHAR ucPCIELaneLow;
+ UCHAR ucPCIELaneHigh;
+
+ USHORT usClassification;
+ ULONG ulCapsAndSettings;
+ USHORT usClassification2;
+ UCHAR ucUnused[4];
+} ATOM_Tonga_State;
+
+typedef struct _ATOM_Tonga_State_Array {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries; /* Number of entries. */
+ ATOM_Tonga_State entries[1]; /* Dynamically allocate entries. */
+} ATOM_Tonga_State_Array;
+
+typedef struct _ATOM_Tonga_MCLK_Dependency_Record {
+ UCHAR ucVddcInd; /* Vddc voltage */
+ USHORT usVddci;
+ USHORT usVddgfxOffset; /* Offset relative to Vddc voltage */
+ USHORT usMvdd;
+ ULONG ulMclk;
+ USHORT usReserved;
+} ATOM_Tonga_MCLK_Dependency_Record;
+
+typedef struct _ATOM_Tonga_MCLK_Dependency_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries; /* Number of entries. */
+ ATOM_Tonga_MCLK_Dependency_Record entries[1]; /* Dynamically allocate entries. */
+} ATOM_Tonga_MCLK_Dependency_Table;
+
+typedef struct _ATOM_Tonga_SCLK_Dependency_Record {
+ UCHAR ucVddInd; /* Base voltage */
+ USHORT usVddcOffset; /* Offset relative to base voltage */
+ ULONG ulSclk;
+ USHORT usEdcCurrent;
+ UCHAR ucReliabilityTemperature;
+ UCHAR ucCKSVOffsetandDisable; /* Bits 0~6: Voltage offset for CKS, Bit 7: Disable/enable for the SCLK level. */
+} ATOM_Tonga_SCLK_Dependency_Record;
+
+typedef struct _ATOM_Tonga_SCLK_Dependency_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries; /* Number of entries. */
+ ATOM_Tonga_SCLK_Dependency_Record entries[1]; /* Dynamically allocate entries. */
+} ATOM_Tonga_SCLK_Dependency_Table;
+
+typedef struct _ATOM_Polaris_SCLK_Dependency_Record {
+ UCHAR ucVddInd; /* Base voltage */
+ USHORT usVddcOffset; /* Offset relative to base voltage */
+ ULONG ulSclk;
+ USHORT usEdcCurrent;
+ UCHAR ucReliabilityTemperature;
+ UCHAR ucCKSVOffsetandDisable; /* Bits 0~6: Voltage offset for CKS, Bit 7: Disable/enable for the SCLK level. */
+ ULONG ulSclkOffset;
+} ATOM_Polaris_SCLK_Dependency_Record;
+
+typedef struct _ATOM_Polaris_SCLK_Dependency_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries; /* Number of entries. */
+ ATOM_Polaris_SCLK_Dependency_Record entries[1]; /* Dynamically allocate entries. */
+} ATOM_Polaris_SCLK_Dependency_Table;
+
+typedef struct _ATOM_Tonga_PCIE_Record {
+ UCHAR ucPCIEGenSpeed;
+ UCHAR usPCIELaneWidth;
+ UCHAR ucReserved[2];
+} ATOM_Tonga_PCIE_Record;
+
+typedef struct _ATOM_Tonga_PCIE_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries; /* Number of entries. */
+ ATOM_Tonga_PCIE_Record entries[1]; /* Dynamically allocate entries. */
+} ATOM_Tonga_PCIE_Table;
+
+typedef struct _ATOM_Polaris10_PCIE_Record {
+ UCHAR ucPCIEGenSpeed;
+ UCHAR usPCIELaneWidth;
+ UCHAR ucReserved[2];
+ ULONG ulPCIE_Sclk;
+} ATOM_Polaris10_PCIE_Record;
+
+typedef struct _ATOM_Polaris10_PCIE_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries; /* Number of entries. */
+ ATOM_Polaris10_PCIE_Record entries[1]; /* Dynamically allocate entries. */
+} ATOM_Polaris10_PCIE_Table;
+
+
+typedef struct _ATOM_Tonga_MM_Dependency_Record {
+ UCHAR ucVddcInd; /* VDDC voltage */
+ USHORT usVddgfxOffset; /* Offset relative to VDDC voltage */
+ ULONG ulDClk; /* UVD D-clock */
+ ULONG ulVClk; /* UVD V-clock */
+ ULONG ulEClk; /* VCE clock */
+ ULONG ulAClk; /* ACP clock */
+ ULONG ulSAMUClk; /* SAMU clock */
+} ATOM_Tonga_MM_Dependency_Record;
+
+typedef struct _ATOM_Tonga_MM_Dependency_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries; /* Number of entries. */
+ ATOM_Tonga_MM_Dependency_Record entries[1]; /* Dynamically allocate entries. */
+} ATOM_Tonga_MM_Dependency_Table;
+
+typedef struct _ATOM_Tonga_Voltage_Lookup_Record {
+ USHORT usVdd; /* Base voltage */
+ USHORT usCACLow;
+ USHORT usCACMid;
+ USHORT usCACHigh;
+} ATOM_Tonga_Voltage_Lookup_Record;
+
+typedef struct _ATOM_Tonga_Voltage_Lookup_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries; /* Number of entries. */
+ ATOM_Tonga_Voltage_Lookup_Record entries[1]; /* Dynamically allocate entries. */
+} ATOM_Tonga_Voltage_Lookup_Table;
+
+typedef struct _ATOM_Tonga_Fan_Table {
+ UCHAR ucRevId; /* Change this if the table format changes or version changes so that the other fields are not the same. */
+ UCHAR ucTHyst; /* Temperature hysteresis. Integer. */
+ USHORT usTMin; /* The temperature, in 0.01 centigrades, below which we just run at a minimal PWM. */
+ USHORT usTMed; /* The middle temperature where we change slopes. */
+ USHORT usTHigh; /* The high point above TMed for adjusting the second slope. */
+ USHORT usPWMMin; /* The minimum PWM value in percent (0.01% increments). */
+ USHORT usPWMMed; /* The PWM value (in percent) at TMed. */
+ USHORT usPWMHigh; /* The PWM value at THigh. */
+ USHORT usTMax; /* The max temperature */
+ UCHAR ucFanControlMode; /* Legacy or Fuzzy Fan mode */
+ USHORT usFanPWMMax; /* Maximum allowed fan power in percent */
+ USHORT usFanOutputSensitivity; /* Sensitivity of fan reaction to temepature changes */
+ USHORT usFanRPMMax; /* The default value in RPM */
+ ULONG ulMinFanSCLKAcousticLimit; /* Minimum Fan Controller SCLK Frequency Acoustic Limit. */
+ UCHAR ucTargetTemperature; /* Advanced fan controller target temperature. */
+ UCHAR ucMinimumPWMLimit; /* The minimum PWM that the advanced fan controller can set. This should be set to the highest PWM that will run the fan at its lowest RPM. */
+ USHORT usReserved;
+} ATOM_Tonga_Fan_Table;
+
+typedef struct _ATOM_Fiji_Fan_Table {
+ UCHAR ucRevId; /* Change this if the table format changes or version changes so that the other fields are not the same. */
+ UCHAR ucTHyst; /* Temperature hysteresis. Integer. */
+ USHORT usTMin; /* The temperature, in 0.01 centigrades, below which we just run at a minimal PWM. */
+ USHORT usTMed; /* The middle temperature where we change slopes. */
+ USHORT usTHigh; /* The high point above TMed for adjusting the second slope. */
+ USHORT usPWMMin; /* The minimum PWM value in percent (0.01% increments). */
+ USHORT usPWMMed; /* The PWM value (in percent) at TMed. */
+ USHORT usPWMHigh; /* The PWM value at THigh. */
+ USHORT usTMax; /* The max temperature */
+ UCHAR ucFanControlMode; /* Legacy or Fuzzy Fan mode */
+ USHORT usFanPWMMax; /* Maximum allowed fan power in percent */
+ USHORT usFanOutputSensitivity; /* Sensitivity of fan reaction to temepature changes */
+ USHORT usFanRPMMax; /* The default value in RPM */
+ ULONG ulMinFanSCLKAcousticLimit; /* Minimum Fan Controller SCLK Frequency Acoustic Limit. */
+ UCHAR ucTargetTemperature; /* Advanced fan controller target temperature. */
+ UCHAR ucMinimumPWMLimit; /* The minimum PWM that the advanced fan controller can set. This should be set to the highest PWM that will run the fan at its lowest RPM. */
+ USHORT usFanGainEdge;
+ USHORT usFanGainHotspot;
+ USHORT usFanGainLiquid;
+ USHORT usFanGainVrVddc;
+ USHORT usFanGainVrMvdd;
+ USHORT usFanGainPlx;
+ USHORT usFanGainHbm;
+ USHORT usReserved;
+} ATOM_Fiji_Fan_Table;
+
+typedef struct _ATOM_Tonga_Thermal_Controller {
+ UCHAR ucRevId;
+ UCHAR ucType; /* one of ATOM_TONGA_PP_THERMALCONTROLLER_* */
+ UCHAR ucI2cLine; /* as interpreted by DAL I2C */
+ UCHAR ucI2cAddress;
+ UCHAR ucFanParameters; /* Fan Control Parameters. */
+ UCHAR ucFanMinRPM; /* Fan Minimum RPM (hundreds) -- for display purposes only. */
+ UCHAR ucFanMaxRPM; /* Fan Maximum RPM (hundreds) -- for display purposes only. */
+ UCHAR ucReserved;
+ UCHAR ucFlags; /* to be defined */
+} ATOM_Tonga_Thermal_Controller;
+
+typedef struct _ATOM_Tonga_VCE_State_Record {
+ UCHAR ucVCEClockIndex; /*index into usVCEDependencyTableOffset of 'ATOM_Tonga_MM_Dependency_Table' type */
+ UCHAR ucFlag; /* 2 bits indicates memory p-states */
+ UCHAR ucSCLKIndex; /*index into ATOM_Tonga_SCLK_Dependency_Table */
+ UCHAR ucMCLKIndex; /*index into ATOM_Tonga_MCLK_Dependency_Table */
+} ATOM_Tonga_VCE_State_Record;
+
+typedef struct _ATOM_Tonga_VCE_State_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries;
+ ATOM_Tonga_VCE_State_Record entries[1];
+} ATOM_Tonga_VCE_State_Table;
+
+typedef struct _ATOM_Tonga_PowerTune_Table {
+ UCHAR ucRevId;
+ USHORT usTDP;
+ USHORT usConfigurableTDP;
+ USHORT usTDC;
+ USHORT usBatteryPowerLimit;
+ USHORT usSmallPowerLimit;
+ USHORT usLowCACLeakage;
+ USHORT usHighCACLeakage;
+ USHORT usMaximumPowerDeliveryLimit;
+ USHORT usTjMax;
+ USHORT usPowerTuneDataSetID;
+ USHORT usEDCLimit;
+ USHORT usSoftwareShutdownTemp;
+ USHORT usClockStretchAmount;
+ USHORT usReserve[2];
+} ATOM_Tonga_PowerTune_Table;
+
+typedef struct _ATOM_Fiji_PowerTune_Table {
+ UCHAR ucRevId;
+ USHORT usTDP;
+ USHORT usConfigurableTDP;
+ USHORT usTDC;
+ USHORT usBatteryPowerLimit;
+ USHORT usSmallPowerLimit;
+ USHORT usLowCACLeakage;
+ USHORT usHighCACLeakage;
+ USHORT usMaximumPowerDeliveryLimit;
+ USHORT usTjMax; /* For Fiji, this is also usTemperatureLimitEdge; */
+ USHORT usPowerTuneDataSetID;
+ USHORT usEDCLimit;
+ USHORT usSoftwareShutdownTemp;
+ USHORT usClockStretchAmount;
+ USHORT usTemperatureLimitHotspot; /*The following are added for Fiji */
+ USHORT usTemperatureLimitLiquid1;
+ USHORT usTemperatureLimitLiquid2;
+ USHORT usTemperatureLimitVrVddc;
+ USHORT usTemperatureLimitVrMvdd;
+ USHORT usTemperatureLimitPlx;
+ UCHAR ucLiquid1_I2C_address; /*Liquid */
+ UCHAR ucLiquid2_I2C_address;
+ UCHAR ucLiquid_I2C_Line;
+ UCHAR ucVr_I2C_address; /*VR */
+ UCHAR ucVr_I2C_Line;
+ UCHAR ucPlx_I2C_address; /*PLX */
+ UCHAR ucPlx_I2C_Line;
+ USHORT usReserved;
+} ATOM_Fiji_PowerTune_Table;
+
+#define ATOM_PPM_A_A 1
+#define ATOM_PPM_A_I 2
+typedef struct _ATOM_Tonga_PPM_Table {
+ UCHAR ucRevId;
+ UCHAR ucPpmDesign; /*A+I or A+A */
+ USHORT usCpuCoreNumber;
+ ULONG ulPlatformTDP;
+ ULONG ulSmallACPlatformTDP;
+ ULONG ulPlatformTDC;
+ ULONG ulSmallACPlatformTDC;
+ ULONG ulApuTDP;
+ ULONG ulDGpuTDP;
+ ULONG ulDGpuUlvPower;
+ ULONG ulTjmax;
+} ATOM_Tonga_PPM_Table;
+
+typedef struct _ATOM_Tonga_Hard_Limit_Record {
+ ULONG ulSCLKLimit;
+ ULONG ulMCLKLimit;
+ USHORT usVddcLimit;
+ USHORT usVddciLimit;
+ USHORT usVddgfxLimit;
+} ATOM_Tonga_Hard_Limit_Record;
+
+typedef struct _ATOM_Tonga_Hard_Limit_Table {
+ UCHAR ucRevId;
+ UCHAR ucNumEntries;
+ ATOM_Tonga_Hard_Limit_Record entries[1];
+} ATOM_Tonga_Hard_Limit_Table;
+
+typedef struct _ATOM_Tonga_GPIO_Table {
+ UCHAR ucRevId;
+ UCHAR ucVRHotTriggeredSclkDpmIndex; /* If VRHot signal is triggered SCLK will be limited to this DPM level */
+ UCHAR ucReserve[5];
+} ATOM_Tonga_GPIO_Table;
+
+typedef struct _PPTable_Generic_SubTable_Header {
+ UCHAR ucRevId;
+} PPTable_Generic_SubTable_Header;
+
+
+#pragma pack(pop)
+
+
+#endif
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include <linux/module.h>
+#include <linux/slab.h>
+
+#include "process_pptables_v1_0.h"
+#include "ppatomctrl.h"
+#include "atombios.h"
+#include "pp_debug.h"
+#include "hwmgr.h"
+#include "cgs_common.h"
+#include "pptable_v1_0.h"
+
+/**
+ * Private Function used during initialization.
+ * @param hwmgr Pointer to the hardware manager.
+ * @param setIt A flag indication if the capability should be set (TRUE) or reset (FALSE).
+ * @param cap Which capability to set/reset.
+ */
+static void set_hw_cap(struct pp_hwmgr *hwmgr, bool setIt, enum phm_platform_caps cap)
+{
+ if (setIt)
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps, cap);
+ else
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps, cap);
+}
+
+
+/**
+ * Private Function used during initialization.
+ * @param hwmgr Pointer to the hardware manager.
+ * @param powerplay_caps the bit array (from BIOS) of capability bits.
+ * @exception the current implementation always returns 1.
+ */
+static int set_platform_caps(struct pp_hwmgr *hwmgr, uint32_t powerplay_caps)
+{
+ PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE16____),
+ "ATOM_PP_PLATFORM_CAP_ASPM_L1 is not supported!", continue);
+ PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE64____),
+ "ATOM_PP_PLATFORM_CAP_GEMINIPRIMARY is not supported!", continue);
+ PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE512____),
+ "ATOM_PP_PLATFORM_CAP_SIDEPORTCONTROL is not supported!", continue);
+ PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE1024____),
+ "ATOM_PP_PLATFORM_CAP_TURNOFFPLL_ASPML1 is not supported!", continue);
+ PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE2048____),
+ "ATOM_PP_PLATFORM_CAP_HTLINKCONTROL is not supported!", continue);
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_POWERPLAY),
+ PHM_PlatformCaps_PowerPlaySupport
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_SBIOSPOWERSOURCE),
+ PHM_PlatformCaps_BiosPowerSourceControl
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_HARDWAREDC),
+ PHM_PlatformCaps_AutomaticDCTransition
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_MVDD_CONTROL),
+ PHM_PlatformCaps_EnableMVDDControl
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_VDDCI_CONTROL),
+ PHM_PlatformCaps_ControlVDDCI
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_VDDGFX_CONTROL),
+ PHM_PlatformCaps_ControlVDDGFX
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_BACO),
+ PHM_PlatformCaps_BACO
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_DISABLE_VOLTAGE_ISLAND),
+ PHM_PlatformCaps_DisableVoltageIsland
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_COMBINE_PCC_WITH_THERMAL_SIGNAL),
+ PHM_PlatformCaps_CombinePCCWithThermalSignal
+ );
+
+ set_hw_cap(
+ hwmgr,
+ 0 != (powerplay_caps & ATOM_TONGA_PLATFORM_LOAD_POST_PRODUCTION_FIRMWARE),
+ PHM_PlatformCaps_LoadPostProductionFirmware
+ );
+
+ return 0;
+}
+
+/**
+ * Private Function to get the PowerPlay Table Address.
+ */
+const void *get_powerplay_table(struct pp_hwmgr *hwmgr)
+{
+ int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
+
+ u16 size;
+ u8 frev, crev;
+ void *table_address = (void *)hwmgr->soft_pp_table;
+
+ if (!table_address) {
+ table_address = (ATOM_Tonga_POWERPLAYTABLE *)
+ cgs_atom_get_data_table(hwmgr->device,
+ index, &size, &frev, &crev);
+ hwmgr->soft_pp_table = table_address; /*Cache the result in RAM.*/
+ hwmgr->soft_pp_table_size = size;
+ }
+
+ return table_address;
+}
+
+static int get_vddc_lookup_table(
+ struct pp_hwmgr *hwmgr,
+ phm_ppt_v1_voltage_lookup_table **lookup_table,
+ const ATOM_Tonga_Voltage_Lookup_Table *vddc_lookup_pp_tables,
+ uint32_t max_levels
+ )
+{
+ uint32_t table_size, i;
+ phm_ppt_v1_voltage_lookup_table *table;
+ phm_ppt_v1_voltage_lookup_record *record;
+ ATOM_Tonga_Voltage_Lookup_Record *atom_record;
+
+ PP_ASSERT_WITH_CODE((0 != vddc_lookup_pp_tables->ucNumEntries),
+ "Invalid CAC Leakage PowerPlay Table!", return 1);
+
+ table_size = sizeof(uint32_t) +
+ sizeof(phm_ppt_v1_voltage_lookup_record) * max_levels;
+
+ table = kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == table)
+ return -ENOMEM;
+
+ memset(table, 0x00, table_size);
+
+ table->count = vddc_lookup_pp_tables->ucNumEntries;
+
+ for (i = 0; i < vddc_lookup_pp_tables->ucNumEntries; i++) {
+ record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ phm_ppt_v1_voltage_lookup_record,
+ entries, table, i);
+ atom_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_Voltage_Lookup_Record,
+ entries, vddc_lookup_pp_tables, i);
+ record->us_calculated = 0;
+ record->us_vdd = atom_record->usVdd;
+ record->us_cac_low = atom_record->usCACLow;
+ record->us_cac_mid = atom_record->usCACMid;
+ record->us_cac_high = atom_record->usCACHigh;
+ }
+
+ *lookup_table = table;
+
+ return 0;
+}
+
+/**
+ * Private Function used during initialization.
+ * Initialize Platform Power Management Parameter table
+ * @param hwmgr Pointer to the hardware manager.
+ * @param atom_ppm_table Pointer to PPM table in VBIOS
+ */
+static int get_platform_power_management_table(
+ struct pp_hwmgr *hwmgr,
+ ATOM_Tonga_PPM_Table *atom_ppm_table)
+{
+ struct phm_ppm_table *ptr = kzalloc(sizeof(ATOM_Tonga_PPM_Table), GFP_KERNEL);
+ struct phm_ppt_v1_information *pp_table_information =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+ if (NULL == ptr)
+ return -ENOMEM;
+
+ ptr->ppm_design
+ = atom_ppm_table->ucPpmDesign;
+ ptr->cpu_core_number
+ = atom_ppm_table->usCpuCoreNumber;
+ ptr->platform_tdp
+ = atom_ppm_table->ulPlatformTDP;
+ ptr->small_ac_platform_tdp
+ = atom_ppm_table->ulSmallACPlatformTDP;
+ ptr->platform_tdc
+ = atom_ppm_table->ulPlatformTDC;
+ ptr->small_ac_platform_tdc
+ = atom_ppm_table->ulSmallACPlatformTDC;
+ ptr->apu_tdp
+ = atom_ppm_table->ulApuTDP;
+ ptr->dgpu_tdp
+ = atom_ppm_table->ulDGpuTDP;
+ ptr->dgpu_ulv_power
+ = atom_ppm_table->ulDGpuUlvPower;
+ ptr->tj_max
+ = atom_ppm_table->ulTjmax;
+
+ pp_table_information->ppm_parameter_table = ptr;
+
+ return 0;
+}
+
+/**
+ * Private Function used during initialization.
+ * Initialize TDP limits for DPM2
+ * @param hwmgr Pointer to the hardware manager.
+ * @param powerplay_table Pointer to the PowerPlay Table.
+ */
+static int init_dpm_2_parameters(
+ struct pp_hwmgr *hwmgr,
+ const ATOM_Tonga_POWERPLAYTABLE *powerplay_table
+ )
+{
+ int result = 0;
+ struct phm_ppt_v1_information *pp_table_information = (struct phm_ppt_v1_information *)(hwmgr->pptable);
+ ATOM_Tonga_PPM_Table *atom_ppm_table;
+ uint32_t disable_ppm = 0;
+ uint32_t disable_power_control = 0;
+
+ pp_table_information->us_ulv_voltage_offset =
+ le16_to_cpu(powerplay_table->usUlvVoltageOffset);
+
+ pp_table_information->ppm_parameter_table = NULL;
+ pp_table_information->vddc_lookup_table = NULL;
+ pp_table_information->vddgfx_lookup_table = NULL;
+ /* TDP limits */
+ hwmgr->platform_descriptor.TDPODLimit =
+ le16_to_cpu(powerplay_table->usPowerControlLimit);
+ hwmgr->platform_descriptor.TDPAdjustment = 0;
+ hwmgr->platform_descriptor.VidAdjustment = 0;
+ hwmgr->platform_descriptor.VidAdjustmentPolarity = 0;
+ hwmgr->platform_descriptor.VidMinLimit = 0;
+ hwmgr->platform_descriptor.VidMaxLimit = 1500000;
+ hwmgr->platform_descriptor.VidStep = 6250;
+
+ disable_power_control = 0;
+ if (0 == disable_power_control) {
+ /* enable TDP overdrive (PowerControl) feature as well if supported */
+ if (hwmgr->platform_descriptor.TDPODLimit != 0)
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerControl);
+ }
+
+ if (0 != powerplay_table->usVddcLookupTableOffset) {
+ const ATOM_Tonga_Voltage_Lookup_Table *pVddcCACTable =
+ (ATOM_Tonga_Voltage_Lookup_Table *)(((unsigned long)powerplay_table) +
+ le16_to_cpu(powerplay_table->usVddcLookupTableOffset));
+
+ result = get_vddc_lookup_table(hwmgr,
+ &pp_table_information->vddc_lookup_table, pVddcCACTable, 16);
+ }
+
+ if (0 != powerplay_table->usVddgfxLookupTableOffset) {
+ const ATOM_Tonga_Voltage_Lookup_Table *pVddgfxCACTable =
+ (ATOM_Tonga_Voltage_Lookup_Table *)(((unsigned long)powerplay_table) +
+ le16_to_cpu(powerplay_table->usVddgfxLookupTableOffset));
+
+ result = get_vddc_lookup_table(hwmgr,
+ &pp_table_information->vddgfx_lookup_table, pVddgfxCACTable, 16);
+ }
+
+ disable_ppm = 0;
+ if (0 == disable_ppm) {
+ atom_ppm_table = (ATOM_Tonga_PPM_Table *)
+ (((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usPPMTableOffset));
+
+ if (0 != powerplay_table->usPPMTableOffset) {
+ if (get_platform_power_management_table(hwmgr, atom_ppm_table) == 0) {
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_EnablePlatformPowerManagement);
+ }
+ }
+ }
+
+ return result;
+}
+
+static int get_valid_clk(
+ struct pp_hwmgr *hwmgr,
+ struct phm_clock_array **clk_table,
+ phm_ppt_v1_clock_voltage_dependency_table const *clk_volt_pp_table
+ )
+{
+ uint32_t table_size, i;
+ struct phm_clock_array *table;
+ phm_ppt_v1_clock_voltage_dependency_record *dep_record;
+
+ PP_ASSERT_WITH_CODE((0 != clk_volt_pp_table->count),
+ "Invalid PowerPlay Table!", return -1);
+
+ table_size = sizeof(uint32_t) +
+ sizeof(uint32_t) * clk_volt_pp_table->count;
+
+ table = kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == table)
+ return -ENOMEM;
+
+ memset(table, 0x00, table_size);
+
+ table->count = (uint32_t)clk_volt_pp_table->count;
+
+ for (i = 0; i < table->count; i++) {
+ dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ phm_ppt_v1_clock_voltage_dependency_record,
+ entries, clk_volt_pp_table, i);
+ table->values[i] = (uint32_t)dep_record->clk;
+ }
+ *clk_table = table;
+
+ return 0;
+}
+
+static int get_hard_limits(
+ struct pp_hwmgr *hwmgr,
+ struct phm_clock_and_voltage_limits *limits,
+ ATOM_Tonga_Hard_Limit_Table const *limitable
+ )
+{
+ PP_ASSERT_WITH_CODE((0 != limitable->ucNumEntries), "Invalid PowerPlay Table!", return -1);
+
+ /* currently we always take entries[0] parameters */
+ limits->sclk = (uint32_t)limitable->entries[0].ulSCLKLimit;
+ limits->mclk = (uint32_t)limitable->entries[0].ulMCLKLimit;
+ limits->vddc = (uint16_t)limitable->entries[0].usVddcLimit;
+ limits->vddci = (uint16_t)limitable->entries[0].usVddciLimit;
+ limits->vddgfx = (uint16_t)limitable->entries[0].usVddgfxLimit;
+
+ return 0;
+}
+
+static int get_mclk_voltage_dependency_table(
+ struct pp_hwmgr *hwmgr,
+ phm_ppt_v1_clock_voltage_dependency_table **pp_tonga_mclk_dep_table,
+ ATOM_Tonga_MCLK_Dependency_Table const *mclk_dep_table
+ )
+{
+ uint32_t table_size, i;
+ phm_ppt_v1_clock_voltage_dependency_table *mclk_table;
+ phm_ppt_v1_clock_voltage_dependency_record *mclk_table_record;
+ ATOM_Tonga_MCLK_Dependency_Record *mclk_dep_record;
+
+ PP_ASSERT_WITH_CODE((0 != mclk_dep_table->ucNumEntries),
+ "Invalid PowerPlay Table!", return -1);
+
+ table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record)
+ * mclk_dep_table->ucNumEntries;
+
+ mclk_table = kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == mclk_table)
+ return -ENOMEM;
+
+ memset(mclk_table, 0x00, table_size);
+
+ mclk_table->count = (uint32_t)mclk_dep_table->ucNumEntries;
+
+ for (i = 0; i < mclk_dep_table->ucNumEntries; i++) {
+ mclk_table_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ phm_ppt_v1_clock_voltage_dependency_record,
+ entries, mclk_table, i);
+ mclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_MCLK_Dependency_Record,
+ entries, mclk_dep_table, i);
+ mclk_table_record->vddInd = mclk_dep_record->ucVddcInd;
+ mclk_table_record->vdd_offset = mclk_dep_record->usVddgfxOffset;
+ mclk_table_record->vddci = mclk_dep_record->usVddci;
+ mclk_table_record->mvdd = mclk_dep_record->usMvdd;
+ mclk_table_record->clk = mclk_dep_record->ulMclk;
+ }
+
+ *pp_tonga_mclk_dep_table = mclk_table;
+
+ return 0;
+}
+
+static int get_sclk_voltage_dependency_table(
+ struct pp_hwmgr *hwmgr,
+ phm_ppt_v1_clock_voltage_dependency_table **pp_tonga_sclk_dep_table,
+ PPTable_Generic_SubTable_Header const *sclk_dep_table
+ )
+{
+ uint32_t table_size, i;
+ phm_ppt_v1_clock_voltage_dependency_table *sclk_table;
+ phm_ppt_v1_clock_voltage_dependency_record *sclk_table_record;
+
+ if (sclk_dep_table->ucRevId < 1) {
+ const ATOM_Tonga_SCLK_Dependency_Table *tonga_table =
+ (ATOM_Tonga_SCLK_Dependency_Table *)sclk_dep_table;
+ ATOM_Tonga_SCLK_Dependency_Record *sclk_dep_record;
+
+ PP_ASSERT_WITH_CODE((0 != tonga_table->ucNumEntries),
+ "Invalid PowerPlay Table!", return -1);
+
+ table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record)
+ * tonga_table->ucNumEntries;
+
+ sclk_table = kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == sclk_table)
+ return -ENOMEM;
+
+ memset(sclk_table, 0x00, table_size);
+
+ sclk_table->count = (uint32_t)tonga_table->ucNumEntries;
+
+ for (i = 0; i < tonga_table->ucNumEntries; i++) {
+ sclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_SCLK_Dependency_Record,
+ entries, tonga_table, i);
+ sclk_table_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ phm_ppt_v1_clock_voltage_dependency_record,
+ entries, sclk_table, i);
+ sclk_table_record->vddInd = sclk_dep_record->ucVddInd;
+ sclk_table_record->vdd_offset = sclk_dep_record->usVddcOffset;
+ sclk_table_record->clk = sclk_dep_record->ulSclk;
+ sclk_table_record->cks_enable =
+ (((sclk_dep_record->ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0;
+ sclk_table_record->cks_voffset = (sclk_dep_record->ucCKSVOffsetandDisable & 0x7F);
+ }
+ } else {
+ const ATOM_Polaris_SCLK_Dependency_Table *polaris_table =
+ (ATOM_Polaris_SCLK_Dependency_Table *)sclk_dep_table;
+ ATOM_Polaris_SCLK_Dependency_Record *sclk_dep_record;
+
+ PP_ASSERT_WITH_CODE((0 != polaris_table->ucNumEntries),
+ "Invalid PowerPlay Table!", return -1);
+
+ table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record)
+ * polaris_table->ucNumEntries;
+
+ sclk_table = kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == sclk_table)
+ return -ENOMEM;
+
+ memset(sclk_table, 0x00, table_size);
+
+ sclk_table->count = (uint32_t)polaris_table->ucNumEntries;
+
+ for (i = 0; i < polaris_table->ucNumEntries; i++) {
+ sclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Polaris_SCLK_Dependency_Record,
+ entries, polaris_table, i);
+ sclk_table_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ phm_ppt_v1_clock_voltage_dependency_record,
+ entries, sclk_table, i);
+ sclk_table_record->vddInd = sclk_dep_record->ucVddInd;
+ sclk_table_record->vdd_offset = sclk_dep_record->usVddcOffset;
+ sclk_table_record->clk = sclk_dep_record->ulSclk;
+ sclk_table_record->cks_enable =
+ (((sclk_dep_record->ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0;
+ sclk_table_record->cks_voffset = (sclk_dep_record->ucCKSVOffsetandDisable & 0x7F);
+ sclk_table_record->sclk_offset = sclk_dep_record->ulSclkOffset;
+ }
+ }
+ *pp_tonga_sclk_dep_table = sclk_table;
+
+ return 0;
+}
+
+static int get_pcie_table(
+ struct pp_hwmgr *hwmgr,
+ phm_ppt_v1_pcie_table **pp_tonga_pcie_table,
+ PPTable_Generic_SubTable_Header const *ptable
+ )
+{
+ uint32_t table_size, i, pcie_count;
+ phm_ppt_v1_pcie_table *pcie_table;
+ struct phm_ppt_v1_information *pp_table_information =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+ phm_ppt_v1_pcie_record *pcie_record;
+
+ if (ptable->ucRevId < 1) {
+ const ATOM_Tonga_PCIE_Table *atom_pcie_table = (ATOM_Tonga_PCIE_Table *)ptable;
+ ATOM_Tonga_PCIE_Record *atom_pcie_record;
+
+ PP_ASSERT_WITH_CODE((atom_pcie_table->ucNumEntries != 0),
+ "Invalid PowerPlay Table!", return -1);
+
+ table_size = sizeof(uint32_t) +
+ sizeof(phm_ppt_v1_pcie_record) * atom_pcie_table->ucNumEntries;
+
+ pcie_table = kzalloc(table_size, GFP_KERNEL);
+
+ if (pcie_table == NULL)
+ return -ENOMEM;
+
+ memset(pcie_table, 0x00, table_size);
+
+ /*
+ * Make sure the number of pcie entries are less than or equal to sclk dpm levels.
+ * Since first PCIE entry is for ULV, #pcie has to be <= SclkLevel + 1.
+ */
+ pcie_count = (pp_table_information->vdd_dep_on_sclk->count) + 1;
+ if ((uint32_t)atom_pcie_table->ucNumEntries <= pcie_count)
+ pcie_count = (uint32_t)atom_pcie_table->ucNumEntries;
+ else
+ printk(KERN_ERR "[ powerplay ] Number of Pcie Entries exceed the number of SCLK Dpm Levels! \
+ Disregarding the excess entries... \n");
+
+ pcie_table->count = pcie_count;
+ for (i = 0; i < pcie_count; i++) {
+ pcie_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ phm_ppt_v1_pcie_record,
+ entries, pcie_table, i);
+ atom_pcie_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_PCIE_Record,
+ entries, atom_pcie_table, i);
+ pcie_record->gen_speed = atom_pcie_record->ucPCIEGenSpeed;
+ pcie_record->lane_width = atom_pcie_record->usPCIELaneWidth;
+ }
+
+ *pp_tonga_pcie_table = pcie_table;
+ } else {
+ /* Polaris10/Polaris11 and newer. */
+ const ATOM_Polaris10_PCIE_Table *atom_pcie_table = (ATOM_Polaris10_PCIE_Table *)ptable;
+ ATOM_Polaris10_PCIE_Record *atom_pcie_record;
+
+ PP_ASSERT_WITH_CODE((atom_pcie_table->ucNumEntries != 0),
+ "Invalid PowerPlay Table!", return -1);
+
+ table_size = sizeof(uint32_t) +
+ sizeof(phm_ppt_v1_pcie_record) * atom_pcie_table->ucNumEntries;
+
+ pcie_table = kzalloc(table_size, GFP_KERNEL);
+
+ if (pcie_table == NULL)
+ return -ENOMEM;
+
+ memset(pcie_table, 0x00, table_size);
+
+ /*
+ * Make sure the number of pcie entries are less than or equal to sclk dpm levels.
+ * Since first PCIE entry is for ULV, #pcie has to be <= SclkLevel + 1.
+ */
+ pcie_count = (pp_table_information->vdd_dep_on_sclk->count) + 1;
+ if ((uint32_t)atom_pcie_table->ucNumEntries <= pcie_count)
+ pcie_count = (uint32_t)atom_pcie_table->ucNumEntries;
+ else
+ printk(KERN_ERR "[ powerplay ] Number of Pcie Entries exceed the number of SCLK Dpm Levels! \
+ Disregarding the excess entries... \n");
+
+ pcie_table->count = pcie_count;
+
+ for (i = 0; i < pcie_count; i++) {
+ pcie_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ phm_ppt_v1_pcie_record,
+ entries, pcie_table, i);
+ atom_pcie_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Polaris10_PCIE_Record,
+ entries, atom_pcie_table, i);
+ pcie_record->gen_speed = atom_pcie_record->ucPCIEGenSpeed;
+ pcie_record->lane_width = atom_pcie_record->usPCIELaneWidth;
+ pcie_record->pcie_sclk = atom_pcie_record->ulPCIE_Sclk;
+ }
+
+ *pp_tonga_pcie_table = pcie_table;
+ }
+
+ return 0;
+}
+
+static int get_cac_tdp_table(
+ struct pp_hwmgr *hwmgr,
+ struct phm_cac_tdp_table **cac_tdp_table,
+ const PPTable_Generic_SubTable_Header * table
+ )
+{
+ uint32_t table_size;
+ struct phm_cac_tdp_table *tdp_table;
+
+ table_size = sizeof(uint32_t) + sizeof(struct phm_cac_tdp_table);
+ tdp_table = kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == tdp_table)
+ return -ENOMEM;
+
+ memset(tdp_table, 0x00, table_size);
+
+ hwmgr->dyn_state.cac_dtp_table = kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == hwmgr->dyn_state.cac_dtp_table) {
+ kfree(tdp_table);
+ return -ENOMEM;
+ }
+
+ memset(hwmgr->dyn_state.cac_dtp_table, 0x00, table_size);
+
+ if (table->ucRevId < 3) {
+ const ATOM_Tonga_PowerTune_Table *tonga_table =
+ (ATOM_Tonga_PowerTune_Table *)table;
+ tdp_table->usTDP = tonga_table->usTDP;
+ tdp_table->usConfigurableTDP =
+ tonga_table->usConfigurableTDP;
+ tdp_table->usTDC = tonga_table->usTDC;
+ tdp_table->usBatteryPowerLimit =
+ tonga_table->usBatteryPowerLimit;
+ tdp_table->usSmallPowerLimit =
+ tonga_table->usSmallPowerLimit;
+ tdp_table->usLowCACLeakage =
+ tonga_table->usLowCACLeakage;
+ tdp_table->usHighCACLeakage =
+ tonga_table->usHighCACLeakage;
+ tdp_table->usMaximumPowerDeliveryLimit =
+ tonga_table->usMaximumPowerDeliveryLimit;
+ tdp_table->usDefaultTargetOperatingTemp =
+ tonga_table->usTjMax;
+ tdp_table->usTargetOperatingTemp =
+ tonga_table->usTjMax; /*Set the initial temp to the same as default */
+ tdp_table->usPowerTuneDataSetID =
+ tonga_table->usPowerTuneDataSetID;
+ tdp_table->usSoftwareShutdownTemp =
+ tonga_table->usSoftwareShutdownTemp;
+ tdp_table->usClockStretchAmount =
+ tonga_table->usClockStretchAmount;
+ } else { /* Fiji and newer */
+ const ATOM_Fiji_PowerTune_Table *fijitable =
+ (ATOM_Fiji_PowerTune_Table *)table;
+ tdp_table->usTDP = fijitable->usTDP;
+ tdp_table->usConfigurableTDP = fijitable->usConfigurableTDP;
+ tdp_table->usTDC = fijitable->usTDC;
+ tdp_table->usBatteryPowerLimit = fijitable->usBatteryPowerLimit;
+ tdp_table->usSmallPowerLimit = fijitable->usSmallPowerLimit;
+ tdp_table->usLowCACLeakage = fijitable->usLowCACLeakage;
+ tdp_table->usHighCACLeakage = fijitable->usHighCACLeakage;
+ tdp_table->usMaximumPowerDeliveryLimit =
+ fijitable->usMaximumPowerDeliveryLimit;
+ tdp_table->usDefaultTargetOperatingTemp =
+ fijitable->usTjMax;
+ tdp_table->usTargetOperatingTemp =
+ fijitable->usTjMax; /*Set the initial temp to the same as default */
+ tdp_table->usPowerTuneDataSetID =
+ fijitable->usPowerTuneDataSetID;
+ tdp_table->usSoftwareShutdownTemp =
+ fijitable->usSoftwareShutdownTemp;
+ tdp_table->usClockStretchAmount =
+ fijitable->usClockStretchAmount;
+ tdp_table->usTemperatureLimitHotspot =
+ fijitable->usTemperatureLimitHotspot;
+ tdp_table->usTemperatureLimitLiquid1 =
+ fijitable->usTemperatureLimitLiquid1;
+ tdp_table->usTemperatureLimitLiquid2 =
+ fijitable->usTemperatureLimitLiquid2;
+ tdp_table->usTemperatureLimitVrVddc =
+ fijitable->usTemperatureLimitVrVddc;
+ tdp_table->usTemperatureLimitVrMvdd =
+ fijitable->usTemperatureLimitVrMvdd;
+ tdp_table->usTemperatureLimitPlx =
+ fijitable->usTemperatureLimitPlx;
+ tdp_table->ucLiquid1_I2C_address =
+ fijitable->ucLiquid1_I2C_address;
+ tdp_table->ucLiquid2_I2C_address =
+ fijitable->ucLiquid2_I2C_address;
+ tdp_table->ucLiquid_I2C_Line =
+ fijitable->ucLiquid_I2C_Line;
+ tdp_table->ucVr_I2C_address = fijitable->ucVr_I2C_address;
+ tdp_table->ucVr_I2C_Line = fijitable->ucVr_I2C_Line;
+ tdp_table->ucPlx_I2C_address = fijitable->ucPlx_I2C_address;
+ tdp_table->ucPlx_I2C_Line = fijitable->ucPlx_I2C_Line;
+ }
+
+ *cac_tdp_table = tdp_table;
+
+ return 0;
+}
+
+static int get_mm_clock_voltage_table(
+ struct pp_hwmgr *hwmgr,
+ phm_ppt_v1_mm_clock_voltage_dependency_table **tonga_mm_table,
+ const ATOM_Tonga_MM_Dependency_Table * mm_dependency_table
+ )
+{
+ uint32_t table_size, i;
+ const ATOM_Tonga_MM_Dependency_Record *mm_dependency_record;
+ phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table;
+ phm_ppt_v1_mm_clock_voltage_dependency_record *mm_table_record;
+
+ PP_ASSERT_WITH_CODE((0 != mm_dependency_table->ucNumEntries),
+ "Invalid PowerPlay Table!", return -1);
+ table_size = sizeof(uint32_t) +
+ sizeof(phm_ppt_v1_mm_clock_voltage_dependency_record)
+ * mm_dependency_table->ucNumEntries;
+ mm_table = kzalloc(table_size, GFP_KERNEL);
+
+ if (NULL == mm_table)
+ return -ENOMEM;
+
+ memset(mm_table, 0x00, table_size);
+
+ mm_table->count = mm_dependency_table->ucNumEntries;
+
+ for (i = 0; i < mm_dependency_table->ucNumEntries; i++) {
+ mm_dependency_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_MM_Dependency_Record,
+ entries, mm_dependency_table, i);
+ mm_table_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ phm_ppt_v1_mm_clock_voltage_dependency_record,
+ entries, mm_table, i);
+ mm_table_record->vddcInd = mm_dependency_record->ucVddcInd;
+ mm_table_record->vddgfx_offset = mm_dependency_record->usVddgfxOffset;
+ mm_table_record->aclk = mm_dependency_record->ulAClk;
+ mm_table_record->samclock = mm_dependency_record->ulSAMUClk;
+ mm_table_record->eclk = mm_dependency_record->ulEClk;
+ mm_table_record->vclk = mm_dependency_record->ulVClk;
+ mm_table_record->dclk = mm_dependency_record->ulDClk;
+ }
+
+ *tonga_mm_table = mm_table;
+
+ return 0;
+}
+
+/**
+ * Private Function used during initialization.
+ * Initialize clock voltage dependency
+ * @param hwmgr Pointer to the hardware manager.
+ * @param powerplay_table Pointer to the PowerPlay Table.
+ */
+static int init_clock_voltage_dependency(
+ struct pp_hwmgr *hwmgr,
+ const ATOM_Tonga_POWERPLAYTABLE *powerplay_table
+ )
+{
+ int result = 0;
+ struct phm_ppt_v1_information *pp_table_information =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+ const ATOM_Tonga_MM_Dependency_Table *mm_dependency_table =
+ (const ATOM_Tonga_MM_Dependency_Table *)(((unsigned long) powerplay_table) +
+ le16_to_cpu(powerplay_table->usMMDependencyTableOffset));
+ const PPTable_Generic_SubTable_Header *pPowerTuneTable =
+ (const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) +
+ le16_to_cpu(powerplay_table->usPowerTuneTableOffset));
+ const ATOM_Tonga_MCLK_Dependency_Table *mclk_dep_table =
+ (const ATOM_Tonga_MCLK_Dependency_Table *)(((unsigned long) powerplay_table) +
+ le16_to_cpu(powerplay_table->usMclkDependencyTableOffset));
+ const PPTable_Generic_SubTable_Header *sclk_dep_table =
+ (const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) +
+ le16_to_cpu(powerplay_table->usSclkDependencyTableOffset));
+ const ATOM_Tonga_Hard_Limit_Table *pHardLimits =
+ (const ATOM_Tonga_Hard_Limit_Table *)(((unsigned long) powerplay_table) +
+ le16_to_cpu(powerplay_table->usHardLimitTableOffset));
+ const PPTable_Generic_SubTable_Header *pcie_table =
+ (const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) +
+ le16_to_cpu(powerplay_table->usPCIETableOffset));
+
+ pp_table_information->vdd_dep_on_sclk = NULL;
+ pp_table_information->vdd_dep_on_mclk = NULL;
+ pp_table_information->mm_dep_table = NULL;
+ pp_table_information->pcie_table = NULL;
+
+ if (powerplay_table->usMMDependencyTableOffset != 0)
+ result = get_mm_clock_voltage_table(hwmgr,
+ &pp_table_information->mm_dep_table, mm_dependency_table);
+
+ if (result == 0 && powerplay_table->usPowerTuneTableOffset != 0)
+ result = get_cac_tdp_table(hwmgr,
+ &pp_table_information->cac_dtp_table, pPowerTuneTable);
+
+ if (result == 0 && powerplay_table->usSclkDependencyTableOffset != 0)
+ result = get_sclk_voltage_dependency_table(hwmgr,
+ &pp_table_information->vdd_dep_on_sclk, sclk_dep_table);
+
+ if (result == 0 && powerplay_table->usMclkDependencyTableOffset != 0)
+ result = get_mclk_voltage_dependency_table(hwmgr,
+ &pp_table_information->vdd_dep_on_mclk, mclk_dep_table);
+
+ if (result == 0 && powerplay_table->usPCIETableOffset != 0)
+ result = get_pcie_table(hwmgr,
+ &pp_table_information->pcie_table, pcie_table);
+
+ if (result == 0 && powerplay_table->usHardLimitTableOffset != 0)
+ result = get_hard_limits(hwmgr,
+ &pp_table_information->max_clock_voltage_on_dc, pHardLimits);
+
+ hwmgr->dyn_state.max_clock_voltage_on_dc.sclk =
+ pp_table_information->max_clock_voltage_on_dc.sclk;
+ hwmgr->dyn_state.max_clock_voltage_on_dc.mclk =
+ pp_table_information->max_clock_voltage_on_dc.mclk;
+ hwmgr->dyn_state.max_clock_voltage_on_dc.vddc =
+ pp_table_information->max_clock_voltage_on_dc.vddc;
+ hwmgr->dyn_state.max_clock_voltage_on_dc.vddci =
+ pp_table_information->max_clock_voltage_on_dc.vddci;
+
+ if (result == 0 && (NULL != pp_table_information->vdd_dep_on_mclk)
+ && (0 != pp_table_information->vdd_dep_on_mclk->count))
+ result = get_valid_clk(hwmgr, &pp_table_information->valid_mclk_values,
+ pp_table_information->vdd_dep_on_mclk);
+
+ if (result == 0 && (NULL != pp_table_information->vdd_dep_on_sclk)
+ && (0 != pp_table_information->vdd_dep_on_sclk->count))
+ result = get_valid_clk(hwmgr, &pp_table_information->valid_sclk_values,
+ pp_table_information->vdd_dep_on_sclk);
+
+ return result;
+}
+
+/** Retrieves the (signed) Overdrive limits from VBIOS.
+ * The max engine clock, memory clock and max temperature come from the firmware info table.
+ *
+ * The information is placed into the platform descriptor.
+ *
+ * @param hwmgr source of the VBIOS table and owner of the platform descriptor to be updated.
+ * @param powerplay_table the address of the PowerPlay table.
+ *
+ * @return 1 as long as the firmware info table was present and of a supported version.
+ */
+static int init_over_drive_limits(
+ struct pp_hwmgr *hwmgr,
+ const ATOM_Tonga_POWERPLAYTABLE *powerplay_table)
+{
+ hwmgr->platform_descriptor.overdriveLimit.engineClock =
+ le16_to_cpu(powerplay_table->ulMaxODEngineClock);
+ hwmgr->platform_descriptor.overdriveLimit.memoryClock =
+ le16_to_cpu(powerplay_table->ulMaxODMemoryClock);
+
+ hwmgr->platform_descriptor.minOverdriveVDDC = 0;
+ hwmgr->platform_descriptor.maxOverdriveVDDC = 0;
+ hwmgr->platform_descriptor.overdriveVDDCStep = 0;
+
+ if (hwmgr->platform_descriptor.overdriveLimit.engineClock > 0 \
+ && hwmgr->platform_descriptor.overdriveLimit.memoryClock > 0) {
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_ACOverdriveSupport);
+ }
+
+ return 0;
+}
+
+/**
+ * Private Function used during initialization.
+ * Inspect the PowerPlay table for obvious signs of corruption.
+ * @param hwmgr Pointer to the hardware manager.
+ * @param powerplay_table Pointer to the PowerPlay Table.
+ * @exception This implementation always returns 1.
+ */
+static int init_thermal_controller(
+ struct pp_hwmgr *hwmgr,
+ const ATOM_Tonga_POWERPLAYTABLE *powerplay_table
+ )
+{
+ const PPTable_Generic_SubTable_Header *fan_table;
+ ATOM_Tonga_Thermal_Controller *thermal_controller;
+
+ thermal_controller = (ATOM_Tonga_Thermal_Controller *)
+ (((unsigned long)powerplay_table) +
+ le16_to_cpu(powerplay_table->usThermalControllerOffset));
+ PP_ASSERT_WITH_CODE((0 != powerplay_table->usThermalControllerOffset),
+ "Thermal controller table not set!", return -1);
+
+ hwmgr->thermal_controller.ucType = thermal_controller->ucType;
+ hwmgr->thermal_controller.ucI2cLine = thermal_controller->ucI2cLine;
+ hwmgr->thermal_controller.ucI2cAddress = thermal_controller->ucI2cAddress;
+
+ hwmgr->thermal_controller.fanInfo.bNoFan =
+ (0 != (thermal_controller->ucFanParameters & ATOM_TONGA_PP_FANPARAMETERS_NOFAN));
+
+ hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution =
+ thermal_controller->ucFanParameters &
+ ATOM_TONGA_PP_FANPARAMETERS_TACHOMETER_PULSES_PER_REVOLUTION_MASK;
+
+ hwmgr->thermal_controller.fanInfo.ulMinRPM
+ = thermal_controller->ucFanMinRPM * 100UL;
+ hwmgr->thermal_controller.fanInfo.ulMaxRPM
+ = thermal_controller->ucFanMaxRPM * 100UL;
+
+ set_hw_cap(
+ hwmgr,
+ ATOM_TONGA_PP_THERMALCONTROLLER_NONE != hwmgr->thermal_controller.ucType,
+ PHM_PlatformCaps_ThermalController
+ );
+
+ if (0 == powerplay_table->usFanTableOffset)
+ return 0;
+
+ fan_table = (const PPTable_Generic_SubTable_Header *)
+ (((unsigned long)powerplay_table) +
+ le16_to_cpu(powerplay_table->usFanTableOffset));
+
+ PP_ASSERT_WITH_CODE((0 != powerplay_table->usFanTableOffset),
+ "Fan table not set!", return -1);
+ PP_ASSERT_WITH_CODE((0 < fan_table->ucRevId),
+ "Unsupported fan table format!", return -1);
+
+ hwmgr->thermal_controller.advanceFanControlParameters.ulCycleDelay
+ = 100000;
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_MicrocodeFanControl);
+
+ if (fan_table->ucRevId < 8) {
+ const ATOM_Tonga_Fan_Table *tonga_fan_table =
+ (ATOM_Tonga_Fan_Table *)fan_table;
+ hwmgr->thermal_controller.advanceFanControlParameters.ucTHyst
+ = tonga_fan_table->ucTHyst;
+ hwmgr->thermal_controller.advanceFanControlParameters.usTMin
+ = tonga_fan_table->usTMin;
+ hwmgr->thermal_controller.advanceFanControlParameters.usTMed
+ = tonga_fan_table->usTMed;
+ hwmgr->thermal_controller.advanceFanControlParameters.usTHigh
+ = tonga_fan_table->usTHigh;
+ hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin
+ = tonga_fan_table->usPWMMin;
+ hwmgr->thermal_controller.advanceFanControlParameters.usPWMMed
+ = tonga_fan_table->usPWMMed;
+ hwmgr->thermal_controller.advanceFanControlParameters.usPWMHigh
+ = tonga_fan_table->usPWMHigh;
+ hwmgr->thermal_controller.advanceFanControlParameters.usTMax
+ = 10900; /* hard coded */
+ hwmgr->thermal_controller.advanceFanControlParameters.usTMax
+ = tonga_fan_table->usTMax;
+ hwmgr->thermal_controller.advanceFanControlParameters.ucFanControlMode
+ = tonga_fan_table->ucFanControlMode;
+ hwmgr->thermal_controller.advanceFanControlParameters.usDefaultMaxFanPWM
+ = tonga_fan_table->usFanPWMMax;
+ hwmgr->thermal_controller.advanceFanControlParameters.usDefaultFanOutputSensitivity
+ = 4836;
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity
+ = tonga_fan_table->usFanOutputSensitivity;
+ hwmgr->thermal_controller.advanceFanControlParameters.usDefaultMaxFanRPM
+ = tonga_fan_table->usFanRPMMax;
+ hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit
+ = (tonga_fan_table->ulMinFanSCLKAcousticLimit / 100); /* PPTable stores it in 10Khz unit for 2 decimal places. SMC wants MHz. */
+ hwmgr->thermal_controller.advanceFanControlParameters.ucTargetTemperature
+ = tonga_fan_table->ucTargetTemperature;
+ hwmgr->thermal_controller.advanceFanControlParameters.ucMinimumPWMLimit
+ = tonga_fan_table->ucMinimumPWMLimit;
+ } else {
+ const ATOM_Fiji_Fan_Table *fiji_fan_table =
+ (ATOM_Fiji_Fan_Table *)fan_table;
+ hwmgr->thermal_controller.advanceFanControlParameters.ucTHyst
+ = fiji_fan_table->ucTHyst;
+ hwmgr->thermal_controller.advanceFanControlParameters.usTMin
+ = fiji_fan_table->usTMin;
+ hwmgr->thermal_controller.advanceFanControlParameters.usTMed
+ = fiji_fan_table->usTMed;
+ hwmgr->thermal_controller.advanceFanControlParameters.usTHigh
+ = fiji_fan_table->usTHigh;
+ hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin
+ = fiji_fan_table->usPWMMin;
+ hwmgr->thermal_controller.advanceFanControlParameters.usPWMMed
+ = fiji_fan_table->usPWMMed;
+ hwmgr->thermal_controller.advanceFanControlParameters.usPWMHigh
+ = fiji_fan_table->usPWMHigh;
+ hwmgr->thermal_controller.advanceFanControlParameters.usTMax
+ = fiji_fan_table->usTMax;
+ hwmgr->thermal_controller.advanceFanControlParameters.ucFanControlMode
+ = fiji_fan_table->ucFanControlMode;
+ hwmgr->thermal_controller.advanceFanControlParameters.usDefaultMaxFanPWM
+ = fiji_fan_table->usFanPWMMax;
+ hwmgr->thermal_controller.advanceFanControlParameters.usDefaultFanOutputSensitivity
+ = 4836;
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity
+ = fiji_fan_table->usFanOutputSensitivity;
+ hwmgr->thermal_controller.advanceFanControlParameters.usDefaultMaxFanRPM
+ = fiji_fan_table->usFanRPMMax;
+ hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit
+ = (fiji_fan_table->ulMinFanSCLKAcousticLimit / 100); /* PPTable stores it in 10Khz unit for 2 decimal places. SMC wants MHz. */
+ hwmgr->thermal_controller.advanceFanControlParameters.ucTargetTemperature
+ = fiji_fan_table->ucTargetTemperature;
+ hwmgr->thermal_controller.advanceFanControlParameters.ucMinimumPWMLimit
+ = fiji_fan_table->ucMinimumPWMLimit;
+
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanGainEdge
+ = fiji_fan_table->usFanGainEdge;
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHotspot
+ = fiji_fan_table->usFanGainHotspot;
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanGainLiquid
+ = fiji_fan_table->usFanGainLiquid;
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrVddc
+ = fiji_fan_table->usFanGainVrVddc;
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrMvdd
+ = fiji_fan_table->usFanGainVrMvdd;
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanGainPlx
+ = fiji_fan_table->usFanGainPlx;
+ hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHbm
+ = fiji_fan_table->usFanGainHbm;
+ }
+
+ return 0;
+}
+
+/**
+ * Private Function used during initialization.
+ * Inspect the PowerPlay table for obvious signs of corruption.
+ * @param hwmgr Pointer to the hardware manager.
+ * @param powerplay_table Pointer to the PowerPlay Table.
+ * @exception 2 if the powerplay table is incorrect.
+ */
+static int check_powerplay_tables(
+ struct pp_hwmgr *hwmgr,
+ const ATOM_Tonga_POWERPLAYTABLE *powerplay_table
+ )
+{
+ const ATOM_Tonga_State_Array *state_arrays;
+
+ state_arrays = (ATOM_Tonga_State_Array *)(((unsigned long)powerplay_table) +
+ le16_to_cpu(powerplay_table->usStateArrayOffset));
+
+ PP_ASSERT_WITH_CODE((ATOM_Tonga_TABLE_REVISION_TONGA <=
+ powerplay_table->sHeader.ucTableFormatRevision),
+ "Unsupported PPTable format!", return -1);
+ PP_ASSERT_WITH_CODE((0 != powerplay_table->usStateArrayOffset),
+ "State table is not set!", return -1);
+ PP_ASSERT_WITH_CODE((0 < powerplay_table->sHeader.usStructureSize),
+ "Invalid PowerPlay Table!", return -1);
+ PP_ASSERT_WITH_CODE((0 < state_arrays->ucNumEntries),
+ "Invalid PowerPlay Table!", return -1);
+
+ return 0;
+}
+
+int pp_tables_v1_0_initialize(struct pp_hwmgr *hwmgr)
+{
+ int result = 0;
+ const ATOM_Tonga_POWERPLAYTABLE *powerplay_table;
+
+ hwmgr->pptable = kzalloc(sizeof(struct phm_ppt_v1_information), GFP_KERNEL);
+
+ PP_ASSERT_WITH_CODE((NULL != hwmgr->pptable),
+ "Failed to allocate hwmgr->pptable!", return -ENOMEM);
+
+ memset(hwmgr->pptable, 0x00, sizeof(struct phm_ppt_v1_information));
+
+ powerplay_table = get_powerplay_table(hwmgr);
+
+ PP_ASSERT_WITH_CODE((NULL != powerplay_table),
+ "Missing PowerPlay Table!", return -1);
+
+ result = check_powerplay_tables(hwmgr, powerplay_table);
+
+ PP_ASSERT_WITH_CODE((result == 0),
+ "check_powerplay_tables failed", return result);
+
+ result = set_platform_caps(hwmgr,
+ le32_to_cpu(powerplay_table->ulPlatformCaps));
+
+ PP_ASSERT_WITH_CODE((result == 0),
+ "set_platform_caps failed", return result);
+
+ result = init_thermal_controller(hwmgr, powerplay_table);
+
+ PP_ASSERT_WITH_CODE((result == 0),
+ "init_thermal_controller failed", return result);
+
+ result = init_over_drive_limits(hwmgr, powerplay_table);
+
+ PP_ASSERT_WITH_CODE((result == 0),
+ "init_over_drive_limits failed", return result);
+
+ result = init_clock_voltage_dependency(hwmgr, powerplay_table);
+
+ PP_ASSERT_WITH_CODE((result == 0),
+ "init_clock_voltage_dependency failed", return result);
+
+ result = init_dpm_2_parameters(hwmgr, powerplay_table);
+
+ PP_ASSERT_WITH_CODE((result == 0),
+ "init_dpm_2_parameters failed", return result);
+
+ return result;
+}
+
+int pp_tables_v1_0_uninitialize(struct pp_hwmgr *hwmgr)
+{
+ struct phm_ppt_v1_information *pp_table_information =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+ kfree(pp_table_information->vdd_dep_on_sclk);
+ pp_table_information->vdd_dep_on_sclk = NULL;
+
+ kfree(pp_table_information->vdd_dep_on_mclk);
+ pp_table_information->vdd_dep_on_mclk = NULL;
+
+ kfree(pp_table_information->valid_mclk_values);
+ pp_table_information->valid_mclk_values = NULL;
+
+ kfree(pp_table_information->valid_sclk_values);
+ pp_table_information->valid_sclk_values = NULL;
+
+ kfree(pp_table_information->vddc_lookup_table);
+ pp_table_information->vddc_lookup_table = NULL;
+
+ kfree(pp_table_information->vddgfx_lookup_table);
+ pp_table_information->vddgfx_lookup_table = NULL;
+
+ kfree(pp_table_information->mm_dep_table);
+ pp_table_information->mm_dep_table = NULL;
+
+ kfree(pp_table_information->cac_dtp_table);
+ pp_table_information->cac_dtp_table = NULL;
+
+ kfree(hwmgr->dyn_state.cac_dtp_table);
+ hwmgr->dyn_state.cac_dtp_table = NULL;
+
+ kfree(pp_table_information->ppm_parameter_table);
+ pp_table_information->ppm_parameter_table = NULL;
+
+ kfree(pp_table_information->pcie_table);
+ pp_table_information->pcie_table = NULL;
+
+ kfree(hwmgr->pptable);
+ hwmgr->pptable = NULL;
+
+ return 0;
+}
+
+const struct pp_table_func pptable_v1_0_funcs = {
+ .pptable_init = pp_tables_v1_0_initialize,
+ .pptable_fini = pp_tables_v1_0_uninitialize,
+};
+
+int get_number_of_powerplay_table_entries_v1_0(struct pp_hwmgr *hwmgr)
+{
+ ATOM_Tonga_State_Array const *state_arrays;
+ const ATOM_Tonga_POWERPLAYTABLE *pp_table = get_powerplay_table(hwmgr);
+
+ PP_ASSERT_WITH_CODE((NULL != pp_table),
+ "Missing PowerPlay Table!", return -1);
+ PP_ASSERT_WITH_CODE((pp_table->sHeader.ucTableFormatRevision >=
+ ATOM_Tonga_TABLE_REVISION_TONGA),
+ "Incorrect PowerPlay table revision!", return -1);
+
+ state_arrays = (ATOM_Tonga_State_Array *)(((unsigned long)pp_table) +
+ le16_to_cpu(pp_table->usStateArrayOffset));
+
+ return (uint32_t)(state_arrays->ucNumEntries);
+}
+
+/**
+* Private function to convert flags stored in the BIOS to software flags in PowerPlay.
+*/
+static uint32_t make_classification_flags(struct pp_hwmgr *hwmgr,
+ uint16_t classification, uint16_t classification2)
+{
+ uint32_t result = 0;
+
+ if (classification & ATOM_PPLIB_CLASSIFICATION_BOOT)
+ result |= PP_StateClassificationFlag_Boot;
+
+ if (classification & ATOM_PPLIB_CLASSIFICATION_THERMAL)
+ result |= PP_StateClassificationFlag_Thermal;
+
+ if (classification & ATOM_PPLIB_CLASSIFICATION_LIMITEDPOWERSOURCE)
+ result |= PP_StateClassificationFlag_LimitedPowerSource;
+
+ if (classification & ATOM_PPLIB_CLASSIFICATION_REST)
+ result |= PP_StateClassificationFlag_Rest;
+
+ if (classification & ATOM_PPLIB_CLASSIFICATION_FORCED)
+ result |= PP_StateClassificationFlag_Forced;
+
+ if (classification & ATOM_PPLIB_CLASSIFICATION_ACPI)
+ result |= PP_StateClassificationFlag_ACPI;
+
+ if (classification2 & ATOM_PPLIB_CLASSIFICATION2_LIMITEDPOWERSOURCE_2)
+ result |= PP_StateClassificationFlag_LimitedPowerSource_2;
+
+ return result;
+}
+
+static int ppt_get_num_of_vce_state_table_entries_v1_0(struct pp_hwmgr *hwmgr)
+{
+ const ATOM_Tonga_POWERPLAYTABLE *pp_table = get_powerplay_table(hwmgr);
+ const ATOM_Tonga_VCE_State_Table *vce_state_table =
+ (ATOM_Tonga_VCE_State_Table *)(((unsigned long)pp_table) + le16_to_cpu(pp_table->usVCEStateTableOffset));
+
+ if (vce_state_table == NULL)
+ return 0;
+
+ return vce_state_table->ucNumEntries;
+}
+
+static int ppt_get_vce_state_table_entry_v1_0(struct pp_hwmgr *hwmgr, uint32_t i,
+ struct pp_vce_state *vce_state, void **clock_info, uint32_t *flag)
+{
+ const ATOM_Tonga_VCE_State_Record *vce_state_record;
+ ATOM_Tonga_SCLK_Dependency_Record *sclk_dep_record;
+ ATOM_Tonga_MCLK_Dependency_Record *mclk_dep_record;
+ ATOM_Tonga_MM_Dependency_Record *mm_dep_record;
+ const ATOM_Tonga_POWERPLAYTABLE *pptable = get_powerplay_table(hwmgr);
+ const ATOM_Tonga_VCE_State_Table *vce_state_table = (ATOM_Tonga_VCE_State_Table *)(((unsigned long)pptable)
+ + le16_to_cpu(pptable->usVCEStateTableOffset));
+ const ATOM_Tonga_SCLK_Dependency_Table *sclk_dep_table = (ATOM_Tonga_SCLK_Dependency_Table *)(((unsigned long)pptable)
+ + le16_to_cpu(pptable->usSclkDependencyTableOffset));
+ const ATOM_Tonga_MCLK_Dependency_Table *mclk_dep_table = (ATOM_Tonga_MCLK_Dependency_Table *)(((unsigned long)pptable)
+ + le16_to_cpu(pptable->usMclkDependencyTableOffset));
+ const ATOM_Tonga_MM_Dependency_Table *mm_dep_table = (ATOM_Tonga_MM_Dependency_Table *)(((unsigned long)pptable)
+ + le16_to_cpu(pptable->usMMDependencyTableOffset));
+
+ PP_ASSERT_WITH_CODE((i < vce_state_table->ucNumEntries),
+ "Requested state entry ID is out of range!",
+ return -EINVAL);
+
+ vce_state_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_VCE_State_Record,
+ entries, vce_state_table, i);
+ sclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_SCLK_Dependency_Record,
+ entries, sclk_dep_table,
+ vce_state_record->ucSCLKIndex);
+ mm_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_MM_Dependency_Record,
+ entries, mm_dep_table,
+ vce_state_record->ucVCEClockIndex);
+ *flag = vce_state_record->ucFlag;
+
+ vce_state->evclk = mm_dep_record->ulEClk;
+ vce_state->ecclk = mm_dep_record->ulEClk;
+ vce_state->sclk = sclk_dep_record->ulSclk;
+
+ if (vce_state_record->ucMCLKIndex >= mclk_dep_table->ucNumEntries)
+ mclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_MCLK_Dependency_Record,
+ entries, mclk_dep_table,
+ mclk_dep_table->ucNumEntries - 1);
+ else
+ mclk_dep_record = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_MCLK_Dependency_Record,
+ entries, mclk_dep_table,
+ vce_state_record->ucMCLKIndex);
+
+ vce_state->mclk = mclk_dep_record->ulMclk;
+ return 0;
+}
+
+/**
+* Create a Power State out of an entry in the PowerPlay table.
+* This function is called by the hardware back-end.
+* @param hwmgr Pointer to the hardware manager.
+* @param entry_index The index of the entry to be extracted from the table.
+* @param power_state The address of the PowerState instance being created.
+* @return -1 if the entry cannot be retrieved.
+*/
+int get_powerplay_table_entry_v1_0(struct pp_hwmgr *hwmgr,
+ uint32_t entry_index, struct pp_power_state *power_state,
+ int (*call_back_func)(struct pp_hwmgr *, void *,
+ struct pp_power_state *, void *, uint32_t))
+{
+ int result = 0;
+ const ATOM_Tonga_State_Array *state_arrays;
+ const ATOM_Tonga_State *state_entry;
+ const ATOM_Tonga_POWERPLAYTABLE *pp_table = get_powerplay_table(hwmgr);
+ int i, j;
+ uint32_t flags = 0;
+
+ PP_ASSERT_WITH_CODE((NULL != pp_table), "Missing PowerPlay Table!", return -1;);
+ power_state->classification.bios_index = entry_index;
+
+ if (pp_table->sHeader.ucTableFormatRevision >=
+ ATOM_Tonga_TABLE_REVISION_TONGA) {
+ state_arrays = (ATOM_Tonga_State_Array *)(((unsigned long)pp_table) +
+ le16_to_cpu(pp_table->usStateArrayOffset));
+
+ PP_ASSERT_WITH_CODE((0 < pp_table->usStateArrayOffset),
+ "Invalid PowerPlay Table State Array Offset.", return -1);
+ PP_ASSERT_WITH_CODE((0 < state_arrays->ucNumEntries),
+ "Invalid PowerPlay Table State Array.", return -1);
+ PP_ASSERT_WITH_CODE((entry_index <= state_arrays->ucNumEntries),
+ "Invalid PowerPlay Table State Array Entry.", return -1);
+
+ state_entry = GET_FLEXIBLE_ARRAY_MEMBER_ADDR(
+ ATOM_Tonga_State, entries,
+ state_arrays, entry_index);
+
+ result = call_back_func(hwmgr, (void *)state_entry, power_state,
+ (void *)pp_table,
+ make_classification_flags(hwmgr,
+ le16_to_cpu(state_entry->usClassification),
+ le16_to_cpu(state_entry->usClassification2)));
+ }
+
+ if (!result && (power_state->classification.flags &
+ PP_StateClassificationFlag_Boot))
+ result = hwmgr->hwmgr_func->patch_boot_state(hwmgr, &(power_state->hardware));
+
+ hwmgr->num_vce_state_tables = i = ppt_get_num_of_vce_state_table_entries_v1_0(hwmgr);
+
+ if ((i != 0) && (i <= PP_MAX_VCE_LEVELS)) {
+ for (j = 0; j < i; j++)
+ ppt_get_vce_state_table_entry_v1_0(hwmgr, j, &(hwmgr->vce_states[j]), NULL, &flags);
+ }
+
+ return result;
+}
+
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef _PROCESSPPTABLES_V1_0_H
+#define _PROCESSPPTABLES_V1_0_H
+
+#include "hwmgr.h"
+
+extern const struct pp_table_func pptable_v1_0_funcs;
+extern int get_number_of_powerplay_table_entries_v1_0(struct pp_hwmgr *hwmgr);
+extern int get_powerplay_table_entry_v1_0(struct pp_hwmgr *hwmgr, uint32_t entry_index,
+ struct pp_power_state *power_state, int (*call_back_func)(struct pp_hwmgr *, void *,
+ struct pp_power_state *, void *, uint32_t));
+
+#endif
+
int get_vce_state_table_entry(struct pp_hwmgr *hwmgr,
unsigned long i,
- struct PP_VCEState *vce_state,
+ struct pp_vce_state *vce_state,
void **clock_info,
unsigned long *flag)
{
#include "tonga_hwmgr.h"
#include "pptable.h"
#include "processpptables.h"
-#include "tonga_processpptables.h"
-#include "tonga_pptable.h"
+#include "process_pptables_v1_0.h"
+#include "pptable_v1_0.h"
#include "pp_debug.h"
#include "tonga_ppsmc.h"
#include "cgs_common.h"
return i - 1;
}
+
/**
* @brief PhwTonga_GetVoltageOrder
* Returns index of requested voltage record in lookup(table)
return i-1;
}
-bool tonga_is_dpm_running(struct pp_hwmgr *hwmgr)
+static bool tonga_is_dpm_running(struct pp_hwmgr *hwmgr)
{
/*
* We return the status of Voltage Control instead of checking SCLK/MCLK DPM
phm_cap_set(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_DisableMemoryTransition);
+ tonga_initialize_power_tune_defaults(hwmgr);
+
data->mclk_strobe_mode_threshold = 40000;
data->mclk_stutter_mode_threshold = 30000;
data->mclk_edc_enable_threshold = 40000;
}
-int tonga_update_sclk_threshold(struct pp_hwmgr *hwmgr)
+static int tonga_update_sclk_threshold(struct pp_hwmgr *hwmgr)
{
tonga_hwmgr *data = (tonga_hwmgr *)(hwmgr->backend);
* @param hwmgr the address of the powerplay hardware manager.
* @return always 0
*/
-int tonga_process_firmware_header(struct pp_hwmgr *hwmgr)
+static int tonga_process_firmware_header(struct pp_hwmgr *hwmgr)
{
tonga_hwmgr *data = (tonga_hwmgr *)(hwmgr->backend);
struct tonga_smumgr *tonga_smu = (struct tonga_smumgr *)(hwmgr->smumgr->backend);
return 0;
}
-/**
- * Convert a voltage value in mv unit to VID number required by SMU firmware
- */
-static uint8_t convert_to_vid(uint16_t vddc)
-{
- return (uint8_t) ((6200 - (vddc * VOLTAGE_SCALE)) / 25);
-}
-
-
/**
* Preparation of vddc and vddgfx CAC tables for SMC.
*
graphic_level->VoltageDownHyst = 0;
graphic_level->PowerThrottle = 0;
- threshold = engine_clock * data->fast_watemark_threshold / 100;
+ threshold = engine_clock * data->fast_watermark_threshold / 100;
/*
*get the DAL clock. do it in funture.
PECI_GetMinClockSettings(hwmgr->peci, &minClocks);
* @param pInput the pointer to input data (PowerState)
* @return always 0
*/
-int tonga_init_smc_table(struct pp_hwmgr *hwmgr)
+static int tonga_init_smc_table(struct pp_hwmgr *hwmgr)
{
int result;
tonga_hwmgr *data = (tonga_hwmgr *)(hwmgr->backend);
PP_ASSERT_WITH_CODE(0 == result,
"Failed to initialize Boot Level!", return result;);
+ result = tonga_populate_bapm_parameters_in_dpm_table(hwmgr);
+ PP_ASSERT_WITH_CODE(result == 0,
+ "Failed to populate BAPM Parameters!", return result);
+
if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
PHM_PlatformCaps_ClockStretcher)) {
result = tonga_populate_clock_stretcher_data_table(hwmgr);
return 0;
}
-int tonga_initialize_mc_reg_table(struct pp_hwmgr *hwmgr)
+static int tonga_initialize_mc_reg_table(struct pp_hwmgr *hwmgr)
{
int result;
tonga_hwmgr *data = (tonga_hwmgr *)(hwmgr->backend);
return 0;
}
+static void tonga_set_dpm_event_sources(struct pp_hwmgr *hwmgr, uint32_t sources)
+{
+ bool protection;
+ enum DPM_EVENT_SRC src;
+
+ switch (sources) {
+ default:
+ printk(KERN_ERR "Unknown throttling event sources.");
+ /* fall through */
+ case 0:
+ protection = false;
+ /* src is unused */
+ break;
+ case (1 << PHM_AutoThrottleSource_Thermal):
+ protection = true;
+ src = DPM_EVENT_SRC_DIGITAL;
+ break;
+ case (1 << PHM_AutoThrottleSource_External):
+ protection = true;
+ src = DPM_EVENT_SRC_EXTERNAL;
+ break;
+ case (1 << PHM_AutoThrottleSource_External) |
+ (1 << PHM_AutoThrottleSource_Thermal):
+ protection = true;
+ src = DPM_EVENT_SRC_DIGITAL_OR_EXTERNAL;
+ break;
+ }
+ /* Order matters - don't enable thermal protection for the wrong source. */
+ if (protection) {
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, CG_THERMAL_CTRL,
+ DPM_EVENT_SRC, src);
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT,
+ THERMAL_PROTECTION_DIS,
+ !phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_ThermalController));
+ } else
+ PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT,
+ THERMAL_PROTECTION_DIS, 1);
+}
+
+static int tonga_enable_auto_throttle_source(struct pp_hwmgr *hwmgr,
+ PHM_AutoThrottleSource source)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+
+ if (!(data->active_auto_throttle_sources & (1 << source))) {
+ data->active_auto_throttle_sources |= 1 << source;
+ tonga_set_dpm_event_sources(hwmgr, data->active_auto_throttle_sources);
+ }
+ return 0;
+}
+
+static int tonga_enable_thermal_auto_throttle(struct pp_hwmgr *hwmgr)
+{
+ return tonga_enable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal);
+}
+
+static int tonga_disable_auto_throttle_source(struct pp_hwmgr *hwmgr,
+ PHM_AutoThrottleSource source)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+
+ if (data->active_auto_throttle_sources & (1 << source)) {
+ data->active_auto_throttle_sources &= ~(1 << source);
+ tonga_set_dpm_event_sources(hwmgr, data->active_auto_throttle_sources);
+ }
+ return 0;
+}
+
+static int tonga_disable_thermal_auto_throttle(struct pp_hwmgr *hwmgr)
+{
+ return tonga_disable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal);
+}
int tonga_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
{
PP_ASSERT_WITH_CODE((0 == tmp_result),
"Failed to initialize ARB table index!", result = tmp_result);
+ tmp_result = tonga_populate_pm_fuses(hwmgr);
+ PP_ASSERT_WITH_CODE((tmp_result == 0),
+ "Failed to populate PM fuses!", result = tmp_result);
+
tmp_result = tonga_populate_initial_mc_reg_table(hwmgr);
PP_ASSERT_WITH_CODE((0 == tmp_result),
"Failed to populate initialize MC Reg table!", result = tmp_result);
PP_ASSERT_WITH_CODE((0 == tmp_result),
"Failed to start DPM!", result = tmp_result);
+ tmp_result = tonga_enable_smc_cac(hwmgr);
+ PP_ASSERT_WITH_CODE((tmp_result == 0),
+ "Failed to enable SMC CAC!", result = tmp_result);
+
+ tmp_result = tonga_enable_power_containment(hwmgr);
+ PP_ASSERT_WITH_CODE((tmp_result == 0),
+ "Failed to enable power containment!", result = tmp_result);
+
+ tmp_result = tonga_power_control_set_level(hwmgr);
+ PP_ASSERT_WITH_CODE((tmp_result == 0),
+ "Failed to power control set level!", result = tmp_result);
+
+ tmp_result = tonga_enable_thermal_auto_throttle(hwmgr);
+ PP_ASSERT_WITH_CODE((0 == tmp_result),
+ "Failed to enable thermal auto throttle!", result = tmp_result);
+
return result;
}
PP_ASSERT_WITH_CODE((0 == tmp_result),
"SMC is still running!", return 0);
+ tmp_result = tonga_disable_thermal_auto_throttle(hwmgr);
+ PP_ASSERT_WITH_CODE((tmp_result == 0),
+ "Failed to disable thermal auto throttle!", result = tmp_result);
+
tmp_result = tonga_stop_dpm(hwmgr);
PP_ASSERT_WITH_CODE((0 == tmp_result),
"Failed to stop DPM!", result = tmp_result);
tonga_ps = cast_phw_tonga_power_state(&(ps->hardware));
- result = tonga_get_powerplay_table_entry(hwmgr, entry_index, ps,
+ result = get_powerplay_table_entry_v1_0(hwmgr, entry_index, ps,
tonga_get_pp_table_entry_callback_func);
/* This is the earliest time we have all the dependency table and the VBIOS boot state
.get_sclk = tonga_dpm_get_sclk,
.patch_boot_state = tonga_dpm_patch_boot_state,
.get_pp_table_entry = tonga_get_pp_table_entry,
- .get_num_of_pp_table_entries = tonga_get_number_of_powerplay_table_entries,
+ .get_num_of_pp_table_entries = get_number_of_powerplay_table_entries_v1_0,
.print_current_perforce_level = tonga_print_current_perforce_level,
.powerdown_uvd = tonga_phm_powerdown_uvd,
.powergate_uvd = tonga_phm_powergate_uvd,
int tonga_hwmgr_init(struct pp_hwmgr *hwmgr)
{
hwmgr->hwmgr_func = &tonga_hwmgr_funcs;
- hwmgr->pptable_func = &tonga_pptable_funcs;
+ hwmgr->pptable_func = &pptable_v1_0_funcs;
pp_tonga_thermal_initialize(hwmgr);
return 0;
}
bool dll_defaule_on;
bool performance_request_registered;
+
/* ----------------- Low Power Features ---------------------*/
phw_tonga_bacos bacos;
phw_tonga_ulv_parm ulv;
bool enable_tdc_limit_feature;
bool enable_pkg_pwr_tracking_feature;
bool disable_uvd_power_tune_feature;
- phw_tonga_pt_defaults *power_tune_defaults;
+ struct tonga_pt_defaults *power_tune_defaults;
SMU72_Discrete_PmFuses power_tune_table;
- uint32_t ul_dte_tj_offset; /* Fudge factor in DPM table to correct HW DTE errors */
- uint32_t fast_watemark_threshold; /* use fast watermark if clock is equal or above this. In percentage of the target high sclk. */
+ uint32_t dte_tj_offset; /* Fudge factor in DPM table to correct HW DTE errors */
+ uint32_t fast_watermark_threshold; /* use fast watermark if clock is equal or above this. In percentage of the target high sclk. */
+
+
+ bool enable_dte_feature;
+
/* ----------------- Phase Shedding ---------------------*/
bool vddc_phase_shed_control;
--- /dev/null
+/*
+ * Copyright 2015 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "hwmgr.h"
+#include "smumgr.h"
+#include "tonga_hwmgr.h"
+#include "tonga_powertune.h"
+#include "tonga_smumgr.h"
+#include "smu72_discrete.h"
+#include "pp_debug.h"
+#include "tonga_ppsmc.h"
+
+#define VOLTAGE_SCALE 4
+#define POWERTUNE_DEFAULT_SET_MAX 1
+
+struct tonga_pt_defaults tonga_power_tune_data_set_array[POWERTUNE_DEFAULT_SET_MAX] = {
+/* sviLoadLIneEn, SviLoadLineVddC, TDC_VDDC_ThrottleReleaseLimitPerc, TDC_MAWt, TdcWaterfallCtl, DTEAmbientTempBase, DisplayCac, BAPM_TEMP_GRADIENT */
+ {1, 0xF, 0xFD, 0x19, 5, 45, 0, 0xB0000,
+ {0x79, 0x253, 0x25D, 0xAE, 0x72, 0x80, 0x83, 0x86, 0x6F, 0xC8, 0xC9, 0xC9, 0x2F, 0x4D, 0x61},
+ {0x17C, 0x172, 0x180, 0x1BC, 0x1B3, 0x1BD, 0x206, 0x200, 0x203, 0x25D, 0x25A, 0x255, 0x2C3, 0x2C5, 0x2B4 } },
+};
+
+void tonga_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *tonga_hwmgr = (struct tonga_hwmgr *)(hwmgr->backend);
+ struct phm_ppt_v1_information *table_info =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+ uint32_t tmp = 0;
+
+ if (table_info &&
+ table_info->cac_dtp_table->usPowerTuneDataSetID <= POWERTUNE_DEFAULT_SET_MAX &&
+ table_info->cac_dtp_table->usPowerTuneDataSetID)
+ tonga_hwmgr->power_tune_defaults =
+ &tonga_power_tune_data_set_array
+ [table_info->cac_dtp_table->usPowerTuneDataSetID - 1];
+ else
+ tonga_hwmgr->power_tune_defaults = &tonga_power_tune_data_set_array[0];
+
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_CAC);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_SQRamping);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_DBRamping);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_TDRamping);
+ phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_TCPRamping);
+
+ tonga_hwmgr->dte_tj_offset = tmp;
+
+ if (!tmp) {
+ phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_CAC);
+
+ tonga_hwmgr->fast_watermark_threshold = 100;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
+ tmp = 1;
+ tonga_hwmgr->enable_dte_feature = tmp ? false : true;
+ tonga_hwmgr->enable_tdc_limit_feature = tmp ? true : false;
+ tonga_hwmgr->enable_pkg_pwr_tracking_feature = tmp ? true : false;
+ }
+ }
+}
+
+
+int tonga_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ struct tonga_pt_defaults *defaults = data->power_tune_defaults;
+ SMU72_Discrete_DpmTable *dpm_table = &(data->smc_state_table);
+ struct phm_ppt_v1_information *table_info =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+ struct phm_cac_tdp_table *cac_dtp_table = table_info->cac_dtp_table;
+ int i, j, k;
+ uint16_t *pdef1;
+ uint16_t *pdef2;
+
+
+ /* TDP number of fraction bits are changed from 8 to 7 for Fiji
+ * as requested by SMC team
+ */
+ dpm_table->DefaultTdp = PP_HOST_TO_SMC_US(
+ (uint16_t)(cac_dtp_table->usTDP * 256));
+ dpm_table->TargetTdp = PP_HOST_TO_SMC_US(
+ (uint16_t)(cac_dtp_table->usConfigurableTDP * 256));
+
+ PP_ASSERT_WITH_CODE(cac_dtp_table->usTargetOperatingTemp <= 255,
+ "Target Operating Temp is out of Range!",
+ );
+
+ dpm_table->GpuTjMax = (uint8_t)(cac_dtp_table->usTargetOperatingTemp);
+ dpm_table->GpuTjHyst = 8;
+
+ dpm_table->DTEAmbientTempBase = defaults->dte_ambient_temp_base;
+
+ dpm_table->BAPM_TEMP_GRADIENT = PP_HOST_TO_SMC_UL(defaults->bamp_temp_gradient);
+ pdef1 = defaults->bapmti_r;
+ pdef2 = defaults->bapmti_rc;
+
+ for (i = 0; i < SMU72_DTE_ITERATIONS; i++) {
+ for (j = 0; j < SMU72_DTE_SOURCES; j++) {
+ for (k = 0; k < SMU72_DTE_SINKS; k++) {
+ dpm_table->BAPMTI_R[i][j][k] = PP_HOST_TO_SMC_US(*pdef1);
+ dpm_table->BAPMTI_RC[i][j][k] = PP_HOST_TO_SMC_US(*pdef2);
+ pdef1++;
+ pdef2++;
+ }
+ }
+ }
+
+ return 0;
+}
+
+static int tonga_populate_svi_load_line(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ const struct tonga_pt_defaults *defaults = data->power_tune_defaults;
+
+ data->power_tune_table.SviLoadLineEn = defaults->svi_load_line_en;
+ data->power_tune_table.SviLoadLineVddC = defaults->svi_load_line_vddC;
+ data->power_tune_table.SviLoadLineTrimVddC = 3;
+ data->power_tune_table.SviLoadLineOffsetVddC = 0;
+
+ return 0;
+}
+
+static int tonga_populate_tdc_limit(struct pp_hwmgr *hwmgr)
+{
+ uint16_t tdc_limit;
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ struct phm_ppt_v1_information *table_info =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+ const struct tonga_pt_defaults *defaults = data->power_tune_defaults;
+
+ /* TDC number of fraction bits are changed from 8 to 7
+ * for Fiji as requested by SMC team
+ */
+ tdc_limit = (uint16_t)(table_info->cac_dtp_table->usTDC * 256);
+ data->power_tune_table.TDC_VDDC_PkgLimit =
+ CONVERT_FROM_HOST_TO_SMC_US(tdc_limit);
+ data->power_tune_table.TDC_VDDC_ThrottleReleaseLimitPerc =
+ defaults->tdc_vddc_throttle_release_limit_perc;
+ data->power_tune_table.TDC_MAWt = defaults->tdc_mawt;
+
+ return 0;
+}
+
+static int tonga_populate_dw8(struct pp_hwmgr *hwmgr, uint32_t fuse_table_offset)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ const struct tonga_pt_defaults *defaults = data->power_tune_defaults;
+ uint32_t temp;
+
+ if (tonga_read_smc_sram_dword(hwmgr->smumgr,
+ fuse_table_offset +
+ offsetof(SMU72_Discrete_PmFuses, TdcWaterfallCtl),
+ (uint32_t *)&temp, data->sram_end))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to read PmFuses.DW6 (SviLoadLineEn) from SMC Failed!",
+ return -EINVAL);
+ else
+ data->power_tune_table.TdcWaterfallCtl = defaults->tdc_waterfall_ctl;
+
+ return 0;
+}
+
+static int tonga_populate_temperature_scaler(struct pp_hwmgr *hwmgr)
+{
+ int i;
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+
+ /* Currently not used. Set all to zero. */
+ for (i = 0; i < 16; i++)
+ data->power_tune_table.LPMLTemperatureScaler[i] = 0;
+
+ return 0;
+}
+
+static int tonga_populate_fuzzy_fan(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+
+ if ((hwmgr->thermal_controller.advanceFanControlParameters.
+ usFanOutputSensitivity & (1 << 15)) ||
+ (hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity == 0))
+ hwmgr->thermal_controller.advanceFanControlParameters.
+ usFanOutputSensitivity = hwmgr->thermal_controller.
+ advanceFanControlParameters.usDefaultFanOutputSensitivity;
+
+ data->power_tune_table.FuzzyFan_PwmSetDelta =
+ PP_HOST_TO_SMC_US(hwmgr->thermal_controller.
+ advanceFanControlParameters.usFanOutputSensitivity);
+ return 0;
+}
+
+static int tonga_populate_gnb_lpml(struct pp_hwmgr *hwmgr)
+{
+ int i;
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+
+ /* Currently not used. Set all to zero. */
+ for (i = 0; i < 16; i++)
+ data->power_tune_table.GnbLPML[i] = 0;
+
+ return 0;
+}
+
+static int tonga_min_max_vgnb_lpml_id_from_bapm_vddc(struct pp_hwmgr *hwmgr)
+{
+ return 0;
+}
+
+static int tonga_populate_bapm_vddc_base_leakage_sidd(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ struct phm_ppt_v1_information *table_info =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+ uint16_t hi_sidd = data->power_tune_table.BapmVddCBaseLeakageHiSidd;
+ uint16_t lo_sidd = data->power_tune_table.BapmVddCBaseLeakageLoSidd;
+ struct phm_cac_tdp_table *cac_table = table_info->cac_dtp_table;
+
+ hi_sidd = (uint16_t)(cac_table->usHighCACLeakage / 100 * 256);
+ lo_sidd = (uint16_t)(cac_table->usLowCACLeakage / 100 * 256);
+
+ data->power_tune_table.BapmVddCBaseLeakageHiSidd =
+ CONVERT_FROM_HOST_TO_SMC_US(hi_sidd);
+ data->power_tune_table.BapmVddCBaseLeakageLoSidd =
+ CONVERT_FROM_HOST_TO_SMC_US(lo_sidd);
+
+ return 0;
+}
+
+int tonga_populate_pm_fuses(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ uint32_t pm_fuse_table_offset;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
+ if (tonga_read_smc_sram_dword(hwmgr->smumgr,
+ SMU72_FIRMWARE_HEADER_LOCATION +
+ offsetof(SMU72_Firmware_Header, PmFuseTable),
+ &pm_fuse_table_offset, data->sram_end))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to get pm_fuse_table_offset Failed!",
+ return -EINVAL);
+
+ /* DW6 */
+ if (tonga_populate_svi_load_line(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate SviLoadLine Failed!",
+ return -EINVAL);
+ /* DW7 */
+ if (tonga_populate_tdc_limit(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate TDCLimit Failed!", return -EINVAL);
+ /* DW8 */
+ if (tonga_populate_dw8(hwmgr, pm_fuse_table_offset))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate TdcWaterfallCtl Failed !",
+ return -EINVAL);
+
+ /* DW9-DW12 */
+ if (tonga_populate_temperature_scaler(hwmgr) != 0)
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate LPMLTemperatureScaler Failed!",
+ return -EINVAL);
+
+ /* DW13-DW14 */
+ if (tonga_populate_fuzzy_fan(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate Fuzzy Fan Control parameters Failed!",
+ return -EINVAL);
+
+ /* DW15-DW18 */
+ if (tonga_populate_gnb_lpml(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate GnbLPML Failed!",
+ return -EINVAL);
+
+ /* DW19 */
+ if (tonga_min_max_vgnb_lpml_id_from_bapm_vddc(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate GnbLPML Min and Max Vid Failed!",
+ return -EINVAL);
+
+ /* DW20 */
+ if (tonga_populate_bapm_vddc_base_leakage_sidd(hwmgr))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to populate BapmVddCBaseLeakage Hi and Lo Sidd Failed!",
+ return -EINVAL);
+
+ if (tonga_copy_bytes_to_smc(hwmgr->smumgr, pm_fuse_table_offset,
+ (uint8_t *)&data->power_tune_table,
+ sizeof(struct SMU72_Discrete_PmFuses), data->sram_end))
+ PP_ASSERT_WITH_CODE(false,
+ "Attempt to download PmFuseTable Failed!",
+ return -EINVAL);
+ }
+ return 0;
+}
+
+int tonga_enable_smc_cac(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ int result = 0;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_CAC)) {
+ int smc_result;
+
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_EnableCac));
+ PP_ASSERT_WITH_CODE((smc_result == 0),
+ "Failed to enable CAC in SMC.", result = -1);
+
+ data->cac_enabled = (smc_result == 0) ? true : false;
+ }
+ return result;
+}
+
+int tonga_disable_smc_cac(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ int result = 0;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_CAC) && data->cac_enabled) {
+ int smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_DisableCac));
+ PP_ASSERT_WITH_CODE((smc_result == 0),
+ "Failed to disable CAC in SMC.", result = -1);
+
+ data->cac_enabled = false;
+ }
+ return result;
+}
+
+int tonga_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+
+ if (data->power_containment_features &
+ POWERCONTAINMENT_FEATURE_PkgPwrLimit)
+ return smum_send_msg_to_smc_with_parameter(hwmgr->smumgr,
+ PPSMC_MSG_PkgPwrSetLimit, n);
+ return 0;
+}
+
+static int tonga_set_overdriver_target_tdp(struct pp_hwmgr *pHwMgr, uint32_t target_tdp)
+{
+ return smum_send_msg_to_smc_with_parameter(pHwMgr->smumgr,
+ PPSMC_MSG_OverDriveSetTargetTdp, target_tdp);
+}
+
+int tonga_enable_power_containment(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ struct phm_ppt_v1_information *table_info =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+ int smc_result;
+ int result = 0;
+
+ data->power_containment_features = 0;
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
+ if (data->enable_dte_feature) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_EnableDTE));
+ PP_ASSERT_WITH_CODE((smc_result == 0),
+ "Failed to enable DTE in SMC.", result = -1;);
+ if (smc_result == 0)
+ data->power_containment_features |= POWERCONTAINMENT_FEATURE_DTE;
+ }
+
+ if (data->enable_tdc_limit_feature) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_TDCLimitEnable));
+ PP_ASSERT_WITH_CODE((smc_result == 0),
+ "Failed to enable TDCLimit in SMC.", result = -1;);
+ if (smc_result == 0)
+ data->power_containment_features |=
+ POWERCONTAINMENT_FEATURE_TDCLimit;
+ }
+
+ if (data->enable_pkg_pwr_tracking_feature) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_PkgPwrLimitEnable));
+ PP_ASSERT_WITH_CODE((smc_result == 0),
+ "Failed to enable PkgPwrTracking in SMC.", result = -1;);
+ if (smc_result == 0) {
+ struct phm_cac_tdp_table *cac_table =
+ table_info->cac_dtp_table;
+ uint32_t default_limit =
+ (uint32_t)(cac_table->usMaximumPowerDeliveryLimit * 256);
+
+ data->power_containment_features |=
+ POWERCONTAINMENT_FEATURE_PkgPwrLimit;
+
+ if (tonga_set_power_limit(hwmgr, default_limit))
+ printk(KERN_ERR "Failed to set Default Power Limit in SMC!");
+ }
+ }
+ }
+ return result;
+}
+
+int tonga_disable_power_containment(struct pp_hwmgr *hwmgr)
+{
+ struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend);
+ int result = 0;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment) &&
+ data->power_containment_features) {
+ int smc_result;
+
+ if (data->power_containment_features &
+ POWERCONTAINMENT_FEATURE_TDCLimit) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_TDCLimitDisable));
+ PP_ASSERT_WITH_CODE((smc_result == 0),
+ "Failed to disable TDCLimit in SMC.",
+ result = smc_result);
+ }
+
+ if (data->power_containment_features &
+ POWERCONTAINMENT_FEATURE_DTE) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_DisableDTE));
+ PP_ASSERT_WITH_CODE((smc_result == 0),
+ "Failed to disable DTE in SMC.",
+ result = smc_result);
+ }
+
+ if (data->power_containment_features &
+ POWERCONTAINMENT_FEATURE_PkgPwrLimit) {
+ smc_result = smum_send_msg_to_smc(hwmgr->smumgr,
+ (uint16_t)(PPSMC_MSG_PkgPwrLimitDisable));
+ PP_ASSERT_WITH_CODE((smc_result == 0),
+ "Failed to disable PkgPwrTracking in SMC.",
+ result = smc_result);
+ }
+ data->power_containment_features = 0;
+ }
+
+ return result;
+}
+
+int tonga_power_control_set_level(struct pp_hwmgr *hwmgr)
+{
+ struct phm_ppt_v1_information *table_info =
+ (struct phm_ppt_v1_information *)(hwmgr->pptable);
+ struct phm_cac_tdp_table *cac_table = table_info->cac_dtp_table;
+ int adjust_percent, target_tdp;
+ int result = 0;
+
+ if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+ PHM_PlatformCaps_PowerContainment)) {
+ /* adjustment percentage has already been validated */
+ adjust_percent = hwmgr->platform_descriptor.TDPAdjustmentPolarity ?
+ hwmgr->platform_descriptor.TDPAdjustment :
+ (-1 * hwmgr->platform_descriptor.TDPAdjustment);
+ /* SMC requested that target_tdp to be 7 bit fraction in DPM table
+ * but message to be 8 bit fraction for messages
+ */
+ target_tdp = ((100 + adjust_percent) * (int)(cac_table->usTDP * 256)) / 100;
+ result = tonga_set_overdriver_target_tdp(hwmgr, (uint32_t)target_tdp);
+ }
+
+ return result;
+}
};
typedef enum _phw_tonga_ptc_config_reg_type phw_tonga_ptc_config_reg_type;
+/* PowerContainment Features */
+#define POWERCONTAINMENT_FEATURE_DTE 0x00000001
+
+
/* PowerContainment Features */
#define POWERCONTAINMENT_FEATURE_BAPM 0x00000001
#define POWERCONTAINMENT_FEATURE_TDCLimit 0x00000002
#define POWERCONTAINMENT_FEATURE_PkgPwrLimit 0x00000004
-struct _phw_tonga_pt_config_reg {
+struct tonga_pt_config_reg {
uint32_t Offset;
uint32_t Mask;
uint32_t Shift;
uint32_t Value;
phw_tonga_ptc_config_reg_type Type;
};
-typedef struct _phw_tonga_pt_config_reg phw_tonga_pt_config_reg;
-struct _phw_tonga_pt_defaults {
+struct tonga_pt_defaults {
uint8_t svi_load_line_en;
uint8_t svi_load_line_vddC;
uint8_t tdc_vddc_throttle_release_limit_perc;
uint16_t bapmti_r[SMU72_DTE_ITERATIONS * SMU72_DTE_SOURCES * SMU72_DTE_SINKS];
uint16_t bapmti_rc[SMU72_DTE_ITERATIONS * SMU72_DTE_SOURCES * SMU72_DTE_SINKS];
};
-typedef struct _phw_tonga_pt_defaults phw_tonga_pt_defaults;
+
+
+
+void tonga_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr);
+int tonga_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr);
+int tonga_populate_pm_fuses(struct pp_hwmgr *hwmgr);
+int tonga_enable_smc_cac(struct pp_hwmgr *hwmgr);
+int tonga_disable_smc_cac(struct pp_hwmgr *hwmgr);
+int tonga_enable_power_containment(struct pp_hwmgr *hwmgr);
+int tonga_disable_power_containment(struct pp_hwmgr *hwmgr);
+int tonga_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n);
+int tonga_power_control_set_level(struct pp_hwmgr *hwmgr);
#endif
+++ /dev/null
-/*
- * Copyright 2015 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- */
-
-#ifndef TONGA_PPTABLE_H
-#define TONGA_PPTABLE_H
-
-/** \file
- * This is a PowerPlay table header file
- */
-#pragma pack(push, 1)
-
-#include "hwmgr.h"
-
-#define ATOM_TONGA_PP_FANPARAMETERS_TACHOMETER_PULSES_PER_REVOLUTION_MASK 0x0f
-#define ATOM_TONGA_PP_FANPARAMETERS_NOFAN 0x80 /* No fan is connected to this controller. */
-
-#define ATOM_TONGA_PP_THERMALCONTROLLER_NONE 0
-#define ATOM_TONGA_PP_THERMALCONTROLLER_LM96163 17
-#define ATOM_TONGA_PP_THERMALCONTROLLER_TONGA 21
-#define ATOM_TONGA_PP_THERMALCONTROLLER_FIJI 22
-
-/*
- * Thermal controller 'combo type' to use an external controller for Fan control and an internal controller for thermal.
- * We probably should reserve the bit 0x80 for this use.
- * To keep the number of these types low we should also use the same code for all ASICs (i.e. do not distinguish RV6xx and RV7xx Internal here).
- * The driver can pick the correct internal controller based on the ASIC.
- */
-
-#define ATOM_TONGA_PP_THERMALCONTROLLER_ADT7473_WITH_INTERNAL 0x89 /* ADT7473 Fan Control + Internal Thermal Controller */
-#define ATOM_TONGA_PP_THERMALCONTROLLER_EMC2103_WITH_INTERNAL 0x8D /* EMC2103 Fan Control + Internal Thermal Controller */
-
-/*/* ATOM_TONGA_POWERPLAYTABLE::ulPlatformCaps */
-#define ATOM_TONGA_PP_PLATFORM_CAP_VDDGFX_CONTROL 0x1 /* This cap indicates whether vddgfx will be a separated power rail. */
-#define ATOM_TONGA_PP_PLATFORM_CAP_POWERPLAY 0x2 /* This cap indicates whether this is a mobile part and CCC need to show Powerplay page. */
-#define ATOM_TONGA_PP_PLATFORM_CAP_SBIOSPOWERSOURCE 0x4 /* This cap indicates whether power source notificaiton is done by SBIOS directly. */
-#define ATOM_TONGA_PP_PLATFORM_CAP_DISABLE_VOLTAGE_ISLAND 0x8 /* Enable the option to overwrite voltage island feature to be disabled, regardless of VddGfx power rail support. */
-#define ____RETIRE16____ 0x10
-#define ATOM_TONGA_PP_PLATFORM_CAP_HARDWAREDC 0x20 /* This cap indicates whether power source notificaiton is done by GPIO directly. */
-#define ____RETIRE64____ 0x40
-#define ____RETIRE128____ 0x80
-#define ____RETIRE256____ 0x100
-#define ____RETIRE512____ 0x200
-#define ____RETIRE1024____ 0x400
-#define ____RETIRE2048____ 0x800
-#define ATOM_TONGA_PP_PLATFORM_CAP_MVDD_CONTROL 0x1000 /* This cap indicates dynamic MVDD is required. Uncheck to disable it. */
-#define ____RETIRE2000____ 0x2000
-#define ____RETIRE4000____ 0x4000
-#define ATOM_TONGA_PP_PLATFORM_CAP_VDDCI_CONTROL 0x8000 /* This cap indicates dynamic VDDCI is required. Uncheck to disable it. */
-#define ____RETIRE10000____ 0x10000
-#define ATOM_TONGA_PP_PLATFORM_CAP_BACO 0x20000 /* Enable to indicate the driver supports BACO state. */
-
-#define ATOM_TONGA_PP_PLATFORM_CAP_OUTPUT_THERMAL2GPIO17 0x100000 /* Enable to indicate the driver supports thermal2GPIO17. */
-#define ATOM_TONGA_PP_PLATFORM_COMBINE_PCC_WITH_THERMAL_SIGNAL 0x1000000 /* Enable to indicate if thermal and PCC are sharing the same GPIO */
-#define ATOM_TONGA_PLATFORM_LOAD_POST_PRODUCTION_FIRMWARE 0x2000000
-
-/* ATOM_PPLIB_NONCLOCK_INFO::usClassification */
-#define ATOM_PPLIB_CLASSIFICATION_UI_MASK 0x0007
-#define ATOM_PPLIB_CLASSIFICATION_UI_SHIFT 0
-#define ATOM_PPLIB_CLASSIFICATION_UI_NONE 0
-#define ATOM_PPLIB_CLASSIFICATION_UI_BATTERY 1
-#define ATOM_PPLIB_CLASSIFICATION_UI_BALANCED 3
-#define ATOM_PPLIB_CLASSIFICATION_UI_PERFORMANCE 5
-/* 2, 4, 6, 7 are reserved */
-
-#define ATOM_PPLIB_CLASSIFICATION_BOOT 0x0008
-#define ATOM_PPLIB_CLASSIFICATION_THERMAL 0x0010
-#define ATOM_PPLIB_CLASSIFICATION_LIMITEDPOWERSOURCE 0x0020
-#define ATOM_PPLIB_CLASSIFICATION_REST 0x0040
-#define ATOM_PPLIB_CLASSIFICATION_FORCED 0x0080
-#define ATOM_PPLIB_CLASSIFICATION_ACPI 0x1000
-
-/* ATOM_PPLIB_NONCLOCK_INFO::usClassification2 */
-#define ATOM_PPLIB_CLASSIFICATION2_LIMITEDPOWERSOURCE_2 0x0001
-
-#define ATOM_Tonga_DISALLOW_ON_DC 0x00004000
-#define ATOM_Tonga_ENABLE_VARIBRIGHT 0x00008000
-
-#define ATOM_Tonga_TABLE_REVISION_TONGA 7
-
-typedef struct _ATOM_Tonga_POWERPLAYTABLE {
- ATOM_COMMON_TABLE_HEADER sHeader;
-
- UCHAR ucTableRevision;
- USHORT usTableSize; /*the size of header structure */
-
- ULONG ulGoldenPPID;
- ULONG ulGoldenRevision;
- USHORT usFormatID;
-
- USHORT usVoltageTime; /*in microseconds */
- ULONG ulPlatformCaps; /*See ATOM_Tonga_CAPS_* */
-
- ULONG ulMaxODEngineClock; /*For Overdrive. */
- ULONG ulMaxODMemoryClock; /*For Overdrive. */
-
- USHORT usPowerControlLimit;
- USHORT usUlvVoltageOffset; /*in mv units */
-
- USHORT usStateArrayOffset; /*points to ATOM_Tonga_State_Array */
- USHORT usFanTableOffset; /*points to ATOM_Tonga_Fan_Table */
- USHORT usThermalControllerOffset; /*points to ATOM_Tonga_Thermal_Controller */
- USHORT usReserv; /*CustomThermalPolicy removed for Tonga. Keep this filed as reserved. */
-
- USHORT usMclkDependencyTableOffset; /*points to ATOM_Tonga_MCLK_Dependency_Table */
- USHORT usSclkDependencyTableOffset; /*points to ATOM_Tonga_SCLK_Dependency_Table */
- USHORT usVddcLookupTableOffset; /*points to ATOM_Tonga_Voltage_Lookup_Table */
- USHORT usVddgfxLookupTableOffset; /*points to ATOM_Tonga_Voltage_Lookup_Table */
-
- USHORT usMMDependencyTableOffset; /*points to ATOM_Tonga_MM_Dependency_Table */
-
- USHORT usVCEStateTableOffset; /*points to ATOM_Tonga_VCE_State_Table; */
-
- USHORT usPPMTableOffset; /*points to ATOM_Tonga_PPM_Table */
- USHORT usPowerTuneTableOffset; /*points to ATOM_PowerTune_Table */
-
- USHORT usHardLimitTableOffset; /*points to ATOM_Tonga_Hard_Limit_Table */
-
- USHORT usPCIETableOffset; /*points to ATOM_Tonga_PCIE_Table */
-
- USHORT usGPIOTableOffset; /*points to ATOM_Tonga_GPIO_Table */
-
- USHORT usReserved[6]; /*TODO: modify reserved size to fit structure aligning */
-} ATOM_Tonga_POWERPLAYTABLE;
-
-typedef struct _ATOM_Tonga_State {
- UCHAR ucEngineClockIndexHigh;
- UCHAR ucEngineClockIndexLow;
-
- UCHAR ucMemoryClockIndexHigh;
- UCHAR ucMemoryClockIndexLow;
-
- UCHAR ucPCIEGenLow;
- UCHAR ucPCIEGenHigh;
-
- UCHAR ucPCIELaneLow;
- UCHAR ucPCIELaneHigh;
-
- USHORT usClassification;
- ULONG ulCapsAndSettings;
- USHORT usClassification2;
- UCHAR ucUnused[4];
-} ATOM_Tonga_State;
-
-typedef struct _ATOM_Tonga_State_Array {
- UCHAR ucRevId;
- UCHAR ucNumEntries; /* Number of entries. */
- ATOM_Tonga_State states[1]; /* Dynamically allocate entries. */
-} ATOM_Tonga_State_Array;
-
-typedef struct _ATOM_Tonga_MCLK_Dependency_Record {
- UCHAR ucVddcInd; /* Vddc voltage */
- USHORT usVddci;
- USHORT usVddgfxOffset; /* Offset relative to Vddc voltage */
- USHORT usMvdd;
- ULONG ulMclk;
- USHORT usReserved;
-} ATOM_Tonga_MCLK_Dependency_Record;
-
-typedef struct _ATOM_Tonga_MCLK_Dependency_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries; /* Number of entries. */
- ATOM_Tonga_MCLK_Dependency_Record entries[1]; /* Dynamically allocate entries. */
-} ATOM_Tonga_MCLK_Dependency_Table;
-
-typedef struct _ATOM_Tonga_SCLK_Dependency_Record {
- UCHAR ucVddInd; /* Base voltage */
- USHORT usVddcOffset; /* Offset relative to base voltage */
- ULONG ulSclk;
- USHORT usEdcCurrent;
- UCHAR ucReliabilityTemperature;
- UCHAR ucCKSVOffsetandDisable; /* Bits 0~6: Voltage offset for CKS, Bit 7: Disable/enable for the SCLK level. */
-} ATOM_Tonga_SCLK_Dependency_Record;
-
-typedef struct _ATOM_Tonga_SCLK_Dependency_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries; /* Number of entries. */
- ATOM_Tonga_SCLK_Dependency_Record entries[1]; /* Dynamically allocate entries. */
-} ATOM_Tonga_SCLK_Dependency_Table;
-
-typedef struct _ATOM_Polaris_SCLK_Dependency_Record {
- UCHAR ucVddInd; /* Base voltage */
- USHORT usVddcOffset; /* Offset relative to base voltage */
- ULONG ulSclk;
- USHORT usEdcCurrent;
- UCHAR ucReliabilityTemperature;
- UCHAR ucCKSVOffsetandDisable; /* Bits 0~6: Voltage offset for CKS, Bit 7: Disable/enable for the SCLK level. */
- ULONG ulSclkOffset;
-} ATOM_Polaris_SCLK_Dependency_Record;
-
-typedef struct _ATOM_Polaris_SCLK_Dependency_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries; /* Number of entries. */
- ATOM_Polaris_SCLK_Dependency_Record entries[1]; /* Dynamically allocate entries. */
-} ATOM_Polaris_SCLK_Dependency_Table;
-
-typedef struct _ATOM_Tonga_PCIE_Record {
- UCHAR ucPCIEGenSpeed;
- UCHAR usPCIELaneWidth;
- UCHAR ucReserved[2];
-} ATOM_Tonga_PCIE_Record;
-
-typedef struct _ATOM_Tonga_PCIE_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries; /* Number of entries. */
- ATOM_Tonga_PCIE_Record entries[1]; /* Dynamically allocate entries. */
-} ATOM_Tonga_PCIE_Table;
-
-typedef struct _ATOM_Polaris10_PCIE_Record {
- UCHAR ucPCIEGenSpeed;
- UCHAR usPCIELaneWidth;
- UCHAR ucReserved[2];
- ULONG ulPCIE_Sclk;
-} ATOM_Polaris10_PCIE_Record;
-
-typedef struct _ATOM_Polaris10_PCIE_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries; /* Number of entries. */
- ATOM_Polaris10_PCIE_Record entries[1]; /* Dynamically allocate entries. */
-} ATOM_Polaris10_PCIE_Table;
-
-
-typedef struct _ATOM_Tonga_MM_Dependency_Record {
- UCHAR ucVddcInd; /* VDDC voltage */
- USHORT usVddgfxOffset; /* Offset relative to VDDC voltage */
- ULONG ulDClk; /* UVD D-clock */
- ULONG ulVClk; /* UVD V-clock */
- ULONG ulEClk; /* VCE clock */
- ULONG ulAClk; /* ACP clock */
- ULONG ulSAMUClk; /* SAMU clock */
-} ATOM_Tonga_MM_Dependency_Record;
-
-typedef struct _ATOM_Tonga_MM_Dependency_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries; /* Number of entries. */
- ATOM_Tonga_MM_Dependency_Record entries[1]; /* Dynamically allocate entries. */
-} ATOM_Tonga_MM_Dependency_Table;
-
-typedef struct _ATOM_Tonga_Voltage_Lookup_Record {
- USHORT usVdd; /* Base voltage */
- USHORT usCACLow;
- USHORT usCACMid;
- USHORT usCACHigh;
-} ATOM_Tonga_Voltage_Lookup_Record;
-
-typedef struct _ATOM_Tonga_Voltage_Lookup_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries; /* Number of entries. */
- ATOM_Tonga_Voltage_Lookup_Record entries[1]; /* Dynamically allocate entries. */
-} ATOM_Tonga_Voltage_Lookup_Table;
-
-typedef struct _ATOM_Tonga_Fan_Table {
- UCHAR ucRevId; /* Change this if the table format changes or version changes so that the other fields are not the same. */
- UCHAR ucTHyst; /* Temperature hysteresis. Integer. */
- USHORT usTMin; /* The temperature, in 0.01 centigrades, below which we just run at a minimal PWM. */
- USHORT usTMed; /* The middle temperature where we change slopes. */
- USHORT usTHigh; /* The high point above TMed for adjusting the second slope. */
- USHORT usPWMMin; /* The minimum PWM value in percent (0.01% increments). */
- USHORT usPWMMed; /* The PWM value (in percent) at TMed. */
- USHORT usPWMHigh; /* The PWM value at THigh. */
- USHORT usTMax; /* The max temperature */
- UCHAR ucFanControlMode; /* Legacy or Fuzzy Fan mode */
- USHORT usFanPWMMax; /* Maximum allowed fan power in percent */
- USHORT usFanOutputSensitivity; /* Sensitivity of fan reaction to temepature changes */
- USHORT usFanRPMMax; /* The default value in RPM */
- ULONG ulMinFanSCLKAcousticLimit; /* Minimum Fan Controller SCLK Frequency Acoustic Limit. */
- UCHAR ucTargetTemperature; /* Advanced fan controller target temperature. */
- UCHAR ucMinimumPWMLimit; /* The minimum PWM that the advanced fan controller can set. This should be set to the highest PWM that will run the fan at its lowest RPM. */
- USHORT usReserved;
-} ATOM_Tonga_Fan_Table;
-
-typedef struct _ATOM_Fiji_Fan_Table {
- UCHAR ucRevId; /* Change this if the table format changes or version changes so that the other fields are not the same. */
- UCHAR ucTHyst; /* Temperature hysteresis. Integer. */
- USHORT usTMin; /* The temperature, in 0.01 centigrades, below which we just run at a minimal PWM. */
- USHORT usTMed; /* The middle temperature where we change slopes. */
- USHORT usTHigh; /* The high point above TMed for adjusting the second slope. */
- USHORT usPWMMin; /* The minimum PWM value in percent (0.01% increments). */
- USHORT usPWMMed; /* The PWM value (in percent) at TMed. */
- USHORT usPWMHigh; /* The PWM value at THigh. */
- USHORT usTMax; /* The max temperature */
- UCHAR ucFanControlMode; /* Legacy or Fuzzy Fan mode */
- USHORT usFanPWMMax; /* Maximum allowed fan power in percent */
- USHORT usFanOutputSensitivity; /* Sensitivity of fan reaction to temepature changes */
- USHORT usFanRPMMax; /* The default value in RPM */
- ULONG ulMinFanSCLKAcousticLimit; /* Minimum Fan Controller SCLK Frequency Acoustic Limit. */
- UCHAR ucTargetTemperature; /* Advanced fan controller target temperature. */
- UCHAR ucMinimumPWMLimit; /* The minimum PWM that the advanced fan controller can set. This should be set to the highest PWM that will run the fan at its lowest RPM. */
- USHORT usFanGainEdge;
- USHORT usFanGainHotspot;
- USHORT usFanGainLiquid;
- USHORT usFanGainVrVddc;
- USHORT usFanGainVrMvdd;
- USHORT usFanGainPlx;
- USHORT usFanGainHbm;
- USHORT usReserved;
-} ATOM_Fiji_Fan_Table;
-
-typedef struct _ATOM_Tonga_Thermal_Controller {
- UCHAR ucRevId;
- UCHAR ucType; /* one of ATOM_TONGA_PP_THERMALCONTROLLER_* */
- UCHAR ucI2cLine; /* as interpreted by DAL I2C */
- UCHAR ucI2cAddress;
- UCHAR ucFanParameters; /* Fan Control Parameters. */
- UCHAR ucFanMinRPM; /* Fan Minimum RPM (hundreds) -- for display purposes only. */
- UCHAR ucFanMaxRPM; /* Fan Maximum RPM (hundreds) -- for display purposes only. */
- UCHAR ucReserved;
- UCHAR ucFlags; /* to be defined */
-} ATOM_Tonga_Thermal_Controller;
-
-typedef struct _ATOM_Tonga_VCE_State_Record {
- UCHAR ucVCEClockIndex; /*index into usVCEDependencyTableOffset of 'ATOM_Tonga_MM_Dependency_Table' type */
- UCHAR ucFlag; /* 2 bits indicates memory p-states */
- UCHAR ucSCLKIndex; /*index into ATOM_Tonga_SCLK_Dependency_Table */
- UCHAR ucMCLKIndex; /*index into ATOM_Tonga_MCLK_Dependency_Table */
-} ATOM_Tonga_VCE_State_Record;
-
-typedef struct _ATOM_Tonga_VCE_State_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries;
- ATOM_Tonga_VCE_State_Record entries[1];
-} ATOM_Tonga_VCE_State_Table;
-
-typedef struct _ATOM_Tonga_PowerTune_Table {
- UCHAR ucRevId;
- USHORT usTDP;
- USHORT usConfigurableTDP;
- USHORT usTDC;
- USHORT usBatteryPowerLimit;
- USHORT usSmallPowerLimit;
- USHORT usLowCACLeakage;
- USHORT usHighCACLeakage;
- USHORT usMaximumPowerDeliveryLimit;
- USHORT usTjMax;
- USHORT usPowerTuneDataSetID;
- USHORT usEDCLimit;
- USHORT usSoftwareShutdownTemp;
- USHORT usClockStretchAmount;
- USHORT usReserve[2];
-} ATOM_Tonga_PowerTune_Table;
-
-typedef struct _ATOM_Fiji_PowerTune_Table {
- UCHAR ucRevId;
- USHORT usTDP;
- USHORT usConfigurableTDP;
- USHORT usTDC;
- USHORT usBatteryPowerLimit;
- USHORT usSmallPowerLimit;
- USHORT usLowCACLeakage;
- USHORT usHighCACLeakage;
- USHORT usMaximumPowerDeliveryLimit;
- USHORT usTjMax; /* For Fiji, this is also usTemperatureLimitEdge; */
- USHORT usPowerTuneDataSetID;
- USHORT usEDCLimit;
- USHORT usSoftwareShutdownTemp;
- USHORT usClockStretchAmount;
- USHORT usTemperatureLimitHotspot; /*The following are added for Fiji */
- USHORT usTemperatureLimitLiquid1;
- USHORT usTemperatureLimitLiquid2;
- USHORT usTemperatureLimitVrVddc;
- USHORT usTemperatureLimitVrMvdd;
- USHORT usTemperatureLimitPlx;
- UCHAR ucLiquid1_I2C_address; /*Liquid */
- UCHAR ucLiquid2_I2C_address;
- UCHAR ucLiquid_I2C_Line;
- UCHAR ucVr_I2C_address; /*VR */
- UCHAR ucVr_I2C_Line;
- UCHAR ucPlx_I2C_address; /*PLX */
- UCHAR ucPlx_I2C_Line;
- USHORT usReserved;
-} ATOM_Fiji_PowerTune_Table;
-
-#define ATOM_PPM_A_A 1
-#define ATOM_PPM_A_I 2
-typedef struct _ATOM_Tonga_PPM_Table {
- UCHAR ucRevId;
- UCHAR ucPpmDesign; /*A+I or A+A */
- USHORT usCpuCoreNumber;
- ULONG ulPlatformTDP;
- ULONG ulSmallACPlatformTDP;
- ULONG ulPlatformTDC;
- ULONG ulSmallACPlatformTDC;
- ULONG ulApuTDP;
- ULONG ulDGpuTDP;
- ULONG ulDGpuUlvPower;
- ULONG ulTjmax;
-} ATOM_Tonga_PPM_Table;
-
-typedef struct _ATOM_Tonga_Hard_Limit_Record {
- ULONG ulSCLKLimit;
- ULONG ulMCLKLimit;
- USHORT usVddcLimit;
- USHORT usVddciLimit;
- USHORT usVddgfxLimit;
-} ATOM_Tonga_Hard_Limit_Record;
-
-typedef struct _ATOM_Tonga_Hard_Limit_Table {
- UCHAR ucRevId;
- UCHAR ucNumEntries;
- ATOM_Tonga_Hard_Limit_Record entries[1];
-} ATOM_Tonga_Hard_Limit_Table;
-
-typedef struct _ATOM_Tonga_GPIO_Table {
- UCHAR ucRevId;
- UCHAR ucVRHotTriggeredSclkDpmIndex; /* If VRHot signal is triggered SCLK will be limited to this DPM level */
- UCHAR ucReserve[5];
-} ATOM_Tonga_GPIO_Table;
-
-typedef struct _PPTable_Generic_SubTable_Header {
- UCHAR ucRevId;
-} PPTable_Generic_SubTable_Header;
-
-
-#pragma pack(pop)
-
-
-#endif
+++ /dev/null
-/*
- * Copyright 2015 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- */
-#include <linux/module.h>
-#include <linux/slab.h>
-
-#include "tonga_processpptables.h"
-#include "ppatomctrl.h"
-#include "atombios.h"
-#include "pp_debug.h"
-#include "hwmgr.h"
-#include "cgs_common.h"
-#include "tonga_pptable.h"
-
-/**
- * Private Function used during initialization.
- * @param hwmgr Pointer to the hardware manager.
- * @param setIt A flag indication if the capability should be set (TRUE) or reset (FALSE).
- * @param cap Which capability to set/reset.
- */
-static void set_hw_cap(struct pp_hwmgr *hwmgr, bool setIt, enum phm_platform_caps cap)
-{
- if (setIt)
- phm_cap_set(hwmgr->platform_descriptor.platformCaps, cap);
- else
- phm_cap_unset(hwmgr->platform_descriptor.platformCaps, cap);
-}
-
-
-/**
- * Private Function used during initialization.
- * @param hwmgr Pointer to the hardware manager.
- * @param powerplay_caps the bit array (from BIOS) of capability bits.
- * @exception the current implementation always returns 1.
- */
-static int set_platform_caps(struct pp_hwmgr *hwmgr, uint32_t powerplay_caps)
-{
- PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE16____),
- "ATOM_PP_PLATFORM_CAP_ASPM_L1 is not supported!", continue);
- PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE64____),
- "ATOM_PP_PLATFORM_CAP_GEMINIPRIMARY is not supported!", continue);
- PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE512____),
- "ATOM_PP_PLATFORM_CAP_SIDEPORTCONTROL is not supported!", continue);
- PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE1024____),
- "ATOM_PP_PLATFORM_CAP_TURNOFFPLL_ASPML1 is not supported!", continue);
- PP_ASSERT_WITH_CODE((~powerplay_caps & ____RETIRE2048____),
- "ATOM_PP_PLATFORM_CAP_HTLINKCONTROL is not supported!", continue);
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_POWERPLAY),
- PHM_PlatformCaps_PowerPlaySupport
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_SBIOSPOWERSOURCE),
- PHM_PlatformCaps_BiosPowerSourceControl
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_HARDWAREDC),
- PHM_PlatformCaps_AutomaticDCTransition
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_MVDD_CONTROL),
- PHM_PlatformCaps_EnableMVDDControl
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_VDDCI_CONTROL),
- PHM_PlatformCaps_ControlVDDCI
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_VDDGFX_CONTROL),
- PHM_PlatformCaps_ControlVDDGFX
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_BACO),
- PHM_PlatformCaps_BACO
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_CAP_DISABLE_VOLTAGE_ISLAND),
- PHM_PlatformCaps_DisableVoltageIsland
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PP_PLATFORM_COMBINE_PCC_WITH_THERMAL_SIGNAL),
- PHM_PlatformCaps_CombinePCCWithThermalSignal
- );
-
- set_hw_cap(
- hwmgr,
- 0 != (powerplay_caps & ATOM_TONGA_PLATFORM_LOAD_POST_PRODUCTION_FIRMWARE),
- PHM_PlatformCaps_LoadPostProductionFirmware
- );
-
- return 0;
-}
-
-/**
- * Private Function to get the PowerPlay Table Address.
- */
-const void *get_powerplay_table(struct pp_hwmgr *hwmgr)
-{
- int index = GetIndexIntoMasterTable(DATA, PowerPlayInfo);
-
- u16 size;
- u8 frev, crev;
- void *table_address = (void *)hwmgr->soft_pp_table;
-
- if (!table_address) {
- table_address = (ATOM_Tonga_POWERPLAYTABLE *)
- cgs_atom_get_data_table(hwmgr->device,
- index, &size, &frev, &crev);
- hwmgr->soft_pp_table = table_address; /*Cache the result in RAM.*/
- hwmgr->soft_pp_table_size = size;
- }
-
- return table_address;
-}
-
-static int get_vddc_lookup_table(
- struct pp_hwmgr *hwmgr,
- phm_ppt_v1_voltage_lookup_table **lookup_table,
- const ATOM_Tonga_Voltage_Lookup_Table *vddc_lookup_pp_tables,
- uint32_t max_levels
- )
-{
- uint32_t table_size, i;
- phm_ppt_v1_voltage_lookup_table *table;
-
- PP_ASSERT_WITH_CODE((0 != vddc_lookup_pp_tables->ucNumEntries),
- "Invalid CAC Leakage PowerPlay Table!", return 1);
-
- table_size = sizeof(uint32_t) +
- sizeof(phm_ppt_v1_voltage_lookup_record) * max_levels;
-
- table = kzalloc(table_size, GFP_KERNEL);
-
- if (NULL == table)
- return -ENOMEM;
-
- memset(table, 0x00, table_size);
-
- table->count = vddc_lookup_pp_tables->ucNumEntries;
-
- for (i = 0; i < vddc_lookup_pp_tables->ucNumEntries; i++) {
- table->entries[i].us_calculated = 0;
- table->entries[i].us_vdd =
- vddc_lookup_pp_tables->entries[i].usVdd;
- table->entries[i].us_cac_low =
- vddc_lookup_pp_tables->entries[i].usCACLow;
- table->entries[i].us_cac_mid =
- vddc_lookup_pp_tables->entries[i].usCACMid;
- table->entries[i].us_cac_high =
- vddc_lookup_pp_tables->entries[i].usCACHigh;
- }
-
- *lookup_table = table;
-
- return 0;
-}
-
-/**
- * Private Function used during initialization.
- * Initialize Platform Power Management Parameter table
- * @param hwmgr Pointer to the hardware manager.
- * @param atom_ppm_table Pointer to PPM table in VBIOS
- */
-static int get_platform_power_management_table(
- struct pp_hwmgr *hwmgr,
- ATOM_Tonga_PPM_Table *atom_ppm_table)
-{
- struct phm_ppm_table *ptr = kzalloc(sizeof(ATOM_Tonga_PPM_Table), GFP_KERNEL);
- struct phm_ppt_v1_information *pp_table_information =
- (struct phm_ppt_v1_information *)(hwmgr->pptable);
-
- if (NULL == ptr)
- return -ENOMEM;
-
- ptr->ppm_design
- = atom_ppm_table->ucPpmDesign;
- ptr->cpu_core_number
- = atom_ppm_table->usCpuCoreNumber;
- ptr->platform_tdp
- = atom_ppm_table->ulPlatformTDP;
- ptr->small_ac_platform_tdp
- = atom_ppm_table->ulSmallACPlatformTDP;
- ptr->platform_tdc
- = atom_ppm_table->ulPlatformTDC;
- ptr->small_ac_platform_tdc
- = atom_ppm_table->ulSmallACPlatformTDC;
- ptr->apu_tdp
- = atom_ppm_table->ulApuTDP;
- ptr->dgpu_tdp
- = atom_ppm_table->ulDGpuTDP;
- ptr->dgpu_ulv_power
- = atom_ppm_table->ulDGpuUlvPower;
- ptr->tj_max
- = atom_ppm_table->ulTjmax;
-
- pp_table_information->ppm_parameter_table = ptr;
-
- return 0;
-}
-
-/**
- * Private Function used during initialization.
- * Initialize TDP limits for DPM2
- * @param hwmgr Pointer to the hardware manager.
- * @param powerplay_table Pointer to the PowerPlay Table.
- */
-static int init_dpm_2_parameters(
- struct pp_hwmgr *hwmgr,
- const ATOM_Tonga_POWERPLAYTABLE *powerplay_table
- )
-{
- int result = 0;
- struct phm_ppt_v1_information *pp_table_information = (struct phm_ppt_v1_information *)(hwmgr->pptable);
- ATOM_Tonga_PPM_Table *atom_ppm_table;
- uint32_t disable_ppm = 0;
- uint32_t disable_power_control = 0;
-
- pp_table_information->us_ulv_voltage_offset =
- le16_to_cpu(powerplay_table->usUlvVoltageOffset);
-
- pp_table_information->ppm_parameter_table = NULL;
- pp_table_information->vddc_lookup_table = NULL;
- pp_table_information->vddgfx_lookup_table = NULL;
- /* TDP limits */
- hwmgr->platform_descriptor.TDPODLimit =
- le16_to_cpu(powerplay_table->usPowerControlLimit);
- hwmgr->platform_descriptor.TDPAdjustment = 0;
- hwmgr->platform_descriptor.VidAdjustment = 0;
- hwmgr->platform_descriptor.VidAdjustmentPolarity = 0;
- hwmgr->platform_descriptor.VidMinLimit = 0;
- hwmgr->platform_descriptor.VidMaxLimit = 1500000;
- hwmgr->platform_descriptor.VidStep = 6250;
-
- disable_power_control = 0;
- if (0 == disable_power_control) {
- /* enable TDP overdrive (PowerControl) feature as well if supported */
- if (hwmgr->platform_descriptor.TDPODLimit != 0)
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_PowerControl);
- }
-
- if (0 != powerplay_table->usVddcLookupTableOffset) {
- const ATOM_Tonga_Voltage_Lookup_Table *pVddcCACTable =
- (ATOM_Tonga_Voltage_Lookup_Table *)(((unsigned long)powerplay_table) +
- le16_to_cpu(powerplay_table->usVddcLookupTableOffset));
-
- result = get_vddc_lookup_table(hwmgr,
- &pp_table_information->vddc_lookup_table, pVddcCACTable, 16);
- }
-
- if (0 != powerplay_table->usVddgfxLookupTableOffset) {
- const ATOM_Tonga_Voltage_Lookup_Table *pVddgfxCACTable =
- (ATOM_Tonga_Voltage_Lookup_Table *)(((unsigned long)powerplay_table) +
- le16_to_cpu(powerplay_table->usVddgfxLookupTableOffset));
-
- result = get_vddc_lookup_table(hwmgr,
- &pp_table_information->vddgfx_lookup_table, pVddgfxCACTable, 16);
- }
-
- disable_ppm = 0;
- if (0 == disable_ppm) {
- atom_ppm_table = (ATOM_Tonga_PPM_Table *)
- (((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usPPMTableOffset));
-
- if (0 != powerplay_table->usPPMTableOffset) {
- if (get_platform_power_management_table(hwmgr, atom_ppm_table) == 0) {
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_EnablePlatformPowerManagement);
- }
- }
- }
-
- return result;
-}
-
-static int get_valid_clk(
- struct pp_hwmgr *hwmgr,
- struct phm_clock_array **clk_table,
- const phm_ppt_v1_clock_voltage_dependency_table * clk_volt_pp_table
- )
-{
- uint32_t table_size, i;
- struct phm_clock_array *table;
-
- PP_ASSERT_WITH_CODE((0 != clk_volt_pp_table->count),
- "Invalid PowerPlay Table!", return -1);
-
- table_size = sizeof(uint32_t) +
- sizeof(uint32_t) * clk_volt_pp_table->count;
-
- table = kzalloc(table_size, GFP_KERNEL);
-
- if (NULL == table)
- return -ENOMEM;
-
- memset(table, 0x00, table_size);
-
- table->count = (uint32_t)clk_volt_pp_table->count;
-
- for (i = 0; i < table->count; i++)
- table->values[i] = (uint32_t)clk_volt_pp_table->entries[i].clk;
-
- *clk_table = table;
-
- return 0;
-}
-
-static int get_hard_limits(
- struct pp_hwmgr *hwmgr,
- struct phm_clock_and_voltage_limits *limits,
- const ATOM_Tonga_Hard_Limit_Table * limitable
- )
-{
- PP_ASSERT_WITH_CODE((0 != limitable->ucNumEntries), "Invalid PowerPlay Table!", return -1);
-
- /* currently we always take entries[0] parameters */
- limits->sclk = (uint32_t)limitable->entries[0].ulSCLKLimit;
- limits->mclk = (uint32_t)limitable->entries[0].ulMCLKLimit;
- limits->vddc = (uint16_t)limitable->entries[0].usVddcLimit;
- limits->vddci = (uint16_t)limitable->entries[0].usVddciLimit;
- limits->vddgfx = (uint16_t)limitable->entries[0].usVddgfxLimit;
-
- return 0;
-}
-
-static int get_mclk_voltage_dependency_table(
- struct pp_hwmgr *hwmgr,
- phm_ppt_v1_clock_voltage_dependency_table **pp_tonga_mclk_dep_table,
- const ATOM_Tonga_MCLK_Dependency_Table * mclk_dep_table
- )
-{
- uint32_t table_size, i;
- phm_ppt_v1_clock_voltage_dependency_table *mclk_table;
-
- PP_ASSERT_WITH_CODE((0 != mclk_dep_table->ucNumEntries),
- "Invalid PowerPlay Table!", return -1);
-
- table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record)
- * mclk_dep_table->ucNumEntries;
-
- mclk_table = kzalloc(table_size, GFP_KERNEL);
-
- if (NULL == mclk_table)
- return -ENOMEM;
-
- memset(mclk_table, 0x00, table_size);
-
- mclk_table->count = (uint32_t)mclk_dep_table->ucNumEntries;
-
- for (i = 0; i < mclk_dep_table->ucNumEntries; i++) {
- mclk_table->entries[i].vddInd =
- mclk_dep_table->entries[i].ucVddcInd;
- mclk_table->entries[i].vdd_offset =
- mclk_dep_table->entries[i].usVddgfxOffset;
- mclk_table->entries[i].vddci =
- mclk_dep_table->entries[i].usVddci;
- mclk_table->entries[i].mvdd =
- mclk_dep_table->entries[i].usMvdd;
- mclk_table->entries[i].clk =
- mclk_dep_table->entries[i].ulMclk;
- }
-
- *pp_tonga_mclk_dep_table = mclk_table;
-
- return 0;
-}
-
-static int get_sclk_voltage_dependency_table(
- struct pp_hwmgr *hwmgr,
- phm_ppt_v1_clock_voltage_dependency_table **pp_tonga_sclk_dep_table,
- const PPTable_Generic_SubTable_Header *sclk_dep_table
- )
-{
- uint32_t table_size, i;
- phm_ppt_v1_clock_voltage_dependency_table *sclk_table;
-
- if (sclk_dep_table->ucRevId < 1) {
- const ATOM_Tonga_SCLK_Dependency_Table *tonga_table =
- (ATOM_Tonga_SCLK_Dependency_Table *)sclk_dep_table;
-
- PP_ASSERT_WITH_CODE((0 != tonga_table->ucNumEntries),
- "Invalid PowerPlay Table!", return -1);
-
- table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record)
- * tonga_table->ucNumEntries;
-
- sclk_table = kzalloc(table_size, GFP_KERNEL);
-
- if (NULL == sclk_table)
- return -ENOMEM;
-
- memset(sclk_table, 0x00, table_size);
-
- sclk_table->count = (uint32_t)tonga_table->ucNumEntries;
-
- for (i = 0; i < tonga_table->ucNumEntries; i++) {
- sclk_table->entries[i].vddInd =
- tonga_table->entries[i].ucVddInd;
- sclk_table->entries[i].vdd_offset =
- tonga_table->entries[i].usVddcOffset;
- sclk_table->entries[i].clk =
- tonga_table->entries[i].ulSclk;
- sclk_table->entries[i].cks_enable =
- (((tonga_table->entries[i].ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0;
- sclk_table->entries[i].cks_voffset =
- (tonga_table->entries[i].ucCKSVOffsetandDisable & 0x7F);
- }
- } else {
- const ATOM_Polaris_SCLK_Dependency_Table *polaris_table =
- (ATOM_Polaris_SCLK_Dependency_Table *)sclk_dep_table;
-
- PP_ASSERT_WITH_CODE((0 != polaris_table->ucNumEntries),
- "Invalid PowerPlay Table!", return -1);
-
- table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record)
- * polaris_table->ucNumEntries;
-
- sclk_table = kzalloc(table_size, GFP_KERNEL);
-
- if (NULL == sclk_table)
- return -ENOMEM;
-
- memset(sclk_table, 0x00, table_size);
-
- sclk_table->count = (uint32_t)polaris_table->ucNumEntries;
-
- for (i = 0; i < polaris_table->ucNumEntries; i++) {
- sclk_table->entries[i].vddInd =
- polaris_table->entries[i].ucVddInd;
- sclk_table->entries[i].vdd_offset =
- polaris_table->entries[i].usVddcOffset;
- sclk_table->entries[i].clk =
- polaris_table->entries[i].ulSclk;
- sclk_table->entries[i].cks_enable =
- (((polaris_table->entries[i].ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0;
- sclk_table->entries[i].cks_voffset =
- (polaris_table->entries[i].ucCKSVOffsetandDisable & 0x7F);
- sclk_table->entries[i].sclk_offset = polaris_table->entries[i].ulSclkOffset;
- }
- }
- *pp_tonga_sclk_dep_table = sclk_table;
-
- return 0;
-}
-
-static int get_pcie_table(
- struct pp_hwmgr *hwmgr,
- phm_ppt_v1_pcie_table **pp_tonga_pcie_table,
- const PPTable_Generic_SubTable_Header * pTable
- )
-{
- uint32_t table_size, i, pcie_count;
- phm_ppt_v1_pcie_table *pcie_table;
- struct phm_ppt_v1_information *pp_table_information =
- (struct phm_ppt_v1_information *)(hwmgr->pptable);
-
- if (pTable->ucRevId < 1) {
- const ATOM_Tonga_PCIE_Table *atom_pcie_table = (ATOM_Tonga_PCIE_Table *)pTable;
- PP_ASSERT_WITH_CODE((atom_pcie_table->ucNumEntries != 0),
- "Invalid PowerPlay Table!", return -1);
-
- table_size = sizeof(uint32_t) +
- sizeof(phm_ppt_v1_pcie_record) * atom_pcie_table->ucNumEntries;
-
- pcie_table = kzalloc(table_size, GFP_KERNEL);
-
- if (pcie_table == NULL)
- return -ENOMEM;
-
- memset(pcie_table, 0x00, table_size);
-
- /*
- * Make sure the number of pcie entries are less than or equal to sclk dpm levels.
- * Since first PCIE entry is for ULV, #pcie has to be <= SclkLevel + 1.
- */
- pcie_count = (pp_table_information->vdd_dep_on_sclk->count) + 1;
- if ((uint32_t)atom_pcie_table->ucNumEntries <= pcie_count)
- pcie_count = (uint32_t)atom_pcie_table->ucNumEntries;
- else
- printk(KERN_ERR "[ powerplay ] Number of Pcie Entries exceed the number of SCLK Dpm Levels! \
- Disregarding the excess entries... \n");
-
- pcie_table->count = pcie_count;
-
- for (i = 0; i < pcie_count; i++) {
- pcie_table->entries[i].gen_speed =
- atom_pcie_table->entries[i].ucPCIEGenSpeed;
- pcie_table->entries[i].lane_width =
- atom_pcie_table->entries[i].usPCIELaneWidth;
- }
-
- *pp_tonga_pcie_table = pcie_table;
- } else {
- /* Polaris10/Polaris11 and newer. */
- const ATOM_Polaris10_PCIE_Table *atom_pcie_table = (ATOM_Polaris10_PCIE_Table *)pTable;
- PP_ASSERT_WITH_CODE((atom_pcie_table->ucNumEntries != 0),
- "Invalid PowerPlay Table!", return -1);
-
- table_size = sizeof(uint32_t) +
- sizeof(phm_ppt_v1_pcie_record) * atom_pcie_table->ucNumEntries;
-
- pcie_table = kzalloc(table_size, GFP_KERNEL);
-
- if (pcie_table == NULL)
- return -ENOMEM;
-
- memset(pcie_table, 0x00, table_size);
-
- /*
- * Make sure the number of pcie entries are less than or equal to sclk dpm levels.
- * Since first PCIE entry is for ULV, #pcie has to be <= SclkLevel + 1.
- */
- pcie_count = (pp_table_information->vdd_dep_on_sclk->count) + 1;
- if ((uint32_t)atom_pcie_table->ucNumEntries <= pcie_count)
- pcie_count = (uint32_t)atom_pcie_table->ucNumEntries;
- else
- printk(KERN_ERR "[ powerplay ] Number of Pcie Entries exceed the number of SCLK Dpm Levels! \
- Disregarding the excess entries... \n");
-
- pcie_table->count = pcie_count;
-
- for (i = 0; i < pcie_count; i++) {
- pcie_table->entries[i].gen_speed =
- atom_pcie_table->entries[i].ucPCIEGenSpeed;
- pcie_table->entries[i].lane_width =
- atom_pcie_table->entries[i].usPCIELaneWidth;
- pcie_table->entries[i].pcie_sclk =
- atom_pcie_table->entries[i].ulPCIE_Sclk;
- }
-
- *pp_tonga_pcie_table = pcie_table;
- }
-
- return 0;
-}
-
-static int get_cac_tdp_table(
- struct pp_hwmgr *hwmgr,
- struct phm_cac_tdp_table **cac_tdp_table,
- const PPTable_Generic_SubTable_Header * table
- )
-{
- uint32_t table_size;
- struct phm_cac_tdp_table *tdp_table;
-
- table_size = sizeof(uint32_t) + sizeof(struct phm_cac_tdp_table);
- tdp_table = kzalloc(table_size, GFP_KERNEL);
-
- if (NULL == tdp_table)
- return -ENOMEM;
-
- memset(tdp_table, 0x00, table_size);
-
- hwmgr->dyn_state.cac_dtp_table = kzalloc(table_size, GFP_KERNEL);
-
- if (NULL == hwmgr->dyn_state.cac_dtp_table) {
- kfree(tdp_table);
- return -ENOMEM;
- }
-
- memset(hwmgr->dyn_state.cac_dtp_table, 0x00, table_size);
-
- if (table->ucRevId < 3) {
- const ATOM_Tonga_PowerTune_Table *tonga_table =
- (ATOM_Tonga_PowerTune_Table *)table;
- tdp_table->usTDP = tonga_table->usTDP;
- tdp_table->usConfigurableTDP =
- tonga_table->usConfigurableTDP;
- tdp_table->usTDC = tonga_table->usTDC;
- tdp_table->usBatteryPowerLimit =
- tonga_table->usBatteryPowerLimit;
- tdp_table->usSmallPowerLimit =
- tonga_table->usSmallPowerLimit;
- tdp_table->usLowCACLeakage =
- tonga_table->usLowCACLeakage;
- tdp_table->usHighCACLeakage =
- tonga_table->usHighCACLeakage;
- tdp_table->usMaximumPowerDeliveryLimit =
- tonga_table->usMaximumPowerDeliveryLimit;
- tdp_table->usDefaultTargetOperatingTemp =
- tonga_table->usTjMax;
- tdp_table->usTargetOperatingTemp =
- tonga_table->usTjMax; /*Set the initial temp to the same as default */
- tdp_table->usPowerTuneDataSetID =
- tonga_table->usPowerTuneDataSetID;
- tdp_table->usSoftwareShutdownTemp =
- tonga_table->usSoftwareShutdownTemp;
- tdp_table->usClockStretchAmount =
- tonga_table->usClockStretchAmount;
- } else { /* Fiji and newer */
- const ATOM_Fiji_PowerTune_Table *fijitable =
- (ATOM_Fiji_PowerTune_Table *)table;
- tdp_table->usTDP = fijitable->usTDP;
- tdp_table->usConfigurableTDP = fijitable->usConfigurableTDP;
- tdp_table->usTDC = fijitable->usTDC;
- tdp_table->usBatteryPowerLimit = fijitable->usBatteryPowerLimit;
- tdp_table->usSmallPowerLimit = fijitable->usSmallPowerLimit;
- tdp_table->usLowCACLeakage = fijitable->usLowCACLeakage;
- tdp_table->usHighCACLeakage = fijitable->usHighCACLeakage;
- tdp_table->usMaximumPowerDeliveryLimit =
- fijitable->usMaximumPowerDeliveryLimit;
- tdp_table->usDefaultTargetOperatingTemp =
- fijitable->usTjMax;
- tdp_table->usTargetOperatingTemp =
- fijitable->usTjMax; /*Set the initial temp to the same as default */
- tdp_table->usPowerTuneDataSetID =
- fijitable->usPowerTuneDataSetID;
- tdp_table->usSoftwareShutdownTemp =
- fijitable->usSoftwareShutdownTemp;
- tdp_table->usClockStretchAmount =
- fijitable->usClockStretchAmount;
- tdp_table->usTemperatureLimitHotspot =
- fijitable->usTemperatureLimitHotspot;
- tdp_table->usTemperatureLimitLiquid1 =
- fijitable->usTemperatureLimitLiquid1;
- tdp_table->usTemperatureLimitLiquid2 =
- fijitable->usTemperatureLimitLiquid2;
- tdp_table->usTemperatureLimitVrVddc =
- fijitable->usTemperatureLimitVrVddc;
- tdp_table->usTemperatureLimitVrMvdd =
- fijitable->usTemperatureLimitVrMvdd;
- tdp_table->usTemperatureLimitPlx =
- fijitable->usTemperatureLimitPlx;
- tdp_table->ucLiquid1_I2C_address =
- fijitable->ucLiquid1_I2C_address;
- tdp_table->ucLiquid2_I2C_address =
- fijitable->ucLiquid2_I2C_address;
- tdp_table->ucLiquid_I2C_Line =
- fijitable->ucLiquid_I2C_Line;
- tdp_table->ucVr_I2C_address = fijitable->ucVr_I2C_address;
- tdp_table->ucVr_I2C_Line = fijitable->ucVr_I2C_Line;
- tdp_table->ucPlx_I2C_address = fijitable->ucPlx_I2C_address;
- tdp_table->ucPlx_I2C_Line = fijitable->ucPlx_I2C_Line;
- }
-
- *cac_tdp_table = tdp_table;
-
- return 0;
-}
-
-static int get_mm_clock_voltage_table(
- struct pp_hwmgr *hwmgr,
- phm_ppt_v1_mm_clock_voltage_dependency_table **tonga_mm_table,
- const ATOM_Tonga_MM_Dependency_Table * mm_dependency_table
- )
-{
- uint32_t table_size, i;
- const ATOM_Tonga_MM_Dependency_Record *mm_dependency_record;
- phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table;
-
- PP_ASSERT_WITH_CODE((0 != mm_dependency_table->ucNumEntries),
- "Invalid PowerPlay Table!", return -1);
- table_size = sizeof(uint32_t) +
- sizeof(phm_ppt_v1_mm_clock_voltage_dependency_record)
- * mm_dependency_table->ucNumEntries;
- mm_table = kzalloc(table_size, GFP_KERNEL);
-
- if (NULL == mm_table)
- return -ENOMEM;
-
- memset(mm_table, 0x00, table_size);
-
- mm_table->count = mm_dependency_table->ucNumEntries;
-
- for (i = 0; i < mm_dependency_table->ucNumEntries; i++) {
- mm_dependency_record = &mm_dependency_table->entries[i];
- mm_table->entries[i].vddcInd = mm_dependency_record->ucVddcInd;
- mm_table->entries[i].vddgfx_offset = mm_dependency_record->usVddgfxOffset;
- mm_table->entries[i].aclk = mm_dependency_record->ulAClk;
- mm_table->entries[i].samclock = mm_dependency_record->ulSAMUClk;
- mm_table->entries[i].eclk = mm_dependency_record->ulEClk;
- mm_table->entries[i].vclk = mm_dependency_record->ulVClk;
- mm_table->entries[i].dclk = mm_dependency_record->ulDClk;
- }
-
- *tonga_mm_table = mm_table;
-
- return 0;
-}
-
-/**
- * Private Function used during initialization.
- * Initialize clock voltage dependency
- * @param hwmgr Pointer to the hardware manager.
- * @param powerplay_table Pointer to the PowerPlay Table.
- */
-static int init_clock_voltage_dependency(
- struct pp_hwmgr *hwmgr,
- const ATOM_Tonga_POWERPLAYTABLE *powerplay_table
- )
-{
- int result = 0;
- struct phm_ppt_v1_information *pp_table_information =
- (struct phm_ppt_v1_information *)(hwmgr->pptable);
-
- const ATOM_Tonga_MM_Dependency_Table *mm_dependency_table =
- (const ATOM_Tonga_MM_Dependency_Table *)(((unsigned long) powerplay_table) +
- le16_to_cpu(powerplay_table->usMMDependencyTableOffset));
- const PPTable_Generic_SubTable_Header *pPowerTuneTable =
- (const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) +
- le16_to_cpu(powerplay_table->usPowerTuneTableOffset));
- const ATOM_Tonga_MCLK_Dependency_Table *mclk_dep_table =
- (const ATOM_Tonga_MCLK_Dependency_Table *)(((unsigned long) powerplay_table) +
- le16_to_cpu(powerplay_table->usMclkDependencyTableOffset));
- const PPTable_Generic_SubTable_Header *sclk_dep_table =
- (const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) +
- le16_to_cpu(powerplay_table->usSclkDependencyTableOffset));
- const ATOM_Tonga_Hard_Limit_Table *pHardLimits =
- (const ATOM_Tonga_Hard_Limit_Table *)(((unsigned long) powerplay_table) +
- le16_to_cpu(powerplay_table->usHardLimitTableOffset));
- const PPTable_Generic_SubTable_Header *pcie_table =
- (const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) +
- le16_to_cpu(powerplay_table->usPCIETableOffset));
-
- pp_table_information->vdd_dep_on_sclk = NULL;
- pp_table_information->vdd_dep_on_mclk = NULL;
- pp_table_information->mm_dep_table = NULL;
- pp_table_information->pcie_table = NULL;
-
- if (powerplay_table->usMMDependencyTableOffset != 0)
- result = get_mm_clock_voltage_table(hwmgr,
- &pp_table_information->mm_dep_table, mm_dependency_table);
-
- if (result == 0 && powerplay_table->usPowerTuneTableOffset != 0)
- result = get_cac_tdp_table(hwmgr,
- &pp_table_information->cac_dtp_table, pPowerTuneTable);
-
- if (result == 0 && powerplay_table->usSclkDependencyTableOffset != 0)
- result = get_sclk_voltage_dependency_table(hwmgr,
- &pp_table_information->vdd_dep_on_sclk, sclk_dep_table);
-
- if (result == 0 && powerplay_table->usMclkDependencyTableOffset != 0)
- result = get_mclk_voltage_dependency_table(hwmgr,
- &pp_table_information->vdd_dep_on_mclk, mclk_dep_table);
-
- if (result == 0 && powerplay_table->usPCIETableOffset != 0)
- result = get_pcie_table(hwmgr,
- &pp_table_information->pcie_table, pcie_table);
-
- if (result == 0 && powerplay_table->usHardLimitTableOffset != 0)
- result = get_hard_limits(hwmgr,
- &pp_table_information->max_clock_voltage_on_dc, pHardLimits);
-
- hwmgr->dyn_state.max_clock_voltage_on_dc.sclk =
- pp_table_information->max_clock_voltage_on_dc.sclk;
- hwmgr->dyn_state.max_clock_voltage_on_dc.mclk =
- pp_table_information->max_clock_voltage_on_dc.mclk;
- hwmgr->dyn_state.max_clock_voltage_on_dc.vddc =
- pp_table_information->max_clock_voltage_on_dc.vddc;
- hwmgr->dyn_state.max_clock_voltage_on_dc.vddci =
- pp_table_information->max_clock_voltage_on_dc.vddci;
-
- if (result == 0 && (NULL != pp_table_information->vdd_dep_on_mclk)
- && (0 != pp_table_information->vdd_dep_on_mclk->count))
- result = get_valid_clk(hwmgr, &pp_table_information->valid_mclk_values,
- pp_table_information->vdd_dep_on_mclk);
-
- if (result == 0 && (NULL != pp_table_information->vdd_dep_on_sclk)
- && (0 != pp_table_information->vdd_dep_on_sclk->count))
- result = get_valid_clk(hwmgr, &pp_table_information->valid_sclk_values,
- pp_table_information->vdd_dep_on_sclk);
-
- return result;
-}
-
-/** Retrieves the (signed) Overdrive limits from VBIOS.
- * The max engine clock, memory clock and max temperature come from the firmware info table.
- *
- * The information is placed into the platform descriptor.
- *
- * @param hwmgr source of the VBIOS table and owner of the platform descriptor to be updated.
- * @param powerplay_table the address of the PowerPlay table.
- *
- * @return 1 as long as the firmware info table was present and of a supported version.
- */
-static int init_over_drive_limits(
- struct pp_hwmgr *hwmgr,
- const ATOM_Tonga_POWERPLAYTABLE *powerplay_table)
-{
- hwmgr->platform_descriptor.overdriveLimit.engineClock =
- le16_to_cpu(powerplay_table->ulMaxODEngineClock);
- hwmgr->platform_descriptor.overdriveLimit.memoryClock =
- le16_to_cpu(powerplay_table->ulMaxODMemoryClock);
-
- hwmgr->platform_descriptor.minOverdriveVDDC = 0;
- hwmgr->platform_descriptor.maxOverdriveVDDC = 0;
- hwmgr->platform_descriptor.overdriveVDDCStep = 0;
-
- if (hwmgr->platform_descriptor.overdriveLimit.engineClock > 0 \
- && hwmgr->platform_descriptor.overdriveLimit.memoryClock > 0) {
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_ACOverdriveSupport);
- }
-
- return 0;
-}
-
-/**
- * Private Function used during initialization.
- * Inspect the PowerPlay table for obvious signs of corruption.
- * @param hwmgr Pointer to the hardware manager.
- * @param powerplay_table Pointer to the PowerPlay Table.
- * @exception This implementation always returns 1.
- */
-static int init_thermal_controller(
- struct pp_hwmgr *hwmgr,
- const ATOM_Tonga_POWERPLAYTABLE *powerplay_table
- )
-{
- const PPTable_Generic_SubTable_Header *fan_table;
- ATOM_Tonga_Thermal_Controller *thermal_controller;
-
- thermal_controller = (ATOM_Tonga_Thermal_Controller *)
- (((unsigned long)powerplay_table) +
- le16_to_cpu(powerplay_table->usThermalControllerOffset));
- PP_ASSERT_WITH_CODE((0 != powerplay_table->usThermalControllerOffset),
- "Thermal controller table not set!", return -1);
-
- hwmgr->thermal_controller.ucType = thermal_controller->ucType;
- hwmgr->thermal_controller.ucI2cLine = thermal_controller->ucI2cLine;
- hwmgr->thermal_controller.ucI2cAddress = thermal_controller->ucI2cAddress;
-
- hwmgr->thermal_controller.fanInfo.bNoFan =
- (0 != (thermal_controller->ucFanParameters & ATOM_TONGA_PP_FANPARAMETERS_NOFAN));
-
- hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution =
- thermal_controller->ucFanParameters &
- ATOM_TONGA_PP_FANPARAMETERS_TACHOMETER_PULSES_PER_REVOLUTION_MASK;
-
- hwmgr->thermal_controller.fanInfo.ulMinRPM
- = thermal_controller->ucFanMinRPM * 100UL;
- hwmgr->thermal_controller.fanInfo.ulMaxRPM
- = thermal_controller->ucFanMaxRPM * 100UL;
-
- set_hw_cap(
- hwmgr,
- ATOM_TONGA_PP_THERMALCONTROLLER_NONE != hwmgr->thermal_controller.ucType,
- PHM_PlatformCaps_ThermalController
- );
-
- if (0 == powerplay_table->usFanTableOffset)
- return 0;
-
- fan_table = (const PPTable_Generic_SubTable_Header *)
- (((unsigned long)powerplay_table) +
- le16_to_cpu(powerplay_table->usFanTableOffset));
-
- PP_ASSERT_WITH_CODE((0 != powerplay_table->usFanTableOffset),
- "Fan table not set!", return -1);
- PP_ASSERT_WITH_CODE((0 < fan_table->ucRevId),
- "Unsupported fan table format!", return -1);
-
- hwmgr->thermal_controller.advanceFanControlParameters.ulCycleDelay
- = 100000;
- phm_cap_set(hwmgr->platform_descriptor.platformCaps,
- PHM_PlatformCaps_MicrocodeFanControl);
-
- if (fan_table->ucRevId < 8) {
- const ATOM_Tonga_Fan_Table *tonga_fan_table =
- (ATOM_Tonga_Fan_Table *)fan_table;
- hwmgr->thermal_controller.advanceFanControlParameters.ucTHyst
- = tonga_fan_table->ucTHyst;
- hwmgr->thermal_controller.advanceFanControlParameters.usTMin
- = tonga_fan_table->usTMin;
- hwmgr->thermal_controller.advanceFanControlParameters.usTMed
- = tonga_fan_table->usTMed;
- hwmgr->thermal_controller.advanceFanControlParameters.usTHigh
- = tonga_fan_table->usTHigh;
- hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin
- = tonga_fan_table->usPWMMin;
- hwmgr->thermal_controller.advanceFanControlParameters.usPWMMed
- = tonga_fan_table->usPWMMed;
- hwmgr->thermal_controller.advanceFanControlParameters.usPWMHigh
- = tonga_fan_table->usPWMHigh;
- hwmgr->thermal_controller.advanceFanControlParameters.usTMax
- = 10900; /* hard coded */
- hwmgr->thermal_controller.advanceFanControlParameters.usTMax
- = tonga_fan_table->usTMax;
- hwmgr->thermal_controller.advanceFanControlParameters.ucFanControlMode
- = tonga_fan_table->ucFanControlMode;
- hwmgr->thermal_controller.advanceFanControlParameters.usDefaultMaxFanPWM
- = tonga_fan_table->usFanPWMMax;
- hwmgr->thermal_controller.advanceFanControlParameters.usDefaultFanOutputSensitivity
- = 4836;
- hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity
- = tonga_fan_table->usFanOutputSensitivity;
- hwmgr->thermal_controller.advanceFanControlParameters.usDefaultMaxFanRPM
- = tonga_fan_table->usFanRPMMax;
- hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit
- = (tonga_fan_table->ulMinFanSCLKAcousticLimit / 100); /* PPTable stores it in 10Khz unit for 2 decimal places. SMC wants MHz. */
- hwmgr->thermal_controller.advanceFanControlParameters.ucTargetTemperature
- = tonga_fan_table->ucTargetTemperature;
- hwmgr->thermal_controller.advanceFanControlParameters.ucMinimumPWMLimit
- = tonga_fan_table->ucMinimumPWMLimit;
- } else {
- const ATOM_Fiji_Fan_Table *fiji_fan_table =
- (ATOM_Fiji_Fan_Table *)fan_table;
- hwmgr->thermal_controller.advanceFanControlParameters.ucTHyst
- = fiji_fan_table->ucTHyst;
- hwmgr->thermal_controller.advanceFanControlParameters.usTMin
- = fiji_fan_table->usTMin;
- hwmgr->thermal_controller.advanceFanControlParameters.usTMed
- = fiji_fan_table->usTMed;
- hwmgr->thermal_controller.advanceFanControlParameters.usTHigh
- = fiji_fan_table->usTHigh;
- hwmgr->thermal_controller.advanceFanControlParameters.usPWMMin
- = fiji_fan_table->usPWMMin;
- hwmgr->thermal_controller.advanceFanControlParameters.usPWMMed
- = fiji_fan_table->usPWMMed;
- hwmgr->thermal_controller.advanceFanControlParameters.usPWMHigh
- = fiji_fan_table->usPWMHigh;
- hwmgr->thermal_controller.advanceFanControlParameters.usTMax
- = fiji_fan_table->usTMax;
- hwmgr->thermal_controller.advanceFanControlParameters.ucFanControlMode
- = fiji_fan_table->ucFanControlMode;
- hwmgr->thermal_controller.advanceFanControlParameters.usDefaultMaxFanPWM
- = fiji_fan_table->usFanPWMMax;
- hwmgr->thermal_controller.advanceFanControlParameters.usDefaultFanOutputSensitivity
- = 4836;
- hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity
- = fiji_fan_table->usFanOutputSensitivity;
- hwmgr->thermal_controller.advanceFanControlParameters.usDefaultMaxFanRPM
- = fiji_fan_table->usFanRPMMax;
- hwmgr->thermal_controller.advanceFanControlParameters.ulMinFanSCLKAcousticLimit
- = (fiji_fan_table->ulMinFanSCLKAcousticLimit / 100); /* PPTable stores it in 10Khz unit for 2 decimal places. SMC wants MHz. */
- hwmgr->thermal_controller.advanceFanControlParameters.ucTargetTemperature
- = fiji_fan_table->ucTargetTemperature;
- hwmgr->thermal_controller.advanceFanControlParameters.ucMinimumPWMLimit
- = fiji_fan_table->ucMinimumPWMLimit;
-
- hwmgr->thermal_controller.advanceFanControlParameters.usFanGainEdge
- = fiji_fan_table->usFanGainEdge;
- hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHotspot
- = fiji_fan_table->usFanGainHotspot;
- hwmgr->thermal_controller.advanceFanControlParameters.usFanGainLiquid
- = fiji_fan_table->usFanGainLiquid;
- hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrVddc
- = fiji_fan_table->usFanGainVrVddc;
- hwmgr->thermal_controller.advanceFanControlParameters.usFanGainVrMvdd
- = fiji_fan_table->usFanGainVrMvdd;
- hwmgr->thermal_controller.advanceFanControlParameters.usFanGainPlx
- = fiji_fan_table->usFanGainPlx;
- hwmgr->thermal_controller.advanceFanControlParameters.usFanGainHbm
- = fiji_fan_table->usFanGainHbm;
- }
-
- return 0;
-}
-
-/**
- * Private Function used during initialization.
- * Inspect the PowerPlay table for obvious signs of corruption.
- * @param hwmgr Pointer to the hardware manager.
- * @param powerplay_table Pointer to the PowerPlay Table.
- * @exception 2 if the powerplay table is incorrect.
- */
-static int check_powerplay_tables(
- struct pp_hwmgr *hwmgr,
- const ATOM_Tonga_POWERPLAYTABLE *powerplay_table
- )
-{
- const ATOM_Tonga_State_Array *state_arrays;
-
- state_arrays = (ATOM_Tonga_State_Array *)(((unsigned long)powerplay_table) +
- le16_to_cpu(powerplay_table->usStateArrayOffset));
-
- PP_ASSERT_WITH_CODE((ATOM_Tonga_TABLE_REVISION_TONGA <=
- powerplay_table->sHeader.ucTableFormatRevision),
- "Unsupported PPTable format!", return -1);
- PP_ASSERT_WITH_CODE((0 != powerplay_table->usStateArrayOffset),
- "State table is not set!", return -1);
- PP_ASSERT_WITH_CODE((0 < powerplay_table->sHeader.usStructureSize),
- "Invalid PowerPlay Table!", return -1);
- PP_ASSERT_WITH_CODE((0 < state_arrays->ucNumEntries),
- "Invalid PowerPlay Table!", return -1);
-
- return 0;
-}
-
-int tonga_pp_tables_initialize(struct pp_hwmgr *hwmgr)
-{
- int result = 0;
- const ATOM_Tonga_POWERPLAYTABLE *powerplay_table;
-
- hwmgr->pptable = kzalloc(sizeof(struct phm_ppt_v1_information), GFP_KERNEL);
-
- PP_ASSERT_WITH_CODE((NULL != hwmgr->pptable),
- "Failed to allocate hwmgr->pptable!", return -ENOMEM);
-
- memset(hwmgr->pptable, 0x00, sizeof(struct phm_ppt_v1_information));
-
- powerplay_table = get_powerplay_table(hwmgr);
-
- PP_ASSERT_WITH_CODE((NULL != powerplay_table),
- "Missing PowerPlay Table!", return -1);
-
- result = check_powerplay_tables(hwmgr, powerplay_table);
-
- PP_ASSERT_WITH_CODE((result == 0),
- "check_powerplay_tables failed", return result);
-
- result = set_platform_caps(hwmgr,
- le32_to_cpu(powerplay_table->ulPlatformCaps));
-
- PP_ASSERT_WITH_CODE((result == 0),
- "set_platform_caps failed", return result);
-
- result = init_thermal_controller(hwmgr, powerplay_table);
-
- PP_ASSERT_WITH_CODE((result == 0),
- "init_thermal_controller failed", return result);
-
- result = init_over_drive_limits(hwmgr, powerplay_table);
-
- PP_ASSERT_WITH_CODE((result == 0),
- "init_over_drive_limits failed", return result);
-
- result = init_clock_voltage_dependency(hwmgr, powerplay_table);
-
- PP_ASSERT_WITH_CODE((result == 0),
- "init_clock_voltage_dependency failed", return result);
-
- result = init_dpm_2_parameters(hwmgr, powerplay_table);
-
- PP_ASSERT_WITH_CODE((result == 0),
- "init_dpm_2_parameters failed", return result);
-
- return result;
-}
-
-int tonga_pp_tables_uninitialize(struct pp_hwmgr *hwmgr)
-{
- struct phm_ppt_v1_information *pp_table_information =
- (struct phm_ppt_v1_information *)(hwmgr->pptable);
-
- kfree(pp_table_information->vdd_dep_on_sclk);
- pp_table_information->vdd_dep_on_sclk = NULL;
-
- kfree(pp_table_information->vdd_dep_on_mclk);
- pp_table_information->vdd_dep_on_mclk = NULL;
-
- kfree(pp_table_information->valid_mclk_values);
- pp_table_information->valid_mclk_values = NULL;
-
- kfree(pp_table_information->valid_sclk_values);
- pp_table_information->valid_sclk_values = NULL;
-
- kfree(pp_table_information->vddc_lookup_table);
- pp_table_information->vddc_lookup_table = NULL;
-
- kfree(pp_table_information->vddgfx_lookup_table);
- pp_table_information->vddgfx_lookup_table = NULL;
-
- kfree(pp_table_information->mm_dep_table);
- pp_table_information->mm_dep_table = NULL;
-
- kfree(pp_table_information->cac_dtp_table);
- pp_table_information->cac_dtp_table = NULL;
-
- kfree(hwmgr->dyn_state.cac_dtp_table);
- hwmgr->dyn_state.cac_dtp_table = NULL;
-
- kfree(pp_table_information->ppm_parameter_table);
- pp_table_information->ppm_parameter_table = NULL;
-
- kfree(pp_table_information->pcie_table);
- pp_table_information->pcie_table = NULL;
-
- kfree(hwmgr->pptable);
- hwmgr->pptable = NULL;
-
- return 0;
-}
-
-const struct pp_table_func tonga_pptable_funcs = {
- .pptable_init = tonga_pp_tables_initialize,
- .pptable_fini = tonga_pp_tables_uninitialize,
-};
-
-int tonga_get_number_of_powerplay_table_entries(struct pp_hwmgr *hwmgr)
-{
- const ATOM_Tonga_State_Array * state_arrays;
- const ATOM_Tonga_POWERPLAYTABLE *pp_table = get_powerplay_table(hwmgr);
-
- PP_ASSERT_WITH_CODE((NULL != pp_table),
- "Missing PowerPlay Table!", return -1);
- PP_ASSERT_WITH_CODE((pp_table->sHeader.ucTableFormatRevision >=
- ATOM_Tonga_TABLE_REVISION_TONGA),
- "Incorrect PowerPlay table revision!", return -1);
-
- state_arrays = (ATOM_Tonga_State_Array *)(((unsigned long)pp_table) +
- le16_to_cpu(pp_table->usStateArrayOffset));
-
- return (uint32_t)(state_arrays->ucNumEntries);
-}
-
-/**
-* Private function to convert flags stored in the BIOS to software flags in PowerPlay.
-*/
-static uint32_t make_classification_flags(struct pp_hwmgr *hwmgr,
- uint16_t classification, uint16_t classification2)
-{
- uint32_t result = 0;
-
- if (classification & ATOM_PPLIB_CLASSIFICATION_BOOT)
- result |= PP_StateClassificationFlag_Boot;
-
- if (classification & ATOM_PPLIB_CLASSIFICATION_THERMAL)
- result |= PP_StateClassificationFlag_Thermal;
-
- if (classification & ATOM_PPLIB_CLASSIFICATION_LIMITEDPOWERSOURCE)
- result |= PP_StateClassificationFlag_LimitedPowerSource;
-
- if (classification & ATOM_PPLIB_CLASSIFICATION_REST)
- result |= PP_StateClassificationFlag_Rest;
-
- if (classification & ATOM_PPLIB_CLASSIFICATION_FORCED)
- result |= PP_StateClassificationFlag_Forced;
-
- if (classification & ATOM_PPLIB_CLASSIFICATION_ACPI)
- result |= PP_StateClassificationFlag_ACPI;
-
- if (classification2 & ATOM_PPLIB_CLASSIFICATION2_LIMITEDPOWERSOURCE_2)
- result |= PP_StateClassificationFlag_LimitedPowerSource_2;
-
- return result;
-}
-
-/**
-* Create a Power State out of an entry in the PowerPlay table.
-* This function is called by the hardware back-end.
-* @param hwmgr Pointer to the hardware manager.
-* @param entry_index The index of the entry to be extracted from the table.
-* @param power_state The address of the PowerState instance being created.
-* @return -1 if the entry cannot be retrieved.
-*/
-int tonga_get_powerplay_table_entry(struct pp_hwmgr *hwmgr,
- uint32_t entry_index, struct pp_power_state *power_state,
- int (*call_back_func)(struct pp_hwmgr *, void *,
- struct pp_power_state *, void *, uint32_t))
-{
- int result = 0;
- const ATOM_Tonga_State_Array * state_arrays;
- const ATOM_Tonga_State *state_entry;
- const ATOM_Tonga_POWERPLAYTABLE *pp_table = get_powerplay_table(hwmgr);
-
- PP_ASSERT_WITH_CODE((NULL != pp_table), "Missing PowerPlay Table!", return -1;);
- power_state->classification.bios_index = entry_index;
-
- if (pp_table->sHeader.ucTableFormatRevision >=
- ATOM_Tonga_TABLE_REVISION_TONGA) {
- state_arrays = (ATOM_Tonga_State_Array *)(((unsigned long)pp_table) +
- le16_to_cpu(pp_table->usStateArrayOffset));
-
- PP_ASSERT_WITH_CODE((0 < pp_table->usStateArrayOffset),
- "Invalid PowerPlay Table State Array Offset.", return -1);
- PP_ASSERT_WITH_CODE((0 < state_arrays->ucNumEntries),
- "Invalid PowerPlay Table State Array.", return -1);
- PP_ASSERT_WITH_CODE((entry_index <= state_arrays->ucNumEntries),
- "Invalid PowerPlay Table State Array Entry.", return -1);
-
- state_entry = &(state_arrays->states[entry_index]);
-
- result = call_back_func(hwmgr, (void *)state_entry, power_state,
- (void *)pp_table,
- make_classification_flags(hwmgr,
- le16_to_cpu(state_entry->usClassification),
- le16_to_cpu(state_entry->usClassification2)));
- }
-
- if (!result && (power_state->classification.flags &
- PP_StateClassificationFlag_Boot))
- result = hwmgr->hwmgr_func->patch_boot_state(hwmgr, &(power_state->hardware));
-
- return result;
-}
+++ /dev/null
-/*
- * Copyright 2015 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- */
-#ifndef TONGA_PROCESSPPTABLES_H
-#define TONGA_PROCESSPPTABLES_H
-
-#include "hwmgr.h"
-
-extern const struct pp_table_func tonga_pptable_funcs;
-extern int tonga_get_number_of_powerplay_table_entries(struct pp_hwmgr *hwmgr);
-extern int tonga_get_powerplay_table_entry(struct pp_hwmgr *hwmgr, uint32_t entry_index,
- struct pp_power_state *power_state, int (*call_back_func)(struct pp_hwmgr *, void *,
- struct pp_power_state *, void *, uint32_t));
-
-#endif
-
struct cgs_device *device;
uint32_t chip_family;
uint32_t chip_id;
- uint32_t rev_id;
- bool powercontainment_enabled;
};
+
enum amd_pp_display_config_type{
AMD_PP_DisplayConfigType_None = 0,
AMD_PP_DisplayConfigType_DP54 ,
extern int phm_setup_asic(struct pp_hwmgr *hwmgr);
extern int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr);
extern int phm_disable_dynamic_state_management(struct pp_hwmgr *hwmgr);
-extern void phm_init_dynamic_caps(struct pp_hwmgr *hwmgr);
extern bool phm_is_hw_access_blocked(struct pp_hwmgr *hwmgr);
extern int phm_block_hw_access(struct pp_hwmgr *hwmgr, bool block);
extern int phm_set_power_state(struct pp_hwmgr *hwmgr,
#include "hwmgr_ppt.h"
#include "ppatomctrl.h"
#include "hwmgr_ppt.h"
+#include "power_state.h"
struct pp_instance;
struct pp_hwmgr;
-struct pp_hw_power_state;
-struct pp_power_state;
-struct PP_VCEState;
struct phm_fan_speed_info;
struct pp_atomctrl_voltage_table;
+extern int amdgpu_powercontainment;
+extern int amdgpu_sclk_deep_sleep_en;
+extern unsigned amdgpu_pp_feature_mask;
+
+#define VOLTAGE_SCALE 4
+
+uint8_t convert_to_vid(uint16_t vddc);
enum DISPLAY_GAP {
DISPLAY_GAP_VBLANK_OR_WM = 0, /* Wait for vblank or MCHG watermark. */
};
typedef enum DISPLAY_GAP DISPLAY_GAP;
-
struct vi_dpm_level {
bool enabled;
uint32_t value;
#define PCIE_PERF_REQ_GEN2 3
#define PCIE_PERF_REQ_GEN3 4
+enum PP_FEATURE_MASK {
+ PP_SCLK_DPM_MASK = 0x1,
+ PP_MCLK_DPM_MASK = 0x2,
+ PP_PCIE_DPM_MASK = 0x4,
+ PP_SCLK_DEEP_SLEEP_MASK = 0x8,
+ PP_POWER_CONTAINMENT_MASK = 0x10,
+ PP_UVD_HANDSHAKE_MASK = 0x20,
+ PP_SMC_VOLTAGE_CONTROL_MASK = 0x40,
+ PP_VBI_TIME_SUPPORT_MASK = 0x80,
+ PP_ULV_MASK = 0x100,
+ PP_ENABLE_GFX_CG_THRU_SMU = 0x200
+};
+
enum PHM_BackEnd_Magic {
PHM_Dummy_Magic = 0xAA5555AA,
PHM_RV770_Magic = 0xDCBAABCD,
int (*pptable_get_vce_state_table_entry)(
struct pp_hwmgr *hwmgr,
unsigned long i,
- struct PP_VCEState *vce_state,
+ struct pp_vce_state *vce_state,
void **clock_info,
unsigned long *flag);
};
uint32_t NB;
};
+#define PP_MAX_VCE_LEVELS 6
+
+enum PP_VCE_LEVEL {
+ PP_VCE_LEVEL_AC_ALL = 0, /* AC, All cases */
+ PP_VCE_LEVEL_DC_EE = 1, /* DC, entropy encoding */
+ PP_VCE_LEVEL_DC_LL_LOW = 2, /* DC, low latency queue, res <= 720 */
+ PP_VCE_LEVEL_DC_LL_HIGH = 3, /* DC, low latency queue, 1080 >= res > 720 */
+ PP_VCE_LEVEL_DC_GP_LOW = 4, /* DC, general purpose queue, res <= 720 */
+ PP_VCE_LEVEL_DC_GP_HIGH = 5, /* DC, general purpose queue, 1080 >= res > 720 */
+};
+
+
+enum PP_TABLE_VERSION {
+ PP_TABLE_V0 = 0,
+ PP_TABLE_V1,
+ PP_TABLE_V2,
+ PP_TABLE_MAX
+};
+
/**
* The main hardware manager structure.
*/
struct pp_hwmgr {
uint32_t chip_family;
uint32_t chip_id;
- uint32_t hw_revision;
- uint32_t sub_sys_id;
- uint32_t sub_vendor_id;
+ uint32_t pp_table_version;
void *device;
struct pp_smumgr *smumgr;
const void *soft_pp_table;
uint32_t soft_pp_table_size;
void *hardcode_pp_table;
bool need_pp_table_upload;
+
+ struct pp_vce_state vce_states[PP_MAX_VCE_LEVELS];
+ uint32_t num_vce_state_tables;
+
enum amd_dpm_forced_level dpm_level;
bool block_hw_access;
struct phm_gfx_arbiter gfx_arbiter;
uint32_t num_ps;
struct pp_thermal_controller_info thermal_controller;
bool fan_ctrl_is_in_default_mode;
- bool powercontainment_enabled;
uint32_t fan_ctrl_default_mode;
uint32_t tmin;
struct phm_microcode_version_info microcode_version_info;
struct pp_power_state *boot_ps;
struct pp_power_state *uvd_ps;
struct amd_pp_display_configuration display_config;
+ uint32_t feature_mask;
};
extern int phm_wait_on_register(struct pp_hwmgr *hwmgr, uint32_t index,
uint32_t value, uint32_t mask);
-extern int phm_wait_for_register_unequal(struct pp_hwmgr *hwmgr,
- uint32_t index, uint32_t value, uint32_t mask);
-extern uint32_t phm_read_indirect_register(struct pp_hwmgr *hwmgr,
- uint32_t indirect_port, uint32_t index);
-
-extern void phm_write_indirect_register(struct pp_hwmgr *hwmgr,
- uint32_t indirect_port,
- uint32_t index,
- uint32_t value);
extern void phm_wait_on_indirect_register(struct pp_hwmgr *hwmgr,
uint32_t indirect_port,
uint32_t value,
uint32_t mask);
-extern void phm_wait_for_indirect_register_unequal(
- struct pp_hwmgr *hwmgr,
- uint32_t indirect_port,
- uint32_t index,
- uint32_t value,
- uint32_t mask);
+
extern bool phm_cf_want_uvd_power_gating(struct pp_hwmgr *hwmgr);
extern bool phm_cf_want_vce_power_gating(struct pp_hwmgr *hwmgr);
extern int phm_reset_single_dpm_table(void *table, uint32_t count, int max);
extern void phm_setup_pcie_table_entry(void *table, uint32_t index, uint32_t pcie_gen, uint32_t pcie_lanes);
extern int32_t phm_get_dpm_level_enable_mask_value(void *table);
+extern uint8_t phm_get_voltage_id(struct pp_atomctrl_voltage_table *voltage_table,
+ uint32_t voltage);
extern uint8_t phm_get_voltage_index(struct phm_ppt_v1_voltage_lookup_table *lookup_table, uint16_t voltage);
extern uint16_t phm_find_closest_vddci(struct pp_atomctrl_voltage_table *vddci_table, uint16_t vddci);
extern int phm_find_boot_level(void *table, uint32_t value, uint32_t *boot_level);
extern uint32_t phm_get_lowest_enabled_level(struct pp_hwmgr *hwmgr, uint32_t mask);
extern void phm_apply_dal_min_voltage_request(struct pp_hwmgr *hwmgr);
+extern int phm_get_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
+ uint32_t sclk, uint16_t id, uint16_t *voltage);
+
#define PHM_ENTIRE_REGISTER_MASK 0xFFFFFFFFU
#define PHM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT
PHM_FIELD_SHIFT(reg, field))
-#define PHM_WAIT_REGISTER_GIVEN_INDEX(hwmgr, index, value, mask) \
- phm_wait_on_register(hwmgr, index, value, mask)
-
-#define PHM_WAIT_REGISTER_UNEQUAL_GIVEN_INDEX(hwmgr, index, value, mask) \
- phm_wait_for_register_unequal(hwmgr, index, value, mask)
-
-#define PHM_WAIT_INDIRECT_REGISTER_GIVEN_INDEX(hwmgr, port, index, value, mask) \
- phm_wait_on_indirect_register(hwmgr, mm##port##_INDEX, index, value, mask)
-
-#define PHM_WAIT_INDIRECT_REGISTER_UNEQUAL_GIVEN_INDEX(hwmgr, port, index, value, mask) \
- phm_wait_for_indirect_register_unequal(hwmgr, mm##port##_INDEX, index, value, mask)
-
-#define PHM_WAIT_VFPF_INDIRECT_REGISTER_GIVEN_INDEX(hwmgr, port, index, value, mask) \
- phm_wait_on_indirect_register(hwmgr, mm##port##_INDEX_0, index, value, mask)
-
-#define PHM_WAIT_VFPF_INDIRECT_REGISTER_UNEQUAL_GIVEN_INDEX(hwmgr, port, index, value, mask) \
- phm_wait_for_indirect_register_unequal(hwmgr, mm##port##_INDEX_0, index, value, mask)
-
-/* Operations on named registers. */
-
-#define PHM_WAIT_REGISTER(hwmgr, reg, value, mask) \
- PHM_WAIT_REGISTER_GIVEN_INDEX(hwmgr, mm##reg, value, mask)
-
-#define PHM_WAIT_REGISTER_UNEQUAL(hwmgr, reg, value, mask) \
- PHM_WAIT_REGISTER_UNEQUAL_GIVEN_INDEX(hwmgr, mm##reg, value, mask)
-
-#define PHM_WAIT_INDIRECT_REGISTER(hwmgr, port, reg, value, mask) \
- PHM_WAIT_INDIRECT_REGISTER_GIVEN_INDEX(hwmgr, port, ix##reg, value, mask)
-
-#define PHM_WAIT_INDIRECT_REGISTER_UNEQUAL(hwmgr, port, reg, value, mask) \
- PHM_WAIT_INDIRECT_REGISTER_UNEQUAL_GIVEN_INDEX(hwmgr, port, ix##reg, value, mask)
-
-#define PHM_WAIT_VFPF_INDIRECT_REGISTER(hwmgr, port, reg, value, mask) \
- PHM_WAIT_VFPF_INDIRECT_REGISTER_GIVEN_INDEX(hwmgr, port, ix##reg, value, mask)
-
-#define PHM_WAIT_VFPF_INDIRECT_REGISTER_UNEQUAL(hwmgr, port, reg, value, mask) \
- PHM_WAIT_VFPF_INDIRECT_REGISTER_UNEQUAL_GIVEN_INDEX(hwmgr, port, ix##reg, value, mask)
-
/* Operations on named fields. */
#define PHM_READ_FIELD(device, reg, field) \
PHM_SET_FIELD(cgs_read_ind_register(device, port, ix##reg), \
reg, field, fieldval))
-#define PHM_WAIT_FIELD(hwmgr, reg, field, fieldval) \
- PHM_WAIT_REGISTER(hwmgr, reg, (fieldval) \
- << PHM_FIELD_SHIFT(reg, field), PHM_FIELD_MASK(reg, field))
-
-#define PHM_WAIT_INDIRECT_FIELD(hwmgr, port, reg, field, fieldval) \
- PHM_WAIT_INDIRECT_REGISTER(hwmgr, port, reg, (fieldval) \
- << PHM_FIELD_SHIFT(reg, field), PHM_FIELD_MASK(reg, field))
-
-#define PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, port, reg, field, fieldval) \
- PHM_WAIT_VFPF_INDIRECT_REGISTER(hwmgr, port, reg, (fieldval) \
- << PHM_FIELD_SHIFT(reg, field), PHM_FIELD_MASK(reg, field))
+#define PHM_WAIT_INDIRECT_REGISTER_GIVEN_INDEX(hwmgr, port, index, value, mask) \
+ phm_wait_on_indirect_register(hwmgr, mm##port##_INDEX, index, value, mask)
-#define PHM_WAIT_FIELD_UNEQUAL(hwmgr, reg, field, fieldval) \
- PHM_WAIT_REGISTER_UNEQUAL(hwmgr, reg, (fieldval) \
- << PHM_FIELD_SHIFT(reg, field), PHM_FIELD_MASK(reg, field))
-#define PHM_WAIT_INDIRECT_FIELD_UNEQUAL(hwmgr, port, reg, field, fieldval) \
- PHM_WAIT_INDIRECT_REGISTER_UNEQUAL(hwmgr, port, reg, (fieldval) \
- << PHM_FIELD_SHIFT(reg, field), PHM_FIELD_MASK(reg, field))
+#define PHM_WAIT_INDIRECT_REGISTER(hwmgr, port, reg, value, mask) \
+ PHM_WAIT_INDIRECT_REGISTER_GIVEN_INDEX(hwmgr, port, ix##reg, value, mask)
-#define PHM_WAIT_VFPF_INDIRECT_FIELD_UNEQUAL(hwmgr, port, reg, field, fieldval) \
- PHM_WAIT_VFPF_INDIRECT_REGISTER_UNEQUAL(hwmgr, port, reg, (fieldval) \
+#define PHM_WAIT_INDIRECT_FIELD(hwmgr, port, reg, field, fieldval) \
+ PHM_WAIT_INDIRECT_REGISTER(hwmgr, port, reg, (fieldval) \
<< PHM_FIELD_SHIFT(reg, field), PHM_FIELD_MASK(reg, field))
-/* Operations on arrays of registers & fields. */
-
-#define PHM_READ_ARRAY_REGISTER(device, reg, offset) \
- cgs_read_register(device, mm##reg + (offset))
-
-#define PHM_WRITE_ARRAY_REGISTER(device, reg, offset, value) \
- cgs_write_register(device, mm##reg + (offset), value)
-
-#define PHM_WAIT_ARRAY_REGISTER(hwmgr, reg, offset, value, mask) \
- PHM_WAIT_REGISTER_GIVEN_INDEX(hwmgr, mm##reg + (offset), value, mask)
-
-#define PHM_WAIT_ARRAY_REGISTER_UNEQUAL(hwmgr, reg, offset, value, mask) \
- PHM_WAIT_REGISTER_UNEQUAL_GIVEN_INDEX(hwmgr, mm##reg + (offset), value, mask)
-
-#define PHM_READ_ARRAY_FIELD(hwmgr, reg, offset, field) \
- PHM_GET_FIELD(PHM_READ_ARRAY_REGISTER(hwmgr->device, reg, offset), reg, field)
-
-#define PHM_WRITE_ARRAY_FIELD(hwmgr, reg, offset, field, fieldvalue) \
- PHM_WRITE_ARRAY_REGISTER(hwmgr->device, reg, offset, \
- PHM_SET_FIELD(PHM_READ_ARRAY_REGISTER(hwmgr->device, reg, offset), \
- reg, field, fieldvalue))
-
-#define PHM_WAIT_ARRAY_FIELD(hwmgr, reg, offset, field, fieldvalue) \
- PHM_WAIT_REGISTER_GIVEN_INDEX(hwmgr, mm##reg + (offset), \
- (fieldvalue) << PHM_FIELD_SHIFT(reg, field), \
- PHM_FIELD_MASK(reg, field))
-
-#define PHM_WAIT_ARRAY_FIELD_UNEQUAL(hwmgr, reg, offset, field, fieldvalue) \
- PHM_WAIT_REGISTER_UNEQUAL_GIVEN_INDEX(hwmgr, mm##reg + (offset), \
- (fieldvalue) << PHM_FIELD_SHIFT(reg, field), \
- PHM_FIELD_MASK(reg, field))
#endif /* _HWMGR_H_ */
/*Structure to hold a VCE state entry*/
-struct PP_VCEState {
+struct pp_vce_state {
uint32_t evclk;
uint32_t ecclk;
uint32_t sclk;
PP_MMProfilingState_Stopped
};
-struct PP_Clock_Engine_Request {
- unsigned long clientType;
- unsigned long ctxid;
+struct pp_clock_engine_request {
+ unsigned long client_type;
+ unsigned long ctx_id;
uint64_t context_handle;
unsigned long sclk;
- unsigned long sclkHardMin;
+ unsigned long sclk_hard_min;
unsigned long mclk;
unsigned long iclk;
unsigned long evclk;
unsigned long ecclk;
- unsigned long ecclkHardMin;
+ unsigned long ecclk_hard_min;
unsigned long vclk;
unsigned long dclk;
- unsigned long samclk;
- unsigned long acpclk;
- unsigned long sclkOverdrive;
- unsigned long mclkOverdrive;
+ unsigned long sclk_over_drive;
+ unsigned long mclk_over_drive;
unsigned long sclk_threshold;
unsigned long flag;
unsigned long vclk_ceiling;
unsigned long dclk_ceiling;
unsigned long num_cus;
- unsigned long pmflag;
- enum PP_MMProfilingState MMProfilingState;
+ unsigned long pm_flag;
+ enum PP_MMProfilingState mm_profiling_state;
};
#endif
} while (0)
+#define GET_FLEXIBLE_ARRAY_MEMBER_ADDR(type, member, ptr, n) \
+ (type *)((char *)&(ptr)->member + (sizeof(type) * (n)))
+
#endif /* PP_DEBUG_H */
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef SMU71_H
+#define SMU71_H
+
+#if !defined(SMC_MICROCODE)
+#pragma pack(push, 1)
+#endif
+
+#define SMU__NUM_PCIE_DPM_LEVELS 8
+#define SMU__NUM_SCLK_DPM_STATE 8
+#define SMU__NUM_MCLK_DPM_LEVELS 4
+#define SMU__VARIANT__ICELAND 1
+#define SMU__DGPU_ONLY 1
+#define SMU__DYNAMIC_MCARB_SETTINGS 1
+
+enum SID_OPTION {
+ SID_OPTION_HI,
+ SID_OPTION_LO,
+ SID_OPTION_COUNT
+};
+
+typedef struct {
+ uint32_t high;
+ uint32_t low;
+} data_64_t;
+
+typedef struct {
+ data_64_t high;
+ data_64_t low;
+} data_128_t;
+
+#define SMU7_CONTEXT_ID_SMC 1
+#define SMU7_CONTEXT_ID_VBIOS 2
+
+#define SMU71_MAX_LEVELS_VDDC 8
+#define SMU71_MAX_LEVELS_VDDCI 4
+#define SMU71_MAX_LEVELS_MVDD 4
+#define SMU71_MAX_LEVELS_VDDNB 8
+
+#define SMU71_MAX_LEVELS_GRAPHICS SMU__NUM_SCLK_DPM_STATE
+#define SMU71_MAX_LEVELS_MEMORY SMU__NUM_MCLK_DPM_LEVELS
+#define SMU71_MAX_LEVELS_GIO SMU__NUM_LCLK_DPM_LEVELS
+#define SMU71_MAX_LEVELS_LINK SMU__NUM_PCIE_DPM_LEVELS
+#define SMU71_MAX_ENTRIES_SMIO 32
+
+#define DPM_NO_LIMIT 0
+#define DPM_NO_UP 1
+#define DPM_GO_DOWN 2
+#define DPM_GO_UP 3
+
+#define SMU7_FIRST_DPM_GRAPHICS_LEVEL 0
+#define SMU7_FIRST_DPM_MEMORY_LEVEL 0
+
+#define GPIO_CLAMP_MODE_VRHOT 1
+#define GPIO_CLAMP_MODE_THERM 2
+#define GPIO_CLAMP_MODE_DC 4
+
+#define SCRATCH_B_TARG_PCIE_INDEX_SHIFT 0
+#define SCRATCH_B_TARG_PCIE_INDEX_MASK (0x7<<SCRATCH_B_TARG_PCIE_INDEX_SHIFT)
+#define SCRATCH_B_CURR_PCIE_INDEX_SHIFT 3
+#define SCRATCH_B_CURR_PCIE_INDEX_MASK (0x7<<SCRATCH_B_CURR_PCIE_INDEX_SHIFT)
+#define SCRATCH_B_TARG_UVD_INDEX_SHIFT 6
+#define SCRATCH_B_TARG_UVD_INDEX_MASK (0x7<<SCRATCH_B_TARG_UVD_INDEX_SHIFT)
+#define SCRATCH_B_CURR_UVD_INDEX_SHIFT 9
+#define SCRATCH_B_CURR_UVD_INDEX_MASK (0x7<<SCRATCH_B_CURR_UVD_INDEX_SHIFT)
+#define SCRATCH_B_TARG_VCE_INDEX_SHIFT 12
+#define SCRATCH_B_TARG_VCE_INDEX_MASK (0x7<<SCRATCH_B_TARG_VCE_INDEX_SHIFT)
+#define SCRATCH_B_CURR_VCE_INDEX_SHIFT 15
+#define SCRATCH_B_CURR_VCE_INDEX_MASK (0x7<<SCRATCH_B_CURR_VCE_INDEX_SHIFT)
+#define SCRATCH_B_TARG_ACP_INDEX_SHIFT 18
+#define SCRATCH_B_TARG_ACP_INDEX_MASK (0x7<<SCRATCH_B_TARG_ACP_INDEX_SHIFT)
+#define SCRATCH_B_CURR_ACP_INDEX_SHIFT 21
+#define SCRATCH_B_CURR_ACP_INDEX_MASK (0x7<<SCRATCH_B_CURR_ACP_INDEX_SHIFT)
+#define SCRATCH_B_TARG_SAMU_INDEX_SHIFT 24
+#define SCRATCH_B_TARG_SAMU_INDEX_MASK (0x7<<SCRATCH_B_TARG_SAMU_INDEX_SHIFT)
+#define SCRATCH_B_CURR_SAMU_INDEX_SHIFT 27
+#define SCRATCH_B_CURR_SAMU_INDEX_MASK (0x7<<SCRATCH_B_CURR_SAMU_INDEX_SHIFT)
+
+
+#if defined SMU__DGPU_ONLY
+#define SMU71_DTE_ITERATIONS 5
+#define SMU71_DTE_SOURCES 3
+#define SMU71_DTE_SINKS 1
+#define SMU71_NUM_CPU_TES 0
+#define SMU71_NUM_GPU_TES 1
+#define SMU71_NUM_NON_TES 2
+
+#endif
+
+#if defined SMU__FUSION_ONLY
+#define SMU7_DTE_ITERATIONS 5
+#define SMU7_DTE_SOURCES 5
+#define SMU7_DTE_SINKS 3
+#define SMU7_NUM_CPU_TES 2
+#define SMU7_NUM_GPU_TES 1
+#define SMU7_NUM_NON_TES 2
+
+#endif
+
+struct SMU71_PIDController
+{
+ uint32_t Ki;
+ int32_t LFWindupUpperLim;
+ int32_t LFWindupLowerLim;
+ uint32_t StatePrecision;
+ uint32_t LfPrecision;
+ uint32_t LfOffset;
+ uint32_t MaxState;
+ uint32_t MaxLfFraction;
+ uint32_t StateShift;
+};
+
+typedef struct SMU71_PIDController SMU71_PIDController;
+
+struct SMU7_LocalDpmScoreboard
+{
+ uint32_t PercentageBusy;
+
+ int32_t PIDError;
+ int32_t PIDIntegral;
+ int32_t PIDOutput;
+
+ uint32_t SigmaDeltaAccum;
+ uint32_t SigmaDeltaOutput;
+ uint32_t SigmaDeltaLevel;
+
+ uint32_t UtilizationSetpoint;
+
+ uint8_t TdpClampMode;
+ uint8_t TdcClampMode;
+ uint8_t ThermClampMode;
+ uint8_t VoltageBusy;
+
+ int8_t CurrLevel;
+ int8_t TargLevel;
+ uint8_t LevelChangeInProgress;
+ uint8_t UpHyst;
+
+ uint8_t DownHyst;
+ uint8_t VoltageDownHyst;
+ uint8_t DpmEnable;
+ uint8_t DpmRunning;
+
+ uint8_t DpmForce;
+ uint8_t DpmForceLevel;
+ uint8_t DisplayWatermark;
+ uint8_t McArbIndex;
+
+ uint32_t MinimumPerfSclk;
+
+ uint8_t AcpiReq;
+ uint8_t AcpiAck;
+ uint8_t GfxClkSlow;
+ uint8_t GpioClampMode;
+
+ uint8_t FpsFilterWeight;
+ uint8_t EnabledLevelsChange;
+ uint8_t DteClampMode;
+ uint8_t FpsClampMode;
+
+ uint16_t LevelResidencyCounters [SMU71_MAX_LEVELS_GRAPHICS];
+ uint16_t LevelSwitchCounters [SMU71_MAX_LEVELS_GRAPHICS];
+
+ void (*TargetStateCalculator)(uint8_t);
+ void (*SavedTargetStateCalculator)(uint8_t);
+
+ uint16_t AutoDpmInterval;
+ uint16_t AutoDpmRange;
+
+ uint8_t FpsEnabled;
+ uint8_t MaxPerfLevel;
+ uint8_t AllowLowClkInterruptToHost;
+ uint8_t FpsRunning;
+
+ uint32_t MaxAllowedFrequency;
+};
+
+typedef struct SMU7_LocalDpmScoreboard SMU7_LocalDpmScoreboard;
+
+#define SMU7_MAX_VOLTAGE_CLIENTS 12
+
+struct SMU7_VoltageScoreboard
+{
+ uint16_t CurrentVoltage;
+ uint16_t HighestVoltage;
+ uint16_t MaxVid;
+ uint8_t HighestVidOffset;
+ uint8_t CurrentVidOffset;
+#if defined (SMU__DGPU_ONLY)
+ uint8_t CurrentPhases;
+ uint8_t HighestPhases;
+#else
+ uint8_t AvsOffset;
+ uint8_t AvsOffsetApplied;
+#endif
+ uint8_t ControllerBusy;
+ uint8_t CurrentVid;
+ uint16_t RequestedVoltage[SMU7_MAX_VOLTAGE_CLIENTS];
+#if defined (SMU__DGPU_ONLY)
+ uint8_t RequestedPhases[SMU7_MAX_VOLTAGE_CLIENTS];
+#endif
+ uint8_t EnabledRequest[SMU7_MAX_VOLTAGE_CLIENTS];
+ uint8_t TargetIndex;
+ uint8_t Delay;
+ uint8_t ControllerEnable;
+ uint8_t ControllerRunning;
+ uint16_t CurrentStdVoltageHiSidd;
+ uint16_t CurrentStdVoltageLoSidd;
+#if defined (SMU__DGPU_ONLY)
+ uint16_t RequestedVddci;
+ uint16_t CurrentVddci;
+ uint16_t HighestVddci;
+ uint8_t CurrentVddciVid;
+ uint8_t TargetVddciIndex;
+#endif
+};
+
+typedef struct SMU7_VoltageScoreboard SMU7_VoltageScoreboard;
+
+// -------------------------------------------------------------------------------------------------------------------------
+#define SMU7_MAX_PCIE_LINK_SPEEDS 3 /* 0:Gen1 1:Gen2 2:Gen3 */
+
+struct SMU7_PCIeLinkSpeedScoreboard
+{
+ uint8_t DpmEnable;
+ uint8_t DpmRunning;
+ uint8_t DpmForce;
+ uint8_t DpmForceLevel;
+
+ uint8_t CurrentLinkSpeed;
+ uint8_t EnabledLevelsChange;
+ uint16_t AutoDpmInterval;
+
+ uint16_t AutoDpmRange;
+ uint16_t AutoDpmCount;
+
+ uint8_t DpmMode;
+ uint8_t AcpiReq;
+ uint8_t AcpiAck;
+ uint8_t CurrentLinkLevel;
+
+};
+
+typedef struct SMU7_PCIeLinkSpeedScoreboard SMU7_PCIeLinkSpeedScoreboard;
+
+// -------------------------------------------------------- CAC table ------------------------------------------------------
+#define SMU7_LKGE_LUT_NUM_OF_TEMP_ENTRIES 16
+#define SMU7_LKGE_LUT_NUM_OF_VOLT_ENTRIES 16
+
+#define SMU7_SCALE_I 7
+#define SMU7_SCALE_R 12
+
+struct SMU7_PowerScoreboard
+{
+ uint16_t MinVoltage;
+ uint16_t MaxVoltage;
+
+ uint32_t AvgGpuPower;
+
+ uint16_t VddcLeakagePower[SID_OPTION_COUNT];
+ uint16_t VddcSclkConstantPower[SID_OPTION_COUNT];
+ uint16_t VddcSclkDynamicPower[SID_OPTION_COUNT];
+ uint16_t VddcNonSclkDynamicPower[SID_OPTION_COUNT];
+ uint16_t VddcTotalPower[SID_OPTION_COUNT];
+ uint16_t VddcTotalCurrent[SID_OPTION_COUNT];
+ uint16_t VddcLoadVoltage[SID_OPTION_COUNT];
+ uint16_t VddcNoLoadVoltage[SID_OPTION_COUNT];
+
+ uint16_t DisplayPhyPower;
+ uint16_t PciePhyPower;
+
+ uint16_t VddciTotalPower;
+ uint16_t Vddr1TotalPower;
+
+ uint32_t RocPower;
+
+ uint32_t last_power;
+ uint32_t enableWinAvg;
+
+ uint32_t lkg_acc;
+ uint16_t VoltLkgeScaler;
+ uint16_t TempLkgeScaler;
+
+ uint32_t uvd_cac_dclk;
+ uint32_t uvd_cac_vclk;
+ uint32_t vce_cac_eclk;
+ uint32_t samu_cac_samclk;
+ uint32_t display_cac_dispclk;
+ uint32_t acp_cac_aclk;
+ uint32_t unb_cac;
+
+ uint32_t WinTime;
+
+ uint16_t GpuPwr_MAWt;
+ uint16_t FilteredVddcTotalPower;
+
+ uint8_t CalculationRepeats;
+ uint8_t WaterfallUp;
+ uint8_t WaterfallDown;
+ uint8_t WaterfallLimit;
+};
+
+typedef struct SMU7_PowerScoreboard SMU7_PowerScoreboard;
+
+// --------------------------------------------------------------------------------------------------
+
+struct SMU7_ThermalScoreboard
+{
+ int16_t GpuLimit;
+ int16_t GpuHyst;
+ uint16_t CurrGnbTemp;
+ uint16_t FilteredGnbTemp;
+ uint8_t ControllerEnable;
+ uint8_t ControllerRunning;
+ uint8_t WaterfallUp;
+ uint8_t WaterfallDown;
+ uint8_t WaterfallLimit;
+ uint8_t padding[3];
+};
+
+typedef struct SMU7_ThermalScoreboard SMU7_ThermalScoreboard;
+
+// For FeatureEnables:
+#define SMU7_SCLK_DPM_CONFIG_MASK 0x01
+#define SMU7_VOLTAGE_CONTROLLER_CONFIG_MASK 0x02
+#define SMU7_THERMAL_CONTROLLER_CONFIG_MASK 0x04
+#define SMU7_MCLK_DPM_CONFIG_MASK 0x08
+#define SMU7_UVD_DPM_CONFIG_MASK 0x10
+#define SMU7_VCE_DPM_CONFIG_MASK 0x20
+#define SMU7_ACP_DPM_CONFIG_MASK 0x40
+#define SMU7_SAMU_DPM_CONFIG_MASK 0x80
+#define SMU7_PCIEGEN_DPM_CONFIG_MASK 0x100
+
+#define SMU7_ACP_MCLK_HANDSHAKE_DISABLE 0x00000001
+#define SMU7_ACP_SCLK_HANDSHAKE_DISABLE 0x00000002
+#define SMU7_UVD_MCLK_HANDSHAKE_DISABLE 0x00000100
+#define SMU7_UVD_SCLK_HANDSHAKE_DISABLE 0x00000200
+#define SMU7_VCE_MCLK_HANDSHAKE_DISABLE 0x00010000
+#define SMU7_VCE_SCLK_HANDSHAKE_DISABLE 0x00020000
+
+// All 'soft registers' should be uint32_t.
+struct SMU71_SoftRegisters
+{
+ uint32_t RefClockFrequency;
+ uint32_t PmTimerPeriod;
+ uint32_t FeatureEnables;
+#if defined (SMU__DGPU_ONLY)
+ uint32_t PreVBlankGap;
+ uint32_t VBlankTimeout;
+ uint32_t TrainTimeGap;
+ uint32_t MvddSwitchTime;
+ uint32_t LongestAcpiTrainTime;
+ uint32_t AcpiDelay;
+ uint32_t G5TrainTime;
+ uint32_t DelayMpllPwron;
+ uint32_t VoltageChangeTimeout;
+#endif
+ uint32_t HandshakeDisables;
+
+ uint8_t DisplayPhy1Config;
+ uint8_t DisplayPhy2Config;
+ uint8_t DisplayPhy3Config;
+ uint8_t DisplayPhy4Config;
+
+ uint8_t DisplayPhy5Config;
+ uint8_t DisplayPhy6Config;
+ uint8_t DisplayPhy7Config;
+ uint8_t DisplayPhy8Config;
+
+ uint32_t AverageGraphicsActivity;
+ uint32_t AverageMemoryActivity;
+ uint32_t AverageGioActivity;
+
+ uint8_t SClkDpmEnabledLevels;
+ uint8_t MClkDpmEnabledLevels;
+ uint8_t LClkDpmEnabledLevels;
+ uint8_t PCIeDpmEnabledLevels;
+
+ uint32_t DRAM_LOG_ADDR_H;
+ uint32_t DRAM_LOG_ADDR_L;
+ uint32_t DRAM_LOG_PHY_ADDR_H;
+ uint32_t DRAM_LOG_PHY_ADDR_L;
+ uint32_t DRAM_LOG_BUFF_SIZE;
+ uint32_t UlvEnterCount;
+ uint32_t UlvTime;
+ uint32_t UcodeLoadStatus;
+ uint8_t DPMFreezeAndForced;
+ uint8_t Activity_Weight;
+ uint8_t Reserved8[2];
+ uint32_t Reserved;
+};
+
+typedef struct SMU71_SoftRegisters SMU71_SoftRegisters;
+
+struct SMU71_Firmware_Header
+{
+ uint32_t Digest[5];
+ uint32_t Version;
+ uint32_t HeaderSize;
+ uint32_t Flags;
+ uint32_t EntryPoint;
+ uint32_t CodeSize;
+ uint32_t ImageSize;
+
+ uint32_t Rtos;
+ uint32_t SoftRegisters;
+ uint32_t DpmTable;
+ uint32_t FanTable;
+ uint32_t CacConfigTable;
+ uint32_t CacStatusTable;
+
+ uint32_t mcRegisterTable;
+
+ uint32_t mcArbDramTimingTable;
+
+ uint32_t PmFuseTable;
+ uint32_t Globals;
+ uint32_t UvdDpmTable;
+ uint32_t AcpDpmTable;
+ uint32_t VceDpmTable;
+ uint32_t SamuDpmTable;
+ uint32_t UlvSettings;
+ uint32_t Reserved[37];
+ uint32_t Signature;
+};
+
+typedef struct SMU71_Firmware_Header SMU71_Firmware_Header;
+
+struct SMU7_HystController_Data
+{
+ uint8_t waterfall_up;
+ uint8_t waterfall_down;
+ uint8_t pstate;
+ uint8_t clamp_mode;
+};
+
+typedef struct SMU7_HystController_Data SMU7_HystController_Data;
+
+#define SMU71_FIRMWARE_HEADER_LOCATION 0x20000
+
+enum DisplayConfig {
+ PowerDown = 1,
+ DP54x4,
+ DP54x2,
+ DP54x1,
+ DP27x4,
+ DP27x2,
+ DP27x1,
+ HDMI297,
+ HDMI162,
+ LVDS,
+ DP324x4,
+ DP324x2,
+ DP324x1
+};
+
+//#define SX_BLOCK_COUNT 8
+//#define MC_BLOCK_COUNT 1
+//#define CPL_BLOCK_COUNT 27
+
+#if defined SMU__VARIANT__ICELAND
+ #define SX_BLOCK_COUNT 8
+ #define MC_BLOCK_COUNT 1
+ #define CPL_BLOCK_COUNT 29
+#endif
+
+struct SMU7_Local_Cac {
+ uint8_t BlockId;
+ uint8_t SignalId;
+ uint8_t Threshold;
+ uint8_t Padding;
+};
+
+typedef struct SMU7_Local_Cac SMU7_Local_Cac;
+
+struct SMU7_Local_Cac_Table {
+ SMU7_Local_Cac SxLocalCac[SX_BLOCK_COUNT];
+ SMU7_Local_Cac CplLocalCac[CPL_BLOCK_COUNT];
+ SMU7_Local_Cac McLocalCac[MC_BLOCK_COUNT];
+};
+
+typedef struct SMU7_Local_Cac_Table SMU7_Local_Cac_Table;
+
+#if !defined(SMC_MICROCODE)
+#pragma pack(pop)
+#endif
+
+#endif
+
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef SMU71_DISCRETE_H
+#define SMU71_DISCRETE_H
+
+#include "smu71.h"
+
+#if !defined(SMC_MICROCODE)
+#pragma pack(push, 1)
+#endif
+
+#define VDDC_ON_SVI2 0x1
+#define VDDCI_ON_SVI2 0x2
+#define MVDD_ON_SVI2 0x4
+
+struct SMU71_Discrete_VoltageLevel
+{
+ uint16_t Voltage;
+ uint16_t StdVoltageHiSidd;
+ uint16_t StdVoltageLoSidd;
+ uint8_t Smio;
+ uint8_t padding;
+};
+
+typedef struct SMU71_Discrete_VoltageLevel SMU71_Discrete_VoltageLevel;
+
+struct SMU71_Discrete_GraphicsLevel
+{
+ uint32_t MinVddc;
+ uint32_t MinVddcPhases;
+
+ uint32_t SclkFrequency;
+
+ uint8_t pcieDpmLevel;
+ uint8_t DeepSleepDivId;
+ uint16_t ActivityLevel;
+
+ uint32_t CgSpllFuncCntl3;
+ uint32_t CgSpllFuncCntl4;
+ uint32_t SpllSpreadSpectrum;
+ uint32_t SpllSpreadSpectrum2;
+ uint32_t CcPwrDynRm;
+ uint32_t CcPwrDynRm1;
+ uint8_t SclkDid;
+ uint8_t DisplayWatermark;
+ uint8_t EnabledForActivity;
+ uint8_t EnabledForThrottle;
+ uint8_t UpHyst;
+ uint8_t DownHyst;
+ uint8_t VoltageDownHyst;
+ uint8_t PowerThrottle;
+};
+
+typedef struct SMU71_Discrete_GraphicsLevel SMU71_Discrete_GraphicsLevel;
+
+struct SMU71_Discrete_ACPILevel
+{
+ uint32_t Flags;
+ uint32_t MinVddc;
+ uint32_t MinVddcPhases;
+ uint32_t SclkFrequency;
+ uint8_t SclkDid;
+ uint8_t DisplayWatermark;
+ uint8_t DeepSleepDivId;
+ uint8_t padding;
+ uint32_t CgSpllFuncCntl;
+ uint32_t CgSpllFuncCntl2;
+ uint32_t CgSpllFuncCntl3;
+ uint32_t CgSpllFuncCntl4;
+ uint32_t SpllSpreadSpectrum;
+ uint32_t SpllSpreadSpectrum2;
+ uint32_t CcPwrDynRm;
+ uint32_t CcPwrDynRm1;
+};
+
+typedef struct SMU71_Discrete_ACPILevel SMU71_Discrete_ACPILevel;
+
+struct SMU71_Discrete_Ulv
+{
+ uint32_t CcPwrDynRm;
+ uint32_t CcPwrDynRm1;
+ uint16_t VddcOffset;
+ uint8_t VddcOffsetVid;
+ uint8_t VddcPhase;
+ uint32_t Reserved;
+};
+
+typedef struct SMU71_Discrete_Ulv SMU71_Discrete_Ulv;
+
+struct SMU71_Discrete_MemoryLevel
+{
+ uint32_t MinVddc;
+ uint32_t MinVddcPhases;
+ uint32_t MinVddci;
+ uint32_t MinMvdd;
+
+ uint32_t MclkFrequency;
+
+ uint8_t EdcReadEnable;
+ uint8_t EdcWriteEnable;
+ uint8_t RttEnable;
+ uint8_t StutterEnable;
+
+ uint8_t StrobeEnable;
+ uint8_t StrobeRatio;
+ uint8_t EnabledForThrottle;
+ uint8_t EnabledForActivity;
+
+ uint8_t UpHyst;
+ uint8_t DownHyst;
+ uint8_t VoltageDownHyst;
+ uint8_t padding;
+
+ uint16_t ActivityLevel;
+ uint8_t DisplayWatermark;
+ uint8_t padding1;
+
+ uint32_t MpllFuncCntl;
+ uint32_t MpllFuncCntl_1;
+ uint32_t MpllFuncCntl_2;
+ uint32_t MpllAdFuncCntl;
+ uint32_t MpllDqFuncCntl;
+ uint32_t MclkPwrmgtCntl;
+ uint32_t DllCntl;
+ uint32_t MpllSs1;
+ uint32_t MpllSs2;
+};
+
+typedef struct SMU71_Discrete_MemoryLevel SMU71_Discrete_MemoryLevel;
+
+struct SMU71_Discrete_LinkLevel
+{
+ uint8_t PcieGenSpeed; ///< 0:PciE-gen1 1:PciE-gen2 2:PciE-gen3
+ uint8_t PcieLaneCount; ///< 1=x1, 2=x2, 3=x4, 4=x8, 5=x12, 6=x16
+ uint8_t EnabledForActivity;
+ uint8_t SPC;
+ uint32_t DownThreshold;
+ uint32_t UpThreshold;
+ uint32_t Reserved;
+};
+
+typedef struct SMU71_Discrete_LinkLevel SMU71_Discrete_LinkLevel;
+
+
+#ifdef SMU__DYNAMIC_MCARB_SETTINGS
+// MC ARB DRAM Timing registers.
+struct SMU71_Discrete_MCArbDramTimingTableEntry
+{
+ uint32_t McArbDramTiming;
+ uint32_t McArbDramTiming2;
+ uint8_t McArbBurstTime;
+ uint8_t padding[3];
+};
+
+typedef struct SMU71_Discrete_MCArbDramTimingTableEntry SMU71_Discrete_MCArbDramTimingTableEntry;
+
+struct SMU71_Discrete_MCArbDramTimingTable
+{
+ SMU71_Discrete_MCArbDramTimingTableEntry entries[SMU__NUM_SCLK_DPM_STATE][SMU__NUM_MCLK_DPM_LEVELS];
+};
+
+typedef struct SMU71_Discrete_MCArbDramTimingTable SMU71_Discrete_MCArbDramTimingTable;
+#endif
+
+// UVD VCLK/DCLK state (level) definition.
+struct SMU71_Discrete_UvdLevel
+{
+ uint32_t VclkFrequency;
+ uint32_t DclkFrequency;
+ uint16_t MinVddc;
+ uint8_t MinVddcPhases;
+ uint8_t VclkDivider;
+ uint8_t DclkDivider;
+ uint8_t padding[3];
+};
+
+typedef struct SMU71_Discrete_UvdLevel SMU71_Discrete_UvdLevel;
+
+// Clocks for other external blocks (VCE, ACP, SAMU).
+struct SMU71_Discrete_ExtClkLevel
+{
+ uint32_t Frequency;
+ uint16_t MinVoltage;
+ uint8_t MinPhases;
+ uint8_t Divider;
+};
+
+typedef struct SMU71_Discrete_ExtClkLevel SMU71_Discrete_ExtClkLevel;
+
+// Everything that we need to keep track of about the current state.
+// Use this instead of copies of the GraphicsLevel and MemoryLevel structures to keep track of state parameters
+// that need to be checked later.
+// We don't need to cache everything about a state, just a few parameters.
+struct SMU71_Discrete_StateInfo
+{
+ uint32_t SclkFrequency;
+ uint32_t MclkFrequency;
+ uint32_t VclkFrequency;
+ uint32_t DclkFrequency;
+ uint32_t SamclkFrequency;
+ uint32_t AclkFrequency;
+ uint32_t EclkFrequency;
+ uint16_t MvddVoltage;
+ uint16_t padding16;
+ uint8_t DisplayWatermark;
+ uint8_t McArbIndex;
+ uint8_t McRegIndex;
+ uint8_t SeqIndex;
+ uint8_t SclkDid;
+ int8_t SclkIndex;
+ int8_t MclkIndex;
+ uint8_t PCIeGen;
+
+};
+
+typedef struct SMU71_Discrete_StateInfo SMU71_Discrete_StateInfo;
+
+
+struct SMU71_Discrete_DpmTable
+{
+ // Multi-DPM controller settings
+ SMU71_PIDController GraphicsPIDController;
+ SMU71_PIDController MemoryPIDController;
+ SMU71_PIDController LinkPIDController;
+
+ uint32_t SystemFlags;
+
+ // SMIO masks for voltage and phase controls
+ uint32_t SmioMaskVddcVid;
+ uint32_t SmioMaskVddcPhase;
+ uint32_t SmioMaskVddciVid;
+ uint32_t SmioMaskMvddVid;
+
+ uint32_t VddcLevelCount;
+ uint32_t VddciLevelCount;
+ uint32_t MvddLevelCount;
+
+ SMU71_Discrete_VoltageLevel VddcLevel [SMU71_MAX_LEVELS_VDDC];
+ SMU71_Discrete_VoltageLevel VddciLevel [SMU71_MAX_LEVELS_VDDCI];
+ SMU71_Discrete_VoltageLevel MvddLevel [SMU71_MAX_LEVELS_MVDD];
+
+ uint8_t GraphicsDpmLevelCount;
+ uint8_t MemoryDpmLevelCount;
+ uint8_t LinkLevelCount;
+ uint8_t MasterDeepSleepControl;
+
+ uint32_t Reserved[5];
+
+ // State table entries for each DPM state
+ SMU71_Discrete_GraphicsLevel GraphicsLevel [SMU71_MAX_LEVELS_GRAPHICS];
+ SMU71_Discrete_MemoryLevel MemoryACPILevel;
+ SMU71_Discrete_MemoryLevel MemoryLevel [SMU71_MAX_LEVELS_MEMORY];
+ SMU71_Discrete_LinkLevel LinkLevel [SMU71_MAX_LEVELS_LINK];
+ SMU71_Discrete_ACPILevel ACPILevel;
+
+ uint32_t SclkStepSize;
+ uint32_t Smio [SMU71_MAX_ENTRIES_SMIO];
+
+ uint8_t GraphicsBootLevel;
+ uint8_t GraphicsVoltageChangeEnable;
+ uint8_t GraphicsThermThrottleEnable;
+ uint8_t GraphicsInterval;
+
+ uint8_t VoltageInterval;
+ uint8_t ThermalInterval;
+ uint16_t TemperatureLimitHigh;
+
+ uint16_t TemperatureLimitLow;
+ uint8_t MemoryBootLevel;
+ uint8_t MemoryVoltageChangeEnable;
+
+ uint8_t MemoryInterval;
+ uint8_t MemoryThermThrottleEnable;
+ uint8_t MergedVddci;
+ uint8_t padding2;
+
+ uint16_t VoltageResponseTime;
+ uint16_t PhaseResponseTime;
+
+ uint8_t PCIeBootLinkLevel;
+ uint8_t PCIeGenInterval;
+ uint8_t DTEInterval;
+ uint8_t DTEMode;
+
+ uint8_t SVI2Enable;
+ uint8_t VRHotGpio;
+ uint8_t AcDcGpio;
+ uint8_t ThermGpio;
+
+ uint32_t DisplayCac;
+
+ uint16_t MaxPwr;
+ uint16_t NomPwr;
+
+ uint16_t FpsHighThreshold;
+ uint16_t FpsLowThreshold;
+
+ uint16_t BAPMTI_R [SMU71_DTE_ITERATIONS][SMU71_DTE_SOURCES][SMU71_DTE_SINKS];
+ uint16_t BAPMTI_RC [SMU71_DTE_ITERATIONS][SMU71_DTE_SOURCES][SMU71_DTE_SINKS];
+
+ uint8_t DTEAmbientTempBase;
+ uint8_t DTETjOffset;
+ uint8_t GpuTjMax;
+ uint8_t GpuTjHyst;
+
+ uint16_t BootVddc;
+ uint16_t BootVddci;
+
+ uint16_t BootMVdd;
+ uint16_t padding;
+
+ uint32_t BAPM_TEMP_GRADIENT;
+
+ uint32_t LowSclkInterruptThreshold;
+ uint32_t VddGfxReChkWait;
+
+ uint16_t PPM_PkgPwrLimit;
+ uint16_t PPM_TemperatureLimit;
+
+ uint16_t DefaultTdp;
+ uint16_t TargetTdp;
+};
+
+typedef struct SMU71_Discrete_DpmTable SMU71_Discrete_DpmTable;
+
+// --------------------------------------------------- AC Timing Parameters ------------------------------------------------
+#define SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE 16
+#define SMU71_DISCRETE_MC_REGISTER_ARRAY_SET_COUNT SMU71_MAX_LEVELS_MEMORY
+
+struct SMU71_Discrete_MCRegisterAddress
+{
+ uint16_t s0;
+ uint16_t s1;
+};
+
+typedef struct SMU71_Discrete_MCRegisterAddress SMU71_Discrete_MCRegisterAddress;
+
+struct SMU71_Discrete_MCRegisterSet
+{
+ uint32_t value[SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE];
+};
+
+typedef struct SMU71_Discrete_MCRegisterSet SMU71_Discrete_MCRegisterSet;
+
+struct SMU71_Discrete_MCRegisters
+{
+ uint8_t last;
+ uint8_t reserved[3];
+ SMU71_Discrete_MCRegisterAddress address[SMU71_DISCRETE_MC_REGISTER_ARRAY_SIZE];
+ SMU71_Discrete_MCRegisterSet data[SMU71_DISCRETE_MC_REGISTER_ARRAY_SET_COUNT];
+};
+
+typedef struct SMU71_Discrete_MCRegisters SMU71_Discrete_MCRegisters;
+
+
+// --------------------------------------------------- Fan Table -----------------------------------------------------------
+struct SMU71_Discrete_FanTable
+{
+ uint16_t FdoMode;
+ int16_t TempMin;
+ int16_t TempMed;
+ int16_t TempMax;
+ int16_t Slope1;
+ int16_t Slope2;
+ int16_t FdoMin;
+ int16_t HystUp;
+ int16_t HystDown;
+ int16_t HystSlope;
+ int16_t TempRespLim;
+ int16_t TempCurr;
+ int16_t SlopeCurr;
+ int16_t PwmCurr;
+ uint32_t RefreshPeriod;
+ int16_t FdoMax;
+ uint8_t TempSrc;
+ int8_t Padding;
+};
+
+typedef struct SMU71_Discrete_FanTable SMU71_Discrete_FanTable;
+
+#define SMU7_DISCRETE_GPIO_SCLK_DEBUG 4
+#define SMU7_DISCRETE_GPIO_SCLK_DEBUG_BIT (0x1 << SMU7_DISCRETE_GPIO_SCLK_DEBUG)
+
+struct SMU71_MclkDpmScoreboard
+{
+
+ uint32_t PercentageBusy;
+
+ int32_t PIDError;
+ int32_t PIDIntegral;
+ int32_t PIDOutput;
+
+ uint32_t SigmaDeltaAccum;
+ uint32_t SigmaDeltaOutput;
+ uint32_t SigmaDeltaLevel;
+
+ uint32_t UtilizationSetpoint;
+
+ uint8_t TdpClampMode;
+ uint8_t TdcClampMode;
+ uint8_t ThermClampMode;
+ uint8_t VoltageBusy;
+
+ int8_t CurrLevel;
+ int8_t TargLevel;
+ uint8_t LevelChangeInProgress;
+ uint8_t UpHyst;
+
+ uint8_t DownHyst;
+ uint8_t VoltageDownHyst;
+ uint8_t DpmEnable;
+ uint8_t DpmRunning;
+
+ uint8_t DpmForce;
+ uint8_t DpmForceLevel;
+ uint8_t DisplayWatermark;
+ uint8_t McArbIndex;
+
+ uint32_t MinimumPerfMclk;
+
+ uint8_t AcpiReq;
+ uint8_t AcpiAck;
+ uint8_t MclkSwitchInProgress;
+ uint8_t MclkSwitchCritical;
+
+ uint8_t TargetMclkIndex;
+ uint8_t TargetMvddIndex;
+ uint8_t MclkSwitchResult;
+
+ uint8_t EnabledLevelsChange;
+
+ uint16_t LevelResidencyCounters [SMU71_MAX_LEVELS_MEMORY];
+ uint16_t LevelSwitchCounters [SMU71_MAX_LEVELS_MEMORY];
+
+ void (*TargetStateCalculator)(uint8_t);
+ void (*SavedTargetStateCalculator)(uint8_t);
+
+ uint16_t AutoDpmInterval;
+ uint16_t AutoDpmRange;
+
+ uint16_t MclkSwitchingTime;
+ uint8_t padding[2];
+};
+
+typedef struct SMU71_MclkDpmScoreboard SMU71_MclkDpmScoreboard;
+
+struct SMU71_UlvScoreboard
+{
+ uint8_t EnterUlv;
+ uint8_t ExitUlv;
+ uint8_t UlvActive;
+ uint8_t WaitingForUlv;
+ uint8_t UlvEnable;
+ uint8_t UlvRunning;
+ uint8_t UlvMasterEnable;
+ uint8_t padding;
+ uint32_t UlvAbortedCount;
+ uint32_t UlvTimeStamp;
+};
+
+typedef struct SMU71_UlvScoreboard SMU71_UlvScoreboard;
+
+struct SMU71_VddGfxScoreboard
+{
+ uint8_t VddGfxEnable;
+ uint8_t VddGfxActive;
+ uint8_t padding[2];
+
+ uint32_t VddGfxEnteredCount;
+ uint32_t VddGfxAbortedCount;
+};
+
+typedef struct SMU71_VddGfxScoreboard SMU71_VddGfxScoreboard;
+
+struct SMU71_AcpiScoreboard {
+ uint32_t SavedInterruptMask[2];
+ uint8_t LastACPIRequest;
+ uint8_t CgBifResp;
+ uint8_t RequestType;
+ uint8_t Padding;
+ SMU71_Discrete_ACPILevel D0Level;
+};
+
+typedef struct SMU71_AcpiScoreboard SMU71_AcpiScoreboard;
+
+
+struct SMU71_Discrete_PmFuses {
+ // dw0-dw1
+ uint8_t BapmVddCVidHiSidd[8];
+
+ // dw2-dw3
+ uint8_t BapmVddCVidLoSidd[8];
+
+ // dw4-dw5
+ uint8_t VddCVid[8];
+
+ // dw6
+ uint8_t SviLoadLineEn;
+ uint8_t SviLoadLineVddC;
+ uint8_t SviLoadLineTrimVddC;
+ uint8_t SviLoadLineOffsetVddC;
+
+ // dw7
+ uint16_t TDC_VDDC_PkgLimit;
+ uint8_t TDC_VDDC_ThrottleReleaseLimitPerc;
+ uint8_t TDC_MAWt;
+
+ // dw8
+ uint8_t TdcWaterfallCtl;
+ uint8_t LPMLTemperatureMin;
+ uint8_t LPMLTemperatureMax;
+ uint8_t Reserved;
+
+ // dw9-dw12
+ uint8_t LPMLTemperatureScaler[16];
+
+ // dw13-dw14
+ int16_t FuzzyFan_ErrorSetDelta;
+ int16_t FuzzyFan_ErrorRateSetDelta;
+ int16_t FuzzyFan_PwmSetDelta;
+ uint16_t Reserved6;
+
+ // dw15
+ uint8_t GnbLPML[16];
+
+ // dw15
+ uint8_t GnbLPMLMaxVid;
+ uint8_t GnbLPMLMinVid;
+ uint8_t Reserved1[2];
+
+ // dw16
+ uint16_t BapmVddCBaseLeakageHiSidd;
+ uint16_t BapmVddCBaseLeakageLoSidd;
+};
+
+typedef struct SMU71_Discrete_PmFuses SMU71_Discrete_PmFuses;
+
+struct SMU71_Discrete_Log_Header_Table {
+ uint32_t version;
+ uint32_t asic_id;
+ uint16_t flags;
+ uint16_t entry_size;
+ uint32_t total_size;
+ uint32_t num_of_entries;
+ uint8_t type;
+ uint8_t mode;
+ uint8_t filler_0[2];
+ uint32_t filler_1[2];
+};
+
+typedef struct SMU71_Discrete_Log_Header_Table SMU71_Discrete_Log_Header_Table;
+
+struct SMU71_Discrete_Log_Cntl {
+ uint8_t Enabled;
+ uint8_t Type;
+ uint8_t padding[2];
+ uint32_t BufferSize;
+ uint32_t SamplesLogged;
+ uint32_t SampleSize;
+ uint32_t AddrL;
+ uint32_t AddrH;
+};
+
+typedef struct SMU71_Discrete_Log_Cntl SMU71_Discrete_Log_Cntl;
+
+#if defined SMU__DGPU_ONLY
+ #define CAC_ACC_NW_NUM_OF_SIGNALS 83
+#endif
+
+
+struct SMU71_Discrete_Cac_Collection_Table {
+ uint32_t temperature;
+ uint32_t cac_acc_nw[CAC_ACC_NW_NUM_OF_SIGNALS];
+ uint32_t filler[4];
+};
+
+typedef struct SMU71_Discrete_Cac_Collection_Table SMU71_Discrete_Cac_Collection_Table;
+
+struct SMU71_Discrete_Cac_Verification_Table {
+ uint32_t VddcTotalPower;
+ uint32_t VddcLeakagePower;
+ uint32_t VddcConstantPower;
+ uint32_t VddcGfxDynamicPower;
+ uint32_t VddcUvdDynamicPower;
+ uint32_t VddcVceDynamicPower;
+ uint32_t VddcAcpDynamicPower;
+ uint32_t VddcPcieDynamicPower;
+ uint32_t VddcDceDynamicPower;
+ uint32_t VddcCurrent;
+ uint32_t VddcVoltage;
+ uint32_t VddciTotalPower;
+ uint32_t VddciLeakagePower;
+ uint32_t VddciConstantPower;
+ uint32_t VddciDynamicPower;
+ uint32_t Vddr1TotalPower;
+ uint32_t Vddr1LeakagePower;
+ uint32_t Vddr1ConstantPower;
+ uint32_t Vddr1DynamicPower;
+ uint32_t spare[8];
+ uint32_t temperature;
+};
+
+typedef struct SMU71_Discrete_Cac_Verification_Table SMU71_Discrete_Cac_Verification_Table;
+
+#if !defined(SMC_MICROCODE)
+#pragma pack(pop)
+#endif
+
+
+#endif
+
struct pp_smumgr {
uint32_t chip_family;
uint32_t chip_id;
- uint32_t hw_revision;
void *device;
void *backend;
uint32_t usec_timeout;
extern int smu_free_memory(void *device, void *handle);
+extern int cz_smum_init(struct pp_smumgr *smumgr);
+extern int iceland_smum_init(struct pp_smumgr *smumgr);
+extern int tonga_smum_init(struct pp_smumgr *smumgr);
+extern int fiji_smum_init(struct pp_smumgr *smumgr);
+extern int polaris10_smum_init(struct pp_smumgr *smumgr);
+
#define SMUM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT
#define SMUM_FIELD_MASK(reg, field) reg##__##field##_MASK
# Makefile for the 'smu manager' sub-component of powerplay.
# It provides the smu management services for the driver.
-SMU_MGR = smumgr.o cz_smumgr.o tonga_smumgr.o fiji_smumgr.o polaris10_smumgr.o
+SMU_MGR = smumgr.o cz_smumgr.o tonga_smumgr.o fiji_smumgr.o \
+ polaris10_smumgr.o iceland_smumgr.o
AMD_PP_SMUMGR = $(addprefix $(AMD_PP_PATH)/smumgr/,$(SMU_MGR))
if (result != 0)
return result;
- result = SMUM_WAIT_FIELD_UNEQUAL(smumgr,
+ return SMUM_WAIT_FIELD_UNEQUAL(smumgr,
SMU_MP1_SRBM2P_RESP_0, CONTENT, 0);
-
- if (result != 0)
- return result;
-
- return 0;
}
static int cz_set_smc_sram_address(struct pp_smumgr *smumgr,
if (0 != (3 & smc_address)) {
printk(KERN_ERR "[ powerplay ] SMC address must be 4 byte aligned\n");
- return -1;
+ return -EINVAL;
}
if (limit <= (smc_address + 3)) {
printk(KERN_ERR "[ powerplay ] SMC address beyond the SMC RAM area\n");
- return -1;
+ return -EINVAL;
}
cgs_write_register(smumgr->device, mmMP0PUB_IND_INDEX_0,
return -EINVAL;
result = cz_set_smc_sram_address(smumgr, smc_address, limit);
- cgs_write_register(smumgr->device, mmMP0PUB_IND_DATA_0, value);
+ if (!result)
+ cgs_write_register(smumgr->device, mmMP0PUB_IND_DATA_0, value);
- return 0;
+ return result;
}
static int cz_send_msg_to_smc_with_parameter(struct pp_smumgr *smumgr,
static int cz_request_smu_load_fw(struct pp_smumgr *smumgr)
{
struct cz_smumgr *cz_smu = (struct cz_smumgr *)(smumgr->backend);
- int result = 0;
uint32_t smc_address;
if (!smumgr->reload_fw) {
cz_send_msg_to_smc_with_parameter(smumgr, PPSMC_MSG_ExecuteJob,
cz_smu->toc_entry_power_profiling_index);
- result = cz_send_msg_to_smc_with_parameter(smumgr,
+ return cz_send_msg_to_smc_with_parameter(smumgr,
PPSMC_MSG_ExecuteJob,
cz_smu->toc_entry_initialize_index);
-
- return result;
}
static int cz_check_fw_load_finish(struct pp_smumgr *smumgr,
if (smumgr == NULL || smumgr->device == NULL)
return -EINVAL;
- return cgs_read_register(smumgr->device,
- mmSMU_MP1_SRBM2P_ARG_0);
-
cgs_write_register(smumgr->device, mmMP0PUB_IND_INDEX, index);
for (i = 0; i < smumgr->usec_timeout; i++) {
if (smumgr->chip_id == CHIP_STONEY)
fw_to_check &= ~(UCODE_ID_SDMA1_MASK | UCODE_ID_CP_MEC_JT2_MASK);
- cz_request_smu_load_fw(smumgr);
+ ret = cz_request_smu_load_fw(smumgr);
+ if (ret)
+ printk(KERN_ERR "[ powerplay] SMU firmware load failed\n");
+
cz_check_fw_load_finish(smumgr, fw_to_check);
ret = cz_load_mec_firmware(smumgr);
cz_smu_populate_single_ucode_load_task(smumgr,
CZ_SCRATCH_ENTRY_UCODE_ID_SDMA0, false);
- if (smumgr->chip_id == CHIP_STONEY)
- cz_smu_populate_single_ucode_load_task(smumgr,
- CZ_SCRATCH_ENTRY_UCODE_ID_SDMA0, false);
- else
+ if (smumgr->chip_id != CHIP_STONEY)
cz_smu_populate_single_ucode_load_task(smumgr,
CZ_SCRATCH_ENTRY_UCODE_ID_SDMA1, false);
cz_smu_populate_single_ucode_load_task(smumgr,
CZ_SCRATCH_ENTRY_UCODE_ID_CP_ME, false);
cz_smu_populate_single_ucode_load_task(smumgr,
CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1, false);
- if (smumgr->chip_id == CHIP_STONEY)
- cz_smu_populate_single_ucode_load_task(smumgr,
- CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT1, false);
- else
+ if (smumgr->chip_id != CHIP_STONEY)
cz_smu_populate_single_ucode_load_task(smumgr,
CZ_SCRATCH_ENTRY_UCODE_ID_CP_MEC_JT2, false);
cz_smu_populate_single_ucode_load_task(smumgr,
struct cz_smumgr *cz_smu = (struct cz_smumgr *)smumgr->backend;
cz_smu->toc_entry_used_count = 0;
-
cz_smu_initialize_toc_empty_job_list(smumgr);
-
cz_smu_construct_toc_for_rlc_aram_save(smumgr);
-
cz_smu_construct_toc_for_vddgfx_enter(smumgr);
-
cz_smu_construct_toc_for_vddgfx_exit(smumgr);
-
cz_smu_construct_toc_for_power_profiling(smumgr);
-
cz_smu_construct_toc_for_bootup(smumgr);
-
cz_smu_construct_toc_for_clock_table(smumgr);
return 0;
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+#include <linux/types.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/gfp.h>
+
+#include "smumgr.h"
+#include "iceland_smumgr.h"
+#include "pp_debug.h"
+#include "smu_ucode_xfer_vi.h"
+#include "ppsmc.h"
+#include "smu/smu_7_1_1_d.h"
+#include "smu/smu_7_1_1_sh_mask.h"
+#include "cgs_common.h"
+
+#define ICELAND_SMC_SIZE 0x20000
+#define BUFFER_SIZE 80000
+#define MAX_STRING_SIZE 15
+#define BUFFER_SIZETWO 131072 /*128 *1024*/
+
+/**
+ * Set the address for reading/writing the SMC SRAM space.
+ * @param smumgr the address of the powerplay hardware manager.
+ * @param smcAddress the address in the SMC RAM to access.
+ */
+static int iceland_set_smc_sram_address(struct pp_smumgr *smumgr,
+ uint32_t smcAddress, uint32_t limit)
+{
+ if (smumgr == NULL || smumgr->device == NULL)
+ return -EINVAL;
+ PP_ASSERT_WITH_CODE((0 == (3 & smcAddress)),
+ "SMC address must be 4 byte aligned.",
+ return -1;);
+
+ PP_ASSERT_WITH_CODE((limit > (smcAddress + 3)),
+ "SMC address is beyond the SMC RAM area.",
+ return -1;);
+
+ cgs_write_register(smumgr->device, mmSMC_IND_INDEX_0, smcAddress);
+ SMUM_WRITE_FIELD(smumgr->device, SMC_IND_ACCESS_CNTL, AUTO_INCREMENT_IND_0, 0);
+
+ return 0;
+}
+
+/**
+ * Copy bytes from an array into the SMC RAM space.
+ *
+ * @param smumgr the address of the powerplay SMU manager.
+ * @param smcStartAddress the start address in the SMC RAM to copy bytes to.
+ * @param src the byte array to copy the bytes from.
+ * @param byteCount the number of bytes to copy.
+ */
+int iceland_copy_bytes_to_smc(struct pp_smumgr *smumgr,
+ uint32_t smcStartAddress, const uint8_t *src,
+ uint32_t byteCount, uint32_t limit)
+{
+ uint32_t addr;
+ uint32_t data, orig_data;
+ int result = 0;
+ uint32_t extra_shift;
+
+ if (smumgr == NULL || smumgr->device == NULL)
+ return -EINVAL;
+ PP_ASSERT_WITH_CODE((0 == (3 & smcStartAddress)),
+ "SMC address must be 4 byte aligned.",
+ return 0;);
+
+ PP_ASSERT_WITH_CODE((limit > (smcStartAddress + byteCount)),
+ "SMC address is beyond the SMC RAM area.",
+ return 0;);
+
+ addr = smcStartAddress;
+
+ while (byteCount >= 4) {
+ /*
+ * Bytes are written into the
+ * SMC address space with the MSB first
+ */
+ data = (src[0] << 24) + (src[1] << 16) + (src[2] << 8) + src[3];
+
+ result = iceland_set_smc_sram_address(smumgr, addr, limit);
+
+ if (result)
+ goto out;
+
+ cgs_write_register(smumgr->device, mmSMC_IND_DATA_0, data);
+
+ src += 4;
+ byteCount -= 4;
+ addr += 4;
+ }
+
+ if (0 != byteCount) {
+ /* Now write odd bytes left, do a read modify write cycle */
+ data = 0;
+
+ result = iceland_set_smc_sram_address(smumgr, addr, limit);
+ if (result)
+ goto out;
+
+ orig_data = cgs_read_register(smumgr->device,
+ mmSMC_IND_DATA_0);
+ extra_shift = 8 * (4 - byteCount);
+
+ while (byteCount > 0) {
+ data = (data << 8) + *src++;
+ byteCount--;
+ }
+
+ data <<= extra_shift;
+ data |= (orig_data & ~((~0UL) << extra_shift));
+
+ result = iceland_set_smc_sram_address(smumgr, addr, limit);
+ if (result)
+ goto out;
+
+ cgs_write_register(smumgr->device, mmSMC_IND_DATA_0, data);
+ }
+
+out:
+ return result;
+}
+
+/**
+ * Deassert the reset'pin' (set it to high).
+ *
+ * @param smumgr the address of the powerplay hardware manager.
+ */
+static int iceland_start_smc(struct pp_smumgr *smumgr)
+{
+ SMUM_WRITE_INDIRECT_FIELD(smumgr->device, CGS_IND_REG__SMC,
+ SMC_SYSCON_RESET_CNTL, rst_reg, 0);
+
+ return 0;
+}
+
+static void iceland_pp_reset_smc(struct pp_smumgr *smumgr)
+{
+ SMUM_WRITE_INDIRECT_FIELD(smumgr->device, CGS_IND_REG__SMC,
+ SMC_SYSCON_RESET_CNTL,
+ rst_reg, 1);
+}
+
+int iceland_program_jump_on_start(struct pp_smumgr *smumgr)
+{
+ static const unsigned char pData[] = { 0xE0, 0x00, 0x80, 0x40 };
+
+ iceland_copy_bytes_to_smc(smumgr, 0x0, pData, 4, sizeof(pData)+1);
+
+ return 0;
+}
+
+/**
+ * Return if the SMC is currently running.
+ *
+ * @param smumgr the address of the powerplay hardware manager.
+ */
+bool iceland_is_smc_ram_running(struct pp_smumgr *smumgr)
+{
+ uint32_t val1, val2;
+
+ val1 = SMUM_READ_INDIRECT_FIELD(smumgr->device, CGS_IND_REG__SMC,
+ SMC_SYSCON_CLOCK_CNTL_0, ck_disable);
+ val2 = cgs_read_ind_register(smumgr->device, CGS_IND_REG__SMC,
+ ixSMC_PC_C);
+
+ return ((0 == val1) && (0x20100 <= val2));
+}
+
+/**
+ * Send a message to the SMC, and wait for its response.
+ *
+ * @param smumgr the address of the powerplay hardware manager.
+ * @param msg the message to send.
+ * @return The response that came from the SMC.
+ */
+static int iceland_send_msg_to_smc(struct pp_smumgr *smumgr, uint16_t msg)
+{
+ if (smumgr == NULL || smumgr->device == NULL)
+ return -EINVAL;
+
+ if (!iceland_is_smc_ram_running(smumgr))
+ return -EINVAL;
+
+ SMUM_WAIT_FIELD_UNEQUAL(smumgr, SMC_RESP_0, SMC_RESP, 0);
+ PP_ASSERT_WITH_CODE(
+ 1 == SMUM_READ_FIELD(smumgr->device, SMC_RESP_0, SMC_RESP),
+ "Failed to send Previous Message.",
+ );
+
+ cgs_write_register(smumgr->device, mmSMC_MESSAGE_0, msg);
+
+ SMUM_WAIT_FIELD_UNEQUAL(smumgr, SMC_RESP_0, SMC_RESP, 0);
+ PP_ASSERT_WITH_CODE(
+ 1 == SMUM_READ_FIELD(smumgr->device, SMC_RESP_0, SMC_RESP),
+ "Failed to send Message.",
+ );
+
+ return 0;
+}
+
+/**
+ * Send a message to the SMC with parameter
+ *
+ * @param smumgr: the address of the powerplay hardware manager.
+ * @param msg: the message to send.
+ * @param parameter: the parameter to send
+ * @return The response that came from the SMC.
+ */
+static int iceland_send_msg_to_smc_with_parameter(struct pp_smumgr *smumgr,
+ uint16_t msg, uint32_t parameter)
+{
+ if (smumgr == NULL || smumgr->device == NULL)
+ return -EINVAL;
+
+ cgs_write_register(smumgr->device, mmSMC_MSG_ARG_0, parameter);
+
+ return iceland_send_msg_to_smc(smumgr, msg);
+}
+
+/*
+ * Read a 32bit value from the SMC SRAM space.
+ * ALL PARAMETERS ARE IN HOST BYTE ORDER.
+ * @param smumgr the address of the powerplay hardware manager.
+ * @param smcAddress the address in the SMC RAM to access.
+ * @param value and output parameter for the data read from the SMC SRAM.
+ */
+int iceland_read_smc_sram_dword(struct pp_smumgr *smumgr,
+ uint32_t smcAddress, uint32_t *value,
+ uint32_t limit)
+{
+ int result;
+
+ result = iceland_set_smc_sram_address(smumgr, smcAddress, limit);
+
+ if (0 != result)
+ return result;
+
+ *value = cgs_read_register(smumgr->device, mmSMC_IND_DATA_0);
+
+ return 0;
+}
+
+/*
+ * Write a 32bit value to the SMC SRAM space.
+ * ALL PARAMETERS ARE IN HOST BYTE ORDER.
+ * @param smumgr the address of the powerplay hardware manager.
+ * @param smcAddress the address in the SMC RAM to access.
+ * @param value to write to the SMC SRAM.
+ */
+int iceland_write_smc_sram_dword(struct pp_smumgr *smumgr,
+ uint32_t smcAddress, uint32_t value,
+ uint32_t limit)
+{
+ int result;
+
+ result = iceland_set_smc_sram_address(smumgr, smcAddress, limit);
+
+ if (0 != result)
+ return result;
+
+ cgs_write_register(smumgr->device, mmSMC_IND_DATA_0, value);
+
+ return 0;
+}
+
+static int iceland_smu_fini(struct pp_smumgr *smumgr)
+{
+ struct iceland_smumgr *priv = (struct iceland_smumgr *)(smumgr->backend);
+
+ smu_free_memory(smumgr->device, (void *)priv->header_buffer.handle);
+
+ if (smumgr->backend != NULL) {
+ kfree(smumgr->backend);
+ smumgr->backend = NULL;
+ }
+
+ cgs_rel_firmware(smumgr->device, CGS_UCODE_ID_SMU);
+ return 0;
+}
+
+static enum cgs_ucode_id iceland_convert_fw_type_to_cgs(uint32_t fw_type)
+{
+ enum cgs_ucode_id result = CGS_UCODE_ID_MAXIMUM;
+
+ switch (fw_type) {
+ case UCODE_ID_SMU:
+ result = CGS_UCODE_ID_SMU;
+ break;
+ case UCODE_ID_SDMA0:
+ result = CGS_UCODE_ID_SDMA0;
+ break;
+ case UCODE_ID_SDMA1:
+ result = CGS_UCODE_ID_SDMA1;
+ break;
+ case UCODE_ID_CP_CE:
+ result = CGS_UCODE_ID_CP_CE;
+ break;
+ case UCODE_ID_CP_PFP:
+ result = CGS_UCODE_ID_CP_PFP;
+ break;
+ case UCODE_ID_CP_ME:
+ result = CGS_UCODE_ID_CP_ME;
+ break;
+ case UCODE_ID_CP_MEC:
+ result = CGS_UCODE_ID_CP_MEC;
+ break;
+ case UCODE_ID_CP_MEC_JT1:
+ result = CGS_UCODE_ID_CP_MEC_JT1;
+ break;
+ case UCODE_ID_CP_MEC_JT2:
+ result = CGS_UCODE_ID_CP_MEC_JT2;
+ break;
+ case UCODE_ID_RLC_G:
+ result = CGS_UCODE_ID_RLC_G;
+ break;
+ default:
+ break;
+ }
+
+ return result;
+}
+
+/**
+ * Convert the PPIRI firmware type to SMU type mask.
+ * For MEC, we need to check all MEC related type
+ */
+static uint16_t iceland_get_mask_for_firmware_type(uint16_t firmwareType)
+{
+ uint16_t result = 0;
+
+ switch (firmwareType) {
+ case UCODE_ID_SDMA0:
+ result = UCODE_ID_SDMA0_MASK;
+ break;
+ case UCODE_ID_SDMA1:
+ result = UCODE_ID_SDMA1_MASK;
+ break;
+ case UCODE_ID_CP_CE:
+ result = UCODE_ID_CP_CE_MASK;
+ break;
+ case UCODE_ID_CP_PFP:
+ result = UCODE_ID_CP_PFP_MASK;
+ break;
+ case UCODE_ID_CP_ME:
+ result = UCODE_ID_CP_ME_MASK;
+ break;
+ case UCODE_ID_CP_MEC:
+ case UCODE_ID_CP_MEC_JT1:
+ case UCODE_ID_CP_MEC_JT2:
+ result = UCODE_ID_CP_MEC_MASK;
+ break;
+ case UCODE_ID_RLC_G:
+ result = UCODE_ID_RLC_G_MASK;
+ break;
+ default:
+ break;
+ }
+
+ return result;
+}
+
+/**
+ * Check if the FW has been loaded,
+ * SMU will not return if loading has not finished.
+*/
+static int iceland_check_fw_load_finish(struct pp_smumgr *smumgr, uint32_t fwType)
+{
+ uint16_t fwMask = iceland_get_mask_for_firmware_type(fwType);
+
+ if (0 != SMUM_WAIT_VFPF_INDIRECT_REGISTER(smumgr, SMC_IND,
+ SOFT_REGISTERS_TABLE_27, fwMask, fwMask)) {
+ pr_err("[ powerplay ] check firmware loading failed\n");
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+/* Populate one firmware image to the data structure */
+static int iceland_populate_single_firmware_entry(struct pp_smumgr *smumgr,
+ uint16_t firmware_type,
+ struct SMU_Entry *pentry)
+{
+ int result;
+ struct cgs_firmware_info info = {0};
+
+ result = cgs_get_firmware_info(
+ smumgr->device,
+ iceland_convert_fw_type_to_cgs(firmware_type),
+ &info);
+
+ if (result == 0) {
+ pentry->version = 0;
+ pentry->id = (uint16_t)firmware_type;
+ pentry->image_addr_high = smu_upper_32_bits(info.mc_addr);
+ pentry->image_addr_low = smu_lower_32_bits(info.mc_addr);
+ pentry->meta_data_addr_high = 0;
+ pentry->meta_data_addr_low = 0;
+ pentry->data_size_byte = info.image_size;
+ pentry->num_register_entries = 0;
+
+ if (firmware_type == UCODE_ID_RLC_G)
+ pentry->flags = 1;
+ else
+ pentry->flags = 0;
+ } else {
+ return result;
+ }
+
+ return result;
+}
+
+static void iceland_pp_stop_smc_clock(struct pp_smumgr *smumgr)
+{
+ SMUM_WRITE_INDIRECT_FIELD(smumgr->device, CGS_IND_REG__SMC,
+ SMC_SYSCON_CLOCK_CNTL_0,
+ ck_disable, 1);
+}
+
+static void iceland_start_smc_clock(struct pp_smumgr *smumgr)
+{
+ SMUM_WRITE_INDIRECT_FIELD(smumgr->device, CGS_IND_REG__SMC,
+ SMC_SYSCON_CLOCK_CNTL_0,
+ ck_disable, 0);
+}
+
+int iceland_smu_start_smc(struct pp_smumgr *smumgr)
+{
+ /* set smc instruct start point at 0x0 */
+ iceland_program_jump_on_start(smumgr);
+
+ /* enable smc clock */
+ iceland_start_smc_clock(smumgr);
+
+ /* de-assert reset */
+ iceland_start_smc(smumgr);
+
+ SMUM_WAIT_INDIRECT_FIELD(smumgr, SMC_IND, FIRMWARE_FLAGS,
+ INTERRUPTS_ENABLED, 1);
+
+ return 0;
+}
+
+/**
+ * Upload the SMC firmware to the SMC microcontroller.
+ *
+ * @param smumgr the address of the powerplay hardware manager.
+ * @param pFirmware the data structure containing the various sections of the firmware.
+ */
+int iceland_smu_upload_firmware_image(struct pp_smumgr *smumgr)
+{
+ const uint8_t *src;
+ uint32_t byte_count, val;
+ uint32_t data;
+ struct cgs_firmware_info info = {0};
+
+ if (smumgr == NULL || smumgr->device == NULL)
+ return -EINVAL;
+
+ /* load SMC firmware */
+ cgs_get_firmware_info(smumgr->device,
+ iceland_convert_fw_type_to_cgs(UCODE_ID_SMU), &info);
+
+ if (info.image_size & 3) {
+ pr_err("[ powerplay ] SMC ucode is not 4 bytes aligned\n");
+ return -EINVAL;
+ }
+
+ if (info.image_size > ICELAND_SMC_SIZE) {
+ pr_err("[ powerplay ] SMC address is beyond the SMC RAM area\n");
+ return -EINVAL;
+ }
+
+ /* wait for smc boot up */
+ SMUM_WAIT_INDIRECT_FIELD_UNEQUAL(smumgr, SMC_IND,
+ RCU_UC_EVENTS, boot_seq_done, 0);
+
+ /* clear firmware interrupt enable flag */
+ val = cgs_read_ind_register(smumgr->device, CGS_IND_REG__SMC,
+ ixSMC_SYSCON_MISC_CNTL);
+ cgs_write_ind_register(smumgr->device, CGS_IND_REG__SMC,
+ ixSMC_SYSCON_MISC_CNTL, val | 1);
+
+ /* stop smc clock */
+ iceland_pp_stop_smc_clock(smumgr);
+
+ /* reset smc */
+ iceland_pp_reset_smc(smumgr);
+
+ cgs_write_register(smumgr->device, mmSMC_IND_INDEX_0,
+ info.ucode_start_address);
+
+ SMUM_WRITE_FIELD(smumgr->device, SMC_IND_ACCESS_CNTL,
+ AUTO_INCREMENT_IND_0, 1);
+
+ byte_count = info.image_size;
+ src = (const uint8_t *)info.kptr;
+
+ while (byte_count >= 4) {
+ data = (src[0] << 24) + (src[1] << 16) + (src[2] << 8) + src[3];
+ cgs_write_register(smumgr->device, mmSMC_IND_DATA_0, data);
+ src += 4;
+ byte_count -= 4;
+ }
+
+ SMUM_WRITE_FIELD(smumgr->device, SMC_IND_ACCESS_CNTL,
+ AUTO_INCREMENT_IND_0, 0);
+
+ return 0;
+}
+
+static int iceland_request_smu_reload_fw(struct pp_smumgr *smumgr)
+{
+ struct iceland_smumgr *iceland_smu =
+ (struct iceland_smumgr *)(smumgr->backend);
+ uint16_t fw_to_load;
+ int result = 0;
+ struct SMU_DRAMData_TOC *toc;
+
+ toc = (struct SMU_DRAMData_TOC *)iceland_smu->pHeader;
+ toc->num_entries = 0;
+ toc->structure_version = 1;
+
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry(smumgr,
+ UCODE_ID_RLC_G,
+ &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n",
+ return -1);
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry(smumgr,
+ UCODE_ID_CP_CE,
+ &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n",
+ return -1);
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry
+ (smumgr, UCODE_ID_CP_PFP, &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n", return -1);
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry
+ (smumgr, UCODE_ID_CP_ME, &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n", return -1);
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry
+ (smumgr, UCODE_ID_CP_MEC, &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n", return -1);
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry
+ (smumgr, UCODE_ID_CP_MEC_JT1, &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n", return -1);
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry
+ (smumgr, UCODE_ID_CP_MEC_JT2, &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n", return -1);
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry
+ (smumgr, UCODE_ID_SDMA0, &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n", return -1);
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_populate_single_firmware_entry
+ (smumgr, UCODE_ID_SDMA1, &toc->entry[toc->num_entries++]),
+ "Failed to Get Firmware Entry.\n", return -1);
+
+ if (!iceland_is_smc_ram_running(smumgr)) {
+ result = iceland_smu_upload_firmware_image(smumgr);
+ if (result)
+ return result;
+
+ result = iceland_smu_start_smc(smumgr);
+ if (result)
+ return result;
+ }
+
+ iceland_send_msg_to_smc_with_parameter(smumgr,
+ PPSMC_MSG_DRV_DRAM_ADDR_HI,
+ iceland_smu->header_buffer.mc_addr_high);
+
+ iceland_send_msg_to_smc_with_parameter(smumgr,
+ PPSMC_MSG_DRV_DRAM_ADDR_LO,
+ iceland_smu->header_buffer.mc_addr_low);
+
+ fw_to_load = UCODE_ID_RLC_G_MASK
+ + UCODE_ID_SDMA0_MASK
+ + UCODE_ID_SDMA1_MASK
+ + UCODE_ID_CP_CE_MASK
+ + UCODE_ID_CP_ME_MASK
+ + UCODE_ID_CP_PFP_MASK
+ + UCODE_ID_CP_MEC_MASK
+ + UCODE_ID_CP_MEC_JT1_MASK
+ + UCODE_ID_CP_MEC_JT2_MASK;
+
+ PP_ASSERT_WITH_CODE(
+ 0 == iceland_send_msg_to_smc_with_parameter(
+ smumgr, PPSMC_MSG_LoadUcodes, fw_to_load),
+ "Fail to Request SMU Load uCode", return 0);
+
+ return result;
+}
+
+static int iceland_request_smu_load_specific_fw(struct pp_smumgr *smumgr,
+ uint32_t firmwareType)
+{
+ return 0;
+}
+
+static int iceland_start_smu(struct pp_smumgr *smumgr)
+{
+ int result;
+
+ result = iceland_smu_upload_firmware_image(smumgr);
+ if (result)
+ return result;
+
+ result = iceland_smu_start_smc(smumgr);
+ if (result)
+ return result;
+
+ result = iceland_request_smu_reload_fw(smumgr);
+
+ return result;
+}
+
+/**
+ * Write a 32bit value to the SMC SRAM space.
+ * ALL PARAMETERS ARE IN HOST BYTE ORDER.
+ * @param smumgr the address of the powerplay hardware manager.
+ * @param smcAddress the address in the SMC RAM to access.
+ * @param value to write to the SMC SRAM.
+ */
+static int iceland_smu_init(struct pp_smumgr *smumgr)
+{
+ struct iceland_smumgr *iceland_smu;
+ uint64_t mc_addr = 0;
+
+ /* Allocate memory for backend private data */
+ iceland_smu = (struct iceland_smumgr *)(smumgr->backend);
+ iceland_smu->header_buffer.data_size =
+ ((sizeof(struct SMU_DRAMData_TOC) / 4096) + 1) * 4096;
+
+ smu_allocate_memory(smumgr->device,
+ iceland_smu->header_buffer.data_size,
+ CGS_GPU_MEM_TYPE__VISIBLE_CONTIG_FB,
+ PAGE_SIZE,
+ &mc_addr,
+ &iceland_smu->header_buffer.kaddr,
+ &iceland_smu->header_buffer.handle);
+
+ iceland_smu->pHeader = iceland_smu->header_buffer.kaddr;
+ iceland_smu->header_buffer.mc_addr_high = smu_upper_32_bits(mc_addr);
+ iceland_smu->header_buffer.mc_addr_low = smu_lower_32_bits(mc_addr);
+
+ PP_ASSERT_WITH_CODE((NULL != iceland_smu->pHeader),
+ "Out of memory.",
+ kfree(smumgr->backend);
+ cgs_free_gpu_mem(smumgr->device,
+ (cgs_handle_t)iceland_smu->header_buffer.handle);
+ return -1);
+
+ return 0;
+}
+
+static const struct pp_smumgr_func iceland_smu_funcs = {
+ .smu_init = &iceland_smu_init,
+ .smu_fini = &iceland_smu_fini,
+ .start_smu = &iceland_start_smu,
+ .check_fw_load_finish = &iceland_check_fw_load_finish,
+ .request_smu_load_fw = &iceland_request_smu_reload_fw,
+ .request_smu_load_specific_fw = &iceland_request_smu_load_specific_fw,
+ .send_msg_to_smc = &iceland_send_msg_to_smc,
+ .send_msg_to_smc_with_parameter = &iceland_send_msg_to_smc_with_parameter,
+ .download_pptable_settings = NULL,
+ .upload_pptable_settings = NULL,
+};
+
+int iceland_smum_init(struct pp_smumgr *smumgr)
+{
+ struct iceland_smumgr *iceland_smu = NULL;
+
+ iceland_smu = kzalloc(sizeof(struct iceland_smumgr), GFP_KERNEL);
+
+ if (iceland_smu == NULL)
+ return -ENOMEM;
+
+ smumgr->backend = iceland_smu;
+ smumgr->smumgr_funcs = &iceland_smu_funcs;
+
+ return 0;
+}
--- /dev/null
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Author: Huang Rui <ray.huang@amd.com>
+ *
+ */
+
+#ifndef _ICELAND_SMUMGR_H_
+#define _ICELAND_SMUMGR_H_
+
+struct iceland_buffer_entry {
+ uint32_t data_size;
+ uint32_t mc_addr_low;
+ uint32_t mc_addr_high;
+ void *kaddr;
+ unsigned long handle;
+};
+
+/* Iceland only has header_buffer, don't have smu buffer. */
+struct iceland_smumgr {
+ uint8_t *pHeader;
+ uint8_t *pMecImage;
+ uint32_t ulSoftRegsStart;
+
+ struct iceland_buffer_entry header_buffer;
+};
+
+extern int iceland_smum_init(struct pp_smumgr *smumgr);
+extern int iceland_copy_bytes_to_smc(struct pp_smumgr *smumgr,
+ uint32_t smcStartAddress,
+ const uint8_t *src,
+ uint32_t byteCount, uint32_t limit);
+
+extern int iceland_smu_start_smc(struct pp_smumgr *smumgr);
+
+extern int iceland_read_smc_sram_dword(struct pp_smumgr *smumgr,
+ uint32_t smcAddress,
+ uint32_t *value, uint32_t limit);
+extern int iceland_write_smc_sram_dword(struct pp_smumgr *smumgr,
+ uint32_t smcAddress,
+ uint32_t value, uint32_t limit);
+
+extern bool iceland_is_smc_ram_running(struct pp_smumgr *smumgr);
+extern int iceland_smu_upload_firmware_image(struct pp_smumgr *smumgr);
+
+#endif
#include "cgs_common.h"
#define POLARIS10_SMC_SIZE 0x20000
-#define VOLTAGE_SCALE 4
/* Microcode file is stored in this buffer */
#define BUFFER_SIZE 80000
return 0;
}
-static const struct pp_smumgr_func ellsemere_smu_funcs = {
+static const struct pp_smumgr_func polaris10_smu_funcs = {
.smu_init = polaris10_smu_init,
.smu_fini = polaris10_smu_fini,
.start_smu = polaris10_start_smu,
return -1;
smumgr->backend = polaris10_smu;
- smumgr->smumgr_funcs = &ellsemere_smu_funcs;
+ smumgr->smumgr_funcs = &polaris10_smu_funcs;
return 0;
}
#include <polaris10_ppsmc.h>
#include <pp_endian.h>
+#include "smu74.h"
struct polaris10_avfs {
enum AVFS_BTC_STATUS avfs_btc_status;
uint32_t avfs_btc_param;
};
+struct polaris10_pt_defaults {
+ uint8_t SviLoadLineEn;
+ uint8_t SviLoadLineVddC;
+ uint8_t TDC_VDDC_ThrottleReleaseLimitPerc;
+ uint8_t TDC_MAWt;
+ uint8_t TdcWaterfallCtl;
+ uint8_t DTEAmbientTempBase;
+
+ uint32_t DisplayCac;
+ uint32_t BAPM_TEMP_GRADIENT;
+ uint16_t BAPMTI_R[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS];
+ uint16_t BAPMTI_RC[SMU74_DTE_ITERATIONS * SMU74_DTE_SOURCES * SMU74_DTE_SINKS];
+};
+
struct polaris10_buffer_entry {
uint32_t data_size;
uint32_t mc_addr_low;
unsigned long handle;
};
+struct polaris10_range_table {
+ uint32_t trans_lower_frequency; /* in 10khz */
+ uint32_t trans_upper_frequency;
+};
+
struct polaris10_smumgr {
uint8_t *header;
uint8_t *mec_image;
#include "smumgr.h"
#include "cgs_common.h"
#include "linux/delay.h"
-#include "cz_smumgr.h"
-#include "tonga_smumgr.h"
-#include "fiji_smumgr.h"
-#include "polaris10_smumgr.h"
+
int smum_init(struct amd_pp_init *pp_init, struct pp_instance *handle)
{
smumgr->device = pp_init->device;
smumgr->chip_family = pp_init->chip_family;
smumgr->chip_id = pp_init->chip_id;
- smumgr->hw_revision = pp_init->rev_id;
smumgr->usec_timeout = AMD_MAX_USEC_TIMEOUT;
smumgr->reload_fw = 1;
handle->smu_mgr = smumgr;
break;
case AMDGPU_FAMILY_VI:
switch (smumgr->chip_id) {
+ case CHIP_TOPAZ:
+ iceland_smum_init(smumgr);
+ break;
case CHIP_TONGA:
tonga_smum_init(smumgr);
break;
return 0;
}
-static const int edid_i2c_addr = 0x7e;
-static const int packet_i2c_addr = 0x70;
-static const int cec_i2c_addr = 0x78;
-
static int adv7511_probe(struct i2c_client *i2c, const struct i2c_device_id *id)
{
struct adv7511_link_config link_config;
struct adv7511 *adv7511;
struct device *dev = &i2c->dev;
+ unsigned int main_i2c_addr = i2c->addr << 1;
+ unsigned int edid_i2c_addr = main_i2c_addr + 4;
unsigned int val;
int ret;
regmap_write(adv7511->regmap, ADV7511_REG_EDID_I2C_ADDR, edid_i2c_addr);
regmap_write(adv7511->regmap, ADV7511_REG_PACKET_I2C_ADDR,
- packet_i2c_addr);
- regmap_write(adv7511->regmap, ADV7511_REG_CEC_I2C_ADDR, cec_i2c_addr);
+ main_i2c_addr - 0xa);
+ regmap_write(adv7511->regmap, ADV7511_REG_CEC_I2C_ADDR,
+ main_i2c_addr - 2);
+
adv7511_packet_disable(adv7511, 0xffff);
adv7511->i2c_main = i2c;
i2c_unregister_device(adv->i2c_cec);
}
-static const int cec_i2c_addr = 0x78;
-
int adv7533_init_cec(struct adv7511 *adv)
{
int ret;
- adv->i2c_cec = i2c_new_dummy(adv->i2c_main->adapter, cec_i2c_addr >> 1);
+ adv->i2c_cec = i2c_new_dummy(adv->i2c_main->adapter,
+ adv->i2c_main->addr - 1);
if (!adv->i2c_cec)
return -ENOMEM;
return 0;
}
+int analogix_dp_enable_psr(struct device *dev)
+{
+ struct analogix_dp_device *dp = dev_get_drvdata(dev);
+ struct edp_vsc_psr psr_vsc;
+
+ if (!dp->psr_support)
+ return -EINVAL;
+
+ /* Prepare VSC packet as per EDP 1.4 spec, Table 6.9 */
+ memset(&psr_vsc, 0, sizeof(psr_vsc));
+ psr_vsc.sdp_header.HB0 = 0;
+ psr_vsc.sdp_header.HB1 = 0x7;
+ psr_vsc.sdp_header.HB2 = 0x2;
+ psr_vsc.sdp_header.HB3 = 0x8;
+
+ psr_vsc.DB0 = 0;
+ psr_vsc.DB1 = EDP_VSC_PSR_STATE_ACTIVE | EDP_VSC_PSR_CRC_VALUES_VALID;
+
+ analogix_dp_send_psr_spd(dp, &psr_vsc);
+ return 0;
+}
+EXPORT_SYMBOL_GPL(analogix_dp_enable_psr);
+
+int analogix_dp_disable_psr(struct device *dev)
+{
+ struct analogix_dp_device *dp = dev_get_drvdata(dev);
+ struct edp_vsc_psr psr_vsc;
+
+ if (!dp->psr_support)
+ return -EINVAL;
+
+ /* Prepare VSC packet as per EDP 1.4 spec, Table 6.9 */
+ memset(&psr_vsc, 0, sizeof(psr_vsc));
+ psr_vsc.sdp_header.HB0 = 0;
+ psr_vsc.sdp_header.HB1 = 0x7;
+ psr_vsc.sdp_header.HB2 = 0x2;
+ psr_vsc.sdp_header.HB3 = 0x8;
+
+ psr_vsc.DB0 = 0;
+ psr_vsc.DB1 = 0;
+
+ analogix_dp_send_psr_spd(dp, &psr_vsc);
+ return 0;
+}
+EXPORT_SYMBOL_GPL(analogix_dp_disable_psr);
+
+static bool analogix_dp_detect_sink_psr(struct analogix_dp_device *dp)
+{
+ unsigned char psr_version;
+
+ analogix_dp_read_byte_from_dpcd(dp, DP_PSR_SUPPORT, &psr_version);
+ dev_dbg(dp->dev, "Panel PSR version : %x\n", psr_version);
+
+ return (psr_version & DP_PSR_IS_SUPPORTED) ? true : false;
+}
+
+static void analogix_dp_enable_sink_psr(struct analogix_dp_device *dp)
+{
+ unsigned char psr_en;
+
+ /* Disable psr function */
+ analogix_dp_read_byte_from_dpcd(dp, DP_PSR_EN_CFG, &psr_en);
+ psr_en &= ~DP_PSR_ENABLE;
+ analogix_dp_write_byte_to_dpcd(dp, DP_PSR_EN_CFG, psr_en);
+
+ /* Main-Link transmitter remains active during PSR active states */
+ psr_en = DP_PSR_MAIN_LINK_ACTIVE | DP_PSR_CRC_VERIFICATION;
+ analogix_dp_write_byte_to_dpcd(dp, DP_PSR_EN_CFG, psr_en);
+
+ /* Enable psr function */
+ psr_en = DP_PSR_ENABLE | DP_PSR_MAIN_LINK_ACTIVE |
+ DP_PSR_CRC_VERIFICATION;
+ analogix_dp_write_byte_to_dpcd(dp, DP_PSR_EN_CFG, psr_en);
+
+ analogix_dp_enable_psr_crc(dp);
+}
+
static unsigned char analogix_dp_calc_edid_check_sum(unsigned char *edid_data)
{
int i;
/* Enable video */
analogix_dp_start_video(dp);
+
+ dp->psr_support = analogix_dp_detect_sink_psr(dp);
+ if (dp->psr_support)
+ analogix_dp_enable_sink_psr(dp);
+}
+
+/*
+ * This function is a bit of a catch-all for panel preparation, hopefully
+ * simplifying the logic of functions that need to prepare/unprepare the panel
+ * below.
+ *
+ * If @prepare is true, this function will prepare the panel. Conversely, if it
+ * is false, the panel will be unprepared.
+ *
+ * If @is_modeset_prepare is true, the function will disregard the current state
+ * of the panel and either prepare/unprepare the panel based on @prepare. Once
+ * it finishes, it will update dp->panel_is_modeset to reflect the current state
+ * of the panel.
+ */
+static int analogix_dp_prepare_panel(struct analogix_dp_device *dp,
+ bool prepare, bool is_modeset_prepare)
+{
+ int ret = 0;
+
+ if (!dp->plat_data->panel)
+ return 0;
+
+ mutex_lock(&dp->panel_lock);
+
+ /*
+ * Exit early if this is a temporary prepare/unprepare and we're already
+ * modeset (since we neither want to prepare twice or unprepare early).
+ */
+ if (dp->panel_is_modeset && !is_modeset_prepare)
+ goto out;
+
+ if (prepare)
+ ret = drm_panel_prepare(dp->plat_data->panel);
+ else
+ ret = drm_panel_unprepare(dp->plat_data->panel);
+
+ if (ret)
+ goto out;
+
+ if (is_modeset_prepare)
+ dp->panel_is_modeset = prepare;
+
+out:
+ mutex_unlock(&dp->panel_lock);
+ return ret;
}
int analogix_dp_get_modes(struct drm_connector *connector)
{
struct analogix_dp_device *dp = to_dp(connector);
struct edid *edid = (struct edid *)dp->edid;
- int num_modes = 0;
+ int ret, num_modes = 0;
+
+ ret = analogix_dp_prepare_panel(dp, true, false);
+ if (ret) {
+ DRM_ERROR("Failed to prepare panel (%d)\n", ret);
+ return 0;
+ }
if (analogix_dp_handle_edid(dp) == 0) {
drm_mode_connector_update_edid_property(&dp->connector, edid);
if (dp->plat_data->get_modes)
num_modes += dp->plat_data->get_modes(dp->plat_data, connector);
+ ret = analogix_dp_prepare_panel(dp, false, false);
+ if (ret)
+ DRM_ERROR("Failed to unprepare panel (%d)\n", ret);
+
return num_modes;
}
analogix_dp_detect(struct drm_connector *connector, bool force)
{
struct analogix_dp_device *dp = to_dp(connector);
+ enum drm_connector_status status = connector_status_disconnected;
+ int ret;
- if (analogix_dp_detect_hpd(dp))
+ ret = analogix_dp_prepare_panel(dp, true, false);
+ if (ret) {
+ DRM_ERROR("Failed to prepare panel (%d)\n", ret);
return connector_status_disconnected;
+ }
+
+ if (!analogix_dp_detect_hpd(dp))
+ status = connector_status_connected;
- return connector_status_connected;
+ ret = analogix_dp_prepare_panel(dp, false, false);
+ if (ret)
+ DRM_ERROR("Failed to unprepare panel (%d)\n", ret);
+
+ return status;
}
static void analogix_dp_connector_destroy(struct drm_connector *connector)
return 0;
}
+static void analogix_dp_bridge_pre_enable(struct drm_bridge *bridge)
+{
+ struct analogix_dp_device *dp = bridge->driver_private;
+ int ret;
+
+ ret = analogix_dp_prepare_panel(dp, true, true);
+ if (ret)
+ DRM_ERROR("failed to setup the panel ret = %d\n", ret);
+}
+
static void analogix_dp_bridge_enable(struct drm_bridge *bridge)
{
struct analogix_dp_device *dp = bridge->driver_private;
static void analogix_dp_bridge_disable(struct drm_bridge *bridge)
{
struct analogix_dp_device *dp = bridge->driver_private;
+ int ret;
if (dp->dpms_mode != DRM_MODE_DPMS_ON)
return;
pm_runtime_put_sync(dp->dev);
+ ret = analogix_dp_prepare_panel(dp, false, true);
+ if (ret)
+ DRM_ERROR("failed to setup the panel ret = %d\n", ret);
+
dp->dpms_mode = DRM_MODE_DPMS_OFF;
}
}
static const struct drm_bridge_funcs analogix_dp_bridge_funcs = {
+ .pre_enable = analogix_dp_bridge_pre_enable,
.enable = analogix_dp_bridge_enable,
.disable = analogix_dp_bridge_disable,
- .pre_enable = analogix_dp_bridge_nop,
.post_disable = analogix_dp_bridge_nop,
.mode_set = analogix_dp_bridge_mode_set,
.attach = analogix_dp_bridge_attach,
dp->dev = &pdev->dev;
dp->dpms_mode = DRM_MODE_DPMS_OFF;
+ mutex_init(&dp->panel_lock);
+ dp->panel_is_modeset = false;
+
/*
* platform dp driver need containor_of the plat_data to get
* the driver private data, so we need to store the point of
phy_power_on(dp->phy);
- if (dp->plat_data->panel) {
- if (drm_panel_prepare(dp->plat_data->panel)) {
- DRM_ERROR("failed to setup the panel\n");
- return -EBUSY;
- }
- }
-
analogix_dp_init_dp(dp);
ret = devm_request_threaded_irq(&pdev->dev, dp->irq,
int hpd_gpio;
bool force_hpd;
unsigned char edid[EDID_BLOCK_LENGTH * 2];
+ bool psr_support;
+
+ struct mutex panel_lock;
+ bool panel_is_modeset;
struct analogix_dp_plat_data *plat_data;
};
void analogix_dp_config_video_slave_mode(struct analogix_dp_device *dp);
void analogix_dp_enable_scrambling(struct analogix_dp_device *dp);
void analogix_dp_disable_scrambling(struct analogix_dp_device *dp);
+void analogix_dp_enable_psr_crc(struct analogix_dp_device *dp);
+void analogix_dp_send_psr_spd(struct analogix_dp_device *dp,
+ struct edp_vsc_psr *vsc);
+
#endif /* _ANALOGIX_DP_CORE_H */
u32 analogix_dp_get_lane0_link_training(struct analogix_dp_device *dp)
{
- u32 reg;
-
- reg = readl(dp->reg_base + ANALOGIX_DP_LN0_LINK_TRAINING_CTL);
- return reg;
+ return readl(dp->reg_base + ANALOGIX_DP_LN0_LINK_TRAINING_CTL);
}
u32 analogix_dp_get_lane1_link_training(struct analogix_dp_device *dp)
{
- u32 reg;
-
- reg = readl(dp->reg_base + ANALOGIX_DP_LN1_LINK_TRAINING_CTL);
- return reg;
+ return readl(dp->reg_base + ANALOGIX_DP_LN1_LINK_TRAINING_CTL);
}
u32 analogix_dp_get_lane2_link_training(struct analogix_dp_device *dp)
{
- u32 reg;
-
- reg = readl(dp->reg_base + ANALOGIX_DP_LN2_LINK_TRAINING_CTL);
- return reg;
+ return readl(dp->reg_base + ANALOGIX_DP_LN2_LINK_TRAINING_CTL);
}
u32 analogix_dp_get_lane3_link_training(struct analogix_dp_device *dp)
{
- u32 reg;
-
- reg = readl(dp->reg_base + ANALOGIX_DP_LN3_LINK_TRAINING_CTL);
- return reg;
+ return readl(dp->reg_base + ANALOGIX_DP_LN3_LINK_TRAINING_CTL);
}
void analogix_dp_reset_macro(struct analogix_dp_device *dp)
reg |= SCRAMBLING_DISABLE;
writel(reg, dp->reg_base + ANALOGIX_DP_TRAINING_PTN_SET);
}
+
+void analogix_dp_enable_psr_crc(struct analogix_dp_device *dp)
+{
+ writel(PSR_VID_CRC_ENABLE, dp->reg_base + ANALOGIX_DP_CRC_CON);
+}
+
+void analogix_dp_send_psr_spd(struct analogix_dp_device *dp,
+ struct edp_vsc_psr *vsc)
+{
+ unsigned int val;
+
+ /* don't send info frame */
+ val = readl(dp->reg_base + ANALOGIX_DP_PKT_SEND_CTL);
+ val &= ~IF_EN;
+ writel(val, dp->reg_base + ANALOGIX_DP_PKT_SEND_CTL);
+
+ /* configure single frame update mode */
+ writel(PSR_FRAME_UP_TYPE_BURST | PSR_CRC_SEL_HARDWARE,
+ dp->reg_base + ANALOGIX_DP_PSR_FRAME_UPDATE_CTRL);
+
+ /* configure VSC HB0~HB3 */
+ writel(vsc->sdp_header.HB0, dp->reg_base + ANALOGIX_DP_SPD_HB0);
+ writel(vsc->sdp_header.HB1, dp->reg_base + ANALOGIX_DP_SPD_HB1);
+ writel(vsc->sdp_header.HB2, dp->reg_base + ANALOGIX_DP_SPD_HB2);
+ writel(vsc->sdp_header.HB3, dp->reg_base + ANALOGIX_DP_SPD_HB3);
+
+ /* configure reused VSC PB0~PB3, magic number from vendor */
+ writel(0x00, dp->reg_base + ANALOGIX_DP_SPD_PB0);
+ writel(0x16, dp->reg_base + ANALOGIX_DP_SPD_PB1);
+ writel(0xCE, dp->reg_base + ANALOGIX_DP_SPD_PB2);
+ writel(0x5D, dp->reg_base + ANALOGIX_DP_SPD_PB3);
+
+ /* configure DB0 / DB1 values */
+ writel(vsc->DB0, dp->reg_base + ANALOGIX_DP_VSC_SHADOW_DB0);
+ writel(vsc->DB1, dp->reg_base + ANALOGIX_DP_VSC_SHADOW_DB1);
+
+ /* set reuse spd inforframe */
+ val = readl(dp->reg_base + ANALOGIX_DP_VIDEO_CTL_3);
+ val |= REUSE_SPD_EN;
+ writel(val, dp->reg_base + ANALOGIX_DP_VIDEO_CTL_3);
+
+ /* mark info frame update */
+ val = readl(dp->reg_base + ANALOGIX_DP_PKT_SEND_CTL);
+ val = (val | IF_UP) & ~IF_EN;
+ writel(val, dp->reg_base + ANALOGIX_DP_PKT_SEND_CTL);
+
+ /* send info frame */
+ val = readl(dp->reg_base + ANALOGIX_DP_PKT_SEND_CTL);
+ val |= IF_EN;
+ writel(val, dp->reg_base + ANALOGIX_DP_PKT_SEND_CTL);
+}
#define ANALOGIX_DP_VIDEO_CTL_8 0x3C
#define ANALOGIX_DP_VIDEO_CTL_10 0x44
+#define ANALOGIX_DP_SPDIF_AUDIO_CTL_0 0xD8
+
#define ANALOGIX_DP_PLL_REG_1 0xfc
#define ANALOGIX_DP_PLL_REG_2 0x9e4
#define ANALOGIX_DP_PLL_REG_3 0x9e8
#define ANALOGIX_DP_PD 0x12c
+#define ANALOGIX_DP_IF_TYPE 0x244
+#define ANALOGIX_DP_IF_PKT_DB1 0x254
+#define ANALOGIX_DP_IF_PKT_DB2 0x258
+#define ANALOGIX_DP_SPD_HB0 0x2F8
+#define ANALOGIX_DP_SPD_HB1 0x2FC
+#define ANALOGIX_DP_SPD_HB2 0x300
+#define ANALOGIX_DP_SPD_HB3 0x304
+#define ANALOGIX_DP_SPD_PB0 0x308
+#define ANALOGIX_DP_SPD_PB1 0x30C
+#define ANALOGIX_DP_SPD_PB2 0x310
+#define ANALOGIX_DP_SPD_PB3 0x314
+#define ANALOGIX_DP_PSR_FRAME_UPDATE_CTRL 0x318
+#define ANALOGIX_DP_VSC_SHADOW_DB0 0x31C
+#define ANALOGIX_DP_VSC_SHADOW_DB1 0x320
+
#define ANALOGIX_DP_LANE_MAP 0x35C
#define ANALOGIX_DP_ANALOG_CTL_1 0x370
#define ANALOGIX_DP_SOC_GENERAL_CTL 0x800
+#define ANALOGIX_DP_CRC_CON 0x890
+
/* ANALOGIX_DP_TX_SW_RESET */
#define RESET_DP_TX (0x1 << 0)
#define VID_CHK_UPDATE_TYPE_SHIFT (4)
#define VID_CHK_UPDATE_TYPE_1 (0x1 << 4)
#define VID_CHK_UPDATE_TYPE_0 (0x0 << 4)
+#define REUSE_SPD_EN (0x1 << 3)
/* ANALOGIX_DP_VIDEO_CTL_8 */
#define VID_HRES_TH(x) (((x) & 0xf) << 4)
#define REF_CLK_27M (0x0 << 0)
#define REF_CLK_MASK (0x1 << 0)
+/* ANALOGIX_DP_PSR_FRAME_UPDATE_CTRL */
+#define PSR_FRAME_UP_TYPE_BURST (0x1 << 0)
+#define PSR_FRAME_UP_TYPE_SINGLE (0x0 << 0)
+#define PSR_CRC_SEL_HARDWARE (0x1 << 1)
+#define PSR_CRC_SEL_MANUALLY (0x0 << 1)
+
/* ANALOGIX_DP_LANE_MAP */
#define LANE3_MAP_LOGIC_LANE_0 (0x0 << 6)
#define LANE3_MAP_LOGIC_LANE_1 (0x1 << 6)
#define VIDEO_MODE_SLAVE_MODE (0x1 << 0)
#define VIDEO_MODE_MASTER_MODE (0x0 << 0)
+/* ANALOGIX_DP_PKT_SEND_CTL */
+#define IF_UP (0x1 << 4)
+#define IF_EN (0x1 << 0)
+
+/* ANALOGIX_DP_CRC_CON */
+#define PSR_VID_CRC_FLUSH (0x1 << 2)
+#define PSR_VID_CRC_ENABLE (0x1 << 0)
+
#endif /* _ANALOGIX_DP_REG_H */
*/
/*
- * AVI data byte 1 differences: Colorspace in bits 4,5 rather than 5,6,
- * active aspect present in bit 6 rather than 4.
+ * AVI data byte 1 differences: Colorspace in bits 0,1 rather than 5,6,
+ * scan info in bits 4,5 rather than 0,1 and active aspect present in
+ * bit 6 rather than 4.
*/
- val = (frame.colorspace & 3) << 4 | (frame.scan_mode & 0x3);
+ val = (frame.scan_mode & 3) << 4 | (frame.colorspace & 3);
if (frame.active_aspect & 15)
val |= HDMI_FC_AVICONF0_ACTIVE_FMT_INFO_PRESENT;
if (frame.top_bar || frame.bottom_bar)
goto out;
}
- prop = drm_property_find(dev, prop_id);
+ prop = drm_mode_obj_find_prop_id(obj, prop_id);
if (!prop) {
drm_mode_object_unreference(obj);
ret = -ENOENT;
* drm_atomic_helper_wait_for_fences - wait for fences stashed in plane state
* @dev: DRM device
* @state: atomic state object with old state structures
+ * @pre_swap: if true, do an interruptible wait
*
* For implicit sync, driver should fish the exclusive fence out from the
* incoming fb's and stash it in the drm_plane_state. This is called after
* drm_atomic_helper_swap_state() so it uses the current plane state (and
* just uses the atomic state to find the changed planes)
+ *
+ * Returns zero if success or < 0 if fence_wait() fails.
*/
-void drm_atomic_helper_wait_for_fences(struct drm_device *dev,
- struct drm_atomic_state *state)
+int drm_atomic_helper_wait_for_fences(struct drm_device *dev,
+ struct drm_atomic_state *state,
+ bool pre_swap)
{
struct drm_plane *plane;
struct drm_plane_state *plane_state;
- int i;
+ int i, ret;
for_each_plane_in_state(state, plane, plane_state, i) {
- if (!plane->state->fence)
+ if (!pre_swap)
+ plane_state = plane->state;
+
+ if (!plane_state->fence)
continue;
- WARN_ON(!plane->state->fb);
+ WARN_ON(!plane_state->fb);
+
+ /*
+ * If waiting for fences pre-swap (ie: nonblock), userspace can
+ * still interrupt the operation. Instead of blocking until the
+ * timer expires, make the wait interruptible.
+ */
+ ret = fence_wait(plane_state->fence, pre_swap);
+ if (ret)
+ return ret;
- fence_wait(plane->state->fence, false);
- fence_put(plane->state->fence);
- plane->state->fence = NULL;
+ fence_put(plane_state->fence);
+ plane_state->fence = NULL;
}
+
+ return 0;
}
EXPORT_SYMBOL(drm_atomic_helper_wait_for_fences);
funcs = dev->mode_config.helper_private;
- drm_atomic_helper_wait_for_fences(dev, state);
+ drm_atomic_helper_wait_for_fences(dev, state, false);
drm_atomic_helper_wait_for_dependencies(state);
if (ret)
return ret;
+ if (!nonblock) {
+ ret = drm_atomic_helper_wait_for_fences(dev, state, true);
+ if (ret)
+ return ret;
+ }
+
/*
* This is the point of no return - everything below never fails except
* when the hw goes bonghits. Which means we can commit the new state on
{ DRM_PLANE_TYPE_CURSOR, "Cursor" },
};
-static const struct drm_prop_enum_list drm_encoder_enum_list[] = {
- { DRM_MODE_ENCODER_NONE, "None" },
- { DRM_MODE_ENCODER_DAC, "DAC" },
- { DRM_MODE_ENCODER_TMDS, "TMDS" },
- { DRM_MODE_ENCODER_LVDS, "LVDS" },
- { DRM_MODE_ENCODER_TVDAC, "TV" },
- { DRM_MODE_ENCODER_VIRTUAL, "Virtual" },
- { DRM_MODE_ENCODER_DSI, "DSI" },
- { DRM_MODE_ENCODER_DPMST, "DP MST" },
- { DRM_MODE_ENCODER_DPI, "DPI" },
-};
-
/*
* Optional properties
*/
-/*
- * Internal function to assign a slot in the object idr and optionally
- * register the object into the idr.
- */
-int drm_mode_object_get_reg(struct drm_device *dev,
- struct drm_mode_object *obj,
- uint32_t obj_type,
- bool register_obj,
- void (*obj_free_cb)(struct kref *kref))
-{
- int ret;
-
- mutex_lock(&dev->mode_config.idr_mutex);
- ret = idr_alloc(&dev->mode_config.crtc_idr, register_obj ? obj : NULL, 1, 0, GFP_KERNEL);
- if (ret >= 0) {
- /*
- * Set up the object linking under the protection of the idr
- * lock so that other users can't see inconsistent state.
- */
- obj->id = ret;
- obj->type = obj_type;
- if (obj_free_cb) {
- obj->free_cb = obj_free_cb;
- kref_init(&obj->refcount);
- }
- }
- mutex_unlock(&dev->mode_config.idr_mutex);
-
- return ret < 0 ? ret : 0;
-}
-
-/**
- * drm_mode_object_get - allocate a new modeset identifier
- * @dev: DRM device
- * @obj: object pointer, used to generate unique ID
- * @obj_type: object type
- *
- * Create a unique identifier based on @ptr in @dev's identifier space. Used
- * for tracking modes, CRTCs and connectors. Note that despite the _get postfix
- * modeset identifiers are _not_ reference counted. Hence don't use this for
- * reference counted modeset objects like framebuffers.
- *
- * Returns:
- * Zero on success, error code on failure.
- */
-int drm_mode_object_get(struct drm_device *dev,
- struct drm_mode_object *obj, uint32_t obj_type)
-{
- return drm_mode_object_get_reg(dev, obj, obj_type, true, NULL);
-}
-
-void drm_mode_object_register(struct drm_device *dev,
- struct drm_mode_object *obj)
-{
- mutex_lock(&dev->mode_config.idr_mutex);
- idr_replace(&dev->mode_config.crtc_idr, obj, obj->id);
- mutex_unlock(&dev->mode_config.idr_mutex);
-}
-
-/**
- * drm_mode_object_unregister - free a modeset identifer
- * @dev: DRM device
- * @object: object to free
- *
- * Free @id from @dev's unique identifier pool.
- * This function can be called multiple times, and guards against
- * multiple removals.
- * These modeset identifiers are _not_ reference counted. Hence don't use this
- * for reference counted modeset objects like framebuffers.
- */
-void drm_mode_object_unregister(struct drm_device *dev,
- struct drm_mode_object *object)
-{
- mutex_lock(&dev->mode_config.idr_mutex);
- if (object->id) {
- idr_remove(&dev->mode_config.crtc_idr, object->id);
- object->id = 0;
- }
- mutex_unlock(&dev->mode_config.idr_mutex);
-}
-
-struct drm_mode_object *__drm_mode_object_find(struct drm_device *dev,
- uint32_t id, uint32_t type)
-{
- struct drm_mode_object *obj = NULL;
-
- mutex_lock(&dev->mode_config.idr_mutex);
- obj = idr_find(&dev->mode_config.crtc_idr, id);
- if (obj && type != DRM_MODE_OBJECT_ANY && obj->type != type)
- obj = NULL;
- if (obj && obj->id != id)
- obj = NULL;
-
- if (obj && obj->free_cb) {
- if (!kref_get_unless_zero(&obj->refcount))
- obj = NULL;
- }
- mutex_unlock(&dev->mode_config.idr_mutex);
-
- return obj;
-}
-
-/**
- * drm_mode_object_find - look up a drm object with static lifetime
- * @dev: drm device
- * @id: id of the mode object
- * @type: type of the mode object
- *
- * This function is used to look up a modeset object. It will acquire a
- * reference for reference counted objects. This reference must be dropped again
- * by callind drm_mode_object_unreference().
- */
-struct drm_mode_object *drm_mode_object_find(struct drm_device *dev,
- uint32_t id, uint32_t type)
-{
- struct drm_mode_object *obj = NULL;
-
- obj = __drm_mode_object_find(dev, id, type);
- return obj;
-}
-EXPORT_SYMBOL(drm_mode_object_find);
-
-/**
- * drm_mode_object_unreference - decr the object refcnt
- * @obj: mode_object
- *
- * This functions decrements the object's refcount if it is a refcounted modeset
- * object. It is a no-op on any other object. This is used to drop references
- * acquired with drm_mode_object_reference().
- */
-void drm_mode_object_unreference(struct drm_mode_object *obj)
-{
- if (obj->free_cb) {
- DRM_DEBUG("OBJ ID: %d (%d)\n", obj->id, atomic_read(&obj->refcount.refcount));
- kref_put(&obj->refcount, obj->free_cb);
- }
-}
-EXPORT_SYMBOL(drm_mode_object_unreference);
-
-/**
- * drm_mode_object_reference - incr the object refcnt
- * @obj: mode_object
- *
- * This functions increments the object's refcount if it is a refcounted modeset
- * object. It is a no-op on any other object. References should be dropped again
- * by calling drm_mode_object_unreference().
- */
-void drm_mode_object_reference(struct drm_mode_object *obj)
-{
- if (obj->free_cb) {
- DRM_DEBUG("OBJ ID: %d (%d)\n", obj->id, atomic_read(&obj->refcount.refcount));
- kref_get(&obj->refcount);
- }
-}
-EXPORT_SYMBOL(drm_mode_object_reference);
-
/**
* drm_crtc_force_disable - Forcibly turn off a CRTC
* @crtc: CRTC to turn off
}
EXPORT_SYMBOL(drm_crtc_cleanup);
-static int drm_encoder_register_all(struct drm_device *dev)
-{
- struct drm_encoder *encoder;
- int ret = 0;
-
- drm_for_each_encoder(encoder, dev) {
- if (encoder->funcs->late_register)
- ret = encoder->funcs->late_register(encoder);
- if (ret)
- return ret;
- }
-
- return 0;
-}
-
-static void drm_encoder_unregister_all(struct drm_device *dev)
-{
- struct drm_encoder *encoder;
-
- drm_for_each_encoder(encoder, dev) {
- if (encoder->funcs->early_unregister)
- encoder->funcs->early_unregister(encoder);
- }
-}
-
-/**
- * drm_encoder_init - Init a preallocated encoder
- * @dev: drm device
- * @encoder: the encoder to init
- * @funcs: callbacks for this encoder
- * @encoder_type: user visible type of the encoder
- * @name: printf style format string for the encoder name, or NULL for default name
- *
- * Initialises a preallocated encoder. Encoder should be
- * subclassed as part of driver encoder objects.
- *
- * Returns:
- * Zero on success, error code on failure.
- */
-int drm_encoder_init(struct drm_device *dev,
- struct drm_encoder *encoder,
- const struct drm_encoder_funcs *funcs,
- int encoder_type, const char *name, ...)
-{
- int ret;
-
- drm_modeset_lock_all(dev);
-
- ret = drm_mode_object_get(dev, &encoder->base, DRM_MODE_OBJECT_ENCODER);
- if (ret)
- goto out_unlock;
-
- encoder->dev = dev;
- encoder->encoder_type = encoder_type;
- encoder->funcs = funcs;
- if (name) {
- va_list ap;
-
- va_start(ap, name);
- encoder->name = kvasprintf(GFP_KERNEL, name, ap);
- va_end(ap);
- } else {
- encoder->name = kasprintf(GFP_KERNEL, "%s-%d",
- drm_encoder_enum_list[encoder_type].name,
- encoder->base.id);
- }
- if (!encoder->name) {
- ret = -ENOMEM;
- goto out_put;
- }
-
- list_add_tail(&encoder->head, &dev->mode_config.encoder_list);
- encoder->index = dev->mode_config.num_encoder++;
-
-out_put:
- if (ret)
- drm_mode_object_unregister(dev, &encoder->base);
-
-out_unlock:
- drm_modeset_unlock_all(dev);
-
- return ret;
-}
-EXPORT_SYMBOL(drm_encoder_init);
-
-/**
- * drm_encoder_cleanup - cleans up an initialised encoder
- * @encoder: encoder to cleanup
- *
- * Cleans up the encoder but doesn't free the object.
- */
-void drm_encoder_cleanup(struct drm_encoder *encoder)
-{
- struct drm_device *dev = encoder->dev;
-
- /* Note that the encoder_list is considered to be static; should we
- * remove the drm_encoder at runtime we would have to decrement all
- * the indices on the drm_encoder after us in the encoder_list.
- */
-
- drm_modeset_lock_all(dev);
- drm_mode_object_unregister(dev, &encoder->base);
- kfree(encoder->name);
- list_del(&encoder->head);
- dev->mode_config.num_encoder--;
- drm_modeset_unlock_all(dev);
-
- memset(encoder, 0, sizeof(*encoder));
-}
-EXPORT_SYMBOL(drm_encoder_cleanup);
-
static unsigned int drm_num_planes(struct drm_device *dev)
{
unsigned int num = 0;
return 0;
}
-/* helper for getconnector and getproperties ioctls */
-int drm_mode_object_get_properties(struct drm_mode_object *obj, bool atomic,
- uint32_t __user *prop_ptr,
- uint64_t __user *prop_values,
- uint32_t *arg_count_props)
-{
- int props_count;
- int i, ret, copied;
-
- props_count = obj->properties->count;
- if (!atomic)
- props_count -= obj->properties->atomic_count;
-
- if ((*arg_count_props >= props_count) && props_count) {
- for (i = 0, copied = 0; copied < props_count; i++) {
- struct drm_property *prop = obj->properties->properties[i];
- uint64_t val;
-
- if ((prop->flags & DRM_MODE_PROP_ATOMIC) && !atomic)
- continue;
-
- ret = drm_object_property_get_value(obj, prop, &val);
- if (ret)
- return ret;
-
- if (put_user(prop->base.id, prop_ptr + copied))
- return -EFAULT;
-
- if (put_user(val, prop_values + copied))
- return -EFAULT;
-
- copied++;
- }
- }
- *arg_count_props = props_count;
-
- return 0;
-}
-
-static struct drm_crtc *drm_encoder_get_crtc(struct drm_encoder *encoder)
-{
- struct drm_connector *connector;
- struct drm_device *dev = encoder->dev;
- bool uses_atomic = false;
-
- /* For atomic drivers only state objects are synchronously updated and
- * protected by modeset locks, so check those first. */
- drm_for_each_connector(connector, dev) {
- if (!connector->state)
- continue;
-
- uses_atomic = true;
-
- if (connector->state->best_encoder != encoder)
- continue;
-
- return connector->state->crtc;
- }
-
- /* Don't return stale data (e.g. pending async disable). */
- if (uses_atomic)
- return NULL;
-
- return encoder->crtc;
-}
-
-/**
- * drm_mode_getencoder - get encoder configuration
- * @dev: drm device for the ioctl
- * @data: data pointer for the ioctl
- * @file_priv: drm file for the ioctl call
- *
- * Construct a encoder configuration structure to return to the user.
- *
- * Called by the user via ioctl.
- *
- * Returns:
- * Zero on success, negative errno on failure.
- */
-int drm_mode_getencoder(struct drm_device *dev, void *data,
- struct drm_file *file_priv)
-{
- struct drm_mode_get_encoder *enc_resp = data;
- struct drm_encoder *encoder;
- struct drm_crtc *crtc;
-
- if (!drm_core_check_feature(dev, DRIVER_MODESET))
- return -EINVAL;
-
- encoder = drm_encoder_find(dev, enc_resp->encoder_id);
- if (!encoder)
- return -ENOENT;
-
- drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
- crtc = drm_encoder_get_crtc(encoder);
- if (crtc)
- enc_resp->crtc_id = crtc->base.id;
- else
- enc_resp->crtc_id = 0;
- drm_modeset_unlock(&dev->mode_config.connection_mutex);
-
- enc_resp->encoder_type = encoder->encoder_type;
- enc_resp->encoder_id = encoder->base.id;
- enc_resp->possible_crtcs = encoder->possible_crtcs;
- enc_resp->possible_clones = encoder->possible_clones;
-
- return 0;
-}
-
/**
* drm_mode_getplane_res - enumerate all plane resources
* @dev: DRM device
return drm_mode_cursor_common(dev, req, file_priv);
}
-/**
- * drm_mode_legacy_fb_format - compute drm fourcc code from legacy description
- * @bpp: bits per pixels
- * @depth: bit depth per pixel
- *
- * Computes a drm fourcc pixel format code for the given @bpp/@depth values.
- * Useful in fbdev emulation code, since that deals in those values.
- */
-uint32_t drm_mode_legacy_fb_format(uint32_t bpp, uint32_t depth)
+int drm_mode_crtc_set_obj_prop(struct drm_mode_object *obj,
+ struct drm_property *property,
+ uint64_t value)
{
- uint32_t fmt;
-
- switch (bpp) {
- case 8:
- fmt = DRM_FORMAT_C8;
- break;
- case 16:
- if (depth == 15)
- fmt = DRM_FORMAT_XRGB1555;
- else
- fmt = DRM_FORMAT_RGB565;
- break;
- case 24:
- fmt = DRM_FORMAT_RGB888;
- break;
- case 32:
- if (depth == 24)
- fmt = DRM_FORMAT_XRGB8888;
- else if (depth == 30)
- fmt = DRM_FORMAT_XRGB2101010;
- else
- fmt = DRM_FORMAT_ARGB8888;
- break;
- default:
- DRM_ERROR("bad bpp, assuming x8r8g8b8 pixel format\n");
- fmt = DRM_FORMAT_XRGB8888;
- break;
- }
+ int ret = -EINVAL;
+ struct drm_crtc *crtc = obj_to_crtc(obj);
- return fmt;
-}
-EXPORT_SYMBOL(drm_mode_legacy_fb_format);
+ if (crtc->funcs->set_property)
+ ret = crtc->funcs->set_property(crtc, property, value);
+ if (!ret)
+ drm_object_property_set_value(obj, property, value);
-static bool drm_property_type_valid(struct drm_property *property)
-{
- if (property->flags & DRM_MODE_PROP_EXTENDED_TYPE)
- return !(property->flags & DRM_MODE_PROP_LEGACY_TYPE);
- return !!(property->flags & DRM_MODE_PROP_LEGACY_TYPE);
+ return ret;
}
/**
- * drm_property_create - create a new property type
- * @dev: drm device
- * @flags: flags specifying the property type
- * @name: name of the property
- * @num_values: number of pre-defined values
- *
- * This creates a new generic drm property which can then be attached to a drm
- * object with drm_object_attach_property. The returned property object must be
- * freed with drm_property_destroy.
+ * drm_mode_plane_set_obj_prop - set the value of a property
+ * @plane: drm plane object to set property value for
+ * @property: property to set
+ * @value: value the property should be set to
*
- * Note that the DRM core keeps a per-device list of properties and that, if
- * drm_mode_config_cleanup() is called, it will destroy all properties created
- * by the driver.
+ * This functions sets a given property on a given plane object. This function
+ * calls the driver's ->set_property callback and changes the software state of
+ * the property if the callback succeeds.
*
* Returns:
- * A pointer to the newly created property on success, NULL on failure.
+ * Zero on success, error code on failure.
*/
-struct drm_property *drm_property_create(struct drm_device *dev, int flags,
- const char *name, int num_values)
+int drm_mode_plane_set_obj_prop(struct drm_plane *plane,
+ struct drm_property *property,
+ uint64_t value)
{
- struct drm_property *property = NULL;
- int ret;
-
- property = kzalloc(sizeof(struct drm_property), GFP_KERNEL);
- if (!property)
- return NULL;
-
- property->dev = dev;
-
- if (num_values) {
- property->values = kcalloc(num_values, sizeof(uint64_t),
- GFP_KERNEL);
- if (!property->values)
- goto fail;
- }
-
- ret = drm_mode_object_get(dev, &property->base, DRM_MODE_OBJECT_PROPERTY);
- if (ret)
- goto fail;
-
- property->flags = flags;
- property->num_values = num_values;
- INIT_LIST_HEAD(&property->enum_list);
-
- if (name) {
- strncpy(property->name, name, DRM_PROP_NAME_LEN);
- property->name[DRM_PROP_NAME_LEN-1] = '\0';
- }
-
- list_add_tail(&property->head, &dev->mode_config.property_list);
+ int ret = -EINVAL;
+ struct drm_mode_object *obj = &plane->base;
- WARN_ON(!drm_property_type_valid(property));
+ if (plane->funcs->set_property)
+ ret = plane->funcs->set_property(plane, property, value);
+ if (!ret)
+ drm_object_property_set_value(obj, property, value);
- return property;
-fail:
- kfree(property->values);
- kfree(property);
- return NULL;
+ return ret;
}
-EXPORT_SYMBOL(drm_property_create);
+EXPORT_SYMBOL(drm_mode_plane_set_obj_prop);
/**
- * drm_property_create_enum - create a new enumeration property type
- * @dev: drm device
- * @flags: flags specifying the property type
- * @name: name of the property
- * @props: enumeration lists with property values
- * @num_values: number of pre-defined values
- *
- * This creates a new generic drm property which can then be attached to a drm
- * object with drm_object_attach_property. The returned property object must be
- * freed with drm_property_destroy.
+ * drm_mode_crtc_set_gamma_size - set the gamma table size
+ * @crtc: CRTC to set the gamma table size for
+ * @gamma_size: size of the gamma table
*
- * Userspace is only allowed to set one of the predefined values for enumeration
- * properties.
+ * Drivers which support gamma tables should set this to the supported gamma
+ * table size when initializing the CRTC. Currently the drm core only supports a
+ * fixed gamma table size.
*
* Returns:
- * A pointer to the newly created property on success, NULL on failure.
+ * Zero on success, negative errno on failure.
*/
-struct drm_property *drm_property_create_enum(struct drm_device *dev, int flags,
- const char *name,
- const struct drm_prop_enum_list *props,
- int num_values)
+int drm_mode_crtc_set_gamma_size(struct drm_crtc *crtc,
+ int gamma_size)
{
- struct drm_property *property;
- int i, ret;
-
- flags |= DRM_MODE_PROP_ENUM;
+ uint16_t *r_base, *g_base, *b_base;
+ int i;
- property = drm_property_create(dev, flags, name, num_values);
- if (!property)
- return NULL;
-
- for (i = 0; i < num_values; i++) {
- ret = drm_property_add_enum(property, i,
- props[i].type,
- props[i].name);
- if (ret) {
- drm_property_destroy(dev, property);
- return NULL;
- }
- }
-
- return property;
-}
-EXPORT_SYMBOL(drm_property_create_enum);
-
-/**
- * drm_property_create_bitmask - create a new bitmask property type
- * @dev: drm device
- * @flags: flags specifying the property type
- * @name: name of the property
- * @props: enumeration lists with property bitflags
- * @num_props: size of the @props array
- * @supported_bits: bitmask of all supported enumeration values
- *
- * This creates a new bitmask drm property which can then be attached to a drm
- * object with drm_object_attach_property. The returned property object must be
- * freed with drm_property_destroy.
- *
- * Compared to plain enumeration properties userspace is allowed to set any
- * or'ed together combination of the predefined property bitflag values
- *
- * Returns:
- * A pointer to the newly created property on success, NULL on failure.
- */
-struct drm_property *drm_property_create_bitmask(struct drm_device *dev,
- int flags, const char *name,
- const struct drm_prop_enum_list *props,
- int num_props,
- uint64_t supported_bits)
-{
- struct drm_property *property;
- int i, ret, index = 0;
- int num_values = hweight64(supported_bits);
-
- flags |= DRM_MODE_PROP_BITMASK;
-
- property = drm_property_create(dev, flags, name, num_values);
- if (!property)
- return NULL;
- for (i = 0; i < num_props; i++) {
- if (!(supported_bits & (1ULL << props[i].type)))
- continue;
-
- if (WARN_ON(index >= num_values)) {
- drm_property_destroy(dev, property);
- return NULL;
- }
-
- ret = drm_property_add_enum(property, index++,
- props[i].type,
- props[i].name);
- if (ret) {
- drm_property_destroy(dev, property);
- return NULL;
- }
- }
-
- return property;
-}
-EXPORT_SYMBOL(drm_property_create_bitmask);
-
-static struct drm_property *property_create_range(struct drm_device *dev,
- int flags, const char *name,
- uint64_t min, uint64_t max)
-{
- struct drm_property *property;
-
- property = drm_property_create(dev, flags, name, 2);
- if (!property)
- return NULL;
-
- property->values[0] = min;
- property->values[1] = max;
-
- return property;
-}
-
-/**
- * drm_property_create_range - create a new unsigned ranged property type
- * @dev: drm device
- * @flags: flags specifying the property type
- * @name: name of the property
- * @min: minimum value of the property
- * @max: maximum value of the property
- *
- * This creates a new generic drm property which can then be attached to a drm
- * object with drm_object_attach_property. The returned property object must be
- * freed with drm_property_destroy.
- *
- * Userspace is allowed to set any unsigned integer value in the (min, max)
- * range inclusive.
- *
- * Returns:
- * A pointer to the newly created property on success, NULL on failure.
- */
-struct drm_property *drm_property_create_range(struct drm_device *dev, int flags,
- const char *name,
- uint64_t min, uint64_t max)
-{
- return property_create_range(dev, DRM_MODE_PROP_RANGE | flags,
- name, min, max);
-}
-EXPORT_SYMBOL(drm_property_create_range);
-
-/**
- * drm_property_create_signed_range - create a new signed ranged property type
- * @dev: drm device
- * @flags: flags specifying the property type
- * @name: name of the property
- * @min: minimum value of the property
- * @max: maximum value of the property
- *
- * This creates a new generic drm property which can then be attached to a drm
- * object with drm_object_attach_property. The returned property object must be
- * freed with drm_property_destroy.
- *
- * Userspace is allowed to set any signed integer value in the (min, max)
- * range inclusive.
- *
- * Returns:
- * A pointer to the newly created property on success, NULL on failure.
- */
-struct drm_property *drm_property_create_signed_range(struct drm_device *dev,
- int flags, const char *name,
- int64_t min, int64_t max)
-{
- return property_create_range(dev, DRM_MODE_PROP_SIGNED_RANGE | flags,
- name, I642U64(min), I642U64(max));
-}
-EXPORT_SYMBOL(drm_property_create_signed_range);
-
-/**
- * drm_property_create_object - create a new object property type
- * @dev: drm device
- * @flags: flags specifying the property type
- * @name: name of the property
- * @type: object type from DRM_MODE_OBJECT_* defines
- *
- * This creates a new generic drm property which can then be attached to a drm
- * object with drm_object_attach_property. The returned property object must be
- * freed with drm_property_destroy.
- *
- * Userspace is only allowed to set this to any property value of the given
- * @type. Only useful for atomic properties, which is enforced.
- *
- * Returns:
- * A pointer to the newly created property on success, NULL on failure.
- */
-struct drm_property *drm_property_create_object(struct drm_device *dev,
- int flags, const char *name, uint32_t type)
-{
- struct drm_property *property;
-
- flags |= DRM_MODE_PROP_OBJECT;
-
- if (WARN_ON(!(flags & DRM_MODE_PROP_ATOMIC)))
- return NULL;
-
- property = drm_property_create(dev, flags, name, 1);
- if (!property)
- return NULL;
-
- property->values[0] = type;
-
- return property;
-}
-EXPORT_SYMBOL(drm_property_create_object);
-
-/**
- * drm_property_create_bool - create a new boolean property type
- * @dev: drm device
- * @flags: flags specifying the property type
- * @name: name of the property
- *
- * This creates a new generic drm property which can then be attached to a drm
- * object with drm_object_attach_property. The returned property object must be
- * freed with drm_property_destroy.
- *
- * This is implemented as a ranged property with only {0, 1} as valid values.
- *
- * Returns:
- * A pointer to the newly created property on success, NULL on failure.
- */
-struct drm_property *drm_property_create_bool(struct drm_device *dev, int flags,
- const char *name)
-{
- return drm_property_create_range(dev, flags, name, 0, 1);
-}
-EXPORT_SYMBOL(drm_property_create_bool);
-
-/**
- * drm_property_add_enum - add a possible value to an enumeration property
- * @property: enumeration property to change
- * @index: index of the new enumeration
- * @value: value of the new enumeration
- * @name: symbolic name of the new enumeration
- *
- * This functions adds enumerations to a property.
- *
- * It's use is deprecated, drivers should use one of the more specific helpers
- * to directly create the property with all enumerations already attached.
- *
- * Returns:
- * Zero on success, error code on failure.
- */
-int drm_property_add_enum(struct drm_property *property, int index,
- uint64_t value, const char *name)
-{
- struct drm_property_enum *prop_enum;
-
- if (!(drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
- drm_property_type_is(property, DRM_MODE_PROP_BITMASK)))
- return -EINVAL;
-
- /*
- * Bitmask enum properties have the additional constraint of values
- * from 0 to 63
- */
- if (drm_property_type_is(property, DRM_MODE_PROP_BITMASK) &&
- (value > 63))
- return -EINVAL;
-
- if (!list_empty(&property->enum_list)) {
- list_for_each_entry(prop_enum, &property->enum_list, head) {
- if (prop_enum->value == value) {
- strncpy(prop_enum->name, name, DRM_PROP_NAME_LEN);
- prop_enum->name[DRM_PROP_NAME_LEN-1] = '\0';
- return 0;
- }
- }
- }
-
- prop_enum = kzalloc(sizeof(struct drm_property_enum), GFP_KERNEL);
- if (!prop_enum)
- return -ENOMEM;
-
- strncpy(prop_enum->name, name, DRM_PROP_NAME_LEN);
- prop_enum->name[DRM_PROP_NAME_LEN-1] = '\0';
- prop_enum->value = value;
-
- property->values[index] = value;
- list_add_tail(&prop_enum->head, &property->enum_list);
- return 0;
-}
-EXPORT_SYMBOL(drm_property_add_enum);
-
-/**
- * drm_property_destroy - destroy a drm property
- * @dev: drm device
- * @property: property to destry
- *
- * This function frees a property including any attached resources like
- * enumeration values.
- */
-void drm_property_destroy(struct drm_device *dev, struct drm_property *property)
-{
- struct drm_property_enum *prop_enum, *pt;
-
- list_for_each_entry_safe(prop_enum, pt, &property->enum_list, head) {
- list_del(&prop_enum->head);
- kfree(prop_enum);
- }
-
- if (property->num_values)
- kfree(property->values);
- drm_mode_object_unregister(dev, &property->base);
- list_del(&property->head);
- kfree(property);
-}
-EXPORT_SYMBOL(drm_property_destroy);
-
-/**
- * drm_object_attach_property - attach a property to a modeset object
- * @obj: drm modeset object
- * @property: property to attach
- * @init_val: initial value of the property
- *
- * This attaches the given property to the modeset object with the given initial
- * value. Currently this function cannot fail since the properties are stored in
- * a statically sized array.
- */
-void drm_object_attach_property(struct drm_mode_object *obj,
- struct drm_property *property,
- uint64_t init_val)
-{
- int count = obj->properties->count;
-
- if (count == DRM_OBJECT_MAX_PROPERTY) {
- WARN(1, "Failed to attach object property (type: 0x%x). Please "
- "increase DRM_OBJECT_MAX_PROPERTY by 1 for each time "
- "you see this message on the same object type.\n",
- obj->type);
- return;
- }
-
- obj->properties->properties[count] = property;
- obj->properties->values[count] = init_val;
- obj->properties->count++;
- if (property->flags & DRM_MODE_PROP_ATOMIC)
- obj->properties->atomic_count++;
-}
-EXPORT_SYMBOL(drm_object_attach_property);
-
-/**
- * drm_object_property_set_value - set the value of a property
- * @obj: drm mode object to set property value for
- * @property: property to set
- * @val: value the property should be set to
- *
- * This functions sets a given property on a given object. This function only
- * changes the software state of the property, it does not call into the
- * driver's ->set_property callback.
- *
- * Returns:
- * Zero on success, error code on failure.
- */
-int drm_object_property_set_value(struct drm_mode_object *obj,
- struct drm_property *property, uint64_t val)
-{
- int i;
-
- for (i = 0; i < obj->properties->count; i++) {
- if (obj->properties->properties[i] == property) {
- obj->properties->values[i] = val;
- return 0;
- }
- }
-
- return -EINVAL;
-}
-EXPORT_SYMBOL(drm_object_property_set_value);
-
-/**
- * drm_object_property_get_value - retrieve the value of a property
- * @obj: drm mode object to get property value from
- * @property: property to retrieve
- * @val: storage for the property value
- *
- * This function retrieves the softare state of the given property for the given
- * property. Since there is no driver callback to retrieve the current property
- * value this might be out of sync with the hardware, depending upon the driver
- * and property.
- *
- * Returns:
- * Zero on success, error code on failure.
- */
-int drm_object_property_get_value(struct drm_mode_object *obj,
- struct drm_property *property, uint64_t *val)
-{
- int i;
-
- /* read-only properties bypass atomic mechanism and still store
- * their value in obj->properties->values[].. mostly to avoid
- * having to deal w/ EDID and similar props in atomic paths:
- */
- if (drm_core_check_feature(property->dev, DRIVER_ATOMIC) &&
- !(property->flags & DRM_MODE_PROP_IMMUTABLE))
- return drm_atomic_get_property(obj, property, val);
-
- for (i = 0; i < obj->properties->count; i++) {
- if (obj->properties->properties[i] == property) {
- *val = obj->properties->values[i];
- return 0;
- }
- }
-
- return -EINVAL;
-}
-EXPORT_SYMBOL(drm_object_property_get_value);
-
-/**
- * drm_mode_getproperty_ioctl - get the property metadata
- * @dev: DRM device
- * @data: ioctl data
- * @file_priv: DRM file info
- *
- * This function retrieves the metadata for a given property, like the different
- * possible values for an enum property or the limits for a range property.
- *
- * Blob properties are special
- *
- * Called by the user via ioctl.
- *
- * Returns:
- * Zero on success, negative errno on failure.
- */
-int drm_mode_getproperty_ioctl(struct drm_device *dev,
- void *data, struct drm_file *file_priv)
-{
- struct drm_mode_get_property *out_resp = data;
- struct drm_property *property;
- int enum_count = 0;
- int value_count = 0;
- int ret = 0, i;
- int copied;
- struct drm_property_enum *prop_enum;
- struct drm_mode_property_enum __user *enum_ptr;
- uint64_t __user *values_ptr;
-
- if (!drm_core_check_feature(dev, DRIVER_MODESET))
- return -EINVAL;
-
- drm_modeset_lock_all(dev);
- property = drm_property_find(dev, out_resp->prop_id);
- if (!property) {
- ret = -ENOENT;
- goto done;
- }
-
- if (drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
- drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
- list_for_each_entry(prop_enum, &property->enum_list, head)
- enum_count++;
- }
-
- value_count = property->num_values;
-
- strncpy(out_resp->name, property->name, DRM_PROP_NAME_LEN);
- out_resp->name[DRM_PROP_NAME_LEN-1] = 0;
- out_resp->flags = property->flags;
-
- if ((out_resp->count_values >= value_count) && value_count) {
- values_ptr = (uint64_t __user *)(unsigned long)out_resp->values_ptr;
- for (i = 0; i < value_count; i++) {
- if (copy_to_user(values_ptr + i, &property->values[i], sizeof(uint64_t))) {
- ret = -EFAULT;
- goto done;
- }
- }
- }
- out_resp->count_values = value_count;
-
- if (drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
- drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
- if ((out_resp->count_enum_blobs >= enum_count) && enum_count) {
- copied = 0;
- enum_ptr = (struct drm_mode_property_enum __user *)(unsigned long)out_resp->enum_blob_ptr;
- list_for_each_entry(prop_enum, &property->enum_list, head) {
-
- if (copy_to_user(&enum_ptr[copied].value, &prop_enum->value, sizeof(uint64_t))) {
- ret = -EFAULT;
- goto done;
- }
-
- if (copy_to_user(&enum_ptr[copied].name,
- &prop_enum->name, DRM_PROP_NAME_LEN)) {
- ret = -EFAULT;
- goto done;
- }
- copied++;
- }
- }
- out_resp->count_enum_blobs = enum_count;
- }
-
- /*
- * NOTE: The idea seems to have been to use this to read all the blob
- * property values. But nothing ever added them to the corresponding
- * list, userspace always used the special-purpose get_blob ioctl to
- * read the value for a blob property. It also doesn't make a lot of
- * sense to return values here when everything else is just metadata for
- * the property itself.
- */
- if (drm_property_type_is(property, DRM_MODE_PROP_BLOB))
- out_resp->count_enum_blobs = 0;
-done:
- drm_modeset_unlock_all(dev);
- return ret;
-}
-
-static void drm_property_free_blob(struct kref *kref)
-{
- struct drm_property_blob *blob =
- container_of(kref, struct drm_property_blob, base.refcount);
-
- mutex_lock(&blob->dev->mode_config.blob_lock);
- list_del(&blob->head_global);
- mutex_unlock(&blob->dev->mode_config.blob_lock);
-
- drm_mode_object_unregister(blob->dev, &blob->base);
-
- kfree(blob);
-}
-
-/**
- * drm_property_create_blob - Create new blob property
- *
- * Creates a new blob property for a specified DRM device, optionally
- * copying data.
- *
- * @dev: DRM device to create property for
- * @length: Length to allocate for blob data
- * @data: If specified, copies data into blob
- *
- * Returns:
- * New blob property with a single reference on success, or an ERR_PTR
- * value on failure.
- */
-struct drm_property_blob *
-drm_property_create_blob(struct drm_device *dev, size_t length,
- const void *data)
-{
- struct drm_property_blob *blob;
- int ret;
-
- if (!length || length > ULONG_MAX - sizeof(struct drm_property_blob))
- return ERR_PTR(-EINVAL);
-
- blob = kzalloc(sizeof(struct drm_property_blob)+length, GFP_KERNEL);
- if (!blob)
- return ERR_PTR(-ENOMEM);
-
- /* This must be explicitly initialised, so we can safely call list_del
- * on it in the removal handler, even if it isn't in a file list. */
- INIT_LIST_HEAD(&blob->head_file);
- blob->length = length;
- blob->dev = dev;
-
- if (data)
- memcpy(blob->data, data, length);
-
- ret = drm_mode_object_get_reg(dev, &blob->base, DRM_MODE_OBJECT_BLOB,
- true, drm_property_free_blob);
- if (ret) {
- kfree(blob);
- return ERR_PTR(-EINVAL);
- }
-
- mutex_lock(&dev->mode_config.blob_lock);
- list_add_tail(&blob->head_global,
- &dev->mode_config.property_blob_list);
- mutex_unlock(&dev->mode_config.blob_lock);
-
- return blob;
-}
-EXPORT_SYMBOL(drm_property_create_blob);
-
-/**
- * drm_property_unreference_blob - Unreference a blob property
- *
- * Drop a reference on a blob property. May free the object.
- *
- * @blob: Pointer to blob property
- */
-void drm_property_unreference_blob(struct drm_property_blob *blob)
-{
- if (!blob)
- return;
-
- drm_mode_object_unreference(&blob->base);
-}
-EXPORT_SYMBOL(drm_property_unreference_blob);
-
-/**
- * drm_property_destroy_user_blobs - destroy all blobs created by this client
- * @dev: DRM device
- * @file_priv: destroy all blobs owned by this file handle
- */
-void drm_property_destroy_user_blobs(struct drm_device *dev,
- struct drm_file *file_priv)
-{
- struct drm_property_blob *blob, *bt;
-
- /*
- * When the file gets released that means no one else can access the
- * blob list any more, so no need to grab dev->blob_lock.
- */
- list_for_each_entry_safe(blob, bt, &file_priv->blobs, head_file) {
- list_del_init(&blob->head_file);
- drm_property_unreference_blob(blob);
- }
-}
-
-/**
- * drm_property_reference_blob - Take a reference on an existing property
- *
- * Take a new reference on an existing blob property.
- *
- * @blob: Pointer to blob property
- */
-struct drm_property_blob *drm_property_reference_blob(struct drm_property_blob *blob)
-{
- drm_mode_object_reference(&blob->base);
- return blob;
-}
-EXPORT_SYMBOL(drm_property_reference_blob);
-
-/**
- * drm_property_lookup_blob - look up a blob property and take a reference
- * @dev: drm device
- * @id: id of the blob property
- *
- * If successful, this takes an additional reference to the blob property.
- * callers need to make sure to eventually unreference the returned property
- * again, using @drm_property_unreference_blob.
- */
-struct drm_property_blob *drm_property_lookup_blob(struct drm_device *dev,
- uint32_t id)
-{
- struct drm_mode_object *obj;
- struct drm_property_blob *blob = NULL;
-
- obj = __drm_mode_object_find(dev, id, DRM_MODE_OBJECT_BLOB);
- if (obj)
- blob = obj_to_blob(obj);
- return blob;
-}
-EXPORT_SYMBOL(drm_property_lookup_blob);
-
-/**
- * drm_property_replace_global_blob - atomically replace existing blob property
- * @dev: drm device
- * @replace: location of blob property pointer to be replaced
- * @length: length of data for new blob, or 0 for no data
- * @data: content for new blob, or NULL for no data
- * @obj_holds_id: optional object for property holding blob ID
- * @prop_holds_id: optional property holding blob ID
- * @return 0 on success or error on failure
- *
- * This function will atomically replace a global property in the blob list,
- * optionally updating a property which holds the ID of that property. It is
- * guaranteed to be atomic: no caller will be allowed to see intermediate
- * results, and either the entire operation will succeed and clean up the
- * previous property, or it will fail and the state will be unchanged.
- *
- * If length is 0 or data is NULL, no new blob will be created, and the holding
- * property, if specified, will be set to 0.
- *
- * Access to the replace pointer is assumed to be protected by the caller, e.g.
- * by holding the relevant modesetting object lock for its parent.
- *
- * For example, a drm_connector has a 'PATH' property, which contains the ID
- * of a blob property with the value of the MST path information. Calling this
- * function with replace pointing to the connector's path_blob_ptr, length and
- * data set for the new path information, obj_holds_id set to the connector's
- * base object, and prop_holds_id set to the path property name, will perform
- * a completely atomic update. The access to path_blob_ptr is protected by the
- * caller holding a lock on the connector.
- */
-int drm_property_replace_global_blob(struct drm_device *dev,
- struct drm_property_blob **replace,
- size_t length,
- const void *data,
- struct drm_mode_object *obj_holds_id,
- struct drm_property *prop_holds_id)
-{
- struct drm_property_blob *new_blob = NULL;
- struct drm_property_blob *old_blob = NULL;
- int ret;
-
- WARN_ON(replace == NULL);
-
- old_blob = *replace;
-
- if (length && data) {
- new_blob = drm_property_create_blob(dev, length, data);
- if (IS_ERR(new_blob))
- return PTR_ERR(new_blob);
- }
-
- /* This does not need to be synchronised with blob_lock, as the
- * get_properties ioctl locks all modesetting objects, and
- * obj_holds_id must be locked before calling here, so we cannot
- * have its value out of sync with the list membership modified
- * below under blob_lock. */
- if (obj_holds_id) {
- ret = drm_object_property_set_value(obj_holds_id,
- prop_holds_id,
- new_blob ?
- new_blob->base.id : 0);
- if (ret != 0)
- goto err_created;
- }
-
- drm_property_unreference_blob(old_blob);
- *replace = new_blob;
-
- return 0;
-
-err_created:
- drm_property_unreference_blob(new_blob);
- return ret;
-}
-EXPORT_SYMBOL(drm_property_replace_global_blob);
-
-/**
- * drm_mode_getblob_ioctl - get the contents of a blob property value
- * @dev: DRM device
- * @data: ioctl data
- * @file_priv: DRM file info
- *
- * This function retrieves the contents of a blob property. The value stored in
- * an object's blob property is just a normal modeset object id.
- *
- * Called by the user via ioctl.
- *
- * Returns:
- * Zero on success, negative errno on failure.
- */
-int drm_mode_getblob_ioctl(struct drm_device *dev,
- void *data, struct drm_file *file_priv)
-{
- struct drm_mode_get_blob *out_resp = data;
- struct drm_property_blob *blob;
- int ret = 0;
- void __user *blob_ptr;
-
- if (!drm_core_check_feature(dev, DRIVER_MODESET))
- return -EINVAL;
-
- blob = drm_property_lookup_blob(dev, out_resp->blob_id);
- if (!blob)
- return -ENOENT;
-
- if (out_resp->length == blob->length) {
- blob_ptr = (void __user *)(unsigned long)out_resp->data;
- if (copy_to_user(blob_ptr, blob->data, blob->length)) {
- ret = -EFAULT;
- goto unref;
- }
- }
- out_resp->length = blob->length;
-unref:
- drm_property_unreference_blob(blob);
-
- return ret;
-}
-
-/**
- * drm_mode_createblob_ioctl - create a new blob property
- * @dev: DRM device
- * @data: ioctl data
- * @file_priv: DRM file info
- *
- * This function creates a new blob property with user-defined values. In order
- * to give us sensible validation and checking when creating, rather than at
- * every potential use, we also require a type to be provided upfront.
- *
- * Called by the user via ioctl.
- *
- * Returns:
- * Zero on success, negative errno on failure.
- */
-int drm_mode_createblob_ioctl(struct drm_device *dev,
- void *data, struct drm_file *file_priv)
-{
- struct drm_mode_create_blob *out_resp = data;
- struct drm_property_blob *blob;
- void __user *blob_ptr;
- int ret = 0;
-
- if (!drm_core_check_feature(dev, DRIVER_MODESET))
- return -EINVAL;
-
- blob = drm_property_create_blob(dev, out_resp->length, NULL);
- if (IS_ERR(blob))
- return PTR_ERR(blob);
-
- blob_ptr = (void __user *)(unsigned long)out_resp->data;
- if (copy_from_user(blob->data, blob_ptr, out_resp->length)) {
- ret = -EFAULT;
- goto out_blob;
- }
-
- /* Dropping the lock between create_blob and our access here is safe
- * as only the same file_priv can remove the blob; at this point, it is
- * not associated with any file_priv. */
- mutex_lock(&dev->mode_config.blob_lock);
- out_resp->blob_id = blob->base.id;
- list_add_tail(&blob->head_file, &file_priv->blobs);
- mutex_unlock(&dev->mode_config.blob_lock);
-
- return 0;
-
-out_blob:
- drm_property_unreference_blob(blob);
- return ret;
-}
-
-/**
- * drm_mode_destroyblob_ioctl - destroy a user blob property
- * @dev: DRM device
- * @data: ioctl data
- * @file_priv: DRM file info
- *
- * Destroy an existing user-defined blob property.
- *
- * Called by the user via ioctl.
- *
- * Returns:
- * Zero on success, negative errno on failure.
- */
-int drm_mode_destroyblob_ioctl(struct drm_device *dev,
- void *data, struct drm_file *file_priv)
-{
- struct drm_mode_destroy_blob *out_resp = data;
- struct drm_property_blob *blob = NULL, *bt;
- bool found = false;
- int ret = 0;
-
- if (!drm_core_check_feature(dev, DRIVER_MODESET))
- return -EINVAL;
-
- blob = drm_property_lookup_blob(dev, out_resp->blob_id);
- if (!blob)
- return -ENOENT;
-
- mutex_lock(&dev->mode_config.blob_lock);
- /* Ensure the property was actually created by this user. */
- list_for_each_entry(bt, &file_priv->blobs, head_file) {
- if (bt == blob) {
- found = true;
- break;
- }
- }
-
- if (!found) {
- ret = -EPERM;
- goto err;
- }
-
- /* We must drop head_file here, because we may not be the last
- * reference on the blob. */
- list_del_init(&blob->head_file);
- mutex_unlock(&dev->mode_config.blob_lock);
-
- /* One reference from lookup, and one from the filp. */
- drm_property_unreference_blob(blob);
- drm_property_unreference_blob(blob);
-
- return 0;
-
-err:
- mutex_unlock(&dev->mode_config.blob_lock);
- drm_property_unreference_blob(blob);
-
- return ret;
-}
-
-/* Some properties could refer to dynamic refcnt'd objects, or things that
- * need special locking to handle lifetime issues (ie. to ensure the prop
- * value doesn't become invalid part way through the property update due to
- * race). The value returned by reference via 'obj' should be passed back
- * to drm_property_change_valid_put() after the property is set (and the
- * object to which the property is attached has a chance to take it's own
- * reference).
- */
-bool drm_property_change_valid_get(struct drm_property *property,
- uint64_t value, struct drm_mode_object **ref)
-{
- int i;
-
- if (property->flags & DRM_MODE_PROP_IMMUTABLE)
- return false;
-
- *ref = NULL;
-
- if (drm_property_type_is(property, DRM_MODE_PROP_RANGE)) {
- if (value < property->values[0] || value > property->values[1])
- return false;
- return true;
- } else if (drm_property_type_is(property, DRM_MODE_PROP_SIGNED_RANGE)) {
- int64_t svalue = U642I64(value);
-
- if (svalue < U642I64(property->values[0]) ||
- svalue > U642I64(property->values[1]))
- return false;
- return true;
- } else if (drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
- uint64_t valid_mask = 0;
-
- for (i = 0; i < property->num_values; i++)
- valid_mask |= (1ULL << property->values[i]);
- return !(value & ~valid_mask);
- } else if (drm_property_type_is(property, DRM_MODE_PROP_BLOB)) {
- struct drm_property_blob *blob;
-
- if (value == 0)
- return true;
-
- blob = drm_property_lookup_blob(property->dev, value);
- if (blob) {
- *ref = &blob->base;
- return true;
- } else {
- return false;
- }
- } else if (drm_property_type_is(property, DRM_MODE_PROP_OBJECT)) {
- /* a zero value for an object property translates to null: */
- if (value == 0)
- return true;
-
- *ref = __drm_mode_object_find(property->dev, value,
- property->values[0]);
- return *ref != NULL;
- }
-
- for (i = 0; i < property->num_values; i++)
- if (property->values[i] == value)
- return true;
- return false;
-}
-
-void drm_property_change_valid_put(struct drm_property *property,
- struct drm_mode_object *ref)
-{
- if (!ref)
- return;
-
- if (drm_property_type_is(property, DRM_MODE_PROP_OBJECT)) {
- drm_mode_object_unreference(ref);
- } else if (drm_property_type_is(property, DRM_MODE_PROP_BLOB))
- drm_property_unreference_blob(obj_to_blob(ref));
-}
-
-static int drm_mode_crtc_set_obj_prop(struct drm_mode_object *obj,
- struct drm_property *property,
- uint64_t value)
-{
- int ret = -EINVAL;
- struct drm_crtc *crtc = obj_to_crtc(obj);
-
- if (crtc->funcs->set_property)
- ret = crtc->funcs->set_property(crtc, property, value);
- if (!ret)
- drm_object_property_set_value(obj, property, value);
-
- return ret;
-}
-
-/**
- * drm_mode_plane_set_obj_prop - set the value of a property
- * @plane: drm plane object to set property value for
- * @property: property to set
- * @value: value the property should be set to
- *
- * This functions sets a given property on a given plane object. This function
- * calls the driver's ->set_property callback and changes the software state of
- * the property if the callback succeeds.
- *
- * Returns:
- * Zero on success, error code on failure.
- */
-int drm_mode_plane_set_obj_prop(struct drm_plane *plane,
- struct drm_property *property,
- uint64_t value)
-{
- int ret = -EINVAL;
- struct drm_mode_object *obj = &plane->base;
-
- if (plane->funcs->set_property)
- ret = plane->funcs->set_property(plane, property, value);
- if (!ret)
- drm_object_property_set_value(obj, property, value);
-
- return ret;
-}
-EXPORT_SYMBOL(drm_mode_plane_set_obj_prop);
-
-/**
- * drm_mode_obj_get_properties_ioctl - get the current value of a object's property
- * @dev: DRM device
- * @data: ioctl data
- * @file_priv: DRM file info
- *
- * This function retrieves the current value for an object's property. Compared
- * to the connector specific ioctl this one is extended to also work on crtc and
- * plane objects.
- *
- * Called by the user via ioctl.
- *
- * Returns:
- * Zero on success, negative errno on failure.
- */
-int drm_mode_obj_get_properties_ioctl(struct drm_device *dev, void *data,
- struct drm_file *file_priv)
-{
- struct drm_mode_obj_get_properties *arg = data;
- struct drm_mode_object *obj;
- int ret = 0;
-
- if (!drm_core_check_feature(dev, DRIVER_MODESET))
- return -EINVAL;
-
- drm_modeset_lock_all(dev);
-
- obj = drm_mode_object_find(dev, arg->obj_id, arg->obj_type);
- if (!obj) {
- ret = -ENOENT;
- goto out;
- }
- if (!obj->properties) {
- ret = -EINVAL;
- goto out_unref;
- }
-
- ret = drm_mode_object_get_properties(obj, file_priv->atomic,
- (uint32_t __user *)(unsigned long)(arg->props_ptr),
- (uint64_t __user *)(unsigned long)(arg->prop_values_ptr),
- &arg->count_props);
-
-out_unref:
- drm_mode_object_unreference(obj);
-out:
- drm_modeset_unlock_all(dev);
- return ret;
-}
-
-int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
- struct drm_file *file_priv)
-{
- struct drm_mode_obj_set_property *arg = data;
- struct drm_mode_object *arg_obj;
- struct drm_mode_object *prop_obj;
- struct drm_property *property;
- int i, ret = -EINVAL;
- struct drm_mode_object *ref;
-
- if (!drm_core_check_feature(dev, DRIVER_MODESET))
- return -EINVAL;
-
- drm_modeset_lock_all(dev);
-
- arg_obj = drm_mode_object_find(dev, arg->obj_id, arg->obj_type);
- if (!arg_obj) {
- ret = -ENOENT;
- goto out;
- }
- if (!arg_obj->properties)
- goto out_unref;
-
- for (i = 0; i < arg_obj->properties->count; i++)
- if (arg_obj->properties->properties[i]->base.id == arg->prop_id)
- break;
-
- if (i == arg_obj->properties->count)
- goto out_unref;
-
- prop_obj = drm_mode_object_find(dev, arg->prop_id,
- DRM_MODE_OBJECT_PROPERTY);
- if (!prop_obj) {
- ret = -ENOENT;
- goto out_unref;
- }
- property = obj_to_property(prop_obj);
-
- if (!drm_property_change_valid_get(property, arg->value, &ref))
- goto out_unref;
-
- switch (arg_obj->type) {
- case DRM_MODE_OBJECT_CONNECTOR:
- ret = drm_mode_connector_set_obj_prop(arg_obj, property,
- arg->value);
- break;
- case DRM_MODE_OBJECT_CRTC:
- ret = drm_mode_crtc_set_obj_prop(arg_obj, property, arg->value);
- break;
- case DRM_MODE_OBJECT_PLANE:
- ret = drm_mode_plane_set_obj_prop(obj_to_plane(arg_obj),
- property, arg->value);
- break;
- }
-
- drm_property_change_valid_put(property, ref);
-
-out_unref:
- drm_mode_object_unreference(arg_obj);
-out:
- drm_modeset_unlock_all(dev);
- return ret;
-}
-
-/**
- * drm_mode_crtc_set_gamma_size - set the gamma table size
- * @crtc: CRTC to set the gamma table size for
- * @gamma_size: size of the gamma table
- *
- * Drivers which support gamma tables should set this to the supported gamma
- * table size when initializing the CRTC. Currently the drm core only supports a
- * fixed gamma table size.
- *
- * Returns:
- * Zero on success, negative errno on failure.
- */
-int drm_mode_crtc_set_gamma_size(struct drm_crtc *crtc,
- int gamma_size)
-{
- uint16_t *r_base, *g_base, *b_base;
- int i;
-
- crtc->gamma_size = gamma_size;
+ crtc->gamma_size = gamma_size;
crtc->gamma_store = kcalloc(gamma_size, sizeof(uint16_t) * 3,
GFP_KERNEL);
int drm_mode_page_flip_ioctl(struct drm_device *dev,
void *data, struct drm_file *file_priv)
{
- struct drm_mode_crtc_page_flip *page_flip = data;
+ struct drm_mode_crtc_page_flip_target *page_flip = data;
struct drm_crtc *crtc;
struct drm_framebuffer *fb = NULL;
struct drm_pending_vblank_event *e = NULL;
+ u32 target_vblank = page_flip->sequence;
int ret = -EINVAL;
- if (page_flip->flags & ~DRM_MODE_PAGE_FLIP_FLAGS ||
- page_flip->reserved != 0)
+ if (page_flip->flags & ~DRM_MODE_PAGE_FLIP_FLAGS)
+ return -EINVAL;
+
+ if (page_flip->sequence != 0 && !(page_flip->flags & DRM_MODE_PAGE_FLIP_TARGET))
+ return -EINVAL;
+
+ /* Only one of the DRM_MODE_PAGE_FLIP_TARGET_ABSOLUTE/RELATIVE flags
+ * can be specified
+ */
+ if ((page_flip->flags & DRM_MODE_PAGE_FLIP_TARGET) == DRM_MODE_PAGE_FLIP_TARGET)
return -EINVAL;
if ((page_flip->flags & DRM_MODE_PAGE_FLIP_ASYNC) && !dev->mode_config.async_page_flip)
if (!crtc)
return -ENOENT;
+ if (crtc->funcs->page_flip_target) {
+ u32 current_vblank;
+ int r;
+
+ r = drm_crtc_vblank_get(crtc);
+ if (r)
+ return r;
+
+ current_vblank = drm_crtc_vblank_count(crtc);
+
+ switch (page_flip->flags & DRM_MODE_PAGE_FLIP_TARGET) {
+ case DRM_MODE_PAGE_FLIP_TARGET_ABSOLUTE:
+ if ((int)(target_vblank - current_vblank) > 1) {
+ DRM_DEBUG("Invalid absolute flip target %u, "
+ "must be <= %u\n", target_vblank,
+ current_vblank + 1);
+ drm_crtc_vblank_put(crtc);
+ return -EINVAL;
+ }
+ break;
+ case DRM_MODE_PAGE_FLIP_TARGET_RELATIVE:
+ if (target_vblank != 0 && target_vblank != 1) {
+ DRM_DEBUG("Invalid relative flip target %u, "
+ "must be 0 or 1\n", target_vblank);
+ drm_crtc_vblank_put(crtc);
+ return -EINVAL;
+ }
+ target_vblank += current_vblank;
+ break;
+ default:
+ target_vblank = current_vblank +
+ !(page_flip->flags & DRM_MODE_PAGE_FLIP_ASYNC);
+ break;
+ }
+ } else if (crtc->funcs->page_flip == NULL ||
+ (page_flip->flags & DRM_MODE_PAGE_FLIP_TARGET)) {
+ return -EINVAL;
+ }
+
drm_modeset_lock_crtc(crtc, crtc->primary);
if (crtc->primary->fb == NULL) {
/* The framebuffer is currently unbound, presumably
goto out;
}
- if (crtc->funcs->page_flip == NULL)
- goto out;
-
fb = drm_framebuffer_lookup(dev, page_flip->fb_id);
if (!fb) {
ret = -ENOENT;
}
crtc->primary->old_fb = crtc->primary->fb;
- ret = crtc->funcs->page_flip(crtc, fb, e, page_flip->flags);
+ if (crtc->funcs->page_flip_target)
+ ret = crtc->funcs->page_flip_target(crtc, fb, e,
+ page_flip->flags,
+ target_vblank);
+ else
+ ret = crtc->funcs->page_flip(crtc, fb, e, page_flip->flags);
if (ret) {
if (page_flip->flags & DRM_MODE_PAGE_FLIP_EVENT)
drm_event_cancel_free(dev, &e->base);
}
out:
+ if (ret && crtc->funcs->page_flip_target)
+ drm_crtc_vblank_put(crtc);
if (fb)
drm_framebuffer_unreference(fb);
if (crtc->primary->old_fb)
/* drm_crtc.c */
-int drm_mode_object_get_reg(struct drm_device *dev,
- struct drm_mode_object *obj,
- uint32_t obj_type,
- bool register_obj,
- void (*obj_free_cb)(struct kref *kref));
-void drm_mode_object_register(struct drm_device *dev,
- struct drm_mode_object *obj);
-int drm_mode_object_get(struct drm_device *dev,
- struct drm_mode_object *obj, uint32_t obj_type);
-struct drm_mode_object *__drm_mode_object_find(struct drm_device *dev,
- uint32_t id, uint32_t type);
-void drm_mode_object_unregister(struct drm_device *dev,
- struct drm_mode_object *object);
-int drm_mode_object_get_properties(struct drm_mode_object *obj, bool atomic,
- uint32_t __user *prop_ptr,
- uint64_t __user *prop_values,
- uint32_t *arg_count_props);
-bool drm_property_change_valid_get(struct drm_property *property,
- uint64_t value,
- struct drm_mode_object **ref);
-void drm_property_change_valid_put(struct drm_property *property,
- struct drm_mode_object *ref);
-
+int drm_mode_crtc_set_obj_prop(struct drm_mode_object *obj,
+ struct drm_property *property,
+ uint64_t value);
int drm_plane_check_pixel_format(const struct drm_plane *plane,
u32 format);
int drm_crtc_check_viewport(const struct drm_crtc *crtc,
const struct drm_framebuffer *fb);
void drm_fb_release(struct drm_file *file_priv);
-void drm_property_destroy_user_blobs(struct drm_device *dev,
- struct drm_file *file_priv);
/* dumb buffer support IOCTLs */
int drm_mode_create_dumb_ioctl(struct drm_device *dev,
void *data, struct drm_file *file_priv);
/* IOCTLs */
-int drm_mode_obj_get_properties_ioctl(struct drm_device *dev, void *data,
- struct drm_file *file_priv);
-int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
- struct drm_file *file_priv);
-
int drm_mode_getresources(struct drm_device *dev,
void *data, struct drm_file *file_priv);
int drm_mode_getplane_res(struct drm_device *dev, void *data,
void *data, struct drm_file *file_priv);
int drm_mode_cursor2_ioctl(struct drm_device *dev,
void *data, struct drm_file *file_priv);
+int drm_mode_gamma_get_ioctl(struct drm_device *dev,
+ void *data, struct drm_file *file_priv);
+int drm_mode_gamma_set_ioctl(struct drm_device *dev,
+ void *data, struct drm_file *file_priv);
+
+int drm_mode_page_flip_ioctl(struct drm_device *dev,
+ void *data, struct drm_file *file_priv);
+
+/* drm_property.c */
+void drm_property_destroy_user_blobs(struct drm_device *dev,
+ struct drm_file *file_priv);
+bool drm_property_change_valid_get(struct drm_property *property,
+ uint64_t value,
+ struct drm_mode_object **ref);
+void drm_property_change_valid_put(struct drm_property *property,
+ struct drm_mode_object *ref);
+
+/* IOCTL */
int drm_mode_getproperty_ioctl(struct drm_device *dev,
void *data, struct drm_file *file_priv);
int drm_mode_getblob_ioctl(struct drm_device *dev,
void *data, struct drm_file *file_priv);
int drm_mode_destroyblob_ioctl(struct drm_device *dev,
void *data, struct drm_file *file_priv);
+
+/* drm_mode_object.c */
+int drm_mode_object_get_reg(struct drm_device *dev,
+ struct drm_mode_object *obj,
+ uint32_t obj_type,
+ bool register_obj,
+ void (*obj_free_cb)(struct kref *kref));
+void drm_mode_object_register(struct drm_device *dev,
+ struct drm_mode_object *obj);
+int drm_mode_object_get(struct drm_device *dev,
+ struct drm_mode_object *obj, uint32_t obj_type);
+struct drm_mode_object *__drm_mode_object_find(struct drm_device *dev,
+ uint32_t id, uint32_t type);
+void drm_mode_object_unregister(struct drm_device *dev,
+ struct drm_mode_object *object);
+int drm_mode_object_get_properties(struct drm_mode_object *obj, bool atomic,
+ uint32_t __user *prop_ptr,
+ uint64_t __user *prop_values,
+ uint32_t *arg_count_props);
+struct drm_property *drm_mode_obj_find_prop_id(struct drm_mode_object *obj,
+ uint32_t prop_id);
+
+/* IOCTL */
+
+int drm_mode_obj_get_properties_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv);
+
+/* drm_encoder.c */
+int drm_encoder_register_all(struct drm_device *dev);
+void drm_encoder_unregister_all(struct drm_device *dev);
+
+/* IOCTL */
int drm_mode_getencoder(struct drm_device *dev,
void *data, struct drm_file *file_priv);
-int drm_mode_gamma_get_ioctl(struct drm_device *dev,
- void *data, struct drm_file *file_priv);
-int drm_mode_gamma_set_ioctl(struct drm_device *dev,
- void *data, struct drm_file *file_priv);
-
-int drm_mode_page_flip_ioctl(struct drm_device *dev,
- void *data, struct drm_file *file_priv);
/* drm_connector.c */
void drm_connector_ida_init(void);
}
EXPORT_SYMBOL(drm_dp_link_configure);
+/**
+ * drm_dp_downstream_max_clock() - extract branch device max
+ * pixel rate for legacy VGA
+ * converter or max TMDS clock
+ * rate for others
+ * @dpcd: DisplayPort configuration data
+ * @port_cap: port capabilities
+ *
+ * Returns max clock in kHz on success or 0 if max clock not defined
+ */
+int drm_dp_downstream_max_clock(const u8 dpcd[DP_RECEIVER_CAP_SIZE],
+ const u8 port_cap[4])
+{
+ int type = port_cap[0] & DP_DS_PORT_TYPE_MASK;
+ bool detailed_cap_info = dpcd[DP_DOWNSTREAMPORT_PRESENT] &
+ DP_DETAILED_CAP_INFO_AVAILABLE;
+
+ if (!detailed_cap_info)
+ return 0;
+
+ switch (type) {
+ case DP_DS_PORT_TYPE_VGA:
+ return port_cap[1] * 8 * 1000;
+ case DP_DS_PORT_TYPE_DVI:
+ case DP_DS_PORT_TYPE_HDMI:
+ case DP_DS_PORT_TYPE_DP_DUALMODE:
+ return port_cap[1] * 2500;
+ default:
+ return 0;
+ }
+}
+EXPORT_SYMBOL(drm_dp_downstream_max_clock);
+
+/**
+ * drm_dp_downstream_max_bpc() - extract branch device max
+ * bits per component
+ * @dpcd: DisplayPort configuration data
+ * @port_cap: port capabilities
+ *
+ * Returns max bpc on success or 0 if max bpc not defined
+ */
+int drm_dp_downstream_max_bpc(const u8 dpcd[DP_RECEIVER_CAP_SIZE],
+ const u8 port_cap[4])
+{
+ int type = port_cap[0] & DP_DS_PORT_TYPE_MASK;
+ bool detailed_cap_info = dpcd[DP_DOWNSTREAMPORT_PRESENT] &
+ DP_DETAILED_CAP_INFO_AVAILABLE;
+ int bpc;
+
+ if (!detailed_cap_info)
+ return 0;
+
+ switch (type) {
+ case DP_DS_PORT_TYPE_VGA:
+ case DP_DS_PORT_TYPE_DVI:
+ case DP_DS_PORT_TYPE_HDMI:
+ case DP_DS_PORT_TYPE_DP_DUALMODE:
+ bpc = port_cap[2] & DP_DS_MAX_BPC_MASK;
+
+ switch (bpc) {
+ case DP_DS_8BPC:
+ return 8;
+ case DP_DS_10BPC:
+ return 10;
+ case DP_DS_12BPC:
+ return 12;
+ case DP_DS_16BPC:
+ return 16;
+ }
+ default:
+ return 0;
+ }
+}
+EXPORT_SYMBOL(drm_dp_downstream_max_bpc);
+
+/**
+ * drm_dp_downstream_id() - identify branch device
+ * @aux: DisplayPort AUX channel
+ * @id: DisplayPort branch device id
+ *
+ * Returns branch device id on success or NULL on failure
+ */
+int drm_dp_downstream_id(struct drm_dp_aux *aux, char id[6])
+{
+ return drm_dp_dpcd_read(aux, DP_BRANCH_ID, id, 6);
+}
+EXPORT_SYMBOL(drm_dp_downstream_id);
+
+/**
+ * drm_dp_downstream_debug() - debug DP branch devices
+ * @m: pointer for debugfs file
+ * @dpcd: DisplayPort configuration data
+ * @port_cap: port capabilities
+ * @aux: DisplayPort AUX channel
+ *
+ */
+void drm_dp_downstream_debug(struct seq_file *m,
+ const u8 dpcd[DP_RECEIVER_CAP_SIZE],
+ const u8 port_cap[4], struct drm_dp_aux *aux)
+{
+ bool detailed_cap_info = dpcd[DP_DOWNSTREAMPORT_PRESENT] &
+ DP_DETAILED_CAP_INFO_AVAILABLE;
+ int clk;
+ int bpc;
+ char id[6];
+ int len;
+ uint8_t rev[2];
+ int type = port_cap[0] & DP_DS_PORT_TYPE_MASK;
+ bool branch_device = dpcd[DP_DOWNSTREAMPORT_PRESENT] &
+ DP_DWN_STRM_PORT_PRESENT;
+
+ seq_printf(m, "\tDP branch device present: %s\n",
+ branch_device ? "yes" : "no");
+
+ if (!branch_device)
+ return;
+
+ switch (type) {
+ case DP_DS_PORT_TYPE_DP:
+ seq_puts(m, "\t\tType: DisplayPort\n");
+ break;
+ case DP_DS_PORT_TYPE_VGA:
+ seq_puts(m, "\t\tType: VGA\n");
+ break;
+ case DP_DS_PORT_TYPE_DVI:
+ seq_puts(m, "\t\tType: DVI\n");
+ break;
+ case DP_DS_PORT_TYPE_HDMI:
+ seq_puts(m, "\t\tType: HDMI\n");
+ break;
+ case DP_DS_PORT_TYPE_NON_EDID:
+ seq_puts(m, "\t\tType: others without EDID support\n");
+ break;
+ case DP_DS_PORT_TYPE_DP_DUALMODE:
+ seq_puts(m, "\t\tType: DP++\n");
+ break;
+ case DP_DS_PORT_TYPE_WIRELESS:
+ seq_puts(m, "\t\tType: Wireless\n");
+ break;
+ default:
+ seq_puts(m, "\t\tType: N/A\n");
+ }
+
+ drm_dp_downstream_id(aux, id);
+ seq_printf(m, "\t\tID: %s\n", id);
+
+ len = drm_dp_dpcd_read(aux, DP_BRANCH_HW_REV, &rev[0], 1);
+ if (len > 0)
+ seq_printf(m, "\t\tHW: %d.%d\n",
+ (rev[0] & 0xf0) >> 4, rev[0] & 0xf);
+
+ len = drm_dp_dpcd_read(aux, DP_BRANCH_SW_REV, &rev, 2);
+ if (len > 0)
+ seq_printf(m, "\t\tSW: %d.%d\n", rev[0], rev[1]);
+
+ if (detailed_cap_info) {
+ clk = drm_dp_downstream_max_clock(dpcd, port_cap);
+
+ if (clk > 0) {
+ if (type == DP_DS_PORT_TYPE_VGA)
+ seq_printf(m, "\t\tMax dot clock: %d kHz\n", clk);
+ else
+ seq_printf(m, "\t\tMax TMDS clock: %d kHz\n", clk);
+ }
+
+ bpc = drm_dp_downstream_max_bpc(dpcd, port_cap);
+
+ if (bpc > 0)
+ seq_printf(m, "\t\tMax bpc: %d\n", bpc);
+ }
+}
+EXPORT_SYMBOL(drm_dp_downstream_debug);
+
/*
* I2C-over-AUX implementation
*/
--- /dev/null
+/*
+ * Copyright (c) 2016 Intel Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission. The copyright holders make no representations
+ * about the suitability of this software for any purpose. It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#include <linux/export.h>
+#include <drm/drmP.h>
+#include <drm/drm_encoder.h>
+
+#include "drm_crtc_internal.h"
+
+/**
+ * DOC: overview
+ *
+ * Encoders represent the connecting element between the CRTC (as the overall
+ * pixel pipeline, represented by struct &drm_crtc) and the connectors (as the
+ * generic sink entity, represented by struct &drm_connector). Encoders are
+ * objects exposed to userspace, originally to allow userspace to infer cloning
+ * and connector/CRTC restrictions. Unfortunately almost all drivers get this
+ * wrong, making the uabi pretty much useless. On top of that the exposed
+ * restrictions are too simple for todays hardware, and the recommend way to
+ * infer restrictions is by using the DRM_MODE_ATOMIC_TEST_ONLY flag for the
+ * atomic IOCTL.
+ *
+ * Otherwise encoders aren't used in the uapi at all (any modeset request from
+ * userspace directly connects a connector with a CRTC), drivers are therefore
+ * free to use them however they wish. Modeset helper libraries make strong use
+ * of encoders to facilitate code sharing. But for more complex settings it is
+ * usually better to move shared code into a separate &drm_bridge. Compared to
+ * encoders bridges also have the benefit of not being purely an internal
+ * abstraction since they are not exposed to userspace at all.
+ *
+ * Encoders are initialized with drm_encoder_init() and cleaned up using
+ * drm_encoder_cleanup().
+ */
+static const struct drm_prop_enum_list drm_encoder_enum_list[] = {
+ { DRM_MODE_ENCODER_NONE, "None" },
+ { DRM_MODE_ENCODER_DAC, "DAC" },
+ { DRM_MODE_ENCODER_TMDS, "TMDS" },
+ { DRM_MODE_ENCODER_LVDS, "LVDS" },
+ { DRM_MODE_ENCODER_TVDAC, "TV" },
+ { DRM_MODE_ENCODER_VIRTUAL, "Virtual" },
+ { DRM_MODE_ENCODER_DSI, "DSI" },
+ { DRM_MODE_ENCODER_DPMST, "DP MST" },
+ { DRM_MODE_ENCODER_DPI, "DPI" },
+};
+
+int drm_encoder_register_all(struct drm_device *dev)
+{
+ struct drm_encoder *encoder;
+ int ret = 0;
+
+ drm_for_each_encoder(encoder, dev) {
+ if (encoder->funcs->late_register)
+ ret = encoder->funcs->late_register(encoder);
+ if (ret)
+ return ret;
+ }
+
+ return 0;
+}
+
+void drm_encoder_unregister_all(struct drm_device *dev)
+{
+ struct drm_encoder *encoder;
+
+ drm_for_each_encoder(encoder, dev) {
+ if (encoder->funcs->early_unregister)
+ encoder->funcs->early_unregister(encoder);
+ }
+}
+
+/**
+ * drm_encoder_init - Init a preallocated encoder
+ * @dev: drm device
+ * @encoder: the encoder to init
+ * @funcs: callbacks for this encoder
+ * @encoder_type: user visible type of the encoder
+ * @name: printf style format string for the encoder name, or NULL for default name
+ *
+ * Initialises a preallocated encoder. Encoder should be subclassed as part of
+ * driver encoder objects. At driver unload time drm_encoder_cleanup() should be
+ * called from the driver's destroy hook in &drm_encoder_funcs.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_encoder_init(struct drm_device *dev,
+ struct drm_encoder *encoder,
+ const struct drm_encoder_funcs *funcs,
+ int encoder_type, const char *name, ...)
+{
+ int ret;
+
+ drm_modeset_lock_all(dev);
+
+ ret = drm_mode_object_get(dev, &encoder->base, DRM_MODE_OBJECT_ENCODER);
+ if (ret)
+ goto out_unlock;
+
+ encoder->dev = dev;
+ encoder->encoder_type = encoder_type;
+ encoder->funcs = funcs;
+ if (name) {
+ va_list ap;
+
+ va_start(ap, name);
+ encoder->name = kvasprintf(GFP_KERNEL, name, ap);
+ va_end(ap);
+ } else {
+ encoder->name = kasprintf(GFP_KERNEL, "%s-%d",
+ drm_encoder_enum_list[encoder_type].name,
+ encoder->base.id);
+ }
+ if (!encoder->name) {
+ ret = -ENOMEM;
+ goto out_put;
+ }
+
+ list_add_tail(&encoder->head, &dev->mode_config.encoder_list);
+ encoder->index = dev->mode_config.num_encoder++;
+
+out_put:
+ if (ret)
+ drm_mode_object_unregister(dev, &encoder->base);
+
+out_unlock:
+ drm_modeset_unlock_all(dev);
+
+ return ret;
+}
+EXPORT_SYMBOL(drm_encoder_init);
+
+/**
+ * drm_encoder_cleanup - cleans up an initialised encoder
+ * @encoder: encoder to cleanup
+ *
+ * Cleans up the encoder but doesn't free the object.
+ */
+void drm_encoder_cleanup(struct drm_encoder *encoder)
+{
+ struct drm_device *dev = encoder->dev;
+
+ /* Note that the encoder_list is considered to be static; should we
+ * remove the drm_encoder at runtime we would have to decrement all
+ * the indices on the drm_encoder after us in the encoder_list.
+ */
+
+ drm_modeset_lock_all(dev);
+ drm_mode_object_unregister(dev, &encoder->base);
+ kfree(encoder->name);
+ list_del(&encoder->head);
+ dev->mode_config.num_encoder--;
+ drm_modeset_unlock_all(dev);
+
+ memset(encoder, 0, sizeof(*encoder));
+}
+EXPORT_SYMBOL(drm_encoder_cleanup);
+
+static struct drm_crtc *drm_encoder_get_crtc(struct drm_encoder *encoder)
+{
+ struct drm_connector *connector;
+ struct drm_device *dev = encoder->dev;
+ bool uses_atomic = false;
+
+ /* For atomic drivers only state objects are synchronously updated and
+ * protected by modeset locks, so check those first. */
+ drm_for_each_connector(connector, dev) {
+ if (!connector->state)
+ continue;
+
+ uses_atomic = true;
+
+ if (connector->state->best_encoder != encoder)
+ continue;
+
+ return connector->state->crtc;
+ }
+
+ /* Don't return stale data (e.g. pending async disable). */
+ if (uses_atomic)
+ return NULL;
+
+ return encoder->crtc;
+}
+
+int drm_mode_getencoder(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_mode_get_encoder *enc_resp = data;
+ struct drm_encoder *encoder;
+ struct drm_crtc *crtc;
+
+ if (!drm_core_check_feature(dev, DRIVER_MODESET))
+ return -EINVAL;
+
+ encoder = drm_encoder_find(dev, enc_resp->encoder_id);
+ if (!encoder)
+ return -ENOENT;
+
+ drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
+ crtc = drm_encoder_get_crtc(encoder);
+ if (crtc)
+ enc_resp->crtc_id = crtc->base.id;
+ else
+ enc_resp->crtc_id = 0;
+ drm_modeset_unlock(&dev->mode_config.connection_mutex);
+
+ enc_resp->encoder_type = encoder->encoder_type;
+ enc_resp->encoder_id = encoder->base.id;
+ enc_resp->possible_crtcs = encoder->possible_crtcs;
+ enc_resp->possible_clones = encoder->possible_clones;
+
+ return 0;
+}
return isascii(c) && isprint(c) ? c : '?';
}
+/**
+ * drm_mode_legacy_fb_format - compute drm fourcc code from legacy description
+ * @bpp: bits per pixels
+ * @depth: bit depth per pixel
+ *
+ * Computes a drm fourcc pixel format code for the given @bpp/@depth values.
+ * Useful in fbdev emulation code, since that deals in those values.
+ */
+uint32_t drm_mode_legacy_fb_format(uint32_t bpp, uint32_t depth)
+{
+ uint32_t fmt;
+
+ switch (bpp) {
+ case 8:
+ fmt = DRM_FORMAT_C8;
+ break;
+ case 16:
+ if (depth == 15)
+ fmt = DRM_FORMAT_XRGB1555;
+ else
+ fmt = DRM_FORMAT_RGB565;
+ break;
+ case 24:
+ fmt = DRM_FORMAT_RGB888;
+ break;
+ case 32:
+ if (depth == 24)
+ fmt = DRM_FORMAT_XRGB8888;
+ else if (depth == 30)
+ fmt = DRM_FORMAT_XRGB2101010;
+ else
+ fmt = DRM_FORMAT_ARGB8888;
+ break;
+ default:
+ DRM_ERROR("bad bpp, assuming x8r8g8b8 pixel format\n");
+ fmt = DRM_FORMAT_XRGB8888;
+ break;
+ }
+
+ return fmt;
+}
+EXPORT_SYMBOL(drm_mode_legacy_fb_format);
+
/**
* drm_get_format_name - return a string for drm fourcc format
* @format: format to compute name of
int drm_global_item_ref(struct drm_global_reference *ref)
{
- int ret;
+ int ret = 0;
struct drm_global_item *item = &glob[ref->global_type];
mutex_lock(&item->mutex);
if (item->refcount == 0) {
- item->object = kzalloc(ref->size, GFP_KERNEL);
- if (unlikely(item->object == NULL)) {
+ ref->object = kzalloc(ref->size, GFP_KERNEL);
+ if (unlikely(ref->object == NULL)) {
ret = -ENOMEM;
- goto out_err;
+ goto error_unlock;
}
-
- ref->object = item->object;
ret = ref->init(ref);
if (unlikely(ret != 0))
- goto out_err;
+ goto error_free;
+ item->object = ref->object;
+ } else {
+ ref->object = item->object;
}
+
++item->refcount;
- ref->object = item->object;
mutex_unlock(&item->mutex);
return 0;
-out_err:
+
+error_free:
+ kfree(ref->object);
+ ref->object = NULL;
+error_unlock:
mutex_unlock(&item->mutex);
- item->object = NULL;
return ret;
}
EXPORT_SYMBOL(drm_global_item_ref);
unsigned long add)
{
int ret;
- unsigned long mask = (1 << bits) - 1;
+ unsigned long mask = (1UL << bits) - 1;
unsigned long first, unshifted_key;
unshifted_key = hash_long(seed, bits);
return 0;
}
+#if defined(CONFIG_X86) || defined(CONFIG_IA64)
typedef struct drm_mode_fb_cmd232 {
u32 fb_id;
u32 width;
return 0;
}
+#endif
static drm_ioctl_compat_t *drm_compat_ioctls[] = {
[DRM_IOCTL_NR(DRM_IOCTL_VERSION32)] = compat_drm_version,
[DRM_IOCTL_NR(DRM_IOCTL_UPDATE_DRAW32)] = compat_drm_update_draw,
#endif
[DRM_IOCTL_NR(DRM_IOCTL_WAIT_VBLANK32)] = compat_drm_wait_vblank,
+#if defined(CONFIG_X86) || defined(CONFIG_IA64)
[DRM_IOCTL_NR(DRM_IOCTL_MODE_ADDFB232)] = compat_drm_mode_addfb2,
+#endif
};
/**
static int drm_getcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
{
struct drm_get_cap *req = data;
+ struct drm_crtc *crtc;
req->value = 0;
switch (req->capability) {
case DRM_CAP_ASYNC_PAGE_FLIP:
req->value = dev->mode_config.async_page_flip;
break;
+ case DRM_CAP_PAGE_FLIP_TARGET:
+ req->value = 1;
+ drm_for_each_crtc(crtc, dev) {
+ if (!crtc->funcs->page_flip_target)
+ req->value = 0;
+ }
+ break;
case DRM_CAP_CURSOR_WIDTH:
if (dev->mode_config.cursor_width)
req->value = dev->mode_config.cursor_width;
--- /dev/null
+/*
+ * Copyright (c) 2016 Intel Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission. The copyright holders make no representations
+ * about the suitability of this software for any purpose. It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#include <linux/export.h>
+#include <drm/drmP.h>
+#include <drm/drm_mode_object.h>
+
+#include "drm_crtc_internal.h"
+
+/*
+ * Internal function to assign a slot in the object idr and optionally
+ * register the object into the idr.
+ */
+int drm_mode_object_get_reg(struct drm_device *dev,
+ struct drm_mode_object *obj,
+ uint32_t obj_type,
+ bool register_obj,
+ void (*obj_free_cb)(struct kref *kref))
+{
+ int ret;
+
+ mutex_lock(&dev->mode_config.idr_mutex);
+ ret = idr_alloc(&dev->mode_config.crtc_idr, register_obj ? obj : NULL, 1, 0, GFP_KERNEL);
+ if (ret >= 0) {
+ /*
+ * Set up the object linking under the protection of the idr
+ * lock so that other users can't see inconsistent state.
+ */
+ obj->id = ret;
+ obj->type = obj_type;
+ if (obj_free_cb) {
+ obj->free_cb = obj_free_cb;
+ kref_init(&obj->refcount);
+ }
+ }
+ mutex_unlock(&dev->mode_config.idr_mutex);
+
+ return ret < 0 ? ret : 0;
+}
+
+/**
+ * drm_mode_object_get - allocate a new modeset identifier
+ * @dev: DRM device
+ * @obj: object pointer, used to generate unique ID
+ * @obj_type: object type
+ *
+ * Create a unique identifier based on @ptr in @dev's identifier space. Used
+ * for tracking modes, CRTCs and connectors. Note that despite the _get postfix
+ * modeset identifiers are _not_ reference counted. Hence don't use this for
+ * reference counted modeset objects like framebuffers.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_mode_object_get(struct drm_device *dev,
+ struct drm_mode_object *obj, uint32_t obj_type)
+{
+ return drm_mode_object_get_reg(dev, obj, obj_type, true, NULL);
+}
+
+void drm_mode_object_register(struct drm_device *dev,
+ struct drm_mode_object *obj)
+{
+ mutex_lock(&dev->mode_config.idr_mutex);
+ idr_replace(&dev->mode_config.crtc_idr, obj, obj->id);
+ mutex_unlock(&dev->mode_config.idr_mutex);
+}
+
+/**
+ * drm_mode_object_unregister - free a modeset identifer
+ * @dev: DRM device
+ * @object: object to free
+ *
+ * Free @id from @dev's unique identifier pool.
+ * This function can be called multiple times, and guards against
+ * multiple removals.
+ * These modeset identifiers are _not_ reference counted. Hence don't use this
+ * for reference counted modeset objects like framebuffers.
+ */
+void drm_mode_object_unregister(struct drm_device *dev,
+ struct drm_mode_object *object)
+{
+ mutex_lock(&dev->mode_config.idr_mutex);
+ if (object->id) {
+ idr_remove(&dev->mode_config.crtc_idr, object->id);
+ object->id = 0;
+ }
+ mutex_unlock(&dev->mode_config.idr_mutex);
+}
+
+struct drm_mode_object *__drm_mode_object_find(struct drm_device *dev,
+ uint32_t id, uint32_t type)
+{
+ struct drm_mode_object *obj = NULL;
+
+ mutex_lock(&dev->mode_config.idr_mutex);
+ obj = idr_find(&dev->mode_config.crtc_idr, id);
+ if (obj && type != DRM_MODE_OBJECT_ANY && obj->type != type)
+ obj = NULL;
+ if (obj && obj->id != id)
+ obj = NULL;
+
+ if (obj && obj->free_cb) {
+ if (!kref_get_unless_zero(&obj->refcount))
+ obj = NULL;
+ }
+ mutex_unlock(&dev->mode_config.idr_mutex);
+
+ return obj;
+}
+
+/**
+ * drm_mode_object_find - look up a drm object with static lifetime
+ * @dev: drm device
+ * @id: id of the mode object
+ * @type: type of the mode object
+ *
+ * This function is used to look up a modeset object. It will acquire a
+ * reference for reference counted objects. This reference must be dropped again
+ * by callind drm_mode_object_unreference().
+ */
+struct drm_mode_object *drm_mode_object_find(struct drm_device *dev,
+ uint32_t id, uint32_t type)
+{
+ struct drm_mode_object *obj = NULL;
+
+ obj = __drm_mode_object_find(dev, id, type);
+ return obj;
+}
+EXPORT_SYMBOL(drm_mode_object_find);
+
+/**
+ * drm_mode_object_unreference - decr the object refcnt
+ * @obj: mode_object
+ *
+ * This function decrements the object's refcount if it is a refcounted modeset
+ * object. It is a no-op on any other object. This is used to drop references
+ * acquired with drm_mode_object_reference().
+ */
+void drm_mode_object_unreference(struct drm_mode_object *obj)
+{
+ if (obj->free_cb) {
+ DRM_DEBUG("OBJ ID: %d (%d)\n", obj->id, atomic_read(&obj->refcount.refcount));
+ kref_put(&obj->refcount, obj->free_cb);
+ }
+}
+EXPORT_SYMBOL(drm_mode_object_unreference);
+
+/**
+ * drm_mode_object_reference - incr the object refcnt
+ * @obj: mode_object
+ *
+ * This function increments the object's refcount if it is a refcounted modeset
+ * object. It is a no-op on any other object. References should be dropped again
+ * by calling drm_mode_object_unreference().
+ */
+void drm_mode_object_reference(struct drm_mode_object *obj)
+{
+ if (obj->free_cb) {
+ DRM_DEBUG("OBJ ID: %d (%d)\n", obj->id, atomic_read(&obj->refcount.refcount));
+ kref_get(&obj->refcount);
+ }
+}
+EXPORT_SYMBOL(drm_mode_object_reference);
+
+/**
+ * drm_object_attach_property - attach a property to a modeset object
+ * @obj: drm modeset object
+ * @property: property to attach
+ * @init_val: initial value of the property
+ *
+ * This attaches the given property to the modeset object with the given initial
+ * value. Currently this function cannot fail since the properties are stored in
+ * a statically sized array.
+ */
+void drm_object_attach_property(struct drm_mode_object *obj,
+ struct drm_property *property,
+ uint64_t init_val)
+{
+ int count = obj->properties->count;
+
+ if (count == DRM_OBJECT_MAX_PROPERTY) {
+ WARN(1, "Failed to attach object property (type: 0x%x). Please "
+ "increase DRM_OBJECT_MAX_PROPERTY by 1 for each time "
+ "you see this message on the same object type.\n",
+ obj->type);
+ return;
+ }
+
+ obj->properties->properties[count] = property;
+ obj->properties->values[count] = init_val;
+ obj->properties->count++;
+}
+EXPORT_SYMBOL(drm_object_attach_property);
+
+/**
+ * drm_object_property_set_value - set the value of a property
+ * @obj: drm mode object to set property value for
+ * @property: property to set
+ * @val: value the property should be set to
+ *
+ * This function sets a given property on a given object. This function only
+ * changes the software state of the property, it does not call into the
+ * driver's ->set_property callback.
+ *
+ * Note that atomic drivers should not have any need to call this, the core will
+ * ensure consistency of values reported back to userspace through the
+ * appropriate ->atomic_get_property callback. Only legacy drivers should call
+ * this function to update the tracked value (after clamping and other
+ * restrictions have been applied).
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_object_property_set_value(struct drm_mode_object *obj,
+ struct drm_property *property, uint64_t val)
+{
+ int i;
+
+ for (i = 0; i < obj->properties->count; i++) {
+ if (obj->properties->properties[i] == property) {
+ obj->properties->values[i] = val;
+ return 0;
+ }
+ }
+
+ return -EINVAL;
+}
+EXPORT_SYMBOL(drm_object_property_set_value);
+
+/**
+ * drm_object_property_get_value - retrieve the value of a property
+ * @obj: drm mode object to get property value from
+ * @property: property to retrieve
+ * @val: storage for the property value
+ *
+ * This function retrieves the softare state of the given property for the given
+ * property. Since there is no driver callback to retrieve the current property
+ * value this might be out of sync with the hardware, depending upon the driver
+ * and property.
+ *
+ * Atomic drivers should never call this function directly, the core will read
+ * out property values through the various ->atomic_get_property callbacks.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_object_property_get_value(struct drm_mode_object *obj,
+ struct drm_property *property, uint64_t *val)
+{
+ int i;
+
+ /* read-only properties bypass atomic mechanism and still store
+ * their value in obj->properties->values[].. mostly to avoid
+ * having to deal w/ EDID and similar props in atomic paths:
+ */
+ if (drm_core_check_feature(property->dev, DRIVER_ATOMIC) &&
+ !(property->flags & DRM_MODE_PROP_IMMUTABLE))
+ return drm_atomic_get_property(obj, property, val);
+
+ for (i = 0; i < obj->properties->count; i++) {
+ if (obj->properties->properties[i] == property) {
+ *val = obj->properties->values[i];
+ return 0;
+ }
+
+ }
+
+ return -EINVAL;
+}
+EXPORT_SYMBOL(drm_object_property_get_value);
+
+/* helper for getconnector and getproperties ioctls */
+int drm_mode_object_get_properties(struct drm_mode_object *obj, bool atomic,
+ uint32_t __user *prop_ptr,
+ uint64_t __user *prop_values,
+ uint32_t *arg_count_props)
+{
+ int i, ret, count;
+
+ for (i = 0, count = 0; i < obj->properties->count; i++) {
+ struct drm_property *prop = obj->properties->properties[i];
+ uint64_t val;
+
+ if ((prop->flags & DRM_MODE_PROP_ATOMIC) && !atomic)
+ continue;
+
+ if (*arg_count_props > count) {
+ ret = drm_object_property_get_value(obj, prop, &val);
+ if (ret)
+ return ret;
+
+ if (put_user(prop->base.id, prop_ptr + count))
+ return -EFAULT;
+
+ if (put_user(val, prop_values + count))
+ return -EFAULT;
+ }
+
+ count++;
+ }
+ *arg_count_props = count;
+
+ return 0;
+}
+
+/**
+ * drm_mode_obj_get_properties_ioctl - get the current value of a object's property
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * This function retrieves the current value for an object's property. Compared
+ * to the connector specific ioctl this one is extended to also work on crtc and
+ * plane objects.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int drm_mode_obj_get_properties_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_mode_obj_get_properties *arg = data;
+ struct drm_mode_object *obj;
+ int ret = 0;
+
+ if (!drm_core_check_feature(dev, DRIVER_MODESET))
+ return -EINVAL;
+
+ drm_modeset_lock_all(dev);
+
+ obj = drm_mode_object_find(dev, arg->obj_id, arg->obj_type);
+ if (!obj) {
+ ret = -ENOENT;
+ goto out;
+ }
+ if (!obj->properties) {
+ ret = -EINVAL;
+ goto out_unref;
+ }
+
+ ret = drm_mode_object_get_properties(obj, file_priv->atomic,
+ (uint32_t __user *)(unsigned long)(arg->props_ptr),
+ (uint64_t __user *)(unsigned long)(arg->prop_values_ptr),
+ &arg->count_props);
+
+out_unref:
+ drm_mode_object_unreference(obj);
+out:
+ drm_modeset_unlock_all(dev);
+ return ret;
+}
+
+struct drm_property *drm_mode_obj_find_prop_id(struct drm_mode_object *obj,
+ uint32_t prop_id)
+{
+ int i;
+
+ for (i = 0; i < obj->properties->count; i++)
+ if (obj->properties->properties[i]->base.id == prop_id)
+ return obj->properties->properties[i];
+
+ return NULL;
+}
+
+int drm_mode_obj_set_property_ioctl(struct drm_device *dev, void *data,
+ struct drm_file *file_priv)
+{
+ struct drm_mode_obj_set_property *arg = data;
+ struct drm_mode_object *arg_obj;
+ struct drm_property *property;
+ int ret = -EINVAL;
+ struct drm_mode_object *ref;
+
+ if (!drm_core_check_feature(dev, DRIVER_MODESET))
+ return -EINVAL;
+
+ drm_modeset_lock_all(dev);
+
+ arg_obj = drm_mode_object_find(dev, arg->obj_id, arg->obj_type);
+ if (!arg_obj) {
+ ret = -ENOENT;
+ goto out;
+ }
+
+ if (!arg_obj->properties)
+ goto out_unref;
+
+ property = drm_mode_obj_find_prop_id(arg_obj, arg->prop_id);
+ if (!property)
+ goto out_unref;
+
+ if (!drm_property_change_valid_get(property, arg->value, &ref))
+ goto out_unref;
+
+ switch (arg_obj->type) {
+ case DRM_MODE_OBJECT_CONNECTOR:
+ ret = drm_mode_connector_set_obj_prop(arg_obj, property,
+ arg->value);
+ break;
+ case DRM_MODE_OBJECT_CRTC:
+ ret = drm_mode_crtc_set_obj_prop(arg_obj, property, arg->value);
+ break;
+ case DRM_MODE_OBJECT_PLANE:
+ ret = drm_mode_plane_set_obj_prop(obj_to_plane(arg_obj),
+ property, arg->value);
+ break;
+ }
+
+ drm_property_change_valid_put(property, ref);
+
+out_unref:
+ drm_mode_object_unreference(arg_obj);
+out:
+ drm_modeset_unlock_all(dev);
+ return ret;
+}
{
bool poll = false;
struct drm_connector *connector;
+ unsigned long delay = DRM_OUTPUT_POLL_PERIOD;
WARN_ON(!mutex_is_locked(&dev->mode_config.mutex));
poll = true;
}
+ if (dev->mode_config.delayed_event) {
+ poll = true;
+ delay = 0;
+ }
+
if (poll)
- schedule_delayed_work(&dev->mode_config.output_poll_work, DRM_OUTPUT_POLL_PERIOD);
+ schedule_delayed_work(&dev->mode_config.output_poll_work, delay);
}
EXPORT_SYMBOL(drm_kms_helper_poll_enable_locked);
--- /dev/null
+/*
+ * Copyright (c) 2016 Intel Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission. The copyright holders make no representations
+ * about the suitability of this software for any purpose. It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#include <linux/export.h>
+#include <drm/drmP.h>
+#include <drm/drm_property.h>
+
+#include "drm_crtc_internal.h"
+
+/**
+ * DOC: overview
+ *
+ * Properties as represented by &drm_property are used to extend the modeset
+ * interface exposed to userspace. For the atomic modeset IOCTL properties are
+ * even the only way to transport metadata about the desired new modeset
+ * configuration from userspace to the kernel. Properties have a well-defined
+ * value range, which is enforced by the drm core. See the documentation of the
+ * flags member of struct &drm_property for an overview of the different
+ * property types and ranges.
+ *
+ * Properties don't store the current value directly, but need to be
+ * instatiated by attaching them to a &drm_mode_object with
+ * drm_object_attach_property().
+ *
+ * Property values are only 64bit. To support bigger piles of data (like gamma
+ * tables, color correction matrizes or large structures) a property can instead
+ * point at a &drm_property_blob with that additional data
+ *
+ * Properties are defined by their symbolic name, userspace must keep a
+ * per-object mapping from those names to the property ID used in the atomic
+ * IOCTL and in the get/set property IOCTL.
+ */
+
+static bool drm_property_type_valid(struct drm_property *property)
+{
+ if (property->flags & DRM_MODE_PROP_EXTENDED_TYPE)
+ return !(property->flags & DRM_MODE_PROP_LEGACY_TYPE);
+ return !!(property->flags & DRM_MODE_PROP_LEGACY_TYPE);
+}
+
+/**
+ * drm_property_create - create a new property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @num_values: number of pre-defined values
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy(), which is done automatically when calling
+ * drm_mode_config_cleanup().
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
+struct drm_property *drm_property_create(struct drm_device *dev, int flags,
+ const char *name, int num_values)
+{
+ struct drm_property *property = NULL;
+ int ret;
+
+ property = kzalloc(sizeof(struct drm_property), GFP_KERNEL);
+ if (!property)
+ return NULL;
+
+ property->dev = dev;
+
+ if (num_values) {
+ property->values = kcalloc(num_values, sizeof(uint64_t),
+ GFP_KERNEL);
+ if (!property->values)
+ goto fail;
+ }
+
+ ret = drm_mode_object_get(dev, &property->base, DRM_MODE_OBJECT_PROPERTY);
+ if (ret)
+ goto fail;
+
+ property->flags = flags;
+ property->num_values = num_values;
+ INIT_LIST_HEAD(&property->enum_list);
+
+ if (name) {
+ strncpy(property->name, name, DRM_PROP_NAME_LEN);
+ property->name[DRM_PROP_NAME_LEN-1] = '\0';
+ }
+
+ list_add_tail(&property->head, &dev->mode_config.property_list);
+
+ WARN_ON(!drm_property_type_valid(property));
+
+ return property;
+fail:
+ kfree(property->values);
+ kfree(property);
+ return NULL;
+}
+EXPORT_SYMBOL(drm_property_create);
+
+/**
+ * drm_property_create_enum - create a new enumeration property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @props: enumeration lists with property values
+ * @num_values: number of pre-defined values
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy(), which is done automatically when calling
+ * drm_mode_config_cleanup().
+ *
+ * Userspace is only allowed to set one of the predefined values for enumeration
+ * properties.
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
+struct drm_property *drm_property_create_enum(struct drm_device *dev, int flags,
+ const char *name,
+ const struct drm_prop_enum_list *props,
+ int num_values)
+{
+ struct drm_property *property;
+ int i, ret;
+
+ flags |= DRM_MODE_PROP_ENUM;
+
+ property = drm_property_create(dev, flags, name, num_values);
+ if (!property)
+ return NULL;
+
+ for (i = 0; i < num_values; i++) {
+ ret = drm_property_add_enum(property, i,
+ props[i].type,
+ props[i].name);
+ if (ret) {
+ drm_property_destroy(dev, property);
+ return NULL;
+ }
+ }
+
+ return property;
+}
+EXPORT_SYMBOL(drm_property_create_enum);
+
+/**
+ * drm_property_create_bitmask - create a new bitmask property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @props: enumeration lists with property bitflags
+ * @num_props: size of the @props array
+ * @supported_bits: bitmask of all supported enumeration values
+ *
+ * This creates a new bitmask drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy(), which is done automatically when calling
+ * drm_mode_config_cleanup().
+ *
+ * Compared to plain enumeration properties userspace is allowed to set any
+ * or'ed together combination of the predefined property bitflag values
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
+struct drm_property *drm_property_create_bitmask(struct drm_device *dev,
+ int flags, const char *name,
+ const struct drm_prop_enum_list *props,
+ int num_props,
+ uint64_t supported_bits)
+{
+ struct drm_property *property;
+ int i, ret, index = 0;
+ int num_values = hweight64(supported_bits);
+
+ flags |= DRM_MODE_PROP_BITMASK;
+
+ property = drm_property_create(dev, flags, name, num_values);
+ if (!property)
+ return NULL;
+ for (i = 0; i < num_props; i++) {
+ if (!(supported_bits & (1ULL << props[i].type)))
+ continue;
+
+ if (WARN_ON(index >= num_values)) {
+ drm_property_destroy(dev, property);
+ return NULL;
+ }
+
+ ret = drm_property_add_enum(property, index++,
+ props[i].type,
+ props[i].name);
+ if (ret) {
+ drm_property_destroy(dev, property);
+ return NULL;
+ }
+ }
+
+ return property;
+}
+EXPORT_SYMBOL(drm_property_create_bitmask);
+
+static struct drm_property *property_create_range(struct drm_device *dev,
+ int flags, const char *name,
+ uint64_t min, uint64_t max)
+{
+ struct drm_property *property;
+
+ property = drm_property_create(dev, flags, name, 2);
+ if (!property)
+ return NULL;
+
+ property->values[0] = min;
+ property->values[1] = max;
+
+ return property;
+}
+
+/**
+ * drm_property_create_range - create a new unsigned ranged property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @min: minimum value of the property
+ * @max: maximum value of the property
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy(), which is done automatically when calling
+ * drm_mode_config_cleanup().
+ *
+ * Userspace is allowed to set any unsigned integer value in the (min, max)
+ * range inclusive.
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
+struct drm_property *drm_property_create_range(struct drm_device *dev, int flags,
+ const char *name,
+ uint64_t min, uint64_t max)
+{
+ return property_create_range(dev, DRM_MODE_PROP_RANGE | flags,
+ name, min, max);
+}
+EXPORT_SYMBOL(drm_property_create_range);
+
+/**
+ * drm_property_create_signed_range - create a new signed ranged property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @min: minimum value of the property
+ * @max: maximum value of the property
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy(), which is done automatically when calling
+ * drm_mode_config_cleanup().
+ *
+ * Userspace is allowed to set any signed integer value in the (min, max)
+ * range inclusive.
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
+struct drm_property *drm_property_create_signed_range(struct drm_device *dev,
+ int flags, const char *name,
+ int64_t min, int64_t max)
+{
+ return property_create_range(dev, DRM_MODE_PROP_SIGNED_RANGE | flags,
+ name, I642U64(min), I642U64(max));
+}
+EXPORT_SYMBOL(drm_property_create_signed_range);
+
+/**
+ * drm_property_create_object - create a new object property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ * @type: object type from DRM_MODE_OBJECT_* defines
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy(), which is done automatically when calling
+ * drm_mode_config_cleanup().
+ *
+ * Userspace is only allowed to set this to any property value of the given
+ * @type. Only useful for atomic properties, which is enforced.
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
+struct drm_property *drm_property_create_object(struct drm_device *dev,
+ int flags, const char *name,
+ uint32_t type)
+{
+ struct drm_property *property;
+
+ flags |= DRM_MODE_PROP_OBJECT;
+
+ if (WARN_ON(!(flags & DRM_MODE_PROP_ATOMIC)))
+ return NULL;
+
+ property = drm_property_create(dev, flags, name, 1);
+ if (!property)
+ return NULL;
+
+ property->values[0] = type;
+
+ return property;
+}
+EXPORT_SYMBOL(drm_property_create_object);
+
+/**
+ * drm_property_create_bool - create a new boolean property type
+ * @dev: drm device
+ * @flags: flags specifying the property type
+ * @name: name of the property
+ *
+ * This creates a new generic drm property which can then be attached to a drm
+ * object with drm_object_attach_property. The returned property object must be
+ * freed with drm_property_destroy(), which is done automatically when calling
+ * drm_mode_config_cleanup().
+ *
+ * This is implemented as a ranged property with only {0, 1} as valid values.
+ *
+ * Returns:
+ * A pointer to the newly created property on success, NULL on failure.
+ */
+struct drm_property *drm_property_create_bool(struct drm_device *dev, int flags,
+ const char *name)
+{
+ return drm_property_create_range(dev, flags, name, 0, 1);
+}
+EXPORT_SYMBOL(drm_property_create_bool);
+
+/**
+ * drm_property_add_enum - add a possible value to an enumeration property
+ * @property: enumeration property to change
+ * @index: index of the new enumeration
+ * @value: value of the new enumeration
+ * @name: symbolic name of the new enumeration
+ *
+ * This functions adds enumerations to a property.
+ *
+ * It's use is deprecated, drivers should use one of the more specific helpers
+ * to directly create the property with all enumerations already attached.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int drm_property_add_enum(struct drm_property *property, int index,
+ uint64_t value, const char *name)
+{
+ struct drm_property_enum *prop_enum;
+
+ if (!(drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
+ drm_property_type_is(property, DRM_MODE_PROP_BITMASK)))
+ return -EINVAL;
+
+ /*
+ * Bitmask enum properties have the additional constraint of values
+ * from 0 to 63
+ */
+ if (drm_property_type_is(property, DRM_MODE_PROP_BITMASK) &&
+ (value > 63))
+ return -EINVAL;
+
+ if (!list_empty(&property->enum_list)) {
+ list_for_each_entry(prop_enum, &property->enum_list, head) {
+ if (prop_enum->value == value) {
+ strncpy(prop_enum->name, name, DRM_PROP_NAME_LEN);
+ prop_enum->name[DRM_PROP_NAME_LEN-1] = '\0';
+ return 0;
+ }
+ }
+ }
+
+ prop_enum = kzalloc(sizeof(struct drm_property_enum), GFP_KERNEL);
+ if (!prop_enum)
+ return -ENOMEM;
+
+ strncpy(prop_enum->name, name, DRM_PROP_NAME_LEN);
+ prop_enum->name[DRM_PROP_NAME_LEN-1] = '\0';
+ prop_enum->value = value;
+
+ property->values[index] = value;
+ list_add_tail(&prop_enum->head, &property->enum_list);
+ return 0;
+}
+EXPORT_SYMBOL(drm_property_add_enum);
+
+/**
+ * drm_property_destroy - destroy a drm property
+ * @dev: drm device
+ * @property: property to destry
+ *
+ * This function frees a property including any attached resources like
+ * enumeration values.
+ */
+void drm_property_destroy(struct drm_device *dev, struct drm_property *property)
+{
+ struct drm_property_enum *prop_enum, *pt;
+
+ list_for_each_entry_safe(prop_enum, pt, &property->enum_list, head) {
+ list_del(&prop_enum->head);
+ kfree(prop_enum);
+ }
+
+ if (property->num_values)
+ kfree(property->values);
+ drm_mode_object_unregister(dev, &property->base);
+ list_del(&property->head);
+ kfree(property);
+}
+EXPORT_SYMBOL(drm_property_destroy);
+
+int drm_mode_getproperty_ioctl(struct drm_device *dev,
+ void *data, struct drm_file *file_priv)
+{
+ struct drm_mode_get_property *out_resp = data;
+ struct drm_property *property;
+ int enum_count = 0;
+ int value_count = 0;
+ int ret = 0, i;
+ int copied;
+ struct drm_property_enum *prop_enum;
+ struct drm_mode_property_enum __user *enum_ptr;
+ uint64_t __user *values_ptr;
+
+ if (!drm_core_check_feature(dev, DRIVER_MODESET))
+ return -EINVAL;
+
+ drm_modeset_lock_all(dev);
+ property = drm_property_find(dev, out_resp->prop_id);
+ if (!property) {
+ ret = -ENOENT;
+ goto done;
+ }
+
+ if (drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
+ drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
+ list_for_each_entry(prop_enum, &property->enum_list, head)
+ enum_count++;
+ }
+
+ value_count = property->num_values;
+
+ strncpy(out_resp->name, property->name, DRM_PROP_NAME_LEN);
+ out_resp->name[DRM_PROP_NAME_LEN-1] = 0;
+ out_resp->flags = property->flags;
+
+ if ((out_resp->count_values >= value_count) && value_count) {
+ values_ptr = (uint64_t __user *)(unsigned long)out_resp->values_ptr;
+ for (i = 0; i < value_count; i++) {
+ if (copy_to_user(values_ptr + i, &property->values[i], sizeof(uint64_t))) {
+ ret = -EFAULT;
+ goto done;
+ }
+ }
+ }
+ out_resp->count_values = value_count;
+
+ if (drm_property_type_is(property, DRM_MODE_PROP_ENUM) ||
+ drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
+ if ((out_resp->count_enum_blobs >= enum_count) && enum_count) {
+ copied = 0;
+ enum_ptr = (struct drm_mode_property_enum __user *)(unsigned long)out_resp->enum_blob_ptr;
+ list_for_each_entry(prop_enum, &property->enum_list, head) {
+
+ if (copy_to_user(&enum_ptr[copied].value, &prop_enum->value, sizeof(uint64_t))) {
+ ret = -EFAULT;
+ goto done;
+ }
+
+ if (copy_to_user(&enum_ptr[copied].name,
+ &prop_enum->name, DRM_PROP_NAME_LEN)) {
+ ret = -EFAULT;
+ goto done;
+ }
+ copied++;
+ }
+ }
+ out_resp->count_enum_blobs = enum_count;
+ }
+
+ /*
+ * NOTE: The idea seems to have been to use this to read all the blob
+ * property values. But nothing ever added them to the corresponding
+ * list, userspace always used the special-purpose get_blob ioctl to
+ * read the value for a blob property. It also doesn't make a lot of
+ * sense to return values here when everything else is just metadata for
+ * the property itself.
+ */
+ if (drm_property_type_is(property, DRM_MODE_PROP_BLOB))
+ out_resp->count_enum_blobs = 0;
+done:
+ drm_modeset_unlock_all(dev);
+ return ret;
+}
+
+static void drm_property_free_blob(struct kref *kref)
+{
+ struct drm_property_blob *blob =
+ container_of(kref, struct drm_property_blob, base.refcount);
+
+ mutex_lock(&blob->dev->mode_config.blob_lock);
+ list_del(&blob->head_global);
+ mutex_unlock(&blob->dev->mode_config.blob_lock);
+
+ drm_mode_object_unregister(blob->dev, &blob->base);
+
+ kfree(blob);
+}
+
+/**
+ * drm_property_create_blob - Create new blob property
+ * @dev: DRM device to create property for
+ * @length: Length to allocate for blob data
+ * @data: If specified, copies data into blob
+ *
+ * Creates a new blob property for a specified DRM device, optionally
+ * copying data. Note that blob properties are meant to be invariant, hence the
+ * data must be filled out before the blob is used as the value of any property.
+ *
+ * Returns:
+ * New blob property with a single reference on success, or an ERR_PTR
+ * value on failure.
+ */
+struct drm_property_blob *
+drm_property_create_blob(struct drm_device *dev, size_t length,
+ const void *data)
+{
+ struct drm_property_blob *blob;
+ int ret;
+
+ if (!length || length > ULONG_MAX - sizeof(struct drm_property_blob))
+ return ERR_PTR(-EINVAL);
+
+ blob = kzalloc(sizeof(struct drm_property_blob)+length, GFP_KERNEL);
+ if (!blob)
+ return ERR_PTR(-ENOMEM);
+
+ /* This must be explicitly initialised, so we can safely call list_del
+ * on it in the removal handler, even if it isn't in a file list. */
+ INIT_LIST_HEAD(&blob->head_file);
+ blob->length = length;
+ blob->dev = dev;
+
+ if (data)
+ memcpy(blob->data, data, length);
+
+ ret = drm_mode_object_get_reg(dev, &blob->base, DRM_MODE_OBJECT_BLOB,
+ true, drm_property_free_blob);
+ if (ret) {
+ kfree(blob);
+ return ERR_PTR(-EINVAL);
+ }
+
+ mutex_lock(&dev->mode_config.blob_lock);
+ list_add_tail(&blob->head_global,
+ &dev->mode_config.property_blob_list);
+ mutex_unlock(&dev->mode_config.blob_lock);
+
+ return blob;
+}
+EXPORT_SYMBOL(drm_property_create_blob);
+
+/**
+ * drm_property_unreference_blob - Unreference a blob property
+ * @blob: Pointer to blob property
+ *
+ * Drop a reference on a blob property. May free the object.
+ */
+void drm_property_unreference_blob(struct drm_property_blob *blob)
+{
+ if (!blob)
+ return;
+
+ drm_mode_object_unreference(&blob->base);
+}
+EXPORT_SYMBOL(drm_property_unreference_blob);
+
+void drm_property_destroy_user_blobs(struct drm_device *dev,
+ struct drm_file *file_priv)
+{
+ struct drm_property_blob *blob, *bt;
+
+ /*
+ * When the file gets released that means no one else can access the
+ * blob list any more, so no need to grab dev->blob_lock.
+ */
+ list_for_each_entry_safe(blob, bt, &file_priv->blobs, head_file) {
+ list_del_init(&blob->head_file);
+ drm_property_unreference_blob(blob);
+ }
+}
+
+/**
+ * drm_property_reference_blob - Take a reference on an existing property
+ * @blob: Pointer to blob property
+ *
+ * Take a new reference on an existing blob property. Returns @blob, which
+ * allows this to be used as a shorthand in assignments.
+ */
+struct drm_property_blob *drm_property_reference_blob(struct drm_property_blob *blob)
+{
+ drm_mode_object_reference(&blob->base);
+ return blob;
+}
+EXPORT_SYMBOL(drm_property_reference_blob);
+
+/**
+ * drm_property_lookup_blob - look up a blob property and take a reference
+ * @dev: drm device
+ * @id: id of the blob property
+ *
+ * If successful, this takes an additional reference to the blob property.
+ * callers need to make sure to eventually unreference the returned property
+ * again, using @drm_property_unreference_blob.
+ *
+ * Return:
+ * NULL on failure, pointer to the blob on success.
+ */
+struct drm_property_blob *drm_property_lookup_blob(struct drm_device *dev,
+ uint32_t id)
+{
+ struct drm_mode_object *obj;
+ struct drm_property_blob *blob = NULL;
+
+ obj = __drm_mode_object_find(dev, id, DRM_MODE_OBJECT_BLOB);
+ if (obj)
+ blob = obj_to_blob(obj);
+ return blob;
+}
+EXPORT_SYMBOL(drm_property_lookup_blob);
+
+/**
+ * drm_property_replace_global_blob - replace existing blob property
+ * @dev: drm device
+ * @replace: location of blob property pointer to be replaced
+ * @length: length of data for new blob, or 0 for no data
+ * @data: content for new blob, or NULL for no data
+ * @obj_holds_id: optional object for property holding blob ID
+ * @prop_holds_id: optional property holding blob ID
+ * @return 0 on success or error on failure
+ *
+ * This function will replace a global property in the blob list, optionally
+ * updating a property which holds the ID of that property.
+ *
+ * If length is 0 or data is NULL, no new blob will be created, and the holding
+ * property, if specified, will be set to 0.
+ *
+ * Access to the replace pointer is assumed to be protected by the caller, e.g.
+ * by holding the relevant modesetting object lock for its parent.
+ *
+ * For example, a drm_connector has a 'PATH' property, which contains the ID
+ * of a blob property with the value of the MST path information. Calling this
+ * function with replace pointing to the connector's path_blob_ptr, length and
+ * data set for the new path information, obj_holds_id set to the connector's
+ * base object, and prop_holds_id set to the path property name, will perform
+ * a completely atomic update. The access to path_blob_ptr is protected by the
+ * caller holding a lock on the connector.
+ */
+int drm_property_replace_global_blob(struct drm_device *dev,
+ struct drm_property_blob **replace,
+ size_t length,
+ const void *data,
+ struct drm_mode_object *obj_holds_id,
+ struct drm_property *prop_holds_id)
+{
+ struct drm_property_blob *new_blob = NULL;
+ struct drm_property_blob *old_blob = NULL;
+ int ret;
+
+ WARN_ON(replace == NULL);
+
+ old_blob = *replace;
+
+ if (length && data) {
+ new_blob = drm_property_create_blob(dev, length, data);
+ if (IS_ERR(new_blob))
+ return PTR_ERR(new_blob);
+ }
+
+ if (obj_holds_id) {
+ ret = drm_object_property_set_value(obj_holds_id,
+ prop_holds_id,
+ new_blob ?
+ new_blob->base.id : 0);
+ if (ret != 0)
+ goto err_created;
+ }
+
+ drm_property_unreference_blob(old_blob);
+ *replace = new_blob;
+
+ return 0;
+
+err_created:
+ drm_property_unreference_blob(new_blob);
+ return ret;
+}
+EXPORT_SYMBOL(drm_property_replace_global_blob);
+
+int drm_mode_getblob_ioctl(struct drm_device *dev,
+ void *data, struct drm_file *file_priv)
+{
+ struct drm_mode_get_blob *out_resp = data;
+ struct drm_property_blob *blob;
+ int ret = 0;
+ void __user *blob_ptr;
+
+ if (!drm_core_check_feature(dev, DRIVER_MODESET))
+ return -EINVAL;
+
+ blob = drm_property_lookup_blob(dev, out_resp->blob_id);
+ if (!blob)
+ return -ENOENT;
+
+ if (out_resp->length == blob->length) {
+ blob_ptr = (void __user *)(unsigned long)out_resp->data;
+ if (copy_to_user(blob_ptr, blob->data, blob->length)) {
+ ret = -EFAULT;
+ goto unref;
+ }
+ }
+ out_resp->length = blob->length;
+unref:
+ drm_property_unreference_blob(blob);
+
+ return ret;
+}
+
+int drm_mode_createblob_ioctl(struct drm_device *dev,
+ void *data, struct drm_file *file_priv)
+{
+ struct drm_mode_create_blob *out_resp = data;
+ struct drm_property_blob *blob;
+ void __user *blob_ptr;
+ int ret = 0;
+
+ if (!drm_core_check_feature(dev, DRIVER_MODESET))
+ return -EINVAL;
+
+ blob = drm_property_create_blob(dev, out_resp->length, NULL);
+ if (IS_ERR(blob))
+ return PTR_ERR(blob);
+
+ blob_ptr = (void __user *)(unsigned long)out_resp->data;
+ if (copy_from_user(blob->data, blob_ptr, out_resp->length)) {
+ ret = -EFAULT;
+ goto out_blob;
+ }
+
+ /* Dropping the lock between create_blob and our access here is safe
+ * as only the same file_priv can remove the blob; at this point, it is
+ * not associated with any file_priv. */
+ mutex_lock(&dev->mode_config.blob_lock);
+ out_resp->blob_id = blob->base.id;
+ list_add_tail(&blob->head_file, &file_priv->blobs);
+ mutex_unlock(&dev->mode_config.blob_lock);
+
+ return 0;
+
+out_blob:
+ drm_property_unreference_blob(blob);
+ return ret;
+}
+
+int drm_mode_destroyblob_ioctl(struct drm_device *dev,
+ void *data, struct drm_file *file_priv)
+{
+ struct drm_mode_destroy_blob *out_resp = data;
+ struct drm_property_blob *blob = NULL, *bt;
+ bool found = false;
+ int ret = 0;
+
+ if (!drm_core_check_feature(dev, DRIVER_MODESET))
+ return -EINVAL;
+
+ blob = drm_property_lookup_blob(dev, out_resp->blob_id);
+ if (!blob)
+ return -ENOENT;
+
+ mutex_lock(&dev->mode_config.blob_lock);
+ /* Ensure the property was actually created by this user. */
+ list_for_each_entry(bt, &file_priv->blobs, head_file) {
+ if (bt == blob) {
+ found = true;
+ break;
+ }
+ }
+
+ if (!found) {
+ ret = -EPERM;
+ goto err;
+ }
+
+ /* We must drop head_file here, because we may not be the last
+ * reference on the blob. */
+ list_del_init(&blob->head_file);
+ mutex_unlock(&dev->mode_config.blob_lock);
+
+ /* One reference from lookup, and one from the filp. */
+ drm_property_unreference_blob(blob);
+ drm_property_unreference_blob(blob);
+
+ return 0;
+
+err:
+ mutex_unlock(&dev->mode_config.blob_lock);
+ drm_property_unreference_blob(blob);
+
+ return ret;
+}
+
+/* Some properties could refer to dynamic refcnt'd objects, or things that
+ * need special locking to handle lifetime issues (ie. to ensure the prop
+ * value doesn't become invalid part way through the property update due to
+ * race). The value returned by reference via 'obj' should be passed back
+ * to drm_property_change_valid_put() after the property is set (and the
+ * object to which the property is attached has a chance to take it's own
+ * reference).
+ */
+bool drm_property_change_valid_get(struct drm_property *property,
+ uint64_t value, struct drm_mode_object **ref)
+{
+ int i;
+
+ if (property->flags & DRM_MODE_PROP_IMMUTABLE)
+ return false;
+
+ *ref = NULL;
+
+ if (drm_property_type_is(property, DRM_MODE_PROP_RANGE)) {
+ if (value < property->values[0] || value > property->values[1])
+ return false;
+ return true;
+ } else if (drm_property_type_is(property, DRM_MODE_PROP_SIGNED_RANGE)) {
+ int64_t svalue = U642I64(value);
+
+ if (svalue < U642I64(property->values[0]) ||
+ svalue > U642I64(property->values[1]))
+ return false;
+ return true;
+ } else if (drm_property_type_is(property, DRM_MODE_PROP_BITMASK)) {
+ uint64_t valid_mask = 0;
+
+ for (i = 0; i < property->num_values; i++)
+ valid_mask |= (1ULL << property->values[i]);
+ return !(value & ~valid_mask);
+ } else if (drm_property_type_is(property, DRM_MODE_PROP_BLOB)) {
+ struct drm_property_blob *blob;
+
+ if (value == 0)
+ return true;
+
+ blob = drm_property_lookup_blob(property->dev, value);
+ if (blob) {
+ *ref = &blob->base;
+ return true;
+ } else {
+ return false;
+ }
+ } else if (drm_property_type_is(property, DRM_MODE_PROP_OBJECT)) {
+ /* a zero value for an object property translates to null: */
+ if (value == 0)
+ return true;
+
+ *ref = __drm_mode_object_find(property->dev, value,
+ property->values[0]);
+ return *ref != NULL;
+ }
+
+ for (i = 0; i < property->num_values; i++)
+ if (property->values[i] == value)
+ return true;
+ return false;
+}
+
+void drm_property_change_valid_put(struct drm_property *property,
+ struct drm_mode_object *ref)
+{
+ if (!ref)
+ return;
+
+ if (drm_property_type_is(property, DRM_MODE_PROP_OBJECT)) {
+ drm_mode_object_unreference(ref);
+ } else if (drm_property_type_is(property, DRM_MODE_PROP_BLOB))
+ drm_property_unreference_blob(obj_to_blob(ref));
+}
config DRM_I2C_NXP_TDA998X
tristate "NXP Semiconductors TDA998X HDMI encoder"
default m if DRM_TILCDC
+ select SND_SOC_HDMI_CODEC if SND_SOC
help
Support for NXP Semiconductors TDA998X HDMI encoders.
#include <linux/module.h>
#include <linux/irq.h>
#include <sound/asoundef.h>
+#include <sound/hdmi-codec.h>
#include <drm/drmP.h>
#include <drm/drm_atomic_helper.h>
#define DBG(fmt, ...) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
+struct tda998x_audio_port {
+ u8 format; /* AFMT_xxx */
+ u8 config; /* AP value */
+};
+
struct tda998x_priv {
struct i2c_client *cec;
struct i2c_client *hdmi;
u8 vip_cntrl_0;
u8 vip_cntrl_1;
u8 vip_cntrl_2;
- struct tda998x_encoder_params params;
+ struct tda998x_audio_params audio_params;
+
+ struct platform_device *audio_pdev;
+ struct mutex audio_mutex;
wait_queue_head_t wq_edid;
volatile int wq_edid_wait;
struct drm_encoder encoder;
struct drm_connector connector;
+
+ struct tda998x_audio_port audio_port[2];
};
#define conn_to_tda998x_priv(x) \
reg_set(priv, REG_DIP_IF_FLAGS, bit);
}
-static void
-tda998x_write_aif(struct tda998x_priv *priv, struct tda998x_encoder_params *p)
+static int tda998x_write_aif(struct tda998x_priv *priv,
+ struct hdmi_audio_infoframe *cea)
{
union hdmi_infoframe frame;
- hdmi_audio_infoframe_init(&frame.audio);
-
- frame.audio.channels = p->audio_frame[1] & 0x07;
- frame.audio.channel_allocation = p->audio_frame[4];
- frame.audio.level_shift_value = (p->audio_frame[5] & 0x78) >> 3;
- frame.audio.downmix_inhibit = (p->audio_frame[5] & 0x80) >> 7;
-
- /*
- * L-PCM and IEC61937 compressed audio shall always set sample
- * frequency to "refer to stream". For others, see the HDMI
- * specification.
- */
- frame.audio.sample_frequency = (p->audio_frame[2] & 0x1c) >> 2;
+ frame.audio = *cea;
tda998x_write_if(priv, DIP_IF_FLAGS_IF4, REG_IF4_HB0, &frame);
+
+ return 0;
}
static void
}
}
-static void
+static int
tda998x_configure_audio(struct tda998x_priv *priv,
- struct drm_display_mode *mode, struct tda998x_encoder_params *p)
+ struct tda998x_audio_params *params,
+ unsigned mode_clock)
{
u8 buf[6], clksel_aip, clksel_fs, cts_n, adiv;
u32 n;
/* Enable audio ports */
- reg_write(priv, REG_ENA_AP, p->audio_cfg);
- reg_write(priv, REG_ENA_ACLK, p->audio_clk_cfg);
+ reg_write(priv, REG_ENA_AP, params->config);
/* Set audio input source */
- switch (p->audio_format) {
+ switch (params->format) {
case AFMT_SPDIF:
+ reg_write(priv, REG_ENA_ACLK, 0);
reg_write(priv, REG_MUX_AP, MUX_AP_SELECT_SPDIF);
clksel_aip = AIP_CLKSEL_AIP_SPDIF;
clksel_fs = AIP_CLKSEL_FS_FS64SPDIF;
break;
case AFMT_I2S:
+ reg_write(priv, REG_ENA_ACLK, 1);
reg_write(priv, REG_MUX_AP, MUX_AP_SELECT_I2S);
clksel_aip = AIP_CLKSEL_AIP_I2S;
clksel_fs = AIP_CLKSEL_FS_ACLK;
- cts_n = CTS_N_M(3) | CTS_N_K(3);
+ switch (params->sample_width) {
+ case 16:
+ cts_n = CTS_N_M(3) | CTS_N_K(1);
+ break;
+ case 18:
+ case 20:
+ case 24:
+ cts_n = CTS_N_M(3) | CTS_N_K(2);
+ break;
+ default:
+ case 32:
+ cts_n = CTS_N_M(3) | CTS_N_K(3);
+ break;
+ }
break;
default:
- BUG();
- return;
+ dev_err(&priv->hdmi->dev, "Unsupported I2S format\n");
+ return -EINVAL;
}
reg_write(priv, REG_AIP_CLKSEL, clksel_aip);
* assume 100MHz requires larger divider.
*/
adiv = AUDIO_DIV_SERCLK_8;
- if (mode->clock > 100000)
+ if (mode_clock > 100000)
adiv++; /* AUDIO_DIV_SERCLK_16 */
/* S/PDIF asks for a larger divider */
- if (p->audio_format == AFMT_SPDIF)
+ if (params->format == AFMT_SPDIF)
adiv++; /* AUDIO_DIV_SERCLK_16 or _32 */
reg_write(priv, REG_AUDIO_DIV, adiv);
* This is the approximate value of N, which happens to be
* the recommended values for non-coherent clocks.
*/
- n = 128 * p->audio_sample_rate / 1000;
+ n = 128 * params->sample_rate / 1000;
/* Write the CTS and N values */
buf[0] = 0x44;
reg_set(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_CTS);
reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_RST_CTS);
- /* Write the channel status */
- buf[0] = IEC958_AES0_CON_NOT_COPYRIGHT;
- buf[1] = 0x00;
- buf[2] = IEC958_AES3_CON_FS_NOTID;
- buf[3] = IEC958_AES4_CON_ORIGFS_NOTID |
- IEC958_AES4_CON_MAX_WORDLEN_24;
+ /* Write the channel status
+ * The REG_CH_STAT_B-registers skip IEC958 AES2 byte, because
+ * there is a separate register for each I2S wire.
+ */
+ buf[0] = params->status[0];
+ buf[1] = params->status[1];
+ buf[2] = params->status[3];
+ buf[3] = params->status[4];
reg_write_range(priv, REG_CH_STAT_B(0), buf, 4);
tda998x_audio_mute(priv, true);
msleep(20);
tda998x_audio_mute(priv, false);
- /* Write the audio information packet */
- tda998x_write_aif(priv, p);
+ return tda998x_write_aif(priv, ¶ms->cea);
}
/* DRM encoder functions */
VIP_CNTRL_2_SWAP_F(p->swap_f) |
(p->mirr_f ? VIP_CNTRL_2_MIRR_F : 0);
- priv->params = *p;
+ priv->audio_params = p->audio_params;
}
static void tda998x_encoder_dpms(struct drm_encoder *encoder, int mode)
tda998x_write_avi(priv, adjusted_mode);
- if (priv->params.audio_cfg)
- tda998x_configure_audio(priv, adjusted_mode,
- &priv->params);
+ if (priv->audio_params.format != AFMT_UNUSED) {
+ mutex_lock(&priv->audio_mutex);
+ tda998x_configure_audio(priv,
+ &priv->audio_params,
+ adjusted_mode->clock);
+ mutex_unlock(&priv->audio_mutex);
+ }
}
}
drm_mode_connector_update_edid_property(connector, edid);
n = drm_add_edid_modes(connector, edid);
priv->is_hdmi_sink = drm_detect_hdmi_monitor(edid);
+ drm_edid_to_eld(connector, edid);
+
kfree(edid);
return n;
cec_write(priv, REG_CEC_RXSHPDINTENA, 0);
reg_clear(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
+ if (priv->audio_pdev)
+ platform_device_unregister(priv->audio_pdev);
+
if (priv->hdmi->irq)
free_irq(priv->hdmi->irq, priv);
i2c_unregister_device(priv->cec);
}
+static int tda998x_audio_hw_params(struct device *dev, void *data,
+ struct hdmi_codec_daifmt *daifmt,
+ struct hdmi_codec_params *params)
+{
+ struct tda998x_priv *priv = dev_get_drvdata(dev);
+ int i, ret;
+ struct tda998x_audio_params audio = {
+ .sample_width = params->sample_width,
+ .sample_rate = params->sample_rate,
+ .cea = params->cea,
+ };
+
+ if (!priv->encoder.crtc)
+ return -ENODEV;
+
+ memcpy(audio.status, params->iec.status,
+ min(sizeof(audio.status), sizeof(params->iec.status)));
+
+ switch (daifmt->fmt) {
+ case HDMI_I2S:
+ if (daifmt->bit_clk_inv || daifmt->frame_clk_inv ||
+ daifmt->bit_clk_master || daifmt->frame_clk_master) {
+ dev_err(dev, "%s: Bad flags %d %d %d %d\n", __func__,
+ daifmt->bit_clk_inv, daifmt->frame_clk_inv,
+ daifmt->bit_clk_master,
+ daifmt->frame_clk_master);
+ return -EINVAL;
+ }
+ for (i = 0; i < ARRAY_SIZE(priv->audio_port); i++)
+ if (priv->audio_port[i].format == AFMT_I2S)
+ audio.config = priv->audio_port[i].config;
+ audio.format = AFMT_I2S;
+ break;
+ case HDMI_SPDIF:
+ for (i = 0; i < ARRAY_SIZE(priv->audio_port); i++)
+ if (priv->audio_port[i].format == AFMT_SPDIF)
+ audio.config = priv->audio_port[i].config;
+ audio.format = AFMT_SPDIF;
+ break;
+ default:
+ dev_err(dev, "%s: Invalid format %d\n", __func__, daifmt->fmt);
+ return -EINVAL;
+ }
+
+ if (audio.config == 0) {
+ dev_err(dev, "%s: No audio configutation found\n", __func__);
+ return -EINVAL;
+ }
+
+ mutex_lock(&priv->audio_mutex);
+ ret = tda998x_configure_audio(priv,
+ &audio,
+ priv->encoder.crtc->hwmode.clock);
+
+ if (ret == 0)
+ priv->audio_params = audio;
+ mutex_unlock(&priv->audio_mutex);
+
+ return ret;
+}
+
+static void tda998x_audio_shutdown(struct device *dev, void *data)
+{
+ struct tda998x_priv *priv = dev_get_drvdata(dev);
+
+ mutex_lock(&priv->audio_mutex);
+
+ reg_write(priv, REG_ENA_AP, 0);
+
+ priv->audio_params.format = AFMT_UNUSED;
+
+ mutex_unlock(&priv->audio_mutex);
+}
+
+int tda998x_audio_digital_mute(struct device *dev, void *data, bool enable)
+{
+ struct tda998x_priv *priv = dev_get_drvdata(dev);
+
+ mutex_lock(&priv->audio_mutex);
+
+ tda998x_audio_mute(priv, enable);
+
+ mutex_unlock(&priv->audio_mutex);
+ return 0;
+}
+
+static int tda998x_audio_get_eld(struct device *dev, void *data,
+ uint8_t *buf, size_t len)
+{
+ struct tda998x_priv *priv = dev_get_drvdata(dev);
+ struct drm_mode_config *config = &priv->encoder.dev->mode_config;
+ struct drm_connector *connector;
+ int ret = -ENODEV;
+
+ mutex_lock(&config->mutex);
+ list_for_each_entry(connector, &config->connector_list, head) {
+ if (&priv->encoder == connector->encoder) {
+ memcpy(buf, connector->eld,
+ min(sizeof(connector->eld), len));
+ ret = 0;
+ }
+ }
+ mutex_unlock(&config->mutex);
+
+ return ret;
+}
+
+static const struct hdmi_codec_ops audio_codec_ops = {
+ .hw_params = tda998x_audio_hw_params,
+ .audio_shutdown = tda998x_audio_shutdown,
+ .digital_mute = tda998x_audio_digital_mute,
+ .get_eld = tda998x_audio_get_eld,
+};
+
+static int tda998x_audio_codec_init(struct tda998x_priv *priv,
+ struct device *dev)
+{
+ struct hdmi_codec_pdata codec_data = {
+ .ops = &audio_codec_ops,
+ .max_i2s_channels = 2,
+ };
+ int i;
+
+ for (i = 0; i < ARRAY_SIZE(priv->audio_port); i++) {
+ if (priv->audio_port[i].format == AFMT_I2S &&
+ priv->audio_port[i].config != 0)
+ codec_data.i2s = 1;
+ if (priv->audio_port[i].format == AFMT_SPDIF &&
+ priv->audio_port[i].config != 0)
+ codec_data.spdif = 1;
+ }
+
+ priv->audio_pdev = platform_device_register_data(
+ dev, HDMI_CODEC_DRV_NAME, PLATFORM_DEVID_AUTO,
+ &codec_data, sizeof(codec_data));
+
+ return PTR_ERR_OR_ZERO(priv->audio_pdev);
+}
+
/* I2C driver functions */
+static int tda998x_get_audio_ports(struct tda998x_priv *priv,
+ struct device_node *np)
+{
+ const u32 *port_data;
+ u32 size;
+ int i;
+
+ port_data = of_get_property(np, "audio-ports", &size);
+ if (!port_data)
+ return 0;
+
+ size /= sizeof(u32);
+ if (size > 2 * ARRAY_SIZE(priv->audio_port) || size % 2 != 0) {
+ dev_err(&priv->hdmi->dev,
+ "Bad number of elements in audio-ports dt-property\n");
+ return -EINVAL;
+ }
+
+ size /= 2;
+
+ for (i = 0; i < size; i++) {
+ u8 afmt = be32_to_cpup(&port_data[2*i]);
+ u8 ena_ap = be32_to_cpup(&port_data[2*i+1]);
+
+ if (afmt != AFMT_SPDIF && afmt != AFMT_I2S) {
+ dev_err(&priv->hdmi->dev,
+ "Bad audio format %u\n", afmt);
+ return -EINVAL;
+ }
+
+ priv->audio_port[i].format = afmt;
+ priv->audio_port[i].config = ena_ap;
+ }
+
+ if (priv->audio_port[0].format == priv->audio_port[1].format) {
+ dev_err(&priv->hdmi->dev,
+ "There can only be on I2S port and one SPDIF port\n");
+ return -EINVAL;
+ }
+ return 0;
+}
+
static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
{
struct device_node *np = client->dev.of_node;
if (!np)
return 0; /* non-DT */
- /* get the optional video properties */
+ /* get the device tree parameters */
ret = of_property_read_u32(np, "video-ports", &video);
if (ret == 0) {
priv->vip_cntrl_0 = video >> 16;
priv->vip_cntrl_2 = video;
}
- return 0;
+ mutex_init(&priv->audio_mutex); /* Protect access from audio thread */
+ ret = tda998x_get_audio_ports(priv, np);
+ if (ret)
+ goto fail;
+
+ if (priv->audio_port[0].format != AFMT_UNUSED)
+ tda998x_audio_codec_init(priv, &client->dev);
+
+ return 0;
fail:
/* if encoder_init fails, the encoder slave is never registered,
* so cleanup here:
# Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
subdir-ccflags-$(CONFIG_DRM_I915_WERROR) := -Werror
+subdir-ccflags-y += \
+ $(call as-instr,movntdqa (%eax)$(comma)%xmm0,-DCONFIG_AS_MOVNTDQA)
# Please keep these build lists sorted!
# core driver code
i915-y := i915_drv.o \
i915_irq.o \
+ i915_memcpy.o \
+ i915_mm.o \
i915_params.o \
i915_pci.o \
i915_suspend.o \
+ i915_sw_fence.o \
i915_sysfs.o \
intel_csr.o \
intel_device_info.o \
include $(src)/gvt/Makefile
endif
-obj-$(CONFIG_DRM_I915) += i915.o
+obj-$(CONFIG_DRM_I915) += i915.o
CFLAGS_i915_trace_points.o := -I$(src)
* general bitmasking mechanism.
*/
-#define STD_MI_OPCODE_MASK 0xFF800000
-#define STD_3D_OPCODE_MASK 0xFFFF0000
-#define STD_2D_OPCODE_MASK 0xFFC00000
-#define STD_MFX_OPCODE_MASK 0xFFFF0000
+#define STD_MI_OPCODE_SHIFT (32 - 9)
+#define STD_3D_OPCODE_SHIFT (32 - 16)
+#define STD_2D_OPCODE_SHIFT (32 - 10)
+#define STD_MFX_OPCODE_SHIFT (32 - 16)
+#define MIN_OPCODE_SHIFT 16
#define CMD(op, opm, f, lm, fl, ...) \
{ \
.flags = (fl) | ((f) ? CMD_DESC_FIXED : 0), \
- .cmd = { (op), (opm) }, \
+ .cmd = { (op), ~0u << (opm) }, \
.length = { (lm) }, \
__VA_ARGS__ \
}
/* Convenience macros to compress the tables */
-#define SMI STD_MI_OPCODE_MASK
-#define S3D STD_3D_OPCODE_MASK
-#define S2D STD_2D_OPCODE_MASK
-#define SMFX STD_MFX_OPCODE_MASK
+#define SMI STD_MI_OPCODE_SHIFT
+#define S3D STD_3D_OPCODE_SHIFT
+#define S2D STD_2D_OPCODE_SHIFT
+#define SMFX STD_MFX_OPCODE_SHIFT
#define F true
#define S CMD_DESC_SKIP
#define R CMD_DESC_REJECT
CMD( MI_LOAD_SCAN_LINES_EXCL, SMI, !F, 0x3F, R ),
};
+static const struct drm_i915_cmd_descriptor noop_desc =
+ CMD(MI_NOOP, SMI, F, 1, S);
+
#undef CMD
#undef SMI
#undef S3D
REG32(GEN7_GPGPU_DISPATCHDIMX),
REG32(GEN7_GPGPU_DISPATCHDIMY),
REG32(GEN7_GPGPU_DISPATCHDIMZ),
+ REG64_IDX(RING_TIMESTAMP, BSD_RING_BASE),
REG64_IDX(GEN7_SO_NUM_PRIMS_WRITTEN, 0),
REG64_IDX(GEN7_SO_NUM_PRIMS_WRITTEN, 1),
REG64_IDX(GEN7_SO_NUM_PRIMS_WRITTEN, 2),
REG32(GEN7_L3SQCREG1),
REG32(GEN7_L3CNTLREG2),
REG32(GEN7_L3CNTLREG3),
+ REG64_IDX(RING_TIMESTAMP, BLT_RING_BASE),
};
static const struct drm_i915_reg_descriptor hsw_render_regs[] = {
};
static const struct drm_i915_reg_descriptor gen7_blt_regs[] = {
+ REG64_IDX(RING_TIMESTAMP, RENDER_RING_BASE),
+ REG64_IDX(RING_TIMESTAMP, BSD_RING_BASE),
REG32(BCS_SWCTRL),
+ REG64_IDX(RING_TIMESTAMP, BLT_RING_BASE),
};
static const struct drm_i915_reg_descriptor ivb_master_regs[] = {
* non-opcode bits being set. But if we don't include those bits, some 3D
* commands may hash to the same bucket due to not including opcode bits that
* make the command unique. For now, we will risk hashing to the same bucket.
- *
- * If we attempt to generate a perfect hash, we should be able to look at bits
- * 31:29 of a command from a batch buffer and use the full mask for that
- * client. The existing INSTR_CLIENT_MASK/SHIFT defines can be used for this.
*/
-#define CMD_HASH_MASK STD_MI_OPCODE_MASK
+static inline u32 cmd_header_key(u32 x)
+{
+ u32 shift;
+
+ switch (x >> INSTR_CLIENT_SHIFT) {
+ default:
+ case INSTR_MI_CLIENT:
+ shift = STD_MI_OPCODE_SHIFT;
+ break;
+ case INSTR_RC_CLIENT:
+ shift = STD_3D_OPCODE_SHIFT;
+ break;
+ case INSTR_BC_CLIENT:
+ shift = STD_2D_OPCODE_SHIFT;
+ break;
+ }
+
+ return x >> shift;
+}
static int init_hash_table(struct intel_engine_cs *engine,
const struct drm_i915_cmd_table *cmd_tables,
desc_node->desc = desc;
hash_add(engine->cmd_hash, &desc_node->node,
- desc->cmd.value & CMD_HASH_MASK);
+ cmd_header_key(desc->cmd.value));
}
}
* Optionally initializes fields related to batch buffer command parsing in the
* struct intel_engine_cs based on whether the platform requires software
* command parsing.
- *
- * Return: non-zero if initialization fails
*/
-int intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
+void intel_engine_init_cmd_parser(struct intel_engine_cs *engine)
{
const struct drm_i915_cmd_table *cmd_tables;
int cmd_table_count;
int ret;
if (!IS_GEN7(engine->i915))
- return 0;
+ return;
switch (engine->id) {
case RCS:
break;
default:
MISSING_CASE(engine->id);
- BUG();
+ return;
}
- BUG_ON(!validate_cmds_sorted(engine, cmd_tables, cmd_table_count));
- BUG_ON(!validate_regs_sorted(engine));
-
- WARN_ON(!hash_empty(engine->cmd_hash));
+ if (!validate_cmds_sorted(engine, cmd_tables, cmd_table_count)) {
+ DRM_ERROR("%s: command descriptions are not sorted\n",
+ engine->name);
+ return;
+ }
+ if (!validate_regs_sorted(engine)) {
+ DRM_ERROR("%s: registers are not sorted\n", engine->name);
+ return;
+ }
ret = init_hash_table(engine, cmd_tables, cmd_table_count);
if (ret) {
- DRM_ERROR("CMD: cmd_parser_init failed!\n");
+ DRM_ERROR("%s: initialised failed!\n", engine->name);
fini_hash_table(engine);
- return ret;
+ return;
}
engine->needs_cmd_parser = true;
-
- return 0;
}
/**
struct cmd_node *desc_node;
hash_for_each_possible(engine->cmd_hash, desc_node, node,
- cmd_header & CMD_HASH_MASK) {
+ cmd_header_key(cmd_header)) {
const struct drm_i915_cmd_descriptor *desc = desc_node->desc;
- u32 masked_cmd = desc->cmd.mask & cmd_header;
- u32 masked_value = desc->cmd.value & desc->cmd.mask;
-
- if (masked_cmd == masked_value)
+ if (((cmd_header ^ desc->cmd.value) & desc->cmd.mask) == 0)
return desc;
}
static const struct drm_i915_cmd_descriptor*
find_cmd(struct intel_engine_cs *engine,
u32 cmd_header,
+ const struct drm_i915_cmd_descriptor *desc,
struct drm_i915_cmd_descriptor *default_desc)
{
- const struct drm_i915_cmd_descriptor *desc;
u32 mask;
+ if (((cmd_header ^ desc->cmd.value) & desc->cmd.mask) == 0)
+ return desc;
+
desc = find_cmd_in_table(engine, cmd_header);
if (desc)
return desc;
if (!mask)
return NULL;
- BUG_ON(!default_desc);
- default_desc->flags = CMD_DESC_SKIP;
+ default_desc->cmd.value = cmd_header;
+ default_desc->cmd.mask = ~0u << MIN_OPCODE_SHIFT;
default_desc->length.mask = mask;
-
+ default_desc->flags = CMD_DESC_SKIP;
return default_desc;
}
static const struct drm_i915_reg_descriptor *
-find_reg(const struct drm_i915_reg_descriptor *table,
- int count, u32 addr)
+__find_reg(const struct drm_i915_reg_descriptor *table, int count, u32 addr)
{
- int i;
-
- for (i = 0; i < count; i++) {
- if (i915_mmio_reg_offset(table[i].addr) == addr)
- return &table[i];
+ int start = 0, end = count;
+ while (start < end) {
+ int mid = start + (end - start) / 2;
+ int ret = addr - i915_mmio_reg_offset(table[mid].addr);
+ if (ret < 0)
+ end = mid;
+ else if (ret > 0)
+ start = mid + 1;
+ else
+ return &table[mid];
}
-
return NULL;
}
static const struct drm_i915_reg_descriptor *
-find_reg_in_tables(const struct drm_i915_reg_table *tables,
- int count, bool is_master, u32 addr)
+find_reg(const struct intel_engine_cs *engine, bool is_master, u32 addr)
{
- int i;
- const struct drm_i915_reg_table *table;
- const struct drm_i915_reg_descriptor *reg;
+ const struct drm_i915_reg_table *table = engine->reg_tables;
+ int count = engine->reg_table_count;
- for (i = 0; i < count; i++) {
- table = &tables[i];
+ do {
if (!table->master || is_master) {
- reg = find_reg(table->regs, table->num_regs,
- addr);
+ const struct drm_i915_reg_descriptor *reg;
+
+ reg = __find_reg(table->regs, table->num_regs, addr);
if (reg != NULL)
return reg;
}
- }
+ } while (table++, --count);
return NULL;
}
-static u32 *vmap_batch(struct drm_i915_gem_object *obj,
- unsigned start, unsigned len)
-{
- int i;
- void *addr = NULL;
- struct sg_page_iter sg_iter;
- int first_page = start >> PAGE_SHIFT;
- int last_page = (len + start + 4095) >> PAGE_SHIFT;
- int npages = last_page - first_page;
- struct page **pages;
-
- pages = drm_malloc_ab(npages, sizeof(*pages));
- if (pages == NULL) {
- DRM_DEBUG_DRIVER("Failed to get space for pages\n");
- goto finish;
- }
-
- i = 0;
- for_each_sg_page(obj->pages->sgl, &sg_iter, obj->pages->nents, first_page) {
- pages[i++] = sg_page_iter_page(&sg_iter);
- if (i == npages)
- break;
- }
-
- addr = vmap(pages, i, 0, PAGE_KERNEL);
- if (addr == NULL) {
- DRM_DEBUG_DRIVER("Failed to vmap pages\n");
- goto finish;
- }
-
-finish:
- if (pages)
- drm_free_large(pages);
- return (u32*)addr;
-}
-
-/* Returns a vmap'd pointer to dest_obj, which the caller must unmap */
-static u32 *copy_batch(struct drm_i915_gem_object *dest_obj,
+/* Returns a vmap'd pointer to dst_obj, which the caller must unmap */
+static u32 *copy_batch(struct drm_i915_gem_object *dst_obj,
struct drm_i915_gem_object *src_obj,
u32 batch_start_offset,
- u32 batch_len)
+ u32 batch_len,
+ bool *needs_clflush_after)
{
- int needs_clflush = 0;
- void *src_base, *src;
- void *dst = NULL;
+ unsigned int src_needs_clflush;
+ unsigned int dst_needs_clflush;
+ void *dst, *src;
int ret;
- if (batch_len > dest_obj->base.size ||
- batch_len + batch_start_offset > src_obj->base.size)
- return ERR_PTR(-E2BIG);
-
- if (WARN_ON(dest_obj->pages_pin_count == 0))
- return ERR_PTR(-ENODEV);
-
- ret = i915_gem_obj_prepare_shmem_read(src_obj, &needs_clflush);
- if (ret) {
- DRM_DEBUG_DRIVER("CMD: failed to prepare shadow batch\n");
+ ret = i915_gem_obj_prepare_shmem_read(src_obj, &src_needs_clflush);
+ if (ret)
return ERR_PTR(ret);
- }
- src_base = vmap_batch(src_obj, batch_start_offset, batch_len);
- if (!src_base) {
- DRM_DEBUG_DRIVER("CMD: Failed to vmap batch\n");
- ret = -ENOMEM;
+ ret = i915_gem_obj_prepare_shmem_write(dst_obj, &dst_needs_clflush);
+ if (ret) {
+ dst = ERR_PTR(ret);
goto unpin_src;
}
- ret = i915_gem_object_set_to_cpu_domain(dest_obj, true);
- if (ret) {
- DRM_DEBUG_DRIVER("CMD: Failed to set shadow batch to CPU\n");
- goto unmap_src;
+ dst = i915_gem_object_pin_map(dst_obj, I915_MAP_WB);
+ if (IS_ERR(dst))
+ goto unpin_dst;
+
+ src = ERR_PTR(-ENODEV);
+ if (src_needs_clflush &&
+ i915_memcpy_from_wc((void *)(uintptr_t)batch_start_offset, NULL, 0)) {
+ src = i915_gem_object_pin_map(src_obj, I915_MAP_WC);
+ if (!IS_ERR(src)) {
+ i915_memcpy_from_wc(dst,
+ src + batch_start_offset,
+ ALIGN(batch_len, 16));
+ i915_gem_object_unpin_map(src_obj);
+ }
}
-
- dst = vmap_batch(dest_obj, 0, batch_len);
- if (!dst) {
- DRM_DEBUG_DRIVER("CMD: Failed to vmap shadow batch\n");
- ret = -ENOMEM;
- goto unmap_src;
+ if (IS_ERR(src)) {
+ void *ptr;
+ int offset, n;
+
+ offset = offset_in_page(batch_start_offset);
+
+ /* We can avoid clflushing partial cachelines before the write
+ * if we only every write full cache-lines. Since we know that
+ * both the source and destination are in multiples of
+ * PAGE_SIZE, we can simply round up to the next cacheline.
+ * We don't care about copying too much here as we only
+ * validate up to the end of the batch.
+ */
+ if (dst_needs_clflush & CLFLUSH_BEFORE)
+ batch_len = roundup(batch_len,
+ boot_cpu_data.x86_clflush_size);
+
+ ptr = dst;
+ for (n = batch_start_offset >> PAGE_SHIFT; batch_len; n++) {
+ int len = min_t(int, batch_len, PAGE_SIZE - offset);
+
+ src = kmap_atomic(i915_gem_object_get_page(src_obj, n));
+ if (src_needs_clflush)
+ drm_clflush_virt_range(src + offset, len);
+ memcpy(ptr, src + offset, len);
+ kunmap_atomic(src);
+
+ ptr += len;
+ batch_len -= len;
+ offset = 0;
+ }
}
- src = src_base + offset_in_page(batch_start_offset);
- if (needs_clflush)
- drm_clflush_virt_range(src, batch_len);
-
- memcpy(dst, src, batch_len);
+ /* dst_obj is returned with vmap pinned */
+ *needs_clflush_after = dst_needs_clflush & CLFLUSH_AFTER;
-unmap_src:
- vunmap(src_base);
+unpin_dst:
+ i915_gem_obj_finish_shmem_access(dst_obj);
unpin_src:
- i915_gem_object_unpin_pages(src_obj);
-
- return ret ? ERR_PTR(ret) : dst;
+ i915_gem_obj_finish_shmem_access(src_obj);
+ return dst;
}
/**
const bool is_master,
bool *oacontrol_set)
{
+ if (desc->flags & CMD_DESC_SKIP)
+ return true;
+
if (desc->flags & CMD_DESC_REJECT) {
DRM_DEBUG_DRIVER("CMD: Rejected command: 0x%08X\n", *cmd);
return false;
offset += step) {
const u32 reg_addr = cmd[offset] & desc->reg.mask;
const struct drm_i915_reg_descriptor *reg =
- find_reg_in_tables(engine->reg_tables,
- engine->reg_table_count,
- is_master,
- reg_addr);
+ find_reg(engine, is_master, reg_addr);
if (!reg) {
DRM_DEBUG_DRIVER("CMD: Rejected register 0x%08X in command: 0x%08X (exec_id=%d)\n",
u32 batch_len,
bool is_master)
{
- u32 *cmd, *batch_base, *batch_end;
- struct drm_i915_cmd_descriptor default_desc = { 0 };
+ u32 *cmd, *batch_end;
+ struct drm_i915_cmd_descriptor default_desc = noop_desc;
+ const struct drm_i915_cmd_descriptor *desc = &default_desc;
bool oacontrol_set = false; /* OACONTROL tracking. See check_cmd() */
+ bool needs_clflush_after = false;
int ret = 0;
- batch_base = copy_batch(shadow_batch_obj, batch_obj,
- batch_start_offset, batch_len);
- if (IS_ERR(batch_base)) {
+ cmd = copy_batch(shadow_batch_obj, batch_obj,
+ batch_start_offset, batch_len,
+ &needs_clflush_after);
+ if (IS_ERR(cmd)) {
DRM_DEBUG_DRIVER("CMD: Failed to copy batch\n");
- return PTR_ERR(batch_base);
+ return PTR_ERR(cmd);
}
/*
* large or larger and copy_batch() will write MI_NOPs to the extra
* space. Parsing should be faster in some cases this way.
*/
- batch_end = batch_base + (batch_len / sizeof(*batch_end));
-
- cmd = batch_base;
+ batch_end = cmd + (batch_len / sizeof(*batch_end));
while (cmd < batch_end) {
- const struct drm_i915_cmd_descriptor *desc;
u32 length;
if (*cmd == MI_BATCH_BUFFER_END)
break;
- desc = find_cmd(engine, *cmd, &default_desc);
+ desc = find_cmd(engine, *cmd, desc, &default_desc);
if (!desc) {
DRM_DEBUG_DRIVER("CMD: Unrecognized command: 0x%08X\n",
*cmd);
ret = -EINVAL;
}
- vunmap(batch_base);
+ if (ret == 0 && needs_clflush_after)
+ drm_clflush_virt_range(shadow_batch_obj->mapping, batch_len);
+ i915_gem_object_unpin_map(shadow_batch_obj);
return ret;
}
#include <drm/i915_drm.h>
#include "i915_drv.h"
-enum {
- ACTIVE_LIST,
- INACTIVE_LIST,
- PINNED_LIST,
-};
+static inline struct drm_i915_private *node_to_i915(struct drm_info_node *node)
+{
+ return to_i915(node->minor->dev);
+}
/* As the drm_debugfs_init() routines are called before dev->dev_private is
* allocated we need to hook into the minor for release. */
node->minor = minor;
node->dent = ent;
- node->info_ent = (void *) key;
+ node->info_ent = (void *)key;
mutex_lock(&minor->debugfs_lock);
list_add(&node->list, &minor->debugfs_list);
static int i915_capabilities(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- const struct intel_device_info *info = INTEL_INFO(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ const struct intel_device_info *info = INTEL_INFO(dev_priv);
- seq_printf(m, "gen: %d\n", info->gen);
- seq_printf(m, "pch: %d\n", INTEL_PCH_TYPE(dev));
+ seq_printf(m, "gen: %d\n", INTEL_GEN(dev_priv));
+ seq_printf(m, "pch: %d\n", INTEL_PCH_TYPE(dev_priv));
#define PRINT_FLAG(x) seq_printf(m, #x ": %s\n", yesno(info->x))
#define SEP_SEMICOLON ;
DEV_INFO_FOR_EACH_FLAG(PRINT_FLAG, SEP_SEMICOLON);
static char get_global_flag(struct drm_i915_gem_object *obj)
{
- return i915_gem_obj_to_ggtt(obj) ? 'g' : ' ';
+ return i915_gem_object_to_ggtt(obj, NULL) ? 'g' : ' ';
}
static char get_pin_mapped_flag(struct drm_i915_gem_object *obj)
seq_printf(m, "%x ",
i915_gem_active_get_seqno(&obj->last_read[id],
&obj->base.dev->struct_mutex));
- seq_printf(m, "] %x %x%s%s%s",
+ seq_printf(m, "] %x %s%s%s",
i915_gem_active_get_seqno(&obj->last_write,
&obj->base.dev->struct_mutex),
- i915_gem_active_get_seqno(&obj->last_fence,
- &obj->base.dev->struct_mutex),
- i915_cache_level_str(to_i915(obj->base.dev), obj->cache_level),
+ i915_cache_level_str(dev_priv, obj->cache_level),
obj->dirty ? " dirty" : "",
obj->madv == I915_MADV_DONTNEED ? " purgeable" : "");
if (obj->base.name)
seq_printf(m, " (pinned x %d)", pin_count);
if (obj->pin_display)
seq_printf(m, " (display)");
- if (obj->fence_reg != I915_FENCE_REG_NONE)
- seq_printf(m, " (fence: %d)", obj->fence_reg);
list_for_each_entry(vma, &obj->vma_list, obj_link) {
if (!drm_mm_node_allocated(&vma->node))
continue;
vma->node.start, vma->node.size);
if (i915_vma_is_ggtt(vma))
seq_printf(m, ", type: %u", vma->ggtt_view.type);
+ if (vma->fence)
+ seq_printf(m, " , fence: %d%s",
+ vma->fence->id,
+ i915_gem_active_isset(&vma->last_fence) ? "*" : "");
seq_puts(m, ")");
}
if (obj->stolen)
}
engine = i915_gem_active_get_engine(&obj->last_write,
- &obj->base.dev->struct_mutex);
+ &dev_priv->drm.struct_mutex);
if (engine)
seq_printf(m, " (%s)", engine->name);
seq_printf(m, " (frontbuffer: 0x%03x)", frontbuffer_bits);
}
-static int i915_gem_object_list_info(struct seq_file *m, void *data)
-{
- struct drm_info_node *node = m->private;
- uintptr_t list = (uintptr_t) node->info_ent->data;
- struct list_head *head;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
- struct i915_ggtt *ggtt = &dev_priv->ggtt;
- struct i915_vma *vma;
- u64 total_obj_size, total_gtt_size;
- int count, ret;
-
- ret = mutex_lock_interruptible(&dev->struct_mutex);
- if (ret)
- return ret;
-
- /* FIXME: the user of this interface might want more than just GGTT */
- switch (list) {
- case ACTIVE_LIST:
- seq_puts(m, "Active:\n");
- head = &ggtt->base.active_list;
- break;
- case INACTIVE_LIST:
- seq_puts(m, "Inactive:\n");
- head = &ggtt->base.inactive_list;
- break;
- default:
- mutex_unlock(&dev->struct_mutex);
- return -EINVAL;
- }
-
- total_obj_size = total_gtt_size = count = 0;
- list_for_each_entry(vma, head, vm_link) {
- seq_printf(m, " ");
- describe_obj(m, vma->obj);
- seq_printf(m, "\n");
- total_obj_size += vma->obj->base.size;
- total_gtt_size += vma->node.size;
- count++;
- }
- mutex_unlock(&dev->struct_mutex);
-
- seq_printf(m, "Total %d objects, %llu bytes, %llu GTT size\n",
- count, total_obj_size, total_gtt_size);
- return 0;
-}
-
static int obj_rank_by_stolen(void *priv,
struct list_head *A, struct list_head *B)
{
static int i915_gem_stolen_list_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct drm_i915_gem_object *obj;
u64 total_obj_size, total_gtt_size;
LIST_HEAD(stolen);
return 0;
}
-#define count_objects(list, member) do { \
- list_for_each_entry(obj, list, member) { \
- size += i915_gem_obj_total_ggtt_size(obj); \
- ++count; \
- if (obj->map_and_fenceable) { \
- mappable_size += i915_gem_obj_ggtt_size(obj); \
- ++mappable_count; \
- } \
- } \
-} while (0)
-
struct file_stats {
struct drm_i915_file_private *file_priv;
unsigned long count;
for (n = 0; n < ARRAY_SIZE(ctx->engine); n++) {
if (ctx->engine[n].state)
- per_file_stats(0, ctx->engine[n].state, data);
+ per_file_stats(0, ctx->engine[n].state->obj, data);
if (ctx->engine[n].ring)
- per_file_stats(0, ctx->engine[n].ring->obj, data);
+ per_file_stats(0, ctx->engine[n].ring->vma->obj, data);
}
return 0;
static void print_context_stats(struct seq_file *m,
struct drm_i915_private *dev_priv)
{
+ struct drm_device *dev = &dev_priv->drm;
struct file_stats stats;
struct drm_file *file;
memset(&stats, 0, sizeof(stats));
- mutex_lock(&dev_priv->drm.struct_mutex);
+ mutex_lock(&dev->struct_mutex);
if (dev_priv->kernel_context)
per_file_ctx_stats(0, dev_priv->kernel_context, &stats);
- list_for_each_entry(file, &dev_priv->drm.filelist, lhead) {
+ list_for_each_entry(file, &dev->filelist, lhead) {
struct drm_i915_file_private *fpriv = file->driver_priv;
idr_for_each(&fpriv->context_idr, per_file_ctx_stats, &stats);
}
- mutex_unlock(&dev_priv->drm.struct_mutex);
+ mutex_unlock(&dev->struct_mutex);
print_file_stats(m, "[k]contexts", stats);
}
-#define count_vmas(list, member) do { \
- list_for_each_entry(vma, list, member) { \
- size += i915_gem_obj_total_ggtt_size(vma->obj); \
- ++count; \
- if (vma->obj->map_and_fenceable) { \
- mappable_size += i915_gem_obj_ggtt_size(vma->obj); \
- ++mappable_count; \
- } \
- } \
-} while (0)
-
-static int i915_gem_object_info(struct seq_file *m, void* data)
+static int i915_gem_object_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct i915_ggtt *ggtt = &dev_priv->ggtt;
- u32 count, mappable_count, purgeable_count;
- u64 size, mappable_size, purgeable_size;
- unsigned long pin_mapped_count = 0, pin_mapped_purgeable_count = 0;
- u64 pin_mapped_size = 0, pin_mapped_purgeable_size = 0;
+ u32 count, mapped_count, purgeable_count, dpy_count;
+ u64 size, mapped_size, purgeable_size, dpy_size;
struct drm_i915_gem_object *obj;
struct drm_file *file;
- struct i915_vma *vma;
int ret;
ret = mutex_lock_interruptible(&dev->struct_mutex);
dev_priv->mm.object_count,
dev_priv->mm.object_memory);
- size = count = mappable_size = mappable_count = 0;
- count_objects(&dev_priv->mm.bound_list, global_list);
- seq_printf(m, "%u [%u] objects, %llu [%llu] bytes in gtt\n",
- count, mappable_count, size, mappable_size);
-
- size = count = mappable_size = mappable_count = 0;
- count_vmas(&ggtt->base.active_list, vm_link);
- seq_printf(m, " %u [%u] active objects, %llu [%llu] bytes\n",
- count, mappable_count, size, mappable_size);
+ size = count = 0;
+ mapped_size = mapped_count = 0;
+ purgeable_size = purgeable_count = 0;
+ list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) {
+ size += obj->base.size;
+ ++count;
- size = count = mappable_size = mappable_count = 0;
- count_vmas(&ggtt->base.inactive_list, vm_link);
- seq_printf(m, " %u [%u] inactive objects, %llu [%llu] bytes\n",
- count, mappable_count, size, mappable_size);
+ if (obj->madv == I915_MADV_DONTNEED) {
+ purgeable_size += obj->base.size;
+ ++purgeable_count;
+ }
- size = count = purgeable_size = purgeable_count = 0;
- list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) {
- size += obj->base.size, ++count;
- if (obj->madv == I915_MADV_DONTNEED)
- purgeable_size += obj->base.size, ++purgeable_count;
if (obj->mapping) {
- pin_mapped_count++;
- pin_mapped_size += obj->base.size;
- if (obj->pages_pin_count == 0) {
- pin_mapped_purgeable_count++;
- pin_mapped_purgeable_size += obj->base.size;
- }
+ mapped_count++;
+ mapped_size += obj->base.size;
}
}
seq_printf(m, "%u unbound objects, %llu bytes\n", count, size);
- size = count = mappable_size = mappable_count = 0;
+ size = count = dpy_size = dpy_count = 0;
list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
- if (obj->fault_mappable) {
- size += i915_gem_obj_ggtt_size(obj);
- ++count;
- }
+ size += obj->base.size;
+ ++count;
+
if (obj->pin_display) {
- mappable_size += i915_gem_obj_ggtt_size(obj);
- ++mappable_count;
+ dpy_size += obj->base.size;
+ ++dpy_count;
}
+
if (obj->madv == I915_MADV_DONTNEED) {
purgeable_size += obj->base.size;
++purgeable_count;
}
+
if (obj->mapping) {
- pin_mapped_count++;
- pin_mapped_size += obj->base.size;
- if (obj->pages_pin_count == 0) {
- pin_mapped_purgeable_count++;
- pin_mapped_purgeable_size += obj->base.size;
- }
+ mapped_count++;
+ mapped_size += obj->base.size;
}
}
+ seq_printf(m, "%u bound objects, %llu bytes\n",
+ count, size);
seq_printf(m, "%u purgeable objects, %llu bytes\n",
purgeable_count, purgeable_size);
- seq_printf(m, "%u pinned mappable objects, %llu bytes\n",
- mappable_count, mappable_size);
- seq_printf(m, "%u fault mappable objects, %llu bytes\n",
- count, size);
- seq_printf(m,
- "%lu [%lu] pin mapped objects, %llu [%llu] bytes [purgeable]\n",
- pin_mapped_count, pin_mapped_purgeable_count,
- pin_mapped_size, pin_mapped_purgeable_size);
+ seq_printf(m, "%u mapped objects, %llu bytes\n",
+ mapped_count, mapped_size);
+ seq_printf(m, "%u display objects (pinned), %llu bytes\n",
+ dpy_count, dpy_size);
seq_printf(m, "%llu [%llu] gtt total\n",
ggtt->base.total, ggtt->mappable_end - ggtt->base.start);
print_context_stats(m, dev_priv);
list_for_each_entry_reverse(file, &dev->filelist, lhead) {
struct file_stats stats;
+ struct drm_i915_file_private *file_priv = file->driver_priv;
+ struct drm_i915_gem_request *request;
struct task_struct *task;
memset(&stats, 0, sizeof(stats));
* still alive (e.g. get_pid(current) => fork() => exit()).
* Therefore, we need to protect this ->comm access using RCU.
*/
+ mutex_lock(&dev->struct_mutex);
+ request = list_first_entry_or_null(&file_priv->mm.request_list,
+ struct drm_i915_gem_request,
+ client_list);
rcu_read_lock();
- task = pid_task(file->pid, PIDTYPE_PID);
+ task = pid_task(request && request->ctx->pid ?
+ request->ctx->pid : file->pid,
+ PIDTYPE_PID);
print_file_stats(m, task ? task->comm : "<unknown>", stats);
rcu_read_unlock();
+ mutex_unlock(&dev->struct_mutex);
}
mutex_unlock(&dev->filelist_mutex);
static int i915_gem_gtt_info(struct seq_file *m, void *data)
{
struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- uintptr_t list = (uintptr_t) node->info_ent->data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(node);
+ struct drm_device *dev = &dev_priv->drm;
+ bool show_pin_display_only = !!node->info_ent->data;
struct drm_i915_gem_object *obj;
u64 total_obj_size, total_gtt_size;
int count, ret;
total_obj_size = total_gtt_size = count = 0;
list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
- if (list == PINNED_LIST && !i915_gem_obj_is_pinned(obj))
+ if (show_pin_display_only && !obj->pin_display)
continue;
seq_puts(m, " ");
static int i915_gem_pageflip_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_crtc *crtc;
int ret;
intel_crtc_get_vblank_counter(crtc));
seq_printf(m, "%d prepares\n", atomic_read(&work->pending));
- if (INTEL_INFO(dev)->gen >= 4)
+ if (INTEL_GEN(dev_priv) >= 4)
addr = I915_HI_DISPBASE(I915_READ(DSPSURF(crtc->plane)));
else
addr = I915_READ(DSPADDR(crtc->plane));
static int i915_gem_batch_pool_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct drm_i915_gem_object *obj;
struct intel_engine_cs *engine;
int total = 0;
static int i915_gem_request_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_engine_cs *engine;
struct drm_i915_gem_request *req;
int ret, any;
seq_printf(m, "%s requests: %d\n", engine->name, count);
list_for_each_entry(req, &engine->request_list, link) {
+ struct pid *pid = req->ctx->pid;
struct task_struct *task;
rcu_read_lock();
- task = NULL;
- if (req->pid)
- task = pid_task(req->pid, PIDTYPE_PID);
+ task = pid ? pid_task(pid, PIDTYPE_PID) : NULL;
seq_printf(m, " %x @ %d: %s [%d]\n",
req->fence.seqno,
(int) (jiffies - req->emitted_jiffies),
seq_printf(m, "Current sequence (%s): %x\n",
engine->name, intel_engine_get_seqno(engine));
- seq_printf(m, "Current user interrupts (%s): %lx\n",
- engine->name, READ_ONCE(engine->breadcrumbs.irq_wakeups));
spin_lock(&b->lock);
for (rb = rb_first(&b->waiters); rb; rb = rb_next(rb)) {
static int i915_gem_seqno_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
struct intel_engine_cs *engine;
- int ret;
-
- ret = mutex_lock_interruptible(&dev->struct_mutex);
- if (ret)
- return ret;
- intel_runtime_pm_get(dev_priv);
for_each_engine(engine, dev_priv)
i915_ring_seqno_info(m, engine);
- intel_runtime_pm_put(dev_priv);
- mutex_unlock(&dev->struct_mutex);
-
return 0;
}
static int i915_interrupt_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
struct intel_engine_cs *engine;
- int ret, i, pipe;
+ int i, pipe;
- ret = mutex_lock_interruptible(&dev->struct_mutex);
- if (ret)
- return ret;
intel_runtime_pm_get(dev_priv);
- if (IS_CHERRYVIEW(dev)) {
+ if (IS_CHERRYVIEW(dev_priv)) {
seq_printf(m, "Master Interrupt Control:\t%08x\n",
I915_READ(GEN8_MASTER_IRQ));
I915_READ(GEN8_PCU_IIR));
seq_printf(m, "PCU interrupt enable:\t%08x\n",
I915_READ(GEN8_PCU_IER));
- } else if (INTEL_INFO(dev)->gen >= 8) {
+ } else if (INTEL_GEN(dev_priv) >= 8) {
seq_printf(m, "Master Interrupt Control:\t%08x\n",
I915_READ(GEN8_MASTER_IRQ));
I915_READ(GEN8_PCU_IIR));
seq_printf(m, "PCU interrupt enable:\t%08x\n",
I915_READ(GEN8_PCU_IER));
- } else if (IS_VALLEYVIEW(dev)) {
+ } else if (IS_VALLEYVIEW(dev_priv)) {
seq_printf(m, "Display IER:\t%08x\n",
I915_READ(VLV_IER));
seq_printf(m, "Display IIR:\t%08x\n",
seq_printf(m, "DPINVGTT:\t%08x\n",
I915_READ(DPINVGTT));
- } else if (!HAS_PCH_SPLIT(dev)) {
+ } else if (!HAS_PCH_SPLIT(dev_priv)) {
seq_printf(m, "Interrupt enable: %08x\n",
I915_READ(IER));
seq_printf(m, "Interrupt identity: %08x\n",
I915_READ(GTIMR));
}
for_each_engine(engine, dev_priv) {
- if (INTEL_INFO(dev)->gen >= 6) {
+ if (INTEL_GEN(dev_priv) >= 6) {
seq_printf(m,
"Graphics Interrupt mask (%s): %08x\n",
engine->name, I915_READ_IMR(engine));
i915_ring_seqno_info(m, engine);
}
intel_runtime_pm_put(dev_priv);
- mutex_unlock(&dev->struct_mutex);
return 0;
}
static int i915_gem_fence_regs_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
int i, ret;
ret = mutex_lock_interruptible(&dev->struct_mutex);
seq_printf(m, "Total fences = %d\n", dev_priv->num_fence_regs);
for (i = 0; i < dev_priv->num_fence_regs; i++) {
- struct drm_i915_gem_object *obj = dev_priv->fence_regs[i].obj;
+ struct i915_vma *vma = dev_priv->fence_regs[i].vma;
seq_printf(m, "Fence %d, pin count = %d, object = ",
i, dev_priv->fence_regs[i].pin_count);
- if (obj == NULL)
+ if (!vma)
seq_puts(m, "unused");
else
- describe_obj(m, obj);
+ describe_obj(m, vma->obj);
seq_putc(m, '\n');
}
static int i915_hws_info(struct seq_file *m, void *data)
{
struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(node);
struct intel_engine_cs *engine;
const u32 *hws;
int i;
loff_t *ppos)
{
struct i915_error_state_file_priv *error_priv = filp->private_data;
- struct drm_device *dev = error_priv->dev;
- int ret;
DRM_DEBUG_DRIVER("Resetting error state\n");
-
- ret = mutex_lock_interruptible(&dev->struct_mutex);
- if (ret)
- return ret;
-
- i915_destroy_error_state(dev);
- mutex_unlock(&dev->struct_mutex);
+ i915_destroy_error_state(error_priv->dev);
return cnt;
}
static int i915_error_state_open(struct inode *inode, struct file *file)
{
- struct drm_device *dev = inode->i_private;
+ struct drm_i915_private *dev_priv = inode->i_private;
struct i915_error_state_file_priv *error_priv;
error_priv = kzalloc(sizeof(*error_priv), GFP_KERNEL);
if (!error_priv)
return -ENOMEM;
- error_priv->dev = dev;
+ error_priv->dev = &dev_priv->drm;
- i915_error_state_get(dev, error_priv);
+ i915_error_state_get(&dev_priv->drm, error_priv);
file->private_data = error_priv;
ssize_t ret_count = 0;
int ret;
- ret = i915_error_state_buf_init(&error_str, to_i915(error_priv->dev), count, *pos);
+ ret = i915_error_state_buf_init(&error_str,
+ to_i915(error_priv->dev), count, *pos);
if (ret)
return ret;
static int
i915_next_seqno_get(void *data, u64 *val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
int ret;
- ret = mutex_lock_interruptible(&dev->struct_mutex);
+ ret = mutex_lock_interruptible(&dev_priv->drm.struct_mutex);
if (ret)
return ret;
*val = dev_priv->next_seqno;
- mutex_unlock(&dev->struct_mutex);
+ mutex_unlock(&dev_priv->drm.struct_mutex);
return 0;
}
static int
i915_next_seqno_set(void *data, u64 val)
{
- struct drm_device *dev = data;
+ struct drm_i915_private *dev_priv = data;
+ struct drm_device *dev = &dev_priv->drm;
int ret;
ret = mutex_lock_interruptible(&dev->struct_mutex);
static int i915_frequency_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
int ret = 0;
intel_runtime_pm_get(dev_priv);
- if (IS_GEN5(dev)) {
+ if (IS_GEN5(dev_priv)) {
u16 rgvswctl = I915_READ16(MEMSWCTL);
u16 rgvstat = I915_READ16(MEMSTAT_ILK);
MEMSTAT_VID_SHIFT);
seq_printf(m, "Current P-state: %d\n",
(rgvstat & MEMSTAT_PSTATE_MASK) >> MEMSTAT_PSTATE_SHIFT);
- } else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) {
+ } else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
u32 freq_sts;
mutex_lock(&dev_priv->rps.hw_lock);
"efficient (RPe) frequency: %d MHz\n",
intel_gpu_freq(dev_priv, dev_priv->rps.efficient_freq));
mutex_unlock(&dev_priv->rps.hw_lock);
- } else if (INTEL_INFO(dev)->gen >= 6) {
+ } else if (INTEL_GEN(dev_priv) >= 6) {
u32 rp_state_limits;
u32 gt_perf_status;
u32 rp_state_cap;
int max_freq;
rp_state_limits = I915_READ(GEN6_RP_STATE_LIMITS);
- if (IS_BROXTON(dev)) {
+ if (IS_BROXTON(dev_priv)) {
rp_state_cap = I915_READ(BXT_RP_STATE_CAP);
gt_perf_status = I915_READ(BXT_GT_PERF_STATUS);
} else {
intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
reqf = I915_READ(GEN6_RPNSWREQ);
- if (IS_GEN9(dev))
+ if (IS_GEN9(dev_priv))
reqf >>= 23;
else {
reqf &= ~GEN6_TURBO_DISABLE;
- if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
reqf >>= 24;
else
reqf >>= 25;
rpdownei = I915_READ(GEN6_RP_CUR_DOWN_EI) & GEN6_CURIAVG_MASK;
rpcurdown = I915_READ(GEN6_RP_CUR_DOWN) & GEN6_CURBSYTAVG_MASK;
rpprevdown = I915_READ(GEN6_RP_PREV_DOWN) & GEN6_CURBSYTAVG_MASK;
- if (IS_GEN9(dev))
+ if (IS_GEN9(dev_priv))
cagf = (rpstat & GEN9_CAGF_MASK) >> GEN9_CAGF_SHIFT;
- else if (IS_HASWELL(dev) || IS_BROADWELL(dev))
+ else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
cagf = (rpstat & HSW_CAGF_MASK) >> HSW_CAGF_SHIFT;
else
cagf = (rpstat & GEN6_CAGF_MASK) >> GEN6_CAGF_SHIFT;
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
mutex_unlock(&dev->struct_mutex);
- if (IS_GEN6(dev) || IS_GEN7(dev)) {
+ if (IS_GEN6(dev_priv) || IS_GEN7(dev_priv)) {
pm_ier = I915_READ(GEN6_PMIER);
pm_imr = I915_READ(GEN6_PMIMR);
pm_isr = I915_READ(GEN6_PMISR);
seq_printf(m, "pm_intr_keep: 0x%08x\n", dev_priv->rps.pm_intr_keep);
seq_printf(m, "GT_PERF_STATUS: 0x%08x\n", gt_perf_status);
seq_printf(m, "Render p-state ratio: %d\n",
- (gt_perf_status & (IS_GEN9(dev) ? 0x1ff00 : 0xff00)) >> 8);
+ (gt_perf_status & (IS_GEN9(dev_priv) ? 0x1ff00 : 0xff00)) >> 8);
seq_printf(m, "Render p-state VID: %d\n",
gt_perf_status & 0xff);
seq_printf(m, "Render p-state limit: %d\n",
seq_printf(m, "Down threshold: %d%%\n",
dev_priv->rps.down_threshold);
- max_freq = (IS_BROXTON(dev) ? rp_state_cap >> 0 :
+ max_freq = (IS_BROXTON(dev_priv) ? rp_state_cap >> 0 :
rp_state_cap >> 16) & 0xff;
- max_freq *= (IS_SKYLAKE(dev) || IS_KABYLAKE(dev) ?
+ max_freq *= (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ?
GEN9_FREQ_SCALER : 1);
seq_printf(m, "Lowest (RPN) frequency: %dMHz\n",
intel_gpu_freq(dev_priv, max_freq));
max_freq = (rp_state_cap & 0xff00) >> 8;
- max_freq *= (IS_SKYLAKE(dev) || IS_KABYLAKE(dev) ?
+ max_freq *= (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ?
GEN9_FREQ_SCALER : 1);
seq_printf(m, "Nominal (RP1) frequency: %dMHz\n",
intel_gpu_freq(dev_priv, max_freq));
- max_freq = (IS_BROXTON(dev) ? rp_state_cap >> 16 :
+ max_freq = (IS_BROXTON(dev_priv) ? rp_state_cap >> 16 :
rp_state_cap >> 0) & 0xff;
- max_freq *= (IS_SKYLAKE(dev) || IS_KABYLAKE(dev) ?
+ max_freq *= (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ?
GEN9_FREQ_SCALER : 1);
seq_printf(m, "Max non-overclocked (RP0) frequency: %dMHz\n",
intel_gpu_freq(dev_priv, max_freq));
static int i915_hangcheck_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
struct intel_engine_cs *engine;
u64 acthd[I915_NUM_ENGINES];
u32 seqno[I915_NUM_ENGINES];
enum intel_engine_id id;
int j;
+ if (test_bit(I915_WEDGED, &dev_priv->gpu_error.flags))
+ seq_printf(m, "Wedged\n");
+ if (test_bit(I915_RESET_IN_PROGRESS, &dev_priv->gpu_error.flags))
+ seq_printf(m, "Reset in progress\n");
+ if (waitqueue_active(&dev_priv->gpu_error.wait_queue))
+ seq_printf(m, "Waiter holding struct mutex\n");
+ if (waitqueue_active(&dev_priv->gpu_error.reset_queue))
+ seq_printf(m, "struct_mutex blocked for reset\n");
+
if (!i915.enable_hangcheck) {
seq_printf(m, "Hangcheck disabled\n");
return 0;
engine->hangcheck.seqno,
seqno[id],
engine->last_submitted_seqno);
- seq_printf(m, "\twaiters? %d\n",
- intel_engine_has_waiter(engine));
- seq_printf(m, "\tuser interrupts = %lx [current %lx]\n",
- engine->hangcheck.user_interrupts,
- READ_ONCE(engine->breadcrumbs.irq_wakeups));
+ seq_printf(m, "\twaiters? %s, fake irq active? %s\n",
+ yesno(intel_engine_has_waiter(engine)),
+ yesno(test_bit(engine->id,
+ &dev_priv->gpu_error.missed_irq_rings)));
seq_printf(m, "\tACTHD = 0x%08llx [current 0x%08llx]\n",
(long long)engine->hangcheck.acthd,
(long long)acthd[id]);
static int ironlake_drpc_info(struct seq_file *m)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
u32 rgvmodectl, rstdbyctl;
u16 crstandvid;
int ret;
static int i915_forcewake_domains(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
struct intel_uncore_forcewake_domain *fw_domain;
spin_lock_irq(&dev_priv->uncore.lock);
static int vlv_drpc_info(struct seq_file *m)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
u32 rpmodectl1, rcctl1, pw_status;
intel_runtime_pm_get(dev_priv);
static int gen6_drpc_info(struct seq_file *m)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
u32 rpmodectl1, gt_core_status, rcctl1, rc6vids = 0;
u32 gen9_powergate_enable = 0, gen9_powergate_status = 0;
unsigned forcewake_count;
rpmodectl1 = I915_READ(GEN6_RP_CONTROL);
rcctl1 = I915_READ(GEN6_RC_CONTROL);
- if (INTEL_INFO(dev)->gen >= 9) {
+ if (INTEL_GEN(dev_priv) >= 9) {
gen9_powergate_enable = I915_READ(GEN9_PG_ENABLE);
gen9_powergate_status = I915_READ(GEN9_PWRGT_DOMAIN_STATUS);
}
yesno(rcctl1 & GEN6_RC_CTL_RC1e_ENABLE));
seq_printf(m, "RC6 Enabled: %s\n",
yesno(rcctl1 & GEN6_RC_CTL_RC6_ENABLE));
- if (INTEL_INFO(dev)->gen >= 9) {
+ if (INTEL_GEN(dev_priv) >= 9) {
seq_printf(m, "Render Well Gating Enabled: %s\n",
yesno(gen9_powergate_enable & GEN9_RENDER_PG_ENABLE));
seq_printf(m, "Media Well Gating Enabled: %s\n",
seq_printf(m, "Core Power Down: %s\n",
yesno(gt_core_status & GEN6_CORE_CPD_STATE_MASK));
- if (INTEL_INFO(dev)->gen >= 9) {
+ if (INTEL_GEN(dev_priv) >= 9) {
seq_printf(m, "Render Power Well: %s\n",
(gen9_powergate_status &
GEN9_PWRGT_RENDER_STATUS_MASK) ? "Up" : "Down");
static int i915_drpc_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
return vlv_drpc_info(m);
- else if (INTEL_INFO(dev)->gen >= 6)
+ else if (INTEL_GEN(dev_priv) >= 6)
return gen6_drpc_info(m);
else
return ironlake_drpc_info(m);
static int i915_frontbuffer_tracking(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
seq_printf(m, "FB tracking busy bits: 0x%08x\n",
dev_priv->fb_tracking.busy_bits);
static int i915_fbc_status(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
- if (!HAS_FBC(dev)) {
+ if (!HAS_FBC(dev_priv)) {
seq_puts(m, "FBC unsupported on this chipset\n");
return 0;
}
seq_printf(m, "FBC disabled: %s\n",
dev_priv->fbc.no_fbc_reason);
- if (INTEL_INFO(dev_priv)->gen >= 7)
+ if (INTEL_GEN(dev_priv) >= 7)
seq_printf(m, "Compressing: %s\n",
yesno(I915_READ(FBC_STATUS2) &
FBC_COMPRESSION_MASK));
static int i915_fbc_fc_get(void *data, u64 *val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
- if (INTEL_INFO(dev)->gen < 7 || !HAS_FBC(dev))
+ if (INTEL_GEN(dev_priv) < 7 || !HAS_FBC(dev_priv))
return -ENODEV;
*val = dev_priv->fbc.false_color;
static int i915_fbc_fc_set(void *data, u64 val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
u32 reg;
- if (INTEL_INFO(dev)->gen < 7 || !HAS_FBC(dev))
+ if (INTEL_GEN(dev_priv) < 7 || !HAS_FBC(dev_priv))
return -ENODEV;
mutex_lock(&dev_priv->fbc.lock);
static int i915_ips_status(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
- if (!HAS_IPS(dev)) {
+ if (!HAS_IPS(dev_priv)) {
seq_puts(m, "not supported\n");
return 0;
}
seq_printf(m, "Enabled by kernel parameter: %s\n",
yesno(i915.enable_ips));
- if (INTEL_INFO(dev)->gen >= 8) {
+ if (INTEL_GEN(dev_priv) >= 8) {
seq_puts(m, "Currently: unknown\n");
} else {
if (I915_READ(IPS_CTL) & IPS_ENABLE)
static int i915_sr_status(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
bool sr_enabled = false;
intel_runtime_pm_get(dev_priv);
- if (HAS_PCH_SPLIT(dev))
+ if (HAS_PCH_SPLIT(dev_priv))
sr_enabled = I915_READ(WM1_LP_ILK) & WM1_LP_SR_EN;
- else if (IS_CRESTLINE(dev) || IS_G4X(dev) ||
- IS_I945G(dev) || IS_I945GM(dev))
+ else if (IS_CRESTLINE(dev_priv) || IS_G4X(dev_priv) ||
+ IS_I945G(dev_priv) || IS_I945GM(dev_priv))
sr_enabled = I915_READ(FW_BLC_SELF) & FW_BLC_SELF_EN;
- else if (IS_I915GM(dev))
+ else if (IS_I915GM(dev_priv))
sr_enabled = I915_READ(INSTPM) & INSTPM_SELF_EN;
- else if (IS_PINEVIEW(dev))
+ else if (IS_PINEVIEW(dev_priv))
sr_enabled = I915_READ(DSPFW3) & PINEVIEW_SELF_REFRESH_EN;
- else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
+ else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
sr_enabled = I915_READ(FW_BLC_SELF_VLV) & FW_CSPWRDWNEN;
intel_runtime_pm_put(dev_priv);
static int i915_emon_status(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
unsigned long temp, chipset, gfx;
int ret;
- if (!IS_GEN5(dev))
+ if (!IS_GEN5(dev_priv))
return -ENODEV;
ret = mutex_lock_interruptible(&dev->struct_mutex);
static int i915_ring_freq_table(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
int ret = 0;
int gpu_freq, ia_freq;
unsigned int max_gpu_freq, min_gpu_freq;
- if (!HAS_CORE_RING_FREQ(dev)) {
+ if (!HAS_LLC(dev_priv)) {
seq_puts(m, "unsupported on this chipset\n");
return 0;
}
if (ret)
goto out;
- if (IS_SKYLAKE(dev) || IS_KABYLAKE(dev)) {
+ if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) {
/* Convert GT frequency to 50 HZ units */
min_gpu_freq =
dev_priv->rps.min_freq_softlimit / GEN9_FREQ_SCALER;
&ia_freq);
seq_printf(m, "%d\t\t%d\t\t\t\t%d\n",
intel_gpu_freq(dev_priv, (gpu_freq *
- (IS_SKYLAKE(dev) || IS_KABYLAKE(dev) ?
+ (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv) ?
GEN9_FREQ_SCALER : 1))),
((ia_freq >> 0) & 0xff) * 100,
((ia_freq >> 8) & 0xff) * 100);
static int i915_opregion(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_opregion *opregion = &dev_priv->opregion;
int ret;
static int i915_vbt(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_opregion *opregion = &dev_priv->opregion;
+ struct intel_opregion *opregion = &node_to_i915(m->private)->opregion;
if (opregion->vbt)
seq_write(m, opregion->vbt, opregion->vbt_size);
static int i915_gem_framebuffer_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_framebuffer *fbdev_fb = NULL;
struct drm_framebuffer *drm_fb;
int ret;
return ret;
#ifdef CONFIG_DRM_FBDEV_EMULATION
- if (to_i915(dev)->fbdev) {
- fbdev_fb = to_intel_framebuffer(to_i915(dev)->fbdev->helper.fb);
+ if (dev_priv->fbdev) {
+ fbdev_fb = to_intel_framebuffer(dev_priv->fbdev->helper.fb);
seq_printf(m, "fbcon size: %d x %d, depth %d, %d bpp, modifier 0x%llx, refcount %d, obj ",
fbdev_fb->base.width,
static int i915_context_status(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_engine_cs *engine;
struct i915_gem_context *ctx;
int ret;
list_for_each_entry(ctx, &dev_priv->context_list, link) {
seq_printf(m, "HW context %u ", ctx->hw_id);
- if (IS_ERR(ctx->file_priv)) {
- seq_puts(m, "(deleted) ");
- } else if (ctx->file_priv) {
- struct pid *pid = ctx->file_priv->file->pid;
+ if (ctx->pid) {
struct task_struct *task;
- task = get_pid_task(pid, PIDTYPE_PID);
+ task = get_pid_task(ctx->pid, PIDTYPE_PID);
if (task) {
seq_printf(m, "(%s [%d]) ",
task->comm, task->pid);
put_task_struct(task);
}
+ } else if (IS_ERR(ctx->file_priv)) {
+ seq_puts(m, "(deleted) ");
} else {
seq_puts(m, "(kernel) ");
}
seq_printf(m, "%s: ", engine->name);
seq_putc(m, ce->initialised ? 'I' : 'i');
if (ce->state)
- describe_obj(m, ce->state);
+ describe_obj(m, ce->state->obj);
if (ce->ring)
describe_ctx_ring(m, ce->ring);
seq_putc(m, '\n');
struct i915_gem_context *ctx,
struct intel_engine_cs *engine)
{
- struct drm_i915_gem_object *ctx_obj = ctx->engine[engine->id].state;
+ struct i915_vma *vma = ctx->engine[engine->id].state;
struct page *page;
- uint32_t *reg_state;
int j;
- unsigned long ggtt_offset = 0;
seq_printf(m, "CONTEXT: %s %u\n", engine->name, ctx->hw_id);
- if (ctx_obj == NULL) {
- seq_puts(m, "\tNot allocated\n");
+ if (!vma) {
+ seq_puts(m, "\tFake context\n");
return;
}
- if (!i915_gem_obj_ggtt_bound(ctx_obj))
- seq_puts(m, "\tNot bound in GGTT\n");
- else
- ggtt_offset = i915_gem_obj_ggtt_offset(ctx_obj);
+ if (vma->flags & I915_VMA_GLOBAL_BIND)
+ seq_printf(m, "\tBound in GGTT at 0x%08x\n",
+ i915_ggtt_offset(vma));
- if (i915_gem_object_get_pages(ctx_obj)) {
- seq_puts(m, "\tFailed to get pages for context object\n");
+ if (i915_gem_object_get_pages(vma->obj)) {
+ seq_puts(m, "\tFailed to get pages for context object\n\n");
return;
}
- page = i915_gem_object_get_page(ctx_obj, LRC_STATE_PN);
- if (!WARN_ON(page == NULL)) {
- reg_state = kmap_atomic(page);
+ page = i915_gem_object_get_page(vma->obj, LRC_STATE_PN);
+ if (page) {
+ u32 *reg_state = kmap_atomic(page);
for (j = 0; j < 0x600 / sizeof(u32) / 4; j += 4) {
- seq_printf(m, "\t[0x%08lx] 0x%08x 0x%08x 0x%08x 0x%08x\n",
- ggtt_offset + 4096 + (j * 4),
+ seq_printf(m,
+ "\t[0x%04x] 0x%08x 0x%08x 0x%08x 0x%08x\n",
+ j * 4,
reg_state[j], reg_state[j + 1],
reg_state[j + 2], reg_state[j + 3]);
}
static int i915_dump_lrc(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = (struct drm_info_node *) m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_engine_cs *engine;
struct i915_gem_context *ctx;
int ret;
static int i915_execlists(struct seq_file *m, void *data)
{
- struct drm_info_node *node = (struct drm_info_node *)m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_engine_cs *engine;
u32 status_pointer;
u8 read_pointer;
status_pointer = I915_READ(RING_CONTEXT_STATUS_PTR(engine));
seq_printf(m, "\tStatus pointer: 0x%08X\n", status_pointer);
- read_pointer = engine->next_context_status_buffer;
+ read_pointer = GEN8_CSB_READ_PTR(status_pointer);
write_pointer = GEN8_CSB_WRITE_PTR(status_pointer);
if (read_pointer > write_pointer)
write_pointer += GEN8_CSB_ENTRIES;
static int i915_swizzle_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
int ret;
ret = mutex_lock_interruptible(&dev->struct_mutex);
seq_printf(m, "bit6 swizzle for Y-tiling = %s\n",
swizzle_string(dev_priv->mm.bit_6_swizzle_y));
- if (IS_GEN3(dev) || IS_GEN4(dev)) {
+ if (IS_GEN3(dev_priv) || IS_GEN4(dev_priv)) {
seq_printf(m, "DDC = 0x%08x\n",
I915_READ(DCC));
seq_printf(m, "DDC2 = 0x%08x\n",
I915_READ16(C0DRB3));
seq_printf(m, "C1DRB3 = 0x%04x\n",
I915_READ16(C1DRB3));
- } else if (INTEL_INFO(dev)->gen >= 6) {
+ } else if (INTEL_GEN(dev_priv) >= 6) {
seq_printf(m, "MAD_DIMM_C0 = 0x%08x\n",
I915_READ(MAD_DIMM_C0));
seq_printf(m, "MAD_DIMM_C1 = 0x%08x\n",
I915_READ(MAD_DIMM_C2));
seq_printf(m, "TILECTL = 0x%08x\n",
I915_READ(TILECTL));
- if (INTEL_INFO(dev)->gen >= 8)
+ if (INTEL_GEN(dev_priv) >= 8)
seq_printf(m, "GAMTARBMODE = 0x%08x\n",
I915_READ(GAMTARBMODE));
else
return 0;
}
-static void gen8_ppgtt_info(struct seq_file *m, struct drm_device *dev)
+static void gen8_ppgtt_info(struct seq_file *m,
+ struct drm_i915_private *dev_priv)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_engine_cs *engine;
struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
int i;
}
}
-static void gen6_ppgtt_info(struct seq_file *m, struct drm_device *dev)
+static void gen6_ppgtt_info(struct seq_file *m,
+ struct drm_i915_private *dev_priv)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_engine_cs *engine;
if (IS_GEN6(dev_priv))
static int i915_ppgtt_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct drm_file *file;
+ int ret;
- int ret = mutex_lock_interruptible(&dev->struct_mutex);
+ mutex_lock(&dev->filelist_mutex);
+ ret = mutex_lock_interruptible(&dev->struct_mutex);
if (ret)
- return ret;
+ goto out_unlock;
+
intel_runtime_pm_get(dev_priv);
- if (INTEL_INFO(dev)->gen >= 8)
- gen8_ppgtt_info(m, dev);
- else if (INTEL_INFO(dev)->gen >= 6)
- gen6_ppgtt_info(m, dev);
+ if (INTEL_GEN(dev_priv) >= 8)
+ gen8_ppgtt_info(m, dev_priv);
+ else if (INTEL_GEN(dev_priv) >= 6)
+ gen6_ppgtt_info(m, dev_priv);
- mutex_lock(&dev->filelist_mutex);
list_for_each_entry_reverse(file, &dev->filelist, lhead) {
struct drm_i915_file_private *file_priv = file->driver_priv;
struct task_struct *task;
task = get_pid_task(file->pid, PIDTYPE_PID);
if (!task) {
ret = -ESRCH;
- goto out_unlock;
+ goto out_rpm;
}
seq_printf(m, "\nproc: %s\n", task->comm);
put_task_struct(task);
idr_for_each(&file_priv->context_idr, per_file_ctx,
(void *)(unsigned long)m);
}
-out_unlock:
- mutex_unlock(&dev->filelist_mutex);
+out_rpm:
intel_runtime_pm_put(dev_priv);
mutex_unlock(&dev->struct_mutex);
-
+out_unlock:
+ mutex_unlock(&dev->filelist_mutex);
return ret;
}
return count;
}
+static const char *rps_power_to_str(unsigned int power)
+{
+ static const char * const strings[] = {
+ [LOW_POWER] = "low power",
+ [BETWEEN] = "mixed",
+ [HIGH_POWER] = "high power",
+ };
+
+ if (power >= ARRAY_SIZE(strings) || !strings[power])
+ return "unknown";
+
+ return strings[power];
+}
+
static int i915_rps_boost_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct drm_file *file;
seq_printf(m, "RPS enabled? %d\n", dev_priv->rps.enabled);
seq_printf(m, "GPU busy? %s [%x]\n",
yesno(dev_priv->gt.awake), dev_priv->gt.active_engines);
seq_printf(m, "CPU waiting? %d\n", count_irq_waiters(dev_priv));
- seq_printf(m, "Frequency requested %d; min hard:%d, soft:%d; max soft:%d, hard:%d\n",
- intel_gpu_freq(dev_priv, dev_priv->rps.cur_freq),
+ seq_printf(m, "Frequency requested %d\n",
+ intel_gpu_freq(dev_priv, dev_priv->rps.cur_freq));
+ seq_printf(m, " min hard:%d, soft:%d; max soft:%d, hard:%d\n",
intel_gpu_freq(dev_priv, dev_priv->rps.min_freq),
intel_gpu_freq(dev_priv, dev_priv->rps.min_freq_softlimit),
intel_gpu_freq(dev_priv, dev_priv->rps.max_freq_softlimit),
intel_gpu_freq(dev_priv, dev_priv->rps.max_freq));
+ seq_printf(m, " idle:%d, efficient:%d, boost:%d\n",
+ intel_gpu_freq(dev_priv, dev_priv->rps.idle_freq),
+ intel_gpu_freq(dev_priv, dev_priv->rps.efficient_freq),
+ intel_gpu_freq(dev_priv, dev_priv->rps.boost_freq));
mutex_lock(&dev->filelist_mutex);
spin_lock(&dev_priv->rps.client_lock);
spin_unlock(&dev_priv->rps.client_lock);
mutex_unlock(&dev->filelist_mutex);
+ if (INTEL_GEN(dev_priv) >= 6 &&
+ dev_priv->rps.enabled &&
+ dev_priv->gt.active_engines) {
+ u32 rpup, rpupei;
+ u32 rpdown, rpdownei;
+
+ intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
+ rpup = I915_READ_FW(GEN6_RP_CUR_UP) & GEN6_RP_EI_MASK;
+ rpupei = I915_READ_FW(GEN6_RP_CUR_UP_EI) & GEN6_RP_EI_MASK;
+ rpdown = I915_READ_FW(GEN6_RP_CUR_DOWN) & GEN6_RP_EI_MASK;
+ rpdownei = I915_READ_FW(GEN6_RP_CUR_DOWN_EI) & GEN6_RP_EI_MASK;
+ intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
+
+ seq_printf(m, "\nRPS Autotuning (current \"%s\" window):\n",
+ rps_power_to_str(dev_priv->rps.power));
+ seq_printf(m, " Avg. up: %d%% [above threshold? %d%%]\n",
+ 100 * rpup / rpupei,
+ dev_priv->rps.up_threshold);
+ seq_printf(m, " Avg. down: %d%% [below threshold? %d%%]\n",
+ 100 * rpdown / rpdownei,
+ dev_priv->rps.down_threshold);
+ } else {
+ seq_puts(m, "\nRPS Autotuning inactive\n");
+ }
+
return 0;
}
static int i915_llc(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
const bool edram = INTEL_GEN(dev_priv) > 8;
- seq_printf(m, "LLC: %s\n", yesno(HAS_LLC(dev)));
+ seq_printf(m, "LLC: %s\n", yesno(HAS_LLC(dev_priv)));
seq_printf(m, "%s: %lluMB\n", edram ? "eDRAM" : "eLLC",
intel_uncore_edram_size(dev_priv)/1024/1024);
static int i915_guc_load_status_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_i915_private *dev_priv = to_i915(node->minor->dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
u32 tmp, i;
struct i915_guc_client *client)
{
struct intel_engine_cs *engine;
+ enum intel_engine_id id;
uint64_t tot = 0;
seq_printf(m, "\tPriority %d, GuC ctx index: %u, PD offset 0x%x\n",
client->wq_size, client->wq_offset, client->wq_tail);
seq_printf(m, "\tWork queue full: %u\n", client->no_wq_space);
- seq_printf(m, "\tFailed to queue: %u\n", client->q_fail);
seq_printf(m, "\tFailed doorbell: %u\n", client->b_fail);
seq_printf(m, "\tLast submission result: %d\n", client->retcode);
- for_each_engine(engine, dev_priv) {
+ for_each_engine_id(engine, dev_priv, id) {
+ u64 submissions = client->submissions[id];
+ tot += submissions;
seq_printf(m, "\tSubmissions: %llu %s\n",
- client->submissions[engine->id],
- engine->name);
- tot += client->submissions[engine->id];
+ submissions, engine->name);
}
seq_printf(m, "\tTotal: %llu\n", tot);
}
static int i915_guc_info(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_guc guc;
struct i915_guc_client client = {};
struct intel_engine_cs *engine;
+ enum intel_engine_id id;
u64 total = 0;
if (!HAS_GUC_SCHED(dev_priv))
seq_printf(m, "GuC last action error code: %d\n", guc.action_err);
seq_printf(m, "\nGuC submissions:\n");
- for_each_engine(engine, dev_priv) {
+ for_each_engine_id(engine, dev_priv, id) {
+ u64 submissions = guc.submissions[id];
+ total += submissions;
seq_printf(m, "\t%-24s: %10llu, last seqno 0x%08x\n",
- engine->name, guc.submissions[engine->id],
- guc.last_seqno[engine->id]);
- total += guc.submissions[engine->id];
+ engine->name, submissions, guc.last_seqno[id]);
}
seq_printf(m, "\t%s: %llu\n", "Total", total);
static int i915_guc_log_dump(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
- struct drm_i915_gem_object *log_obj = dev_priv->guc.log_obj;
- u32 *log;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_i915_gem_object *obj;
int i = 0, pg;
- if (!log_obj)
+ if (!dev_priv->guc.log_vma)
return 0;
- for (pg = 0; pg < log_obj->base.size / PAGE_SIZE; pg++) {
- log = kmap_atomic(i915_gem_object_get_page(log_obj, pg));
+ obj = dev_priv->guc.log_vma->obj;
+ for (pg = 0; pg < obj->base.size / PAGE_SIZE; pg++) {
+ u32 *log = kmap_atomic(i915_gem_object_get_page(obj, pg));
for (i = 0; i < PAGE_SIZE / sizeof(u32); i += 4)
seq_printf(m, "0x%08x 0x%08x 0x%08x 0x%08x\n",
static int i915_edp_psr_status(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
u32 psrperf = 0;
u32 stat[3];
enum pipe pipe;
bool enabled = false;
- if (!HAS_PSR(dev)) {
+ if (!HAS_PSR(dev_priv)) {
seq_puts(m, "PSR not supported\n");
return 0;
}
seq_printf(m, "Re-enable work scheduled: %s\n",
yesno(work_busy(&dev_priv->psr.work.work)));
- if (HAS_DDI(dev))
+ if (HAS_DDI(dev_priv))
enabled = I915_READ(EDP_PSR_CTL) & EDP_PSR_ENABLE;
else {
for_each_pipe(dev_priv, pipe) {
seq_printf(m, "HW Enabled & Active bit: %s", yesno(enabled));
- if (!HAS_DDI(dev))
+ if (!HAS_DDI(dev_priv))
for_each_pipe(dev_priv, pipe) {
if ((stat[pipe] == VLV_EDP_PSR_ACTIVE_NORFB_UP) ||
(stat[pipe] == VLV_EDP_PSR_ACTIVE_SF_UPDATE))
* VLV/CHV PSR has no kind of performance counter
* SKL+ Perf counter is reset to 0 everytime DC state is entered
*/
- if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
+ if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) {
psrperf = I915_READ(EDP_PSR_PERF_CNT) &
EDP_PSR_PERF_CNT_MASK;
static int i915_sink_crc(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_connector *connector;
struct intel_dp *intel_dp = NULL;
int ret;
static int i915_energy_uJ(struct seq_file *m, void *data)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
u64 power;
u32 units;
- if (INTEL_INFO(dev)->gen < 6)
+ if (INTEL_GEN(dev_priv) < 6)
return -ENODEV;
intel_runtime_pm_get(dev_priv);
static int i915_runtime_pm_status(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
if (!HAS_RUNTIME_PM(dev_priv))
seq_puts(m, "Runtime power management not supported\n");
yesno(!intel_irqs_enabled(dev_priv)));
#ifdef CONFIG_PM
seq_printf(m, "Usage count: %d\n",
- atomic_read(&dev->dev->power.usage_count));
+ atomic_read(&dev_priv->drm.dev->power.usage_count));
#else
seq_printf(m, "Device Power Management (CONFIG_PM) disabled\n");
#endif
seq_printf(m, "PCI device power state: %s [%d]\n",
- pci_power_name(dev_priv->drm.pdev->current_state),
- dev_priv->drm.pdev->current_state);
+ pci_power_name(pdev->current_state),
+ pdev->current_state);
return 0;
}
static int i915_power_domain_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
struct i915_power_domains *power_domains = &dev_priv->power_domains;
int i;
static int i915_dmc_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
struct intel_csr *csr;
- if (!HAS_CSR(dev)) {
+ if (!HAS_CSR(dev_priv)) {
seq_puts(m, "not supported\n");
return 0;
}
seq_printf(m, "version: %d.%d\n", CSR_VERSION_MAJOR(csr->version),
CSR_VERSION_MINOR(csr->version));
- if (IS_SKYLAKE(dev) && csr->version >= CSR_VERSION(1, 6)) {
+ if (IS_SKYLAKE(dev_priv) && csr->version >= CSR_VERSION(1, 6)) {
seq_printf(m, "DC3 -> DC5 count: %d\n",
I915_READ(SKL_CSR_DC3_DC5_COUNT));
seq_printf(m, "DC5 -> DC6 count: %d\n",
I915_READ(SKL_CSR_DC5_DC6_COUNT));
- } else if (IS_BROXTON(dev) && csr->version >= CSR_VERSION(1, 4)) {
+ } else if (IS_BROXTON(dev_priv) && csr->version >= CSR_VERSION(1, 4)) {
seq_printf(m, "DC3 -> DC5 count: %d\n",
I915_READ(BXT_CSR_DC3_DC5_COUNT));
}
struct intel_crtc *intel_crtc,
struct intel_encoder *intel_encoder)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct drm_crtc *crtc = &intel_crtc->base;
struct intel_connector *intel_connector;
struct drm_encoder *encoder;
static void intel_crtc_info(struct seq_file *m, struct intel_crtc *intel_crtc)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct drm_crtc *crtc = &intel_crtc->base;
struct intel_encoder *intel_encoder;
struct drm_plane_state *plane_state = crtc->primary->state;
seq_printf(m, "\taudio support: %s\n", yesno(intel_dp->has_audio));
if (intel_connector->base.connector_type == DRM_MODE_CONNECTOR_eDP)
intel_panel_info(m, &intel_connector->panel);
+
+ drm_dp_downstream_debug(m, intel_dp->dpcd, intel_dp->downstream_ports,
+ &intel_dp->aux);
}
static void intel_hdmi_info(struct seq_file *m,
intel_seq_print_mode(m, 2, mode);
}
-static bool cursor_active(struct drm_device *dev, int pipe)
+static bool cursor_active(struct drm_i915_private *dev_priv, int pipe)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
u32 state;
- if (IS_845G(dev) || IS_I865G(dev))
+ if (IS_845G(dev_priv) || IS_I865G(dev_priv))
state = I915_READ(CURCNTR(PIPE_A)) & CURSOR_ENABLE;
else
state = I915_READ(CURCNTR(pipe)) & CURSOR_MODE;
return state;
}
-static bool cursor_position(struct drm_device *dev, int pipe, int *x, int *y)
+static bool cursor_position(struct drm_i915_private *dev_priv,
+ int pipe, int *x, int *y)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
u32 pos;
pos = I915_READ(CURPOS(pipe));
if (pos & (CURSOR_POS_SIGN << CURSOR_Y_SHIFT))
*y = -*y;
- return cursor_active(dev, pipe);
+ return cursor_active(dev_priv, pipe);
}
static const char *plane_type(enum drm_plane_type type)
static void intel_plane_info(struct seq_file *m, struct intel_crtc *intel_crtc)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_plane *intel_plane;
for_each_intel_plane_on_crtc(dev, intel_crtc, intel_plane) {
static int i915_display_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_crtc *crtc;
struct drm_connector *connector;
if (pipe_config->base.active) {
intel_crtc_info(m, crtc);
- active = cursor_position(dev, crtc->pipe, &x, &y);
+ active = cursor_position(dev_priv, crtc->pipe, &x, &y);
seq_printf(m, "\tcursor visible? %s, position (%d, %d), size %dx%d, addr 0x%08x, active? %s\n",
yesno(crtc->cursor_base),
x, y, crtc->base.cursor->state->crtc_w,
static int i915_semaphore_status(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = (struct drm_info_node *) m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_engine_cs *engine;
- int num_rings = hweight32(INTEL_INFO(dev)->ring_mask);
+ int num_rings = INTEL_INFO(dev_priv)->num_rings;
enum intel_engine_id id;
int j, ret;
return ret;
intel_runtime_pm_get(dev_priv);
- if (IS_BROADWELL(dev)) {
+ if (IS_BROADWELL(dev_priv)) {
struct page *page;
uint64_t *seqno;
- page = i915_gem_object_get_page(dev_priv->semaphore_obj, 0);
+ page = i915_gem_object_get_page(dev_priv->semaphore->obj, 0);
seqno = (uint64_t *)kmap_atomic(page);
for_each_engine_id(engine, dev_priv, id) {
static int i915_shared_dplls_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = (struct drm_info_node *) m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
int i;
drm_modeset_lock_all(dev);
int i;
int ret;
struct intel_engine_cs *engine;
- struct drm_info_node *node = (struct drm_info_node *) m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct i915_workarounds *workarounds = &dev_priv->workarounds;
enum intel_engine_id id;
static int i915_ddb_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct skl_ddb_allocation *ddb;
struct skl_ddb_entry *entry;
enum pipe pipe;
int plane;
- if (INTEL_INFO(dev)->gen < 9)
+ if (INTEL_GEN(dev_priv) < 9)
return 0;
drm_modeset_lock_all(dev);
}
static void drrs_status_per_crtc(struct seq_file *m,
- struct drm_device *dev, struct intel_crtc *intel_crtc)
+ struct drm_device *dev,
+ struct intel_crtc *intel_crtc)
{
struct drm_i915_private *dev_priv = to_i915(dev);
struct i915_drrs *drrs = &dev_priv->drrs;
static int i915_drrs_status(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = m->private;
- struct drm_device *dev = node->minor->dev;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_crtc *intel_crtc;
int active_crtc_cnt = 0;
struct pipe_crc_info {
const char *name;
- struct drm_device *dev;
+ struct drm_i915_private *dev_priv;
enum pipe pipe;
};
static int i915_dp_mst_info(struct seq_file *m, void *unused)
{
- struct drm_info_node *node = (struct drm_info_node *) m->private;
- struct drm_device *dev = node->minor->dev;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_encoder *intel_encoder;
struct intel_digital_port *intel_dig_port;
struct drm_connector *connector;
static int i915_pipe_crc_open(struct inode *inode, struct file *filep)
{
struct pipe_crc_info *info = inode->i_private;
- struct drm_i915_private *dev_priv = to_i915(info->dev);
+ struct drm_i915_private *dev_priv = info->dev_priv;
struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[info->pipe];
- if (info->pipe >= INTEL_INFO(info->dev)->num_pipes)
+ if (info->pipe >= INTEL_INFO(dev_priv)->num_pipes)
return -ENODEV;
spin_lock_irq(&pipe_crc->lock);
static int i915_pipe_crc_release(struct inode *inode, struct file *filep)
{
struct pipe_crc_info *info = inode->i_private;
- struct drm_i915_private *dev_priv = to_i915(info->dev);
+ struct drm_i915_private *dev_priv = info->dev_priv;
struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[info->pipe];
spin_lock_irq(&pipe_crc->lock);
loff_t *pos)
{
struct pipe_crc_info *info = filep->private_data;
- struct drm_device *dev = info->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = info->dev_priv;
struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[info->pipe];
char buf[PIPE_CRC_BUFFER_LEN];
int n_entries;
static int i915_pipe_crc_create(struct dentry *root, struct drm_minor *minor,
enum pipe pipe)
{
- struct drm_device *dev = minor->dev;
+ struct drm_i915_private *dev_priv = to_i915(minor->dev);
struct dentry *ent;
struct pipe_crc_info *info = &i915_pipe_crc_data[pipe];
- info->dev = dev;
+ info->dev_priv = dev_priv;
ent = debugfs_create_file(info->name, S_IRUGO, root, info,
&i915_pipe_crc_fops);
if (!ent)
static int display_crc_ctl_show(struct seq_file *m, void *data)
{
- struct drm_device *dev = m->private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = m->private;
int i;
for (i = 0; i < I915_MAX_PIPES; i++)
static int display_crc_ctl_open(struct inode *inode, struct file *file)
{
- struct drm_device *dev = inode->i_private;
-
- return single_open(file, display_crc_ctl_show, dev);
+ return single_open(file, display_crc_ctl_show, inode->i_private);
}
static int i8xx_pipe_crc_ctl_reg(enum intel_pipe_crc_source *source,
return 0;
}
-static int i9xx_pipe_crc_auto_source(struct drm_device *dev, enum pipe pipe,
+static int i9xx_pipe_crc_auto_source(struct drm_i915_private *dev_priv,
+ enum pipe pipe,
enum intel_pipe_crc_source *source)
{
+ struct drm_device *dev = &dev_priv->drm;
struct intel_encoder *encoder;
struct intel_crtc *crtc;
struct intel_digital_port *dig_port;
return ret;
}
-static int vlv_pipe_crc_ctl_reg(struct drm_device *dev,
+static int vlv_pipe_crc_ctl_reg(struct drm_i915_private *dev_priv,
enum pipe pipe,
enum intel_pipe_crc_source *source,
uint32_t *val)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
bool need_stable_symbols = false;
if (*source == INTEL_PIPE_CRC_SOURCE_AUTO) {
- int ret = i9xx_pipe_crc_auto_source(dev, pipe, source);
+ int ret = i9xx_pipe_crc_auto_source(dev_priv, pipe, source);
if (ret)
return ret;
}
need_stable_symbols = true;
break;
case INTEL_PIPE_CRC_SOURCE_DP_D:
- if (!IS_CHERRYVIEW(dev))
+ if (!IS_CHERRYVIEW(dev_priv))
return -EINVAL;
*val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_DP_D_VLV;
need_stable_symbols = true;
return 0;
}
-static int i9xx_pipe_crc_ctl_reg(struct drm_device *dev,
+static int i9xx_pipe_crc_ctl_reg(struct drm_i915_private *dev_priv,
enum pipe pipe,
enum intel_pipe_crc_source *source,
uint32_t *val)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
bool need_stable_symbols = false;
if (*source == INTEL_PIPE_CRC_SOURCE_AUTO) {
- int ret = i9xx_pipe_crc_auto_source(dev, pipe, source);
+ int ret = i9xx_pipe_crc_auto_source(dev_priv, pipe, source);
if (ret)
return ret;
}
*val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_PIPE_I9XX;
break;
case INTEL_PIPE_CRC_SOURCE_TV:
- if (!SUPPORTS_TV(dev))
+ if (!SUPPORTS_TV(dev_priv))
return -EINVAL;
*val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_TV_PRE;
break;
case INTEL_PIPE_CRC_SOURCE_DP_B:
- if (!IS_G4X(dev))
+ if (!IS_G4X(dev_priv))
return -EINVAL;
*val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_DP_B_G4X;
need_stable_symbols = true;
break;
case INTEL_PIPE_CRC_SOURCE_DP_C:
- if (!IS_G4X(dev))
+ if (!IS_G4X(dev_priv))
return -EINVAL;
*val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_DP_C_G4X;
need_stable_symbols = true;
break;
case INTEL_PIPE_CRC_SOURCE_DP_D:
- if (!IS_G4X(dev))
+ if (!IS_G4X(dev_priv))
return -EINVAL;
*val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_DP_D_G4X;
need_stable_symbols = true;
if (need_stable_symbols) {
uint32_t tmp = I915_READ(PORT_DFT2_G4X);
- WARN_ON(!IS_G4X(dev));
+ WARN_ON(!IS_G4X(dev_priv));
I915_WRITE(PORT_DFT_I9XX,
I915_READ(PORT_DFT_I9XX) | DC_BALANCE_RESET);
return 0;
}
-static void vlv_undo_pipe_scramble_reset(struct drm_device *dev,
+static void vlv_undo_pipe_scramble_reset(struct drm_i915_private *dev_priv,
enum pipe pipe)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
uint32_t tmp = I915_READ(PORT_DFT2_G4X);
switch (pipe) {
}
-static void g4x_undo_pipe_scramble_reset(struct drm_device *dev,
+static void g4x_undo_pipe_scramble_reset(struct drm_i915_private *dev_priv,
enum pipe pipe)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
uint32_t tmp = I915_READ(PORT_DFT2_G4X);
if (pipe == PIPE_A)
return 0;
}
-static void hsw_trans_edp_pipe_A_crc_wa(struct drm_device *dev, bool enable)
+static void hsw_trans_edp_pipe_A_crc_wa(struct drm_i915_private *dev_priv,
+ bool enable)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_crtc *crtc =
to_intel_crtc(dev_priv->pipe_to_crtc_mapping[PIPE_A]);
struct intel_crtc_state *pipe_config;
drm_atomic_state_free(state);
}
-static int ivb_pipe_crc_ctl_reg(struct drm_device *dev,
+static int ivb_pipe_crc_ctl_reg(struct drm_i915_private *dev_priv,
enum pipe pipe,
enum intel_pipe_crc_source *source,
uint32_t *val)
*val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_SPRITE_IVB;
break;
case INTEL_PIPE_CRC_SOURCE_PF:
- if (IS_HASWELL(dev) && pipe == PIPE_A)
- hsw_trans_edp_pipe_A_crc_wa(dev, true);
+ if (IS_HASWELL(dev_priv) && pipe == PIPE_A)
+ hsw_trans_edp_pipe_A_crc_wa(dev_priv, true);
*val = PIPE_CRC_ENABLE | PIPE_CRC_SOURCE_PF_IVB;
break;
return 0;
}
-static int pipe_crc_set_source(struct drm_device *dev, enum pipe pipe,
+static int pipe_crc_set_source(struct drm_i915_private *dev_priv,
+ enum pipe pipe,
enum intel_pipe_crc_source source)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_device *dev = &dev_priv->drm;
struct intel_pipe_crc *pipe_crc = &dev_priv->pipe_crc[pipe];
- struct intel_crtc *crtc = to_intel_crtc(intel_get_crtc_for_pipe(dev,
- pipe));
+ struct intel_crtc *crtc =
+ to_intel_crtc(intel_get_crtc_for_pipe(dev, pipe));
enum intel_display_power_domain power_domain;
u32 val = 0; /* shut up gcc */
int ret;
return -EIO;
}
- if (IS_GEN2(dev))
+ if (IS_GEN2(dev_priv))
ret = i8xx_pipe_crc_ctl_reg(&source, &val);
- else if (INTEL_INFO(dev)->gen < 5)
- ret = i9xx_pipe_crc_ctl_reg(dev, pipe, &source, &val);
- else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
- ret = vlv_pipe_crc_ctl_reg(dev, pipe, &source, &val);
- else if (IS_GEN5(dev) || IS_GEN6(dev))
+ else if (INTEL_GEN(dev_priv) < 5)
+ ret = i9xx_pipe_crc_ctl_reg(dev_priv, pipe, &source, &val);
+ else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+ ret = vlv_pipe_crc_ctl_reg(dev_priv, pipe, &source, &val);
+ else if (IS_GEN5(dev_priv) || IS_GEN6(dev_priv))
ret = ilk_pipe_crc_ctl_reg(&source, &val);
else
- ret = ivb_pipe_crc_ctl_reg(dev, pipe, &source, &val);
+ ret = ivb_pipe_crc_ctl_reg(dev_priv, pipe, &source, &val);
if (ret != 0)
goto out;
kfree(entries);
- if (IS_G4X(dev))
- g4x_undo_pipe_scramble_reset(dev, pipe);
- else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
- vlv_undo_pipe_scramble_reset(dev, pipe);
- else if (IS_HASWELL(dev) && pipe == PIPE_A)
- hsw_trans_edp_pipe_A_crc_wa(dev, false);
+ if (IS_G4X(dev_priv))
+ g4x_undo_pipe_scramble_reset(dev_priv, pipe);
+ else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+ vlv_undo_pipe_scramble_reset(dev_priv, pipe);
+ else if (IS_HASWELL(dev_priv) && pipe == PIPE_A)
+ hsw_trans_edp_pipe_A_crc_wa(dev_priv, false);
hsw_enable_ips(crtc);
}
return -EINVAL;
}
-static int display_crc_ctl_parse(struct drm_device *dev, char *buf, size_t len)
+static int display_crc_ctl_parse(struct drm_i915_private *dev_priv,
+ char *buf, size_t len)
{
#define N_WORDS 3
int n_words;
return -EINVAL;
}
- return pipe_crc_set_source(dev, pipe, source);
+ return pipe_crc_set_source(dev_priv, pipe, source);
}
static ssize_t display_crc_ctl_write(struct file *file, const char __user *ubuf,
size_t len, loff_t *offp)
{
struct seq_file *m = file->private_data;
- struct drm_device *dev = m->private;
+ struct drm_i915_private *dev_priv = m->private;
char *tmpbuf;
int ret;
}
tmpbuf[len] = '\0';
- ret = display_crc_ctl_parse(dev, tmpbuf, len);
+ ret = display_crc_ctl_parse(dev_priv, tmpbuf, len);
out:
kfree(tmpbuf);
};
static ssize_t i915_displayport_test_active_write(struct file *file,
- const char __user *ubuf,
- size_t len, loff_t *offp)
+ const char __user *ubuf,
+ size_t len, loff_t *offp)
{
char *input_buffer;
int status = 0;
DRM_DEBUG_DRIVER("Copied %d bytes from user\n", (unsigned int)len);
list_for_each_entry(connector, connector_list, head) {
-
if (connector->connector_type !=
DRM_MODE_CONNECTOR_DisplayPort)
continue;
struct intel_dp *intel_dp;
list_for_each_entry(connector, connector_list, head) {
-
if (connector->connector_type !=
DRM_MODE_CONNECTOR_DisplayPort)
continue;
}
static int i915_displayport_test_active_open(struct inode *inode,
- struct file *file)
+ struct file *file)
{
- struct drm_device *dev = inode->i_private;
+ struct drm_i915_private *dev_priv = inode->i_private;
- return single_open(file, i915_displayport_test_active_show, dev);
+ return single_open(file, i915_displayport_test_active_show,
+ &dev_priv->drm);
}
static const struct file_operations i915_displayport_test_active_fops = {
struct intel_dp *intel_dp;
list_for_each_entry(connector, connector_list, head) {
-
if (connector->connector_type !=
DRM_MODE_CONNECTOR_DisplayPort)
continue;
return 0;
}
static int i915_displayport_test_data_open(struct inode *inode,
- struct file *file)
+ struct file *file)
{
- struct drm_device *dev = inode->i_private;
+ struct drm_i915_private *dev_priv = inode->i_private;
- return single_open(file, i915_displayport_test_data_show, dev);
+ return single_open(file, i915_displayport_test_data_show,
+ &dev_priv->drm);
}
static const struct file_operations i915_displayport_test_data_fops = {
struct intel_dp *intel_dp;
list_for_each_entry(connector, connector_list, head) {
-
if (connector->connector_type !=
DRM_MODE_CONNECTOR_DisplayPort)
continue;
static int i915_displayport_test_type_open(struct inode *inode,
struct file *file)
{
- struct drm_device *dev = inode->i_private;
+ struct drm_i915_private *dev_priv = inode->i_private;
- return single_open(file, i915_displayport_test_type_show, dev);
+ return single_open(file, i915_displayport_test_type_show,
+ &dev_priv->drm);
}
static const struct file_operations i915_displayport_test_type_fops = {
static void wm_latency_show(struct seq_file *m, const uint16_t wm[8])
{
- struct drm_device *dev = m->private;
+ struct drm_i915_private *dev_priv = m->private;
+ struct drm_device *dev = &dev_priv->drm;
int level;
int num_levels;
- if (IS_CHERRYVIEW(dev))
+ if (IS_CHERRYVIEW(dev_priv))
num_levels = 3;
- else if (IS_VALLEYVIEW(dev))
+ else if (IS_VALLEYVIEW(dev_priv))
num_levels = 1;
else
num_levels = ilk_wm_max_level(dev) + 1;
* - WM1+ latency values in 0.5us units
* - latencies are in us on gen9/vlv/chv
*/
- if (INTEL_INFO(dev)->gen >= 9 || IS_VALLEYVIEW(dev) ||
- IS_CHERRYVIEW(dev))
+ if (INTEL_GEN(dev_priv) >= 9 || IS_VALLEYVIEW(dev_priv) ||
+ IS_CHERRYVIEW(dev_priv))
latency *= 10;
else if (level > 0)
latency *= 5;
static int pri_wm_latency_show(struct seq_file *m, void *data)
{
- struct drm_device *dev = m->private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = m->private;
const uint16_t *latencies;
- if (INTEL_INFO(dev)->gen >= 9)
+ if (INTEL_GEN(dev_priv) >= 9)
latencies = dev_priv->wm.skl_latency;
else
- latencies = to_i915(dev)->wm.pri_latency;
+ latencies = dev_priv->wm.pri_latency;
wm_latency_show(m, latencies);
static int spr_wm_latency_show(struct seq_file *m, void *data)
{
- struct drm_device *dev = m->private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = m->private;
const uint16_t *latencies;
- if (INTEL_INFO(dev)->gen >= 9)
+ if (INTEL_GEN(dev_priv) >= 9)
latencies = dev_priv->wm.skl_latency;
else
- latencies = to_i915(dev)->wm.spr_latency;
+ latencies = dev_priv->wm.spr_latency;
wm_latency_show(m, latencies);
static int cur_wm_latency_show(struct seq_file *m, void *data)
{
- struct drm_device *dev = m->private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = m->private;
const uint16_t *latencies;
- if (INTEL_INFO(dev)->gen >= 9)
+ if (INTEL_GEN(dev_priv) >= 9)
latencies = dev_priv->wm.skl_latency;
else
- latencies = to_i915(dev)->wm.cur_latency;
+ latencies = dev_priv->wm.cur_latency;
wm_latency_show(m, latencies);
static int pri_wm_latency_open(struct inode *inode, struct file *file)
{
- struct drm_device *dev = inode->i_private;
+ struct drm_i915_private *dev_priv = inode->i_private;
- if (INTEL_INFO(dev)->gen < 5)
+ if (INTEL_GEN(dev_priv) < 5)
return -ENODEV;
- return single_open(file, pri_wm_latency_show, dev);
+ return single_open(file, pri_wm_latency_show, dev_priv);
}
static int spr_wm_latency_open(struct inode *inode, struct file *file)
{
- struct drm_device *dev = inode->i_private;
+ struct drm_i915_private *dev_priv = inode->i_private;
- if (HAS_GMCH_DISPLAY(dev))
+ if (HAS_GMCH_DISPLAY(dev_priv))
return -ENODEV;
- return single_open(file, spr_wm_latency_show, dev);
+ return single_open(file, spr_wm_latency_show, dev_priv);
}
static int cur_wm_latency_open(struct inode *inode, struct file *file)
{
- struct drm_device *dev = inode->i_private;
+ struct drm_i915_private *dev_priv = inode->i_private;
- if (HAS_GMCH_DISPLAY(dev))
+ if (HAS_GMCH_DISPLAY(dev_priv))
return -ENODEV;
- return single_open(file, cur_wm_latency_show, dev);
+ return single_open(file, cur_wm_latency_show, dev_priv);
}
static ssize_t wm_latency_write(struct file *file, const char __user *ubuf,
size_t len, loff_t *offp, uint16_t wm[8])
{
struct seq_file *m = file->private_data;
- struct drm_device *dev = m->private;
+ struct drm_i915_private *dev_priv = m->private;
+ struct drm_device *dev = &dev_priv->drm;
uint16_t new[8] = { 0 };
int num_levels;
int level;
int ret;
char tmp[32];
- if (IS_CHERRYVIEW(dev))
+ if (IS_CHERRYVIEW(dev_priv))
num_levels = 3;
- else if (IS_VALLEYVIEW(dev))
+ else if (IS_VALLEYVIEW(dev_priv))
num_levels = 1;
else
num_levels = ilk_wm_max_level(dev) + 1;
size_t len, loff_t *offp)
{
struct seq_file *m = file->private_data;
- struct drm_device *dev = m->private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = m->private;
uint16_t *latencies;
- if (INTEL_INFO(dev)->gen >= 9)
+ if (INTEL_GEN(dev_priv) >= 9)
latencies = dev_priv->wm.skl_latency;
else
- latencies = to_i915(dev)->wm.pri_latency;
+ latencies = dev_priv->wm.pri_latency;
return wm_latency_write(file, ubuf, len, offp, latencies);
}
size_t len, loff_t *offp)
{
struct seq_file *m = file->private_data;
- struct drm_device *dev = m->private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = m->private;
uint16_t *latencies;
- if (INTEL_INFO(dev)->gen >= 9)
+ if (INTEL_GEN(dev_priv) >= 9)
latencies = dev_priv->wm.skl_latency;
else
- latencies = to_i915(dev)->wm.spr_latency;
+ latencies = dev_priv->wm.spr_latency;
return wm_latency_write(file, ubuf, len, offp, latencies);
}
size_t len, loff_t *offp)
{
struct seq_file *m = file->private_data;
- struct drm_device *dev = m->private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = m->private;
uint16_t *latencies;
- if (INTEL_INFO(dev)->gen >= 9)
+ if (INTEL_GEN(dev_priv) >= 9)
latencies = dev_priv->wm.skl_latency;
else
- latencies = to_i915(dev)->wm.cur_latency;
+ latencies = dev_priv->wm.cur_latency;
return wm_latency_write(file, ubuf, len, offp, latencies);
}
static int
i915_wedged_get(void *data, u64 *val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
*val = i915_terminally_wedged(&dev_priv->gpu_error);
static int
i915_wedged_set(void *data, u64 val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
/*
* There is no safeguard against this debugfs entry colliding
static int
i915_ring_missed_irq_get(void *data, u64 *val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
*val = dev_priv->gpu_error.missed_irq_rings;
return 0;
static int
i915_ring_missed_irq_set(void *data, u64 val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
+ struct drm_device *dev = &dev_priv->drm;
int ret;
/* Lock against concurrent debugfs callers */
static int
i915_ring_test_irq_get(void *data, u64 *val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
*val = dev_priv->gpu_error.test_irq_rings;
static int
i915_ring_test_irq_set(void *data, u64 val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
val &= INTEL_INFO(dev_priv)->ring_mask;
DRM_DEBUG_DRIVER("Masking interrupts on rings 0x%08llx\n", val);
static int
i915_drop_caches_set(void *data, u64 val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
+ struct drm_device *dev = &dev_priv->drm;
int ret;
DRM_DEBUG("Dropping caches: 0x%08llx\n", val);
return ret;
if (val & DROP_ACTIVE) {
- ret = i915_gem_wait_for_idle(dev_priv, true);
+ ret = i915_gem_wait_for_idle(dev_priv,
+ I915_WAIT_INTERRUPTIBLE |
+ I915_WAIT_LOCKED);
if (ret)
goto unlock;
}
static int
i915_max_freq_get(void *data, u64 *val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
- if (INTEL_INFO(dev)->gen < 6)
+ if (INTEL_GEN(dev_priv) < 6)
return -ENODEV;
*val = intel_gpu_freq(dev_priv, dev_priv->rps.max_freq_softlimit);
static int
i915_max_freq_set(void *data, u64 val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
u32 hw_max, hw_min;
int ret;
- if (INTEL_INFO(dev)->gen < 6)
+ if (INTEL_GEN(dev_priv) < 6)
return -ENODEV;
DRM_DEBUG_DRIVER("Manually setting max freq to %llu\n", val);
static int
i915_min_freq_get(void *data, u64 *val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
if (INTEL_GEN(dev_priv) < 6)
return -ENODEV;
static int
i915_min_freq_set(void *data, u64 val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
u32 hw_max, hw_min;
int ret;
hw_max = dev_priv->rps.max_freq;
hw_min = dev_priv->rps.min_freq;
- if (val < hw_min || val > hw_max || val > dev_priv->rps.max_freq_softlimit) {
+ if (val < hw_min ||
+ val > hw_max || val > dev_priv->rps.max_freq_softlimit) {
mutex_unlock(&dev_priv->rps.hw_lock);
return -EINVAL;
}
static int
i915_cache_sharing_get(void *data, u64 *val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
+ struct drm_device *dev = &dev_priv->drm;
u32 snpcr;
int ret;
- if (!(IS_GEN6(dev) || IS_GEN7(dev)))
+ if (!(IS_GEN6(dev_priv) || IS_GEN7(dev_priv)))
return -ENODEV;
ret = mutex_lock_interruptible(&dev->struct_mutex);
snpcr = I915_READ(GEN6_MBCUNIT_SNPCR);
intel_runtime_pm_put(dev_priv);
- mutex_unlock(&dev_priv->drm.struct_mutex);
+ mutex_unlock(&dev->struct_mutex);
*val = (snpcr & GEN6_MBC_SNPCR_MASK) >> GEN6_MBC_SNPCR_SHIFT;
static int
i915_cache_sharing_set(void *data, u64 val)
{
- struct drm_device *dev = data;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = data;
u32 snpcr;
- if (!(IS_GEN6(dev) || IS_GEN7(dev)))
+ if (!(IS_GEN6(dev_priv) || IS_GEN7(dev_priv)))
return -ENODEV;
if (val > 3)
i915_cache_sharing_get, i915_cache_sharing_set,
"%llu\n");
-struct sseu_dev_status {
- unsigned int slice_total;
- unsigned int subslice_total;
- unsigned int subslice_per_slice;
- unsigned int eu_total;
- unsigned int eu_per_subslice;
-};
-
-static void cherryview_sseu_device_status(struct drm_device *dev,
- struct sseu_dev_status *stat)
+static void cherryview_sseu_device_status(struct drm_i915_private *dev_priv,
+ struct sseu_dev_info *sseu)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
int ss_max = 2;
int ss;
u32 sig1[ss_max], sig2[ss_max];
/* skip disabled subslice */
continue;
- stat->slice_total = 1;
- stat->subslice_per_slice++;
+ sseu->slice_mask = BIT(0);
+ sseu->subslice_mask |= BIT(ss);
eu_cnt = ((sig1[ss] & CHV_EU08_PG_ENABLE) ? 0 : 2) +
((sig1[ss] & CHV_EU19_PG_ENABLE) ? 0 : 2) +
((sig1[ss] & CHV_EU210_PG_ENABLE) ? 0 : 2) +
((sig2[ss] & CHV_EU311_PG_ENABLE) ? 0 : 2);
- stat->eu_total += eu_cnt;
- stat->eu_per_subslice = max(stat->eu_per_subslice, eu_cnt);
+ sseu->eu_total += eu_cnt;
+ sseu->eu_per_subslice = max_t(unsigned int,
+ sseu->eu_per_subslice, eu_cnt);
}
- stat->subslice_total = stat->subslice_per_slice;
}
-static void gen9_sseu_device_status(struct drm_device *dev,
- struct sseu_dev_status *stat)
+static void gen9_sseu_device_status(struct drm_i915_private *dev_priv,
+ struct sseu_dev_info *sseu)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
int s_max = 3, ss_max = 4;
int s, ss;
u32 s_reg[s_max], eu_reg[2*s_max], eu_mask[2];
/* BXT has a single slice and at most 3 subslices. */
- if (IS_BROXTON(dev)) {
+ if (IS_BROXTON(dev_priv)) {
s_max = 1;
ss_max = 3;
}
GEN9_PGCTL_SSB_EU311_ACK;
for (s = 0; s < s_max; s++) {
- unsigned int ss_cnt = 0;
-
if ((s_reg[s] & GEN9_PGCTL_SLICE_ACK) == 0)
/* skip disabled slice */
continue;
- stat->slice_total++;
+ sseu->slice_mask |= BIT(s);
- if (IS_SKYLAKE(dev) || IS_KABYLAKE(dev))
- ss_cnt = INTEL_INFO(dev)->subslice_per_slice;
+ if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))
+ sseu->subslice_mask =
+ INTEL_INFO(dev_priv)->sseu.subslice_mask;
for (ss = 0; ss < ss_max; ss++) {
unsigned int eu_cnt;
- if (IS_BROXTON(dev) &&
- !(s_reg[s] & (GEN9_PGCTL_SS_ACK(ss))))
- /* skip disabled subslice */
- continue;
+ if (IS_BROXTON(dev_priv)) {
+ if (!(s_reg[s] & (GEN9_PGCTL_SS_ACK(ss))))
+ /* skip disabled subslice */
+ continue;
- if (IS_BROXTON(dev))
- ss_cnt++;
+ sseu->subslice_mask |= BIT(ss);
+ }
eu_cnt = 2 * hweight32(eu_reg[2*s + ss/2] &
eu_mask[ss%2]);
- stat->eu_total += eu_cnt;
- stat->eu_per_subslice = max(stat->eu_per_subslice,
- eu_cnt);
+ sseu->eu_total += eu_cnt;
+ sseu->eu_per_subslice = max_t(unsigned int,
+ sseu->eu_per_subslice,
+ eu_cnt);
}
-
- stat->subslice_total += ss_cnt;
- stat->subslice_per_slice = max(stat->subslice_per_slice,
- ss_cnt);
}
}
-static void broadwell_sseu_device_status(struct drm_device *dev,
- struct sseu_dev_status *stat)
+static void broadwell_sseu_device_status(struct drm_i915_private *dev_priv,
+ struct sseu_dev_info *sseu)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
- int s;
u32 slice_info = I915_READ(GEN8_GT_SLICE_INFO);
+ int s;
- stat->slice_total = hweight32(slice_info & GEN8_LSLICESTAT_MASK);
+ sseu->slice_mask = slice_info & GEN8_LSLICESTAT_MASK;
- if (stat->slice_total) {
- stat->subslice_per_slice = INTEL_INFO(dev)->subslice_per_slice;
- stat->subslice_total = stat->slice_total *
- stat->subslice_per_slice;
- stat->eu_per_subslice = INTEL_INFO(dev)->eu_per_subslice;
- stat->eu_total = stat->eu_per_subslice * stat->subslice_total;
+ if (sseu->slice_mask) {
+ sseu->subslice_mask = INTEL_INFO(dev_priv)->sseu.subslice_mask;
+ sseu->eu_per_subslice =
+ INTEL_INFO(dev_priv)->sseu.eu_per_subslice;
+ sseu->eu_total = sseu->eu_per_subslice *
+ sseu_subslice_total(sseu);
/* subtract fused off EU(s) from enabled slice(s) */
- for (s = 0; s < stat->slice_total; s++) {
- u8 subslice_7eu = INTEL_INFO(dev)->subslice_7eu[s];
+ for (s = 0; s < fls(sseu->slice_mask); s++) {
+ u8 subslice_7eu =
+ INTEL_INFO(dev_priv)->sseu.subslice_7eu[s];
- stat->eu_total -= hweight8(subslice_7eu);
+ sseu->eu_total -= hweight8(subslice_7eu);
}
}
}
-static int i915_sseu_status(struct seq_file *m, void *unused)
+static void i915_print_sseu_info(struct seq_file *m, bool is_available_info,
+ const struct sseu_dev_info *sseu)
{
- struct drm_info_node *node = (struct drm_info_node *) m->private;
- struct drm_i915_private *dev_priv = to_i915(node->minor->dev);
- struct drm_device *dev = &dev_priv->drm;
- struct sseu_dev_status stat;
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ const char *type = is_available_info ? "Available" : "Enabled";
- if (INTEL_INFO(dev)->gen < 8)
- return -ENODEV;
+ seq_printf(m, " %s Slice Mask: %04x\n", type,
+ sseu->slice_mask);
+ seq_printf(m, " %s Slice Total: %u\n", type,
+ hweight8(sseu->slice_mask));
+ seq_printf(m, " %s Subslice Total: %u\n", type,
+ sseu_subslice_total(sseu));
+ seq_printf(m, " %s Subslice Mask: %04x\n", type,
+ sseu->subslice_mask);
+ seq_printf(m, " %s Subslice Per Slice: %u\n", type,
+ hweight8(sseu->subslice_mask));
+ seq_printf(m, " %s EU Total: %u\n", type,
+ sseu->eu_total);
+ seq_printf(m, " %s EU Per Subslice: %u\n", type,
+ sseu->eu_per_subslice);
+
+ if (!is_available_info)
+ return;
+
+ seq_printf(m, " Has Pooled EU: %s\n", yesno(HAS_POOLED_EU(dev_priv)));
+ if (HAS_POOLED_EU(dev_priv))
+ seq_printf(m, " Min EU in pool: %u\n", sseu->min_eu_in_pool);
- seq_puts(m, "SSEU Device Info\n");
- seq_printf(m, " Available Slice Total: %u\n",
- INTEL_INFO(dev)->slice_total);
- seq_printf(m, " Available Subslice Total: %u\n",
- INTEL_INFO(dev)->subslice_total);
- seq_printf(m, " Available Subslice Per Slice: %u\n",
- INTEL_INFO(dev)->subslice_per_slice);
- seq_printf(m, " Available EU Total: %u\n",
- INTEL_INFO(dev)->eu_total);
- seq_printf(m, " Available EU Per Subslice: %u\n",
- INTEL_INFO(dev)->eu_per_subslice);
- seq_printf(m, " Has Pooled EU: %s\n", yesno(HAS_POOLED_EU(dev)));
- if (HAS_POOLED_EU(dev))
- seq_printf(m, " Min EU in pool: %u\n",
- INTEL_INFO(dev)->min_eu_in_pool);
seq_printf(m, " Has Slice Power Gating: %s\n",
- yesno(INTEL_INFO(dev)->has_slice_pg));
+ yesno(sseu->has_slice_pg));
seq_printf(m, " Has Subslice Power Gating: %s\n",
- yesno(INTEL_INFO(dev)->has_subslice_pg));
+ yesno(sseu->has_subslice_pg));
seq_printf(m, " Has EU Power Gating: %s\n",
- yesno(INTEL_INFO(dev)->has_eu_pg));
+ yesno(sseu->has_eu_pg));
+}
+
+static int i915_sseu_status(struct seq_file *m, void *unused)
+{
+ struct drm_i915_private *dev_priv = node_to_i915(m->private);
+ struct sseu_dev_info sseu;
+
+ if (INTEL_GEN(dev_priv) < 8)
+ return -ENODEV;
+
+ seq_puts(m, "SSEU Device Info\n");
+ i915_print_sseu_info(m, true, &INTEL_INFO(dev_priv)->sseu);
seq_puts(m, "SSEU Device Status\n");
- memset(&stat, 0, sizeof(stat));
+ memset(&sseu, 0, sizeof(sseu));
intel_runtime_pm_get(dev_priv);
- if (IS_CHERRYVIEW(dev)) {
- cherryview_sseu_device_status(dev, &stat);
- } else if (IS_BROADWELL(dev)) {
- broadwell_sseu_device_status(dev, &stat);
- } else if (INTEL_INFO(dev)->gen >= 9) {
- gen9_sseu_device_status(dev, &stat);
+ if (IS_CHERRYVIEW(dev_priv)) {
+ cherryview_sseu_device_status(dev_priv, &sseu);
+ } else if (IS_BROADWELL(dev_priv)) {
+ broadwell_sseu_device_status(dev_priv, &sseu);
+ } else if (INTEL_GEN(dev_priv) >= 9) {
+ gen9_sseu_device_status(dev_priv, &sseu);
}
intel_runtime_pm_put(dev_priv);
- seq_printf(m, " Enabled Slice Total: %u\n",
- stat.slice_total);
- seq_printf(m, " Enabled Subslice Total: %u\n",
- stat.subslice_total);
- seq_printf(m, " Enabled Subslice Per Slice: %u\n",
- stat.subslice_per_slice);
- seq_printf(m, " Enabled EU Total: %u\n",
- stat.eu_total);
- seq_printf(m, " Enabled EU Per Subslice: %u\n",
- stat.eu_per_subslice);
+ i915_print_sseu_info(m, false, &sseu);
return 0;
}
static int i915_forcewake_open(struct inode *inode, struct file *file)
{
- struct drm_device *dev = inode->i_private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = inode->i_private;
- if (INTEL_INFO(dev)->gen < 6)
+ if (INTEL_GEN(dev_priv) < 6)
return 0;
intel_runtime_pm_get(dev_priv);
static int i915_forcewake_release(struct inode *inode, struct file *file)
{
- struct drm_device *dev = inode->i_private;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = inode->i_private;
- if (INTEL_INFO(dev)->gen < 6)
+ if (INTEL_GEN(dev_priv) < 6)
return 0;
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
static int i915_forcewake_create(struct dentry *root, struct drm_minor *minor)
{
- struct drm_device *dev = minor->dev;
struct dentry *ent;
ent = debugfs_create_file("i915_forcewake_user",
S_IRUSR,
- root, dev,
+ root, to_i915(minor->dev),
&i915_forcewake_fops);
if (!ent)
return -ENOMEM;
const char *name,
const struct file_operations *fops)
{
- struct drm_device *dev = minor->dev;
struct dentry *ent;
ent = debugfs_create_file(name,
S_IRUGO | S_IWUSR,
- root, dev,
+ root, to_i915(minor->dev),
fops);
if (!ent)
return -ENOMEM;
{"i915_capabilities", i915_capabilities, 0},
{"i915_gem_objects", i915_gem_object_info, 0},
{"i915_gem_gtt", i915_gem_gtt_info, 0},
- {"i915_gem_pinned", i915_gem_gtt_info, 0, (void *) PINNED_LIST},
- {"i915_gem_active", i915_gem_object_list_info, 0, (void *) ACTIVE_LIST},
- {"i915_gem_inactive", i915_gem_object_list_info, 0, (void *) INACTIVE_LIST},
+ {"i915_gem_pin_display", i915_gem_gtt_info, 0, (void *)1},
{"i915_gem_stolen", i915_gem_stolen_list_info },
{"i915_gem_pageflip", i915_gem_pageflip_info, 0},
{"i915_gem_request", i915_gem_request_info, 0},
{"i915_dp_test_active", &i915_displayport_test_active_fops}
};
-void intel_display_crc_init(struct drm_device *dev)
+void intel_display_crc_init(struct drm_i915_private *dev_priv)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
enum pipe pipe;
for_each_pipe(dev_priv, pipe) {
drm_debugfs_remove_files(i915_debugfs_list,
I915_DEBUGFS_ENTRIES, minor);
- drm_debugfs_remove_files((struct drm_info_list *) &i915_forcewake_fops,
+ drm_debugfs_remove_files((struct drm_info_list *)&i915_forcewake_fops,
1, minor);
for (i = 0; i < ARRAY_SIZE(i915_pipe_crc_data); i++) {
for (i = 0; i < ARRAY_SIZE(i915_debugfs_files); i++) {
struct drm_info_list *info_list =
- (struct drm_info_list *) i915_debugfs_files[i].fops;
+ (struct drm_info_list *)i915_debugfs_files[i].fops;
drm_debugfs_remove_files(info_list, 1, minor);
}
.release = single_release,
};
+static int i915_panel_show(struct seq_file *m, void *data)
+{
+ struct drm_connector *connector = m->private;
+ struct intel_dp *intel_dp =
+ enc_to_intel_dp(&intel_attached_encoder(connector)->base);
+
+ if (connector->status != connector_status_connected)
+ return -ENODEV;
+
+ seq_printf(m, "Panel power up delay: %d\n",
+ intel_dp->panel_power_up_delay);
+ seq_printf(m, "Panel power down delay: %d\n",
+ intel_dp->panel_power_down_delay);
+ seq_printf(m, "Backlight on delay: %d\n",
+ intel_dp->backlight_on_delay);
+ seq_printf(m, "Backlight off delay: %d\n",
+ intel_dp->backlight_off_delay);
+
+ return 0;
+}
+
+static int i915_panel_open(struct inode *inode, struct file *file)
+{
+ return single_open(file, i915_panel_show, inode->i_private);
+}
+
+static const struct file_operations i915_panel_fops = {
+ .owner = THIS_MODULE,
+ .open = i915_panel_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = single_release,
+};
+
/**
* i915_debugfs_connector_add - add i915 specific connector debugfs files
* @connector: pointer to a registered drm_connector
if (connector->connector_type == DRM_MODE_CONNECTOR_DisplayPort ||
connector->connector_type == DRM_MODE_CONNECTOR_eDP)
- debugfs_create_file("i915_dpcd", S_IRUGO, root, connector,
- &i915_dpcd_fops);
+ debugfs_create_file("i915_dpcd", S_IRUGO, root,
+ connector, &i915_dpcd_fops);
+
+ if (connector->connector_type == DRM_MODE_CONNECTOR_eDP)
+ debugfs_create_file("i915_panel_timings", S_IRUGO, root,
+ connector, &i915_panel_fops);
return 0;
}
const char *fmt, ...)
{
static bool shown_bug_once;
- struct device *dev = dev_priv->drm.dev;
+ struct device *kdev = dev_priv->drm.dev;
bool is_error = level[1] <= KERN_ERR[1];
bool is_debug = level[1] == KERN_DEBUG[1];
struct va_format vaf;
vaf.fmt = fmt;
vaf.va = &args;
- dev_printk(level, dev, "[" DRM_NAME ":%ps] %pV",
+ dev_printk(level, kdev, "[" DRM_NAME ":%ps] %pV",
__builtin_return_address(0), &vaf);
if (is_error && !shown_bug_once) {
- dev_notice(dev, "%s", FDO_BUG_MSG);
+ dev_notice(kdev, "%s", FDO_BUG_MSG);
shown_bug_once = true;
}
struct drm_file *file_priv)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
drm_i915_getparam_t *param = data;
int value;
/* Reject all old ums/dri params. */
return -ENODEV;
case I915_PARAM_CHIPSET_ID:
- value = dev->pdev->device;
+ value = pdev->device;
break;
case I915_PARAM_REVISION:
- value = dev->pdev->revision;
- break;
- case I915_PARAM_HAS_GEM:
- value = 1;
+ value = pdev->revision;
break;
case I915_PARAM_NUM_FENCES_AVAIL:
value = dev_priv->num_fence_regs;
case I915_PARAM_HAS_OVERLAY:
value = dev_priv->overlay ? 1 : 0;
break;
- case I915_PARAM_HAS_PAGEFLIPPING:
- value = 1;
- break;
- case I915_PARAM_HAS_EXECBUF2:
- /* depends on GEM */
- value = 1;
- break;
case I915_PARAM_HAS_BSD:
value = intel_engine_initialized(&dev_priv->engine[VCS]);
break;
case I915_PARAM_HAS_BSD2:
value = intel_engine_initialized(&dev_priv->engine[VCS2]);
break;
- case I915_PARAM_HAS_RELAXED_FENCING:
- value = 1;
- break;
- case I915_PARAM_HAS_COHERENT_RINGS:
- value = 1;
- break;
case I915_PARAM_HAS_EXEC_CONSTANTS:
- value = INTEL_INFO(dev)->gen >= 4;
- break;
- case I915_PARAM_HAS_RELAXED_DELTA:
- value = 1;
- break;
- case I915_PARAM_HAS_GEN7_SOL_RESET:
- value = 1;
+ value = INTEL_GEN(dev_priv) >= 4;
break;
case I915_PARAM_HAS_LLC:
- value = HAS_LLC(dev);
+ value = HAS_LLC(dev_priv);
break;
case I915_PARAM_HAS_WT:
- value = HAS_WT(dev);
+ value = HAS_WT(dev_priv);
break;
case I915_PARAM_HAS_ALIASING_PPGTT:
- value = USES_PPGTT(dev);
- break;
- case I915_PARAM_HAS_WAIT_TIMEOUT:
- value = 1;
+ value = USES_PPGTT(dev_priv);
break;
case I915_PARAM_HAS_SEMAPHORES:
value = i915.semaphores;
break;
- case I915_PARAM_HAS_PRIME_VMAP_FLUSH:
- value = 1;
- break;
case I915_PARAM_HAS_SECURE_BATCHES:
value = capable(CAP_SYS_ADMIN);
break;
- case I915_PARAM_HAS_PINNED_BATCHES:
- value = 1;
- break;
- case I915_PARAM_HAS_EXEC_NO_RELOC:
- value = 1;
- break;
- case I915_PARAM_HAS_EXEC_HANDLE_LUT:
- value = 1;
- break;
case I915_PARAM_CMD_PARSER_VERSION:
value = i915_cmd_parser_get_version(dev_priv);
break;
- case I915_PARAM_HAS_COHERENT_PHYS_GTT:
- value = 1;
- break;
- case I915_PARAM_MMAP_VERSION:
- value = 1;
- break;
case I915_PARAM_SUBSLICE_TOTAL:
- value = INTEL_INFO(dev)->subslice_total;
+ value = sseu_subslice_total(&INTEL_INFO(dev_priv)->sseu);
if (!value)
return -ENODEV;
break;
case I915_PARAM_EU_TOTAL:
- value = INTEL_INFO(dev)->eu_total;
+ value = INTEL_INFO(dev_priv)->sseu.eu_total;
if (!value)
return -ENODEV;
break;
value = i915.enable_hangcheck && intel_has_gpu_reset(dev_priv);
break;
case I915_PARAM_HAS_RESOURCE_STREAMER:
- value = HAS_RESOURCE_STREAMER(dev);
- break;
- case I915_PARAM_HAS_EXEC_SOFTPIN:
- value = 1;
+ value = HAS_RESOURCE_STREAMER(dev_priv);
break;
case I915_PARAM_HAS_POOLED_EU:
- value = HAS_POOLED_EU(dev);
+ value = HAS_POOLED_EU(dev_priv);
break;
case I915_PARAM_MIN_EU_IN_POOL:
- value = INTEL_INFO(dev)->min_eu_in_pool;
+ value = INTEL_INFO(dev_priv)->sseu.min_eu_in_pool;
+ break;
+ case I915_PARAM_MMAP_GTT_VERSION:
+ /* Though we've started our numbering from 1, and so class all
+ * earlier versions as 0, in effect their value is undefined as
+ * the ioctl will report EINVAL for the unknown param!
+ */
+ value = i915_gem_mmap_gtt_version();
+ break;
+ case I915_PARAM_MMAP_VERSION:
+ /* Remember to bump this if the version changes! */
+ case I915_PARAM_HAS_GEM:
+ case I915_PARAM_HAS_PAGEFLIPPING:
+ case I915_PARAM_HAS_EXECBUF2: /* depends on GEM */
+ case I915_PARAM_HAS_RELAXED_FENCING:
+ case I915_PARAM_HAS_COHERENT_RINGS:
+ case I915_PARAM_HAS_RELAXED_DELTA:
+ case I915_PARAM_HAS_GEN7_SOL_RESET:
+ case I915_PARAM_HAS_WAIT_TIMEOUT:
+ case I915_PARAM_HAS_PRIME_VMAP_FLUSH:
+ case I915_PARAM_HAS_PINNED_BATCHES:
+ case I915_PARAM_HAS_EXEC_NO_RELOC:
+ case I915_PARAM_HAS_EXEC_HANDLE_LUT:
+ case I915_PARAM_HAS_COHERENT_PHYS_GTT:
+ case I915_PARAM_HAS_EXEC_SOFTPIN:
+ /* For the time being all of these are always true;
+ * if some supported hardware does not have one of these
+ * features this value needs to be provided from
+ * INTEL_INFO(), a feature macro, or similar.
+ */
+ value = 1;
break;
default:
DRM_DEBUG("Unknown parameter %d\n", param->param);
pr_info("switched on\n");
dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
/* i915 resume handler doesn't set to D0 */
- pci_set_power_state(dev->pdev, PCI_D0);
+ pci_set_power_state(pdev, PCI_D0);
i915_resume_switcheroo(dev);
dev->switch_power_state = DRM_SWITCH_POWER_ON;
} else {
}
mutex_lock(&dev->struct_mutex);
- i915_gem_reset(dev);
i915_gem_cleanup_engines(dev);
i915_gem_context_fini(dev);
mutex_unlock(&dev->struct_mutex);
static int i915_load_modeset_init(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
int ret;
if (i915_inject_load_failure())
* then we do not take part in VGA arbitration and the
* vga_client_register() fails with -ENODEV.
*/
- ret = vga_client_register(dev->pdev, dev, NULL, i915_vga_set_decode);
+ ret = vga_client_register(pdev, dev, NULL, i915_vga_set_decode);
if (ret && ret != -ENODEV)
goto out;
intel_register_dsm_handler();
- ret = vga_switcheroo_register_client(dev->pdev, &i915_switcheroo_ops, false);
+ ret = vga_switcheroo_register_client(pdev, &i915_switcheroo_ops, false);
if (ret)
goto cleanup_vga_client;
cleanup_csr:
intel_csr_ucode_fini(dev_priv);
intel_power_domains_fini(dev_priv);
- vga_switcheroo_unregister_client(dev->pdev);
+ vga_switcheroo_unregister_client(pdev);
cleanup_vga_client:
- vga_client_register(dev->pdev, NULL, NULL, NULL);
+ vga_client_register(pdev, NULL, NULL, NULL);
out:
return ret;
}
mutex_init(&dev_priv->wm.wm_mutex);
mutex_init(&dev_priv->pps_mutex);
+ i915_memcpy_init_early(dev_priv);
+
ret = i915_workqueues_init(dev_priv);
if (ret < 0)
return ret;
intel_init_audio_hooks(dev_priv);
i915_gem_load_init(&dev_priv->drm);
- intel_display_crc_init(&dev_priv->drm);
+ intel_display_crc_init(dev_priv);
intel_device_info_dump(dev_priv);
static int i915_mmio_setup(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
int mmio_bar;
int mmio_size;
mmio_size = 512 * 1024;
else
mmio_size = 2 * 1024 * 1024;
- dev_priv->regs = pci_iomap(dev->pdev, mmio_bar, mmio_size);
+ dev_priv->regs = pci_iomap(pdev, mmio_bar, mmio_size);
if (dev_priv->regs == NULL) {
DRM_ERROR("failed to map registers\n");
static void i915_mmio_cleanup(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
intel_teardown_mchbar(dev);
- pci_iounmap(dev->pdev, dev_priv->regs);
+ pci_iounmap(pdev, dev_priv->regs);
}
/**
*/
static int i915_driver_init_hw(struct drm_i915_private *dev_priv)
{
+ struct pci_dev *pdev = dev_priv->drm.pdev;
struct drm_device *dev = &dev_priv->drm;
int ret;
goto out_ggtt;
}
- pci_set_master(dev->pdev);
+ pci_set_master(pdev);
/* overlay on gen2 is broken and can't address above 1G */
if (IS_GEN2(dev)) {
- ret = dma_set_coherent_mask(&dev->pdev->dev, DMA_BIT_MASK(30));
+ ret = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(30));
if (ret) {
DRM_ERROR("failed to set DMA mask\n");
* which also needs to be handled carefully.
*/
if (IS_BROADWATER(dev) || IS_CRESTLINE(dev)) {
- ret = dma_set_coherent_mask(&dev->pdev->dev, DMA_BIT_MASK(32));
+ ret = dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
if (ret) {
DRM_ERROR("failed to set DMA mask\n");
* stuck interrupts on some machines.
*/
if (!IS_I945G(dev) && !IS_I945GM(dev)) {
- if (pci_enable_msi(dev->pdev) < 0)
+ if (pci_enable_msi(pdev) < 0)
DRM_DEBUG_DRIVER("can't enable MSI");
}
*/
static void i915_driver_cleanup_hw(struct drm_i915_private *dev_priv)
{
- struct drm_device *dev = &dev_priv->drm;
+ struct pci_dev *pdev = dev_priv->drm.pdev;
- if (dev->pdev->msi_enabled)
- pci_disable_msi(dev->pdev);
+ if (pdev->msi_enabled)
+ pci_disable_msi(pdev);
pm_qos_remove_request(&dev_priv->pm_qos);
i915_ggtt_cleanup_hw(dev_priv);
/* Reveal our presence to userspace */
if (drm_dev_register(dev, 0) == 0) {
i915_debugfs_register(dev_priv);
- i915_setup_sysfs(dev);
+ i915_setup_sysfs(dev_priv);
} else
DRM_ERROR("Failed to register driver for userspace access!\n");
acpi_video_unregister();
intel_opregion_unregister(dev_priv);
- i915_teardown_sysfs(&dev_priv->drm);
+ i915_teardown_sysfs(dev_priv);
i915_debugfs_unregister(dev_priv);
drm_dev_unregister(&dev_priv->drm);
intel_runtime_pm_enable(dev_priv);
+ /* Everything is in place, we can now relax! */
+ DRM_INFO("Initialized %s %d.%d.%d %s for %s on minor %d\n",
+ driver.name, driver.major, driver.minor, driver.patchlevel,
+ driver.date, pci_name(pdev), dev_priv->drm.primary->index);
+
intel_runtime_pm_put(dev_priv);
return 0;
void i915_driver_unload(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
intel_fbdev_fini(dev);
kfree(dev_priv->vbt.lfp_lvds_vbt_mode);
dev_priv->vbt.lfp_lvds_vbt_mode = NULL;
- vga_switcheroo_unregister_client(dev->pdev);
- vga_client_register(dev->pdev, NULL, NULL, NULL);
+ vga_switcheroo_unregister_client(pdev);
+ vga_client_register(pdev, NULL, NULL, NULL);
intel_csr_ucode_fini(dev_priv);
static int i915_drm_suspend(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
pci_power_t opregion_target_state;
int error;
drm_kms_helper_poll_disable(dev);
- pci_save_state(dev->pdev);
+ pci_save_state(pdev);
error = i915_gem_suspend(dev);
if (error) {
- dev_err(&dev->pdev->dev,
+ dev_err(&pdev->dev,
"GEM idle failed, resume might fail\n");
goto out;
}
return error;
}
-static int i915_drm_suspend_late(struct drm_device *drm_dev, bool hibernation)
+static int i915_drm_suspend_late(struct drm_device *dev, bool hibernation)
{
- struct drm_i915_private *dev_priv = to_i915(drm_dev);
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
bool fw_csr;
int ret;
goto out;
}
- pci_disable_device(drm_dev->pdev);
+ pci_disable_device(pdev);
/*
* During hibernation on some platforms the BIOS may try to access
* the device even though it's already in D3 and hang the machine. So
* Acer Aspire 1830T
*/
if (!(hibernation && INTEL_INFO(dev_priv)->gen < 6))
- pci_set_power_state(drm_dev->pdev, PCI_D3hot);
+ pci_set_power_state(pdev, PCI_D3hot);
dev_priv->suspended_to_idle = suspend_to_idle(dev_priv);
int ret;
disable_rpm_wakeref_asserts(dev_priv);
+ intel_sanitize_gt_powersave(dev_priv);
ret = i915_ggtt_enable_hw(dev_priv);
if (ret)
i915_gem_resume(dev);
i915_restore_state(dev);
+ intel_pps_unlock_regs_wa(dev_priv);
intel_opregion_setup(dev_priv);
intel_init_pch_refclk(dev);
mutex_lock(&dev->struct_mutex);
if (i915_gem_init_hw(dev)) {
DRM_ERROR("failed to re-initialize GPU, declaring wedged!\n");
- atomic_or(I915_WEDGED, &dev_priv->gpu_error.reset_counter);
+ i915_gem_set_wedged(dev_priv);
}
mutex_unlock(&dev->struct_mutex);
static int i915_drm_resume_early(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
int ret;
/*
* the device powered we can also remove the following set power state
* call.
*/
- ret = pci_set_power_state(dev->pdev, PCI_D0);
+ ret = pci_set_power_state(pdev, PCI_D0);
if (ret) {
DRM_ERROR("failed to set PCI D0 power state (%d)\n", ret);
goto out;
* depend on the device enable refcount we can't anyway depend on them
* disabling/enabling the device.
*/
- if (pci_enable_device(dev->pdev)) {
+ if (pci_enable_device(pdev)) {
ret = -EIO;
goto out;
}
- pci_set_master(dev->pdev);
+ pci_set_master(pdev);
disable_rpm_wakeref_asserts(dev_priv);
* i915_reset - reset chip after a hang
* @dev: drm device to reset
*
- * Reset the chip. Useful if a hang is detected. Returns zero on successful
- * reset or otherwise an error code.
+ * Reset the chip. Useful if a hang is detected. Marks the device as wedged
+ * on failure.
+ *
+ * Caller must hold the struct_mutex.
*
* Procedure is fairly simple:
* - reset the chip using the reset reg
* - re-init interrupt state
* - re-init display
*/
-int i915_reset(struct drm_i915_private *dev_priv)
+void i915_reset(struct drm_i915_private *dev_priv)
{
struct drm_device *dev = &dev_priv->drm;
struct i915_gpu_error *error = &dev_priv->gpu_error;
- unsigned reset_counter;
int ret;
- mutex_lock(&dev->struct_mutex);
+ lockdep_assert_held(&dev->struct_mutex);
- /* Clear any previous failed attempts at recovery. Time to try again. */
- atomic_andnot(I915_WEDGED, &error->reset_counter);
+ if (!test_and_clear_bit(I915_RESET_IN_PROGRESS, &error->flags))
+ return;
- /* Clear the reset-in-progress flag and increment the reset epoch. */
- reset_counter = atomic_inc_return(&error->reset_counter);
- if (WARN_ON(__i915_reset_in_progress(reset_counter))) {
- ret = -EIO;
- goto error;
- }
+ /* Clear any previous failed attempts at recovery. Time to try again. */
+ __clear_bit(I915_WEDGED, &error->flags);
+ error->reset_count++;
pr_notice("drm/i915: Resetting chip after gpu hang\n");
-
- i915_gem_reset(dev);
-
ret = intel_gpu_reset(dev_priv, ALL_ENGINES);
if (ret) {
if (ret != -ENODEV)
goto error;
}
+ i915_gem_reset(dev_priv);
intel_overlay_reset(dev_priv);
/* Ok, now get things going again... */
goto error;
}
- mutex_unlock(&dev->struct_mutex);
-
/*
* rps/rc6 re-init is necessary to restore state lost after the
* reset and the re-install of gt irqs. Skip for ironlake per
* previous concerns that it doesn't respond well to some forms
* of re-init after reset.
*/
+ intel_sanitize_gt_powersave(dev_priv);
intel_autoenable_gt_powersave(dev_priv);
- return 0;
+wakeup:
+ wake_up_bit(&error->flags, I915_RESET_IN_PROGRESS);
+ return;
error:
- atomic_or(I915_WEDGED, &error->reset_counter);
- mutex_unlock(&dev->struct_mutex);
- return ret;
+ i915_gem_set_wedged(dev_priv);
+ goto wakeup;
}
-static int i915_pm_suspend(struct device *dev)
+static int i915_pm_suspend(struct device *kdev)
{
- struct pci_dev *pdev = to_pci_dev(dev);
- struct drm_device *drm_dev = pci_get_drvdata(pdev);
+ struct pci_dev *pdev = to_pci_dev(kdev);
+ struct drm_device *dev = pci_get_drvdata(pdev);
- if (!drm_dev) {
- dev_err(dev, "DRM not initialized, aborting suspend.\n");
+ if (!dev) {
+ dev_err(kdev, "DRM not initialized, aborting suspend.\n");
return -ENODEV;
}
- if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
- return i915_drm_suspend(drm_dev);
+ return i915_drm_suspend(dev);
}
-static int i915_pm_suspend_late(struct device *dev)
+static int i915_pm_suspend_late(struct device *kdev)
{
- struct drm_device *drm_dev = &dev_to_i915(dev)->drm;
+ struct drm_device *dev = &kdev_to_i915(kdev)->drm;
/*
* We have a suspend ordering issue with the snd-hda driver also
* FIXME: This should be solved with a special hdmi sink device or
* similar so that power domains can be employed.
*/
- if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
- return i915_drm_suspend_late(drm_dev, false);
+ return i915_drm_suspend_late(dev, false);
}
-static int i915_pm_poweroff_late(struct device *dev)
+static int i915_pm_poweroff_late(struct device *kdev)
{
- struct drm_device *drm_dev = &dev_to_i915(dev)->drm;
+ struct drm_device *dev = &kdev_to_i915(kdev)->drm;
- if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
- return i915_drm_suspend_late(drm_dev, true);
+ return i915_drm_suspend_late(dev, true);
}
-static int i915_pm_resume_early(struct device *dev)
+static int i915_pm_resume_early(struct device *kdev)
{
- struct drm_device *drm_dev = &dev_to_i915(dev)->drm;
+ struct drm_device *dev = &kdev_to_i915(kdev)->drm;
- if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
- return i915_drm_resume_early(drm_dev);
+ return i915_drm_resume_early(dev);
}
-static int i915_pm_resume(struct device *dev)
+static int i915_pm_resume(struct device *kdev)
{
- struct drm_device *drm_dev = &dev_to_i915(dev)->drm;
+ struct drm_device *dev = &kdev_to_i915(kdev)->drm;
- if (drm_dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
- return i915_drm_resume(drm_dev);
+ return i915_drm_resume(dev);
}
/* freeze: before creating the hibernation_image */
-static int i915_pm_freeze(struct device *dev)
+static int i915_pm_freeze(struct device *kdev)
{
- return i915_pm_suspend(dev);
+ return i915_pm_suspend(kdev);
}
-static int i915_pm_freeze_late(struct device *dev)
+static int i915_pm_freeze_late(struct device *kdev)
{
int ret;
- ret = i915_pm_suspend_late(dev);
+ ret = i915_pm_suspend_late(kdev);
if (ret)
return ret;
- ret = i915_gem_freeze_late(dev_to_i915(dev));
+ ret = i915_gem_freeze_late(kdev_to_i915(kdev));
if (ret)
return ret;
}
/* thaw: called after creating the hibernation image, but before turning off. */
-static int i915_pm_thaw_early(struct device *dev)
+static int i915_pm_thaw_early(struct device *kdev)
{
- return i915_pm_resume_early(dev);
+ return i915_pm_resume_early(kdev);
}
-static int i915_pm_thaw(struct device *dev)
+static int i915_pm_thaw(struct device *kdev)
{
- return i915_pm_resume(dev);
+ return i915_pm_resume(kdev);
}
/* restore: called after loading the hibernation image. */
-static int i915_pm_restore_early(struct device *dev)
+static int i915_pm_restore_early(struct device *kdev)
{
- return i915_pm_resume_early(dev);
+ return i915_pm_resume_early(kdev);
}
-static int i915_pm_restore(struct device *dev)
+static int i915_pm_restore(struct device *kdev)
{
- return i915_pm_resume(dev);
+ return i915_pm_resume(kdev);
}
/*
return ret;
}
-static int intel_runtime_suspend(struct device *device)
+static int intel_runtime_suspend(struct device *kdev)
{
- struct pci_dev *pdev = to_pci_dev(device);
+ struct pci_dev *pdev = to_pci_dev(kdev);
struct drm_device *dev = pci_get_drvdata(pdev);
struct drm_i915_private *dev_priv = to_i915(dev);
int ret;
* Bump the expiration timestamp, otherwise the suspend won't
* be rescheduled.
*/
- pm_runtime_mark_last_busy(device);
+ pm_runtime_mark_last_busy(kdev);
return -EAGAIN;
}
return 0;
}
-static int intel_runtime_resume(struct device *device)
+static int intel_runtime_resume(struct device *kdev)
{
- struct pci_dev *pdev = to_pci_dev(device);
+ struct pci_dev *pdev = to_pci_dev(kdev);
struct drm_device *dev = pci_get_drvdata(pdev);
struct drm_i915_private *dev_priv = to_i915(dev);
int ret = 0;
#define DRIVER_NAME "i915"
#define DRIVER_DESC "Intel Graphics"
-#define DRIVER_DATE "20160808"
+#define DRIVER_DATE "20160919"
#undef WARN_ON
/* Many gcc seem to no see through this and fall over :( */
struct intel_overlay;
struct intel_overlay_error_state;
-#define I915_FENCE_REG_NONE -1
-#define I915_MAX_NUM_FENCES 32
-/* 32 fences + sign bit for FENCE_REG_NONE */
-#define I915_MAX_NUM_FENCE_BITS 6
-
struct drm_i915_fence_reg {
- struct list_head lru_list;
- struct drm_i915_gem_object *obj;
+ struct list_head link;
+ struct drm_i915_private *i915;
+ struct i915_vma *vma;
int pin_count;
+ int id;
+ /**
+ * Whether the tiling parameters for the currently
+ * associated fence register have changed. Note that
+ * for the purposes of tracking tiling changes we also
+ * treat the unfenced register, the register slot that
+ * the object occupies whilst it executes a fenced
+ * command (such as BLT on gen2/3), as a "fence".
+ */
+ bool dirty;
};
struct sdvo_device_mapping {
u8 ddc_pin;
};
-struct intel_display_error_state;
-
-struct drm_i915_error_state {
- struct kref ref;
- struct timeval time;
-
- char error_msg[128];
- bool simulated;
- int iommu;
- u32 reset_count;
- u32 suspend_count;
-
- /* Generic register state */
- u32 eir;
- u32 pgtbl_er;
- u32 ier;
- u32 gtier[4];
- u32 ccid;
- u32 derrmr;
- u32 forcewake;
- u32 error; /* gen6+ */
- u32 err_int; /* gen7 */
- u32 fault_data0; /* gen8, gen9 */
- u32 fault_data1; /* gen8, gen9 */
- u32 done_reg;
- u32 gac_eco;
- u32 gam_ecochk;
- u32 gab_ctl;
- u32 gfx_mode;
- u32 extra_instdone[I915_NUM_INSTDONE_REG];
- u64 fence[I915_MAX_NUM_FENCES];
- struct intel_overlay_error_state *overlay;
- struct intel_display_error_state *display;
- struct drm_i915_error_object *semaphore_obj;
-
- struct drm_i915_error_engine {
- int engine_id;
- /* Software tracked state */
- bool waiting;
- int num_waiters;
- int hangcheck_score;
- enum intel_engine_hangcheck_action hangcheck_action;
- int num_requests;
-
- /* our own tracking of ring head and tail */
- u32 cpu_ring_head;
- u32 cpu_ring_tail;
-
- u32 last_seqno;
- u32 semaphore_seqno[I915_NUM_ENGINES - 1];
-
- /* Register state */
- u32 start;
- u32 tail;
- u32 head;
- u32 ctl;
- u32 hws;
- u32 ipeir;
- u32 ipehr;
- u32 instdone;
- u32 bbstate;
- u32 instpm;
- u32 instps;
- u32 seqno;
- u64 bbaddr;
- u64 acthd;
- u32 fault_reg;
- u64 faddr;
- u32 rc_psmi; /* sleep state */
- u32 semaphore_mboxes[I915_NUM_ENGINES - 1];
-
- struct drm_i915_error_object {
- int page_count;
- u64 gtt_offset;
- u32 *pages[0];
- } *ringbuffer, *batchbuffer, *wa_batchbuffer, *ctx, *hws_page;
-
- struct drm_i915_error_object *wa_ctx;
-
- struct drm_i915_error_request {
- long jiffies;
- u32 seqno;
- u32 tail;
- } *requests;
-
- struct drm_i915_error_waiter {
- char comm[TASK_COMM_LEN];
- pid_t pid;
- u32 seqno;
- } *waiters;
-
- struct {
- u32 gfx_mode;
- union {
- u64 pdp[4];
- u32 pp_dir_base;
- };
- } vm_info;
-
- pid_t pid;
- char comm[TASK_COMM_LEN];
- } engine[I915_NUM_ENGINES];
-
- struct drm_i915_error_buffer {
- u32 size;
- u32 name;
- u32 rseqno[I915_NUM_ENGINES], wseqno;
- u64 gtt_offset;
- u32 read_domains;
- u32 write_domain;
- s32 fence_reg:I915_MAX_NUM_FENCE_BITS;
- s32 pinned:2;
- u32 tiling:2;
- u32 dirty:1;
- u32 purgeable:1;
- u32 userptr:1;
- s32 engine:4;
- u32 cache_level:3;
- } **active_bo, **pinned_bo;
-
- u32 *active_bo_count, *pinned_bo_count;
- u32 vm_count;
-};
-
struct intel_connector;
struct intel_encoder;
struct intel_crtc_state;
struct intel_initial_plane_config *);
int (*crtc_compute_clock)(struct intel_crtc *crtc,
struct intel_crtc_state *crtc_state);
- void (*crtc_enable)(struct drm_crtc *crtc);
- void (*crtc_disable)(struct drm_crtc *crtc);
+ void (*crtc_enable)(struct intel_crtc_state *pipe_config,
+ struct drm_atomic_state *old_state);
+ void (*crtc_disable)(struct intel_crtc_state *old_crtc_state,
+ struct drm_atomic_state *old_state);
+ void (*update_crtcs)(struct drm_atomic_state *state,
+ unsigned int *crtc_vblank_mask);
void (*audio_codec_enable)(struct drm_connector *connector,
struct intel_encoder *encoder,
const struct drm_display_mode *adjusted_mode);
uint16_t val, bool trace);
void (*mmio_writel)(struct drm_i915_private *dev_priv, i915_reg_t r,
uint32_t val, bool trace);
- void (*mmio_writeq)(struct drm_i915_private *dev_priv, i915_reg_t r,
- uint64_t val, bool trace);
};
struct intel_uncore {
func(is_i915g) sep \
func(is_i945gm) sep \
func(is_g33) sep \
- func(need_gfx_hws) sep \
+ func(hws_needs_physical) sep \
func(is_g4x) sep \
func(is_pineview) sep \
func(is_broadwater) sep \
func(is_kabylake) sep \
func(is_preliminary) sep \
func(has_fbc) sep \
+ func(has_psr) sep \
+ func(has_runtime_pm) sep \
+ func(has_csr) sep \
+ func(has_resource_streamer) sep \
+ func(has_rc6) sep \
+ func(has_rc6p) sep \
+ func(has_dp_mst) sep \
+ func(has_gmbus_irq) sep \
+ func(has_hw_contexts) sep \
+ func(has_logical_ring_contexts) sep \
+ func(has_l3_dpf) sep \
+ func(has_gmch_display) sep \
+ func(has_guc) sep \
func(has_pipe_cxsr) sep \
func(has_hotplug) sep \
func(cursor_needs_physical) sep \
#define DEFINE_FLAG(name) u8 name:1
#define SEP_SEMICOLON ;
+struct sseu_dev_info {
+ u8 slice_mask;
+ u8 subslice_mask;
+ u8 eu_total;
+ u8 eu_per_subslice;
+ u8 min_eu_in_pool;
+ /* For each slice, which subslice(s) has(have) 7 EUs (bitfield)? */
+ u8 subslice_7eu[3];
+ u8 has_slice_pg:1;
+ u8 has_subslice_pg:1;
+ u8 has_eu_pg:1;
+};
+
+static inline unsigned int sseu_subslice_total(const struct sseu_dev_info *sseu)
+{
+ return hweight8(sseu->slice_mask) * hweight8(sseu->subslice_mask);
+}
+
struct intel_device_info {
u32 display_mmio_offset;
u16 device_id;
u8 gen;
u16 gen_mask;
u8 ring_mask; /* Rings supported by the HW */
+ u8 num_rings;
DEV_INFO_FOR_EACH_FLAG(DEFINE_FLAG, SEP_SEMICOLON);
+ u16 ddb_size; /* in blocks */
/* Register offsets for the various display pipes and transcoders */
int pipe_offsets[I915_MAX_TRANSCODERS];
int trans_offsets[I915_MAX_TRANSCODERS];
int cursor_offsets[I915_MAX_PIPES];
/* Slice/subslice/EU info */
- u8 slice_total;
- u8 subslice_total;
- u8 subslice_per_slice;
- u8 eu_total;
- u8 eu_per_subslice;
- u8 min_eu_in_pool;
- /* For each slice, which subslice(s) has(have) 7 EUs (bitfield)? */
- u8 subslice_7eu[3];
- u8 has_slice_pg:1;
- u8 has_subslice_pg:1;
- u8 has_eu_pg:1;
+ struct sseu_dev_info sseu;
struct color_luts {
u16 degamma_lut_size;
#undef DEFINE_FLAG
#undef SEP_SEMICOLON
+struct intel_display_error_state;
+
+struct drm_i915_error_state {
+ struct kref ref;
+ struct timeval time;
+
+ char error_msg[128];
+ bool simulated;
+ int iommu;
+ u32 reset_count;
+ u32 suspend_count;
+ struct intel_device_info device_info;
+
+ /* Generic register state */
+ u32 eir;
+ u32 pgtbl_er;
+ u32 ier;
+ u32 gtier[4];
+ u32 ccid;
+ u32 derrmr;
+ u32 forcewake;
+ u32 error; /* gen6+ */
+ u32 err_int; /* gen7 */
+ u32 fault_data0; /* gen8, gen9 */
+ u32 fault_data1; /* gen8, gen9 */
+ u32 done_reg;
+ u32 gac_eco;
+ u32 gam_ecochk;
+ u32 gab_ctl;
+ u32 gfx_mode;
+ u32 extra_instdone[I915_NUM_INSTDONE_REG];
+ u64 fence[I915_MAX_NUM_FENCES];
+ struct intel_overlay_error_state *overlay;
+ struct intel_display_error_state *display;
+ struct drm_i915_error_object *semaphore;
+
+ struct drm_i915_error_engine {
+ int engine_id;
+ /* Software tracked state */
+ bool waiting;
+ int num_waiters;
+ int hangcheck_score;
+ enum intel_engine_hangcheck_action hangcheck_action;
+ struct i915_address_space *vm;
+ int num_requests;
+
+ /* our own tracking of ring head and tail */
+ u32 cpu_ring_head;
+ u32 cpu_ring_tail;
+
+ u32 last_seqno;
+ u32 semaphore_seqno[I915_NUM_ENGINES - 1];
+
+ /* Register state */
+ u32 start;
+ u32 tail;
+ u32 head;
+ u32 ctl;
+ u32 mode;
+ u32 hws;
+ u32 ipeir;
+ u32 ipehr;
+ u32 instdone;
+ u32 bbstate;
+ u32 instpm;
+ u32 instps;
+ u32 seqno;
+ u64 bbaddr;
+ u64 acthd;
+ u32 fault_reg;
+ u64 faddr;
+ u32 rc_psmi; /* sleep state */
+ u32 semaphore_mboxes[I915_NUM_ENGINES - 1];
+
+ struct drm_i915_error_object {
+ int page_count;
+ u64 gtt_offset;
+ u64 gtt_size;
+ u32 *pages[0];
+ } *ringbuffer, *batchbuffer, *wa_batchbuffer, *ctx, *hws_page;
+
+ struct drm_i915_error_object *wa_ctx;
+
+ struct drm_i915_error_request {
+ long jiffies;
+ pid_t pid;
+ u32 seqno;
+ u32 head;
+ u32 tail;
+ } *requests;
+
+ struct drm_i915_error_waiter {
+ char comm[TASK_COMM_LEN];
+ pid_t pid;
+ u32 seqno;
+ } *waiters;
+
+ struct {
+ u32 gfx_mode;
+ union {
+ u64 pdp[4];
+ u32 pp_dir_base;
+ };
+ } vm_info;
+
+ pid_t pid;
+ char comm[TASK_COMM_LEN];
+ } engine[I915_NUM_ENGINES];
+
+ struct drm_i915_error_buffer {
+ u32 size;
+ u32 name;
+ u32 rseqno[I915_NUM_ENGINES], wseqno;
+ u64 gtt_offset;
+ u32 read_domains;
+ u32 write_domain;
+ s32 fence_reg:I915_MAX_NUM_FENCE_BITS;
+ u32 tiling:2;
+ u32 dirty:1;
+ u32 purgeable:1;
+ u32 userptr:1;
+ s32 engine:4;
+ u32 cache_level:3;
+ } *active_bo[I915_NUM_ENGINES], *pinned_bo;
+ u32 active_bo_count[I915_NUM_ENGINES], pinned_bo_count;
+ struct i915_address_space *active_vm[I915_NUM_ENGINES];
+};
+
enum i915_cache_level {
I915_CACHE_NONE = 0,
I915_CACHE_LLC, /* also used for snoopable memory on non-LLC */
struct drm_i915_private *i915;
struct drm_i915_file_private *file_priv;
struct i915_hw_ppgtt *ppgtt;
+ struct pid *pid;
struct i915_ctx_hang_stats hang_stats;
- /* Unique identifier for this context, used by the hw for tracking */
unsigned long flags;
#define CONTEXT_NO_ZEROMAP BIT(0)
#define CONTEXT_NO_ERROR_CAPTURE BIT(1)
- unsigned hw_id;
+
+ /* Unique identifier for this context, used by the hw for tracking */
+ unsigned int hw_id;
u32 user_handle;
u32 ggtt_alignment;
struct intel_context {
- struct drm_i915_gem_object *state;
+ struct i915_vma *state;
struct intel_ring *ring;
- struct i915_vma *lrc_vma;
uint32_t *lrc_reg_state;
u64 lrc_desc;
int pin_count;
struct i915_suspend_saved_registers {
u32 saveDSPARB;
- u32 saveLVDS;
- u32 savePP_ON_DELAYS;
- u32 savePP_OFF_DELAYS;
- u32 savePP_ON;
- u32 savePP_OFF;
- u32 savePP_CONTROL;
- u32 savePP_DIVISOR;
u32 saveFBC_CONTROL;
u32 saveCACHE_MODE_0;
u32 saveMI_ARB_STATE;
bool interrupts_enabled;
u32 pm_iir;
+ /* PM interrupt bits that should never be masked */
u32 pm_intr_keep;
/* Frequencies are stored in potentially platform dependent multiples.
bool interruptible;
/* the indicator for dispatch video commands on two BSD rings */
- unsigned int bsd_engine_dispatch_index;
+ atomic_t bsd_engine_dispatch_index;
/** Bit 6 swizzling required for X tiling */
uint32_t bit_6_swizzle_x;
* State variable controlling the reset flow and count
*
* This is a counter which gets incremented when reset is triggered,
- * and again when reset has been handled. So odd values (lowest bit set)
- * means that reset is in progress and even values that
- * (reset_counter >> 1):th reset was successfully completed.
+ *
+ * Before the reset commences, the I915_RESET_IN_PROGRESS bit is set
+ * meaning that any waiters holding onto the struct_mutex should
+ * relinquish the lock immediately in order for the reset to start.
*
* If reset is not completed succesfully, the I915_WEDGE bit is
* set meaning that hardware is terminally sour and there is no
* naturally enforces the correct ordering between the bail-out of the
* waiter and the gpu reset work code.
*/
- atomic_t reset_counter;
+ unsigned long reset_count;
-#define I915_RESET_IN_PROGRESS_FLAG 1
-#define I915_WEDGED (1 << 31)
+ unsigned long flags;
+#define I915_RESET_IN_PROGRESS 0
+#define I915_WEDGED (BITS_PER_LONG - 1)
/**
* Waitqueue to signal when a hang is detected. Used to for waiters
uint32_t psr_mmio_base;
+ uint32_t pps_mmio_base;
+
wait_queue_head_t gmbus_wait_queue;
struct pci_dev *bridge_dev;
struct i915_gem_context *kernel_context;
struct intel_engine_cs engine[I915_NUM_ENGINES];
- struct drm_i915_gem_object *semaphore_obj;
+ struct i915_vma *semaphore;
u32 next_seqno;
struct drm_dma_handle *status_page_dmah;
enum modeset_restore modeset_restore;
struct mutex modeset_restore_lock;
struct drm_atomic_state *modeset_restore_state;
+ struct drm_modeset_acquire_ctx reset_ctx;
struct list_head vm_list; /* Global list of all address spaces */
struct i915_ggtt ggtt; /* VM representing the global address space */
struct i915_suspend_saved_registers regfile;
struct vlv_s0ix_state vlv_s0ix_state;
+ enum {
+ I915_SKL_SAGV_UNKNOWN = 0,
+ I915_SKL_SAGV_DISABLED,
+ I915_SKL_SAGV_ENABLED,
+ I915_SKL_SAGV_NOT_CONTROLLED
+ } skl_sagv_status;
+
struct {
/*
* Raw watermark latency values:
/* Abstract the submission mechanism (legacy ringbuffer or execlists) away */
struct {
+ void (*resume)(struct drm_i915_private *);
void (*cleanup_engine)(struct intel_engine_cs *engine);
/**
return container_of(dev, struct drm_i915_private, drm);
}
-static inline struct drm_i915_private *dev_to_i915(struct device *dev)
+static inline struct drm_i915_private *kdev_to_i915(struct device *kdev)
{
- return to_i915(dev_get_drvdata(dev));
+ return to_i915(dev_get_drvdata(kdev));
}
static inline struct drm_i915_private *guc_to_i915(struct intel_guc *guc)
for_each_if (((id__) = (engine__)->id, \
intel_engine_initialized(engine__)))
+#define __mask_next_bit(mask) ({ \
+ int __idx = ffs(mask) - 1; \
+ mask &= ~BIT(__idx); \
+ __idx; \
+})
+
/* Iterator over subset of engines selected by mask */
-#define for_each_engine_masked(engine__, dev_priv__, mask__) \
- for ((engine__) = &(dev_priv__)->engine[0]; \
- (engine__) < &(dev_priv__)->engine[I915_NUM_ENGINES]; \
- (engine__)++) \
- for_each_if (((mask__) & intel_engine_flag(engine__)) && \
- intel_engine_initialized(engine__))
+#define for_each_engine_masked(engine__, dev_priv__, mask__, tmp__) \
+ for (tmp__ = mask__ & INTEL_INFO(dev_priv__)->ring_mask; \
+ tmp__ ? (engine__ = &(dev_priv__)->engine[__mask_next_bit(tmp__)]), 1 : 0; )
enum hdmi_force_audio {
HDMI_AUDIO_OFF_DVI = -2, /* no aux data for HDMI-DVI converter */
*/
unsigned int dirty:1;
- /**
- * Fence register bits (if any) for this object. Will be set
- * as needed when mapped into the GTT.
- * Protected by dev->struct_mutex.
- */
- signed int fence_reg:I915_MAX_NUM_FENCE_BITS;
-
/**
* Advice: are the backing pages purgeable?
*/
unsigned int madv:2;
- /**
- * Whether the tiling parameters for the currently associated fence
- * register have changed. Note that for the purposes of tracking
- * tiling changes we also treat the unfenced register, the register
- * slot that the object occupies whilst it executes a fenced
- * command (such as BLT on gen2/3), as a "fence".
- */
- unsigned int fence_dirty:1;
-
- /**
- * Is the object at the current location in the gtt mappable and
- * fenceable? Used to avoid costly recalculations.
- */
- unsigned int map_and_fenceable:1;
-
/**
* Whether the current gtt mapping needs to be mappable (and isn't just
* mappable by accident). Track pin and fault separate for a more
unsigned int cache_dirty:1;
atomic_t frontbuffer_bits;
+ unsigned int frontbuffer_ggtt_origin; /* write once */
/** Current tiling stride for the object, if it's tiled. */
unsigned int tiling_and_stride;
#define TILING_MASK (FENCE_MINIMUM_STRIDE-1)
#define STRIDE_MASK (~TILING_MASK)
- unsigned int has_wc_mmap;
/** Count of VMA actually bound by this object */
unsigned int bind_count;
unsigned int pin_display;
*/
struct i915_gem_active last_read[I915_NUM_ENGINES];
struct i915_gem_active last_write;
- struct i915_gem_active last_fence;
/** References from framebuffers, locks out tiling changes. */
unsigned long framebuffer_references;
return obj->tiling_and_stride & STRIDE_MASK;
}
+static inline struct i915_vma *i915_vma_get(struct i915_vma *vma)
+{
+ i915_gem_object_get(vma->obj);
+ return vma;
+}
+
+static inline void i915_vma_put(struct i915_vma *vma)
+{
+ lockdep_assert_held(&vma->vm->dev->struct_mutex);
+ i915_gem_object_put(vma->obj);
+}
+
/*
* Optimised SGL iterator for GEM objects
*/
BUILD_BUG(); \
__p; \
})
-#define INTEL_INFO(p) (&__I915__(p)->info)
+#define INTEL_INFO(p) (&__I915__(p)->info)
#define INTEL_GEN(p) (INTEL_INFO(p)->gen)
#define INTEL_DEVID(p) (INTEL_INFO(p)->device_id)
#define HAS_EDRAM(dev) (!!(__I915__(dev)->edram_cap & EDRAM_ENABLED))
#define HAS_WT(dev) ((IS_HASWELL(dev) || IS_BROADWELL(dev)) && \
HAS_EDRAM(dev))
-#define I915_NEED_GFX_HWS(dev) (INTEL_INFO(dev)->need_gfx_hws)
+#define HWS_NEEDS_PHYSICAL(dev) (INTEL_INFO(dev)->hws_needs_physical)
-#define HAS_HW_CONTEXTS(dev) (INTEL_INFO(dev)->gen >= 6)
-#define HAS_LOGICAL_RING_CONTEXTS(dev) (INTEL_INFO(dev)->gen >= 8)
+#define HAS_HW_CONTEXTS(dev) (INTEL_INFO(dev)->has_hw_contexts)
+#define HAS_LOGICAL_RING_CONTEXTS(dev) (INTEL_INFO(dev)->has_logical_ring_contexts)
#define USES_PPGTT(dev) (i915.enable_ppgtt)
#define USES_FULL_PPGTT(dev) (i915.enable_ppgtt >= 2)
#define USES_FULL_48BIT_PPGTT(dev) (i915.enable_ppgtt == 3)
* interrupt source and so prevents the other device from working properly.
*/
#define HAS_AUX_IRQ(dev) (INTEL_INFO(dev)->gen >= 5)
-#define HAS_GMBUS_IRQ(dev) (INTEL_INFO(dev)->gen >= 5)
+#define HAS_GMBUS_IRQ(dev) (INTEL_INFO(dev)->has_gmbus_irq)
/* With the 945 and later, Y tiling got adjusted so that it was 32 128-byte
* rows, which changed the alignment requirements and fence programming.
#define HAS_IPS(dev) (IS_HSW_ULT(dev) || IS_BROADWELL(dev))
-#define HAS_DP_MST(dev) (IS_HASWELL(dev) || IS_BROADWELL(dev) || \
- INTEL_INFO(dev)->gen >= 9)
+#define HAS_DP_MST(dev) (INTEL_INFO(dev)->has_dp_mst)
#define HAS_DDI(dev) (INTEL_INFO(dev)->has_ddi)
#define HAS_FPGA_DBG_UNCLAIMED(dev) (INTEL_INFO(dev)->has_fpga_dbg)
-#define HAS_PSR(dev) (IS_HASWELL(dev) || IS_BROADWELL(dev) || \
- IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev) || \
- IS_SKYLAKE(dev) || IS_KABYLAKE(dev))
-#define HAS_RUNTIME_PM(dev) (IS_GEN6(dev) || IS_HASWELL(dev) || \
- IS_BROADWELL(dev) || IS_VALLEYVIEW(dev) || \
- IS_CHERRYVIEW(dev) || IS_SKYLAKE(dev) || \
- IS_KABYLAKE(dev) || IS_BROXTON(dev))
-#define HAS_RC6(dev) (INTEL_INFO(dev)->gen >= 6)
-#define HAS_RC6p(dev) (IS_GEN6(dev) || IS_IVYBRIDGE(dev))
-
-#define HAS_CSR(dev) (IS_GEN9(dev))
+#define HAS_PSR(dev) (INTEL_INFO(dev)->has_psr)
+#define HAS_RUNTIME_PM(dev) (INTEL_INFO(dev)->has_runtime_pm)
+#define HAS_RC6(dev) (INTEL_INFO(dev)->has_rc6)
+#define HAS_RC6p(dev) (INTEL_INFO(dev)->has_rc6p)
+
+#define HAS_CSR(dev) (INTEL_INFO(dev)->has_csr)
/*
* For now, anything with a GuC requires uCode loading, and then supports
* command submission once loaded. But these are logically independent
* properties, so we have separate macros to test them.
*/
-#define HAS_GUC(dev) (IS_GEN9(dev))
+#define HAS_GUC(dev) (INTEL_INFO(dev)->has_guc)
#define HAS_GUC_UCODE(dev) (HAS_GUC(dev))
#define HAS_GUC_SCHED(dev) (HAS_GUC(dev))
-#define HAS_RESOURCE_STREAMER(dev) (IS_HASWELL(dev) || \
- INTEL_INFO(dev)->gen >= 8)
-
-#define HAS_CORE_RING_FREQ(dev) (INTEL_INFO(dev)->gen >= 6 && \
- !IS_VALLEYVIEW(dev) && !IS_CHERRYVIEW(dev) && \
- !IS_BROXTON(dev))
+#define HAS_RESOURCE_STREAMER(dev) (INTEL_INFO(dev)->has_resource_streamer)
#define HAS_POOLED_EU(dev) (INTEL_INFO(dev)->has_pooled_eu)
#define HAS_PCH_NOP(dev) (INTEL_PCH_TYPE(dev) == PCH_NOP)
#define HAS_PCH_SPLIT(dev) (INTEL_PCH_TYPE(dev) != PCH_NONE)
-#define HAS_GMCH_DISPLAY(dev) (INTEL_INFO(dev)->gen < 5 || \
- IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
+#define HAS_GMCH_DISPLAY(dev) (INTEL_INFO(dev)->has_gmch_display)
/* DPF == dynamic parity feature */
-#define HAS_L3_DPF(dev) (IS_IVYBRIDGE(dev) || IS_HASWELL(dev))
+#define HAS_L3_DPF(dev) (INTEL_INFO(dev)->has_l3_dpf)
#define NUM_L3_SLICES(dev) (IS_HSW_GT3(dev) ? 2 : HAS_L3_DPF(dev))
#define GT_FREQUENCY_MULTIPLIER 50
extern int i915_resume_switcheroo(struct drm_device *dev);
int intel_sanitize_enable_ppgtt(struct drm_i915_private *dev_priv,
- int enable_ppgtt);
+ int enable_ppgtt);
bool intel_sanitize_semaphores(struct drm_i915_private *dev_priv, int value);
#endif
extern int intel_gpu_reset(struct drm_i915_private *dev_priv, u32 engine_mask);
extern bool intel_has_gpu_reset(struct drm_i915_private *dev_priv);
-extern int i915_reset(struct drm_i915_private *dev_priv);
+extern void i915_reset(struct drm_i915_private *dev_priv);
extern int intel_guc_reset(struct drm_i915_private *dev_priv);
extern void intel_engine_init_hangcheck(struct intel_engine_cs *engine);
extern unsigned long i915_chipset_val(struct drm_i915_private *dev_priv);
void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file);
void i915_gem_free_object(struct drm_gem_object *obj);
-int __must_check
+struct i915_vma * __must_check
i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj,
const struct i915_ggtt_view *view,
u64 size,
void i915_gem_release_all_mmaps(struct drm_i915_private *dev_priv);
void i915_gem_release_mmap(struct drm_i915_gem_object *obj);
-int i915_gem_obj_prepare_shmem_read(struct drm_i915_gem_object *obj,
- int *needs_clflush);
-
int __must_check i915_gem_object_get_pages(struct drm_i915_gem_object *obj);
static inline int __sg_page_count(struct scatterlist *sg)
obj->pages_pin_count--;
}
+enum i915_map_type {
+ I915_MAP_WB = 0,
+ I915_MAP_WC,
+};
+
/**
* i915_gem_object_pin_map - return a contiguous mapping of the entire object
* @obj - the object to map into kernel address space
+ * @type - the type of mapping, used to select pgprot_t
*
* Calls i915_gem_object_pin_pages() to prevent reaping of the object's
* pages and then returns a contiguous mapping of the backing storage into
- * the kernel address space.
+ * the kernel address space. Based on the @type of mapping, the PTE will be
+ * set to either WriteBack or WriteCombine (via pgprot_t).
*
* The caller must hold the struct_mutex, and is responsible for calling
* i915_gem_object_unpin_map() when the mapping is no longer required.
* Returns the pointer through which to access the mapped object, or an
* ERR_PTR() on error.
*/
-void *__must_check i915_gem_object_pin_map(struct drm_i915_gem_object *obj);
+void *__must_check i915_gem_object_pin_map(struct drm_i915_gem_object *obj,
+ enum i915_map_type type);
/**
* i915_gem_object_unpin_map - releases an earlier mapping
i915_gem_object_unpin_pages(obj);
}
+int i915_gem_obj_prepare_shmem_read(struct drm_i915_gem_object *obj,
+ unsigned int *needs_clflush);
+int i915_gem_obj_prepare_shmem_write(struct drm_i915_gem_object *obj,
+ unsigned int *needs_clflush);
+#define CLFLUSH_BEFORE 0x1
+#define CLFLUSH_AFTER 0x2
+#define CLFLUSH_FLAGS (CLFLUSH_BEFORE | CLFLUSH_AFTER)
+
+static inline void
+i915_gem_obj_finish_shmem_access(struct drm_i915_gem_object *obj)
+{
+ i915_gem_object_unpin_pages(obj);
+}
+
int __must_check i915_mutex_lock_interruptible(struct drm_device *dev);
-int i915_gem_object_sync(struct drm_i915_gem_object *obj,
- struct drm_i915_gem_request *to);
void i915_vma_move_to_active(struct i915_vma *vma,
struct drm_i915_gem_request *req,
unsigned int flags);
struct drm_mode_create_dumb *args);
int i915_gem_mmap_gtt(struct drm_file *file_priv, struct drm_device *dev,
uint32_t handle, uint64_t *offset);
+int i915_gem_mmap_gtt_version(void);
void i915_gem_track_fb(struct drm_i915_gem_object *old,
struct drm_i915_gem_object *new,
void i915_gem_retire_requests(struct drm_i915_private *dev_priv);
-static inline u32 i915_reset_counter(struct i915_gpu_error *error)
-{
- return atomic_read(&error->reset_counter);
-}
-
-static inline bool __i915_reset_in_progress(u32 reset)
-{
- return unlikely(reset & I915_RESET_IN_PROGRESS_FLAG);
-}
-
-static inline bool __i915_reset_in_progress_or_wedged(u32 reset)
-{
- return unlikely(reset & (I915_RESET_IN_PROGRESS_FLAG | I915_WEDGED));
-}
-
-static inline bool __i915_terminally_wedged(u32 reset)
-{
- return unlikely(reset & I915_WEDGED);
-}
-
static inline bool i915_reset_in_progress(struct i915_gpu_error *error)
{
- return __i915_reset_in_progress(i915_reset_counter(error));
+ return unlikely(test_bit(I915_RESET_IN_PROGRESS, &error->flags));
}
-static inline bool i915_reset_in_progress_or_wedged(struct i915_gpu_error *error)
+static inline bool i915_terminally_wedged(struct i915_gpu_error *error)
{
- return __i915_reset_in_progress_or_wedged(i915_reset_counter(error));
+ return unlikely(test_bit(I915_WEDGED, &error->flags));
}
-static inline bool i915_terminally_wedged(struct i915_gpu_error *error)
+static inline bool i915_reset_in_progress_or_wedged(struct i915_gpu_error *error)
{
- return __i915_terminally_wedged(i915_reset_counter(error));
+ return i915_reset_in_progress(error) | i915_terminally_wedged(error);
}
static inline u32 i915_reset_count(struct i915_gpu_error *error)
{
- return ((i915_reset_counter(error) & ~I915_WEDGED) + 1) / 2;
+ return READ_ONCE(error->reset_count);
}
-void i915_gem_reset(struct drm_device *dev);
+void i915_gem_reset(struct drm_i915_private *dev_priv);
+void i915_gem_set_wedged(struct drm_i915_private *dev_priv);
bool i915_gem_clflush_object(struct drm_i915_gem_object *obj, bool force);
int __must_check i915_gem_init(struct drm_device *dev);
int __must_check i915_gem_init_hw(struct drm_device *dev);
void i915_gem_init_swizzling(struct drm_device *dev);
void i915_gem_cleanup_engines(struct drm_device *dev);
int __must_check i915_gem_wait_for_idle(struct drm_i915_private *dev_priv,
- bool interruptible);
+ unsigned int flags);
int __must_check i915_gem_suspend(struct drm_device *dev);
void i915_gem_resume(struct drm_device *dev);
int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf);
bool write);
int __must_check
i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write);
-int __must_check
+struct i915_vma * __must_check
i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
u32 alignment,
const struct i915_ggtt_view *view);
-void i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj,
- const struct i915_ggtt_view *view);
+void i915_gem_object_unpin_from_display_plane(struct i915_vma *vma);
int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj,
int align);
int i915_gem_open(struct drm_device *dev, struct drm_file *file);
struct dma_buf *i915_gem_prime_export(struct drm_device *dev,
struct drm_gem_object *gem_obj, int flags);
-u64 i915_gem_obj_ggtt_offset_view(struct drm_i915_gem_object *o,
- const struct i915_ggtt_view *view);
-u64 i915_gem_obj_offset(struct drm_i915_gem_object *o,
- struct i915_address_space *vm);
-static inline u64
-i915_gem_obj_ggtt_offset(struct drm_i915_gem_object *o)
-{
- return i915_gem_obj_ggtt_offset_view(o, &i915_ggtt_view_normal);
-}
-
-bool i915_gem_obj_ggtt_bound_view(struct drm_i915_gem_object *o,
- const struct i915_ggtt_view *view);
-bool i915_gem_obj_bound(struct drm_i915_gem_object *o,
- struct i915_address_space *vm);
-
struct i915_vma *
i915_gem_obj_to_vma(struct drm_i915_gem_object *obj,
- struct i915_address_space *vm);
-struct i915_vma *
-i915_gem_obj_to_ggtt_view(struct drm_i915_gem_object *obj,
- const struct i915_ggtt_view *view);
+ struct i915_address_space *vm,
+ const struct i915_ggtt_view *view);
struct i915_vma *
i915_gem_obj_lookup_or_create_vma(struct drm_i915_gem_object *obj,
- struct i915_address_space *vm);
-struct i915_vma *
-i915_gem_obj_lookup_or_create_ggtt_vma(struct drm_i915_gem_object *obj,
- const struct i915_ggtt_view *view);
-
-static inline struct i915_vma *
-i915_gem_obj_to_ggtt(struct drm_i915_gem_object *obj)
-{
- return i915_gem_obj_to_ggtt_view(obj, &i915_ggtt_view_normal);
-}
-bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj);
+ struct i915_address_space *vm,
+ const struct i915_ggtt_view *view);
-/* Some GGTT VM helpers */
static inline struct i915_hw_ppgtt *
i915_vm_to_ppgtt(struct i915_address_space *vm)
{
return container_of(vm, struct i915_hw_ppgtt, base);
}
-static inline bool i915_gem_obj_ggtt_bound(struct drm_i915_gem_object *obj)
+static inline struct i915_vma *
+i915_gem_object_to_ggtt(struct drm_i915_gem_object *obj,
+ const struct i915_ggtt_view *view)
{
- return i915_gem_obj_ggtt_bound_view(obj, &i915_ggtt_view_normal);
+ return i915_gem_obj_to_vma(obj, &to_i915(obj->base.dev)->ggtt.base, view);
}
-unsigned long
-i915_gem_obj_ggtt_size(struct drm_i915_gem_object *obj);
-
-void i915_gem_object_ggtt_unpin_view(struct drm_i915_gem_object *obj,
- const struct i915_ggtt_view *view);
-static inline void
-i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj)
+static inline unsigned long
+i915_gem_object_ggtt_offset(struct drm_i915_gem_object *o,
+ const struct i915_ggtt_view *view)
{
- i915_gem_object_ggtt_unpin_view(obj, &i915_ggtt_view_normal);
+ return i915_ggtt_offset(i915_gem_object_to_ggtt(o, view));
}
/* i915_gem_fence.c */
-int __must_check i915_gem_object_get_fence(struct drm_i915_gem_object *obj);
-int __must_check i915_gem_object_put_fence(struct drm_i915_gem_object *obj);
+int __must_check i915_vma_get_fence(struct i915_vma *vma);
+int __must_check i915_vma_put_fence(struct i915_vma *vma);
-bool i915_gem_object_pin_fence(struct drm_i915_gem_object *obj);
-void i915_gem_object_unpin_fence(struct drm_i915_gem_object *obj);
+/**
+ * i915_vma_pin_fence - pin fencing state
+ * @vma: vma to pin fencing for
+ *
+ * This pins the fencing state (whether tiled or untiled) to make sure the
+ * vma (and its object) is ready to be used as a scanout target. Fencing
+ * status must be synchronize first by calling i915_vma_get_fence():
+ *
+ * The resulting fence pin reference must be released again with
+ * i915_vma_unpin_fence().
+ *
+ * Returns:
+ *
+ * True if the vma has a fence, false otherwise.
+ */
+static inline bool
+i915_vma_pin_fence(struct i915_vma *vma)
+{
+ if (vma->fence) {
+ vma->fence->pin_count++;
+ return true;
+ } else
+ return false;
+}
+
+/**
+ * i915_vma_unpin_fence - unpin fencing state
+ * @vma: vma to unpin fencing for
+ *
+ * This releases the fence pin reference acquired through
+ * i915_vma_pin_fence. It will handle both objects with and without an
+ * attached fence correctly, callers do not need to distinguish this.
+ */
+static inline void
+i915_vma_unpin_fence(struct i915_vma *vma)
+{
+ if (vma->fence) {
+ GEM_BUG_ON(vma->fence->pin_count <= 0);
+ vma->fence->pin_count--;
+ }
+}
void i915_gem_restore_fences(struct drm_device *dev);
int __must_check i915_gem_context_init(struct drm_device *dev);
void i915_gem_context_lost(struct drm_i915_private *dev_priv);
void i915_gem_context_fini(struct drm_device *dev);
-void i915_gem_context_reset(struct drm_device *dev);
int i915_gem_context_open(struct drm_device *dev, struct drm_file *file);
void i915_gem_context_close(struct drm_device *dev, struct drm_file *file);
int i915_switch_context(struct drm_i915_gem_request *req);
/* belongs in i915_gem_gtt.h */
static inline void i915_gem_chipset_flush(struct drm_i915_private *dev_priv)
{
+ wmb();
if (INTEL_GEN(dev_priv) < 6)
intel_gtt_chipset_flush();
}
int i915_debugfs_register(struct drm_i915_private *dev_priv);
void i915_debugfs_unregister(struct drm_i915_private *dev_priv);
int i915_debugfs_connector_add(struct drm_connector *connector);
-void intel_display_crc_init(struct drm_device *dev);
+void intel_display_crc_init(struct drm_i915_private *dev_priv);
#else
static inline int i915_debugfs_register(struct drm_i915_private *dev_priv) {return 0;}
static inline void i915_debugfs_unregister(struct drm_i915_private *dev_priv) {}
static inline int i915_debugfs_connector_add(struct drm_connector *connector)
{ return 0; }
-static inline void intel_display_crc_init(struct drm_device *dev) {}
+static inline void intel_display_crc_init(struct drm_i915_private *dev_priv) {}
#endif
/* i915_gpu_error.c */
/* i915_cmd_parser.c */
int i915_cmd_parser_get_version(struct drm_i915_private *dev_priv);
-int intel_engine_init_cmd_parser(struct intel_engine_cs *engine);
+void intel_engine_init_cmd_parser(struct intel_engine_cs *engine);
void intel_engine_cleanup_cmd_parser(struct intel_engine_cs *engine);
bool intel_engine_needs_cmd_parser(struct intel_engine_cs *engine);
int intel_engine_cmd_parser(struct intel_engine_cs *engine,
extern int i915_restore_state(struct drm_device *dev);
/* i915_sysfs.c */
-void i915_setup_sysfs(struct drm_device *dev_priv);
-void i915_teardown_sysfs(struct drm_device *dev_priv);
+void i915_setup_sysfs(struct drm_i915_private *dev_priv);
+void i915_teardown_sysfs(struct drm_i915_private *dev_priv);
/* intel_i2c.c */
extern int intel_setup_gmbus(struct drm_device *dev);
* will be implemented using 2 32-bit writes in an arbitrary order with
* an arbitrary delay between them. This can cause the hardware to
* act upon the intermediate value, possibly leading to corruption and
- * machine death. You have been warned.
+ * machine death. For this reason we do not support I915_WRITE64, or
+ * dev_priv->uncore.funcs.mmio_writeq.
+ *
+ * When reading a 64-bit value as two 32-bit values, the delay may cause
+ * the two reads to mismatch, e.g. a timestamp overflowing. Also note that
+ * occasionally a 64-bit register does not actualy support a full readq
+ * and must be read using two 32-bit reads.
+ *
+ * You have been warned.
*/
-#define I915_WRITE64(reg, val) dev_priv->uncore.funcs.mmio_writeq(dev_priv, (reg), (val), true)
#define I915_READ64(reg) dev_priv->uncore.funcs.mmio_readq(dev_priv, (reg), true)
#define I915_READ64_2x32(lower_reg, upper_reg) ({ \
#undef __raw_write
/* These are untraced mmio-accessors that are only valid to be used inside
- * criticial sections inside IRQ handlers where forcewake is explicitly
+ * critical sections inside IRQ handlers where forcewake is explicitly
* controlled.
* Think twice, and think again, before using these.
* Note: Should only be used between intel_uncore_forcewake_irqlock() and
schedule_timeout_uninterruptible(remaining_jiffies);
}
}
-static inline bool __i915_request_irq_complete(struct drm_i915_gem_request *req)
+
+static inline bool
+__i915_request_irq_complete(struct drm_i915_gem_request *req)
{
struct intel_engine_cs *engine = req->engine;
* is woken.
*/
if (engine->irq_seqno_barrier &&
- READ_ONCE(engine->breadcrumbs.irq_seqno_bh) == current &&
+ rcu_access_pointer(engine->breadcrumbs.irq_seqno_bh) == current &&
cmpxchg_relaxed(&engine->breadcrumbs.irq_posted, 1, 0)) {
struct task_struct *tsk;
* irq_posted == false but we are still running).
*/
rcu_read_lock();
- tsk = READ_ONCE(engine->breadcrumbs.irq_seqno_bh);
+ tsk = rcu_dereference(engine->breadcrumbs.irq_seqno_bh);
if (tsk && tsk != current)
/* Note that if the bottom-half is changed as we
* are sending the wake-up, the new bottom-half will
return true;
}
- /* We need to check whether any gpu reset happened in between
- * the request being submitted and now. If a reset has occurred,
- * the seqno will have been advance past ours and our request
- * is complete. If we are in the process of handling a reset,
- * the request is effectively complete as the rendering will
- * be discarded, but we need to return in order to drop the
- * struct_mutex.
- */
- if (i915_reset_in_progress(&req->i915->gpu_error))
- return true;
-
return false;
}
+void i915_memcpy_init_early(struct drm_i915_private *dev_priv);
+bool i915_memcpy_from_wc(void *dst, const void *src, unsigned long len);
+
+/* i915_mm.c */
+int remap_io_mapping(struct vm_area_struct *vma,
+ unsigned long addr, unsigned long pfn, unsigned long size,
+ struct io_mapping *iomap);
+
+#define ptr_mask_bits(ptr) ({ \
+ unsigned long __v = (unsigned long)(ptr); \
+ (typeof(ptr))(__v & PAGE_MASK); \
+})
+
+#define ptr_unpack_bits(ptr, bits) ({ \
+ unsigned long __v = (unsigned long)(ptr); \
+ (bits) = __v & ~PAGE_MASK; \
+ (typeof(ptr))(__v & PAGE_MASK); \
+})
+
+#define ptr_pack_bits(ptr, bits) \
+ ((typeof(ptr))((unsigned long)(ptr) | (bits)))
+
+#define fetch_and_zero(ptr) ({ \
+ typeof(*ptr) __T = *(ptr); \
+ *(ptr) = (typeof(*ptr))0; \
+ __T; \
+})
+
#endif
.release = i915_gem_object_release_phys,
};
-int
-i915_gem_object_unbind(struct drm_i915_gem_object *obj)
+int i915_gem_object_unbind(struct drm_i915_gem_object *obj)
{
struct i915_vma *vma;
LIST_HEAD(still_in_list);
int ret;
- /* The vma will only be freed if it is marked as closed, and if we wait
- * upon rendering to the vma, we may unbind anything in the list.
+ lockdep_assert_held(&obj->base.dev->struct_mutex);
+
+ /* Closed vma are removed from the obj->vma_list - but they may
+ * still have an active binding on the object. To remove those we
+ * must wait for all rendering to complete to the object (as unbinding
+ * must anyway), and retire the requests.
*/
+ ret = i915_gem_object_wait_rendering(obj, false);
+ if (ret)
+ return ret;
+
+ i915_gem_retire_requests(to_i915(obj->base.dev));
+
while ((vma = list_first_entry_or_null(&obj->vma_list,
struct i915_vma,
obj_link))) {
int ret;
ret = i915_gem_active_wait_unlocked(&active[idx],
- true, NULL, rps);
+ I915_WAIT_INTERRUPTIBLE,
+ NULL, rps);
if (ret)
return ret;
}
* flush the object from the CPU cache.
*/
int i915_gem_obj_prepare_shmem_read(struct drm_i915_gem_object *obj,
- int *needs_clflush)
+ unsigned int *needs_clflush)
{
int ret;
*needs_clflush = 0;
- if (WARN_ON(!i915_gem_object_has_struct_page(obj)))
- return -EINVAL;
+ if (!i915_gem_object_has_struct_page(obj))
+ return -ENODEV;
ret = i915_gem_object_wait_rendering(obj, true);
if (ret)
return ret;
- if (!(obj->base.read_domains & I915_GEM_DOMAIN_CPU)) {
- /* If we're not in the cpu read domain, set ourself into the gtt
- * read domain and manually flush cachelines (if required). This
- * optimizes for the case when the gpu will dirty the data
- * anyway again before the next pread happens. */
+ ret = i915_gem_object_get_pages(obj);
+ if (ret)
+ return ret;
+
+ i915_gem_object_pin_pages(obj);
+
+ i915_gem_object_flush_gtt_write_domain(obj);
+
+ /* If we're not in the cpu read domain, set ourself into the gtt
+ * read domain and manually flush cachelines (if required). This
+ * optimizes for the case when the gpu will dirty the data
+ * anyway again before the next pread happens.
+ */
+ if (!(obj->base.read_domains & I915_GEM_DOMAIN_CPU))
*needs_clflush = !cpu_cache_is_coherent(obj->base.dev,
obj->cache_level);
+
+ if (*needs_clflush && !static_cpu_has(X86_FEATURE_CLFLUSH)) {
+ ret = i915_gem_object_set_to_cpu_domain(obj, false);
+ if (ret)
+ goto err_unpin;
+
+ *needs_clflush = 0;
}
+ /* return with the pages pinned */
+ return 0;
+
+err_unpin:
+ i915_gem_object_unpin_pages(obj);
+ return ret;
+}
+
+int i915_gem_obj_prepare_shmem_write(struct drm_i915_gem_object *obj,
+ unsigned int *needs_clflush)
+{
+ int ret;
+
+ *needs_clflush = 0;
+ if (!i915_gem_object_has_struct_page(obj))
+ return -ENODEV;
+
+ ret = i915_gem_object_wait_rendering(obj, false);
+ if (ret)
+ return ret;
+
ret = i915_gem_object_get_pages(obj);
if (ret)
return ret;
i915_gem_object_pin_pages(obj);
+ i915_gem_object_flush_gtt_write_domain(obj);
+
+ /* If we're not in the cpu write domain, set ourself into the
+ * gtt write domain and manually flush cachelines (as required).
+ * This optimizes for the case when the gpu will use the data
+ * right away and we therefore have to clflush anyway.
+ */
+ if (obj->base.write_domain != I915_GEM_DOMAIN_CPU)
+ *needs_clflush |= cpu_write_needs_clflush(obj) << 1;
+
+ /* Same trick applies to invalidate partially written cachelines read
+ * before writing.
+ */
+ if (!(obj->base.read_domains & I915_GEM_DOMAIN_CPU))
+ *needs_clflush |= !cpu_cache_is_coherent(obj->base.dev,
+ obj->cache_level);
+
+ if (*needs_clflush && !static_cpu_has(X86_FEATURE_CLFLUSH)) {
+ ret = i915_gem_object_set_to_cpu_domain(obj, true);
+ if (ret)
+ goto err_unpin;
+
+ *needs_clflush = 0;
+ }
+
+ if ((*needs_clflush & CLFLUSH_AFTER) == 0)
+ obj->cache_dirty = true;
+
+ intel_fb_obj_invalidate(obj, ORIGIN_CPU);
+ obj->dirty = 1;
+ /* return with the pages pinned */
+ return 0;
+
+err_unpin:
+ i915_gem_object_unpin_pages(obj);
return ret;
}
{
struct drm_i915_private *dev_priv = to_i915(dev);
struct i915_ggtt *ggtt = &dev_priv->ggtt;
+ struct i915_vma *vma;
struct drm_mm_node node;
char __user *user_data;
uint64_t remain;
uint64_t offset;
int ret;
- ret = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, PIN_MAPPABLE);
- if (ret) {
+ vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, PIN_MAPPABLE);
+ if (!IS_ERR(vma)) {
+ node.start = i915_ggtt_offset(vma);
+ node.allocated = false;
+ ret = i915_vma_put_fence(vma);
+ if (ret) {
+ i915_vma_unpin(vma);
+ vma = ERR_PTR(ret);
+ }
+ }
+ if (IS_ERR(vma)) {
ret = insert_mappable_node(dev_priv, &node, PAGE_SIZE);
if (ret)
goto out;
}
i915_gem_object_pin_pages(obj);
- } else {
- node.start = i915_gem_obj_ggtt_offset(obj);
- node.allocated = false;
- ret = i915_gem_object_put_fence(obj);
- if (ret)
- goto out_unpin;
}
ret = i915_gem_object_set_to_gtt_domain(obj, false);
* and write to user memory which may result into page
* faults, and so we cannot perform this under struct_mutex.
*/
- if (slow_user_access(ggtt->mappable, page_base,
+ if (slow_user_access(&ggtt->mappable, page_base,
page_offset, user_data,
page_length, false)) {
ret = -EFAULT;
i915_gem_object_unpin_pages(obj);
remove_mappable_node(&node);
} else {
- i915_gem_object_ggtt_unpin(obj);
+ i915_vma_unpin(vma);
}
out:
return ret;
int needs_clflush = 0;
struct sg_page_iter sg_iter;
- if (!i915_gem_object_has_struct_page(obj))
- return -ENODEV;
-
- user_data = u64_to_user_ptr(args->data_ptr);
- remain = args->size;
-
- obj_do_bit17_swizzling = i915_gem_object_needs_bit17_swizzle(obj);
-
ret = i915_gem_obj_prepare_shmem_read(obj, &needs_clflush);
if (ret)
return ret;
+ obj_do_bit17_swizzling = i915_gem_object_needs_bit17_swizzle(obj);
+ user_data = u64_to_user_ptr(args->data_ptr);
offset = args->offset;
+ remain = args->size;
for_each_sg_page(obj->pages->sgl, &sg_iter, obj->pages->nents,
offset >> PAGE_SHIFT) {
}
out:
- i915_gem_object_unpin_pages(obj);
+ i915_gem_obj_finish_shmem_access(obj);
return ret;
}
{
struct i915_ggtt *ggtt = &i915->ggtt;
struct drm_device *dev = obj->base.dev;
+ struct i915_vma *vma;
struct drm_mm_node node;
uint64_t remain, offset;
char __user *user_data;
if (i915_gem_object_is_tiled(obj))
return -EFAULT;
- ret = i915_gem_object_ggtt_pin(obj, NULL, 0, 0,
+ vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0,
PIN_MAPPABLE | PIN_NONBLOCK);
- if (ret) {
+ if (!IS_ERR(vma)) {
+ node.start = i915_ggtt_offset(vma);
+ node.allocated = false;
+ ret = i915_vma_put_fence(vma);
+ if (ret) {
+ i915_vma_unpin(vma);
+ vma = ERR_PTR(ret);
+ }
+ }
+ if (IS_ERR(vma)) {
ret = insert_mappable_node(i915, &node, PAGE_SIZE);
if (ret)
goto out;
}
i915_gem_object_pin_pages(obj);
- } else {
- node.start = i915_gem_obj_ggtt_offset(obj);
- node.allocated = false;
- ret = i915_gem_object_put_fence(obj);
- if (ret)
- goto out_unpin;
}
ret = i915_gem_object_set_to_gtt_domain(obj, true);
if (ret)
goto out_unpin;
- intel_fb_obj_invalidate(obj, ORIGIN_GTT);
+ intel_fb_obj_invalidate(obj, ORIGIN_CPU);
obj->dirty = true;
user_data = u64_to_user_ptr(args->data_ptr);
* If the object is non-shmem backed, we retry again with the
* path that handles page fault.
*/
- if (fast_user_write(ggtt->mappable, page_base,
+ if (fast_user_write(&ggtt->mappable, page_base,
page_offset, user_data, page_length)) {
hit_slow_path = true;
mutex_unlock(&dev->struct_mutex);
- if (slow_user_access(ggtt->mappable,
+ if (slow_user_access(&ggtt->mappable,
page_base,
page_offset, user_data,
page_length, true)) {
}
}
- intel_fb_obj_flush(obj, false, ORIGIN_GTT);
+ intel_fb_obj_flush(obj, false, ORIGIN_CPU);
out_unpin:
if (node.allocated) {
wmb();
i915_gem_object_unpin_pages(obj);
remove_mappable_node(&node);
} else {
- i915_gem_object_ggtt_unpin(obj);
+ i915_vma_unpin(vma);
}
out:
return ret;
int shmem_page_offset, page_length, ret = 0;
int obj_do_bit17_swizzling, page_do_bit17_swizzling;
int hit_slowpath = 0;
- int needs_clflush_after = 0;
- int needs_clflush_before = 0;
+ unsigned int needs_clflush;
struct sg_page_iter sg_iter;
- user_data = u64_to_user_ptr(args->data_ptr);
- remain = args->size;
-
- obj_do_bit17_swizzling = i915_gem_object_needs_bit17_swizzle(obj);
-
- ret = i915_gem_object_wait_rendering(obj, false);
- if (ret)
- return ret;
-
- if (obj->base.write_domain != I915_GEM_DOMAIN_CPU) {
- /* If we're not in the cpu write domain, set ourself into the gtt
- * write domain and manually flush cachelines (if required). This
- * optimizes for the case when the gpu will use the data
- * right away and we therefore have to clflush anyway. */
- needs_clflush_after = cpu_write_needs_clflush(obj);
- }
- /* Same trick applies to invalidate partially written cachelines read
- * before writing. */
- if ((obj->base.read_domains & I915_GEM_DOMAIN_CPU) == 0)
- needs_clflush_before =
- !cpu_cache_is_coherent(dev, obj->cache_level);
-
- ret = i915_gem_object_get_pages(obj);
+ ret = i915_gem_obj_prepare_shmem_write(obj, &needs_clflush);
if (ret)
return ret;
- intel_fb_obj_invalidate(obj, ORIGIN_CPU);
-
- i915_gem_object_pin_pages(obj);
-
+ obj_do_bit17_swizzling = i915_gem_object_needs_bit17_swizzle(obj);
+ user_data = u64_to_user_ptr(args->data_ptr);
offset = args->offset;
- obj->dirty = 1;
+ remain = args->size;
for_each_sg_page(obj->pages->sgl, &sg_iter, obj->pages->nents,
offset >> PAGE_SHIFT) {
/* If we don't overwrite a cacheline completely we need to be
* careful to have up-to-date data by first clflushing. Don't
* overcomplicate things and flush the entire patch. */
- partial_cacheline_write = needs_clflush_before &&
+ partial_cacheline_write = needs_clflush & CLFLUSH_BEFORE &&
((shmem_page_offset | page_length)
& (boot_cpu_data.x86_clflush_size - 1));
ret = shmem_pwrite_fast(page, shmem_page_offset, page_length,
user_data, page_do_bit17_swizzling,
partial_cacheline_write,
- needs_clflush_after);
+ needs_clflush & CLFLUSH_AFTER);
if (ret == 0)
goto next_page;
ret = shmem_pwrite_slow(page, shmem_page_offset, page_length,
user_data, page_do_bit17_swizzling,
partial_cacheline_write,
- needs_clflush_after);
+ needs_clflush & CLFLUSH_AFTER);
mutex_lock(&dev->struct_mutex);
}
out:
- i915_gem_object_unpin_pages(obj);
+ i915_gem_obj_finish_shmem_access(obj);
if (hit_slowpath) {
/*
* cachelines in-line while writing and the object moved
* out of the cpu write domain while we've dropped the lock.
*/
- if (!needs_clflush_after &&
+ if (!(needs_clflush & CLFLUSH_AFTER) &&
obj->base.write_domain != I915_GEM_DOMAIN_CPU) {
if (i915_gem_clflush_object(obj, obj->pin_display))
- needs_clflush_after = true;
+ needs_clflush |= CLFLUSH_AFTER;
}
}
- if (needs_clflush_after)
+ if (needs_clflush & CLFLUSH_AFTER)
i915_gem_chipset_flush(to_i915(dev));
- else
- obj->cache_dirty = true;
intel_fb_obj_flush(obj, false, ORIGIN_CPU);
return ret;
if (ret == -EFAULT || ret == -ENOSPC) {
if (obj->phys_handle)
ret = i915_gem_phys_pwrite(obj, args, file);
- else if (i915_gem_object_has_struct_page(obj))
- ret = i915_gem_shmem_pwrite(dev, obj, args, file);
else
- ret = -ENODEV;
+ ret = i915_gem_shmem_pwrite(dev, obj, args, file);
}
i915_gem_object_put(obj);
return ret;
}
-static enum fb_op_origin
+static inline enum fb_op_origin
write_origin(struct drm_i915_gem_object *obj, unsigned domain)
{
- return domain == I915_GEM_DOMAIN_GTT && !obj->has_wc_mmap ?
- ORIGIN_GTT : ORIGIN_CPU;
+ return (domain == I915_GEM_DOMAIN_GTT ?
+ obj->frontbuffer_ggtt_origin : ORIGIN_CPU);
}
/**
up_write(&mm->mmap_sem);
/* This may race, but that's ok, it only gets set */
- WRITE_ONCE(obj->has_wc_mmap, true);
+ WRITE_ONCE(obj->frontbuffer_ggtt_origin, ORIGIN_CPU);
}
i915_gem_object_put_unlocked(obj);
if (IS_ERR((void *)addr))
return 0;
}
+static unsigned int tile_row_pages(struct drm_i915_gem_object *obj)
+{
+ u64 size;
+
+ size = i915_gem_object_get_stride(obj);
+ size *= i915_gem_object_get_tiling(obj) == I915_TILING_Y ? 32 : 8;
+
+ return size >> PAGE_SHIFT;
+}
+
+/**
+ * i915_gem_mmap_gtt_version - report the current feature set for GTT mmaps
+ *
+ * A history of the GTT mmap interface:
+ *
+ * 0 - Everything had to fit into the GTT. Both parties of a memcpy had to
+ * aligned and suitable for fencing, and still fit into the available
+ * mappable space left by the pinned display objects. A classic problem
+ * we called the page-fault-of-doom where we would ping-pong between
+ * two objects that could not fit inside the GTT and so the memcpy
+ * would page one object in at the expense of the other between every
+ * single byte.
+ *
+ * 1 - Objects can be any size, and have any compatible fencing (X Y, or none
+ * as set via i915_gem_set_tiling() [DRM_I915_GEM_SET_TILING]). If the
+ * object is too large for the available space (or simply too large
+ * for the mappable aperture!), a view is created instead and faulted
+ * into userspace. (This view is aligned and sized appropriately for
+ * fenced access.)
+ *
+ * Restrictions:
+ *
+ * * snoopable objects cannot be accessed via the GTT. It can cause machine
+ * hangs on some architectures, corruption on others. An attempt to service
+ * a GTT page fault from a snoopable object will generate a SIGBUS.
+ *
+ * * the object must be able to fit into RAM (physical memory, though no
+ * limited to the mappable aperture).
+ *
+ *
+ * Caveats:
+ *
+ * * a new GTT page fault will synchronize rendering from the GPU and flush
+ * all data to system memory. Subsequent access will not be synchronized.
+ *
+ * * all mappings are revoked on runtime device suspend.
+ *
+ * * there are only 8, 16 or 32 fence registers to share between all users
+ * (older machines require fence register for display and blitter access
+ * as well). Contention of the fence registers will cause the previous users
+ * to be unmapped and any new access will generate new page faults.
+ *
+ * * running out of memory while servicing a fault may generate a SIGBUS,
+ * rather than the expected SIGSEGV.
+ */
+int i915_gem_mmap_gtt_version(void)
+{
+ return 1;
+}
+
/**
* i915_gem_fault - fault a page into the GTT
- * @vma: VMA in question
+ * @area: CPU VMA in question
* @vmf: fault info
*
* The fault handler is set up by drm_gem_mmap() when a object is GTT mapped
* from the GTT and/or fence registers to make room. So performance may
* suffer if the GTT working set is large or there are few fence registers
* left.
+ *
+ * The current feature set supported by i915_gem_fault() and thus GTT mmaps
+ * is exposed via I915_PARAM_MMAP_GTT_VERSION (see i915_gem_mmap_gtt_version).
*/
-int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
+int i915_gem_fault(struct vm_area_struct *area, struct vm_fault *vmf)
{
- struct drm_i915_gem_object *obj = to_intel_bo(vma->vm_private_data);
+#define MIN_CHUNK_PAGES ((1 << 20) >> PAGE_SHIFT) /* 1 MiB */
+ struct drm_i915_gem_object *obj = to_intel_bo(area->vm_private_data);
struct drm_device *dev = obj->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct i915_ggtt *ggtt = &dev_priv->ggtt;
- struct i915_ggtt_view view = i915_ggtt_view_normal;
bool write = !!(vmf->flags & FAULT_FLAG_WRITE);
+ struct i915_vma *vma;
pgoff_t page_offset;
- unsigned long pfn;
+ unsigned int flags;
int ret;
/* We don't use vmf->pgoff since that has the fake offset */
- page_offset = ((unsigned long)vmf->virtual_address - vma->vm_start) >>
+ page_offset = ((unsigned long)vmf->virtual_address - area->vm_start) >>
PAGE_SHIFT;
trace_i915_gem_object_fault(obj, page_offset, true, write);
goto err_unlock;
}
- /* Use a partial view if the object is bigger than the aperture. */
- if (obj->base.size >= ggtt->mappable_end &&
- !i915_gem_object_is_tiled(obj)) {
- static const unsigned int chunk_size = 256; // 1 MiB
+ /* If the object is smaller than a couple of partial vma, it is
+ * not worth only creating a single partial vma - we may as well
+ * clear enough space for the full object.
+ */
+ flags = PIN_MAPPABLE;
+ if (obj->base.size > 2 * MIN_CHUNK_PAGES << PAGE_SHIFT)
+ flags |= PIN_NONBLOCK | PIN_NONFAULT;
+
+ /* Now pin it into the GTT as needed */
+ vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, flags);
+ if (IS_ERR(vma)) {
+ struct i915_ggtt_view view;
+ unsigned int chunk_size;
+
+ /* Use a partial view if it is bigger than available space */
+ chunk_size = MIN_CHUNK_PAGES;
+ if (i915_gem_object_is_tiled(obj))
+ chunk_size = max(chunk_size, tile_row_pages(obj));
memset(&view, 0, sizeof(view));
view.type = I915_GGTT_VIEW_PARTIAL;
view.params.partial.offset = rounddown(page_offset, chunk_size);
view.params.partial.size =
- min_t(unsigned int,
- chunk_size,
- (vma->vm_end - vma->vm_start)/PAGE_SIZE -
+ min_t(unsigned int, chunk_size,
+ (area->vm_end - area->vm_start) / PAGE_SIZE -
view.params.partial.offset);
- }
- /* Now pin it into the GTT if needed */
- ret = i915_gem_object_ggtt_pin(obj, &view, 0, 0, PIN_MAPPABLE);
- if (ret)
+ /* If the partial covers the entire object, just create a
+ * normal VMA.
+ */
+ if (chunk_size >= obj->base.size >> PAGE_SHIFT)
+ view.type = I915_GGTT_VIEW_NORMAL;
+
+ /* Userspace is now writing through an untracked VMA, abandon
+ * all hope that the hardware is able to track future writes.
+ */
+ obj->frontbuffer_ggtt_origin = ORIGIN_CPU;
+
+ vma = i915_gem_object_ggtt_pin(obj, &view, 0, 0, PIN_MAPPABLE);
+ }
+ if (IS_ERR(vma)) {
+ ret = PTR_ERR(vma);
goto err_unlock;
+ }
ret = i915_gem_object_set_to_gtt_domain(obj, write);
if (ret)
goto err_unpin;
- ret = i915_gem_object_get_fence(obj);
+ ret = i915_vma_get_fence(vma);
if (ret)
goto err_unpin;
/* Finally, remap it using the new GTT offset */
- pfn = ggtt->mappable_base +
- i915_gem_obj_ggtt_offset_view(obj, &view);
- pfn >>= PAGE_SHIFT;
-
- if (unlikely(view.type == I915_GGTT_VIEW_PARTIAL)) {
- /* Overriding existing pages in partial view does not cause
- * us any trouble as TLBs are still valid because the fault
- * is due to userspace losing part of the mapping or never
- * having accessed it before (at this partials' range).
- */
- unsigned long base = vma->vm_start +
- (view.params.partial.offset << PAGE_SHIFT);
- unsigned int i;
-
- for (i = 0; i < view.params.partial.size; i++) {
- ret = vm_insert_pfn(vma, base + i * PAGE_SIZE, pfn + i);
- if (ret)
- break;
- }
-
- obj->fault_mappable = true;
- } else {
- if (!obj->fault_mappable) {
- unsigned long size = min_t(unsigned long,
- vma->vm_end - vma->vm_start,
- obj->base.size);
- int i;
-
- for (i = 0; i < size >> PAGE_SHIFT; i++) {
- ret = vm_insert_pfn(vma,
- (unsigned long)vma->vm_start + i * PAGE_SIZE,
- pfn + i);
- if (ret)
- break;
- }
+ ret = remap_io_mapping(area,
+ area->vm_start + (vma->ggtt_view.params.partial.offset << PAGE_SHIFT),
+ (ggtt->mappable_base + vma->node.start) >> PAGE_SHIFT,
+ min_t(u64, vma->size, area->vm_end - area->vm_start),
+ &ggtt->mappable);
+ if (ret)
+ goto err_unpin;
- obj->fault_mappable = true;
- } else
- ret = vm_insert_pfn(vma,
- (unsigned long)vmf->virtual_address,
- pfn + page_offset);
- }
+ obj->fault_mappable = true;
err_unpin:
- i915_gem_object_ggtt_unpin_view(obj, &view);
+ __i915_vma_unpin(vma);
err_unlock:
mutex_unlock(&dev->struct_mutex);
err_rpm:
* to claim that space for ourselves, we need to take the big
* struct_mutex to free the requests+objects and allocate our slot.
*/
- err = i915_gem_wait_for_idle(dev_priv, true);
+ err = i915_gem_wait_for_idle(dev_priv, I915_WAIT_INTERRUPTIBLE);
if (err)
return err;
list_del(&obj->global_list);
if (obj->mapping) {
- if (is_vmalloc_addr(obj->mapping))
- vunmap(obj->mapping);
+ void *ptr;
+
+ ptr = ptr_mask_bits(obj->mapping);
+ if (is_vmalloc_addr(ptr))
+ vunmap(ptr);
else
- kunmap(kmap_to_page(obj->mapping));
+ kunmap(kmap_to_page(ptr));
+
obj->mapping = NULL;
}
}
/* The 'mapping' part of i915_gem_object_pin_map() below */
-static void *i915_gem_object_map(const struct drm_i915_gem_object *obj)
+static void *i915_gem_object_map(const struct drm_i915_gem_object *obj,
+ enum i915_map_type type)
{
unsigned long n_pages = obj->base.size >> PAGE_SHIFT;
struct sg_table *sgt = obj->pages;
struct page *stack_pages[32];
struct page **pages = stack_pages;
unsigned long i = 0;
+ pgprot_t pgprot;
void *addr;
/* A single page can always be kmapped */
- if (n_pages == 1)
+ if (n_pages == 1 && type == I915_MAP_WB)
return kmap(sg_page(sgt->sgl));
if (n_pages > ARRAY_SIZE(stack_pages)) {
/* Check that we have the expected number of pages */
GEM_BUG_ON(i != n_pages);
- addr = vmap(pages, n_pages, 0, PAGE_KERNEL);
+ switch (type) {
+ case I915_MAP_WB:
+ pgprot = PAGE_KERNEL;
+ break;
+ case I915_MAP_WC:
+ pgprot = pgprot_writecombine(PAGE_KERNEL_IO);
+ break;
+ }
+ addr = vmap(pages, n_pages, 0, pgprot);
if (pages != stack_pages)
drm_free_large(pages);
}
/* get, pin, and map the pages of the object into kernel space */
-void *i915_gem_object_pin_map(struct drm_i915_gem_object *obj)
+void *i915_gem_object_pin_map(struct drm_i915_gem_object *obj,
+ enum i915_map_type type)
{
+ enum i915_map_type has_type;
+ bool pinned;
+ void *ptr;
int ret;
lockdep_assert_held(&obj->base.dev->struct_mutex);
+ GEM_BUG_ON(!i915_gem_object_has_struct_page(obj));
ret = i915_gem_object_get_pages(obj);
if (ret)
return ERR_PTR(ret);
i915_gem_object_pin_pages(obj);
+ pinned = obj->pages_pin_count > 1;
- if (!obj->mapping) {
- obj->mapping = i915_gem_object_map(obj);
- if (!obj->mapping) {
- i915_gem_object_unpin_pages(obj);
- return ERR_PTR(-ENOMEM);
+ ptr = ptr_unpack_bits(obj->mapping, has_type);
+ if (ptr && has_type != type) {
+ if (pinned) {
+ ret = -EBUSY;
+ goto err;
}
+
+ if (is_vmalloc_addr(ptr))
+ vunmap(ptr);
+ else
+ kunmap(kmap_to_page(ptr));
+
+ ptr = obj->mapping = NULL;
}
- return obj->mapping;
+ if (!ptr) {
+ ptr = i915_gem_object_map(obj, type);
+ if (!ptr) {
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ obj->mapping = ptr_pack_bits(ptr, type);
+ }
+
+ return ptr;
+
+err:
+ i915_gem_object_unpin_pages(obj);
+ return ERR_PTR(ret);
}
static void
if (i915_gem_request_completed(request))
continue;
+ if (!i915_sw_fence_done(&request->submit))
+ break;
+
return request;
}
return NULL;
}
-static void i915_gem_reset_engine_status(struct intel_engine_cs *engine)
+static void reset_request(struct drm_i915_gem_request *request)
+{
+ void *vaddr = request->ring->vaddr;
+ u32 head;
+
+ /* As this request likely depends on state from the lost
+ * context, clear out all the user operations leaving the
+ * breadcrumb at the end (so we get the fence notifications).
+ */
+ head = request->head;
+ if (request->postfix < head) {
+ memset(vaddr + head, 0, request->ring->size - head);
+ head = 0;
+ }
+ memset(vaddr + head, 0, request->postfix - head);
+}
+
+static void i915_gem_reset_engine(struct intel_engine_cs *engine)
{
struct drm_i915_gem_request *request;
+ struct i915_gem_context *incomplete_ctx;
bool ring_hung;
+ /* Ensure irq handler finishes, and not run again. */
+ tasklet_kill(&engine->irq_tasklet);
+ if (engine->irq_seqno_barrier)
+ engine->irq_seqno_barrier(engine);
+
request = i915_gem_find_active_request(engine);
- if (request == NULL)
+ if (!request)
return;
ring_hung = engine->hangcheck.score >= HANGCHECK_SCORE_RING_HUNG;
-
i915_set_reset_status(request->ctx, ring_hung);
+ if (!ring_hung)
+ return;
+
+ DRM_DEBUG_DRIVER("resetting %s to restart from tail of request 0x%x\n",
+ engine->name, request->fence.seqno);
+
+ /* Setup the CS to resume from the breadcrumb of the hung request */
+ engine->reset_hw(engine, request);
+
+ /* Users of the default context do not rely on logical state
+ * preserved between batches. They have to emit full state on
+ * every batch and so it is safe to execute queued requests following
+ * the hang.
+ *
+ * Other contexts preserve state, now corrupt. We want to skip all
+ * queued requests that reference the corrupt context.
+ */
+ incomplete_ctx = request->ctx;
+ if (i915_gem_context_is_default(incomplete_ctx))
+ return;
+
list_for_each_entry_continue(request, &engine->request_list, link)
- i915_set_reset_status(request->ctx, false);
+ if (request->ctx == incomplete_ctx)
+ reset_request(request);
}
-static void i915_gem_reset_engine_cleanup(struct intel_engine_cs *engine)
+void i915_gem_reset(struct drm_i915_private *dev_priv)
{
- struct drm_i915_gem_request *request;
- struct intel_ring *ring;
+ struct intel_engine_cs *engine;
- request = i915_gem_active_peek(&engine->last_request,
- &engine->i915->drm.struct_mutex);
+ i915_gem_retire_requests(dev_priv);
+
+ for_each_engine(engine, dev_priv)
+ i915_gem_reset_engine(engine);
+
+ i915_gem_restore_fences(&dev_priv->drm);
+}
+
+static void nop_submit_request(struct drm_i915_gem_request *request)
+{
+}
+
+static void i915_gem_cleanup_engine(struct intel_engine_cs *engine)
+{
+ engine->submit_request = nop_submit_request;
/* Mark all pending requests as complete so that any concurrent
* (lockless) lookup doesn't try and wait upon the request as we
* reset it.
*/
- if (request)
- intel_engine_init_seqno(engine, request->fence.seqno);
+ intel_engine_init_seqno(engine, engine->last_submitted_seqno);
/*
* Clear the execlists queue up before freeing the requests, as those
*/
if (i915.enable_execlists) {
- /* Ensure irq handler finishes or is cancelled. */
- tasklet_kill(&engine->irq_tasklet);
-
- intel_execlists_cancel_requests(engine);
- }
-
- /*
- * We must free the requests after all the corresponding objects have
- * been moved off active lists. Which is the same order as the normal
- * retire_requests function does. This is important if object hold
- * implicit references on things like e.g. ppgtt address spaces through
- * the request.
- */
- if (request)
- i915_gem_request_retire_upto(request);
- GEM_BUG_ON(intel_engine_is_active(engine));
-
- /* Having flushed all requests from all queues, we know that all
- * ringbuffers must now be empty. However, since we do not reclaim
- * all space when retiring the request (to prevent HEADs colliding
- * with rapid ringbuffer wraparound) the amount of available space
- * upon reset is less than when we start. Do one more pass over
- * all the ringbuffers to reset last_retired_head.
- */
- list_for_each_entry(ring, &engine->buffers, link) {
- ring->last_retired_head = ring->tail;
- intel_ring_update_space(ring);
+ spin_lock(&engine->execlist_lock);
+ INIT_LIST_HEAD(&engine->execlist_queue);
+ i915_gem_request_put(engine->execlist_port[0].request);
+ i915_gem_request_put(engine->execlist_port[1].request);
+ memset(engine->execlist_port, 0, sizeof(engine->execlist_port));
+ spin_unlock(&engine->execlist_lock);
}
engine->i915->gt.active_engines &= ~intel_engine_flag(engine);
}
-void i915_gem_reset(struct drm_device *dev)
+void i915_gem_set_wedged(struct drm_i915_private *dev_priv)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_engine_cs *engine;
- /*
- * Before we free the objects from the requests, we need to inspect
- * them for finding the guilty party. As the requests only borrow
- * their reference to the objects, the inspection must be done first.
- */
- for_each_engine(engine, dev_priv)
- i915_gem_reset_engine_status(engine);
+ lockdep_assert_held(&dev_priv->drm.struct_mutex);
+ set_bit(I915_WEDGED, &dev_priv->gpu_error.flags);
+ i915_gem_context_lost(dev_priv);
for_each_engine(engine, dev_priv)
- i915_gem_reset_engine_cleanup(engine);
+ i915_gem_cleanup_engine(engine);
mod_delayed_work(dev_priv->wq, &dev_priv->gt.idle_work, 0);
- i915_gem_context_reset(dev);
-
- i915_gem_restore_fences(dev);
+ i915_gem_retire_requests(dev_priv);
}
static void
container_of(work, typeof(*dev_priv), gt.idle_work.work);
struct drm_device *dev = &dev_priv->drm;
struct intel_engine_cs *engine;
- unsigned int stuck_engines;
bool rearm_hangcheck;
if (!READ_ONCE(dev_priv->gt.awake))
dev_priv->gt.awake = false;
rearm_hangcheck = false;
- /* As we have disabled hangcheck, we need to unstick any waiters still
- * hanging around. However, as we may be racing against the interrupt
- * handler or the waiters themselves, we skip enabling the fake-irq.
- */
- stuck_engines = intel_kick_waiters(dev_priv);
- if (unlikely(stuck_engines))
- DRM_DEBUG_DRIVER("kicked stuck waiters (%x)...missed irq?\n",
- stuck_engines);
-
if (INTEL_GEN(dev_priv) >= 6)
gen6_rps_idle(dev_priv);
intel_runtime_pm_put(dev_priv);
active = __I915_BO_ACTIVE(obj);
for_each_active(active, idx) {
s64 *timeout = args->timeout_ns >= 0 ? &args->timeout_ns : NULL;
- ret = i915_gem_active_wait_unlocked(&obj->last_read[idx], true,
+ ret = i915_gem_active_wait_unlocked(&obj->last_read[idx],
+ I915_WAIT_INTERRUPTIBLE,
timeout, rps);
if (ret)
break;
return ret;
}
-static int
-__i915_gem_object_sync(struct drm_i915_gem_request *to,
- struct drm_i915_gem_request *from)
-{
- int ret;
-
- if (to->engine == from->engine)
- return 0;
-
- if (!i915.semaphores) {
- ret = i915_wait_request(from,
- from->i915->mm.interruptible,
- NULL,
- NO_WAITBOOST);
- if (ret)
- return ret;
- } else {
- int idx = intel_engine_sync_index(from->engine, to->engine);
- if (from->fence.seqno <= from->engine->semaphore.sync_seqno[idx])
- return 0;
-
- trace_i915_gem_ring_sync_to(to, from);
- ret = to->engine->semaphore.sync_to(to, from);
- if (ret)
- return ret;
-
- from->engine->semaphore.sync_seqno[idx] = from->fence.seqno;
- }
-
- return 0;
-}
-
-/**
- * i915_gem_object_sync - sync an object to a ring.
- *
- * @obj: object which may be in use on another ring.
- * @to: request we are wishing to use
- *
- * This code is meant to abstract object synchronization with the GPU.
- * Conceptually we serialise writes between engines inside the GPU.
- * We only allow one engine to write into a buffer at any time, but
- * multiple readers. To ensure each has a coherent view of memory, we must:
- *
- * - If there is an outstanding write request to the object, the new
- * request must wait for it to complete (either CPU or in hw, requests
- * on the same ring will be naturally ordered).
- *
- * - If we are a write request (pending_write_domain is set), the new
- * request must wait for outstanding read requests to complete.
- *
- * Returns 0 if successful, else propagates up the lower layer error.
- */
-int
-i915_gem_object_sync(struct drm_i915_gem_object *obj,
- struct drm_i915_gem_request *to)
-{
- struct i915_gem_active *active;
- unsigned long active_mask;
- int idx;
-
- lockdep_assert_held(&obj->base.dev->struct_mutex);
-
- active_mask = i915_gem_object_get_active(obj);
- if (!active_mask)
- return 0;
-
- if (obj->base.pending_write_domain) {
- active = obj->last_read;
- } else {
- active_mask = 1;
- active = &obj->last_write;
- }
-
- for_each_active(active_mask, idx) {
- struct drm_i915_gem_request *request;
- int ret;
-
- request = i915_gem_active_peek(&active[idx],
- &obj->base.dev->struct_mutex);
- if (!request)
- continue;
-
- ret = __i915_gem_object_sync(to, request);
- if (ret)
- return ret;
- }
-
- return 0;
-}
-
-static void i915_gem_object_finish_gtt(struct drm_i915_gem_object *obj)
-{
- u32 old_write_domain, old_read_domains;
-
- /* Force a pagefault for domain tracking on next user access */
- i915_gem_release_mmap(obj);
-
- if ((obj->base.read_domains & I915_GEM_DOMAIN_GTT) == 0)
- return;
-
- old_read_domains = obj->base.read_domains;
- old_write_domain = obj->base.write_domain;
-
- obj->base.read_domains &= ~I915_GEM_DOMAIN_GTT;
- obj->base.write_domain &= ~I915_GEM_DOMAIN_GTT;
-
- trace_i915_gem_object_change_domain(obj,
- old_read_domains,
- old_write_domain);
-}
-
static void __i915_vma_iounmap(struct i915_vma *vma)
{
GEM_BUG_ON(i915_vma_is_pinned(vma));
GEM_BUG_ON(obj->bind_count == 0);
GEM_BUG_ON(!obj->pages);
- if (i915_vma_is_ggtt(vma) &&
- vma->ggtt_view.type == I915_GGTT_VIEW_NORMAL) {
- i915_gem_object_finish_gtt(obj);
-
+ if (i915_vma_is_map_and_fenceable(vma)) {
/* release the fence reg _after_ flushing */
- ret = i915_gem_object_put_fence(obj);
+ ret = i915_vma_put_fence(vma);
if (ret)
return ret;
+ /* Force a pagefault for domain tracking on next user access */
+ i915_gem_release_mmap(obj);
+
__i915_vma_iounmap(vma);
+ vma->flags &= ~I915_VMA_CAN_FENCE;
}
if (likely(!vma->vm->closed)) {
drm_mm_remove_node(&vma->node);
list_move_tail(&vma->vm_link, &vma->vm->unbound_list);
- if (i915_vma_is_ggtt(vma)) {
- if (vma->ggtt_view.type == I915_GGTT_VIEW_NORMAL) {
- obj->map_and_fenceable = false;
- } else if (vma->ggtt_view.pages) {
- sg_free_table(vma->ggtt_view.pages);
- kfree(vma->ggtt_view.pages);
- }
- vma->ggtt_view.pages = NULL;
+ if (vma->pages != obj->pages) {
+ GEM_BUG_ON(!vma->pages);
+ sg_free_table(vma->pages);
+ kfree(vma->pages);
}
+ vma->pages = NULL;
/* Since the unbound list is global, only move to that list if
* no more VMAs exist. */
}
int i915_gem_wait_for_idle(struct drm_i915_private *dev_priv,
- bool interruptible)
+ unsigned int flags)
{
struct intel_engine_cs *engine;
int ret;
if (engine->last_context == NULL)
continue;
- ret = intel_engine_idle(engine, interruptible);
+ ret = intel_engine_idle(engine, flags);
if (ret)
return ret;
}
struct drm_i915_private *dev_priv = to_i915(vma->vm->dev);
struct drm_i915_gem_object *obj = vma->obj;
u64 start, end;
- u64 min_alignment;
int ret;
GEM_BUG_ON(vma->flags & (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND));
size = i915_gem_get_ggtt_size(dev_priv, size,
i915_gem_object_get_tiling(obj));
- min_alignment =
- i915_gem_get_ggtt_alignment(dev_priv, size,
- i915_gem_object_get_tiling(obj),
- flags & PIN_MAPPABLE);
- if (alignment == 0)
- alignment = min_alignment;
- if (alignment & (min_alignment - 1)) {
- DRM_DEBUG("Invalid object alignment requested %llu, minimum %llu\n",
- alignment, min_alignment);
- return -EINVAL;
- }
+ alignment = max(max(alignment, vma->display_alignment),
+ i915_gem_get_ggtt_alignment(dev_priv, size,
+ i915_gem_object_get_tiling(obj),
+ flags & PIN_MAPPABLE));
start = flags & PIN_OFFSET_BIAS ? flags & PIN_OFFSET_MASK : 0;
static void
i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj)
{
- uint32_t old_write_domain;
+ struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
if (obj->base.write_domain != I915_GEM_DOMAIN_GTT)
return;
/* No actual flushing is required for the GTT write domain. Writes
- * to it immediately go to main memory as far as we know, so there's
+ * to it "immediately" go to main memory as far as we know, so there's
* no chipset flush. It also doesn't land in render cache.
*
* However, we do have to enforce the order so that all writes through
* the GTT land before any writes to the device, such as updates to
* the GATT itself.
+ *
+ * We also have to wait a bit for the writes to land from the GTT.
+ * An uncached read (i.e. mmio) seems to be ideal for the round-trip
+ * timing. This issue has only been observed when switching quickly
+ * between GTT writes and CPU reads from inside the kernel on recent hw,
+ * and it appears to only affect discrete GTT blocks (i.e. on LLC
+ * system agents we cannot reproduce this behaviour).
*/
wmb();
+ if (INTEL_GEN(dev_priv) >= 6 && !HAS_LLC(dev_priv))
+ POSTING_READ(RING_ACTHD(dev_priv->engine[RCS].mmio_base));
- old_write_domain = obj->base.write_domain;
- obj->base.write_domain = 0;
-
- intel_fb_obj_flush(obj, false, ORIGIN_GTT);
+ intel_fb_obj_flush(obj, false, write_origin(obj, I915_GEM_DOMAIN_GTT));
+ obj->base.write_domain = 0;
trace_i915_gem_object_change_domain(obj,
obj->base.read_domains,
- old_write_domain);
+ I915_GEM_DOMAIN_GTT);
}
/** Flushes the CPU write domain for the object if it's dirty. */
static void
i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj)
{
- uint32_t old_write_domain;
-
if (obj->base.write_domain != I915_GEM_DOMAIN_CPU)
return;
if (i915_gem_clflush_object(obj, obj->pin_display))
i915_gem_chipset_flush(to_i915(obj->base.dev));
- old_write_domain = obj->base.write_domain;
- obj->base.write_domain = 0;
-
intel_fb_obj_flush(obj, false, ORIGIN_CPU);
+ obj->base.write_domain = 0;
trace_i915_gem_object_change_domain(obj,
obj->base.read_domains,
- old_write_domain);
+ I915_GEM_DOMAIN_CPU);
+}
+
+static void i915_gem_object_bump_inactive_ggtt(struct drm_i915_gem_object *obj)
+{
+ struct i915_vma *vma;
+
+ list_for_each_entry(vma, &obj->vma_list, obj_link) {
+ if (!i915_vma_is_ggtt(vma))
+ continue;
+
+ if (i915_vma_is_active(vma))
+ continue;
+
+ if (!drm_mm_node_allocated(&vma->node))
+ continue;
+
+ list_move_tail(&vma->vm_link, &vma->vm->inactive_list);
+ }
}
/**
i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
{
uint32_t old_write_domain, old_read_domains;
- struct i915_vma *vma;
int ret;
ret = i915_gem_object_wait_rendering(obj, !write);
old_write_domain);
/* And bump the LRU for this access */
- vma = i915_gem_obj_to_ggtt(obj);
- if (vma &&
- drm_mm_node_allocated(&vma->node) &&
- !i915_vma_is_active(vma))
- list_move_tail(&vma->vm_link, &vma->vm->inactive_list);
+ i915_gem_object_bump_inactive_ggtt(obj);
return 0;
}
* dropped the fence as all snoopable access is
* supposed to be linear.
*/
- ret = i915_gem_object_put_fence(obj);
- if (ret)
- return ret;
+ list_for_each_entry(vma, &obj->vma_list, obj_link) {
+ ret = i915_vma_put_fence(vma);
+ if (ret)
+ return ret;
+ }
} else {
/* We either have incoherent backing store and
* so no GTT access or the architecture is fully
* Can be called from an uninterruptible phase (modesetting) and allows
* any flushes to be pipelined (for pageflips).
*/
-int
+struct i915_vma *
i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
u32 alignment,
const struct i915_ggtt_view *view)
{
+ struct i915_vma *vma;
u32 old_read_domains, old_write_domain;
int ret;
*/
ret = i915_gem_object_set_cache_level(obj,
HAS_WT(obj->base.dev) ? I915_CACHE_WT : I915_CACHE_NONE);
- if (ret)
+ if (ret) {
+ vma = ERR_PTR(ret);
goto err_unpin_display;
+ }
/* As the user may map the buffer once pinned in the display plane
* (e.g. libkms for the bootup splash), we have to ensure that we
- * always use map_and_fenceable for all scanout buffers.
+ * always use map_and_fenceable for all scanout buffers. However,
+ * it may simply be too big to fit into mappable, in which case
+ * put it anyway and hope that userspace can cope (but always first
+ * try to preserve the existing ABI).
*/
- ret = i915_gem_object_ggtt_pin(obj, view, 0, alignment,
- view->type == I915_GGTT_VIEW_NORMAL ?
- PIN_MAPPABLE : 0);
- if (ret)
+ vma = ERR_PTR(-ENOSPC);
+ if (view->type == I915_GGTT_VIEW_NORMAL)
+ vma = i915_gem_object_ggtt_pin(obj, view, 0, alignment,
+ PIN_MAPPABLE | PIN_NONBLOCK);
+ if (IS_ERR(vma))
+ vma = i915_gem_object_ggtt_pin(obj, view, 0, alignment, 0);
+ if (IS_ERR(vma))
goto err_unpin_display;
+ vma->display_alignment = max_t(u64, vma->display_alignment, alignment);
+
+ WARN_ON(obj->pin_display > i915_vma_pin_count(vma));
+
i915_gem_object_flush_cpu_write_domain(obj);
old_write_domain = obj->base.write_domain;
old_read_domains,
old_write_domain);
- return 0;
+ return vma;
err_unpin_display:
obj->pin_display--;
- return ret;
+ return vma;
}
void
-i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj,
- const struct i915_ggtt_view *view)
+i915_gem_object_unpin_from_display_plane(struct i915_vma *vma)
{
- if (WARN_ON(obj->pin_display == 0))
+ if (WARN_ON(vma->obj->pin_display == 0))
return;
- i915_gem_object_ggtt_unpin_view(obj, view);
+ if (--vma->obj->pin_display == 0)
+ vma->display_alignment = 0;
- obj->pin_display--;
+ /* Bump the LRU to try and avoid premature eviction whilst flipping */
+ if (!i915_vma_is_active(vma))
+ list_move_tail(&vma->vm_link, &vma->vm->inactive_list);
+
+ i915_vma_unpin(vma);
+ WARN_ON(vma->obj->pin_display > i915_vma_pin_count(vma));
}
/**
if (target == NULL)
return 0;
- ret = i915_wait_request(target, true, NULL, NULL);
+ ret = i915_wait_request(target, I915_WAIT_INTERRUPTIBLE, NULL, NULL);
i915_gem_request_put(target);
return ret;
static bool
i915_vma_misplaced(struct i915_vma *vma, u64 size, u64 alignment, u64 flags)
{
- struct drm_i915_gem_object *obj = vma->obj;
-
if (!drm_mm_node_allocated(&vma->node))
return false;
if (alignment && vma->node.start & (alignment - 1))
return true;
- if (flags & PIN_MAPPABLE && !obj->map_and_fenceable)
+ if (flags & PIN_MAPPABLE && !i915_vma_is_map_and_fenceable(vma))
return true;
if (flags & PIN_OFFSET_BIAS &&
u32 fence_size, fence_alignment;
fence_size = i915_gem_get_ggtt_size(dev_priv,
- obj->base.size,
+ vma->size,
i915_gem_object_get_tiling(obj));
fence_alignment = i915_gem_get_ggtt_alignment(dev_priv,
- obj->base.size,
+ vma->size,
i915_gem_object_get_tiling(obj),
true);
mappable = (vma->node.start + fence_size <=
dev_priv->ggtt.mappable_end);
- obj->map_and_fenceable = mappable && fenceable;
+ if (mappable && fenceable)
+ vma->flags |= I915_VMA_CAN_FENCE;
+ else
+ vma->flags &= ~I915_VMA_CAN_FENCE;
}
int __i915_vma_do_pin(struct i915_vma *vma,
return ret;
}
-int
+struct i915_vma *
i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj,
const struct i915_ggtt_view *view,
u64 size,
u64 alignment,
u64 flags)
{
+ struct i915_address_space *vm = &to_i915(obj->base.dev)->ggtt.base;
struct i915_vma *vma;
int ret;
- if (!view)
- view = &i915_ggtt_view_normal;
-
- vma = i915_gem_obj_lookup_or_create_ggtt_vma(obj, view);
+ vma = i915_gem_obj_lookup_or_create_vma(obj, vm, view);
if (IS_ERR(vma))
- return PTR_ERR(vma);
+ return vma;
if (i915_vma_misplaced(vma, size, alignment, flags)) {
if (flags & PIN_NONBLOCK &&
(i915_vma_is_pinned(vma) || i915_vma_is_active(vma)))
- return -ENOSPC;
+ return ERR_PTR(-ENOSPC);
WARN(i915_vma_is_pinned(vma),
"bo is already pinned in ggtt with incorrect alignment:"
- " offset=%08x %08x, req.alignment=%llx, req.map_and_fenceable=%d,"
- " obj->map_and_fenceable=%d\n",
- upper_32_bits(vma->node.start),
- lower_32_bits(vma->node.start),
- alignment,
+ " offset=%08x, req.alignment=%llx,"
+ " req.map_and_fenceable=%d, vma->map_and_fenceable=%d\n",
+ i915_ggtt_offset(vma), alignment,
!!(flags & PIN_MAPPABLE),
- obj->map_and_fenceable);
+ i915_vma_is_map_and_fenceable(vma));
ret = i915_vma_unbind(vma);
if (ret)
- return ret;
+ return ERR_PTR(ret);
}
- return i915_vma_pin(vma, size, alignment, flags | PIN_GLOBAL);
-}
+ ret = i915_vma_pin(vma, size, alignment, flags | PIN_GLOBAL);
+ if (ret)
+ return ERR_PTR(ret);
-void
-i915_gem_object_ggtt_unpin_view(struct drm_i915_gem_object *obj,
- const struct i915_ggtt_view *view)
-{
- i915_vma_unpin(i915_gem_obj_to_ggtt_view(obj, view));
+ return vma;
}
-static __always_inline unsigned __busy_read_flag(unsigned int id)
+static __always_inline unsigned int __busy_read_flag(unsigned int id)
{
/* Note that we could alias engines in the execbuf API, but
* that would be very unwise as it prevents userspace from
static __always_inline unsigned int __busy_write_id(unsigned int id)
{
- return id;
+ /* The uABI guarantees an active writer is also amongst the read
+ * engines. This would be true if we accessed the activity tracking
+ * under the lock, but as we perform the lookup of the object and
+ * its activity locklessly we can not guarantee that the last_write
+ * being active implies that we have set the same engine flag from
+ * last_read - hence we always set both read and write busy for
+ * last_write.
+ */
+ return id | __busy_read_flag(id);
}
-static __always_inline unsigned
+static __always_inline unsigned int
__busy_set_if_active(const struct i915_gem_active *active,
unsigned int (*flag)(unsigned int id))
{
- /* For more discussion about the barriers and locking concerns,
- * see __i915_gem_active_get_rcu().
- */
- do {
- struct drm_i915_gem_request *request;
- unsigned int id;
-
- request = rcu_dereference(active->request);
- if (!request || i915_gem_request_completed(request))
- return 0;
+ struct drm_i915_gem_request *request;
- id = request->engine->exec_id;
+ request = rcu_dereference(active->request);
+ if (!request || i915_gem_request_completed(request))
+ return 0;
- /* Check that the pointer wasn't reassigned and overwritten. */
- if (request == rcu_access_pointer(active->request))
- return flag(id);
- } while (1);
+ /* This is racy. See __i915_gem_active_get_rcu() for an in detail
+ * discussion of how to handle the race correctly, but for reporting
+ * the busy state we err on the side of potentially reporting the
+ * wrong engine as being busy (but we guarantee that the result
+ * is at least self-consistent).
+ *
+ * As we use SLAB_DESTROY_BY_RCU, the request may be reallocated
+ * whilst we are inspecting it, even under the RCU read lock as we are.
+ * This means that there is a small window for the engine and/or the
+ * seqno to have been overwritten. The seqno will always be in the
+ * future compared to the intended, and so we know that if that
+ * seqno is idle (on whatever engine) our request is idle and the
+ * return 0 above is correct.
+ *
+ * The issue is that if the engine is switched, it is just as likely
+ * to report that it is busy (but since the switch happened, we know
+ * the request should be idle). So there is a small chance that a busy
+ * result is actually the wrong engine.
+ *
+ * So why don't we care?
+ *
+ * For starters, the busy ioctl is a heuristic that is by definition
+ * racy. Even with perfect serialisation in the driver, the hardware
+ * state is constantly advancing - the state we report to the user
+ * is stale.
+ *
+ * The critical information for the busy-ioctl is whether the object
+ * is idle as userspace relies on that to detect whether its next
+ * access will stall, or if it has missed submitting commands to
+ * the hardware allowing the GPU to stall. We never generate a
+ * false-positive for idleness, thus busy-ioctl is reliable at the
+ * most fundamental level, and we maintain the guarantee that a
+ * busy object left to itself will eventually become idle (and stay
+ * idle!).
+ *
+ * We allow ourselves the leeway of potentially misreporting the busy
+ * state because that is an optimisation heuristic that is constantly
+ * in flux. Being quickly able to detect the busy/idle state is much
+ * more important than accurate logging of exactly which engines were
+ * busy.
+ *
+ * For accuracy in reporting the engine, we could use
+ *
+ * result = 0;
+ * request = __i915_gem_active_get_rcu(active);
+ * if (request) {
+ * if (!i915_gem_request_completed(request))
+ * result = flag(request->engine->exec_id);
+ * i915_gem_request_put(request);
+ * }
+ *
+ * but that still remains susceptible to both hardware and userspace
+ * races. So we accept making the result of that race slightly worse,
+ * given the rarity of the race and its low impact on the result.
+ */
+ return flag(READ_ONCE(request->engine->exec_id));
}
-static inline unsigned
+static __always_inline unsigned int
busy_check_reader(const struct i915_gem_active *active)
{
return __busy_set_if_active(active, __busy_read_flag);
}
-static inline unsigned
+static __always_inline unsigned int
busy_check_writer(const struct i915_gem_active *active)
{
return __busy_set_if_active(active, __busy_write_id);
* retired and freed. We take a local copy of the pointer,
* but before we add its engine into the busy set, the other
* thread reallocates it and assigns it to a task on another
- * engine with a fresh and incomplete seqno.
- *
- * So after we lookup the engine's id, we double check that
- * the active request is the same and only then do we add it
- * into the busy set.
+ * engine with a fresh and incomplete seqno. Guarding against
+ * that requires careful serialisation and reference counting,
+ * i.e. using __i915_gem_active_get_request_rcu(). We don't,
+ * instead we expect that if the result is busy, which engines
+ * are busy is not completely reliable - we only guarantee
+ * that the object was busy.
*/
rcu_read_lock();
args->busy |= busy_check_reader(&obj->last_read[idx]);
/* For ABI sanity, we only care that the write engine is in
- * the set of read engines. This is ensured by the ordering
- * of setting last_read/last_write in i915_vma_move_to_active,
- * and then in reverse in retire.
+ * the set of read engines. This should be ensured by the
+ * ordering of setting last_read/last_write in
+ * i915_vma_move_to_active(), and then in reverse in retire.
+ * However, for good measure, we always report the last_write
+ * request as a busy read as well as being a busy write.
*
* We don't care that the set of active read/write engines
* may change during construction of the result, as it is
i915_gem_object_retire__read);
init_request_active(&obj->last_write,
i915_gem_object_retire__write);
- init_request_active(&obj->last_fence, NULL);
INIT_LIST_HEAD(&obj->obj_exec_link);
INIT_LIST_HEAD(&obj->vma_list);
INIT_LIST_HEAD(&obj->batch_pool_link);
obj->ops = ops;
- obj->fence_reg = I915_FENCE_REG_NONE;
+ obj->frontbuffer_ggtt_origin = ORIGIN_GTT;
obj->madv = I915_MADV_WILLNEED;
i915_gem_info_add_obj(to_i915(obj->base.dev), obj->base.size);
intel_runtime_pm_put(dev_priv);
}
-struct i915_vma *i915_gem_obj_to_vma(struct drm_i915_gem_object *obj,
- struct i915_address_space *vm)
-{
- struct i915_vma *vma;
- list_for_each_entry(vma, &obj->vma_list, obj_link) {
- if (vma->ggtt_view.type == I915_GGTT_VIEW_NORMAL &&
- vma->vm == vm)
- return vma;
- }
- return NULL;
-}
-
-struct i915_vma *i915_gem_obj_to_ggtt_view(struct drm_i915_gem_object *obj,
- const struct i915_ggtt_view *view)
-{
- struct i915_vma *vma;
-
- GEM_BUG_ON(!view);
-
- list_for_each_entry(vma, &obj->vma_list, obj_link)
- if (i915_vma_is_ggtt(vma) &&
- i915_ggtt_view_equal(&vma->ggtt_view, view))
- return vma;
- return NULL;
-}
-
int i915_gem_suspend(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
if (ret)
goto err;
- ret = i915_gem_wait_for_idle(dev_priv, true);
+ ret = i915_gem_wait_for_idle(dev_priv,
+ I915_WAIT_INTERRUPTIBLE |
+ I915_WAIT_LOCKED);
if (ret)
goto err;
* guarantee that the context image is complete. So let's just reset
* it and start again.
*/
- if (i915.enable_execlists)
- intel_lr_context_reset(dev_priv, dev_priv->kernel_context);
+ dev_priv->gt.resume(dev_priv);
mutex_unlock(&dev->struct_mutex);
}
mutex_lock(&dev->struct_mutex);
if (!i915.enable_execlists) {
+ dev_priv->gt.resume = intel_legacy_submission_resume;
dev_priv->gt.cleanup_engine = intel_engine_cleanup;
} else {
+ dev_priv->gt.resume = intel_lr_context_resume;
dev_priv->gt.cleanup_engine = intel_logical_ring_cleanup;
}
* for all other failure, such as an allocation failure, bail.
*/
DRM_ERROR("Failed to initialize GPU, declaring it wedged\n");
- atomic_or(I915_WEDGED, &dev_priv->gpu_error.reset_counter);
+ i915_gem_set_wedged(dev_priv);
ret = 0;
}
i915_gem_load_init_fences(struct drm_i915_private *dev_priv)
{
struct drm_device *dev = &dev_priv->drm;
+ int i;
if (INTEL_INFO(dev_priv)->gen >= 7 && !IS_VALLEYVIEW(dev_priv) &&
!IS_CHERRYVIEW(dev_priv))
I915_READ(vgtif_reg(avail_rs.fence_num));
/* Initialize fence registers to zero */
+ for (i = 0; i < dev_priv->num_fence_regs; i++) {
+ struct drm_i915_fence_reg *fence = &dev_priv->fence_regs[i];
+
+ fence->i915 = dev_priv;
+ fence->id = i;
+ list_add_tail(&fence->link, &dev_priv->mm.fence_list);
+ }
i915_gem_restore_fences(dev);
i915_gem_detect_bit_6_swizzle(dev);
INIT_LIST_HEAD(&dev_priv->mm.fence_list);
for (i = 0; i < I915_NUM_ENGINES; i++)
init_engine_lists(&dev_priv->engine[i]);
- for (i = 0; i < I915_MAX_NUM_FENCES; i++)
- INIT_LIST_HEAD(&dev_priv->fence_regs[i].lru_list);
INIT_DELAYED_WORK(&dev_priv->gt.retire_work,
i915_gem_retire_work_handler);
INIT_DELAYED_WORK(&dev_priv->gt.idle_work,
dev_priv->relative_constants_mode = I915_EXEC_CONSTANTS_REL_GENERAL;
- INIT_LIST_HEAD(&dev_priv->mm.fence_list);
-
init_waitqueue_head(&dev_priv->pending_flip_queue);
dev_priv->mm.interruptible = true;
+ atomic_set(&dev_priv->mm.bsd_engine_dispatch_index, 0);
+
spin_lock_init(&dev_priv->fb_tracking.lock);
}
int i915_gem_freeze_late(struct drm_i915_private *dev_priv)
{
struct drm_i915_gem_object *obj;
+ struct list_head *phases[] = {
+ &dev_priv->mm.unbound_list,
+ &dev_priv->mm.bound_list,
+ NULL
+ }, **p;
/* Called just before we write the hibernation image.
*
*
* To make sure the hibernation image contains the latest state,
* we update that state just before writing out the image.
+ *
+ * To try and reduce the hibernation image, we manually shrink
+ * the objects as well.
*/
- list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) {
- obj->base.read_domains = I915_GEM_DOMAIN_CPU;
- obj->base.write_domain = I915_GEM_DOMAIN_CPU;
- }
+ i915_gem_shrink_all(dev_priv);
- list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
- obj->base.read_domains = I915_GEM_DOMAIN_CPU;
- obj->base.write_domain = I915_GEM_DOMAIN_CPU;
+ for (p = phases; *p; p++) {
+ list_for_each_entry(obj, *p, global_list) {
+ obj->base.read_domains = I915_GEM_DOMAIN_CPU;
+ obj->base.write_domain = I915_GEM_DOMAIN_CPU;
+ }
}
return 0;
}
}
-/* All the new VM stuff */
-u64 i915_gem_obj_offset(struct drm_i915_gem_object *o,
- struct i915_address_space *vm)
-{
- struct drm_i915_private *dev_priv = to_i915(o->base.dev);
- struct i915_vma *vma;
-
- WARN_ON(vm == &dev_priv->mm.aliasing_ppgtt->base);
-
- list_for_each_entry(vma, &o->vma_list, obj_link) {
- if (i915_vma_is_ggtt(vma) &&
- vma->ggtt_view.type != I915_GGTT_VIEW_NORMAL)
- continue;
- if (vma->vm == vm)
- return vma->node.start;
- }
-
- WARN(1, "%s vma for this object not found.\n",
- i915_is_ggtt(vm) ? "global" : "ppgtt");
- return -1;
-}
-
-u64 i915_gem_obj_ggtt_offset_view(struct drm_i915_gem_object *o,
- const struct i915_ggtt_view *view)
-{
- struct i915_vma *vma;
-
- list_for_each_entry(vma, &o->vma_list, obj_link)
- if (i915_vma_is_ggtt(vma) &&
- i915_ggtt_view_equal(&vma->ggtt_view, view))
- return vma->node.start;
-
- WARN(1, "global vma for this object not found. (view=%u)\n", view->type);
- return -1;
-}
-
-bool i915_gem_obj_bound(struct drm_i915_gem_object *o,
- struct i915_address_space *vm)
-{
- struct i915_vma *vma;
-
- list_for_each_entry(vma, &o->vma_list, obj_link) {
- if (i915_vma_is_ggtt(vma) &&
- vma->ggtt_view.type != I915_GGTT_VIEW_NORMAL)
- continue;
- if (vma->vm == vm && drm_mm_node_allocated(&vma->node))
- return true;
- }
-
- return false;
-}
-
-bool i915_gem_obj_ggtt_bound_view(struct drm_i915_gem_object *o,
- const struct i915_ggtt_view *view)
-{
- struct i915_vma *vma;
-
- list_for_each_entry(vma, &o->vma_list, obj_link)
- if (i915_vma_is_ggtt(vma) &&
- i915_ggtt_view_equal(&vma->ggtt_view, view) &&
- drm_mm_node_allocated(&vma->node))
- return true;
-
- return false;
-}
-
-unsigned long i915_gem_obj_ggtt_size(struct drm_i915_gem_object *o)
-{
- struct i915_vma *vma;
-
- GEM_BUG_ON(list_empty(&o->vma_list));
-
- list_for_each_entry(vma, &o->vma_list, obj_link) {
- if (i915_vma_is_ggtt(vma) &&
- vma->ggtt_view.type == I915_GGTT_VIEW_NORMAL)
- return vma->node.size;
- }
-
- return 0;
-}
-
-bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj)
-{
- struct i915_vma *vma;
- list_for_each_entry(vma, &obj->vma_list, obj_link)
- if (i915_vma_is_pinned(vma))
- return true;
-
- return false;
-}
-
/* Like i915_gem_object_get_page(), but mark the returned page dirty */
struct page *
i915_gem_object_get_dirty_page(struct drm_i915_gem_object *obj, int n)
if (ce->ring)
intel_ring_free(ce->ring);
- i915_gem_object_put(ce->state);
+ i915_vma_put(ce->state);
}
+ put_pid(ctx->pid);
list_del(&ctx->link);
ida_simple_remove(&ctx->i915->context_hw_ida, ctx->hw_id);
ctx->ggtt_alignment = get_context_alignment(dev_priv);
if (dev_priv->hw_context_size) {
- struct drm_i915_gem_object *obj =
- i915_gem_alloc_context_obj(dev, dev_priv->hw_context_size);
+ struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
+
+ obj = i915_gem_alloc_context_obj(dev,
+ dev_priv->hw_context_size);
if (IS_ERR(obj)) {
ret = PTR_ERR(obj);
goto err_out;
}
- ctx->engine[RCS].state = obj;
+
+ vma = i915_vma_create(obj, &dev_priv->ggtt.base, NULL);
+ if (IS_ERR(vma)) {
+ i915_gem_object_put(obj);
+ ret = PTR_ERR(vma);
+ goto err_out;
+ }
+
+ ctx->engine[RCS].state = vma;
}
/* Default context will never have a file_priv */
ret = DEFAULT_CONTEXT_HANDLE;
ctx->file_priv = file_priv;
+ if (file_priv)
+ ctx->pid = get_task_pid(current, PIDTYPE_PID);
+
ctx->user_handle = ret;
/* NB: Mark all slices as needing a remap so that when the context first
* loads it will restore whatever remap state already exists. If there
struct intel_context *ce = &ctx->engine[engine->id];
if (ce->state)
- i915_gem_object_ggtt_unpin(ce->state);
+ i915_vma_unpin(ce->state);
i915_gem_context_put(ctx);
}
}
-void i915_gem_context_reset(struct drm_device *dev)
-{
- struct drm_i915_private *dev_priv = to_i915(dev);
-
- lockdep_assert_held(&dev->struct_mutex);
-
- if (i915.enable_execlists) {
- struct i915_gem_context *ctx;
-
- list_for_each_entry(ctx, &dev_priv->context_list, link)
- intel_lr_context_reset(dev_priv, ctx);
- }
-
- i915_gem_context_lost(dev_priv);
-}
-
int i915_gem_context_init(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
const int num_rings =
/* Use an extended w/a on ivb+ if signalling from other rings */
i915.semaphores ?
- hweight32(INTEL_INFO(dev_priv)->ring_mask) - 1 :
+ INTEL_INFO(dev_priv)->num_rings - 1 :
0;
int len, ret;
intel_ring_emit(ring, MI_NOOP);
intel_ring_emit(ring, MI_SET_CONTEXT);
intel_ring_emit(ring,
- i915_gem_obj_ggtt_offset(req->ctx->engine[RCS].state) |
- flags);
+ i915_ggtt_offset(req->ctx->engine[RCS].state) | flags);
/*
* w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
* WaMiSetContext_Hang:snb,ivb,vlv
MI_STORE_REGISTER_MEM |
MI_SRM_LRM_GLOBAL_GTT);
intel_ring_emit_reg(ring, last_reg);
- intel_ring_emit(ring, engine->scratch.gtt_offset);
+ intel_ring_emit(ring,
+ i915_ggtt_offset(engine->scratch));
intel_ring_emit(ring, MI_NOOP);
}
intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_ENABLE);
struct i915_gem_context *to = req->ctx;
struct intel_engine_cs *engine = req->engine;
struct i915_hw_ppgtt *ppgtt = to->ppgtt ?: req->i915->mm.aliasing_ppgtt;
+ struct i915_vma *vma = to->engine[RCS].state;
struct i915_gem_context *from;
u32 hw_flags;
int ret, i;
if (skip_rcs_switch(ppgtt, engine, to))
return 0;
+ /* Clear this page out of any CPU caches for coherent swap-in/out. */
+ if (!(vma->flags & I915_VMA_GLOBAL_BIND)) {
+ ret = i915_gem_object_set_to_gtt_domain(vma->obj, false);
+ if (ret)
+ return ret;
+ }
+
/* Trying to pin first makes error handling easier. */
- ret = i915_gem_object_ggtt_pin(to->engine[RCS].state, NULL, 0,
- to->ggtt_alignment, 0);
+ ret = i915_vma_pin(vma, 0, to->ggtt_alignment, PIN_GLOBAL);
if (ret)
return ret;
*/
from = engine->last_context;
- /*
- * Clear this page out of any CPU caches for coherent swap-in/out. Note
- * that thanks to write = false in this call and us not setting any gpu
- * write domains when putting a context object onto the active list
- * (when switching away from it), this won't block.
- *
- * XXX: We need a real interface to do this instead of trickery.
- */
- ret = i915_gem_object_set_to_gtt_domain(to->engine[RCS].state, false);
- if (ret)
- goto unpin_out;
-
if (needs_pd_load_pre(ppgtt, engine, to)) {
/* Older GENs and non render rings still want the load first,
* "PP_DCLV followed by PP_DIR_BASE register through Load
trace_switch_mm(engine, to);
ret = ppgtt->switch_mm(ppgtt, req);
if (ret)
- goto unpin_out;
+ goto err;
}
if (!to->engine[RCS].initialised || i915_gem_context_is_default(to))
if (to != from || (hw_flags & MI_FORCE_RESTORE)) {
ret = mi_set_context(req, hw_flags);
if (ret)
- goto unpin_out;
+ goto err;
}
/* The backing object for the context is done after switching to the
* MI_SET_CONTEXT instead of when the next seqno has completed.
*/
if (from != NULL) {
- struct drm_i915_gem_object *obj = from->engine[RCS].state;
-
/* As long as MI_SET_CONTEXT is serializing, ie. it flushes the
* whole damn pipeline, we don't need to explicitly mark the
* object dirty. The only exception is that the context must be
* able to defer doing this until we know the object would be
* swapped, but there is no way to do that yet.
*/
- obj->base.read_domains = I915_GEM_DOMAIN_INSTRUCTION;
- i915_vma_move_to_active(i915_gem_obj_to_ggtt(obj), req, 0);
-
- /* obj is kept alive until the next request by its active ref */
- i915_gem_object_ggtt_unpin(obj);
+ i915_vma_move_to_active(from->engine[RCS].state, req, 0);
+ /* state is kept alive until the next request */
+ i915_vma_unpin(from->engine[RCS].state);
i915_gem_context_put(from);
}
engine->last_context = i915_gem_context_get(to);
return 0;
-unpin_out:
- i915_gem_object_ggtt_unpin(to->engine[RCS].state);
+err:
+ i915_vma_unpin(vma);
return ret;
}
if (ret)
return ERR_PTR(ret);
- addr = i915_gem_object_pin_map(obj);
+ addr = i915_gem_object_pin_map(obj, I915_MAP_WB);
mutex_unlock(&dev->struct_mutex);
return addr;
}
static bool
-mark_free(struct i915_vma *vma, struct list_head *unwind)
+mark_free(struct i915_vma *vma, unsigned int flags, struct list_head *unwind)
{
if (i915_vma_is_pinned(vma))
return false;
if (WARN_ON(!list_empty(&vma->exec_list)))
return false;
+ if (flags & PIN_NONFAULT && vma->obj->fault_mappable)
+ return false;
+
list_add(&vma->exec_list, unwind);
return drm_mm_scan_add_block(&vma->node);
}
phase = phases;
do {
list_for_each_entry(vma, *phase, vm_link)
- if (mark_free(vma, &eviction_list))
+ if (mark_free(vma, flags, &eviction_list))
goto found;
} while (*++phase);
if (ret)
return ret;
- ret = i915_gem_wait_for_idle(dev_priv, true);
+ ret = i915_gem_wait_for_idle(dev_priv,
+ I915_WAIT_INTERRUPTIBLE |
+ I915_WAIT_LOCKED);
if (ret)
return ret;
return ret;
}
- ret = i915_gem_wait_for_idle(dev_priv, true);
+ ret = i915_gem_wait_for_idle(dev_priv,
+ I915_WAIT_INTERRUPTIBLE |
+ I915_WAIT_LOCKED);
if (ret)
return ret;
#include "intel_drv.h"
#include "intel_frontbuffer.h"
+#define DBG_USE_CPU_RELOC 0 /* -1 force GTT relocs; 1 force CPU relocs */
+
#define __EXEC_OBJECT_HAS_PIN (1<<31)
#define __EXEC_OBJECT_HAS_FENCE (1<<30)
#define __EXEC_OBJECT_NEEDS_MAP (1<<29)
};
struct eb_vmas {
+ struct drm_i915_private *i915;
struct list_head vmas;
int and;
union {
};
static struct eb_vmas *
-eb_create(struct drm_i915_gem_execbuffer2 *args)
+eb_create(struct drm_i915_private *i915,
+ struct drm_i915_gem_execbuffer2 *args)
{
struct eb_vmas *eb = NULL;
} else
eb->and = -args->buffer_count;
+ eb->i915 = i915;
INIT_LIST_HEAD(&eb->vmas);
return eb;
}
* from the (obj, vm) we don't run the risk of creating
* duplicated vmas for the same vm.
*/
- vma = i915_gem_obj_lookup_or_create_vma(obj, vm);
- if (IS_ERR(vma)) {
+ vma = i915_gem_obj_lookup_or_create_vma(obj, vm, NULL);
+ if (unlikely(IS_ERR(vma))) {
DRM_DEBUG("Failed to lookup VMA\n");
ret = PTR_ERR(vma);
goto err;
i915_gem_execbuffer_unreserve_vma(struct i915_vma *vma)
{
struct drm_i915_gem_exec_object2 *entry;
- struct drm_i915_gem_object *obj = vma->obj;
if (!drm_mm_node_allocated(&vma->node))
return;
entry = vma->exec_entry;
if (entry->flags & __EXEC_OBJECT_HAS_FENCE)
- i915_gem_object_unpin_fence(obj);
+ i915_vma_unpin_fence(vma);
if (entry->flags & __EXEC_OBJECT_HAS_PIN)
__i915_vma_unpin(vma);
exec_list);
list_del_init(&vma->exec_list);
i915_gem_execbuffer_unreserve_vma(vma);
- i915_gem_object_put(vma->obj);
+ i915_vma_put(vma);
}
kfree(eb);
}
static inline int use_cpu_reloc(struct drm_i915_gem_object *obj)
{
+ if (!i915_gem_object_has_struct_page(obj))
+ return false;
+
+ if (DBG_USE_CPU_RELOC)
+ return DBG_USE_CPU_RELOC > 0;
+
return (HAS_LLC(obj->base.dev) ||
obj->base.write_domain == I915_GEM_DOMAIN_CPU ||
obj->cache_level != I915_CACHE_NONE);
}
static inline uint64_t
-relocation_target(struct drm_i915_gem_relocation_entry *reloc,
+relocation_target(const struct drm_i915_gem_relocation_entry *reloc,
uint64_t target_offset)
{
return gen8_canonical_addr((int)reloc->delta + target_offset);
}
-static int
-relocate_entry_cpu(struct drm_i915_gem_object *obj,
- struct drm_i915_gem_relocation_entry *reloc,
- uint64_t target_offset)
+struct reloc_cache {
+ struct drm_i915_private *i915;
+ struct drm_mm_node node;
+ unsigned long vaddr;
+ unsigned int page;
+ bool use_64bit_reloc;
+};
+
+static void reloc_cache_init(struct reloc_cache *cache,
+ struct drm_i915_private *i915)
{
- struct drm_device *dev = obj->base.dev;
- uint32_t page_offset = offset_in_page(reloc->offset);
- uint64_t delta = relocation_target(reloc, target_offset);
- char *vaddr;
- int ret;
+ cache->page = -1;
+ cache->vaddr = 0;
+ cache->i915 = i915;
+ cache->use_64bit_reloc = INTEL_GEN(cache->i915) >= 8;
+ cache->node.allocated = false;
+}
- ret = i915_gem_object_set_to_cpu_domain(obj, true);
- if (ret)
- return ret;
+static inline void *unmask_page(unsigned long p)
+{
+ return (void *)(uintptr_t)(p & PAGE_MASK);
+}
+
+static inline unsigned int unmask_flags(unsigned long p)
+{
+ return p & ~PAGE_MASK;
+}
- vaddr = kmap_atomic(i915_gem_object_get_dirty_page(obj,
- reloc->offset >> PAGE_SHIFT));
- *(uint32_t *)(vaddr + page_offset) = lower_32_bits(delta);
+#define KMAP 0x4 /* after CLFLUSH_FLAGS */
- if (INTEL_INFO(dev)->gen >= 8) {
- page_offset = offset_in_page(page_offset + sizeof(uint32_t));
+static void reloc_cache_fini(struct reloc_cache *cache)
+{
+ void *vaddr;
+
+ if (!cache->vaddr)
+ return;
- if (page_offset == 0) {
- kunmap_atomic(vaddr);
- vaddr = kmap_atomic(i915_gem_object_get_dirty_page(obj,
- (reloc->offset + sizeof(uint32_t)) >> PAGE_SHIFT));
+ vaddr = unmask_page(cache->vaddr);
+ if (cache->vaddr & KMAP) {
+ if (cache->vaddr & CLFLUSH_AFTER)
+ mb();
+
+ kunmap_atomic(vaddr);
+ i915_gem_obj_finish_shmem_access((struct drm_i915_gem_object *)cache->node.mm);
+ } else {
+ wmb();
+ io_mapping_unmap_atomic((void __iomem *)vaddr);
+ if (cache->node.allocated) {
+ struct i915_ggtt *ggtt = &cache->i915->ggtt;
+
+ ggtt->base.clear_range(&ggtt->base,
+ cache->node.start,
+ cache->node.size,
+ true);
+ drm_mm_remove_node(&cache->node);
+ } else {
+ i915_vma_unpin((struct i915_vma *)cache->node.mm);
}
+ }
+}
+
+static void *reloc_kmap(struct drm_i915_gem_object *obj,
+ struct reloc_cache *cache,
+ int page)
+{
+ void *vaddr;
+
+ if (cache->vaddr) {
+ kunmap_atomic(unmask_page(cache->vaddr));
+ } else {
+ unsigned int flushes;
+ int ret;
+
+ ret = i915_gem_obj_prepare_shmem_write(obj, &flushes);
+ if (ret)
+ return ERR_PTR(ret);
- *(uint32_t *)(vaddr + page_offset) = upper_32_bits(delta);
+ BUILD_BUG_ON(KMAP & CLFLUSH_FLAGS);
+ BUILD_BUG_ON((KMAP | CLFLUSH_FLAGS) & PAGE_MASK);
+
+ cache->vaddr = flushes | KMAP;
+ cache->node.mm = (void *)obj;
+ if (flushes)
+ mb();
}
- kunmap_atomic(vaddr);
+ vaddr = kmap_atomic(i915_gem_object_get_dirty_page(obj, page));
+ cache->vaddr = unmask_flags(cache->vaddr) | (unsigned long)vaddr;
+ cache->page = page;
- return 0;
+ return vaddr;
}
-static int
-relocate_entry_gtt(struct drm_i915_gem_object *obj,
- struct drm_i915_gem_relocation_entry *reloc,
- uint64_t target_offset)
+static void *reloc_iomap(struct drm_i915_gem_object *obj,
+ struct reloc_cache *cache,
+ int page)
{
- struct drm_device *dev = obj->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
- struct i915_ggtt *ggtt = &dev_priv->ggtt;
- uint64_t delta = relocation_target(reloc, target_offset);
- uint64_t offset;
- void __iomem *reloc_page;
- int ret;
+ struct i915_ggtt *ggtt = &cache->i915->ggtt;
+ unsigned long offset;
+ void *vaddr;
- ret = i915_gem_object_set_to_gtt_domain(obj, true);
- if (ret)
- return ret;
+ if (cache->node.allocated) {
+ wmb();
+ ggtt->base.insert_page(&ggtt->base,
+ i915_gem_object_get_dma_address(obj, page),
+ cache->node.start, I915_CACHE_NONE, 0);
+ cache->page = page;
+ return unmask_page(cache->vaddr);
+ }
- ret = i915_gem_object_put_fence(obj);
- if (ret)
- return ret;
+ if (cache->vaddr) {
+ io_mapping_unmap_atomic(unmask_page(cache->vaddr));
+ } else {
+ struct i915_vma *vma;
+ int ret;
+
+ if (use_cpu_reloc(obj))
+ return NULL;
- /* Map the page containing the relocation we're going to perform. */
- offset = i915_gem_obj_ggtt_offset(obj);
- offset += reloc->offset;
- reloc_page = io_mapping_map_atomic_wc(ggtt->mappable,
- offset & PAGE_MASK);
- iowrite32(lower_32_bits(delta), reloc_page + offset_in_page(offset));
-
- if (INTEL_INFO(dev)->gen >= 8) {
- offset += sizeof(uint32_t);
-
- if (offset_in_page(offset) == 0) {
- io_mapping_unmap_atomic(reloc_page);
- reloc_page =
- io_mapping_map_atomic_wc(ggtt->mappable,
- offset);
+ ret = i915_gem_object_set_to_gtt_domain(obj, true);
+ if (ret)
+ return ERR_PTR(ret);
+
+ vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0,
+ PIN_MAPPABLE | PIN_NONBLOCK);
+ if (IS_ERR(vma)) {
+ memset(&cache->node, 0, sizeof(cache->node));
+ ret = drm_mm_insert_node_in_range_generic
+ (&ggtt->base.mm, &cache->node,
+ 4096, 0, 0,
+ 0, ggtt->mappable_end,
+ DRM_MM_SEARCH_DEFAULT,
+ DRM_MM_CREATE_DEFAULT);
+ if (ret)
+ return ERR_PTR(ret);
+ } else {
+ ret = i915_vma_put_fence(vma);
+ if (ret) {
+ i915_vma_unpin(vma);
+ return ERR_PTR(ret);
+ }
+
+ cache->node.start = vma->node.start;
+ cache->node.mm = (void *)vma;
}
+ }
- iowrite32(upper_32_bits(delta),
- reloc_page + offset_in_page(offset));
+ offset = cache->node.start;
+ if (cache->node.allocated) {
+ ggtt->base.insert_page(&ggtt->base,
+ i915_gem_object_get_dma_address(obj, page),
+ offset, I915_CACHE_NONE, 0);
+ } else {
+ offset += page << PAGE_SHIFT;
}
- io_mapping_unmap_atomic(reloc_page);
+ vaddr = io_mapping_map_atomic_wc(&cache->i915->ggtt.mappable, offset);
+ cache->page = page;
+ cache->vaddr = (unsigned long)vaddr;
- return 0;
+ return vaddr;
}
-static void
-clflush_write32(void *addr, uint32_t value)
+static void *reloc_vaddr(struct drm_i915_gem_object *obj,
+ struct reloc_cache *cache,
+ int page)
{
- /* This is not a fast path, so KISS. */
- drm_clflush_virt_range(addr, sizeof(uint32_t));
- *(uint32_t *)addr = value;
- drm_clflush_virt_range(addr, sizeof(uint32_t));
+ void *vaddr;
+
+ if (cache->page == page) {
+ vaddr = unmask_page(cache->vaddr);
+ } else {
+ vaddr = NULL;
+ if ((cache->vaddr & KMAP) == 0)
+ vaddr = reloc_iomap(obj, cache, page);
+ if (!vaddr)
+ vaddr = reloc_kmap(obj, cache, page);
+ }
+
+ return vaddr;
}
-static int
-relocate_entry_clflush(struct drm_i915_gem_object *obj,
- struct drm_i915_gem_relocation_entry *reloc,
- uint64_t target_offset)
+static void clflush_write32(u32 *addr, u32 value, unsigned int flushes)
{
- struct drm_device *dev = obj->base.dev;
- uint32_t page_offset = offset_in_page(reloc->offset);
- uint64_t delta = relocation_target(reloc, target_offset);
- char *vaddr;
- int ret;
+ if (unlikely(flushes & (CLFLUSH_BEFORE | CLFLUSH_AFTER))) {
+ if (flushes & CLFLUSH_BEFORE) {
+ clflushopt(addr);
+ mb();
+ }
- ret = i915_gem_object_set_to_gtt_domain(obj, true);
- if (ret)
- return ret;
+ *addr = value;
- vaddr = kmap_atomic(i915_gem_object_get_dirty_page(obj,
- reloc->offset >> PAGE_SHIFT));
- clflush_write32(vaddr + page_offset, lower_32_bits(delta));
+ /* Writes to the same cacheline are serialised by the CPU
+ * (including clflush). On the write path, we only require
+ * that it hits memory in an orderly fashion and place
+ * mb barriers at the start and end of the relocation phase
+ * to ensure ordering of clflush wrt to the system.
+ */
+ if (flushes & CLFLUSH_AFTER)
+ clflushopt(addr);
+ } else
+ *addr = value;
+}
- if (INTEL_INFO(dev)->gen >= 8) {
- page_offset = offset_in_page(page_offset + sizeof(uint32_t));
+static int
+relocate_entry(struct drm_i915_gem_object *obj,
+ const struct drm_i915_gem_relocation_entry *reloc,
+ struct reloc_cache *cache,
+ u64 target_offset)
+{
+ u64 offset = reloc->offset;
+ bool wide = cache->use_64bit_reloc;
+ void *vaddr;
- if (page_offset == 0) {
- kunmap_atomic(vaddr);
- vaddr = kmap_atomic(i915_gem_object_get_dirty_page(obj,
- (reloc->offset + sizeof(uint32_t)) >> PAGE_SHIFT));
- }
+ target_offset = relocation_target(reloc, target_offset);
+repeat:
+ vaddr = reloc_vaddr(obj, cache, offset >> PAGE_SHIFT);
+ if (IS_ERR(vaddr))
+ return PTR_ERR(vaddr);
- clflush_write32(vaddr + page_offset, upper_32_bits(delta));
- }
+ clflush_write32(vaddr + offset_in_page(offset),
+ lower_32_bits(target_offset),
+ cache->vaddr);
- kunmap_atomic(vaddr);
+ if (wide) {
+ offset += sizeof(u32);
+ target_offset >>= 32;
+ wide = false;
+ goto repeat;
+ }
return 0;
}
static int
i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj,
struct eb_vmas *eb,
- struct drm_i915_gem_relocation_entry *reloc)
+ struct drm_i915_gem_relocation_entry *reloc,
+ struct reloc_cache *cache)
{
struct drm_device *dev = obj->base.dev;
struct drm_gem_object *target_obj;
/* Check that the relocation address is valid... */
if (unlikely(reloc->offset >
- obj->base.size - (INTEL_INFO(dev)->gen >= 8 ? 8 : 4))) {
+ obj->base.size - (cache->use_64bit_reloc ? 8 : 4))) {
DRM_DEBUG("Relocation beyond object bounds: "
"obj %p target %d offset %d size %d.\n",
obj, reloc->target_handle,
if (pagefault_disabled() && !object_is_idle(obj))
return -EFAULT;
- if (use_cpu_reloc(obj))
- ret = relocate_entry_cpu(obj, reloc, target_offset);
- else if (obj->map_and_fenceable)
- ret = relocate_entry_gtt(obj, reloc, target_offset);
- else if (static_cpu_has(X86_FEATURE_CLFLUSH))
- ret = relocate_entry_clflush(obj, reloc, target_offset);
- else {
- WARN_ONCE(1, "Impossible case in relocation handling\n");
- ret = -ENODEV;
- }
-
+ ret = relocate_entry(obj, reloc, cache, target_offset);
if (ret)
return ret;
/* and update the user's relocation entry */
reloc->presumed_offset = target_offset;
-
return 0;
}
struct drm_i915_gem_relocation_entry stack_reloc[N_RELOC(512)];
struct drm_i915_gem_relocation_entry __user *user_relocs;
struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
- int remain, ret;
+ struct reloc_cache cache;
+ int remain, ret = 0;
user_relocs = u64_to_user_ptr(entry->relocs_ptr);
+ reloc_cache_init(&cache, eb->i915);
remain = entry->relocation_count;
while (remain) {
count = ARRAY_SIZE(stack_reloc);
remain -= count;
- if (__copy_from_user_inatomic(r, user_relocs, count*sizeof(r[0])))
- return -EFAULT;
+ if (__copy_from_user_inatomic(r, user_relocs, count*sizeof(r[0]))) {
+ ret = -EFAULT;
+ goto out;
+ }
do {
u64 offset = r->presumed_offset;
- ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, r);
+ ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, r, &cache);
if (ret)
- return ret;
+ goto out;
if (r->presumed_offset != offset &&
- __put_user(r->presumed_offset, &user_relocs->presumed_offset)) {
- return -EFAULT;
+ __put_user(r->presumed_offset,
+ &user_relocs->presumed_offset)) {
+ ret = -EFAULT;
+ goto out;
}
user_relocs++;
} while (--count);
}
- return 0;
+out:
+ reloc_cache_fini(&cache);
+ return ret;
#undef N_RELOC
}
struct drm_i915_gem_relocation_entry *relocs)
{
const struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
- int i, ret;
+ struct reloc_cache cache;
+ int i, ret = 0;
+ reloc_cache_init(&cache, eb->i915);
for (i = 0; i < entry->relocation_count; i++) {
- ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, &relocs[i]);
+ ret = i915_gem_execbuffer_relocate_entry(vma->obj, eb, &relocs[i], &cache);
if (ret)
- return ret;
+ break;
}
+ reloc_cache_fini(&cache);
- return 0;
+ return ret;
}
static int
entry->flags |= __EXEC_OBJECT_HAS_PIN;
if (entry->flags & EXEC_OBJECT_NEEDS_FENCE) {
- ret = i915_gem_object_get_fence(obj);
+ ret = i915_vma_get_fence(vma);
if (ret)
return ret;
- if (i915_gem_object_pin_fence(obj))
+ if (i915_vma_pin_fence(vma))
entry->flags |= __EXEC_OBJECT_HAS_FENCE;
}
eb_vma_misplaced(struct i915_vma *vma)
{
struct drm_i915_gem_exec_object2 *entry = vma->exec_entry;
- struct drm_i915_gem_object *obj = vma->obj;
WARN_ON(entry->flags & __EXEC_OBJECT_NEEDS_MAP &&
!i915_vma_is_ggtt(vma));
return true;
/* avoid costly ping-pong once a batch bo ended up non-mappable */
- if (entry->flags & __EXEC_OBJECT_NEEDS_MAP && !obj->map_and_fenceable)
+ if (entry->flags & __EXEC_OBJECT_NEEDS_MAP &&
+ !i915_vma_is_map_and_fenceable(vma))
return !only_mappable_for_reloc(entry->flags);
if ((entry->flags & EXEC_OBJECT_SUPPORTS_48B_ADDRESS) == 0 &&
vma = list_first_entry(&eb->vmas, struct i915_vma, exec_list);
list_del_init(&vma->exec_list);
i915_gem_execbuffer_unreserve_vma(vma);
- i915_gem_object_put(vma->obj);
+ i915_vma_put(vma);
}
mutex_unlock(&dev->struct_mutex);
{
const unsigned int other_rings = eb_other_engines(req);
struct i915_vma *vma;
- uint32_t flush_domains = 0;
- bool flush_chipset = false;
int ret;
list_for_each_entry(vma, vmas, exec_list) {
struct drm_i915_gem_object *obj = vma->obj;
+ struct reservation_object *resv;
if (obj->flags & other_rings) {
- ret = i915_gem_object_sync(obj, req);
+ ret = i915_gem_request_await_object
+ (req, obj, obj->base.pending_write_domain);
if (ret)
return ret;
}
- if (obj->base.write_domain & I915_GEM_DOMAIN_CPU)
- flush_chipset |= i915_gem_clflush_object(obj, false);
+ resv = i915_gem_object_get_dmabuf_resv(obj);
+ if (resv) {
+ ret = i915_sw_fence_await_reservation
+ (&req->submit, resv, &i915_fence_ops,
+ obj->base.pending_write_domain, 10*HZ,
+ GFP_KERNEL | __GFP_NOWARN);
+ if (ret < 0)
+ return ret;
+ }
- flush_domains |= obj->base.write_domain;
+ if (obj->base.write_domain & I915_GEM_DOMAIN_CPU)
+ i915_gem_clflush_object(obj, false);
}
- if (flush_chipset)
- i915_gem_chipset_flush(req->engine->i915);
-
- if (flush_domains & I915_GEM_DOMAIN_GTT)
- wmb();
+ /* Unconditionally flush any chipset caches (for streaming writes). */
+ i915_gem_chipset_flush(req->engine->i915);
/* Unconditionally invalidate GPU caches and TLBs. */
return req->engine->emit_flush(req, EMIT_INVALIDATE);
i915_gem_validate_context(struct drm_device *dev, struct drm_file *file,
struct intel_engine_cs *engine, const u32 ctx_id)
{
- struct i915_gem_context *ctx = NULL;
+ struct i915_gem_context *ctx;
struct i915_ctx_hang_stats *hs;
- if (engine->id != RCS && ctx_id != DEFAULT_CONTEXT_HANDLE)
- return ERR_PTR(-EINVAL);
-
ctx = i915_gem_context_lookup(file->driver_priv, ctx_id);
if (IS_ERR(ctx))
return ctx;
obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
}
- if (flags & EXEC_OBJECT_NEEDS_FENCE) {
- i915_gem_active_set(&obj->last_fence, req);
- if (flags & __EXEC_OBJECT_HAS_FENCE) {
- struct drm_i915_private *dev_priv = req->i915;
-
- list_move_tail(&dev_priv->fence_regs[obj->fence_reg].lru_list,
- &dev_priv->mm.fence_list);
- }
- }
+ if (flags & EXEC_OBJECT_NEEDS_FENCE)
+ i915_gem_active_set(&vma->last_fence, req);
i915_vma_set_active(vma, idx);
i915_gem_active_set(&vma->last_read[idx], req);
return 0;
}
-static struct i915_vma*
+static struct i915_vma *
i915_gem_execbuffer_parse(struct intel_engine_cs *engine,
struct drm_i915_gem_exec_object2 *shadow_exec_entry,
struct drm_i915_gem_object *batch_obj,
batch_start_offset,
batch_len,
is_master);
- if (ret)
- goto err;
-
- ret = i915_gem_object_ggtt_pin(shadow_batch_obj, NULL, 0, 0, 0);
- if (ret)
- goto err;
+ if (ret) {
+ if (ret == -EACCES) /* unhandled chained batch */
+ vma = NULL;
+ else
+ vma = ERR_PTR(ret);
+ goto out;
+ }
- i915_gem_object_unpin_pages(shadow_batch_obj);
+ vma = i915_gem_object_ggtt_pin(shadow_batch_obj, NULL, 0, 0, 0);
+ if (IS_ERR(vma))
+ goto out;
memset(shadow_exec_entry, 0, sizeof(*shadow_exec_entry));
- vma = i915_gem_obj_to_ggtt(shadow_batch_obj);
vma->exec_entry = shadow_exec_entry;
vma->exec_entry->flags = __EXEC_OBJECT_HAS_PIN;
i915_gem_object_get(shadow_batch_obj);
list_add_tail(&vma->exec_list, &eb->vmas);
- return vma;
-
-err:
+out:
i915_gem_object_unpin_pages(shadow_batch_obj);
- if (ret == -EACCES) /* unhandled chained batch */
- return NULL;
- else
- return ERR_PTR(ret);
+ return vma;
}
static int
params->args_batch_start_offset;
if (exec_len == 0)
- exec_len = params->batch->size;
+ exec_len = params->batch->size - params->args_batch_start_offset;
ret = params->engine->emit_bb_start(params->request,
exec_start, exec_len,
struct drm_i915_file_private *file_priv = file->driver_priv;
/* Check whether the file_priv has already selected one ring. */
- if ((int)file_priv->bsd_engine < 0) {
- /* If not, use the ping-pong mechanism to select one. */
- mutex_lock(&dev_priv->drm.struct_mutex);
- file_priv->bsd_engine = dev_priv->mm.bsd_engine_dispatch_index;
- dev_priv->mm.bsd_engine_dispatch_index ^= 1;
- mutex_unlock(&dev_priv->drm.struct_mutex);
- }
+ if ((int)file_priv->bsd_engine < 0)
+ file_priv->bsd_engine = atomic_fetch_xor(1,
+ &dev_priv->mm.bsd_engine_dispatch_index);
return file_priv->bsd_engine;
}
memset(¶ms_master, 0x00, sizeof(params_master));
- eb = eb_create(args);
+ eb = eb_create(dev_priv, args);
if (eb == NULL) {
i915_gem_context_put(ctx);
mutex_unlock(&dev->struct_mutex);
ret = -EINVAL;
goto err;
}
+ if (args->batch_start_offset > params->batch->size ||
+ args->batch_len > params->batch->size - args->batch_start_offset) {
+ DRM_DEBUG("Attempting to use out-of-bounds batch\n");
+ ret = -EINVAL;
+ goto err;
+ }
params->args_batch_start_offset = args->batch_start_offset;
if (intel_engine_needs_cmd_parser(engine) && args->batch_len) {
* hsw should have this fixed, but bdw mucks it up again. */
if (dispatch_flags & I915_DISPATCH_SECURE) {
struct drm_i915_gem_object *obj = params->batch->obj;
+ struct i915_vma *vma;
/*
* So on first glance it looks freaky that we pin the batch here
* fitting due to fragmentation.
* So this is actually safe.
*/
- ret = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, 0);
- if (ret)
+ vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, 0);
+ if (IS_ERR(vma)) {
+ ret = PTR_ERR(vma);
goto err;
+ }
- params->batch = i915_gem_obj_to_ggtt(obj);
+ params->batch = vma;
}
/* Allocate a request for this batch buffer nice and early. */
goto err_batch_unpin;
}
+ /* Whilst this request exists, batch_obj will be on the
+ * active_list, and so will hold the active reference. Only when this
+ * request is retired will the the batch_obj be moved onto the
+ * inactive_list and lose its active reference. Hence we do not need
+ * to explicitly hold another reference here.
+ */
+ params->request->batch = params->batch;
+
ret = i915_gem_request_add_to_client(params->request, file);
if (ret)
goto err_request;
ret = execbuf_submit(params, args, &eb->vmas);
err_request:
- __i915_add_request(params->request, params->batch->obj, ret == 0);
+ __i915_add_request(params->request, ret == 0);
err_batch_unpin:
/*
* CPU ptes into GTT mmaps (not the GTT ptes themselves) as needed.
*/
-static void i965_write_fence_reg(struct drm_device *dev, int reg,
- struct drm_i915_gem_object *obj)
+#define pipelined 0
+
+static void i965_write_fence_reg(struct drm_i915_fence_reg *fence,
+ struct i915_vma *vma)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
i915_reg_t fence_reg_lo, fence_reg_hi;
int fence_pitch_shift;
+ u64 val;
- if (INTEL_INFO(dev)->gen >= 6) {
- fence_reg_lo = FENCE_REG_GEN6_LO(reg);
- fence_reg_hi = FENCE_REG_GEN6_HI(reg);
+ if (INTEL_INFO(fence->i915)->gen >= 6) {
+ fence_reg_lo = FENCE_REG_GEN6_LO(fence->id);
+ fence_reg_hi = FENCE_REG_GEN6_HI(fence->id);
fence_pitch_shift = GEN6_FENCE_PITCH_SHIFT;
+
} else {
- fence_reg_lo = FENCE_REG_965_LO(reg);
- fence_reg_hi = FENCE_REG_965_HI(reg);
+ fence_reg_lo = FENCE_REG_965_LO(fence->id);
+ fence_reg_hi = FENCE_REG_965_HI(fence->id);
fence_pitch_shift = I965_FENCE_PITCH_SHIFT;
}
- /* To w/a incoherency with non-atomic 64-bit register updates,
- * we split the 64-bit update into two 32-bit writes. In order
- * for a partial fence not to be evaluated between writes, we
- * precede the update with write to turn off the fence register,
- * and only enable the fence as the last step.
- *
- * For extra levels of paranoia, we make sure each step lands
- * before applying the next step.
- */
- I915_WRITE(fence_reg_lo, 0);
- POSTING_READ(fence_reg_lo);
-
- if (obj) {
- u32 size = i915_gem_obj_ggtt_size(obj);
- unsigned int tiling = i915_gem_object_get_tiling(obj);
- unsigned int stride = i915_gem_object_get_stride(obj);
- uint64_t val;
-
- /* Adjust fence size to match tiled area */
- if (tiling != I915_TILING_NONE) {
- uint32_t row_size = stride *
- (tiling == I915_TILING_Y ? 32 : 8);
- size = (size / row_size) * row_size;
- }
-
- val = (uint64_t)((i915_gem_obj_ggtt_offset(obj) + size - 4096) &
- 0xfffff000) << 32;
- val |= i915_gem_obj_ggtt_offset(obj) & 0xfffff000;
- val |= (uint64_t)((stride / 128) - 1) << fence_pitch_shift;
- if (tiling == I915_TILING_Y)
- val |= 1 << I965_FENCE_TILING_Y_SHIFT;
+ val = 0;
+ if (vma) {
+ unsigned int tiling = i915_gem_object_get_tiling(vma->obj);
+ bool is_y_tiled = tiling == I915_TILING_Y;
+ unsigned int stride = i915_gem_object_get_stride(vma->obj);
+ u32 row_size = stride * (is_y_tiled ? 32 : 8);
+ u32 size = rounddown((u32)vma->node.size, row_size);
+
+ val = ((vma->node.start + size - 4096) & 0xfffff000) << 32;
+ val |= vma->node.start & 0xfffff000;
+ val |= (u64)((stride / 128) - 1) << fence_pitch_shift;
+ if (is_y_tiled)
+ val |= BIT(I965_FENCE_TILING_Y_SHIFT);
val |= I965_FENCE_REG_VALID;
+ }
- I915_WRITE(fence_reg_hi, val >> 32);
- POSTING_READ(fence_reg_hi);
+ if (!pipelined) {
+ struct drm_i915_private *dev_priv = fence->i915;
- I915_WRITE(fence_reg_lo, val);
+ /* To w/a incoherency with non-atomic 64-bit register updates,
+ * we split the 64-bit update into two 32-bit writes. In order
+ * for a partial fence not to be evaluated between writes, we
+ * precede the update with write to turn off the fence register,
+ * and only enable the fence as the last step.
+ *
+ * For extra levels of paranoia, we make sure each step lands
+ * before applying the next step.
+ */
+ I915_WRITE(fence_reg_lo, 0);
+ POSTING_READ(fence_reg_lo);
+
+ I915_WRITE(fence_reg_hi, upper_32_bits(val));
+ I915_WRITE(fence_reg_lo, lower_32_bits(val));
POSTING_READ(fence_reg_lo);
- } else {
- I915_WRITE(fence_reg_hi, 0);
- POSTING_READ(fence_reg_hi);
}
}
-static void i915_write_fence_reg(struct drm_device *dev, int reg,
- struct drm_i915_gem_object *obj)
+static void i915_write_fence_reg(struct drm_i915_fence_reg *fence,
+ struct i915_vma *vma)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
u32 val;
- if (obj) {
- u32 size = i915_gem_obj_ggtt_size(obj);
- unsigned int tiling = i915_gem_object_get_tiling(obj);
- unsigned int stride = i915_gem_object_get_stride(obj);
+ val = 0;
+ if (vma) {
+ unsigned int tiling = i915_gem_object_get_tiling(vma->obj);
+ bool is_y_tiled = tiling == I915_TILING_Y;
+ unsigned int stride = i915_gem_object_get_stride(vma->obj);
int pitch_val;
int tile_width;
- WARN((i915_gem_obj_ggtt_offset(obj) & ~I915_FENCE_START_MASK) ||
- (size & -size) != size ||
- (i915_gem_obj_ggtt_offset(obj) & (size - 1)),
- "object 0x%08llx [fenceable? %d] not 1M or pot-size (0x%08x) aligned\n",
- i915_gem_obj_ggtt_offset(obj), obj->map_and_fenceable, size);
+ WARN((vma->node.start & ~I915_FENCE_START_MASK) ||
+ !is_power_of_2(vma->node.size) ||
+ (vma->node.start & (vma->node.size - 1)),
+ "object 0x%08llx [fenceable? %d] not 1M or pot-size (0x%08llx) aligned\n",
+ vma->node.start,
+ i915_vma_is_map_and_fenceable(vma),
+ vma->node.size);
- if (tiling == I915_TILING_Y && HAS_128_BYTE_Y_TILING(dev))
+ if (is_y_tiled && HAS_128_BYTE_Y_TILING(fence->i915))
tile_width = 128;
else
tile_width = 512;
pitch_val = stride / tile_width;
pitch_val = ffs(pitch_val) - 1;
- val = i915_gem_obj_ggtt_offset(obj);
- if (tiling == I915_TILING_Y)
- val |= 1 << I830_FENCE_TILING_Y_SHIFT;
- val |= I915_FENCE_SIZE_BITS(size);
+ val = vma->node.start;
+ if (is_y_tiled)
+ val |= BIT(I830_FENCE_TILING_Y_SHIFT);
+ val |= I915_FENCE_SIZE_BITS(vma->node.size);
val |= pitch_val << I830_FENCE_PITCH_SHIFT;
val |= I830_FENCE_REG_VALID;
- } else
- val = 0;
+ }
+
+ if (!pipelined) {
+ struct drm_i915_private *dev_priv = fence->i915;
+ i915_reg_t reg = FENCE_REG(fence->id);
- I915_WRITE(FENCE_REG(reg), val);
- POSTING_READ(FENCE_REG(reg));
+ I915_WRITE(reg, val);
+ POSTING_READ(reg);
+ }
}
-static void i830_write_fence_reg(struct drm_device *dev, int reg,
- struct drm_i915_gem_object *obj)
+static void i830_write_fence_reg(struct drm_i915_fence_reg *fence,
+ struct i915_vma *vma)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
- uint32_t val;
+ u32 val;
- if (obj) {
- u32 size = i915_gem_obj_ggtt_size(obj);
- unsigned int tiling = i915_gem_object_get_tiling(obj);
- unsigned int stride = i915_gem_object_get_stride(obj);
- uint32_t pitch_val;
+ val = 0;
+ if (vma) {
+ unsigned int tiling = i915_gem_object_get_tiling(vma->obj);
+ bool is_y_tiled = tiling == I915_TILING_Y;
+ unsigned int stride = i915_gem_object_get_stride(vma->obj);
+ u32 pitch_val;
- WARN((i915_gem_obj_ggtt_offset(obj) & ~I830_FENCE_START_MASK) ||
- (size & -size) != size ||
- (i915_gem_obj_ggtt_offset(obj) & (size - 1)),
- "object 0x%08llx not 512K or pot-size 0x%08x aligned\n",
- i915_gem_obj_ggtt_offset(obj), size);
+ WARN((vma->node.start & ~I830_FENCE_START_MASK) ||
+ !is_power_of_2(vma->node.size) ||
+ (vma->node.start & (vma->node.size - 1)),
+ "object 0x%08llx not 512K or pot-size 0x%08llx aligned\n",
+ vma->node.start, vma->node.size);
pitch_val = stride / 128;
pitch_val = ffs(pitch_val) - 1;
- val = i915_gem_obj_ggtt_offset(obj);
- if (tiling == I915_TILING_Y)
- val |= 1 << I830_FENCE_TILING_Y_SHIFT;
- val |= I830_FENCE_SIZE_BITS(size);
+ val = vma->node.start;
+ if (is_y_tiled)
+ val |= BIT(I830_FENCE_TILING_Y_SHIFT);
+ val |= I830_FENCE_SIZE_BITS(vma->node.size);
val |= pitch_val << I830_FENCE_PITCH_SHIFT;
val |= I830_FENCE_REG_VALID;
- } else
- val = 0;
+ }
- I915_WRITE(FENCE_REG(reg), val);
- POSTING_READ(FENCE_REG(reg));
-}
+ if (!pipelined) {
+ struct drm_i915_private *dev_priv = fence->i915;
+ i915_reg_t reg = FENCE_REG(fence->id);
-inline static bool i915_gem_object_needs_mb(struct drm_i915_gem_object *obj)
-{
- return obj && obj->base.read_domains & I915_GEM_DOMAIN_GTT;
+ I915_WRITE(reg, val);
+ POSTING_READ(reg);
+ }
}
-static void i915_gem_write_fence(struct drm_device *dev, int reg,
- struct drm_i915_gem_object *obj)
+static void fence_write(struct drm_i915_fence_reg *fence,
+ struct i915_vma *vma)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
-
- /* Ensure that all CPU reads are completed before installing a fence
- * and all writes before removing the fence.
+ /* Previous access through the fence register is marshalled by
+ * the mb() inside the fault handlers (i915_gem_release_mmaps)
+ * and explicitly managed for internal users.
*/
- if (i915_gem_object_needs_mb(dev_priv->fence_regs[reg].obj))
- mb();
-
- WARN(obj &&
- (!i915_gem_object_get_stride(obj) ||
- !i915_gem_object_get_tiling(obj)),
- "bogus fence setup with stride: 0x%x, tiling mode: %i\n",
- i915_gem_object_get_stride(obj),
- i915_gem_object_get_tiling(obj));
-
- if (IS_GEN2(dev))
- i830_write_fence_reg(dev, reg, obj);
- else if (IS_GEN3(dev))
- i915_write_fence_reg(dev, reg, obj);
- else if (INTEL_INFO(dev)->gen >= 4)
- i965_write_fence_reg(dev, reg, obj);
-
- /* And similarly be paranoid that no direct access to this region
- * is reordered to before the fence is installed.
+
+ if (IS_GEN2(fence->i915))
+ i830_write_fence_reg(fence, vma);
+ else if (IS_GEN3(fence->i915))
+ i915_write_fence_reg(fence, vma);
+ else
+ i965_write_fence_reg(fence, vma);
+
+ /* Access through the fenced region afterwards is
+ * ordered by the posting reads whilst writing the registers.
*/
- if (i915_gem_object_needs_mb(obj))
- mb();
-}
-static inline int fence_number(struct drm_i915_private *dev_priv,
- struct drm_i915_fence_reg *fence)
-{
- return fence - dev_priv->fence_regs;
+ fence->dirty = false;
}
-static void i915_gem_object_update_fence(struct drm_i915_gem_object *obj,
- struct drm_i915_fence_reg *fence,
- bool enable)
+static int fence_update(struct drm_i915_fence_reg *fence,
+ struct i915_vma *vma)
{
- struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
- int reg = fence_number(dev_priv, fence);
+ int ret;
- i915_gem_write_fence(obj->base.dev, reg, enable ? obj : NULL);
+ if (vma) {
+ if (!i915_vma_is_map_and_fenceable(vma))
+ return -EINVAL;
- if (enable) {
- obj->fence_reg = reg;
- fence->obj = obj;
- list_move_tail(&fence->lru_list, &dev_priv->mm.fence_list);
- } else {
- obj->fence_reg = I915_FENCE_REG_NONE;
- fence->obj = NULL;
- list_del_init(&fence->lru_list);
+ if (WARN(!i915_gem_object_get_stride(vma->obj) ||
+ !i915_gem_object_get_tiling(vma->obj),
+ "bogus fence setup with stride: 0x%x, tiling mode: %i\n",
+ i915_gem_object_get_stride(vma->obj),
+ i915_gem_object_get_tiling(vma->obj)))
+ return -EINVAL;
+
+ ret = i915_gem_active_retire(&vma->last_fence,
+ &vma->obj->base.dev->struct_mutex);
+ if (ret)
+ return ret;
}
- obj->fence_dirty = false;
-}
-static inline void i915_gem_object_fence_lost(struct drm_i915_gem_object *obj)
-{
- if (i915_gem_object_is_tiled(obj))
- i915_gem_release_mmap(obj);
+ if (fence->vma) {
+ ret = i915_gem_active_retire(&fence->vma->last_fence,
+ &fence->vma->obj->base.dev->struct_mutex);
+ if (ret)
+ return ret;
+ }
- /* As we do not have an associated fence register, we will force
- * a tiling change if we ever need to acquire one.
- */
- obj->fence_dirty = false;
- obj->fence_reg = I915_FENCE_REG_NONE;
-}
+ if (fence->vma && fence->vma != vma) {
+ /* Ensure that all userspace CPU access is completed before
+ * stealing the fence.
+ */
+ i915_gem_release_mmap(fence->vma->obj);
-static int
-i915_gem_object_wait_fence(struct drm_i915_gem_object *obj)
-{
- return i915_gem_active_retire(&obj->last_fence,
- &obj->base.dev->struct_mutex);
+ fence->vma->fence = NULL;
+ fence->vma = NULL;
+
+ list_move(&fence->link, &fence->i915->mm.fence_list);
+ }
+
+ fence_write(fence, vma);
+
+ if (vma) {
+ if (fence->vma != vma) {
+ vma->fence = fence;
+ fence->vma = vma;
+ }
+
+ list_move_tail(&fence->link, &fence->i915->mm.fence_list);
+ }
+
+ return 0;
}
/**
- * i915_gem_object_put_fence - force-remove fence for an object
- * @obj: object to map through a fence reg
+ * i915_vma_put_fence - force-remove fence for a VMA
+ * @vma: vma to map linearly (not through a fence reg)
*
* This function force-removes any fence from the given object, which is useful
* if the kernel wants to do untiled GTT access.
* 0 on success, negative error code on failure.
*/
int
-i915_gem_object_put_fence(struct drm_i915_gem_object *obj)
+i915_vma_put_fence(struct i915_vma *vma)
{
- struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
- struct drm_i915_fence_reg *fence;
- int ret;
-
- ret = i915_gem_object_wait_fence(obj);
- if (ret)
- return ret;
+ struct drm_i915_fence_reg *fence = vma->fence;
- if (obj->fence_reg == I915_FENCE_REG_NONE)
+ if (!fence)
return 0;
- fence = &dev_priv->fence_regs[obj->fence_reg];
-
- if (WARN_ON(fence->pin_count))
+ if (fence->pin_count)
return -EBUSY;
- i915_gem_object_fence_lost(obj);
- i915_gem_object_update_fence(obj, fence, false);
-
- return 0;
+ return fence_update(fence, NULL);
}
-static struct drm_i915_fence_reg *
-i915_find_fence_reg(struct drm_device *dev)
+static struct drm_i915_fence_reg *fence_find(struct drm_i915_private *dev_priv)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
- struct drm_i915_fence_reg *reg, *avail;
- int i;
-
- /* First try to find a free reg */
- avail = NULL;
- for (i = 0; i < dev_priv->num_fence_regs; i++) {
- reg = &dev_priv->fence_regs[i];
- if (!reg->obj)
- return reg;
-
- if (!reg->pin_count)
- avail = reg;
- }
-
- if (avail == NULL)
- goto deadlock;
+ struct drm_i915_fence_reg *fence;
- /* None available, try to steal one or wait for a user to finish */
- list_for_each_entry(reg, &dev_priv->mm.fence_list, lru_list) {
- if (reg->pin_count)
+ list_for_each_entry(fence, &dev_priv->mm.fence_list, link) {
+ if (fence->pin_count)
continue;
- return reg;
+ return fence;
}
-deadlock:
/* Wait for completion of pending flips which consume fences */
- if (intel_has_pending_fb_unpin(dev))
+ if (intel_has_pending_fb_unpin(&dev_priv->drm))
return ERR_PTR(-EAGAIN);
return ERR_PTR(-EDEADLK);
}
/**
- * i915_gem_object_get_fence - set up fencing for an object
- * @obj: object to map through a fence reg
+ * i915_vma_get_fence - set up fencing for a vma
+ * @vma: vma to map through a fence reg
*
* When mapping objects through the GTT, userspace wants to be able to write
* to them without having to worry about swizzling if the object is tiled.
* 0 on success, negative error code on failure.
*/
int
-i915_gem_object_get_fence(struct drm_i915_gem_object *obj)
+i915_vma_get_fence(struct i915_vma *vma)
{
- struct drm_device *dev = obj->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
- bool enable = i915_gem_object_is_tiled(obj);
- struct drm_i915_fence_reg *reg;
- int ret;
-
- /* Have we updated the tiling parameters upon the object and so
- * will need to serialise the write to the associated fence register?
- */
- if (obj->fence_dirty) {
- ret = i915_gem_object_wait_fence(obj);
- if (ret)
- return ret;
- }
+ struct drm_i915_fence_reg *fence;
+ struct i915_vma *set = i915_gem_object_is_tiled(vma->obj) ? vma : NULL;
/* Just update our place in the LRU if our fence is getting reused. */
- if (obj->fence_reg != I915_FENCE_REG_NONE) {
- reg = &dev_priv->fence_regs[obj->fence_reg];
- if (!obj->fence_dirty) {
- list_move_tail(®->lru_list,
- &dev_priv->mm.fence_list);
+ if (vma->fence) {
+ fence = vma->fence;
+ if (!fence->dirty) {
+ list_move_tail(&fence->link,
+ &fence->i915->mm.fence_list);
return 0;
}
- } else if (enable) {
- if (WARN_ON(!obj->map_and_fenceable))
- return -EINVAL;
-
- reg = i915_find_fence_reg(dev);
- if (IS_ERR(reg))
- return PTR_ERR(reg);
-
- if (reg->obj) {
- struct drm_i915_gem_object *old = reg->obj;
-
- ret = i915_gem_object_wait_fence(old);
- if (ret)
- return ret;
-
- i915_gem_object_fence_lost(old);
- }
+ } else if (set) {
+ fence = fence_find(to_i915(vma->vm->dev));
+ if (IS_ERR(fence))
+ return PTR_ERR(fence);
} else
return 0;
- i915_gem_object_update_fence(obj, reg, enable);
-
- return 0;
-}
-
-/**
- * i915_gem_object_pin_fence - pin fencing state
- * @obj: object to pin fencing for
- *
- * This pins the fencing state (whether tiled or untiled) to make sure the
- * object is ready to be used as a scanout target. Fencing status must be
- * synchronize first by calling i915_gem_object_get_fence():
- *
- * The resulting fence pin reference must be released again with
- * i915_gem_object_unpin_fence().
- *
- * Returns:
- *
- * True if the object has a fence, false otherwise.
- */
-bool
-i915_gem_object_pin_fence(struct drm_i915_gem_object *obj)
-{
- if (obj->fence_reg != I915_FENCE_REG_NONE) {
- struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
- struct i915_vma *ggtt_vma = i915_gem_obj_to_ggtt(obj);
-
- WARN_ON(!ggtt_vma ||
- dev_priv->fence_regs[obj->fence_reg].pin_count >
- i915_vma_pin_count(ggtt_vma));
- dev_priv->fence_regs[obj->fence_reg].pin_count++;
- return true;
- } else
- return false;
-}
-
-/**
- * i915_gem_object_unpin_fence - unpin fencing state
- * @obj: object to unpin fencing for
- *
- * This releases the fence pin reference acquired through
- * i915_gem_object_pin_fence. It will handle both objects with and without an
- * attached fence correctly, callers do not need to distinguish this.
- */
-void
-i915_gem_object_unpin_fence(struct drm_i915_gem_object *obj)
-{
- if (obj->fence_reg != I915_FENCE_REG_NONE) {
- struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
- WARN_ON(dev_priv->fence_regs[obj->fence_reg].pin_count <= 0);
- dev_priv->fence_regs[obj->fence_reg].pin_count--;
- }
+ return fence_update(fence, set);
}
/**
for (i = 0; i < dev_priv->num_fence_regs; i++) {
struct drm_i915_fence_reg *reg = &dev_priv->fence_regs[i];
+ struct i915_vma *vma = reg->vma;
/*
* Commit delayed tiling changes if we have an object still
* attached to the fence, otherwise just clear the fence.
*/
- if (reg->obj) {
- i915_gem_object_update_fence(reg->obj, reg,
- i915_gem_object_get_tiling(reg->obj));
- } else {
- i915_gem_write_fence(dev, i, NULL);
- }
+ if (vma && !i915_gem_object_is_tiled(vma->obj))
+ vma = NULL;
+
+ fence_update(reg, vma);
}
}
#include "i915_trace.h"
#include "intel_drv.h"
+#define I915_GFP_DMA (GFP_KERNEL | __GFP_HIGHMEM)
+
/**
* DOC: Global GTT views
*
has_full_48bit_ppgtt =
IS_BROADWELL(dev_priv) || INTEL_GEN(dev_priv) >= 9;
- if (intel_vgpu_active(dev_priv))
- has_full_ppgtt = false; /* emulation is too hard */
+ if (intel_vgpu_active(dev_priv)) {
+ /* emulation is too hard */
+ has_full_ppgtt = false;
+ has_full_48bit_ppgtt = false;
+ }
if (!has_aliasing_ppgtt)
return 0;
return 0;
}
- if (INTEL_GEN(dev_priv) >= 8 && i915.enable_execlists)
+ if (INTEL_GEN(dev_priv) >= 8 && i915.enable_execlists && has_full_ppgtt)
return has_full_48bit_ppgtt ? 3 : 2;
else
return has_aliasing_ppgtt ? 1 : 0;
{
u32 pte_flags = 0;
+ vma->pages = vma->obj->pages;
+
/* Currently applicable only to VLV */
if (vma->obj->gt_ro)
pte_flags |= PTE_READ_ONLY;
- vma->vm->insert_entries(vma->vm, vma->obj->pages, vma->node.start,
+ vma->vm->insert_entries(vma->vm, vma->pages, vma->node.start,
cache_level, pte_flags);
return 0;
static int __setup_page_dma(struct drm_device *dev,
struct i915_page_dma *p, gfp_t flags)
{
- struct device *device = &dev->pdev->dev;
+ struct device *kdev = &dev->pdev->dev;
p->page = alloc_page(flags);
if (!p->page)
return -ENOMEM;
- p->daddr = dma_map_page(device,
+ p->daddr = dma_map_page(kdev,
p->page, 0, 4096, PCI_DMA_BIDIRECTIONAL);
- if (dma_mapping_error(device, p->daddr)) {
+ if (dma_mapping_error(kdev, p->daddr)) {
__free_page(p->page);
return -EINVAL;
}
static int setup_page_dma(struct drm_device *dev, struct i915_page_dma *p)
{
- return __setup_page_dma(dev, p, GFP_KERNEL);
+ return __setup_page_dma(dev, p, I915_GFP_DMA);
}
static void cleanup_page_dma(struct drm_device *dev, struct i915_page_dma *p)
{
+ struct pci_dev *pdev = dev->pdev;
+
if (WARN_ON(!p->page))
return;
- dma_unmap_page(&dev->pdev->dev, p->daddr, 4096, PCI_DMA_BIDIRECTIONAL);
+ dma_unmap_page(&pdev->dev, p->daddr, 4096, PCI_DMA_BIDIRECTIONAL);
__free_page(p->page);
memset(p, 0, sizeof(*p));
}
fill_page_dma(dev, p, v);
}
-static struct i915_page_scratch *alloc_scratch_page(struct drm_device *dev)
+static int
+setup_scratch_page(struct drm_device *dev,
+ struct i915_page_dma *scratch,
+ gfp_t gfp)
{
- struct i915_page_scratch *sp;
- int ret;
-
- sp = kzalloc(sizeof(*sp), GFP_KERNEL);
- if (sp == NULL)
- return ERR_PTR(-ENOMEM);
-
- ret = __setup_page_dma(dev, px_base(sp), GFP_DMA32 | __GFP_ZERO);
- if (ret) {
- kfree(sp);
- return ERR_PTR(ret);
- }
-
- set_pages_uc(px_page(sp), 1);
-
- return sp;
+ return __setup_page_dma(dev, scratch, gfp | __GFP_ZERO);
}
-static void free_scratch_page(struct drm_device *dev,
- struct i915_page_scratch *sp)
+static void cleanup_scratch_page(struct drm_device *dev,
+ struct i915_page_dma *scratch)
{
- set_pages_wb(px_page(sp), 1);
-
- cleanup_px(dev, sp);
- kfree(sp);
+ cleanup_page_dma(dev, scratch);
}
static struct i915_page_table *alloc_pt(struct drm_device *dev)
{
gen8_pte_t scratch_pte;
- scratch_pte = gen8_pte_encode(px_dma(vm->scratch_page),
+ scratch_pte = gen8_pte_encode(vm->scratch_page.daddr,
I915_CACHE_LLC, true);
fill_px(vm->dev, pt, scratch_pte);
{
gen6_pte_t scratch_pte;
- WARN_ON(px_dma(vm->scratch_page) == 0);
+ WARN_ON(vm->scratch_page.daddr == 0);
- scratch_pte = vm->pte_encode(px_dma(vm->scratch_page),
+ scratch_pte = vm->pte_encode(vm->scratch_page.daddr,
I915_CACHE_LLC, true, 0);
fill32_px(vm->dev, pt, scratch_pte);
bool use_scratch)
{
struct i915_hw_ppgtt *ppgtt = i915_vm_to_ppgtt(vm);
- gen8_pte_t scratch_pte = gen8_pte_encode(px_dma(vm->scratch_page),
+ gen8_pte_t scratch_pte = gen8_pte_encode(vm->scratch_page.daddr,
I915_CACHE_LLC, use_scratch);
if (!USES_FULL_48BIT_PPGTT(vm->dev)) {
struct drm_device *dev = vm->dev;
int ret;
- vm->scratch_page = alloc_scratch_page(dev);
- if (IS_ERR(vm->scratch_page))
- return PTR_ERR(vm->scratch_page);
+ ret = setup_scratch_page(dev, &vm->scratch_page, I915_GFP_DMA);
+ if (ret)
+ return ret;
vm->scratch_pt = alloc_pt(dev);
if (IS_ERR(vm->scratch_pt)) {
free_pt:
free_pt(dev, vm->scratch_pt);
free_scratch_page:
- free_scratch_page(dev, vm->scratch_page);
+ cleanup_scratch_page(dev, &vm->scratch_page);
return ret;
}
free_pdp(dev, vm->scratch_pdp);
free_pd(dev, vm->scratch_pd);
free_pt(dev, vm->scratch_pt);
- free_scratch_page(dev, vm->scratch_page);
+ cleanup_scratch_page(dev, &vm->scratch_page);
}
static void gen8_ppgtt_cleanup_3lvl(struct drm_device *dev,
struct i915_address_space *vm = &ppgtt->base;
uint64_t start = ppgtt->base.start;
uint64_t length = ppgtt->base.total;
- gen8_pte_t scratch_pte = gen8_pte_encode(px_dma(vm->scratch_page),
+ gen8_pte_t scratch_pte = gen8_pte_encode(vm->scratch_page.daddr,
I915_CACHE_LLC, true);
if (!USES_FULL_48BIT_PPGTT(vm->dev)) {
uint32_t pte, pde;
uint32_t start = ppgtt->base.start, length = ppgtt->base.total;
- scratch_pte = vm->pte_encode(px_dma(vm->scratch_page),
+ scratch_pte = vm->pte_encode(vm->scratch_page.daddr,
I915_CACHE_LLC, true, 0);
gen6_for_each_pde(unused, &ppgtt->pd, start, length, pde) {
unsigned first_pte = first_entry % GEN6_PTES;
unsigned last_pte, i;
- scratch_pte = vm->pte_encode(px_dma(vm->scratch_page),
+ scratch_pte = vm->pte_encode(vm->scratch_page.daddr,
I915_CACHE_LLC, true, 0);
while (num_entries) {
static int gen6_init_scratch(struct i915_address_space *vm)
{
struct drm_device *dev = vm->dev;
+ int ret;
- vm->scratch_page = alloc_scratch_page(dev);
- if (IS_ERR(vm->scratch_page))
- return PTR_ERR(vm->scratch_page);
+ ret = setup_scratch_page(dev, &vm->scratch_page, I915_GFP_DMA);
+ if (ret)
+ return ret;
vm->scratch_pt = alloc_pt(dev);
if (IS_ERR(vm->scratch_pt)) {
- free_scratch_page(dev, vm->scratch_page);
+ cleanup_scratch_page(dev, &vm->scratch_page);
return PTR_ERR(vm->scratch_pt);
}
struct drm_device *dev = vm->dev;
free_pt(dev, vm->scratch_pt);
- free_scratch_page(dev, vm->scratch_page);
+ cleanup_scratch_page(dev, &vm->scratch_page);
}
static void gen6_ppgtt_cleanup(struct i915_address_space *vm)
static void gen8_set_pte(void __iomem *addr, gen8_pte_t pte)
{
-#ifdef writeq
writeq(pte, addr);
-#else
- iowrite32((u32)pte, addr);
- iowrite32(pte >> 32, addr + 4);
-#endif
}
static void gen8_ggtt_insert_page(struct i915_address_space *vm,
first_entry, num_entries, max_entries))
num_entries = max_entries;
- scratch_pte = gen8_pte_encode(px_dma(vm->scratch_page),
+ scratch_pte = gen8_pte_encode(vm->scratch_page.daddr,
I915_CACHE_LLC,
use_scratch);
for (i = 0; i < num_entries; i++)
first_entry, num_entries, max_entries))
num_entries = max_entries;
- scratch_pte = vm->pte_encode(px_dma(vm->scratch_page),
+ scratch_pte = vm->pte_encode(vm->scratch_page.daddr,
I915_CACHE_LLC, use_scratch, 0);
for (i = 0; i < num_entries; i++)
if (obj->gt_ro)
pte_flags |= PTE_READ_ONLY;
- vma->vm->insert_entries(vma->vm, vma->ggtt_view.pages,
- vma->node.start,
+ vma->vm->insert_entries(vma->vm, vma->pages, vma->node.start,
cache_level, pte_flags);
/*
if (flags & I915_VMA_GLOBAL_BIND) {
vma->vm->insert_entries(vma->vm,
- vma->ggtt_view.pages,
- vma->node.start,
+ vma->pages, vma->node.start,
cache_level, pte_flags);
}
struct i915_hw_ppgtt *appgtt =
to_i915(vma->vm->dev)->mm.aliasing_ppgtt;
appgtt->base.insert_entries(&appgtt->base,
- vma->ggtt_view.pages,
- vma->node.start,
+ vma->pages, vma->node.start,
cache_level, pte_flags);
}
void i915_gem_gtt_finish_object(struct drm_i915_gem_object *obj)
{
- struct drm_device *dev = obj->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
+ struct device *kdev = &dev_priv->drm.pdev->dev;
struct i915_ggtt *ggtt = &dev_priv->ggtt;
if (unlikely(ggtt->do_idle_maps)) {
- if (i915_gem_wait_for_idle(dev_priv, false)) {
+ if (i915_gem_wait_for_idle(dev_priv, I915_WAIT_LOCKED)) {
DRM_ERROR("Failed to wait for idle; VT'd may hang.\n");
/* Wait a bit, in hopes it avoids the hang */
udelay(10);
}
}
- dma_unmap_sg(&dev->pdev->dev, obj->pages->sgl, obj->pages->nents,
+ dma_unmap_sg(kdev, obj->pages->sgl, obj->pages->nents,
PCI_DMA_BIDIRECTIONAL);
}
if (dev_priv->mm.aliasing_ppgtt) {
struct i915_hw_ppgtt *ppgtt = dev_priv->mm.aliasing_ppgtt;
-
ppgtt->base.cleanup(&ppgtt->base);
kfree(ppgtt);
}
ggtt->base.cleanup(&ggtt->base);
arch_phys_wc_del(ggtt->mtrr);
- io_mapping_free(ggtt->mappable);
+ io_mapping_fini(&ggtt->mappable);
}
static unsigned int gen6_get_total_gtt_size(u16 snb_gmch_ctl)
static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
{
struct pci_dev *pdev = ggtt->base.dev->pdev;
- struct i915_page_scratch *scratch_page;
phys_addr_t phys_addr;
+ int ret;
/* For Modern GENs the PTEs and register space are split in the BAR */
phys_addr = pci_resource_start(pdev, 0) + pci_resource_len(pdev, 0) / 2;
return -ENOMEM;
}
- scratch_page = alloc_scratch_page(ggtt->base.dev);
- if (IS_ERR(scratch_page)) {
+ ret = setup_scratch_page(ggtt->base.dev,
+ &ggtt->base.scratch_page,
+ GFP_DMA32);
+ if (ret) {
DRM_ERROR("Scratch setup failed\n");
/* iounmap will also get called at remove, but meh */
iounmap(ggtt->gsm);
- return PTR_ERR(scratch_page);
+ return ret;
}
- ggtt->base.scratch_page = scratch_page;
-
return 0;
}
struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm);
iounmap(ggtt->gsm);
- free_scratch_page(vm->dev, vm->scratch_page);
+ cleanup_scratch_page(vm->dev, &vm->scratch_page);
}
static int gen8_gmch_probe(struct i915_ggtt *ggtt)
if (!HAS_LLC(dev_priv))
ggtt->base.mm.color_adjust = i915_gtt_color_adjust;
- ggtt->mappable =
- io_mapping_create_wc(ggtt->mappable_base, ggtt->mappable_end);
- if (!ggtt->mappable) {
+ if (!io_mapping_init_wc(&dev_priv->ggtt.mappable,
+ dev_priv->ggtt.mappable_base,
+ dev_priv->ggtt.mappable_end)) {
ret = -EIO;
goto out_gtt_cleanup;
}
{
struct drm_i915_private *dev_priv = to_i915(dev);
struct i915_ggtt *ggtt = &dev_priv->ggtt;
- struct drm_i915_gem_object *obj;
- struct i915_vma *vma;
+ struct drm_i915_gem_object *obj, *on;
i915_check_and_clear_faults(dev_priv);
ggtt->base.clear_range(&ggtt->base, ggtt->base.start, ggtt->base.total,
true);
- /* Cache flush objects bound into GGTT and rebind them. */
- list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
+ ggtt->base.closed = true; /* skip rewriting PTE on VMA unbind */
+
+ /* clflush objects bound into the GGTT and rebind them. */
+ list_for_each_entry_safe(obj, on,
+ &dev_priv->mm.bound_list, global_list) {
+ bool ggtt_bound = false;
+ struct i915_vma *vma;
+
list_for_each_entry(vma, &obj->vma_list, obj_link) {
if (vma->vm != &ggtt->base)
continue;
+ if (!i915_vma_unbind(vma))
+ continue;
+
WARN_ON(i915_vma_bind(vma, obj->cache_level,
PIN_UPDATE));
+ ggtt_bound = true;
}
- if (obj->pin_display)
+ if (ggtt_bound)
WARN_ON(i915_gem_object_set_to_gtt_domain(obj, false));
}
+ ggtt->base.closed = false;
+
if (INTEL_INFO(dev)->gen >= 8) {
if (IS_CHERRYVIEW(dev) || IS_BROXTON(dev))
chv_setup_private_ppat(dev_priv);
GEM_BUG_ON(vma->node.allocated);
GEM_BUG_ON(i915_vma_is_active(vma));
GEM_BUG_ON(!i915_vma_is_closed(vma));
+ GEM_BUG_ON(vma->fence);
list_del(&vma->vm_link);
if (!i915_vma_is_ggtt(vma))
}
static struct i915_vma *
-__i915_gem_vma_create(struct drm_i915_gem_object *obj,
- struct i915_address_space *vm,
- const struct i915_ggtt_view *view)
+__i915_vma_create(struct drm_i915_gem_object *obj,
+ struct i915_address_space *vm,
+ const struct i915_ggtt_view *view)
{
struct i915_vma *vma;
int i;
GEM_BUG_ON(vm->closed);
- if (WARN_ON(i915_is_ggtt(vm) != !!view))
- return ERR_PTR(-EINVAL);
-
vma = kmem_cache_zalloc(to_i915(obj->base.dev)->vmas, GFP_KERNEL);
if (vma == NULL)
return ERR_PTR(-ENOMEM);
- INIT_LIST_HEAD(&vma->obj_link);
INIT_LIST_HEAD(&vma->exec_list);
for (i = 0; i < ARRAY_SIZE(vma->last_read); i++)
init_request_active(&vma->last_read[i], i915_vma_retire);
+ init_request_active(&vma->last_fence, NULL);
list_add(&vma->vm_link, &vm->unbound_list);
vma->vm = vm;
vma->obj = obj;
vma->size = obj->base.size;
- if (i915_is_ggtt(vm)) {
- vma->flags |= I915_VMA_GGTT;
+ if (view) {
vma->ggtt_view = *view;
if (view->type == I915_GGTT_VIEW_PARTIAL) {
vma->size = view->params.partial.size;
intel_rotation_info_size(&view->params.rotated);
vma->size <<= PAGE_SHIFT;
}
+ }
+
+ if (i915_is_ggtt(vm)) {
+ vma->flags |= I915_VMA_GGTT;
} else {
i915_ppgtt_get(i915_vm_to_ppgtt(vm));
}
list_add_tail(&vma->obj_link, &obj->vma_list);
-
return vma;
}
+static inline bool vma_matches(struct i915_vma *vma,
+ struct i915_address_space *vm,
+ const struct i915_ggtt_view *view)
+{
+ if (vma->vm != vm)
+ return false;
+
+ if (!i915_vma_is_ggtt(vma))
+ return true;
+
+ if (!view)
+ return vma->ggtt_view.type == 0;
+
+ if (vma->ggtt_view.type != view->type)
+ return false;
+
+ return memcmp(&vma->ggtt_view.params,
+ &view->params,
+ sizeof(view->params)) == 0;
+}
+
struct i915_vma *
-i915_gem_obj_lookup_or_create_vma(struct drm_i915_gem_object *obj,
- struct i915_address_space *vm)
+i915_vma_create(struct drm_i915_gem_object *obj,
+ struct i915_address_space *vm,
+ const struct i915_ggtt_view *view)
+{
+ GEM_BUG_ON(view && !i915_is_ggtt(vm));
+ GEM_BUG_ON(i915_gem_obj_to_vma(obj, vm, view));
+
+ return __i915_vma_create(obj, vm, view);
+}
+
+struct i915_vma *
+i915_gem_obj_to_vma(struct drm_i915_gem_object *obj,
+ struct i915_address_space *vm,
+ const struct i915_ggtt_view *view)
{
struct i915_vma *vma;
- vma = i915_gem_obj_to_vma(obj, vm);
- if (!vma)
- vma = __i915_gem_vma_create(obj, vm,
- i915_is_ggtt(vm) ? &i915_ggtt_view_normal : NULL);
+ list_for_each_entry_reverse(vma, &obj->vma_list, obj_link)
+ if (vma_matches(vma, vm, view))
+ return vma;
- return vma;
+ return NULL;
}
struct i915_vma *
-i915_gem_obj_lookup_or_create_ggtt_vma(struct drm_i915_gem_object *obj,
- const struct i915_ggtt_view *view)
+i915_gem_obj_lookup_or_create_vma(struct drm_i915_gem_object *obj,
+ struct i915_address_space *vm,
+ const struct i915_ggtt_view *view)
{
- struct drm_device *dev = obj->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
- struct i915_ggtt *ggtt = &dev_priv->ggtt;
- struct i915_vma *vma = i915_gem_obj_to_ggtt_view(obj, view);
+ struct i915_vma *vma;
- GEM_BUG_ON(!view);
+ GEM_BUG_ON(view && !i915_is_ggtt(vm));
+ vma = i915_gem_obj_to_vma(obj, vm, view);
if (!vma)
- vma = __i915_gem_vma_create(obj, &ggtt->base, view);
+ vma = __i915_vma_create(obj, vm, view);
GEM_BUG_ON(i915_vma_is_closed(vma));
return vma;
-
}
static struct scatterlist *
}
static struct sg_table *
-intel_rotate_fb_obj_pages(struct intel_rotation_info *rot_info,
+intel_rotate_fb_obj_pages(const struct intel_rotation_info *rot_info,
struct drm_i915_gem_object *obj)
{
const size_t n_pages = obj->base.size / PAGE_SIZE;
- unsigned int size_pages = rot_info->plane[0].width * rot_info->plane[0].height;
- unsigned int size_pages_uv;
+ unsigned int size = intel_rotation_info_size(rot_info);
struct sgt_iter sgt_iter;
dma_addr_t dma_addr;
unsigned long i;
dma_addr_t *page_addr_list;
struct sg_table *st;
- unsigned int uv_start_page;
struct scatterlist *sg;
int ret = -ENOMEM;
if (!page_addr_list)
return ERR_PTR(ret);
- /* Account for UV plane with NV12. */
- if (rot_info->pixel_format == DRM_FORMAT_NV12)
- size_pages_uv = rot_info->plane[1].width * rot_info->plane[1].height;
- else
- size_pages_uv = 0;
-
/* Allocate target SG list. */
st = kmalloc(sizeof(*st), GFP_KERNEL);
if (!st)
goto err_st_alloc;
- ret = sg_alloc_table(st, size_pages + size_pages_uv, GFP_KERNEL);
+ ret = sg_alloc_table(st, size, GFP_KERNEL);
if (ret)
goto err_sg_alloc;
st->nents = 0;
sg = st->sgl;
- /* Rotate the pages. */
- sg = rotate_pages(page_addr_list, 0,
- rot_info->plane[0].width, rot_info->plane[0].height,
- rot_info->plane[0].width,
- st, sg);
-
- /* Append the UV plane if NV12. */
- if (rot_info->pixel_format == DRM_FORMAT_NV12) {
- uv_start_page = size_pages;
-
- /* Check for tile-row un-alignment. */
- if (offset_in_page(rot_info->uv_offset))
- uv_start_page--;
-
- rot_info->uv_start_page = uv_start_page;
-
- sg = rotate_pages(page_addr_list, rot_info->uv_start_page,
- rot_info->plane[1].width, rot_info->plane[1].height,
- rot_info->plane[1].width,
- st, sg);
+ for (i = 0 ; i < ARRAY_SIZE(rot_info->plane); i++) {
+ sg = rotate_pages(page_addr_list, rot_info->plane[i].offset,
+ rot_info->plane[i].width, rot_info->plane[i].height,
+ rot_info->plane[i].stride, st, sg);
}
- DRM_DEBUG_KMS("Created rotated page mapping for object size %zu (%ux%u tiles, %u pages (%u plane 0)).\n",
- obj->base.size, rot_info->plane[0].width,
- rot_info->plane[0].height, size_pages + size_pages_uv,
- size_pages);
+ DRM_DEBUG_KMS("Created rotated page mapping for object size %zu (%ux%u tiles, %u pages)\n",
+ obj->base.size, rot_info->plane[0].width, rot_info->plane[0].height, size);
drm_free_large(page_addr_list);
err_st_alloc:
drm_free_large(page_addr_list);
- DRM_DEBUG_KMS("Failed to create rotated mapping for object size %zu! (%d) (%ux%u tiles, %u pages (%u plane 0))\n",
- obj->base.size, ret, rot_info->plane[0].width,
- rot_info->plane[0].height, size_pages + size_pages_uv,
- size_pages);
+ DRM_DEBUG_KMS("Failed to create rotated mapping for object size %zu! (%ux%u tiles, %u pages)\n",
+ obj->base.size, rot_info->plane[0].width, rot_info->plane[0].height, size);
+
return ERR_PTR(ret);
}
{
int ret = 0;
- if (vma->ggtt_view.pages)
+ if (vma->pages)
return 0;
if (vma->ggtt_view.type == I915_GGTT_VIEW_NORMAL)
- vma->ggtt_view.pages = vma->obj->pages;
+ vma->pages = vma->obj->pages;
else if (vma->ggtt_view.type == I915_GGTT_VIEW_ROTATED)
- vma->ggtt_view.pages =
+ vma->pages =
intel_rotate_fb_obj_pages(&vma->ggtt_view.params.rotated, vma->obj);
else if (vma->ggtt_view.type == I915_GGTT_VIEW_PARTIAL)
- vma->ggtt_view.pages =
- intel_partial_pages(&vma->ggtt_view, vma->obj);
+ vma->pages = intel_partial_pages(&vma->ggtt_view, vma->obj);
else
WARN_ONCE(1, "GGTT view %u not implemented!\n",
vma->ggtt_view.type);
- if (!vma->ggtt_view.pages) {
+ if (!vma->pages) {
DRM_ERROR("Failed to get pages for GGTT view type %u!\n",
vma->ggtt_view.type);
ret = -EINVAL;
- } else if (IS_ERR(vma->ggtt_view.pages)) {
- ret = PTR_ERR(vma->ggtt_view.pages);
- vma->ggtt_view.pages = NULL;
+ } else if (IS_ERR(vma->pages)) {
+ ret = PTR_ERR(vma->pages);
+ vma->pages = NULL;
DRM_ERROR("Failed to get pages for VMA view type %u (%d)!\n",
vma->ggtt_view.type, ret);
}
{
void __iomem *ptr;
+ /* Access through the GTT requires the device to be awake. */
+ assert_rpm_wakelock_held(to_i915(vma->vm->dev));
+
lockdep_assert_held(&vma->vm->dev->struct_mutex);
- if (WARN_ON(!vma->obj->map_and_fenceable))
+ if (WARN_ON(!i915_vma_is_map_and_fenceable(vma)))
return IO_ERR_PTR(-ENODEV);
GEM_BUG_ON(!i915_vma_is_ggtt(vma));
ptr = vma->iomap;
if (ptr == NULL) {
- ptr = io_mapping_map_wc(i915_vm_to_ggtt(vma->vm)->mappable,
+ ptr = io_mapping_map_wc(&i915_vm_to_ggtt(vma->vm)->mappable,
vma->node.start,
vma->node.size);
if (ptr == NULL)
__i915_vma_pin(vma);
return ptr;
}
+
+void i915_vma_unpin_and_release(struct i915_vma **p_vma)
+{
+ struct i915_vma *vma;
+
+ vma = fetch_and_zero(p_vma);
+ if (!vma)
+ return;
+
+ i915_vma_unpin(vma);
+ i915_vma_put(vma);
+}
#include "i915_gem_request.h"
+#define I915_FENCE_REG_NONE -1
+#define I915_MAX_NUM_FENCES 32
+/* 32 fences + sign bit for FENCE_REG_NONE */
+#define I915_MAX_NUM_FENCE_BITS 6
+
struct drm_i915_file_private;
+struct drm_i915_fence_reg;
typedef uint32_t gen6_pte_t;
typedef uint64_t gen8_pte_t;
};
struct intel_rotation_info {
- unsigned int uv_offset;
- uint32_t pixel_format;
- unsigned int uv_start_page;
struct {
/* tiles */
- unsigned int width, height;
+ unsigned int width, height, stride, offset;
} plane[2];
};
} partial;
struct intel_rotation_info rotated;
} params;
-
- struct sg_table *pages;
};
extern const struct i915_ggtt_view i915_ggtt_view_normal;
struct drm_mm_node node;
struct drm_i915_gem_object *obj;
struct i915_address_space *vm;
+ struct drm_i915_fence_reg *fence;
+ struct sg_table *pages;
void __iomem *iomap;
u64 size;
+ u64 display_alignment;
unsigned int flags;
/**
#define I915_VMA_LOCAL_BIND BIT(7)
#define I915_VMA_BIND_MASK (I915_VMA_GLOBAL_BIND | I915_VMA_LOCAL_BIND | I915_VMA_PIN_OVERFLOW)
-#define I915_VMA_GGTT BIT(8)
-#define I915_VMA_CLOSED BIT(9)
+#define I915_VMA_GGTT BIT(8)
+#define I915_VMA_CAN_FENCE BIT(9)
+#define I915_VMA_CLOSED BIT(10)
unsigned int active;
struct i915_gem_active last_read[I915_NUM_ENGINES];
+ struct i915_gem_active last_fence;
/**
* Support different GGTT views into the same object.
struct drm_i915_gem_exec_object2 *exec_entry;
};
+struct i915_vma *
+i915_vma_create(struct drm_i915_gem_object *obj,
+ struct i915_address_space *vm,
+ const struct i915_ggtt_view *view);
+void i915_vma_unpin_and_release(struct i915_vma **p_vma);
+
static inline bool i915_vma_is_ggtt(const struct i915_vma *vma)
{
return vma->flags & I915_VMA_GGTT;
}
+static inline bool i915_vma_is_map_and_fenceable(const struct i915_vma *vma)
+{
+ return vma->flags & I915_VMA_CAN_FENCE;
+}
+
static inline bool i915_vma_is_closed(const struct i915_vma *vma)
{
return vma->flags & I915_VMA_CLOSED;
return vma->active & BIT(engine);
}
+static inline u32 i915_ggtt_offset(const struct i915_vma *vma)
+{
+ GEM_BUG_ON(!i915_vma_is_ggtt(vma));
+ GEM_BUG_ON(!vma->node.allocated);
+ GEM_BUG_ON(upper_32_bits(vma->node.start));
+ GEM_BUG_ON(upper_32_bits(vma->node.start + vma->node.size - 1));
+ return lower_32_bits(vma->node.start);
+}
+
struct i915_page_dma {
struct page *page;
union {
#define px_page(px) (px_base(px)->page)
#define px_dma(px) (px_base(px)->daddr)
-struct i915_page_scratch {
- struct i915_page_dma base;
-};
-
struct i915_page_table {
struct i915_page_dma base;
bool closed;
- struct i915_page_scratch *scratch_page;
+ struct i915_page_dma scratch_page;
struct i915_page_table *scratch_pt;
struct i915_page_directory *scratch_pd;
struct i915_page_directory_pointer *scratch_pdp; /* GEN8+ & 48b PPGTT */
*/
struct i915_ggtt {
struct i915_address_space base;
+ struct io_mapping mappable; /* Mapping to our CPU mappable region */
size_t stolen_size; /* Total size of stolen memory */
size_t stolen_usable_size; /* Total size minus BIOS reserved */
size_t stolen_reserved_base;
size_t stolen_reserved_size;
u64 mappable_end; /* End offset that we can CPU map */
- struct io_mapping *mappable; /* Mapping to our CPU mappable region */
phys_addr_t mappable_base; /* PA of our GMADR */
/** "Graphics Stolen Memory" holds the global PTEs */
int __must_check i915_gem_gtt_prepare_object(struct drm_i915_gem_object *obj);
void i915_gem_gtt_finish_object(struct drm_i915_gem_object *obj);
-static inline bool
-i915_ggtt_view_equal(const struct i915_ggtt_view *a,
- const struct i915_ggtt_view *b)
-{
- if (WARN_ON(!a || !b))
- return false;
-
- if (a->type != b->type)
- return false;
- if (a->type != I915_GGTT_VIEW_NORMAL)
- return !memcmp(&a->params, &b->params, sizeof(a->params));
- return true;
-}
-
/* Flags used by pin/bind&friends. */
#define PIN_NONBLOCK BIT(0)
#define PIN_MAPPABLE BIT(1)
#define PIN_ZONE_4G BIT(2)
+#define PIN_NONFAULT BIT(3)
#define PIN_MBZ BIT(5) /* I915_VMA_PIN_OVERFLOW */
#define PIN_GLOBAL BIT(6) /* I915_VMA_GLOBAL_BIND */
i915_vma_unpin(vma);
}
+static inline struct page *i915_vma_first_page(struct i915_vma *vma)
+{
+ GEM_BUG_ON(!vma->pages);
+ return sg_page(vma->pages->sgl);
+}
+
#endif
struct render_state {
const struct intel_renderstate_rodata *rodata;
- struct drm_i915_gem_object *obj;
- u64 ggtt_offset;
+ struct i915_vma *vma;
u32 aux_batch_size;
u32 aux_batch_offset;
};
static int render_state_setup(struct render_state *so)
{
- struct drm_device *dev = so->obj->base.dev;
+ struct drm_device *dev = so->vma->vm->dev;
const struct intel_renderstate_rodata *rodata = so->rodata;
const bool has_64bit_reloc = INTEL_GEN(dev) >= 8;
unsigned int i = 0, reloc_index = 0;
u32 *d;
int ret;
- ret = i915_gem_object_set_to_cpu_domain(so->obj, true);
+ ret = i915_gem_object_set_to_cpu_domain(so->vma->obj, true);
if (ret)
return ret;
- page = i915_gem_object_get_dirty_page(so->obj, 0);
+ page = i915_gem_object_get_dirty_page(so->vma->obj, 0);
d = kmap(page);
while (i < rodata->batch_items) {
u32 s = rodata->batch[i];
if (i * 4 == rodata->reloc[reloc_index]) {
- u64 r = s + so->ggtt_offset;
+ u64 r = s + so->vma->node.start;
s = lower_32_bits(r);
if (has_64bit_reloc) {
if (i + 1 >= rodata->batch_items ||
kunmap(page);
- ret = i915_gem_object_set_to_gtt_domain(so->obj, false);
+ ret = i915_gem_object_set_to_gtt_domain(so->vma->obj, false);
if (ret)
return ret;
int i915_gem_render_state_init(struct drm_i915_gem_request *req)
{
struct render_state so;
+ struct drm_i915_gem_object *obj;
int ret;
if (WARN_ON(req->engine->id != RCS))
if (so.rodata->batch_items * 4 > 4096)
return -EINVAL;
- so.obj = i915_gem_object_create(&req->i915->drm, 4096);
- if (IS_ERR(so.obj))
- return PTR_ERR(so.obj);
+ obj = i915_gem_object_create(&req->i915->drm, 4096);
+ if (IS_ERR(obj))
+ return PTR_ERR(obj);
- ret = i915_gem_object_ggtt_pin(so.obj, NULL, 0, 0, 0);
- if (ret)
+ so.vma = i915_vma_create(obj, &req->i915->ggtt.base, NULL);
+ if (IS_ERR(so.vma)) {
+ ret = PTR_ERR(so.vma);
goto err_obj;
+ }
- so.ggtt_offset = i915_gem_obj_ggtt_offset(so.obj);
+ ret = i915_vma_pin(so.vma, 0, 0, PIN_GLOBAL);
+ if (ret)
+ goto err_obj;
ret = render_state_setup(&so);
if (ret)
goto err_unpin;
- ret = req->engine->emit_bb_start(req, so.ggtt_offset,
+ ret = req->engine->emit_bb_start(req, so.vma->node.start,
so.rodata->batch_items * 4,
I915_DISPATCH_SECURE);
if (ret)
if (so.aux_batch_size > 8) {
ret = req->engine->emit_bb_start(req,
- (so.ggtt_offset +
+ (so.vma->node.start +
so.aux_batch_offset),
so.aux_batch_size,
I915_DISPATCH_SECURE);
goto err_unpin;
}
- i915_vma_move_to_active(i915_gem_obj_to_ggtt(so.obj), req, 0);
+ i915_vma_move_to_active(so.vma, req, 0);
err_unpin:
- i915_gem_object_ggtt_unpin(so.obj);
+ i915_vma_unpin(so.vma);
err_obj:
- i915_gem_object_put(so.obj);
+ i915_gem_object_put(obj);
return ret;
}
#ifndef _I915_GEM_RENDER_STATE_H_
#define _I915_GEM_RENDER_STATE_H_
-#include <linux/types.h>
+struct drm_i915_gem_request;
int i915_gem_render_state_init(struct drm_i915_gem_request *req);
list_add_tail(&req->client_list, &file_priv->mm.request_list);
spin_unlock(&file_priv->mm.lock);
- req->pid = get_pid(task_pid(current));
-
return 0;
}
list_del(&request->client_list);
request->file_priv = NULL;
spin_unlock(&file_priv->mm.lock);
-
- put_pid(request->pid);
- request->pid = NULL;
}
void i915_gem_retire_noop(struct i915_gem_active *active,
} while (tmp != req);
}
-static int i915_gem_check_wedge(unsigned int reset_counter, bool interruptible)
+static int i915_gem_check_wedge(struct drm_i915_private *dev_priv)
{
- if (__i915_terminally_wedged(reset_counter))
+ struct i915_gpu_error *error = &dev_priv->gpu_error;
+
+ if (i915_terminally_wedged(error))
return -EIO;
- if (__i915_reset_in_progress(reset_counter)) {
+ if (i915_reset_in_progress(error)) {
/* Non-interruptible callers can't handle -EAGAIN, hence return
* -EIO unconditionally for these.
*/
- if (!interruptible)
+ if (!dev_priv->mm.interruptible)
return -EIO;
return -EAGAIN;
/* Carefully retire all requests without writing to the rings */
for_each_engine(engine, dev_priv) {
- ret = intel_engine_idle(engine, true);
+ ret = intel_engine_idle(engine,
+ I915_WAIT_INTERRUPTIBLE |
+ I915_WAIT_LOCKED);
if (ret)
return ret;
}
return 0;
}
+static int __i915_sw_fence_call
+submit_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
+{
+ struct drm_i915_gem_request *request =
+ container_of(fence, typeof(*request), submit);
+
+ /* Will be called from irq-context when using foreign DMA fences */
+
+ switch (state) {
+ case FENCE_COMPLETE:
+ request->engine->submit_request(request);
+ break;
+
+ case FENCE_FREE:
+ break;
+ }
+
+ return NOTIFY_DONE;
+}
+
/**
* i915_gem_request_alloc - allocate a request structure
*
struct i915_gem_context *ctx)
{
struct drm_i915_private *dev_priv = engine->i915;
- unsigned int reset_counter = i915_reset_counter(&dev_priv->gpu_error);
struct drm_i915_gem_request *req;
u32 seqno;
int ret;
* EIO if the GPU is already wedged, or EAGAIN to drop the struct_mutex
* and restart.
*/
- ret = i915_gem_check_wedge(reset_counter, dev_priv->mm.interruptible);
+ ret = i915_gem_check_wedge(dev_priv);
if (ret)
return ERR_PTR(ret);
if (req && i915_gem_request_completed(req))
i915_gem_request_retire(req);
- req = kmem_cache_zalloc(dev_priv->requests, GFP_KERNEL);
+ /* Beware: Dragons be flying overhead.
+ *
+ * We use RCU to look up requests in flight. The lookups may
+ * race with the request being allocated from the slab freelist.
+ * That is the request we are writing to here, may be in the process
+ * of being read by __i915_gem_active_get_rcu(). As such,
+ * we have to be very careful when overwriting the contents. During
+ * the RCU lookup, we change chase the request->engine pointer,
+ * read the request->fence.seqno and increment the reference count.
+ *
+ * The reference count is incremented atomically. If it is zero,
+ * the lookup knows the request is unallocated and complete. Otherwise,
+ * it is either still in use, or has been reallocated and reset
+ * with fence_init(). This increment is safe for release as we check
+ * that the request we have a reference to and matches the active
+ * request.
+ *
+ * Before we increment the refcount, we chase the request->engine
+ * pointer. We must not call kmem_cache_zalloc() or else we set
+ * that pointer to NULL and cause a crash during the lookup. If
+ * we see the request is completed (based on the value of the
+ * old engine and seqno), the lookup is complete and reports NULL.
+ * If we decide the request is not completed (new engine or seqno),
+ * then we grab a reference and double check that it is still the
+ * active request - which it won't be and restart the lookup.
+ *
+ * Do not use kmem_cache_zalloc() here!
+ */
+ req = kmem_cache_alloc(dev_priv->requests, GFP_KERNEL);
if (!req)
return ERR_PTR(-ENOMEM);
engine->fence_context,
seqno);
+ i915_sw_fence_init(&req->submit, submit_notify);
+
INIT_LIST_HEAD(&req->active_list);
req->i915 = dev_priv;
req->engine = engine;
req->ctx = i915_gem_context_get(ctx);
+ /* No zalloc, must clear what we need by hand */
+ req->previous_context = NULL;
+ req->file_priv = NULL;
+ req->batch = NULL;
+
/*
* Reserve space in the ring buffer for all the commands required to
* eventually emit this request. This is to guarantee that the
if (ret)
goto err_ctx;
+ /* Record the position of the start of the request so that
+ * should we detect the updated seqno part-way through the
+ * GPU processing the request, we never over-estimate the
+ * position of the head.
+ */
+ req->head = req->ring->tail;
+
return req;
err_ctx:
return ERR_PTR(ret);
}
+static int
+i915_gem_request_await_request(struct drm_i915_gem_request *to,
+ struct drm_i915_gem_request *from)
+{
+ int idx, ret;
+
+ GEM_BUG_ON(to == from);
+
+ if (to->engine == from->engine)
+ return 0;
+
+ idx = intel_engine_sync_index(from->engine, to->engine);
+ if (from->fence.seqno <= from->engine->semaphore.sync_seqno[idx])
+ return 0;
+
+ trace_i915_gem_ring_sync_to(to, from);
+ if (!i915.semaphores) {
+ if (!i915_spin_request(from, TASK_INTERRUPTIBLE, 2)) {
+ ret = i915_sw_fence_await_dma_fence(&to->submit,
+ &from->fence, 0,
+ GFP_KERNEL);
+ if (ret < 0)
+ return ret;
+ }
+ } else {
+ ret = to->engine->semaphore.sync_to(to, from);
+ if (ret)
+ return ret;
+ }
+
+ from->engine->semaphore.sync_seqno[idx] = from->fence.seqno;
+ return 0;
+}
+
+/**
+ * i915_gem_request_await_object - set this request to (async) wait upon a bo
+ *
+ * @to: request we are wishing to use
+ * @obj: object which may be in use on another ring.
+ *
+ * This code is meant to abstract object synchronization with the GPU.
+ * Conceptually we serialise writes between engines inside the GPU.
+ * We only allow one engine to write into a buffer at any time, but
+ * multiple readers. To ensure each has a coherent view of memory, we must:
+ *
+ * - If there is an outstanding write request to the object, the new
+ * request must wait for it to complete (either CPU or in hw, requests
+ * on the same ring will be naturally ordered).
+ *
+ * - If we are a write request (pending_write_domain is set), the new
+ * request must wait for outstanding read requests to complete.
+ *
+ * Returns 0 if successful, else propagates up the lower layer error.
+ */
+int
+i915_gem_request_await_object(struct drm_i915_gem_request *to,
+ struct drm_i915_gem_object *obj,
+ bool write)
+{
+ struct i915_gem_active *active;
+ unsigned long active_mask;
+ int idx;
+
+ if (write) {
+ active_mask = i915_gem_object_get_active(obj);
+ active = obj->last_read;
+ } else {
+ active_mask = 1;
+ active = &obj->last_write;
+ }
+
+ for_each_active(active_mask, idx) {
+ struct drm_i915_gem_request *request;
+ int ret;
+
+ request = i915_gem_active_peek(&active[idx],
+ &obj->base.dev->struct_mutex);
+ if (!request)
+ continue;
+
+ ret = i915_gem_request_await_request(to, request);
+ if (ret)
+ return ret;
+ }
+
+ return 0;
+}
+
static void i915_gem_mark_busy(const struct intel_engine_cs *engine)
{
struct drm_i915_private *dev_priv = engine->i915;
* request is not being tracked for completion but the work itself is
* going to happen on the hardware. This would be a Bad Thing(tm).
*/
-void __i915_add_request(struct drm_i915_gem_request *request,
- struct drm_i915_gem_object *obj,
- bool flush_caches)
+void __i915_add_request(struct drm_i915_gem_request *request, bool flush_caches)
{
- struct intel_engine_cs *engine;
- struct intel_ring *ring;
+ struct intel_engine_cs *engine = request->engine;
+ struct intel_ring *ring = request->ring;
+ struct drm_i915_gem_request *prev;
u32 request_start;
u32 reserved_tail;
int ret;
- if (WARN_ON(!request))
- return;
-
- engine = request->engine;
- ring = request->ring;
+ trace_i915_gem_request_add(request);
/*
* To ensure that this call will not fail, space for its emissions
WARN(ret, "engine->emit_flush() failed: %d!\n", ret);
}
- trace_i915_gem_request_add(request);
-
- request->head = request_start;
-
- /* Whilst this request exists, batch_obj will be on the
- * active_list, and so will hold the active reference. Only when this
- * request is retired will the the batch_obj be moved onto the
- * inactive_list and lose its active reference. Hence we do not need
- * to explicitly hold another reference here.
- */
- request->batch_obj = obj;
-
- /* Seal the request and mark it as pending execution. Note that
- * we may inspect this state, without holding any locks, during
- * hangcheck. Hence we apply the barrier to ensure that we do not
- * see a more recent value in the hws than we are tracking.
- */
- request->emitted_jiffies = jiffies;
- request->previous_seqno = engine->last_submitted_seqno;
- engine->last_submitted_seqno = request->fence.seqno;
- i915_gem_active_set(&engine->last_request, request);
- list_add_tail(&request->link, &engine->request_list);
- list_add_tail(&request->ring_link, &ring->request_list);
-
- /* Record the position of the start of the request so that
+ /* Record the position of the start of the breadcrumb so that
* should we detect the updated seqno part-way through the
* GPU processing the request, we never over-estimate the
- * position of the head.
+ * position of the ring's HEAD.
*/
request->postfix = ring->tail;
"for adding the request (%d bytes)\n",
reserved_tail, ret);
+ /* Seal the request and mark it as pending execution. Note that
+ * we may inspect this state, without holding any locks, during
+ * hangcheck. Hence we apply the barrier to ensure that we do not
+ * see a more recent value in the hws than we are tracking.
+ */
+
+ prev = i915_gem_active_raw(&engine->last_request,
+ &request->i915->drm.struct_mutex);
+ if (prev)
+ i915_sw_fence_await_sw_fence(&request->submit, &prev->submit,
+ &request->submitq);
+
+ request->emitted_jiffies = jiffies;
+ request->previous_seqno = engine->last_submitted_seqno;
+ engine->last_submitted_seqno = request->fence.seqno;
+ i915_gem_active_set(&engine->last_request, request);
+ list_add_tail(&request->link, &engine->request_list);
+ list_add_tail(&request->ring_link, &ring->request_list);
+
i915_gem_mark_busy(engine);
- engine->submit_request(request);
+
+ local_bh_disable();
+ i915_sw_fence_commit(&request->submit);
+ local_bh_enable(); /* Kick the execlists tasklet if just scheduled */
+}
+
+static void reset_wait_queue(wait_queue_head_t *q, wait_queue_t *wait)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&q->lock, flags);
+ if (list_empty(&wait->task_list))
+ __add_wait_queue(q, wait);
+ spin_unlock_irqrestore(&q->lock, flags);
}
static unsigned long local_clock_us(unsigned int *cpu)
/**
* i915_wait_request - wait until execution of request has finished
* @req: duh!
- * @interruptible: do an interruptible wait (normally yes)
+ * @flags: how to wait
* @timeout: in - how long to wait (NULL forever); out - how much time remaining
* @rps: client to charge for RPS boosting
*
* errno with remaining time filled in timeout argument.
*/
int i915_wait_request(struct drm_i915_gem_request *req,
- bool interruptible,
+ unsigned int flags,
s64 *timeout,
struct intel_rps_client *rps)
{
- int state = interruptible ? TASK_INTERRUPTIBLE : TASK_UNINTERRUPTIBLE;
+ const int state = flags & I915_WAIT_INTERRUPTIBLE ?
+ TASK_INTERRUPTIBLE : TASK_UNINTERRUPTIBLE;
DEFINE_WAIT(reset);
struct intel_wait wait;
unsigned long timeout_remain;
int ret = 0;
might_sleep();
+#if IS_ENABLED(CONFIG_LOCKDEP)
+ GEM_BUG_ON(!!lockdep_is_held(&req->i915->drm.struct_mutex) !=
+ !!(flags & I915_WAIT_LOCKED));
+#endif
if (i915_gem_request_completed(req))
return 0;
goto complete;
set_current_state(state);
- add_wait_queue(&req->i915->gpu_error.wait_queue, &reset);
+ if (flags & I915_WAIT_LOCKED)
+ add_wait_queue(&req->i915->gpu_error.wait_queue, &reset);
intel_wait_init(&wait, req->fence.seqno);
if (intel_engine_add_wait(req->engine, &wait))
if (__i915_request_irq_complete(req))
break;
+ /* If the GPU is hung, and we hold the lock, reset the GPU
+ * and then check for completion. On a full reset, the engine's
+ * HW seqno will be advanced passed us and we are complete.
+ * If we do a partial reset, we have to wait for the GPU to
+ * resume and update the breadcrumb.
+ *
+ * If we don't hold the mutex, we can just wait for the worker
+ * to come along and update the breadcrumb (either directly
+ * itself, or indirectly by recovering the GPU).
+ */
+ if (flags & I915_WAIT_LOCKED &&
+ i915_reset_in_progress(&req->i915->gpu_error)) {
+ __set_current_state(TASK_RUNNING);
+ i915_reset(req->i915);
+ reset_wait_queue(&req->i915->gpu_error.wait_queue,
+ &reset);
+ continue;
+ }
+
/* Only spin if we know the GPU is processing this request */
if (i915_spin_request(req, state, 2))
break;
}
- remove_wait_queue(&req->i915->gpu_error.wait_queue, &reset);
intel_engine_remove_wait(req->engine, &wait);
+ if (flags & I915_WAIT_LOCKED)
+ remove_wait_queue(&req->i915->gpu_error.wait_queue, &reset);
__set_current_state(TASK_RUNNING);
+
complete:
trace_i915_gem_request_wait_end(req);
return ret;
}
-static void engine_retire_requests(struct intel_engine_cs *engine)
+static bool engine_retire_requests(struct intel_engine_cs *engine)
{
struct drm_i915_gem_request *request, *next;
list_for_each_entry_safe(request, next, &engine->request_list, link) {
if (!i915_gem_request_completed(request))
- break;
+ return false;
i915_gem_request_retire(request);
}
+
+ return true;
}
void i915_gem_retire_requests(struct drm_i915_private *dev_priv)
{
struct intel_engine_cs *engine;
+ unsigned int tmp;
lockdep_assert_held(&dev_priv->drm.struct_mutex);
GEM_BUG_ON(!dev_priv->gt.awake);
- for_each_engine(engine, dev_priv) {
- engine_retire_requests(engine);
- if (!intel_engine_is_active(engine))
+ for_each_engine_masked(engine, dev_priv, dev_priv->gt.active_engines, tmp)
+ if (engine_retire_requests(engine))
dev_priv->gt.active_engines &= ~intel_engine_flag(engine);
- }
if (dev_priv->gt.active_engines == 0)
queue_delayed_work(dev_priv->wq,
#include <linux/fence.h>
#include "i915_gem.h"
+#include "i915_sw_fence.h"
struct intel_wait {
struct rb_node node;
* emission time to be associated with the request for tracking how far ahead
* of the GPU the submission is.
*
+ * When modifying this structure be very aware that we perform a lockless
+ * RCU lookup of it that may race against reallocation of the struct
+ * from the slab freelist. We intentionally do not zero the structure on
+ * allocation so that the lookup can use the dangling pointers (and is
+ * cogniscent that those pointers may be wrong). Instead, everything that
+ * needs to be initialised must be done so explicitly.
+ *
* The requests are reference counted.
*/
struct drm_i915_gem_request {
struct intel_ring *ring;
struct intel_signal_node signaling;
+ struct i915_sw_fence submit;
+ wait_queue_t submitq;
+
/** GEM sequence number associated with the previous request,
* when the HWS breadcrumb is equal to this the GPU is processing
* this request.
*/
u32 previous_seqno;
- /** Position in the ringbuffer of the start of the request */
+ /** Position in the ring of the start of the request */
u32 head;
/**
- * Position in the ringbuffer of the start of the postfix.
- * This is required to calculate the maximum available ringbuffer
- * space without overwriting the postfix.
+ * Position in the ring of the start of the postfix.
+ * This is required to calculate the maximum available ring space
+ * without overwriting the postfix.
*/
u32 postfix;
- /** Position in the ringbuffer of the end of the whole request */
+ /** Position in the ring of the end of the whole request */
u32 tail;
- /** Preallocate space in the ringbuffer for the emitting the request */
+ /** Position in the ring of the end of any workarounds after the tail */
+ u32 wa_tail;
+
+ /** Preallocate space in the ring for the emitting the request */
u32 reserved_space;
/**
/** Batch buffer related to this request if any (used for
* error state dump only).
*/
- struct drm_i915_gem_object *batch_obj;
+ struct i915_vma *batch;
struct list_head active_list;
/** Time at which this request was emitted, in jiffies. */
/** file_priv list entry for this request */
struct list_head client_list;
- /** process identifier submitting this request */
- struct pid *pid;
-
- /**
- * The ELSP only accepts two elements at a time, so we queue
- * context/tail pairs on a given queue (ring->execlist_queue) until the
- * hardware is available. The queue serves a double purpose: we also use
- * it to keep track of the up to 2 contexts currently in the hardware
- * (usually one in execution and the other queued up by the GPU): We
- * only remove elements from the head of the queue when the hardware
- * informs us that an element has been completed.
- *
- * All accesses to the queue are mediated by a spinlock
- * (ring->execlist_lock).
- */
-
- /** Execlist link in the submission queue.*/
+ /** Link in the execlist submission queue, guarded by execlist_lock. */
struct list_head execlist_link;
-
- /** Execlists no. of times this request has been sent to the ELSP */
- int elsp_submitted;
-
- /** Execlists context hardware id. */
- unsigned int ctx_hw_id;
};
extern const struct fence_ops i915_fence_ops;
*pdst = src;
}
-void __i915_add_request(struct drm_i915_gem_request *req,
- struct drm_i915_gem_object *batch_obj,
- bool flush_caches);
+int
+i915_gem_request_await_object(struct drm_i915_gem_request *to,
+ struct drm_i915_gem_object *obj,
+ bool write);
+
+void __i915_add_request(struct drm_i915_gem_request *req, bool flush_caches);
#define i915_add_request(req) \
- __i915_add_request(req, NULL, true)
+ __i915_add_request(req, true)
#define i915_add_request_no_flush(req) \
- __i915_add_request(req, NULL, false)
+ __i915_add_request(req, false)
struct intel_rps_client;
#define NO_WAITBOOST ERR_PTR(-1)
#define IS_RPS_USER(p) (!IS_ERR_OR_NULL(p))
int i915_wait_request(struct drm_i915_gem_request *req,
- bool interruptible,
+ unsigned int flags,
s64 *timeout,
struct intel_rps_client *rps)
__attribute__((nonnull(1)));
+#define I915_WAIT_INTERRUPTIBLE BIT(0)
+#define I915_WAIT_LOCKED BIT(1) /* struct_mutex held, handle GPU reset */
static inline u32 intel_engine_get_seqno(struct intel_engine_cs *engine);
}
/**
- * i915_gem_active_peek - report the active request being monitored
+ * i915_gem_active_raw - return the active request
* @active - the active tracker
*
- * i915_gem_active_peek() returns the current request being tracked if
- * still active, or NULL. It does not obtain a reference on the request
- * for the caller, so the caller must hold struct_mutex.
+ * i915_gem_active_raw() returns the current request being tracked, or NULL.
+ * It does not obtain a reference on the request for the caller, so the caller
+ * must hold struct_mutex.
*/
static inline struct drm_i915_gem_request *
-i915_gem_active_peek(const struct i915_gem_active *active, struct mutex *mutex)
+i915_gem_active_raw(const struct i915_gem_active *active, struct mutex *mutex)
{
- struct drm_i915_gem_request *request;
-
- request = rcu_dereference_protected(active->request,
- lockdep_is_held(mutex));
- if (!request || i915_gem_request_completed(request))
- return NULL;
-
- return request;
+ return rcu_dereference_protected(active->request,
+ lockdep_is_held(mutex));
}
/**
- * i915_gem_active_peek_rcu - report the active request being monitored
+ * i915_gem_active_peek - report the active request being monitored
* @active - the active tracker
*
- * i915_gem_active_peek_rcu() returns the current request being tracked if
+ * i915_gem_active_peek() returns the current request being tracked if
* still active, or NULL. It does not obtain a reference on the request
- * for the caller, and inspection of the request is only valid under
- * the RCU lock.
+ * for the caller, so the caller must hold struct_mutex.
*/
static inline struct drm_i915_gem_request *
-i915_gem_active_peek_rcu(const struct i915_gem_active *active)
+i915_gem_active_peek(const struct i915_gem_active *active, struct mutex *mutex)
{
struct drm_i915_gem_request *request;
- request = rcu_dereference(active->request);
+ request = i915_gem_active_raw(active, mutex);
if (!request || i915_gem_request_completed(request))
return NULL;
* just report the active tracker is idle. If the new request is
* incomplete, then we acquire a reference on it and check that
* it remained the active request.
+ *
+ * It is then imperative that we do not zero the request on
+ * reallocation, so that we can chase the dangling pointers!
+ * See i915_gem_request_alloc().
*/
do {
struct drm_i915_gem_request *request;
if (!request || i915_gem_request_completed(request))
return NULL;
+ /* An especially silly compiler could decide to recompute the
+ * result of i915_gem_request_completed, more specifically
+ * re-emit the load for request->fence.seqno. A race would catch
+ * a later seqno value, which could flip the result from true to
+ * false. Which means part of the instructions below might not
+ * be executed, while later on instructions are executed. Due to
+ * barriers within the refcounting the inconsistency can't reach
+ * past the call to i915_gem_request_get_rcu, but not executing
+ * that while still executing i915_gem_request_put() creates
+ * havoc enough. Prevent this with a compiler barrier.
+ */
+ barrier();
+
request = i915_gem_request_get_rcu(request);
/* What stops the following rcu_access_pointer() from occurring
* incremented) then the following read for rcu_access_pointer()
* must occur after the atomic operation and so confirm
* that this request is the one currently being tracked.
+ *
+ * The corresponding write barrier is part of
+ * rcu_assign_pointer().
*/
if (!request || request == rcu_access_pointer(active->request))
return rcu_pointer_handoff(request);
if (!request)
return 0;
- return i915_wait_request(request, true, NULL, NULL);
+ return i915_wait_request(request,
+ I915_WAIT_INTERRUPTIBLE | I915_WAIT_LOCKED,
+ NULL, NULL);
}
/**
* i915_gem_active_wait_unlocked - waits until the request is completed
* @active - the active request on which to wait
- * @interruptible - whether the wait can be woken by a userspace signal
+ * @flags - how to wait
* @timeout - how long to wait at most
* @rps - userspace client to charge for a waitboost
*
*/
static inline int
i915_gem_active_wait_unlocked(const struct i915_gem_active *active,
- bool interruptible,
+ unsigned int flags,
s64 *timeout,
struct intel_rps_client *rps)
{
request = i915_gem_active_get_unlocked(active);
if (request) {
- ret = i915_wait_request(request, interruptible, timeout, rps);
+ ret = i915_wait_request(request, flags, timeout, rps);
i915_gem_request_put(request);
}
struct drm_i915_gem_request *request;
int ret;
- request = rcu_dereference_protected(active->request,
- lockdep_is_held(mutex));
+ request = i915_gem_active_raw(active, mutex);
if (!request)
return 0;
- ret = i915_wait_request(request, true, NULL, NULL);
+ ret = i915_wait_request(request,
+ I915_WAIT_INTERRUPTIBLE | I915_WAIT_LOCKED,
+ NULL, NULL);
if (ret)
return ret;
unsigned long timeout = jiffies + msecs_to_jiffies_timeout(timeout_ms);
do {
- if (i915_gem_wait_for_idle(dev_priv, false) == 0 &&
+ if (i915_gem_wait_for_idle(dev_priv, 0) == 0 &&
i915_gem_shrinker_lock(&dev_priv->drm, &slu->unlock))
break;
return NOTIFY_DONE;
/* Force everything onto the inactive lists */
- ret = i915_gem_wait_for_idle(dev_priv, false);
+ ret = i915_gem_wait_for_idle(dev_priv, I915_WAIT_LOCKED);
if (ret)
goto out;
static unsigned long i915_stolen_to_physical(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
struct i915_ggtt *ggtt = &dev_priv->ggtt;
struct resource *r;
u32 base;
if (INTEL_INFO(dev)->gen >= 3) {
u32 bsm;
- pci_read_config_dword(dev->pdev, INTEL_BSM, &bsm);
+ pci_read_config_dword(pdev, INTEL_BSM, &bsm);
base = bsm & INTEL_BSM_MASK;
} else if (IS_I865G(dev)) {
+ u32 tseg_size = 0;
u16 toud = 0;
+ u8 tmp;
- /*
- * FIXME is the graphics stolen memory region
- * always at TOUD? Ie. is it always the last
- * one to be allocated by the BIOS?
- */
- pci_bus_read_config_word(dev->pdev->bus, PCI_DEVFN(0, 0),
+ pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0),
+ I845_ESMRAMC, &tmp);
+
+ if (tmp & TSEG_ENABLE) {
+ switch (tmp & I845_TSEG_SIZE_MASK) {
+ case I845_TSEG_SIZE_512K:
+ tseg_size = KB(512);
+ break;
+ case I845_TSEG_SIZE_1M:
+ tseg_size = MB(1);
+ break;
+ }
+ }
+
+ pci_bus_read_config_word(pdev->bus, PCI_DEVFN(0, 0),
I865_TOUD, &toud);
- base = toud << 16;
+ base = (toud << 16) + tseg_size;
} else if (IS_I85X(dev)) {
u32 tseg_size = 0;
u32 tom;
u8 tmp;
- pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0),
+ pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0),
I85X_ESMRAMC, &tmp);
if (tmp & TSEG_ENABLE)
tseg_size = MB(1);
- pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 1),
+ pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 1),
I85X_DRB3, &tmp);
tom = tmp * MB(32);
u32 tom;
u8 tmp;
- pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0),
+ pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0),
I845_ESMRAMC, &tmp);
if (tmp & TSEG_ENABLE) {
}
}
- pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0),
+ pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0),
I830_DRB3, &tmp);
tom = tmp * MB(32);
u32 tom;
u8 tmp;
- pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0),
+ pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0),
I830_ESMRAMC, &tmp);
if (tmp & TSEG_ENABLE) {
tseg_size = KB(512);
}
- pci_bus_read_config_byte(dev->pdev->bus, PCI_DEVFN(0, 0),
+ pci_bus_read_config_byte(pdev->bus, PCI_DEVFN(0, 0),
I830_DRB3, &tmp);
tom = tmp * MB(32);
if (gtt_offset == I915_GTT_OFFSET_NONE)
return obj;
- vma = i915_gem_obj_lookup_or_create_vma(obj, &ggtt->base);
+ vma = i915_gem_obj_lookup_or_create_vma(obj, &ggtt->base, NULL);
if (IS_ERR(vma)) {
ret = PTR_ERR(vma);
goto err;
goto err;
}
+ vma->pages = obj->pages;
vma->flags |= I915_VMA_GLOBAL_BIND;
__i915_vma_set_map_and_fenceable(vma);
list_move_tail(&vma->vm_link, &ggtt->base.inactive_list);
return true;
}
-/* Is the current GTT allocation valid for the change in tiling? */
-static bool
-i915_gem_object_fence_ok(struct drm_i915_gem_object *obj, int tiling_mode)
+static bool i915_vma_fence_prepare(struct i915_vma *vma, int tiling_mode)
{
- struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
+ struct drm_i915_private *dev_priv = to_i915(vma->vm->dev);
u32 size;
- if (tiling_mode == I915_TILING_NONE)
- return true;
-
- if (INTEL_GEN(dev_priv) >= 4)
+ if (!i915_vma_is_map_and_fenceable(vma))
return true;
- if (IS_GEN3(dev_priv)) {
- if (i915_gem_obj_ggtt_offset(obj) & ~I915_FENCE_START_MASK)
+ if (INTEL_GEN(dev_priv) == 3) {
+ if (vma->node.start & ~I915_FENCE_START_MASK)
return false;
} else {
- if (i915_gem_obj_ggtt_offset(obj) & ~I830_FENCE_START_MASK)
+ if (vma->node.start & ~I830_FENCE_START_MASK)
return false;
}
- size = i915_gem_get_ggtt_size(dev_priv, obj->base.size, tiling_mode);
- if (i915_gem_obj_ggtt_size(obj) != size)
+ size = i915_gem_get_ggtt_size(dev_priv, vma->size, tiling_mode);
+ if (vma->node.size < size)
return false;
- if (i915_gem_obj_ggtt_offset(obj) & (size - 1))
+ if (vma->node.start & (size - 1))
return false;
return true;
}
+/* Make the current GTT allocation valid for the change in tiling. */
+static int
+i915_gem_object_fence_prepare(struct drm_i915_gem_object *obj, int tiling_mode)
+{
+ struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
+ struct i915_vma *vma;
+ int ret;
+
+ if (tiling_mode == I915_TILING_NONE)
+ return 0;
+
+ if (INTEL_GEN(dev_priv) >= 4)
+ return 0;
+
+ list_for_each_entry(vma, &obj->vma_list, obj_link) {
+ if (i915_vma_fence_prepare(vma, tiling_mode))
+ continue;
+
+ ret = i915_vma_unbind(vma);
+ if (ret)
+ return ret;
+ }
+
+ return 0;
+}
+
/**
* i915_gem_set_tiling - IOCTL handler to set tiling mode
* @dev: DRM device
struct drm_i915_gem_set_tiling *args = data;
struct drm_i915_private *dev_priv = to_i915(dev);
struct drm_i915_gem_object *obj;
- int ret = 0;
+ int err = 0;
/* Make sure we don't cross-contaminate obj->tiling_and_stride */
BUILD_BUG_ON(I915_TILING_LAST & STRIDE_MASK);
mutex_lock(&dev->struct_mutex);
if (obj->pin_display || obj->framebuffer_references) {
- ret = -EBUSY;
+ err = -EBUSY;
goto err;
}
* has to also include the unfenced register the GPU uses
* whilst executing a fenced command for an untiled object.
*/
- if (obj->map_and_fenceable &&
- !i915_gem_object_fence_ok(obj, args->tiling_mode))
- ret = i915_vma_unbind(i915_gem_obj_to_ggtt(obj));
- if (ret == 0) {
+ err = i915_gem_object_fence_prepare(obj, args->tiling_mode);
+ if (!err) {
+ struct i915_vma *vma;
+
if (obj->pages &&
obj->madv == I915_MADV_WILLNEED &&
dev_priv->quirks & QUIRK_PIN_SWIZZLED_PAGES) {
i915_gem_object_pin_pages(obj);
}
- obj->fence_dirty =
- !i915_gem_active_is_idle(&obj->last_fence,
- &dev->struct_mutex) ||
- obj->fence_reg != I915_FENCE_REG_NONE;
+ list_for_each_entry(vma, &obj->vma_list, obj_link) {
+ if (!vma->fence)
+ continue;
+ vma->fence->dirty = true;
+ }
obj->tiling_and_stride =
args->stride | args->tiling_mode;
intel_runtime_pm_put(dev_priv);
- return ret;
+ return err;
}
/**
for_each_active(active, idx)
i915_gem_active_wait_unlocked(&obj->last_read[idx],
- false, NULL, NULL);
+ 0, NULL, NULL);
}
static void cancel_userptr(struct work_struct *work)
}
}
obj->userptr.work = ERR_PTR(ret);
- if (ret)
- __i915_gem_userptr_set_active(obj, false);
}
obj->userptr.workers--;
* to the vma (discard or cloning) which should prevent the more
* egregious cases from causing harm.
*/
- if (IS_ERR(obj->userptr.work)) {
- /* active flag will have been dropped already by the worker */
- ret = PTR_ERR(obj->userptr.work);
- obj->userptr.work = NULL;
- return ret;
- }
- if (obj->userptr.work)
+
+ if (obj->userptr.work) {
/* active flag should still be held for the pending work */
- return -EAGAIN;
+ if (IS_ERR(obj->userptr.work))
+ return PTR_ERR(obj->userptr.work);
+ else
+ return -EAGAIN;
+ }
/* Let the mmu-notifier know that we have begun and need cancellation */
ret = __i915_gem_userptr_set_active(obj, true);
}
}
-static const char *pin_flag(int pinned)
-{
- if (pinned > 0)
- return " P";
- else if (pinned < 0)
- return " p";
- else
- return "";
-}
-
static const char *tiling_flag(int tiling)
{
switch (tiling) {
{
int i;
- err_printf(m, " %s [%d]:\n", name, count);
+ err_printf(m, "%s [%d]:\n", name, count);
while (count--) {
err_printf(m, " %08x_%08x %8u %02x %02x [ ",
err_printf(m, "%02x ", err->rseqno[i]);
err_printf(m, "] %02x", err->wseqno);
- err_puts(m, pin_flag(err->pinned));
err_puts(m, tiling_flag(err->tiling));
err_puts(m, dirty_flag(err->dirty));
err_puts(m, purgeable_flag(err->purgeable));
err_printf(m, " HEAD: 0x%08x\n", ee->head);
err_printf(m, " TAIL: 0x%08x\n", ee->tail);
err_printf(m, " CTL: 0x%08x\n", ee->ctl);
+ err_printf(m, " MODE: 0x%08x\n", ee->mode);
err_printf(m, " HWS: 0x%08x\n", ee->hws);
err_printf(m, " ACTHD: 0x%08x %08x\n",
(u32)(ee->acthd>>32), (u32)ee->acthd);
err_printf(m, " IPEIR: 0x%08x\n", ee->ipeir);
err_printf(m, " IPEHR: 0x%08x\n", ee->ipehr);
err_printf(m, " INSTDONE: 0x%08x\n", ee->instdone);
+ if (ee->batchbuffer) {
+ u64 start = ee->batchbuffer->gtt_offset;
+ u64 end = start + ee->batchbuffer->gtt_size;
+
+ err_printf(m, " batch: [0x%08x_%08x, 0x%08x_%08x]\n",
+ upper_32_bits(start), lower_32_bits(start),
+ upper_32_bits(end), lower_32_bits(end));
+ }
if (INTEL_GEN(m->i915) >= 4) {
- err_printf(m, " BBADDR: 0x%08x %08x\n",
+ err_printf(m, " BBADDR: 0x%08x_%08x\n",
(u32)(ee->bbaddr>>32), (u32)ee->bbaddr);
err_printf(m, " BB_STATE: 0x%08x\n", ee->bbstate);
err_printf(m, " INSTPS: 0x%08x\n", ee->instps);
}
}
+static void err_print_capabilities(struct drm_i915_error_state_buf *m,
+ const struct intel_device_info *info)
+{
+#define PRINT_FLAG(x) err_printf(m, #x ": %s\n", yesno(info->x))
+#define SEP_SEMICOLON ;
+ DEV_INFO_FOR_EACH_FLAG(PRINT_FLAG, SEP_SEMICOLON);
+#undef PRINT_FLAG
+#undef SEP_SEMICOLON
+}
+
int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
const struct i915_error_state_file_priv *error_priv)
{
struct drm_device *dev = error_priv->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
struct drm_i915_error_state *error = error_priv->error;
struct drm_i915_error_object *obj;
int i, j, offset, elt;
err_printf(m, "Time: %ld s %ld us\n", error->time.tv_sec,
error->time.tv_usec);
err_printf(m, "Kernel: " UTS_RELEASE "\n");
+ err_print_capabilities(m, &error->device_info);
max_hangcheck_score = 0;
for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
if (error->engine[i].hangcheck_score > max_hangcheck_score)
}
err_printf(m, "Reset count: %u\n", error->reset_count);
err_printf(m, "Suspend count: %u\n", error->suspend_count);
- err_printf(m, "PCI ID: 0x%04x\n", dev->pdev->device);
- err_printf(m, "PCI Revision: 0x%02x\n", dev->pdev->revision);
+ err_printf(m, "PCI ID: 0x%04x\n", pdev->device);
+ err_printf(m, "PCI Revision: 0x%02x\n", pdev->revision);
err_printf(m, "PCI Subsystem: %04x:%04x\n",
- dev->pdev->subsystem_vendor,
- dev->pdev->subsystem_device);
+ pdev->subsystem_vendor,
+ pdev->subsystem_device);
err_printf(m, "IOMMU enabled?: %d\n", error->iommu);
if (HAS_CSR(dev)) {
error_print_engine(m, &error->engine[i]);
}
- for (i = 0; i < error->vm_count; i++) {
- err_printf(m, "vm[%d]\n", i);
+ for (i = 0; i < ARRAY_SIZE(error->active_vm); i++) {
+ char buf[128];
+ int len, first = 1;
+
+ if (!error->active_vm[i])
+ break;
+
+ len = scnprintf(buf, sizeof(buf), "Active (");
+ for (j = 0; j < ARRAY_SIZE(error->engine); j++) {
+ if (error->engine[j].vm != error->active_vm[i])
+ continue;
- print_error_buffers(m, "Active",
+ len += scnprintf(buf + len, sizeof(buf), "%s%s",
+ first ? "" : ", ",
+ dev_priv->engine[j].name);
+ first = 0;
+ }
+ scnprintf(buf + len, sizeof(buf), ")");
+ print_error_buffers(m, buf,
error->active_bo[i],
error->active_bo_count[i]);
-
- print_error_buffers(m, "Pinned",
- error->pinned_bo[i],
- error->pinned_bo_count[i]);
}
+ print_error_buffers(m, "Pinned (global)",
+ error->pinned_bo,
+ error->pinned_bo_count);
+
for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
struct drm_i915_error_engine *ee = &error->engine[i];
dev_priv->engine[i].name,
ee->num_requests);
for (j = 0; j < ee->num_requests; j++) {
- err_printf(m, " seqno 0x%08x, emitted %ld, tail 0x%08x\n",
+ err_printf(m, " pid %d, seqno 0x%08x, emitted %ld, head 0x%08x, tail 0x%08x\n",
+ ee->requests[j].pid,
ee->requests[j].seqno,
ee->requests[j].jiffies,
+ ee->requests[j].head,
ee->requests[j].tail);
}
}
- if (ee->num_waiters) {
+ if (IS_ERR(ee->waiters)) {
+ err_printf(m, "%s --- ? waiters [unable to acquire spinlock]\n",
+ dev_priv->engine[i].name);
+ } else if (ee->num_waiters) {
err_printf(m, "%s --- %d waiters\n",
dev_priv->engine[i].name,
ee->num_waiters);
}
}
- if ((obj = error->semaphore_obj)) {
+ if ((obj = error->semaphore)) {
err_printf(m, "Semaphore page = 0x%08x\n",
lower_32_bits(obj->gtt_offset));
for (elt = 0; elt < PAGE_SIZE/16; elt += 4) {
i915_error_object_free(ee->wa_ctx);
kfree(ee->requests);
- kfree(ee->waiters);
+ if (!IS_ERR_OR_NULL(ee->waiters))
+ kfree(ee->waiters);
}
- i915_error_object_free(error->semaphore_obj);
+ i915_error_object_free(error->semaphore);
- for (i = 0; i < error->vm_count; i++)
+ for (i = 0; i < ARRAY_SIZE(error->active_bo); i++)
kfree(error->active_bo[i]);
-
- kfree(error->active_bo);
- kfree(error->active_bo_count);
kfree(error->pinned_bo);
- kfree(error->pinned_bo_count);
+
kfree(error->overlay);
kfree(error->display);
kfree(error);
static struct drm_i915_error_object *
i915_error_object_create(struct drm_i915_private *dev_priv,
- struct drm_i915_gem_object *src,
- struct i915_address_space *vm)
+ struct i915_vma *vma)
{
struct i915_ggtt *ggtt = &dev_priv->ggtt;
+ struct drm_i915_gem_object *src;
struct drm_i915_error_object *dst;
- struct i915_vma *vma = NULL;
int num_pages;
bool use_ggtt;
int i = 0;
u64 reloc_offset;
- if (src == NULL || src->pages == NULL)
+ if (!vma)
+ return NULL;
+
+ src = vma->obj;
+ if (!src->pages)
return NULL;
num_pages = src->base.size >> PAGE_SHIFT;
dst = kmalloc(sizeof(*dst) + num_pages * sizeof(u32 *), GFP_ATOMIC);
- if (dst == NULL)
+ if (!dst)
return NULL;
- if (i915_gem_obj_bound(src, vm))
- dst->gtt_offset = i915_gem_obj_offset(src, vm);
- else
- dst->gtt_offset = -1;
+ dst->gtt_offset = vma->node.start;
+ dst->gtt_size = vma->node.size;
reloc_offset = dst->gtt_offset;
- if (i915_is_ggtt(vm))
- vma = i915_gem_obj_to_ggtt(src);
use_ggtt = (src->cache_level == I915_CACHE_NONE &&
- vma && (vma->flags & I915_VMA_GLOBAL_BIND) &&
+ (vma->flags & I915_VMA_GLOBAL_BIND) &&
reloc_offset + num_pages * PAGE_SIZE <= ggtt->mappable_end);
/* Cannot access stolen address directly, try to use the aperture */
if (src->stolen) {
use_ggtt = true;
- if (!(vma && vma->flags & I915_VMA_GLOBAL_BIND))
+ if (!(vma->flags & I915_VMA_GLOBAL_BIND))
goto unwind;
- reloc_offset = i915_gem_obj_ggtt_offset(src);
+ reloc_offset = vma->node.start;
if (reloc_offset + num_pages * PAGE_SIZE > ggtt->mappable_end)
goto unwind;
}
* captures what the GPU read.
*/
- s = io_mapping_map_atomic_wc(ggtt->mappable,
+ s = io_mapping_map_atomic_wc(&ggtt->mappable,
reloc_offset);
memcpy_fromio(d, s, PAGE_SIZE);
io_mapping_unmap_atomic(s);
kfree(dst);
return NULL;
}
-#define i915_error_ggtt_object_create(dev_priv, src) \
- i915_error_object_create((dev_priv), (src), &(dev_priv)->ggtt.base)
/* The error capture is special as tries to run underneath the normal
* locking rules - so we use the raw version of the i915_gem_active lookup.
err->gtt_offset = vma->node.start;
err->read_domains = obj->base.read_domains;
err->write_domain = obj->base.write_domain;
- err->fence_reg = obj->fence_reg;
- err->pinned = 0;
- if (i915_gem_obj_is_pinned(obj))
- err->pinned = 1;
+ err->fence_reg = vma->fence ? vma->fence->id : -1;
err->tiling = i915_gem_object_get_tiling(obj);
err->dirty = obj->dirty;
err->purgeable = obj->madv != I915_MADV_WILLNEED;
err->cache_level = obj->cache_level;
}
-static u32 capture_active_bo(struct drm_i915_error_buffer *err,
- int count, struct list_head *head)
+static u32 capture_error_bo(struct drm_i915_error_buffer *err,
+ int count, struct list_head *head,
+ bool pinned_only)
{
struct i915_vma *vma;
int i = 0;
list_for_each_entry(vma, head, vm_link) {
+ if (pinned_only && !i915_vma_is_pinned(vma))
+ continue;
+
capture_bo(err++, vma);
if (++i == count)
break;
return i;
}
-static u32 capture_pinned_bo(struct drm_i915_error_buffer *err,
- int count, struct list_head *head,
- struct i915_address_space *vm)
-{
- struct drm_i915_gem_object *obj;
- struct drm_i915_error_buffer * const first = err;
- struct drm_i915_error_buffer * const last = err + count;
-
- list_for_each_entry(obj, head, global_list) {
- struct i915_vma *vma;
-
- if (err == last)
- break;
-
- list_for_each_entry(vma, &obj->vma_list, obj_link)
- if (vma->vm == vm && i915_vma_is_pinned(vma))
- capture_bo(err++, vma);
- }
-
- return err - first;
-}
-
/* Generate a semi-unique error code. The code is not meant to have meaning, The
* code's only purpose is to try to prevent false duplicated bug reports by
* grossly estimating a GPU error state.
struct intel_engine_cs *to;
enum intel_engine_id id;
- if (!error->semaphore_obj)
+ if (!error->semaphore)
return;
for_each_engine_id(to, dev_priv, id) {
signal_offset =
(GEN8_SIGNAL_OFFSET(engine, id) & (PAGE_SIZE - 1)) / 4;
- tmp = error->semaphore_obj->pages[0];
+ tmp = error->semaphore->pages[0];
idx = intel_engine_sync_index(engine, to);
ee->semaphore_mboxes[idx] = tmp[signal_offset];
ee->num_waiters = 0;
ee->waiters = NULL;
- spin_lock(&b->lock);
+ if (RB_EMPTY_ROOT(&b->waiters))
+ return;
+
+ if (!spin_trylock(&b->lock)) {
+ ee->waiters = ERR_PTR(-EDEADLK);
+ return;
+ }
+
count = 0;
for (rb = rb_first(&b->waiters); rb != NULL; rb = rb_next(rb))
count++;
if (!waiter)
return;
- ee->waiters = waiter;
+ if (!spin_trylock(&b->lock)) {
+ kfree(waiter);
+ ee->waiters = ERR_PTR(-EDEADLK);
+ return;
+ }
- spin_lock(&b->lock);
+ ee->waiters = waiter;
for (rb = rb_first(&b->waiters); rb; rb = rb_next(rb)) {
struct intel_wait *w = container_of(rb, typeof(*w), node);
ee->head = I915_READ_HEAD(engine);
ee->tail = I915_READ_TAIL(engine);
ee->ctl = I915_READ_CTL(engine);
+ if (INTEL_GEN(dev_priv) > 2)
+ ee->mode = I915_READ_MODE(engine);
- if (I915_NEED_GFX_HWS(dev_priv)) {
+ if (!HWS_NEEDS_PHYSICAL(dev_priv)) {
i915_reg_t mmio;
if (IS_GEN7(dev_priv)) {
}
}
-
-static void i915_gem_record_active_context(struct intel_engine_cs *engine,
- struct drm_i915_error_state *error,
- struct drm_i915_error_engine *ee)
+static void engine_record_requests(struct intel_engine_cs *engine,
+ struct drm_i915_gem_request *first,
+ struct drm_i915_error_engine *ee)
{
- struct drm_i915_private *dev_priv = engine->i915;
- struct drm_i915_gem_object *obj;
+ struct drm_i915_gem_request *request;
+ int count;
+
+ count = 0;
+ request = first;
+ list_for_each_entry_from(request, &engine->request_list, link)
+ count++;
+ if (!count)
+ return;
- /* Currently render ring is the only HW context user */
- if (engine->id != RCS || !error->ccid)
+ ee->requests = kcalloc(count, sizeof(*ee->requests), GFP_ATOMIC);
+ if (!ee->requests)
return;
- list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
- if (!i915_gem_obj_ggtt_bound(obj))
- continue;
+ ee->num_requests = count;
- if ((error->ccid & PAGE_MASK) == i915_gem_obj_ggtt_offset(obj)) {
- ee->ctx = i915_error_ggtt_object_create(dev_priv, obj);
+ count = 0;
+ request = first;
+ list_for_each_entry_from(request, &engine->request_list, link) {
+ struct drm_i915_error_request *erq;
+
+ if (count >= ee->num_requests) {
+ /*
+ * If the ring request list was changed in
+ * between the point where the error request
+ * list was created and dimensioned and this
+ * point then just exit early to avoid crashes.
+ *
+ * We don't need to communicate that the
+ * request list changed state during error
+ * state capture and that the error state is
+ * slightly incorrect as a consequence since we
+ * are typically only interested in the request
+ * list state at the point of error state
+ * capture, not in any changes happening during
+ * the capture.
+ */
break;
}
+
+ erq = &ee->requests[count++];
+ erq->seqno = request->fence.seqno;
+ erq->jiffies = request->emitted_jiffies;
+ erq->head = request->head;
+ erq->tail = request->tail;
+
+ rcu_read_lock();
+ erq->pid = request->ctx->pid ? pid_nr(request->ctx->pid) : 0;
+ rcu_read_unlock();
}
+ ee->num_requests = count;
}
static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
struct drm_i915_error_state *error)
{
struct i915_ggtt *ggtt = &dev_priv->ggtt;
- struct drm_i915_gem_request *request;
- int i, count;
+ int i;
- if (dev_priv->semaphore_obj) {
- error->semaphore_obj =
- i915_error_ggtt_object_create(dev_priv,
- dev_priv->semaphore_obj);
- }
+ error->semaphore =
+ i915_error_object_create(dev_priv, dev_priv->semaphore);
for (i = 0; i < I915_NUM_ENGINES; i++) {
struct intel_engine_cs *engine = &dev_priv->engine[i];
struct drm_i915_error_engine *ee = &error->engine[i];
+ struct drm_i915_gem_request *request;
ee->pid = -1;
ee->engine_id = -1;
request = i915_gem_find_active_request(engine);
if (request) {
- struct i915_address_space *vm;
struct intel_ring *ring;
+ struct pid *pid;
- vm = request->ctx->ppgtt ?
+ ee->vm = request->ctx->ppgtt ?
&request->ctx->ppgtt->base : &ggtt->base;
/* We need to copy these to an anonymous buffer
*/
ee->batchbuffer =
i915_error_object_create(dev_priv,
- request->batch_obj,
- vm);
+ request->batch);
if (HAS_BROKEN_CS_TLB(dev_priv))
ee->wa_batchbuffer =
- i915_error_ggtt_object_create(dev_priv,
- engine->scratch.obj);
+ i915_error_object_create(dev_priv,
+ engine->scratch);
+
+ ee->ctx =
+ i915_error_object_create(dev_priv,
+ request->ctx->engine[i].state);
- if (request->pid) {
+ pid = request->ctx->pid;
+ if (pid) {
struct task_struct *task;
rcu_read_lock();
- task = pid_task(request->pid, PIDTYPE_PID);
+ task = pid_task(pid, PIDTYPE_PID);
if (task) {
strcpy(ee->comm, task->comm);
ee->pid = task->pid;
ee->cpu_ring_head = ring->head;
ee->cpu_ring_tail = ring->tail;
ee->ringbuffer =
- i915_error_ggtt_object_create(dev_priv,
- ring->obj);
- }
-
- ee->hws_page =
- i915_error_ggtt_object_create(dev_priv,
- engine->status_page.obj);
+ i915_error_object_create(dev_priv, ring->vma);
- ee->wa_ctx = i915_error_ggtt_object_create(dev_priv,
- engine->wa_ctx.obj);
-
- i915_gem_record_active_context(engine, error, ee);
-
- count = 0;
- list_for_each_entry(request, &engine->request_list, link)
- count++;
-
- ee->num_requests = count;
- ee->requests =
- kcalloc(count, sizeof(*ee->requests), GFP_ATOMIC);
- if (!ee->requests) {
- ee->num_requests = 0;
- continue;
+ engine_record_requests(engine, request, ee);
}
- count = 0;
- list_for_each_entry(request, &engine->request_list, link) {
- struct drm_i915_error_request *erq;
-
- if (count >= ee->num_requests) {
- /*
- * If the ring request list was changed in
- * between the point where the error request
- * list was created and dimensioned and this
- * point then just exit early to avoid crashes.
- *
- * We don't need to communicate that the
- * request list changed state during error
- * state capture and that the error state is
- * slightly incorrect as a consequence since we
- * are typically only interested in the request
- * list state at the point of error state
- * capture, not in any changes happening during
- * the capture.
- */
- break;
- }
+ ee->hws_page =
+ i915_error_object_create(dev_priv,
+ engine->status_page.vma);
- erq = &ee->requests[count++];
- erq->seqno = request->fence.seqno;
- erq->jiffies = request->emitted_jiffies;
- erq->tail = request->postfix;
- }
+ ee->wa_ctx =
+ i915_error_object_create(dev_priv, engine->wa_ctx.vma);
}
}
-/* FIXME: Since pin count/bound list is global, we duplicate what we capture per
- * VM.
- */
static void i915_gem_capture_vm(struct drm_i915_private *dev_priv,
struct drm_i915_error_state *error,
struct i915_address_space *vm,
- const int ndx)
+ int idx)
{
- struct drm_i915_error_buffer *active_bo = NULL, *pinned_bo = NULL;
- struct drm_i915_gem_object *obj;
+ struct drm_i915_error_buffer *active_bo;
struct i915_vma *vma;
- int i;
+ int count;
- i = 0;
+ count = 0;
list_for_each_entry(vma, &vm->active_list, vm_link)
- i++;
- error->active_bo_count[ndx] = i;
-
- list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) {
- list_for_each_entry(vma, &obj->vma_list, obj_link)
- if (vma->vm == vm && i915_vma_is_pinned(vma))
- i++;
- }
- error->pinned_bo_count[ndx] = i - error->active_bo_count[ndx];
-
- if (i) {
- active_bo = kcalloc(i, sizeof(*active_bo), GFP_ATOMIC);
- if (active_bo)
- pinned_bo = active_bo + error->active_bo_count[ndx];
- }
+ count++;
+ active_bo = NULL;
+ if (count)
+ active_bo = kcalloc(count, sizeof(*active_bo), GFP_ATOMIC);
if (active_bo)
- error->active_bo_count[ndx] =
- capture_active_bo(active_bo,
- error->active_bo_count[ndx],
- &vm->active_list);
-
- if (pinned_bo)
- error->pinned_bo_count[ndx] =
- capture_pinned_bo(pinned_bo,
- error->pinned_bo_count[ndx],
- &dev_priv->mm.bound_list, vm);
- error->active_bo[ndx] = active_bo;
- error->pinned_bo[ndx] = pinned_bo;
+ count = capture_error_bo(active_bo, count, &vm->active_list, false);
+ else
+ count = 0;
+
+ error->active_vm[idx] = vm;
+ error->active_bo[idx] = active_bo;
+ error->active_bo_count[idx] = count;
}
-static void i915_gem_capture_buffers(struct drm_i915_private *dev_priv,
- struct drm_i915_error_state *error)
+static void i915_capture_active_buffers(struct drm_i915_private *dev_priv,
+ struct drm_i915_error_state *error)
{
- struct i915_address_space *vm;
- int cnt = 0, i = 0;
-
- list_for_each_entry(vm, &dev_priv->vm_list, global_link)
- cnt++;
-
- error->active_bo = kcalloc(cnt, sizeof(*error->active_bo), GFP_ATOMIC);
- error->pinned_bo = kcalloc(cnt, sizeof(*error->pinned_bo), GFP_ATOMIC);
- error->active_bo_count = kcalloc(cnt, sizeof(*error->active_bo_count),
- GFP_ATOMIC);
- error->pinned_bo_count = kcalloc(cnt, sizeof(*error->pinned_bo_count),
- GFP_ATOMIC);
-
- if (error->active_bo == NULL ||
- error->pinned_bo == NULL ||
- error->active_bo_count == NULL ||
- error->pinned_bo_count == NULL) {
- kfree(error->active_bo);
- kfree(error->active_bo_count);
- kfree(error->pinned_bo);
- kfree(error->pinned_bo_count);
-
- error->active_bo = NULL;
- error->active_bo_count = NULL;
- error->pinned_bo = NULL;
- error->pinned_bo_count = NULL;
- } else {
- list_for_each_entry(vm, &dev_priv->vm_list, global_link)
- i915_gem_capture_vm(dev_priv, error, vm, i++);
+ int cnt = 0, i, j;
+
+ BUILD_BUG_ON(ARRAY_SIZE(error->engine) > ARRAY_SIZE(error->active_bo));
+ BUILD_BUG_ON(ARRAY_SIZE(error->active_bo) != ARRAY_SIZE(error->active_vm));
+ BUILD_BUG_ON(ARRAY_SIZE(error->active_bo) != ARRAY_SIZE(error->active_bo_count));
+
+ /* Scan each engine looking for unique active contexts/vm */
+ for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
+ struct drm_i915_error_engine *ee = &error->engine[i];
+ bool found;
- error->vm_count = cnt;
+ if (!ee->vm)
+ continue;
+
+ found = false;
+ for (j = 0; j < i && !found; j++)
+ found = error->engine[j].vm == ee->vm;
+ if (!found)
+ i915_gem_capture_vm(dev_priv, error, ee->vm, cnt++);
}
}
+static void i915_capture_pinned_buffers(struct drm_i915_private *dev_priv,
+ struct drm_i915_error_state *error)
+{
+ struct i915_address_space *vm = &dev_priv->ggtt.base;
+ struct drm_i915_error_buffer *bo;
+ struct i915_vma *vma;
+ int count_inactive, count_active;
+
+ count_inactive = 0;
+ list_for_each_entry(vma, &vm->active_list, vm_link)
+ count_inactive++;
+
+ count_active = 0;
+ list_for_each_entry(vma, &vm->inactive_list, vm_link)
+ count_active++;
+
+ bo = NULL;
+ if (count_inactive + count_active)
+ bo = kcalloc(count_inactive + count_active,
+ sizeof(*bo), GFP_ATOMIC);
+ if (!bo)
+ return;
+
+ count_inactive = capture_error_bo(bo, count_inactive,
+ &vm->active_list, true);
+ count_active = capture_error_bo(bo + count_inactive, count_active,
+ &vm->inactive_list, true);
+ error->pinned_bo_count = count_inactive + count_active;
+ error->pinned_bo = bo;
+}
+
/* Capture all registers which don't fit into another category. */
static void i915_capture_reg_state(struct drm_i915_private *dev_priv,
struct drm_i915_error_state *error)
#endif
error->reset_count = i915_reset_count(&dev_priv->gpu_error);
error->suspend_count = dev_priv->suspend_count;
+
+ memcpy(&error->device_info,
+ INTEL_INFO(dev_priv),
+ sizeof(error->device_info));
}
/**
i915_capture_gen_state(dev_priv, error);
i915_capture_reg_state(dev_priv, error);
- i915_gem_capture_buffers(dev_priv, error);
i915_gem_record_fences(dev_priv, error);
i915_gem_record_rings(dev_priv, error);
+ i915_capture_active_buffers(dev_priv, error);
+ i915_capture_pinned_buffers(dev_priv, error);
do_gettimeofday(&error->time);
#define HOST2GUC_INTERRUPT _MMIO(0xc4c8)
#define HOST2GUC_TRIGGER (1<<0)
-#define DRBMISC1 0x1984
-#define DOORBELL_ENABLE (1<<0)
-
#define GEN8_DRBREGL(x) _MMIO(0x1000 + (x) * 8)
#define GEN8_DRB_VALID (1<<0)
#define GEN8_DRBREGU(x) _MMIO(0x1000 + (x) * 8 + 4)
* WQ_TYPE_INORDER is needed to support legacy submission via GuC, which
* represents in-order queue. The kernel driver packs ring tail pointer and an
* ELSP context descriptor dword into Work Item.
- * See guc_add_workqueue_item()
+ * See guc_wq_item_append()
*
*/
if (ret != -ETIMEDOUT)
ret = -EIO;
- DRM_ERROR("GUC: host2guc action 0x%X failed. ret=%d "
- "status=0x%08X response=0x%08X\n",
- data[0], ret, status,
- I915_READ(SOFT_SCRATCH(15)));
+ DRM_WARN("Action 0x%X failed; ret=%d status=0x%08X response=0x%08X\n",
+ data[0], ret, status, I915_READ(SOFT_SCRATCH(15)));
dev_priv->guc.action_fail += 1;
dev_priv->guc.action_err = ret;
struct i915_guc_client *client,
u16 new_id)
{
- struct sg_table *sg = guc->ctx_pool_obj->pages;
+ struct sg_table *sg = guc->ctx_pool_vma->pages;
void *doorbell_bitmap = guc->doorbell_bitmap;
struct guc_doorbell_info *doorbell;
struct guc_context_desc desc;
/*
* Initialise the process descriptor shared with the GuC firmware.
*/
-static void guc_init_proc_desc(struct intel_guc *guc,
+static void guc_proc_desc_init(struct intel_guc *guc,
struct i915_guc_client *client)
{
struct guc_process_desc *desc;
* write queue, etc).
*/
-static void guc_init_ctx_desc(struct intel_guc *guc,
+static void guc_ctx_desc_init(struct intel_guc *guc,
struct i915_guc_client *client)
{
- struct drm_i915_gem_object *client_obj = client->client_obj;
struct drm_i915_private *dev_priv = guc_to_i915(guc);
struct intel_engine_cs *engine;
struct i915_gem_context *ctx = client->owner;
struct guc_context_desc desc;
struct sg_table *sg;
+ unsigned int tmp;
u32 gfx_addr;
memset(&desc, 0, sizeof(desc));
desc.priority = client->priority;
desc.db_id = client->doorbell_id;
- for_each_engine(engine, dev_priv) {
+ for_each_engine_masked(engine, dev_priv, client->engines, tmp) {
struct intel_context *ce = &ctx->engine[engine->id];
- struct guc_execlist_context *lrc = &desc.lrc[engine->guc_id];
- struct drm_i915_gem_object *obj;
+ uint32_t guc_engine_id = engine->guc_id;
+ struct guc_execlist_context *lrc = &desc.lrc[guc_engine_id];
/* TODO: We have a design issue to be solved here. Only when we
* receive the first batch, we know which engine is used by the
lrc->context_desc = lower_32_bits(ce->lrc_desc);
/* The state page is after PPHWSP */
- gfx_addr = i915_gem_obj_ggtt_offset(ce->state);
- lrc->ring_lcra = gfx_addr + LRC_STATE_PN * PAGE_SIZE;
+ lrc->ring_lcra =
+ i915_ggtt_offset(ce->state) + LRC_STATE_PN * PAGE_SIZE;
lrc->context_id = (client->ctx_index << GUC_ELC_CTXID_OFFSET) |
- (engine->guc_id << GUC_ELC_ENGINE_OFFSET);
-
- obj = ce->ring->obj;
- gfx_addr = i915_gem_obj_ggtt_offset(obj);
+ (guc_engine_id << GUC_ELC_ENGINE_OFFSET);
- lrc->ring_begin = gfx_addr;
- lrc->ring_end = gfx_addr + obj->base.size - 1;
- lrc->ring_next_free_location = gfx_addr;
+ lrc->ring_begin = i915_ggtt_offset(ce->ring->vma);
+ lrc->ring_end = lrc->ring_begin + ce->ring->size - 1;
+ lrc->ring_next_free_location = lrc->ring_begin;
lrc->ring_current_tail_pointer_value = 0;
- desc.engines_used |= (1 << engine->guc_id);
+ desc.engines_used |= (1 << guc_engine_id);
}
+ DRM_DEBUG_DRIVER("Host engines 0x%x => GuC engines used 0x%x\n",
+ client->engines, desc.engines_used);
WARN_ON(desc.engines_used == 0);
/*
* The doorbell, process descriptor, and workqueue are all parts
* of the client object, which the GuC will reference via the GGTT
*/
- gfx_addr = i915_gem_obj_ggtt_offset(client_obj);
- desc.db_trigger_phy = sg_dma_address(client_obj->pages->sgl) +
+ gfx_addr = i915_ggtt_offset(client->vma);
+ desc.db_trigger_phy = sg_dma_address(client->vma->pages->sgl) +
client->doorbell_offset;
desc.db_trigger_cpu = (uintptr_t)client->client_base +
client->doorbell_offset;
desc.desc_private = (uintptr_t)client;
/* Pool context is pinned already */
- sg = guc->ctx_pool_obj->pages;
+ sg = guc->ctx_pool_vma->pages;
sg_pcopy_from_buffer(sg->sgl, sg->nents, &desc, sizeof(desc),
sizeof(desc) * client->ctx_index);
}
-static void guc_fini_ctx_desc(struct intel_guc *guc,
+static void guc_ctx_desc_fini(struct intel_guc *guc,
struct i915_guc_client *client)
{
struct guc_context_desc desc;
memset(&desc, 0, sizeof(desc));
- sg = guc->ctx_pool_obj->pages;
+ sg = guc->ctx_pool_vma->pages;
sg_pcopy_from_buffer(sg->sgl, sg->nents, &desc, sizeof(desc),
sizeof(desc) * client->ctx_index);
}
/**
- * i915_guc_wq_check_space() - check that the GuC can accept a request
+ * i915_guc_wq_reserve() - reserve space in the GuC's workqueue
* @request: request associated with the commands
*
* Return: 0 if space is available
*
* This function must be called (and must return 0) before a request
* is submitted to the GuC via i915_guc_submit() below. Once a result
- * of 0 has been returned, it remains valid until (but only until)
- * the next call to submit().
+ * of 0 has been returned, it must be balanced by a corresponding
+ * call to submit().
*
- * This precheck allows the caller to determine in advance that space
+ * Reservation allows the caller to determine in advance that space
* will be available for the next submission before committing resources
* to it, and helps avoid late failures with complicated recovery paths.
*/
-int i915_guc_wq_check_space(struct drm_i915_gem_request *request)
+int i915_guc_wq_reserve(struct drm_i915_gem_request *request)
{
const size_t wqi_size = sizeof(struct guc_wq_item);
struct i915_guc_client *gc = request->i915->guc.execbuf_client;
- struct guc_process_desc *desc;
+ struct guc_process_desc *desc = gc->client_base + gc->proc_desc_offset;
u32 freespace;
+ int ret;
- GEM_BUG_ON(gc == NULL);
-
- desc = gc->client_base + gc->proc_desc_offset;
-
+ spin_lock(&gc->wq_lock);
freespace = CIRC_SPACE(gc->wq_tail, desc->head, gc->wq_size);
- if (likely(freespace >= wqi_size))
- return 0;
-
- gc->no_wq_space += 1;
+ freespace -= gc->wq_rsvd;
+ if (likely(freespace >= wqi_size)) {
+ gc->wq_rsvd += wqi_size;
+ ret = 0;
+ } else {
+ gc->no_wq_space++;
+ ret = -EAGAIN;
+ }
+ spin_unlock(&gc->wq_lock);
- return -EAGAIN;
+ return ret;
}
-static void guc_add_workqueue_item(struct i915_guc_client *gc,
- struct drm_i915_gem_request *rq)
+/* Construct a Work Item and append it to the GuC's Work Queue */
+static void guc_wq_item_append(struct i915_guc_client *gc,
+ struct drm_i915_gem_request *rq)
{
/* wqi_len is in DWords, and does not include the one-word header */
const size_t wqi_size = sizeof(struct guc_wq_item);
const u32 wqi_len = wqi_size/sizeof(u32) - 1;
+ struct intel_engine_cs *engine = rq->engine;
struct guc_process_desc *desc;
struct guc_wq_item *wqi;
void *base;
desc = gc->client_base + gc->proc_desc_offset;
- /* Free space is guaranteed, see i915_guc_wq_check_space() above */
+ /* Free space is guaranteed, see i915_guc_wq_reserve() above */
freespace = CIRC_SPACE(gc->wq_tail, desc->head, gc->wq_size);
GEM_BUG_ON(freespace < wqi_size);
* workqueue buffer dw by dw.
*/
BUILD_BUG_ON(wqi_size != 16);
+ GEM_BUG_ON(gc->wq_rsvd < wqi_size);
/* postincrement WQ tail for next time */
wq_off = gc->wq_tail;
+ GEM_BUG_ON(wq_off & (wqi_size - 1));
gc->wq_tail += wqi_size;
gc->wq_tail &= gc->wq_size - 1;
- GEM_BUG_ON(wq_off & (wqi_size - 1));
+ gc->wq_rsvd -= wqi_size;
/* WQ starts from the page after doorbell / process_desc */
wq_page = (wq_off + GUC_DB_SIZE) >> PAGE_SHIFT;
wq_off &= PAGE_SIZE - 1;
- base = kmap_atomic(i915_gem_object_get_page(gc->client_obj, wq_page));
+ base = kmap_atomic(i915_gem_object_get_page(gc->vma->obj, wq_page));
wqi = (struct guc_wq_item *)((char *)base + wq_off);
/* Now fill in the 4-word work queue item */
wqi->header = WQ_TYPE_INORDER |
(wqi_len << WQ_LEN_SHIFT) |
- (rq->engine->guc_id << WQ_TARGET_SHIFT) |
+ (engine->guc_id << WQ_TARGET_SHIFT) |
WQ_NO_WCFLUSH_WAIT;
/* The GuC wants only the low-order word of the context descriptor */
- wqi->context_desc = (u32)intel_lr_context_descriptor(rq->ctx,
- rq->engine);
+ wqi->context_desc = (u32)intel_lr_context_descriptor(rq->ctx, engine);
wqi->ring_tail = tail << WQ_RING_TAIL_SHIFT;
wqi->fence_id = rq->fence.seqno;
if (db_ret.db_status == GUC_DOORBELL_DISABLED)
break;
- DRM_ERROR("Cookie mismatch. Expected %d, returned %d\n",
- db_cmp.cookie, db_ret.cookie);
+ DRM_WARN("Cookie mismatch. Expected %d, found %d\n",
+ db_cmp.cookie, db_ret.cookie);
/* update the cookie to newly read cookie from GuC */
db_cmp.cookie = db_ret.cookie;
* Return: 0 on success, otherwise an errno.
* (Note: nonzero really shouldn't happen!)
*
- * The caller must have already called i915_guc_wq_check_space() above
- * with a result of 0 (success) since the last request submission. This
- * guarantees that there is space in the work queue for the new request,
- * so enqueuing the item cannot fail.
+ * The caller must have already called i915_guc_wq_reserve() above with
+ * a result of 0 (success), guaranteeing that there is space in the work
+ * queue for the new request, so enqueuing the item cannot fail.
*
* Bad Things Will Happen if the caller violates this protocol e.g. calls
- * submit() when check() says there's no space, or calls submit() multiple
- * times with no intervening check().
+ * submit() when _reserve() says there's no space, or calls _submit()
+ * a different number of times from (successful) calls to _reserve().
*
* The only error here arises if the doorbell hardware isn't functioning
* as expected, which really shouln't happen.
struct i915_guc_client *client = guc->execbuf_client;
int b_ret;
- guc_add_workqueue_item(client, rq);
+ spin_lock(&client->wq_lock);
+ guc_wq_item_append(client, rq);
b_ret = guc_ring_doorbell(client);
client->submissions[engine_id] += 1;
guc->submissions[engine_id] += 1;
guc->last_seqno[engine_id] = rq->fence.seqno;
+ spin_unlock(&client->wq_lock);
}
/*
*/
/**
- * gem_allocate_guc_obj() - Allocate gem object for GuC usage
- * @dev_priv: driver private data structure
- * @size: size of object
+ * guc_allocate_vma() - Allocate a GGTT VMA for GuC usage
+ * @guc: the guc
+ * @size: size of area to allocate (both virtual space and memory)
*
- * This is a wrapper to create a gem obj. In order to use it inside GuC, the
- * object needs to be pinned lifetime. Also we must pin it to gtt space other
- * than [0, GUC_WOPCM_TOP) because this range is reserved inside GuC.
+ * This is a wrapper to create an object for use with the GuC. In order to
+ * use it inside the GuC, an object needs to be pinned lifetime, so we allocate
+ * both some backing storage and a range inside the Global GTT. We must pin
+ * it in the GGTT somewhere other than than [0, GUC_WOPCM_TOP) because that
+ * range is reserved inside GuC.
*
- * Return: A drm_i915_gem_object if successful, otherwise NULL.
+ * Return: A i915_vma if successful, otherwise an ERR_PTR.
*/
-static struct drm_i915_gem_object *
-gem_allocate_guc_obj(struct drm_i915_private *dev_priv, u32 size)
+static struct i915_vma *guc_allocate_vma(struct intel_guc *guc, u32 size)
{
+ struct drm_i915_private *dev_priv = guc_to_i915(guc);
struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
+ int ret;
obj = i915_gem_object_create(&dev_priv->drm, size);
if (IS_ERR(obj))
- return NULL;
+ return ERR_CAST(obj);
- if (i915_gem_object_get_pages(obj)) {
- i915_gem_object_put(obj);
- return NULL;
- }
+ vma = i915_vma_create(obj, &dev_priv->ggtt.base, NULL);
+ if (IS_ERR(vma))
+ goto err;
- if (i915_gem_object_ggtt_pin(obj, NULL, 0, PAGE_SIZE,
- PIN_OFFSET_BIAS | GUC_WOPCM_TOP)) {
- i915_gem_object_put(obj);
- return NULL;
+ ret = i915_vma_pin(vma, 0, PAGE_SIZE,
+ PIN_GLOBAL | PIN_OFFSET_BIAS | GUC_WOPCM_TOP);
+ if (ret) {
+ vma = ERR_PTR(ret);
+ goto err;
}
/* Invalidate GuC TLB to let GuC take the latest updates to GTT. */
I915_WRITE(GEN8_GTCR, GEN8_GTCR_INVALIDATE);
- return obj;
-}
-
-/**
- * gem_release_guc_obj() - Release gem object allocated for GuC usage
- * @obj: gem obj to be released
- */
-static void gem_release_guc_obj(struct drm_i915_gem_object *obj)
-{
- if (!obj)
- return;
-
- if (i915_gem_obj_is_pinned(obj))
- i915_gem_object_ggtt_unpin(obj);
+ return vma;
+err:
i915_gem_object_put(obj);
+ return vma;
}
static void
kunmap(kmap_to_page(client->client_base));
}
- gem_release_guc_obj(client->client_obj);
+ i915_vma_unpin_and_release(&client->vma);
if (client->ctx_index != GUC_INVALID_CTX_ID) {
- guc_fini_ctx_desc(guc, client);
+ guc_ctx_desc_fini(guc, client);
ida_simple_remove(&guc->ctx_ids, client->ctx_index);
}
kfree(client);
}
+/* Check that a doorbell register is in the expected state */
+static bool guc_doorbell_check(struct intel_guc *guc, uint16_t db_id)
+{
+ struct drm_i915_private *dev_priv = guc_to_i915(guc);
+ i915_reg_t drbreg = GEN8_DRBREGL(db_id);
+ uint32_t value = I915_READ(drbreg);
+ bool enabled = (value & GUC_DOORBELL_ENABLED) != 0;
+ bool expected = test_bit(db_id, guc->doorbell_bitmap);
+
+ if (enabled == expected)
+ return true;
+
+ DRM_DEBUG_DRIVER("Doorbell %d (reg 0x%x) 0x%x, should be %s\n",
+ db_id, drbreg.reg, value,
+ expected ? "active" : "inactive");
+
+ return false;
+}
+
/*
- * Borrow the first client to set up & tear down every doorbell
+ * Borrow the first client to set up & tear down each unused doorbell
* in turn, to ensure that all doorbell h/w is (re)initialised.
*/
static void guc_init_doorbell_hw(struct intel_guc *guc)
{
- struct drm_i915_private *dev_priv = guc_to_i915(guc);
struct i915_guc_client *client = guc->execbuf_client;
- uint16_t db_id, i;
- int err;
+ uint16_t db_id;
+ int i, err;
+ /* Save client's original doorbell selection */
db_id = client->doorbell_id;
for (i = 0; i < GUC_MAX_DOORBELLS; ++i) {
- i915_reg_t drbreg = GEN8_DRBREGL(i);
- u32 value = I915_READ(drbreg);
+ /* Skip if doorbell is OK */
+ if (guc_doorbell_check(guc, i))
+ continue;
err = guc_update_doorbell_id(guc, client, i);
-
- /* Report update failure or unexpectedly active doorbell */
- if (err || (i != db_id && (value & GUC_DOORBELL_ENABLED)))
- DRM_DEBUG_DRIVER("Doorbell %d (reg 0x%x) was 0x%x, err %d\n",
- i, drbreg.reg, value, err);
+ if (err)
+ DRM_DEBUG_DRIVER("Doorbell %d update failed, err %d\n",
+ i, err);
}
/* Restore to original value */
err = guc_update_doorbell_id(guc, client, db_id);
if (err)
- DRM_ERROR("Failed to restore doorbell to %d, err %d\n",
- db_id, err);
-
- for (i = 0; i < GUC_MAX_DOORBELLS; ++i) {
- i915_reg_t drbreg = GEN8_DRBREGL(i);
- u32 value = I915_READ(drbreg);
-
- if (i != db_id && (value & GUC_DOORBELL_ENABLED))
- DRM_DEBUG_DRIVER("Doorbell %d (reg 0x%x) finally 0x%x\n",
- i, drbreg.reg, value);
+ DRM_WARN("Failed to restore doorbell to %d, err %d\n",
+ db_id, err);
- }
+ /* Read back & verify all doorbell registers */
+ for (i = 0; i < GUC_MAX_DOORBELLS; ++i)
+ (void)guc_doorbell_check(guc, i);
}
/**
* guc_client_alloc() - Allocate an i915_guc_client
* @dev_priv: driver private data structure
+ * @engines: The set of engines to enable for this client
* @priority: four levels priority _CRITICAL, _HIGH, _NORMAL and _LOW
* The kernel client to replace ExecList submission is created with
* NORMAL priority. Priority of a client for scheduler can be HIGH,
*/
static struct i915_guc_client *
guc_client_alloc(struct drm_i915_private *dev_priv,
+ uint32_t engines,
uint32_t priority,
struct i915_gem_context *ctx)
{
struct i915_guc_client *client;
struct intel_guc *guc = &dev_priv->guc;
- struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
uint16_t db_id;
client = kzalloc(sizeof(*client), GFP_KERNEL);
if (!client)
return NULL;
- client->doorbell_id = GUC_INVALID_DOORBELL_ID;
- client->priority = priority;
client->owner = ctx;
client->guc = guc;
+ client->engines = engines;
+ client->priority = priority;
+ client->doorbell_id = GUC_INVALID_DOORBELL_ID;
client->ctx_index = (uint32_t)ida_simple_get(&guc->ctx_ids, 0,
GUC_MAX_GPU_CONTEXTS, GFP_KERNEL);
}
/* The first page is doorbell/proc_desc. Two followed pages are wq. */
- obj = gem_allocate_guc_obj(dev_priv, GUC_DB_SIZE + GUC_WQ_SIZE);
- if (!obj)
+ vma = guc_allocate_vma(guc, GUC_DB_SIZE + GUC_WQ_SIZE);
+ if (IS_ERR(vma))
goto err;
/* We'll keep just the first (doorbell/proc) page permanently kmap'd. */
- client->client_obj = obj;
- client->client_base = kmap(i915_gem_object_get_page(obj, 0));
+ client->vma = vma;
+ client->client_base = kmap(i915_vma_first_page(vma));
+
+ spin_lock_init(&client->wq_lock);
client->wq_offset = GUC_DB_SIZE;
client->wq_size = GUC_WQ_SIZE;
else
client->proc_desc_offset = (GUC_DB_SIZE / 2);
- guc_init_proc_desc(guc, client);
- guc_init_ctx_desc(guc, client);
+ guc_proc_desc_init(guc, client);
+ guc_ctx_desc_init(guc, client);
if (guc_init_doorbell(guc, client, db_id))
goto err;
- DRM_DEBUG_DRIVER("new priority %u client %p: ctx_index %u\n",
- priority, client, client->ctx_index);
+ DRM_DEBUG_DRIVER("new priority %u client %p for engine(s) 0x%x: ctx_index %u\n",
+ priority, client, client->engines, client->ctx_index);
DRM_DEBUG_DRIVER("doorbell id %u, cacheline offset 0x%x\n",
client->doorbell_id, client->doorbell_offset);
return client;
err:
- DRM_ERROR("FAILED to create priority %u GuC client!\n", priority);
-
guc_client_free(dev_priv, client);
return NULL;
}
-static void guc_create_log(struct intel_guc *guc)
+static void guc_log_create(struct intel_guc *guc)
{
- struct drm_i915_private *dev_priv = guc_to_i915(guc);
- struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
unsigned long offset;
uint32_t size, flags;
GUC_LOG_ISR_PAGES + 1 +
GUC_LOG_CRASH_PAGES + 1) << PAGE_SHIFT;
- obj = guc->log_obj;
- if (!obj) {
- obj = gem_allocate_guc_obj(dev_priv, size);
- if (!obj) {
+ vma = guc->log_vma;
+ if (!vma) {
+ vma = guc_allocate_vma(guc, size);
+ if (IS_ERR(vma)) {
/* logging will be off */
i915.guc_log_level = -1;
return;
}
- guc->log_obj = obj;
+ guc->log_vma = vma;
}
/* each allocated unit is a page */
(GUC_LOG_ISR_PAGES << GUC_LOG_ISR_SHIFT) |
(GUC_LOG_CRASH_PAGES << GUC_LOG_CRASH_SHIFT);
- offset = i915_gem_obj_ggtt_offset(obj) >> PAGE_SHIFT; /* in pages */
+ offset = i915_ggtt_offset(vma) >> PAGE_SHIFT; /* in pages */
guc->log_flags = (offset << GUC_LOG_BUF_ADDR_SHIFT) | flags;
}
-static void init_guc_policies(struct guc_policies *policies)
+static void guc_policies_init(struct guc_policies *policies)
{
struct guc_policy *policy;
u32 p, i;
policies->is_valid = 1;
}
-static void guc_create_ads(struct intel_guc *guc)
+static void guc_addon_create(struct intel_guc *guc)
{
struct drm_i915_private *dev_priv = guc_to_i915(guc);
- struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
struct guc_ads *ads;
struct guc_policies *policies;
struct guc_mmio_reg_state *reg_state;
sizeof(struct guc_mmio_reg_state) +
GUC_S3_SAVE_SPACE_PAGES * PAGE_SIZE;
- obj = guc->ads_obj;
- if (!obj) {
- obj = gem_allocate_guc_obj(dev_priv, PAGE_ALIGN(size));
- if (!obj)
+ vma = guc->ads_vma;
+ if (!vma) {
+ vma = guc_allocate_vma(guc, PAGE_ALIGN(size));
+ if (IS_ERR(vma))
return;
- guc->ads_obj = obj;
+ guc->ads_vma = vma;
}
- page = i915_gem_object_get_page(obj, 0);
+ page = i915_vma_first_page(vma);
ads = kmap(page);
/*
* to find it.
*/
engine = &dev_priv->engine[RCS];
- ads->golden_context_lrca = engine->status_page.gfx_addr;
+ ads->golden_context_lrca = engine->status_page.ggtt_offset;
for_each_engine(engine, dev_priv)
ads->eng_state_size[engine->guc_id] = intel_lr_context_size(engine);
/* GuC scheduling policies */
policies = (void *)ads + sizeof(struct guc_ads);
- init_guc_policies(policies);
+ guc_policies_init(policies);
- ads->scheduler_policies = i915_gem_obj_ggtt_offset(obj) +
- sizeof(struct guc_ads);
+ ads->scheduler_policies =
+ i915_ggtt_offset(vma) + sizeof(struct guc_ads);
/* MMIO reg state */
reg_state = (void *)policies + sizeof(struct guc_policies);
const size_t poolsize = GUC_MAX_GPU_CONTEXTS * ctxsize;
const size_t gemsize = round_up(poolsize, PAGE_SIZE);
struct intel_guc *guc = &dev_priv->guc;
+ struct i915_vma *vma;
/* Wipe bitmap & delete client in case of reinitialisation */
bitmap_clear(guc->doorbell_bitmap, 0, GUC_MAX_DOORBELLS);
if (!i915.enable_guc_submission)
return 0; /* not enabled */
- if (guc->ctx_pool_obj)
+ if (guc->ctx_pool_vma)
return 0; /* already allocated */
- guc->ctx_pool_obj = gem_allocate_guc_obj(dev_priv, gemsize);
- if (!guc->ctx_pool_obj)
- return -ENOMEM;
+ vma = guc_allocate_vma(guc, gemsize);
+ if (IS_ERR(vma))
+ return PTR_ERR(vma);
+ guc->ctx_pool_vma = vma;
ida_init(&guc->ctx_ids);
- guc_create_log(guc);
- guc_create_ads(guc);
+ guc_log_create(guc);
+ guc_addon_create(guc);
return 0;
}
struct intel_guc *guc = &dev_priv->guc;
struct i915_guc_client *client;
struct intel_engine_cs *engine;
+ struct drm_i915_gem_request *request;
/* client for execbuf submission */
client = guc_client_alloc(dev_priv,
+ INTEL_INFO(dev_priv)->ring_mask,
GUC_CTX_PRIORITY_KMD_NORMAL,
dev_priv->kernel_context);
if (!client) {
- DRM_ERROR("Failed to create execbuf guc_client\n");
+ DRM_ERROR("Failed to create normal GuC client!\n");
return -ENOMEM;
}
guc_init_doorbell_hw(guc);
/* Take over from manual control of ELSP (execlists) */
- for_each_engine(engine, dev_priv)
+ for_each_engine(engine, dev_priv) {
engine->submit_request = i915_guc_submit;
+ /* Replay the current set of previously submitted requests */
+ list_for_each_entry(request, &engine->request_list, link) {
+ client->wq_rsvd += sizeof(struct guc_wq_item);
+ if (i915_sw_fence_done(&request->submit))
+ i915_guc_submit(request);
+ }
+ }
+
return 0;
}
{
struct intel_guc *guc = &dev_priv->guc;
- gem_release_guc_obj(dev_priv->guc.ads_obj);
- guc->ads_obj = NULL;
-
- gem_release_guc_obj(dev_priv->guc.log_obj);
- guc->log_obj = NULL;
+ i915_vma_unpin_and_release(&guc->ads_vma);
+ i915_vma_unpin_and_release(&guc->log_vma);
- if (guc->ctx_pool_obj)
+ if (guc->ctx_pool_vma)
ida_destroy(&guc->ctx_ids);
- gem_release_guc_obj(guc->ctx_pool_obj);
- guc->ctx_pool_obj = NULL;
+ i915_vma_unpin_and_release(&guc->ctx_pool_vma);
}
/**
/* any value greater than GUC_POWER_D0 */
data[1] = GUC_POWER_D1;
/* first page is shared data with GuC */
- data[2] = i915_gem_obj_ggtt_offset(ctx->engine[RCS].state);
+ data[2] = i915_ggtt_offset(ctx->engine[RCS].state);
return host2guc_action(guc, data, ARRAY_SIZE(data));
}
data[0] = HOST2GUC_ACTION_EXIT_S_STATE;
data[1] = GUC_POWER_D0;
/* first page is shared data with GuC */
- data[2] = i915_gem_obj_ggtt_offset(ctx->engine[RCS].state);
+ data[2] = i915_ggtt_offset(ctx->engine[RCS].state);
return host2guc_action(guc, data, ARRAY_SIZE(data));
}
spin_lock_irq(&dev_priv->irq_lock);
dev_priv->rps.interrupts_enabled = false;
- I915_WRITE(GEN6_PMINTRMSK, gen6_sanitize_rps_pm_mask(dev_priv, ~0));
+ I915_WRITE(GEN6_PMINTRMSK, gen6_sanitize_rps_pm_mask(dev_priv, ~0u));
__gen6_disable_pm_irq(dev_priv, dev_priv->pm_rps_events);
I915_WRITE(gen6_pm_ier(dev_priv), I915_READ(gen6_pm_ier(dev_priv)) &
static void notify_ring(struct intel_engine_cs *engine)
{
smp_store_mb(engine->breadcrumbs.irq_posted, true);
- if (intel_engine_wakeup(engine)) {
+ if (intel_engine_wakeup(engine))
trace_i915_gem_request_notify(engine);
- engine->breadcrumbs.irq_wakeups++;
- }
}
static void vlv_c0_read(struct drm_i915_private *dev_priv,
char *error_event[] = { I915_ERROR_UEVENT "=1", NULL };
char *reset_event[] = { I915_RESET_UEVENT "=1", NULL };
char *reset_done_event[] = { I915_ERROR_UEVENT "=0", NULL };
- int ret;
kobject_uevent_env(kobj, KOBJ_CHANGE, error_event);
+ DRM_DEBUG_DRIVER("resetting chip\n");
+ kobject_uevent_env(kobj, KOBJ_CHANGE, reset_event);
+
/*
- * Note that there's only one work item which does gpu resets, so we
- * need not worry about concurrent gpu resets potentially incrementing
- * error->reset_counter twice. We only need to take care of another
- * racing irq/hangcheck declaring the gpu dead for a second time. A
- * quick check for that is good enough: schedule_work ensures the
- * correct ordering between hang detection and this work item, and since
- * the reset in-progress bit is only ever set by code outside of this
- * work we don't need to worry about any other races.
+ * In most cases it's guaranteed that we get here with an RPM
+ * reference held, for example because there is a pending GPU
+ * request that won't finish until the reset is done. This
+ * isn't the case at least when we get here by doing a
+ * simulated reset via debugs, so get an RPM reference.
*/
- if (i915_reset_in_progress(&dev_priv->gpu_error)) {
- DRM_DEBUG_DRIVER("resetting chip\n");
- kobject_uevent_env(kobj, KOBJ_CHANGE, reset_event);
-
- /*
- * In most cases it's guaranteed that we get here with an RPM
- * reference held, for example because there is a pending GPU
- * request that won't finish until the reset is done. This
- * isn't the case at least when we get here by doing a
- * simulated reset via debugs, so get an RPM reference.
- */
- intel_runtime_pm_get(dev_priv);
-
- intel_prepare_reset(dev_priv);
+ intel_runtime_pm_get(dev_priv);
+ intel_prepare_reset(dev_priv);
+ do {
/*
* All state reset _must_ be completed before we update the
* reset counter, for otherwise waiters might miss the reset
* pending state and not properly drop locks, resulting in
* deadlocks with the reset work.
*/
- ret = i915_reset(dev_priv);
+ if (mutex_trylock(&dev_priv->drm.struct_mutex)) {
+ i915_reset(dev_priv);
+ mutex_unlock(&dev_priv->drm.struct_mutex);
+ }
- intel_finish_reset(dev_priv);
+ /* We need to wait for anyone holding the lock to wakeup */
+ } while (wait_on_bit_timeout(&dev_priv->gpu_error.flags,
+ I915_RESET_IN_PROGRESS,
+ TASK_UNINTERRUPTIBLE,
+ HZ));
- intel_runtime_pm_put(dev_priv);
+ intel_finish_reset(dev_priv);
+ intel_runtime_pm_put(dev_priv);
- if (ret == 0)
- kobject_uevent_env(kobj,
- KOBJ_CHANGE, reset_done_event);
+ if (!test_bit(I915_WEDGED, &dev_priv->gpu_error.flags))
+ kobject_uevent_env(kobj,
+ KOBJ_CHANGE, reset_done_event);
- /*
- * Note: The wake_up also serves as a memory barrier so that
- * waiters see the update value of the reset counter atomic_t.
- */
- wake_up_all(&dev_priv->gpu_error.reset_queue);
- }
+ /*
+ * Note: The wake_up also serves as a memory barrier so that
+ * waiters see the updated value of the dev_priv->gpu_error.
+ */
+ wake_up_all(&dev_priv->gpu_error.reset_queue);
}
static void i915_report_and_clear_eir(struct drm_i915_private *dev_priv)
i915_capture_error_state(dev_priv, engine_mask, error_msg);
i915_report_and_clear_eir(dev_priv);
- if (engine_mask) {
- atomic_or(I915_RESET_IN_PROGRESS_FLAG,
- &dev_priv->gpu_error.reset_counter);
+ if (!engine_mask)
+ return;
- /*
- * Wakeup waiting processes so that the reset function
- * i915_reset_and_wakeup doesn't deadlock trying to grab
- * various locks. By bumping the reset counter first, the woken
- * processes will see a reset in progress and back off,
- * releasing their locks and then wait for the reset completion.
- * We must do this for _all_ gpu waiters that might hold locks
- * that the reset work needs to acquire.
- *
- * Note: The wake_up serves as the required memory barrier to
- * ensure that the waiters see the updated value of the reset
- * counter atomic_t.
- */
- i915_error_wake_up(dev_priv);
- }
+ if (test_and_set_bit(I915_RESET_IN_PROGRESS,
+ &dev_priv->gpu_error.flags))
+ return;
+
+ /*
+ * Wakeup waiting processes so that the reset function
+ * i915_reset_and_wakeup doesn't deadlock trying to grab
+ * various locks. By bumping the reset counter first, the woken
+ * processes will see a reset in progress and back off,
+ * releasing their locks and then wait for the reset completion.
+ * We must do this for _all_ gpu waiters that might hold locks
+ * that the reset work needs to acquire.
+ *
+ * Note: The wake_up also provides a memory barrier to ensure that the
+ * waiters see the updated value of the reset flags.
+ */
+ i915_error_wake_up(dev_priv);
i915_reset_and_wakeup(dev_priv);
}
}
}
- DRM_ERROR("No signaller ring found for ring %i, ipehr 0x%08x, offset 0x%016llx\n",
- engine->id, ipehr, offset);
+ DRM_DEBUG_DRIVER("No signaller ring found for ring %i, ipehr 0x%08x, offset 0x%016llx\n",
+ engine->id, ipehr, offset);
- return NULL;
+ return ERR_PTR(-ENODEV);
}
static struct intel_engine_cs *
if (signaller == NULL)
return -1;
+ if (IS_ERR(signaller))
+ return 0;
+
/* Prevent pathological recursion due to driver bugs */
if (signaller->hangcheck.deadlock >= I915_NUM_ENGINES)
return -1;
return HANGCHECK_HUNG;
}
-static unsigned long kick_waiters(struct intel_engine_cs *engine)
-{
- struct drm_i915_private *i915 = engine->i915;
- unsigned long irq_count = READ_ONCE(engine->breadcrumbs.irq_wakeups);
-
- if (engine->hangcheck.user_interrupts == irq_count &&
- !test_and_set_bit(engine->id, &i915->gpu_error.missed_irq_rings)) {
- if (!test_bit(engine->id, &i915->gpu_error.test_irq_rings))
- DRM_ERROR("Hangcheck timer elapsed... %s idle\n",
- engine->name);
-
- intel_engine_enable_fake_irq(engine);
- }
-
- return irq_count;
-}
/*
* This is called when the chip hasn't reported back with completed
* batchbuffers in a long time. We keep track per ring seqno progress and
bool busy = intel_engine_has_waiter(engine);
u64 acthd;
u32 seqno;
- unsigned user_interrupts;
+ u32 submit;
semaphore_clear_deadlocks(dev_priv);
acthd = intel_engine_get_active_head(engine);
seqno = intel_engine_get_seqno(engine);
-
- /* Reset stuck interrupts between batch advances */
- user_interrupts = 0;
+ submit = READ_ONCE(engine->last_submitted_seqno);
if (engine->hangcheck.seqno == seqno) {
- if (!intel_engine_is_active(engine)) {
+ if (i915_seqno_passed(seqno, submit)) {
engine->hangcheck.action = HANGCHECK_IDLE;
- if (busy) {
- /* Safeguard against driver failure */
- user_interrupts = kick_waiters(engine);
- engine->hangcheck.score += BUSY;
- }
} else {
/* We always increment the hangcheck score
* if the engine is busy and still processing
engine->hangcheck.seqno = seqno;
engine->hangcheck.acthd = acthd;
- engine->hangcheck.user_interrupts = user_interrupts;
busy_count += busy;
}
if (hung) {
char msg[80];
+ unsigned int tmp;
int len;
/* If some rings hung but others were still busy, only
hung &= ~stuck;
len = scnprintf(msg, sizeof(msg),
"%s on ", stuck == hung ? "No progress" : "Hang");
- for_each_engine_masked(engine, dev_priv, hung)
+ for_each_engine_masked(engine, dev_priv, hung, tmp)
len += scnprintf(msg + len, sizeof(msg) - len,
"%s, ", engine->name);
msg[len-2] = '\0';
dev_priv->rps.pm_intr_keep |= GEN6_PM_RP_UP_EI_EXPIRED;
if (INTEL_INFO(dev_priv)->gen >= 8)
- dev_priv->rps.pm_intr_keep |= GEN8_PMINTR_REDIRECT_TO_NON_DISP;
+ dev_priv->rps.pm_intr_keep |= GEN8_PMINTR_REDIRECT_TO_GUC;
INIT_DELAYED_WORK(&dev_priv->gpu_error.hangcheck_work,
i915_hangcheck_elapsed);
--- /dev/null
+/*
+ * Copyright © 2016 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <asm/fpu/api.h>
+
+#include "i915_drv.h"
+
+static DEFINE_STATIC_KEY_FALSE(has_movntdqa);
+
+#ifdef CONFIG_AS_MOVNTDQA
+static void __memcpy_ntdqa(void *dst, const void *src, unsigned long len)
+{
+ kernel_fpu_begin();
+
+ len >>= 4;
+ while (len >= 4) {
+ asm("movntdqa (%0), %%xmm0\n"
+ "movntdqa 16(%0), %%xmm1\n"
+ "movntdqa 32(%0), %%xmm2\n"
+ "movntdqa 48(%0), %%xmm3\n"
+ "movaps %%xmm0, (%1)\n"
+ "movaps %%xmm1, 16(%1)\n"
+ "movaps %%xmm2, 32(%1)\n"
+ "movaps %%xmm3, 48(%1)\n"
+ :: "r" (src), "r" (dst) : "memory");
+ src += 64;
+ dst += 64;
+ len -= 4;
+ }
+ while (len--) {
+ asm("movntdqa (%0), %%xmm0\n"
+ "movaps %%xmm0, (%1)\n"
+ :: "r" (src), "r" (dst) : "memory");
+ src += 16;
+ dst += 16;
+ }
+
+ kernel_fpu_end();
+}
+#endif
+
+/**
+ * i915_memcpy_from_wc: perform an accelerated *aligned* read from WC
+ * @dst: destination pointer
+ * @src: source pointer
+ * @len: how many bytes to copy
+ *
+ * i915_memcpy_from_wc copies @len bytes from @src to @dst using
+ * non-temporal instructions where available. Note that all arguments
+ * (@src, @dst) must be aligned to 16 bytes and @len must be a multiple
+ * of 16.
+ *
+ * To test whether accelerated reads from WC are supported, use
+ * i915_memcpy_from_wc(NULL, NULL, 0);
+ *
+ * Returns true if the copy was successful, false if the preconditions
+ * are not met.
+ */
+bool i915_memcpy_from_wc(void *dst, const void *src, unsigned long len)
+{
+ if (unlikely(((unsigned long)dst | (unsigned long)src | len) & 15))
+ return false;
+
+#ifdef CONFIG_AS_MOVNTDQA
+ if (static_branch_likely(&has_movntdqa)) {
+ if (likely(len))
+ __memcpy_ntdqa(dst, src, len);
+ return true;
+ }
+#endif
+
+ return false;
+}
+
+void i915_memcpy_init_early(struct drm_i915_private *dev_priv)
+{
+ if (static_cpu_has(X86_FEATURE_XMM4_1))
+ static_branch_enable(&has_movntdqa);
+}
--- /dev/null
+/*
+ * Copyright © 2014 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/mm.h>
+#include <linux/io-mapping.h>
+
+#include <asm/pgtable.h>
+
+#include "i915_drv.h"
+
+struct remap_pfn {
+ struct mm_struct *mm;
+ unsigned long pfn;
+ pgprot_t prot;
+};
+
+static int remap_pfn(pte_t *pte, pgtable_t token,
+ unsigned long addr, void *data)
+{
+ struct remap_pfn *r = data;
+
+ /* Special PTE are not associated with any struct page */
+ set_pte_at(r->mm, addr, pte, pte_mkspecial(pfn_pte(r->pfn, r->prot)));
+ r->pfn++;
+
+ return 0;
+}
+
+/**
+ * remap_io_mapping - remap an IO mapping to userspace
+ * @vma: user vma to map to
+ * @addr: target user address to start at
+ * @pfn: physical address of kernel memory
+ * @size: size of map area
+ * @iomap: the source io_mapping
+ *
+ * Note: this is only safe if the mm semaphore is held when called.
+ */
+int remap_io_mapping(struct vm_area_struct *vma,
+ unsigned long addr, unsigned long pfn, unsigned long size,
+ struct io_mapping *iomap)
+{
+ struct remap_pfn r;
+ int err;
+
+ GEM_BUG_ON((vma->vm_flags &
+ (VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP)) !=
+ (VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP));
+
+ /* We rely on prevalidation of the io-mapping to skip track_pfn(). */
+ r.mm = vma->vm_mm;
+ r.pfn = pfn;
+ r.prot = __pgprot((pgprot_val(iomap->prot) & _PAGE_CACHE_MASK) |
+ (pgprot_val(vma->vm_page_prot) & ~_PAGE_CACHE_MASK));
+
+ err = apply_to_page_range(r.mm, addr, size, remap_pfn, &r);
+ if (unlikely(err)) {
+ zap_vma_ptes(vma, addr, (r.pfn - pfn) << PAGE_SHIFT);
+ return err;
+ }
+
+ return 0;
+}
.fastboot = 0,
.prefault_disable = 0,
.load_detect_test = 0,
+ .force_reset_modeset_test = 0,
.reset = true,
.invert_brightness = 0,
.disable_display = 0,
"Force-enable the VGA load detect code for testing (default:false). "
"For developers only.");
+module_param_named_unsafe(force_reset_modeset_test, i915.force_reset_modeset_test, bool, 0600);
+MODULE_PARM_DESC(force_reset_modeset_test,
+ "Force a modeset during gpu reset for testing (default:false). "
+ "For developers only.");
+
module_param_named_unsafe(invert_brightness, i915.invert_brightness, int, 0600);
MODULE_PARM_DESC(invert_brightness,
"Invert backlight brightness "
bool fastboot;
bool prefault_disable;
bool load_detect_test;
+ bool force_reset_modeset_test;
bool reset;
bool disable_display;
bool verbose_state_checks;
#define CHV_COLORS \
.color = { .degamma_lut_size = 65, .gamma_lut_size = 257 }
+#define GEN2_FEATURES \
+ .gen = 2, .num_pipes = 1, \
+ .has_overlay = 1, .overlay_needs_physical = 1, \
+ .has_gmch_display = 1, \
+ .hws_needs_physical = 1, \
+ .ring_mask = RENDER_RING, \
+ GEN_DEFAULT_PIPEOFFSETS, \
+ CURSOR_OFFSETS
+
static const struct intel_device_info intel_i830_info = {
- .gen = 2, .is_mobile = 1, .cursor_needs_physical = 1, .num_pipes = 2,
- .has_overlay = 1, .overlay_needs_physical = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ GEN2_FEATURES,
+ .is_mobile = 1, .cursor_needs_physical = 1,
+ .num_pipes = 2, /* legal, last one wins */
};
static const struct intel_device_info intel_845g_info = {
- .gen = 2, .num_pipes = 1,
- .has_overlay = 1, .overlay_needs_physical = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ GEN2_FEATURES,
};
static const struct intel_device_info intel_i85x_info = {
- .gen = 2, .is_i85x = 1, .is_mobile = 1, .num_pipes = 2,
+ GEN2_FEATURES,
+ .is_i85x = 1, .is_mobile = 1,
+ .num_pipes = 2, /* legal, last one wins */
.cursor_needs_physical = 1,
- .has_overlay = 1, .overlay_needs_physical = 1,
.has_fbc = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
};
static const struct intel_device_info intel_i865g_info = {
- .gen = 2, .num_pipes = 1,
- .has_overlay = 1, .overlay_needs_physical = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ GEN2_FEATURES,
};
+#define GEN3_FEATURES \
+ .gen = 3, .num_pipes = 2, \
+ .has_gmch_display = 1, \
+ .ring_mask = RENDER_RING, \
+ GEN_DEFAULT_PIPEOFFSETS, \
+ CURSOR_OFFSETS
+
static const struct intel_device_info intel_i915g_info = {
- .gen = 3, .is_i915g = 1, .cursor_needs_physical = 1, .num_pipes = 2,
+ GEN3_FEATURES,
+ .is_i915g = 1, .cursor_needs_physical = 1,
.has_overlay = 1, .overlay_needs_physical = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ .hws_needs_physical = 1,
};
static const struct intel_device_info intel_i915gm_info = {
- .gen = 3, .is_mobile = 1, .num_pipes = 2,
+ GEN3_FEATURES,
+ .is_mobile = 1,
.cursor_needs_physical = 1,
.has_overlay = 1, .overlay_needs_physical = 1,
.supports_tv = 1,
.has_fbc = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ .hws_needs_physical = 1,
};
static const struct intel_device_info intel_i945g_info = {
- .gen = 3, .has_hotplug = 1, .cursor_needs_physical = 1, .num_pipes = 2,
+ GEN3_FEATURES,
+ .has_hotplug = 1, .cursor_needs_physical = 1,
.has_overlay = 1, .overlay_needs_physical = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ .hws_needs_physical = 1,
};
static const struct intel_device_info intel_i945gm_info = {
- .gen = 3, .is_i945gm = 1, .is_mobile = 1, .num_pipes = 2,
+ GEN3_FEATURES,
+ .is_i945gm = 1, .is_mobile = 1,
.has_hotplug = 1, .cursor_needs_physical = 1,
.has_overlay = 1, .overlay_needs_physical = 1,
.supports_tv = 1,
.has_fbc = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ .hws_needs_physical = 1,
};
+#define GEN4_FEATURES \
+ .gen = 4, .num_pipes = 2, \
+ .has_hotplug = 1, \
+ .has_gmch_display = 1, \
+ .ring_mask = RENDER_RING, \
+ GEN_DEFAULT_PIPEOFFSETS, \
+ CURSOR_OFFSETS
+
static const struct intel_device_info intel_i965g_info = {
- .gen = 4, .is_broadwater = 1, .num_pipes = 2,
- .has_hotplug = 1,
+ GEN4_FEATURES,
+ .is_broadwater = 1,
.has_overlay = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ .hws_needs_physical = 1,
};
static const struct intel_device_info intel_i965gm_info = {
- .gen = 4, .is_crestline = 1, .num_pipes = 2,
- .is_mobile = 1, .has_fbc = 1, .has_hotplug = 1,
+ GEN4_FEATURES,
+ .is_crestline = 1,
+ .is_mobile = 1, .has_fbc = 1,
.has_overlay = 1,
.supports_tv = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ .hws_needs_physical = 1,
};
static const struct intel_device_info intel_g33_info = {
- .gen = 3, .is_g33 = 1, .num_pipes = 2,
- .need_gfx_hws = 1, .has_hotplug = 1,
+ GEN3_FEATURES,
+ .is_g33 = 1,
+ .has_hotplug = 1,
.has_overlay = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
};
static const struct intel_device_info intel_g45_info = {
- .gen = 4, .is_g4x = 1, .need_gfx_hws = 1, .num_pipes = 2,
- .has_pipe_cxsr = 1, .has_hotplug = 1,
+ GEN4_FEATURES,
+ .is_g4x = 1,
+ .has_pipe_cxsr = 1,
.ring_mask = RENDER_RING | BSD_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
};
static const struct intel_device_info intel_gm45_info = {
- .gen = 4, .is_g4x = 1, .num_pipes = 2,
- .is_mobile = 1, .need_gfx_hws = 1, .has_fbc = 1,
- .has_pipe_cxsr = 1, .has_hotplug = 1,
+ GEN4_FEATURES,
+ .is_g4x = 1,
+ .is_mobile = 1, .has_fbc = 1,
+ .has_pipe_cxsr = 1,
.supports_tv = 1,
.ring_mask = RENDER_RING | BSD_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
};
static const struct intel_device_info intel_pineview_info = {
- .gen = 3, .is_g33 = 1, .is_pineview = 1, .is_mobile = 1, .num_pipes = 2,
- .need_gfx_hws = 1, .has_hotplug = 1,
+ GEN3_FEATURES,
+ .is_g33 = 1, .is_pineview = 1, .is_mobile = 1,
+ .has_hotplug = 1,
.has_overlay = 1,
- .ring_mask = RENDER_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
};
+#define GEN5_FEATURES \
+ .gen = 5, .num_pipes = 2, \
+ .has_hotplug = 1, \
+ .has_gmbus_irq = 1, \
+ .ring_mask = RENDER_RING | BSD_RING, \
+ GEN_DEFAULT_PIPEOFFSETS, \
+ CURSOR_OFFSETS
+
static const struct intel_device_info intel_ironlake_d_info = {
- .gen = 5, .num_pipes = 2,
- .need_gfx_hws = 1, .has_hotplug = 1,
- .ring_mask = RENDER_RING | BSD_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ GEN5_FEATURES,
};
static const struct intel_device_info intel_ironlake_m_info = {
- .gen = 5, .is_mobile = 1, .num_pipes = 2,
- .need_gfx_hws = 1, .has_hotplug = 1,
- .has_fbc = 1,
- .ring_mask = RENDER_RING | BSD_RING,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ GEN5_FEATURES,
+ .is_mobile = 1,
};
+#define GEN6_FEATURES \
+ .gen = 6, .num_pipes = 2, \
+ .has_hotplug = 1, \
+ .has_fbc = 1, \
+ .ring_mask = RENDER_RING | BSD_RING | BLT_RING, \
+ .has_llc = 1, \
+ .has_rc6 = 1, \
+ .has_rc6p = 1, \
+ .has_gmbus_irq = 1, \
+ .has_hw_contexts = 1, \
+ GEN_DEFAULT_PIPEOFFSETS, \
+ CURSOR_OFFSETS
+
static const struct intel_device_info intel_sandybridge_d_info = {
- .gen = 6, .num_pipes = 2,
- .need_gfx_hws = 1, .has_hotplug = 1,
- .has_fbc = 1,
- .ring_mask = RENDER_RING | BSD_RING | BLT_RING,
- .has_llc = 1,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ GEN6_FEATURES,
};
static const struct intel_device_info intel_sandybridge_m_info = {
- .gen = 6, .is_mobile = 1, .num_pipes = 2,
- .need_gfx_hws = 1, .has_hotplug = 1,
- .has_fbc = 1,
- .ring_mask = RENDER_RING | BSD_RING | BLT_RING,
- .has_llc = 1,
- GEN_DEFAULT_PIPEOFFSETS,
- CURSOR_OFFSETS,
+ GEN6_FEATURES,
+ .is_mobile = 1,
};
#define GEN7_FEATURES \
.gen = 7, .num_pipes = 3, \
- .need_gfx_hws = 1, .has_hotplug = 1, \
+ .has_hotplug = 1, \
.has_fbc = 1, \
.ring_mask = RENDER_RING | BSD_RING | BLT_RING, \
.has_llc = 1, \
+ .has_rc6 = 1, \
+ .has_rc6p = 1, \
+ .has_gmbus_irq = 1, \
+ .has_hw_contexts = 1, \
GEN_DEFAULT_PIPEOFFSETS, \
IVB_CURSOR_OFFSETS
static const struct intel_device_info intel_ivybridge_d_info = {
GEN7_FEATURES,
.is_ivybridge = 1,
+ .has_l3_dpf = 1,
};
static const struct intel_device_info intel_ivybridge_m_info = {
GEN7_FEATURES,
.is_ivybridge = 1,
.is_mobile = 1,
+ .has_l3_dpf = 1,
};
static const struct intel_device_info intel_ivybridge_q_info = {
GEN7_FEATURES,
.is_ivybridge = 1,
.num_pipes = 0, /* legal, last one wins */
+ .has_l3_dpf = 1,
};
#define VLV_FEATURES \
.gen = 7, .num_pipes = 2, \
- .need_gfx_hws = 1, .has_hotplug = 1, \
+ .has_psr = 1, \
+ .has_runtime_pm = 1, \
+ .has_rc6 = 1, \
+ .has_gmbus_irq = 1, \
+ .has_hw_contexts = 1, \
+ .has_gmch_display = 1, \
+ .has_hotplug = 1, \
.ring_mask = RENDER_RING | BSD_RING | BLT_RING, \
.display_mmio_offset = VLV_DISPLAY_BASE, \
GEN_DEFAULT_PIPEOFFSETS, \
CURSOR_OFFSETS
-static const struct intel_device_info intel_valleyview_m_info = {
- VLV_FEATURES,
- .is_valleyview = 1,
- .is_mobile = 1,
-};
-
-static const struct intel_device_info intel_valleyview_d_info = {
+static const struct intel_device_info intel_valleyview_info = {
VLV_FEATURES,
.is_valleyview = 1,
};
GEN7_FEATURES, \
.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING, \
.has_ddi = 1, \
- .has_fpga_dbg = 1
-
-static const struct intel_device_info intel_haswell_d_info = {
- HSW_FEATURES,
- .is_haswell = 1,
-};
-
-static const struct intel_device_info intel_haswell_m_info = {
+ .has_fpga_dbg = 1, \
+ .has_psr = 1, \
+ .has_resource_streamer = 1, \
+ .has_dp_mst = 1, \
+ .has_rc6p = 0 /* RC6p removed-by HSW */, \
+ .has_runtime_pm = 1
+
+static const struct intel_device_info intel_haswell_info = {
HSW_FEATURES,
.is_haswell = 1,
- .is_mobile = 1,
+ .has_l3_dpf = 1,
};
#define BDW_FEATURES \
HSW_FEATURES, \
- BDW_COLORS
+ BDW_COLORS, \
+ .has_logical_ring_contexts = 1
-static const struct intel_device_info intel_broadwell_d_info = {
+static const struct intel_device_info intel_broadwell_info = {
BDW_FEATURES,
.gen = 8,
.is_broadwell = 1,
};
-static const struct intel_device_info intel_broadwell_m_info = {
- BDW_FEATURES,
- .gen = 8, .is_mobile = 1,
- .is_broadwell = 1,
-};
-
-static const struct intel_device_info intel_broadwell_gt3d_info = {
+static const struct intel_device_info intel_broadwell_gt3_info = {
BDW_FEATURES,
.gen = 8,
.is_broadwell = 1,
.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
};
-static const struct intel_device_info intel_broadwell_gt3m_info = {
- BDW_FEATURES,
- .gen = 8, .is_mobile = 1,
- .is_broadwell = 1,
- .ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
-};
-
static const struct intel_device_info intel_cherryview_info = {
.gen = 8, .num_pipes = 3,
- .need_gfx_hws = 1, .has_hotplug = 1,
+ .has_hotplug = 1,
.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
.is_cherryview = 1,
+ .has_psr = 1,
+ .has_runtime_pm = 1,
+ .has_resource_streamer = 1,
+ .has_rc6 = 1,
+ .has_gmbus_irq = 1,
+ .has_hw_contexts = 1,
+ .has_logical_ring_contexts = 1,
+ .has_gmch_display = 1,
.display_mmio_offset = VLV_DISPLAY_BASE,
GEN_CHV_PIPEOFFSETS,
CURSOR_OFFSETS,
BDW_FEATURES,
.is_skylake = 1,
.gen = 9,
+ .has_csr = 1,
+ .has_guc = 1,
+ .ddb_size = 896,
};
static const struct intel_device_info intel_skylake_gt3_info = {
BDW_FEATURES,
.is_skylake = 1,
.gen = 9,
+ .has_csr = 1,
+ .has_guc = 1,
+ .ddb_size = 896,
.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
};
static const struct intel_device_info intel_broxton_info = {
.is_broxton = 1,
.gen = 9,
- .need_gfx_hws = 1, .has_hotplug = 1,
+ .has_hotplug = 1,
.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING,
.num_pipes = 3,
.has_ddi = 1,
.has_fpga_dbg = 1,
.has_fbc = 1,
+ .has_runtime_pm = 1,
.has_pooled_eu = 0,
+ .has_csr = 1,
+ .has_resource_streamer = 1,
+ .has_rc6 = 1,
+ .has_dp_mst = 1,
+ .has_gmbus_irq = 1,
+ .has_hw_contexts = 1,
+ .has_logical_ring_contexts = 1,
+ .has_guc = 1,
+ .ddb_size = 512,
GEN_DEFAULT_PIPEOFFSETS,
IVB_CURSOR_OFFSETS,
BDW_COLORS,
BDW_FEATURES,
.is_kabylake = 1,
.gen = 9,
+ .has_csr = 1,
+ .has_guc = 1,
+ .ddb_size = 896,
};
static const struct intel_device_info intel_kabylake_gt3_info = {
BDW_FEATURES,
.is_kabylake = 1,
.gen = 9,
+ .has_csr = 1,
+ .has_guc = 1,
+ .ddb_size = 896,
.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
};
INTEL_IVB_Q_IDS(&intel_ivybridge_q_info), /* must be first IVB */
INTEL_IVB_M_IDS(&intel_ivybridge_m_info),
INTEL_IVB_D_IDS(&intel_ivybridge_d_info),
- INTEL_HSW_D_IDS(&intel_haswell_d_info),
- INTEL_HSW_M_IDS(&intel_haswell_m_info),
- INTEL_VLV_M_IDS(&intel_valleyview_m_info),
- INTEL_VLV_D_IDS(&intel_valleyview_d_info),
- INTEL_BDW_GT12M_IDS(&intel_broadwell_m_info),
- INTEL_BDW_GT12D_IDS(&intel_broadwell_d_info),
- INTEL_BDW_GT3M_IDS(&intel_broadwell_gt3m_info),
- INTEL_BDW_GT3D_IDS(&intel_broadwell_gt3d_info),
+ INTEL_HSW_IDS(&intel_haswell_info),
+ INTEL_VLV_IDS(&intel_valleyview_info),
+ INTEL_BDW_GT12_IDS(&intel_broadwell_info),
+ INTEL_BDW_GT3_IDS(&intel_broadwell_gt3_info),
INTEL_CHV_IDS(&intel_cherryview_info),
INTEL_SKL_GT1_IDS(&intel_skylake_info),
INTEL_SKL_GT2_IDS(&intel_skylake_info),
#define VIDEO_DIP_ENABLE_SPD_HSW (1 << 0)
/* Panel power sequencing */
-#define PP_STATUS _MMIO(0x61200)
-#define PP_ON (1 << 31)
+#define PPS_BASE 0x61200
+#define VLV_PPS_BASE (VLV_DISPLAY_BASE + PPS_BASE)
+#define PCH_PPS_BASE 0xC7200
+
+#define _MMIO_PPS(pps_idx, reg) _MMIO(dev_priv->pps_mmio_base - \
+ PPS_BASE + (reg) + \
+ (pps_idx) * 0x100)
+
+#define _PP_STATUS 0x61200
+#define PP_STATUS(pps_idx) _MMIO_PPS(pps_idx, _PP_STATUS)
+#define PP_ON (1 << 31)
/*
* Indicates that all dependencies of the panel are on:
*
* - pipe enabled
* - LVDS/DVOB/DVOC on
*/
-#define PP_READY (1 << 30)
-#define PP_SEQUENCE_NONE (0 << 28)
-#define PP_SEQUENCE_POWER_UP (1 << 28)
-#define PP_SEQUENCE_POWER_DOWN (2 << 28)
-#define PP_SEQUENCE_MASK (3 << 28)
-#define PP_SEQUENCE_SHIFT 28
-#define PP_CYCLE_DELAY_ACTIVE (1 << 27)
-#define PP_SEQUENCE_STATE_MASK 0x0000000f
+#define PP_READY (1 << 30)
+#define PP_SEQUENCE_NONE (0 << 28)
+#define PP_SEQUENCE_POWER_UP (1 << 28)
+#define PP_SEQUENCE_POWER_DOWN (2 << 28)
+#define PP_SEQUENCE_MASK (3 << 28)
+#define PP_SEQUENCE_SHIFT 28
+#define PP_CYCLE_DELAY_ACTIVE (1 << 27)
+#define PP_SEQUENCE_STATE_MASK 0x0000000f
#define PP_SEQUENCE_STATE_OFF_IDLE (0x0 << 0)
#define PP_SEQUENCE_STATE_OFF_S0_1 (0x1 << 0)
#define PP_SEQUENCE_STATE_OFF_S0_2 (0x2 << 0)
#define PP_SEQUENCE_STATE_ON_S1_2 (0xa << 0)
#define PP_SEQUENCE_STATE_ON_S1_3 (0xb << 0)
#define PP_SEQUENCE_STATE_RESET (0xf << 0)
-#define PP_CONTROL _MMIO(0x61204)
-#define POWER_TARGET_ON (1 << 0)
-#define PP_ON_DELAYS _MMIO(0x61208)
-#define PP_OFF_DELAYS _MMIO(0x6120c)
-#define PP_DIVISOR _MMIO(0x61210)
+
+#define _PP_CONTROL 0x61204
+#define PP_CONTROL(pps_idx) _MMIO_PPS(pps_idx, _PP_CONTROL)
+#define PANEL_UNLOCK_REGS (0xabcd << 16)
+#define PANEL_UNLOCK_MASK (0xffff << 16)
+#define BXT_POWER_CYCLE_DELAY_MASK 0x1f0
+#define BXT_POWER_CYCLE_DELAY_SHIFT 4
+#define EDP_FORCE_VDD (1 << 3)
+#define EDP_BLC_ENABLE (1 << 2)
+#define PANEL_POWER_RESET (1 << 1)
+#define PANEL_POWER_OFF (0 << 0)
+#define PANEL_POWER_ON (1 << 0)
+
+#define _PP_ON_DELAYS 0x61208
+#define PP_ON_DELAYS(pps_idx) _MMIO_PPS(pps_idx, _PP_ON_DELAYS)
+#define PANEL_PORT_SELECT_SHIFT 30
+#define PANEL_PORT_SELECT_MASK (3 << 30)
+#define PANEL_PORT_SELECT_LVDS (0 << 30)
+#define PANEL_PORT_SELECT_DPA (1 << 30)
+#define PANEL_PORT_SELECT_DPC (2 << 30)
+#define PANEL_PORT_SELECT_DPD (3 << 30)
+#define PANEL_PORT_SELECT_VLV(port) ((port) << 30)
+#define PANEL_POWER_UP_DELAY_MASK 0x1fff0000
+#define PANEL_POWER_UP_DELAY_SHIFT 16
+#define PANEL_LIGHT_ON_DELAY_MASK 0x1fff
+#define PANEL_LIGHT_ON_DELAY_SHIFT 0
+
+#define _PP_OFF_DELAYS 0x6120C
+#define PP_OFF_DELAYS(pps_idx) _MMIO_PPS(pps_idx, _PP_OFF_DELAYS)
+#define PANEL_POWER_DOWN_DELAY_MASK 0x1fff0000
+#define PANEL_POWER_DOWN_DELAY_SHIFT 16
+#define PANEL_LIGHT_OFF_DELAY_MASK 0x1fff
+#define PANEL_LIGHT_OFF_DELAY_SHIFT 0
+
+#define _PP_DIVISOR 0x61210
+#define PP_DIVISOR(pps_idx) _MMIO_PPS(pps_idx, _PP_DIVISOR)
+#define PP_REFERENCE_DIVIDER_MASK 0xffffff00
+#define PP_REFERENCE_DIVIDER_SHIFT 8
+#define PANEL_POWER_CYCLE_DELAY_MASK 0x1f
+#define PANEL_POWER_CYCLE_DELAY_SHIFT 0
/* Panel fitting */
#define PFIT_CONTROL _MMIO(dev_priv->info.display_mmio_offset + 0x61230)
#define PCH_LVDS _MMIO(0xe1180)
#define LVDS_DETECTED (1 << 1)
-/* vlv has 2 sets of panel control regs. */
-#define _PIPEA_PP_STATUS (VLV_DISPLAY_BASE + 0x61200)
-#define _PIPEA_PP_CONTROL (VLV_DISPLAY_BASE + 0x61204)
-#define _PIPEA_PP_ON_DELAYS (VLV_DISPLAY_BASE + 0x61208)
-#define PANEL_PORT_SELECT_VLV(port) ((port) << 30)
-#define _PIPEA_PP_OFF_DELAYS (VLV_DISPLAY_BASE + 0x6120c)
-#define _PIPEA_PP_DIVISOR (VLV_DISPLAY_BASE + 0x61210)
-
-#define _PIPEB_PP_STATUS (VLV_DISPLAY_BASE + 0x61300)
-#define _PIPEB_PP_CONTROL (VLV_DISPLAY_BASE + 0x61304)
-#define _PIPEB_PP_ON_DELAYS (VLV_DISPLAY_BASE + 0x61308)
-#define _PIPEB_PP_OFF_DELAYS (VLV_DISPLAY_BASE + 0x6130c)
-#define _PIPEB_PP_DIVISOR (VLV_DISPLAY_BASE + 0x61310)
-
-#define VLV_PIPE_PP_STATUS(pipe) _MMIO_PIPE(pipe, _PIPEA_PP_STATUS, _PIPEB_PP_STATUS)
-#define VLV_PIPE_PP_CONTROL(pipe) _MMIO_PIPE(pipe, _PIPEA_PP_CONTROL, _PIPEB_PP_CONTROL)
-#define VLV_PIPE_PP_ON_DELAYS(pipe) _MMIO_PIPE(pipe, _PIPEA_PP_ON_DELAYS, _PIPEB_PP_ON_DELAYS)
-#define VLV_PIPE_PP_OFF_DELAYS(pipe) _MMIO_PIPE(pipe, _PIPEA_PP_OFF_DELAYS, _PIPEB_PP_OFF_DELAYS)
-#define VLV_PIPE_PP_DIVISOR(pipe) _MMIO_PIPE(pipe, _PIPEA_PP_DIVISOR, _PIPEB_PP_DIVISOR)
-
-#define _PCH_PP_STATUS 0xc7200
-#define _PCH_PP_CONTROL 0xc7204
-#define PANEL_UNLOCK_REGS (0xabcd << 16)
-#define PANEL_UNLOCK_MASK (0xffff << 16)
-#define BXT_POWER_CYCLE_DELAY_MASK (0x1f0)
-#define BXT_POWER_CYCLE_DELAY_SHIFT 4
-#define EDP_FORCE_VDD (1 << 3)
-#define EDP_BLC_ENABLE (1 << 2)
-#define PANEL_POWER_RESET (1 << 1)
-#define PANEL_POWER_OFF (0 << 0)
-#define PANEL_POWER_ON (1 << 0)
-#define _PCH_PP_ON_DELAYS 0xc7208
-#define PANEL_PORT_SELECT_MASK (3 << 30)
-#define PANEL_PORT_SELECT_LVDS (0 << 30)
-#define PANEL_PORT_SELECT_DPA (1 << 30)
-#define PANEL_PORT_SELECT_DPC (2 << 30)
-#define PANEL_PORT_SELECT_DPD (3 << 30)
-#define PANEL_POWER_UP_DELAY_MASK (0x1fff0000)
-#define PANEL_POWER_UP_DELAY_SHIFT 16
-#define PANEL_LIGHT_ON_DELAY_MASK (0x1fff)
-#define PANEL_LIGHT_ON_DELAY_SHIFT 0
-
-#define _PCH_PP_OFF_DELAYS 0xc720c
-#define PANEL_POWER_DOWN_DELAY_MASK (0x1fff0000)
-#define PANEL_POWER_DOWN_DELAY_SHIFT 16
-#define PANEL_LIGHT_OFF_DELAY_MASK (0x1fff)
-#define PANEL_LIGHT_OFF_DELAY_SHIFT 0
-
-#define _PCH_PP_DIVISOR 0xc7210
-#define PP_REFERENCE_DIVIDER_MASK (0xffffff00)
-#define PP_REFERENCE_DIVIDER_SHIFT 8
-#define PANEL_POWER_CYCLE_DELAY_MASK (0x1f)
-#define PANEL_POWER_CYCLE_DELAY_SHIFT 0
-
-#define PCH_PP_STATUS _MMIO(_PCH_PP_STATUS)
-#define PCH_PP_CONTROL _MMIO(_PCH_PP_CONTROL)
-#define PCH_PP_ON_DELAYS _MMIO(_PCH_PP_ON_DELAYS)
-#define PCH_PP_OFF_DELAYS _MMIO(_PCH_PP_OFF_DELAYS)
-#define PCH_PP_DIVISOR _MMIO(_PCH_PP_DIVISOR)
-
-/* BXT PPS changes - 2nd set of PPS registers */
-#define _BXT_PP_STATUS2 0xc7300
-#define _BXT_PP_CONTROL2 0xc7304
-#define _BXT_PP_ON_DELAYS2 0xc7308
-#define _BXT_PP_OFF_DELAYS2 0xc730c
-
-#define BXT_PP_STATUS(n) _MMIO_PIPE(n, _PCH_PP_STATUS, _BXT_PP_STATUS2)
-#define BXT_PP_CONTROL(n) _MMIO_PIPE(n, _PCH_PP_CONTROL, _BXT_PP_CONTROL2)
-#define BXT_PP_ON_DELAYS(n) _MMIO_PIPE(n, _PCH_PP_ON_DELAYS, _BXT_PP_ON_DELAYS2)
-#define BXT_PP_OFF_DELAYS(n) _MMIO_PIPE(n, _PCH_PP_OFF_DELAYS, _BXT_PP_OFF_DELAYS2)
-
#define _PCH_DP_B 0xe4100
#define PCH_DP_B _MMIO(_PCH_DP_B)
#define _PCH_DPB_AUX_CH_CTL 0xe4110
#define GEN6_RP_UP_THRESHOLD _MMIO(0xA02C)
#define GEN6_RP_DOWN_THRESHOLD _MMIO(0xA030)
#define GEN6_RP_CUR_UP_EI _MMIO(0xA050)
-#define GEN6_CURICONT_MASK 0xffffff
+#define GEN6_RP_EI_MASK 0xffffff
+#define GEN6_CURICONT_MASK GEN6_RP_EI_MASK
#define GEN6_RP_CUR_UP _MMIO(0xA054)
-#define GEN6_CURBSYTAVG_MASK 0xffffff
+#define GEN6_CURBSYTAVG_MASK GEN6_RP_EI_MASK
#define GEN6_RP_PREV_UP _MMIO(0xA058)
#define GEN6_RP_CUR_DOWN_EI _MMIO(0xA05C)
-#define GEN6_CURIAVG_MASK 0xffffff
+#define GEN6_CURIAVG_MASK GEN6_RP_EI_MASK
#define GEN6_RP_CUR_DOWN _MMIO(0xA060)
#define GEN6_RP_PREV_DOWN _MMIO(0xA064)
#define GEN6_RP_UP_EI _MMIO(0xA068)
#define VLV_RCEDATA _MMIO(0xA0BC)
#define GEN6_RC6pp_THRESHOLD _MMIO(0xA0C0)
#define GEN6_PMINTRMSK _MMIO(0xA168)
-#define GEN8_PMINTR_REDIRECT_TO_NON_DISP (1<<31)
+#define GEN8_PMINTR_REDIRECT_TO_GUC (1<<31)
#define GEN8_MISC_CTRL0 _MMIO(0xA180)
#define VLV_PWRDWNUPCTL _MMIO(0xA294)
#define GEN9_MEDIA_PG_IDLE_HYSTERESIS _MMIO(0xA0C4)
#define GEN6_PCODE_MAILBOX _MMIO(0x138124)
#define GEN6_PCODE_READY (1<<31)
+#define GEN6_PCODE_ERROR_MASK 0xFF
+#define GEN6_PCODE_SUCCESS 0x0
+#define GEN6_PCODE_ILLEGAL_CMD 0x1
+#define GEN6_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE 0x2
+#define GEN6_PCODE_TIMEOUT 0x3
+#define GEN6_PCODE_UNIMPLEMENTED_CMD 0xFF
+#define GEN7_PCODE_TIMEOUT 0x2
+#define GEN7_PCODE_ILLEGAL_DATA 0x3
+#define GEN7_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE 0x10
#define GEN6_PCODE_WRITE_RC6VIDS 0x4
#define GEN6_PCODE_READ_RC6VIDS 0x5
#define GEN6_ENCODE_RC6_VID(mv) (((mv) - 245) / 5)
#define HSW_PCODE_DE_WRITE_FREQ_REQ 0x17
#define DISPLAY_IPS_CONTROL 0x19
#define HSW_PCODE_DYNAMIC_DUTY_CYCLE_CONTROL 0x1A
+#define GEN9_PCODE_SAGV_CONTROL 0x21
+#define GEN9_SAGV_DISABLE 0x0
+#define GEN9_SAGV_IS_DISABLED 0x1
+#define GEN9_SAGV_ENABLE 0x3
#define GEN6_PCODE_DATA _MMIO(0x138128)
#define GEN6_PCODE_FREQ_IA_RATIO_SHIFT 8
#define GEN6_PCODE_FREQ_RING_RATIO_SHIFT 16
if (INTEL_INFO(dev)->gen <= 4)
dev_priv->regfile.saveDSPARB = I915_READ(DSPARB);
- /* LVDS state */
- if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
- dev_priv->regfile.saveLVDS = I915_READ(PCH_LVDS);
- else if (INTEL_INFO(dev)->gen <= 4 && IS_MOBILE(dev) && !IS_I830(dev))
- dev_priv->regfile.saveLVDS = I915_READ(LVDS);
-
- /* Panel power sequencer */
- if (HAS_PCH_SPLIT(dev)) {
- dev_priv->regfile.savePP_CONTROL = I915_READ(PCH_PP_CONTROL);
- dev_priv->regfile.savePP_ON_DELAYS = I915_READ(PCH_PP_ON_DELAYS);
- dev_priv->regfile.savePP_OFF_DELAYS = I915_READ(PCH_PP_OFF_DELAYS);
- dev_priv->regfile.savePP_DIVISOR = I915_READ(PCH_PP_DIVISOR);
- } else if (INTEL_INFO(dev)->gen <= 4) {
- dev_priv->regfile.savePP_CONTROL = I915_READ(PP_CONTROL);
- dev_priv->regfile.savePP_ON_DELAYS = I915_READ(PP_ON_DELAYS);
- dev_priv->regfile.savePP_OFF_DELAYS = I915_READ(PP_OFF_DELAYS);
- dev_priv->regfile.savePP_DIVISOR = I915_READ(PP_DIVISOR);
- }
-
/* save FBC interval */
if (HAS_FBC(dev) && INTEL_INFO(dev)->gen <= 4 && !IS_G4X(dev))
dev_priv->regfile.saveFBC_CONTROL = I915_READ(FBC_CONTROL);
static void i915_restore_display(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
- u32 mask = 0xffffffff;
/* Display arbitration */
if (INTEL_INFO(dev)->gen <= 4)
I915_WRITE(DSPARB, dev_priv->regfile.saveDSPARB);
- mask = ~LVDS_PORT_EN;
-
- /* LVDS state */
- if (HAS_PCH_IBX(dev) || HAS_PCH_CPT(dev))
- I915_WRITE(PCH_LVDS, dev_priv->regfile.saveLVDS & mask);
- else if (INTEL_INFO(dev)->gen <= 4 && IS_MOBILE(dev) && !IS_I830(dev))
- I915_WRITE(LVDS, dev_priv->regfile.saveLVDS & mask);
-
- /* Panel power sequencer */
- if (HAS_PCH_SPLIT(dev)) {
- I915_WRITE(PCH_PP_ON_DELAYS, dev_priv->regfile.savePP_ON_DELAYS);
- I915_WRITE(PCH_PP_OFF_DELAYS, dev_priv->regfile.savePP_OFF_DELAYS);
- I915_WRITE(PCH_PP_DIVISOR, dev_priv->regfile.savePP_DIVISOR);
- I915_WRITE(PCH_PP_CONTROL, dev_priv->regfile.savePP_CONTROL);
- } else if (INTEL_INFO(dev)->gen <= 4) {
- I915_WRITE(PP_ON_DELAYS, dev_priv->regfile.savePP_ON_DELAYS);
- I915_WRITE(PP_OFF_DELAYS, dev_priv->regfile.savePP_OFF_DELAYS);
- I915_WRITE(PP_DIVISOR, dev_priv->regfile.savePP_DIVISOR);
- I915_WRITE(PP_CONTROL, dev_priv->regfile.savePP_CONTROL);
- }
-
/* only restore FBC info on the platform that supports FBC*/
intel_fbc_global_disable(dev_priv);
int i915_save_state(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
int i;
mutex_lock(&dev->struct_mutex);
i915_save_display(dev);
if (IS_GEN4(dev))
- pci_read_config_word(dev->pdev, GCDGMBUS,
+ pci_read_config_word(pdev, GCDGMBUS,
&dev_priv->regfile.saveGCDGMBUS);
/* Cache mode state */
int i915_restore_state(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
int i;
mutex_lock(&dev->struct_mutex);
i915_gem_restore_fences(dev);
if (IS_GEN4(dev))
- pci_write_config_word(dev->pdev, GCDGMBUS,
+ pci_write_config_word(pdev, GCDGMBUS,
dev_priv->regfile.saveGCDGMBUS);
i915_restore_display(dev);
--- /dev/null
+/*
+ * (C) Copyright 2016 Intel Corporation
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation; version 2
+ * of the License.
+ */
+
+#include <linux/slab.h>
+#include <linux/fence.h>
+#include <linux/reservation.h>
+
+#include "i915_sw_fence.h"
+
+static DEFINE_SPINLOCK(i915_sw_fence_lock);
+
+static int __i915_sw_fence_notify(struct i915_sw_fence *fence,
+ enum i915_sw_fence_notify state)
+{
+ i915_sw_fence_notify_t fn;
+
+ fn = (i915_sw_fence_notify_t)(fence->flags & I915_SW_FENCE_MASK);
+ return fn(fence, state);
+}
+
+static void i915_sw_fence_free(struct kref *kref)
+{
+ struct i915_sw_fence *fence = container_of(kref, typeof(*fence), kref);
+
+ WARN_ON(atomic_read(&fence->pending) > 0);
+
+ if (fence->flags & I915_SW_FENCE_MASK)
+ __i915_sw_fence_notify(fence, FENCE_FREE);
+ else
+ kfree(fence);
+}
+
+static void i915_sw_fence_put(struct i915_sw_fence *fence)
+{
+ kref_put(&fence->kref, i915_sw_fence_free);
+}
+
+static struct i915_sw_fence *i915_sw_fence_get(struct i915_sw_fence *fence)
+{
+ kref_get(&fence->kref);
+ return fence;
+}
+
+static void __i915_sw_fence_wake_up_all(struct i915_sw_fence *fence,
+ struct list_head *continuation)
+{
+ wait_queue_head_t *x = &fence->wait;
+ wait_queue_t *pos, *next;
+ unsigned long flags;
+
+ atomic_set_release(&fence->pending, -1); /* 0 -> -1 [done] */
+
+ /*
+ * To prevent unbounded recursion as we traverse the graph of
+ * i915_sw_fences, we move the task_list from this, the next ready
+ * fence, to the tail of the original fence's task_list
+ * (and so added to the list to be woken).
+ */
+
+ spin_lock_irqsave_nested(&x->lock, flags, 1 + !!continuation);
+ if (continuation) {
+ list_for_each_entry_safe(pos, next, &x->task_list, task_list) {
+ if (pos->func == autoremove_wake_function)
+ pos->func(pos, TASK_NORMAL, 0, continuation);
+ else
+ list_move_tail(&pos->task_list, continuation);
+ }
+ } else {
+ LIST_HEAD(extra);
+
+ do {
+ list_for_each_entry_safe(pos, next,
+ &x->task_list, task_list)
+ pos->func(pos, TASK_NORMAL, 0, &extra);
+
+ if (list_empty(&extra))
+ break;
+
+ list_splice_tail_init(&extra, &x->task_list);
+ } while (1);
+ }
+ spin_unlock_irqrestore(&x->lock, flags);
+}
+
+static void __i915_sw_fence_complete(struct i915_sw_fence *fence,
+ struct list_head *continuation)
+{
+ if (!atomic_dec_and_test(&fence->pending))
+ return;
+
+ if (fence->flags & I915_SW_FENCE_MASK &&
+ __i915_sw_fence_notify(fence, FENCE_COMPLETE) != NOTIFY_DONE)
+ return;
+
+ __i915_sw_fence_wake_up_all(fence, continuation);
+}
+
+static void i915_sw_fence_complete(struct i915_sw_fence *fence)
+{
+ if (WARN_ON(i915_sw_fence_done(fence)))
+ return;
+
+ __i915_sw_fence_complete(fence, NULL);
+}
+
+static void i915_sw_fence_await(struct i915_sw_fence *fence)
+{
+ WARN_ON(atomic_inc_return(&fence->pending) <= 1);
+}
+
+void i915_sw_fence_init(struct i915_sw_fence *fence, i915_sw_fence_notify_t fn)
+{
+ BUG_ON((unsigned long)fn & ~I915_SW_FENCE_MASK);
+
+ init_waitqueue_head(&fence->wait);
+ kref_init(&fence->kref);
+ atomic_set(&fence->pending, 1);
+ fence->flags = (unsigned long)fn;
+}
+
+void i915_sw_fence_commit(struct i915_sw_fence *fence)
+{
+ i915_sw_fence_complete(fence);
+ i915_sw_fence_put(fence);
+}
+
+static int i915_sw_fence_wake(wait_queue_t *wq, unsigned mode, int flags, void *key)
+{
+ list_del(&wq->task_list);
+ __i915_sw_fence_complete(wq->private, key);
+ i915_sw_fence_put(wq->private);
+ return 0;
+}
+
+static bool __i915_sw_fence_check_if_after(struct i915_sw_fence *fence,
+ const struct i915_sw_fence * const signaler)
+{
+ wait_queue_t *wq;
+
+ if (__test_and_set_bit(I915_SW_FENCE_CHECKED_BIT, &fence->flags))
+ return false;
+
+ if (fence == signaler)
+ return true;
+
+ list_for_each_entry(wq, &fence->wait.task_list, task_list) {
+ if (wq->func != i915_sw_fence_wake)
+ continue;
+
+ if (__i915_sw_fence_check_if_after(wq->private, signaler))
+ return true;
+ }
+
+ return false;
+}
+
+static void __i915_sw_fence_clear_checked_bit(struct i915_sw_fence *fence)
+{
+ wait_queue_t *wq;
+
+ if (!__test_and_clear_bit(I915_SW_FENCE_CHECKED_BIT, &fence->flags))
+ return;
+
+ list_for_each_entry(wq, &fence->wait.task_list, task_list) {
+ if (wq->func != i915_sw_fence_wake)
+ continue;
+
+ __i915_sw_fence_clear_checked_bit(wq->private);
+ }
+}
+
+static bool i915_sw_fence_check_if_after(struct i915_sw_fence *fence,
+ const struct i915_sw_fence * const signaler)
+{
+ unsigned long flags;
+ bool err;
+
+ if (!IS_ENABLED(CONFIG_I915_SW_FENCE_CHECK_DAG))
+ return false;
+
+ spin_lock_irqsave(&i915_sw_fence_lock, flags);
+ err = __i915_sw_fence_check_if_after(fence, signaler);
+ __i915_sw_fence_clear_checked_bit(fence);
+ spin_unlock_irqrestore(&i915_sw_fence_lock, flags);
+
+ return err;
+}
+
+int i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
+ struct i915_sw_fence *signaler,
+ wait_queue_t *wq)
+{
+ unsigned long flags;
+ int pending;
+
+ if (i915_sw_fence_done(signaler))
+ return 0;
+
+ /* The dependency graph must be acyclic. */
+ if (unlikely(i915_sw_fence_check_if_after(fence, signaler)))
+ return -EINVAL;
+
+ INIT_LIST_HEAD(&wq->task_list);
+ wq->flags = 0;
+ wq->func = i915_sw_fence_wake;
+ wq->private = i915_sw_fence_get(fence);
+
+ i915_sw_fence_await(fence);
+
+ spin_lock_irqsave(&signaler->wait.lock, flags);
+ if (likely(!i915_sw_fence_done(signaler))) {
+ __add_wait_queue_tail(&signaler->wait, wq);
+ pending = 1;
+ } else {
+ i915_sw_fence_wake(wq, 0, 0, NULL);
+ pending = 0;
+ }
+ spin_unlock_irqrestore(&signaler->wait.lock, flags);
+
+ return pending;
+}
+
+struct dma_fence_cb {
+ struct fence_cb base;
+ struct i915_sw_fence *fence;
+ struct fence *dma;
+ struct timer_list timer;
+};
+
+static void timer_i915_sw_fence_wake(unsigned long data)
+{
+ struct dma_fence_cb *cb = (struct dma_fence_cb *)data;
+
+ printk(KERN_WARNING "asynchronous wait on fence %s:%s:%x timed out\n",
+ cb->dma->ops->get_driver_name(cb->dma),
+ cb->dma->ops->get_timeline_name(cb->dma),
+ cb->dma->seqno);
+ fence_put(cb->dma);
+ cb->dma = NULL;
+
+ i915_sw_fence_commit(cb->fence);
+ cb->timer.function = NULL;
+}
+
+static void dma_i915_sw_fence_wake(struct fence *dma, struct fence_cb *data)
+{
+ struct dma_fence_cb *cb = container_of(data, typeof(*cb), base);
+
+ del_timer_sync(&cb->timer);
+ if (cb->timer.function)
+ i915_sw_fence_commit(cb->fence);
+ fence_put(cb->dma);
+
+ kfree(cb);
+}
+
+int i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence,
+ struct fence *dma,
+ unsigned long timeout,
+ gfp_t gfp)
+{
+ struct dma_fence_cb *cb;
+ int ret;
+
+ if (fence_is_signaled(dma))
+ return 0;
+
+ cb = kmalloc(sizeof(*cb), gfp);
+ if (!cb) {
+ if (!gfpflags_allow_blocking(gfp))
+ return -ENOMEM;
+
+ return fence_wait(dma, false);
+ }
+
+ cb->fence = i915_sw_fence_get(fence);
+ i915_sw_fence_await(fence);
+
+ cb->dma = NULL;
+ __setup_timer(&cb->timer,
+ timer_i915_sw_fence_wake, (unsigned long)cb,
+ TIMER_IRQSAFE);
+ if (timeout) {
+ cb->dma = fence_get(dma);
+ mod_timer(&cb->timer, round_jiffies_up(jiffies + timeout));
+ }
+
+ ret = fence_add_callback(dma, &cb->base, dma_i915_sw_fence_wake);
+ if (ret == 0) {
+ ret = 1;
+ } else {
+ dma_i915_sw_fence_wake(dma, &cb->base);
+ if (ret == -ENOENT) /* fence already signaled */
+ ret = 0;
+ }
+
+ return ret;
+}
+
+int i915_sw_fence_await_reservation(struct i915_sw_fence *fence,
+ struct reservation_object *resv,
+ const struct fence_ops *exclude,
+ bool write,
+ unsigned long timeout,
+ gfp_t gfp)
+{
+ struct fence *excl;
+ int ret = 0, pending;
+
+ if (write) {
+ struct fence **shared;
+ unsigned int count, i;
+
+ ret = reservation_object_get_fences_rcu(resv,
+ &excl, &count, &shared);
+ if (ret)
+ return ret;
+
+ for (i = 0; i < count; i++) {
+ if (shared[i]->ops == exclude)
+ continue;
+
+ pending = i915_sw_fence_await_dma_fence(fence,
+ shared[i],
+ timeout,
+ gfp);
+ if (pending < 0) {
+ ret = pending;
+ break;
+ }
+
+ ret |= pending;
+ }
+
+ for (i = 0; i < count; i++)
+ fence_put(shared[i]);
+ kfree(shared);
+ } else {
+ excl = reservation_object_get_excl_rcu(resv);
+ }
+
+ if (ret >= 0 && excl && excl->ops != exclude) {
+ pending = i915_sw_fence_await_dma_fence(fence,
+ excl,
+ timeout,
+ gfp);
+ if (pending < 0)
+ ret = pending;
+ else
+ ret |= pending;
+ }
+
+ fence_put(excl);
+
+ return ret;
+}
--- /dev/null
+/*
+ * i915_sw_fence.h - library routines for N:M synchronisation points
+ *
+ * Copyright (C) 2016 Intel Corporation
+ *
+ * This file is released under the GPLv2.
+ *
+ */
+
+#ifndef _I915_SW_FENCE_H_
+#define _I915_SW_FENCE_H_
+
+#include <linux/gfp.h>
+#include <linux/kref.h>
+#include <linux/notifier.h> /* for NOTIFY_DONE */
+#include <linux/wait.h>
+
+struct completion;
+struct fence;
+struct fence_ops;
+struct reservation_object;
+
+struct i915_sw_fence {
+ wait_queue_head_t wait;
+ unsigned long flags;
+ struct kref kref;
+ atomic_t pending;
+};
+
+#define I915_SW_FENCE_CHECKED_BIT 0 /* used internally for DAG checking */
+#define I915_SW_FENCE_PRIVATE_BIT 1 /* available for use by owner */
+#define I915_SW_FENCE_MASK (~3)
+
+enum i915_sw_fence_notify {
+ FENCE_COMPLETE,
+ FENCE_FREE
+};
+
+typedef int (*i915_sw_fence_notify_t)(struct i915_sw_fence *,
+ enum i915_sw_fence_notify state);
+#define __i915_sw_fence_call __aligned(4)
+
+void i915_sw_fence_init(struct i915_sw_fence *fence, i915_sw_fence_notify_t fn);
+void i915_sw_fence_commit(struct i915_sw_fence *fence);
+
+int i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
+ struct i915_sw_fence *after,
+ wait_queue_t *wq);
+int i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence,
+ struct fence *dma,
+ unsigned long timeout,
+ gfp_t gfp);
+int i915_sw_fence_await_reservation(struct i915_sw_fence *fence,
+ struct reservation_object *resv,
+ const struct fence_ops *exclude,
+ bool write,
+ unsigned long timeout,
+ gfp_t gfp);
+
+static inline bool i915_sw_fence_done(const struct i915_sw_fence *fence)
+{
+ return atomic_read(&fence->pending) < 0;
+}
+
+#endif /* _I915_SW_FENCE_H_ */
#include "intel_drv.h"
#include "i915_drv.h"
-#define dev_to_drm_minor(d) dev_get_drvdata((d))
+static inline struct drm_i915_private *kdev_minor_to_i915(struct device *kdev)
+{
+ struct drm_minor *minor = dev_get_drvdata(kdev);
+ return to_i915(minor->dev);
+}
#ifdef CONFIG_PM
-static u32 calc_residency(struct drm_device *dev,
+static u32 calc_residency(struct drm_i915_private *dev_priv,
i915_reg_t reg)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
u64 raw_time; /* 32b value may overflow during fixed point math */
u64 units = 128ULL, div = 100000ULL;
u32 ret;
intel_runtime_pm_get(dev_priv);
/* On VLV and CHV, residency time is in CZ units rather than 1.28us */
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) {
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
units = 1;
div = dev_priv->czclk_freq;
if (I915_READ(VLV_COUNTER_CONTROL) & VLV_COUNT_RANGE_HIGH)
units <<= 8;
- } else if (IS_BROXTON(dev)) {
+ } else if (IS_BROXTON(dev_priv)) {
units = 1;
div = 1200; /* 833.33ns */
}
static ssize_t
show_rc6_ms(struct device *kdev, struct device_attribute *attr, char *buf)
{
- struct drm_minor *dminor = dev_get_drvdata(kdev);
- u32 rc6_residency = calc_residency(dminor->dev, GEN6_GT_GFX_RC6);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+ u32 rc6_residency = calc_residency(dev_priv, GEN6_GT_GFX_RC6);
return snprintf(buf, PAGE_SIZE, "%u\n", rc6_residency);
}
static ssize_t
show_rc6p_ms(struct device *kdev, struct device_attribute *attr, char *buf)
{
- struct drm_minor *dminor = dev_to_drm_minor(kdev);
- u32 rc6p_residency = calc_residency(dminor->dev, GEN6_GT_GFX_RC6p);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+ u32 rc6p_residency = calc_residency(dev_priv, GEN6_GT_GFX_RC6p);
return snprintf(buf, PAGE_SIZE, "%u\n", rc6p_residency);
}
static ssize_t
show_rc6pp_ms(struct device *kdev, struct device_attribute *attr, char *buf)
{
- struct drm_minor *dminor = dev_to_drm_minor(kdev);
- u32 rc6pp_residency = calc_residency(dminor->dev, GEN6_GT_GFX_RC6pp);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+ u32 rc6pp_residency = calc_residency(dev_priv, GEN6_GT_GFX_RC6pp);
return snprintf(buf, PAGE_SIZE, "%u\n", rc6pp_residency);
}
static ssize_t
show_media_rc6_ms(struct device *kdev, struct device_attribute *attr, char *buf)
{
- struct drm_minor *dminor = dev_get_drvdata(kdev);
- u32 rc6_residency = calc_residency(dminor->dev, VLV_GT_MEDIA_RC6);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+ u32 rc6_residency = calc_residency(dev_priv, VLV_GT_MEDIA_RC6);
return snprintf(buf, PAGE_SIZE, "%u\n", rc6_residency);
}
};
#endif
-static int l3_access_valid(struct drm_device *dev, loff_t offset)
+static int l3_access_valid(struct drm_i915_private *dev_priv, loff_t offset)
{
- if (!HAS_L3_DPF(dev))
+ if (!HAS_L3_DPF(dev_priv))
return -EPERM;
if (offset % 4 != 0)
struct bin_attribute *attr, char *buf,
loff_t offset, size_t count)
{
- struct device *dev = kobj_to_dev(kobj);
- struct drm_minor *dminor = dev_to_drm_minor(dev);
- struct drm_device *drm_dev = dminor->dev;
- struct drm_i915_private *dev_priv = to_i915(drm_dev);
+ struct device *kdev = kobj_to_dev(kobj);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+ struct drm_device *dev = &dev_priv->drm;
int slice = (int)(uintptr_t)attr->private;
int ret;
count = round_down(count, 4);
- ret = l3_access_valid(drm_dev, offset);
+ ret = l3_access_valid(dev_priv, offset);
if (ret)
return ret;
count = min_t(size_t, GEN7_L3LOG_SIZE - offset, count);
- ret = i915_mutex_lock_interruptible(drm_dev);
+ ret = i915_mutex_lock_interruptible(dev);
if (ret)
return ret;
else
memset(buf, 0, count);
- mutex_unlock(&drm_dev->struct_mutex);
+ mutex_unlock(&dev->struct_mutex);
return count;
}
struct bin_attribute *attr, char *buf,
loff_t offset, size_t count)
{
- struct device *dev = kobj_to_dev(kobj);
- struct drm_minor *dminor = dev_to_drm_minor(dev);
- struct drm_device *drm_dev = dminor->dev;
- struct drm_i915_private *dev_priv = to_i915(drm_dev);
+ struct device *kdev = kobj_to_dev(kobj);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+ struct drm_device *dev = &dev_priv->drm;
struct i915_gem_context *ctx;
u32 *temp = NULL; /* Just here to make handling failures easy */
int slice = (int)(uintptr_t)attr->private;
int ret;
- if (!HAS_HW_CONTEXTS(drm_dev))
+ if (!HAS_HW_CONTEXTS(dev_priv))
return -ENXIO;
- ret = l3_access_valid(drm_dev, offset);
+ ret = l3_access_valid(dev_priv, offset);
if (ret)
return ret;
- ret = i915_mutex_lock_interruptible(drm_dev);
+ ret = i915_mutex_lock_interruptible(dev);
if (ret)
return ret;
if (!dev_priv->l3_parity.remap_info[slice]) {
temp = kzalloc(GEN7_L3LOG_SIZE, GFP_KERNEL);
if (!temp) {
- mutex_unlock(&drm_dev->struct_mutex);
+ mutex_unlock(&dev->struct_mutex);
return -ENOMEM;
}
}
list_for_each_entry(ctx, &dev_priv->context_list, link)
ctx->remap_slice |= (1<<slice);
- mutex_unlock(&drm_dev->struct_mutex);
+ mutex_unlock(&dev->struct_mutex);
return count;
}
static ssize_t gt_act_freq_mhz_show(struct device *kdev,
struct device_attribute *attr, char *buf)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
int ret;
intel_runtime_pm_get(dev_priv);
static ssize_t gt_cur_freq_mhz_show(struct device *kdev,
struct device_attribute *attr, char *buf)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
return snprintf(buf, PAGE_SIZE, "%d\n",
intel_gpu_freq(dev_priv,
static ssize_t gt_boost_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_i915_private *dev_priv = to_i915(minor->dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
return snprintf(buf, PAGE_SIZE, "%d\n",
intel_gpu_freq(dev_priv,
struct device_attribute *attr,
const char *buf, size_t count)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
u32 val;
ssize_t ret;
static ssize_t vlv_rpe_freq_mhz_show(struct device *kdev,
struct device_attribute *attr, char *buf)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
return snprintf(buf, PAGE_SIZE, "%d\n",
intel_gpu_freq(dev_priv,
static ssize_t gt_max_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
return snprintf(buf, PAGE_SIZE, "%d\n",
intel_gpu_freq(dev_priv,
struct device_attribute *attr,
const char *buf, size_t count)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
u32 val;
ssize_t ret;
static ssize_t gt_min_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
return snprintf(buf, PAGE_SIZE, "%d\n",
intel_gpu_freq(dev_priv,
struct device_attribute *attr,
const char *buf, size_t count)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
u32 val;
ssize_t ret;
/* For now we have a static number of RP states */
static ssize_t gt_rp_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
{
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
u32 val;
if (attr == &dev_attr_gt_RP0_freq_mhz)
{
struct device *kdev = kobj_to_dev(kobj);
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
+ struct drm_device *dev = &dev_priv->drm;
struct i915_error_state_file_priv error_priv;
struct drm_i915_error_state_buf error_str;
ssize_t ret_count = 0;
loff_t off, size_t count)
{
struct device *kdev = kobj_to_dev(kobj);
- struct drm_minor *minor = dev_to_drm_minor(kdev);
- struct drm_device *dev = minor->dev;
- int ret;
+ struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
DRM_DEBUG_DRIVER("Resetting error state\n");
-
- ret = mutex_lock_interruptible(&dev->struct_mutex);
- if (ret)
- return ret;
-
- i915_destroy_error_state(dev);
- mutex_unlock(&dev->struct_mutex);
+ i915_destroy_error_state(&dev_priv->drm);
return count;
}
.write = error_state_write,
};
-void i915_setup_sysfs(struct drm_device *dev)
+void i915_setup_sysfs(struct drm_i915_private *dev_priv)
{
+ struct device *kdev = dev_priv->drm.primary->kdev;
int ret;
#ifdef CONFIG_PM
- if (HAS_RC6(dev)) {
- ret = sysfs_merge_group(&dev->primary->kdev->kobj,
+ if (HAS_RC6(dev_priv)) {
+ ret = sysfs_merge_group(&kdev->kobj,
&rc6_attr_group);
if (ret)
DRM_ERROR("RC6 residency sysfs setup failed\n");
}
- if (HAS_RC6p(dev)) {
- ret = sysfs_merge_group(&dev->primary->kdev->kobj,
+ if (HAS_RC6p(dev_priv)) {
+ ret = sysfs_merge_group(&kdev->kobj,
&rc6p_attr_group);
if (ret)
DRM_ERROR("RC6p residency sysfs setup failed\n");
}
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) {
- ret = sysfs_merge_group(&dev->primary->kdev->kobj,
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
+ ret = sysfs_merge_group(&kdev->kobj,
&media_rc6_attr_group);
if (ret)
DRM_ERROR("Media RC6 residency sysfs setup failed\n");
}
#endif
- if (HAS_L3_DPF(dev)) {
- ret = device_create_bin_file(dev->primary->kdev, &dpf_attrs);
+ if (HAS_L3_DPF(dev_priv)) {
+ ret = device_create_bin_file(kdev, &dpf_attrs);
if (ret)
DRM_ERROR("l3 parity sysfs setup failed\n");
- if (NUM_L3_SLICES(dev) > 1) {
- ret = device_create_bin_file(dev->primary->kdev,
+ if (NUM_L3_SLICES(dev_priv) > 1) {
+ ret = device_create_bin_file(kdev,
&dpf_attrs_1);
if (ret)
DRM_ERROR("l3 parity slice 1 setup failed\n");
}
ret = 0;
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
- ret = sysfs_create_files(&dev->primary->kdev->kobj, vlv_attrs);
- else if (INTEL_INFO(dev)->gen >= 6)
- ret = sysfs_create_files(&dev->primary->kdev->kobj, gen6_attrs);
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+ ret = sysfs_create_files(&kdev->kobj, vlv_attrs);
+ else if (INTEL_GEN(dev_priv) >= 6)
+ ret = sysfs_create_files(&kdev->kobj, gen6_attrs);
if (ret)
DRM_ERROR("RPS sysfs setup failed\n");
- ret = sysfs_create_bin_file(&dev->primary->kdev->kobj,
+ ret = sysfs_create_bin_file(&kdev->kobj,
&error_state_attr);
if (ret)
DRM_ERROR("error_state sysfs setup failed\n");
}
-void i915_teardown_sysfs(struct drm_device *dev)
+void i915_teardown_sysfs(struct drm_i915_private *dev_priv)
{
- sysfs_remove_bin_file(&dev->primary->kdev->kobj, &error_state_attr);
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
- sysfs_remove_files(&dev->primary->kdev->kobj, vlv_attrs);
+ struct device *kdev = dev_priv->drm.primary->kdev;
+
+ sysfs_remove_bin_file(&kdev->kobj, &error_state_attr);
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+ sysfs_remove_files(&kdev->kobj, vlv_attrs);
else
- sysfs_remove_files(&dev->primary->kdev->kobj, gen6_attrs);
- device_remove_bin_file(dev->primary->kdev, &dpf_attrs_1);
- device_remove_bin_file(dev->primary->kdev, &dpf_attrs);
+ sysfs_remove_files(&kdev->kobj, gen6_attrs);
+ device_remove_bin_file(kdev, &dpf_attrs_1);
+ device_remove_bin_file(kdev, &dpf_attrs);
#ifdef CONFIG_PM
- sysfs_unmerge_group(&dev->primary->kdev->kobj, &rc6_attr_group);
- sysfs_unmerge_group(&dev->primary->kdev->kobj, &rc6p_attr_group);
+ sysfs_unmerge_group(&kdev->kobj, &rc6_attr_group);
+ sysfs_unmerge_group(&kdev->kobj, &rc6p_attr_group);
#endif
}
BUILD_BUG_ON(sizeof(struct vgt_if) != VGT_PVINFO_SIZE);
- if (!IS_HASWELL(dev_priv))
- return;
-
magic = __raw_i915_read64(dev_priv, vgtif_reg(magic));
if (magic != VGT_MAGIC)
return;
{
struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
- struct intel_digital_port *intel_dig_port =
- enc_to_dig_port(&encoder->base);
- enum port port = intel_dig_port->port;
+ enum port port = enc_to_dig_port(&encoder->base)->port;
enum pipe pipe = intel_crtc->pipe;
uint32_t tmp, eldv;
i915_reg_t aud_config, aud_cntrl_st2;
{
struct drm_i915_private *dev_priv = to_i915(connector->dev);
struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
- struct intel_digital_port *intel_dig_port =
- enc_to_dig_port(&encoder->base);
- enum port port = intel_dig_port->port;
+ enum port port = enc_to_dig_port(&encoder->base)->port;
enum pipe pipe = intel_crtc->pipe;
uint8_t *eld = connector->eld;
- uint32_t eldv;
- uint32_t tmp;
+ uint32_t tmp, eldv;
int len, i;
i915_reg_t hdmiw_hdmiedid, aud_config, aud_cntl_st, aud_cntrl_st2;
}
}
-static void i915_audio_component_get_power(struct device *dev)
+static void i915_audio_component_get_power(struct device *kdev)
{
- intel_display_power_get(dev_to_i915(dev), POWER_DOMAIN_AUDIO);
+ intel_display_power_get(kdev_to_i915(kdev), POWER_DOMAIN_AUDIO);
}
-static void i915_audio_component_put_power(struct device *dev)
+static void i915_audio_component_put_power(struct device *kdev)
{
- intel_display_power_put(dev_to_i915(dev), POWER_DOMAIN_AUDIO);
+ intel_display_power_put(kdev_to_i915(kdev), POWER_DOMAIN_AUDIO);
}
-static void i915_audio_component_codec_wake_override(struct device *dev,
+static void i915_audio_component_codec_wake_override(struct device *kdev,
bool enable)
{
- struct drm_i915_private *dev_priv = dev_to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_to_i915(kdev);
u32 tmp;
if (!IS_SKYLAKE(dev_priv) && !IS_KABYLAKE(dev_priv))
return;
- i915_audio_component_get_power(dev);
+ i915_audio_component_get_power(kdev);
/*
* Enable/disable generating the codec wake signal, overriding the
usleep_range(1000, 1500);
}
- i915_audio_component_put_power(dev);
+ i915_audio_component_put_power(kdev);
}
/* Get CDCLK in kHz */
-static int i915_audio_component_get_cdclk_freq(struct device *dev)
+static int i915_audio_component_get_cdclk_freq(struct device *kdev)
{
- struct drm_i915_private *dev_priv = dev_to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_to_i915(kdev);
if (WARN_ON_ONCE(!HAS_DDI(dev_priv)))
return -ENODEV;
return dev_priv->cdclk_freq;
}
-static int i915_audio_component_sync_audio_rate(struct device *dev,
+static int i915_audio_component_sync_audio_rate(struct device *kdev,
int port, int rate)
{
- struct drm_i915_private *dev_priv = dev_to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_to_i915(kdev);
struct intel_encoder *intel_encoder;
struct intel_crtc *crtc;
struct drm_display_mode *mode;
!IS_HASWELL(dev_priv))
return 0;
- i915_audio_component_get_power(dev);
+ i915_audio_component_get_power(kdev);
mutex_lock(&dev_priv->av_mutex);
/* 1. get the pipe */
intel_encoder = dev_priv->dig_port_map[port];
unlock:
mutex_unlock(&dev_priv->av_mutex);
- i915_audio_component_put_power(dev);
+ i915_audio_component_put_power(kdev);
return err;
}
-static int i915_audio_component_get_eld(struct device *dev, int port,
+static int i915_audio_component_get_eld(struct device *kdev, int port,
bool *enabled,
unsigned char *buf, int max_bytes)
{
- struct drm_i915_private *dev_priv = dev_to_i915(dev);
+ struct drm_i915_private *dev_priv = kdev_to_i915(kdev);
struct intel_encoder *intel_encoder;
struct intel_digital_port *intel_dig_port;
const u8 *eld;
.get_eld = i915_audio_component_get_eld,
};
-static int i915_audio_component_bind(struct device *i915_dev,
- struct device *hda_dev, void *data)
+static int i915_audio_component_bind(struct device *i915_kdev,
+ struct device *hda_kdev, void *data)
{
struct i915_audio_component *acomp = data;
- struct drm_i915_private *dev_priv = dev_to_i915(i915_dev);
+ struct drm_i915_private *dev_priv = kdev_to_i915(i915_kdev);
int i;
if (WARN_ON(acomp->ops || acomp->dev))
drm_modeset_lock_all(&dev_priv->drm);
acomp->ops = &i915_audio_component_ops;
- acomp->dev = i915_dev;
+ acomp->dev = i915_kdev;
BUILD_BUG_ON(MAX_PORTS != I915_MAX_PORTS);
for (i = 0; i < ARRAY_SIZE(acomp->aud_sample_rate); i++)
acomp->aud_sample_rate[i] = 0;
return 0;
}
-static void i915_audio_component_unbind(struct device *i915_dev,
- struct device *hda_dev, void *data)
+static void i915_audio_component_unbind(struct device *i915_kdev,
+ struct device *hda_kdev, void *data)
{
struct i915_audio_component *acomp = data;
- struct drm_i915_private *dev_priv = dev_to_i915(i915_dev);
+ struct drm_i915_private *dev_priv = kdev_to_i915(i915_kdev);
drm_modeset_lock_all(&dev_priv->drm);
acomp->ops = NULL;
#include "i915_drv.h"
+static void intel_breadcrumbs_hangcheck(unsigned long data)
+{
+ struct intel_engine_cs *engine = (struct intel_engine_cs *)data;
+ struct intel_breadcrumbs *b = &engine->breadcrumbs;
+
+ if (!b->irq_enabled)
+ return;
+
+ if (time_before(jiffies, b->timeout)) {
+ mod_timer(&b->hangcheck, b->timeout);
+ return;
+ }
+
+ DRM_DEBUG("Hangcheck timer elapsed... %s idle\n", engine->name);
+ set_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings);
+ mod_timer(&engine->breadcrumbs.fake_irq, jiffies + 1);
+
+ /* Ensure that even if the GPU hangs, we get woken up.
+ *
+ * However, note that if no one is waiting, we never notice
+ * a gpu hang. Eventually, we will have to wait for a resource
+ * held by the GPU and so trigger a hangcheck. In the most
+ * pathological case, this will be upon memory starvation! To
+ * prevent this, we also queue the hangcheck from the retire
+ * worker.
+ */
+ i915_queue_hangcheck(engine->i915);
+}
+
+static unsigned long wait_timeout(void)
+{
+ return round_jiffies_up(jiffies + DRM_I915_HANGCHECK_JIFFIES);
+}
+
static void intel_breadcrumbs_fake_irq(unsigned long data)
{
struct intel_engine_cs *engine = (struct intel_engine_cs *)data;
* every jiffie in order to kick the oldest waiter to do the
* coherent seqno check.
*/
- rcu_read_lock();
if (intel_engine_wakeup(engine))
mod_timer(&engine->breadcrumbs.fake_irq, jiffies + 1);
- rcu_read_unlock();
}
static void irq_enable(struct intel_engine_cs *engine)
*/
engine->breadcrumbs.irq_posted = true;
- /* Make sure the current hangcheck doesn't falsely accuse a just
- * started irq handler from missing an interrupt (because the
- * interrupt count still matches the stale value from when
- * the irq handler was disabled, many hangchecks ago).
- */
- engine->breadcrumbs.irq_wakeups++;
-
spin_lock_irq(&engine->i915->irq_lock);
engine->irq_enable(engine);
spin_unlock_irq(&engine->i915->irq_lock);
}
if (!b->irq_enabled ||
- test_bit(engine->id, &i915->gpu_error.missed_irq_rings))
+ test_bit(engine->id, &i915->gpu_error.missed_irq_rings)) {
mod_timer(&b->fake_irq, jiffies + 1);
-
- /* Ensure that even if the GPU hangs, we get woken up.
- *
- * However, note that if no one is waiting, we never notice
- * a gpu hang. Eventually, we will have to wait for a resource
- * held by the GPU and so trigger a hangcheck. In the most
- * pathological case, this will be upon memory starvation!
- */
- i915_queue_hangcheck(i915);
+ } else {
+ /* Ensure we never sleep indefinitely */
+ GEM_BUG_ON(!time_after(b->timeout, jiffies));
+ mod_timer(&b->hangcheck, b->timeout);
+ }
}
static void __intel_breadcrumbs_disable_irq(struct intel_breadcrumbs *b)
}
rb_link_node(&wait->node, parent, p);
rb_insert_color(&wait->node, &b->waiters);
- GEM_BUG_ON(!first && !b->irq_seqno_bh);
+ GEM_BUG_ON(!first && !rcu_access_pointer(b->irq_seqno_bh));
if (completed) {
struct rb_node *next = rb_next(completed);
GEM_BUG_ON(!next && !first);
if (next && next != &wait->node) {
GEM_BUG_ON(first);
+ b->timeout = wait_timeout();
b->first_wait = to_wait(next);
- smp_store_mb(b->irq_seqno_bh, b->first_wait->tsk);
+ rcu_assign_pointer(b->irq_seqno_bh, b->first_wait->tsk);
/* As there is a delay between reading the current
* seqno, processing the completed tasks and selecting
* the next waiter, we may have missed the interrupt
if (first) {
GEM_BUG_ON(rb_first(&b->waiters) != &wait->node);
+ b->timeout = wait_timeout();
b->first_wait = wait;
- smp_store_mb(b->irq_seqno_bh, wait->tsk);
+ rcu_assign_pointer(b->irq_seqno_bh, wait->tsk);
/* After assigning ourselves as the new bottom-half, we must
* perform a cursory check to prevent a missed interrupt.
* Either we miss the interrupt whilst programming the hardware,
*/
__intel_breadcrumbs_enable_irq(b);
}
- GEM_BUG_ON(!b->irq_seqno_bh);
+ GEM_BUG_ON(!rcu_access_pointer(b->irq_seqno_bh));
GEM_BUG_ON(!b->first_wait);
GEM_BUG_ON(rb_first(&b->waiters) != &b->first_wait->node);
return first;
}
-void intel_engine_enable_fake_irq(struct intel_engine_cs *engine)
-{
- mod_timer(&engine->breadcrumbs.fake_irq, jiffies + 1);
-}
-
static inline bool chain_wakeup(struct rb_node *rb, int priority)
{
return rb && to_wait(rb)->tsk->prio <= priority;
const int priority = wakeup_priority(b, wait->tsk);
struct rb_node *next;
- GEM_BUG_ON(b->irq_seqno_bh != wait->tsk);
+ GEM_BUG_ON(rcu_access_pointer(b->irq_seqno_bh) != wait->tsk);
/* We are the current bottom-half. Find the next candidate,
* the first waiter in the queue on the remaining oldest
* the interrupt, or if we have to handle an
* exception rather than a seqno completion.
*/
+ b->timeout = wait_timeout();
b->first_wait = to_wait(next);
- smp_store_mb(b->irq_seqno_bh, b->first_wait->tsk);
+ rcu_assign_pointer(b->irq_seqno_bh, b->first_wait->tsk);
if (b->first_wait->seqno != wait->seqno)
__intel_breadcrumbs_enable_irq(b);
- wake_up_process(b->irq_seqno_bh);
+ wake_up_process(b->first_wait->tsk);
} else {
b->first_wait = NULL;
- WRITE_ONCE(b->irq_seqno_bh, NULL);
+ rcu_assign_pointer(b->irq_seqno_bh, NULL);
__intel_breadcrumbs_disable_irq(b);
}
} else {
GEM_BUG_ON(b->first_wait == wait);
GEM_BUG_ON(rb_first(&b->waiters) !=
(b->first_wait ? &b->first_wait->node : NULL));
- GEM_BUG_ON(!b->irq_seqno_bh ^ RB_EMPTY_ROOT(&b->waiters));
+ GEM_BUG_ON(!rcu_access_pointer(b->irq_seqno_bh) ^ RB_EMPTY_ROOT(&b->waiters));
spin_unlock(&b->lock);
}
*/
intel_engine_remove_wait(engine,
&request->signaling.wait);
+
+ local_bh_disable();
fence_signal(&request->fence);
+ local_bh_enable(); /* kick start the tasklets */
/* Find the next oldest signal. Note that as we have
* not been holding the lock, another client may
setup_timer(&b->fake_irq,
intel_breadcrumbs_fake_irq,
(unsigned long)engine);
+ setup_timer(&b->hangcheck,
+ intel_breadcrumbs_hangcheck,
+ (unsigned long)engine);
/* Spawn a thread to provide a common bottom-half for all signals.
* As this is an asynchronous interface we cannot steal the current
if (!IS_ERR_OR_NULL(b->signaler))
kthread_stop(b->signaler);
+ del_timer_sync(&b->hangcheck);
del_timer_sync(&b->fake_irq);
}
* RCU lock, i.e. as we call wake_up_process() we must be holding the
* rcu_read_lock().
*/
- rcu_read_lock();
for_each_engine(engine, i915)
if (unlikely(intel_engine_wakeup(engine)))
mask |= intel_engine_flag(engine);
- rcu_read_unlock();
return mask;
}
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
int i, pipe = intel_crtc->pipe;
uint16_t coeffs[9] = { 0, };
+ struct intel_crtc_state *intel_crtc_state = to_intel_crtc_state(crtc_state);
if (crtc_state->ctm) {
struct drm_color_ctm *ctm =
(struct drm_color_ctm *)crtc_state->ctm->data;
uint64_t input[9] = { 0, };
- if (intel_crtc->config->limited_color_range) {
+ if (intel_crtc_state->limited_color_range) {
ctm_mult_by_limited(input, ctm->matrix);
} else {
for (i = 0; i < ARRAY_SIZE(input); i++)
* into consideration.
*/
for (i = 0; i < 3; i++) {
- if (intel_crtc->config->limited_color_range)
+ if (intel_crtc_state->limited_color_range)
coeffs[i * 3 + i] =
I9XX_CSC_COEFF_LIMITED_RANGE;
else
if (INTEL_INFO(dev)->gen > 6) {
uint16_t postoff = 0;
- if (intel_crtc->config->limited_color_range)
+ if (intel_crtc_state->limited_color_range)
postoff = (16 * (1 << 12) / 255) & 0x1fff;
I915_WRITE(PIPE_CSC_POSTOFF_HI(pipe), postoff);
} else {
uint32_t mode = CSC_MODE_YUV_TO_RGB;
- if (intel_crtc->config->limited_color_range)
+ if (intel_crtc_state->limited_color_range)
mode |= CSC_BLACK_SCREEN_OFFSET;
I915_WRITE(PIPE_CSC_MODE(pipe), mode);
/* Loads the legacy palette/gamma unit for the CRTC. */
static void i9xx_load_luts_internal(struct drm_crtc *crtc,
- struct drm_property_blob *blob)
+ struct drm_property_blob *blob,
+ struct intel_crtc_state *crtc_state)
{
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
int i;
if (HAS_GMCH_DISPLAY(dev)) {
- if (intel_crtc_has_type(intel_crtc->config, INTEL_OUTPUT_DSI))
+ if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI))
assert_dsi_pll_enabled(dev_priv);
else
assert_pll_enabled(dev_priv, pipe);
static void i9xx_load_luts(struct drm_crtc_state *crtc_state)
{
- i9xx_load_luts_internal(crtc_state->crtc, crtc_state->gamma_lut);
+ i9xx_load_luts_internal(crtc_state->crtc, crtc_state->gamma_lut,
+ to_intel_crtc_state(crtc_state));
}
/* Loads the legacy palette/gamma unit for the CRTC on Haswell. */
* Workaround : Do not read or write the pipe palette/gamma data while
* GAMMA_MODE is configured for split gamma and IPS_CTL has IPS enabled.
*/
- if (IS_HASWELL(dev) && intel_crtc->config->ips_enabled &&
+ if (IS_HASWELL(dev) && intel_crtc_state->ips_enabled &&
(intel_crtc_state->gamma_mode == GAMMA_MODE_MODE_SPLIT)) {
hsw_disable_ips(intel_crtc);
reenable_ips = true;
/* Turn off degamma/gamma on CGM block. */
I915_WRITE(CGM_PIPE_MODE(pipe),
(state->ctm ? CGM_PIPE_MODE_CSC : 0));
- i9xx_load_luts_internal(crtc, state->gamma_lut);
+ i9xx_load_luts_internal(crtc, state->gamma_lut,
+ to_intel_crtc_state(state));
return;
}
* Also program a linear LUT in the legacy block (behind the
* CGM block).
*/
- i9xx_load_luts_internal(crtc, NULL);
+ i9xx_load_luts_internal(crtc, NULL, to_intel_crtc_state(state));
}
void intel_color_load_luts(struct drm_crtc_state *crtc_state)
/* Note: The caller is required to filter out dpms modes not supported by the
* platform. */
-static void intel_crt_set_dpms(struct intel_encoder *encoder, int mode)
+static void intel_crt_set_dpms(struct intel_encoder *encoder,
+ struct intel_crtc_state *crtc_state,
+ int mode)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crt *crt = intel_encoder_to_crt(encoder);
- struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
- const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode;
+ struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+ const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode;
u32 adpa;
if (INTEL_INFO(dev)->gen >= 5)
I915_WRITE(crt->adpa_reg, adpa);
}
-static void intel_disable_crt(struct intel_encoder *encoder)
+static void intel_disable_crt(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
- intel_crt_set_dpms(encoder, DRM_MODE_DPMS_OFF);
+ intel_crt_set_dpms(encoder, old_crtc_state, DRM_MODE_DPMS_OFF);
}
-static void pch_disable_crt(struct intel_encoder *encoder)
+static void pch_disable_crt(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
}
-static void pch_post_disable_crt(struct intel_encoder *encoder)
+static void pch_post_disable_crt(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
- intel_disable_crt(encoder);
+ intel_disable_crt(encoder, old_crtc_state, old_conn_state);
}
-static void intel_enable_crt(struct intel_encoder *encoder)
+static void hsw_post_disable_crt(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
- intel_crt_set_dpms(encoder, DRM_MODE_DPMS_ON);
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+
+ pch_post_disable_crt(encoder, old_crtc_state, old_conn_state);
+
+ lpt_disable_pch_transcoder(dev_priv);
+ lpt_disable_iclkip(dev_priv);
+
+ intel_ddi_fdi_post_disable(encoder, old_crtc_state, old_conn_state);
+}
+
+static void intel_enable_crt(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
+{
+ intel_crt_set_dpms(encoder, pipe_config, DRM_MODE_DPMS_ON);
}
static enum drm_mode_status
}
static bool intel_crt_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
if (HAS_DDI(dev)) {
crt->base.get_config = hsw_crt_get_config;
crt->base.get_hw_state = intel_ddi_get_hw_state;
+ crt->base.post_disable = hsw_post_disable_crt;
} else {
crt->base.get_config = intel_crt_get_config;
crt->base.get_hw_state = intel_crt_get_hw_state;
* low-power state and comes back to normal.
*/
-#define I915_CSR_KBL "i915/kbl_dmc_ver1.bin"
+#define I915_CSR_KBL "i915/kbl_dmc_ver1_01.bin"
MODULE_FIRMWARE(I915_CSR_KBL);
#define KBL_CSR_VERSION_REQUIRED CSR_VERSION(1, 1)
-#define I915_CSR_SKL "i915/skl_dmc_ver1.bin"
+#define I915_CSR_SKL "i915/skl_dmc_ver1_26.bin"
MODULE_FIRMWARE(I915_CSR_SKL);
-#define SKL_CSR_VERSION_REQUIRED CSR_VERSION(1, 23)
+#define SKL_CSR_VERSION_REQUIRED CSR_VERSION(1, 26)
-#define I915_CSR_BXT "i915/bxt_dmc_ver1.bin"
+#define I915_CSR_BXT "i915/bxt_dmc_ver1_07.bin"
MODULE_FIRMWARE(I915_CSR_BXT);
#define BXT_CSR_VERSION_REQUIRED CSR_VERSION(1, 7)
DRM_ERROR("Timeout waiting for DDI BUF %c idle bit\n", port_name(port));
}
+static uint32_t hsw_pll_to_ddi_pll_sel(struct intel_shared_dpll *pll)
+{
+ switch (pll->id) {
+ case DPLL_ID_WRPLL1:
+ return PORT_CLK_SEL_WRPLL1;
+ case DPLL_ID_WRPLL2:
+ return PORT_CLK_SEL_WRPLL2;
+ case DPLL_ID_SPLL:
+ return PORT_CLK_SEL_SPLL;
+ case DPLL_ID_LCPLL_810:
+ return PORT_CLK_SEL_LCPLL_810;
+ case DPLL_ID_LCPLL_1350:
+ return PORT_CLK_SEL_LCPLL_1350;
+ case DPLL_ID_LCPLL_2700:
+ return PORT_CLK_SEL_LCPLL_2700;
+ default:
+ MISSING_CASE(pll->id);
+ return PORT_CLK_SEL_NONE;
+ }
+}
+
/* Starting with Haswell, different DDI ports can work in FDI mode for
* connection to the PCH-located connectors. For this, it is necessary to train
* both the DDI port and PCH receiver for the desired DDI buffer settings.
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
struct intel_encoder *encoder;
- u32 temp, i, rx_ctl_val;
+ u32 temp, i, rx_ctl_val, ddi_pll_sel;
for_each_encoder_on_crtc(dev, crtc, encoder) {
WARN_ON(encoder->type != INTEL_OUTPUT_ANALOG);
I915_WRITE(FDI_RX_CTL(PIPE_A), rx_ctl_val);
/* Configure Port Clock Select */
- I915_WRITE(PORT_CLK_SEL(PORT_E), intel_crtc->config->ddi_pll_sel);
- WARN_ON(intel_crtc->config->ddi_pll_sel != PORT_CLK_SEL_SPLL);
+ ddi_pll_sel = hsw_pll_to_ddi_pll_sel(intel_crtc->config->shared_dpll);
+ I915_WRITE(PORT_CLK_SEL(PORT_E), ddi_pll_sel);
+ WARN_ON(ddi_pll_sel != PORT_CLK_SEL_SPLL);
/* Start the training iterating through available voltages and emphasis,
* testing each value twice. */
int link_clock = 0;
uint32_t dpll_ctl1, dpll;
- dpll = pipe_config->ddi_pll_sel;
+ dpll = intel_get_shared_dpll_id(dev_priv, pipe_config->shared_dpll);
dpll_ctl1 = I915_READ(DPLL_CTRL1);
int link_clock = 0;
u32 val, pll;
- val = pipe_config->ddi_pll_sel;
+ val = hsw_pll_to_ddi_pll_sel(pipe_config->shared_dpll);
switch (val & PORT_CLK_SEL_MASK) {
case PORT_CLK_SEL_LCPLL_810:
link_clock = 81000;
}
void intel_ddi_clk_select(struct intel_encoder *encoder,
- const struct intel_crtc_state *pipe_config)
+ struct intel_shared_dpll *pll)
{
struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
enum port port = intel_ddi_get_encoder_port(encoder);
+ if (WARN_ON(!pll))
+ return;
+
if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) {
- uint32_t dpll = pipe_config->ddi_pll_sel;
uint32_t val;
/* DDI -> PLL mapping */
val &= ~(DPLL_CTRL2_DDI_CLK_OFF(port) |
DPLL_CTRL2_DDI_CLK_SEL_MASK(port));
- val |= (DPLL_CTRL2_DDI_CLK_SEL(dpll, port) |
+ val |= (DPLL_CTRL2_DDI_CLK_SEL(pll->id, port) |
DPLL_CTRL2_DDI_SEL_OVERRIDE(port));
I915_WRITE(DPLL_CTRL2, val);
} else if (INTEL_INFO(dev_priv)->gen < 9) {
- WARN_ON(pipe_config->ddi_pll_sel == PORT_CLK_SEL_NONE);
- I915_WRITE(PORT_CLK_SEL(port), pipe_config->ddi_pll_sel);
+ I915_WRITE(PORT_CLK_SEL(port), hsw_pll_to_ddi_pll_sel(pll));
}
}
-static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder)
+static void intel_ddi_pre_enable_dp(struct intel_encoder *encoder,
+ int link_rate, uint32_t lane_count,
+ struct intel_shared_dpll *pll,
+ bool link_mst)
{
- struct drm_encoder *encoder = &intel_encoder->base;
- struct drm_i915_private *dev_priv = to_i915(encoder->dev);
- struct intel_crtc *crtc = to_intel_crtc(encoder->crtc);
- enum port port = intel_ddi_get_encoder_port(intel_encoder);
- int type = intel_encoder->type;
-
- if (type == INTEL_OUTPUT_HDMI) {
- struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder);
-
- intel_dp_dual_mode_set_tmds_output(intel_hdmi, true);
- }
+ struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+ enum port port = intel_ddi_get_encoder_port(encoder);
- if (type == INTEL_OUTPUT_EDP) {
- struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
+ intel_dp_set_link_params(intel_dp, link_rate, lane_count,
+ link_mst);
+ if (encoder->type == INTEL_OUTPUT_EDP)
intel_edp_panel_on(intel_dp);
- }
-
- intel_ddi_clk_select(intel_encoder, crtc->config);
- if (type == INTEL_OUTPUT_DP || type == INTEL_OUTPUT_EDP) {
- struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
-
- intel_prepare_dp_ddi_buffers(intel_encoder);
-
- intel_dp_set_link_params(intel_dp, crtc->config);
+ intel_ddi_clk_select(encoder, pll);
+ intel_prepare_dp_ddi_buffers(encoder);
+ intel_ddi_init_dp_buf_reg(encoder);
+ intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
+ intel_dp_start_link_train(intel_dp);
+ if (port != PORT_A || INTEL_GEN(dev_priv) >= 9)
+ intel_dp_stop_link_train(intel_dp);
+}
- intel_ddi_init_dp_buf_reg(intel_encoder);
+static void intel_ddi_pre_enable_hdmi(struct intel_encoder *encoder,
+ bool has_hdmi_sink,
+ struct drm_display_mode *adjusted_mode,
+ struct intel_shared_dpll *pll)
+{
+ struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+ struct drm_encoder *drm_encoder = &encoder->base;
+ enum port port = intel_ddi_get_encoder_port(encoder);
+ int level = intel_ddi_hdmi_level(dev_priv, port);
- intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
- intel_dp_start_link_train(intel_dp);
- if (port != PORT_A || INTEL_INFO(dev_priv)->gen >= 9)
- intel_dp_stop_link_train(intel_dp);
- } else if (type == INTEL_OUTPUT_HDMI) {
- struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder);
- int level = intel_ddi_hdmi_level(dev_priv, port);
+ intel_dp_dual_mode_set_tmds_output(intel_hdmi, true);
+ intel_ddi_clk_select(encoder, pll);
+ intel_prepare_hdmi_ddi_buffers(encoder);
+ if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))
+ skl_ddi_set_iboost(encoder, level);
+ else if (IS_BROXTON(dev_priv))
+ bxt_ddi_vswing_sequence(dev_priv, level, port,
+ INTEL_OUTPUT_HDMI);
- intel_prepare_hdmi_ddi_buffers(intel_encoder);
+ intel_hdmi->set_infoframes(drm_encoder,
+ has_hdmi_sink,
+ adjusted_mode);
+}
- if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv))
- skl_ddi_set_iboost(intel_encoder, level);
- else if (IS_BROXTON(dev_priv))
- bxt_ddi_vswing_sequence(dev_priv, level, port,
- INTEL_OUTPUT_HDMI);
+static void intel_ddi_pre_enable(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
+{
+ struct drm_encoder *encoder = &intel_encoder->base;
+ struct intel_crtc *crtc = to_intel_crtc(encoder->crtc);
+ int type = intel_encoder->type;
- intel_hdmi->set_infoframes(encoder,
- crtc->config->has_hdmi_sink,
- &crtc->config->base.adjusted_mode);
+ if (type == INTEL_OUTPUT_DP || type == INTEL_OUTPUT_EDP) {
+ intel_ddi_pre_enable_dp(intel_encoder,
+ crtc->config->port_clock,
+ crtc->config->lane_count,
+ crtc->config->shared_dpll,
+ intel_crtc_has_type(crtc->config,
+ INTEL_OUTPUT_DP_MST));
+ }
+ if (type == INTEL_OUTPUT_HDMI) {
+ intel_ddi_pre_enable_hdmi(intel_encoder,
+ crtc->config->has_hdmi_sink,
+ &crtc->config->base.adjusted_mode,
+ crtc->config->shared_dpll);
}
}
-static void intel_ddi_post_disable(struct intel_encoder *intel_encoder)
+static void intel_ddi_post_disable(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct drm_encoder *encoder = &intel_encoder->base;
struct drm_device *dev = encoder->dev;
uint32_t val;
bool wait = false;
+ /* old_crtc_state and old_conn_state are NULL when called from DP_MST */
+
val = I915_READ(DDI_BUF_CTL(port));
if (val & DDI_BUF_CTL_ENABLE) {
val &= ~DDI_BUF_CTL_ENABLE;
}
}
-static void intel_enable_ddi(struct intel_encoder *intel_encoder)
+void intel_ddi_fdi_post_disable(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
+{
+ struct drm_i915_private *dev_priv = to_i915(intel_encoder->base.dev);
+ uint32_t val;
+
+ /*
+ * Bspec lists this as both step 13 (before DDI_BUF_CTL disable)
+ * and step 18 (after clearing PORT_CLK_SEL). Based on a BUN,
+ * step 13 is the correct place for it. Step 18 is where it was
+ * originally before the BUN.
+ */
+ val = I915_READ(FDI_RX_CTL(PIPE_A));
+ val &= ~FDI_RX_ENABLE;
+ I915_WRITE(FDI_RX_CTL(PIPE_A), val);
+
+ intel_ddi_post_disable(intel_encoder, old_crtc_state, old_conn_state);
+
+ val = I915_READ(FDI_RX_MISC(PIPE_A));
+ val &= ~(FDI_RX_PWRDN_LANE1_MASK | FDI_RX_PWRDN_LANE0_MASK);
+ val |= FDI_RX_PWRDN_LANE1_VAL(2) | FDI_RX_PWRDN_LANE0_VAL(2);
+ I915_WRITE(FDI_RX_MISC(PIPE_A), val);
+
+ val = I915_READ(FDI_RX_CTL(PIPE_A));
+ val &= ~FDI_PCDCLK;
+ I915_WRITE(FDI_RX_CTL(PIPE_A), val);
+
+ val = I915_READ(FDI_RX_CTL(PIPE_A));
+ val &= ~FDI_RX_PLL_ENABLE;
+ I915_WRITE(FDI_RX_CTL(PIPE_A), val);
+}
+
+static void intel_enable_ddi(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_encoder *encoder = &intel_encoder->base;
struct drm_crtc *crtc = encoder->crtc;
intel_edp_backlight_on(intel_dp);
intel_psr_enable(intel_dp);
- intel_edp_drrs_enable(intel_dp);
+ intel_edp_drrs_enable(intel_dp, pipe_config);
}
if (intel_crtc->config->has_audio) {
}
}
-static void intel_disable_ddi(struct intel_encoder *intel_encoder)
+static void intel_disable_ddi(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct drm_encoder *encoder = &intel_encoder->base;
struct drm_crtc *crtc = encoder->crtc;
if (type == INTEL_OUTPUT_EDP) {
struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
- intel_edp_drrs_disable(intel_dp);
+ intel_edp_drrs_disable(intel_dp, old_crtc_state);
intel_psr_disable(intel_dp);
intel_edp_backlight_off(intel_dp);
}
}
}
-static void bxt_ddi_pre_pll_enable(struct intel_encoder *encoder)
+static void bxt_ddi_pre_pll_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
struct drm_i915_private *dev_priv = to_i915(dport->base.base.dev);
udelay(600);
}
-void intel_ddi_fdi_disable(struct drm_crtc *crtc)
-{
- struct drm_i915_private *dev_priv = to_i915(crtc->dev);
- struct intel_encoder *intel_encoder = intel_ddi_get_crtc_encoder(crtc);
- uint32_t val;
-
- /*
- * Bspec lists this as both step 13 (before DDI_BUF_CTL disable)
- * and step 18 (after clearing PORT_CLK_SEL). Based on a BUN,
- * step 13 is the correct place for it. Step 18 is where it was
- * originally before the BUN.
- */
- val = I915_READ(FDI_RX_CTL(PIPE_A));
- val &= ~FDI_RX_ENABLE;
- I915_WRITE(FDI_RX_CTL(PIPE_A), val);
-
- intel_ddi_post_disable(intel_encoder);
-
- val = I915_READ(FDI_RX_MISC(PIPE_A));
- val &= ~(FDI_RX_PWRDN_LANE1_MASK | FDI_RX_PWRDN_LANE0_MASK);
- val |= FDI_RX_PWRDN_LANE1_VAL(2) | FDI_RX_PWRDN_LANE0_VAL(2);
- I915_WRITE(FDI_RX_MISC(PIPE_A), val);
-
- val = I915_READ(FDI_RX_CTL(PIPE_A));
- val &= ~FDI_PCDCLK;
- I915_WRITE(FDI_RX_CTL(PIPE_A), val);
-
- val = I915_READ(FDI_RX_CTL(PIPE_A));
- val &= ~FDI_RX_PLL_ENABLE;
- I915_WRITE(FDI_RX_CTL(PIPE_A), val);
-}
-
void intel_ddi_get_config(struct intel_encoder *encoder,
struct intel_crtc_state *pipe_config)
{
}
static bool intel_ddi_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
int type = encoder->type;
pipe_config->cpu_transcoder = TRANSCODER_EDP;
if (type == INTEL_OUTPUT_HDMI)
- ret = intel_hdmi_compute_config(encoder, pipe_config);
+ ret = intel_hdmi_compute_config(encoder, pipe_config, conn_state);
else
- ret = intel_dp_compute_config(encoder, pipe_config);
+ ret = intel_dp_compute_config(encoder, pipe_config, conn_state);
if (IS_BROXTON(dev_priv) && ret)
pipe_config->lane_lat_optim_mask =
return connector;
}
+struct intel_shared_dpll *
+intel_ddi_get_link_dpll(struct intel_dp *intel_dp, int clock)
+{
+ struct intel_connector *connector = intel_dp->attached_connector;
+ struct intel_encoder *encoder = connector->encoder;
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+ struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
+ struct intel_shared_dpll *pll = NULL;
+ struct intel_shared_dpll_config tmp_pll_config;
+ enum intel_dpll_id dpll_id;
+
+ if (IS_BROXTON(dev_priv)) {
+ dpll_id = (enum intel_dpll_id)dig_port->port;
+ /*
+ * Select the required PLL. This works for platforms where
+ * there is no shared DPLL.
+ */
+ pll = &dev_priv->shared_dplls[dpll_id];
+ if (WARN_ON(pll->active_mask)) {
+
+ DRM_ERROR("Shared DPLL in use. active_mask:%x\n",
+ pll->active_mask);
+ return NULL;
+ }
+ tmp_pll_config = pll->config;
+ if (!bxt_ddi_dp_set_dpll_hw_state(clock,
+ &pll->config.hw_state)) {
+ DRM_ERROR("Could not setup DPLL\n");
+ pll->config = tmp_pll_config;
+ return NULL;
+ }
+ } else if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) {
+ pll = skl_find_link_pll(dev_priv, clock);
+ } else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) {
+ pll = hsw_ddi_dp_get_dpll(encoder, clock);
+ }
+ return pll;
+}
+
void intel_ddi_init(struct drm_device *dev, enum port port)
{
struct drm_i915_private *dev_priv = to_i915(dev);
static void cherryview_sseu_info_init(struct drm_i915_private *dev_priv)
{
- struct intel_device_info *info = mkwrite_device_info(dev_priv);
+ struct sseu_dev_info *sseu = &mkwrite_device_info(dev_priv)->sseu;
u32 fuse, eu_dis;
fuse = I915_READ(CHV_FUSE_GT);
- info->slice_total = 1;
+ sseu->slice_mask = BIT(0);
if (!(fuse & CHV_FGT_DISABLE_SS0)) {
- info->subslice_per_slice++;
+ sseu->subslice_mask |= BIT(0);
eu_dis = fuse & (CHV_FGT_EU_DIS_SS0_R0_MASK |
CHV_FGT_EU_DIS_SS0_R1_MASK);
- info->eu_total += 8 - hweight32(eu_dis);
+ sseu->eu_total += 8 - hweight32(eu_dis);
}
if (!(fuse & CHV_FGT_DISABLE_SS1)) {
- info->subslice_per_slice++;
+ sseu->subslice_mask |= BIT(1);
eu_dis = fuse & (CHV_FGT_EU_DIS_SS1_R0_MASK |
CHV_FGT_EU_DIS_SS1_R1_MASK);
- info->eu_total += 8 - hweight32(eu_dis);
+ sseu->eu_total += 8 - hweight32(eu_dis);
}
- info->subslice_total = info->subslice_per_slice;
/*
* CHV expected to always have a uniform distribution of EU
* across subslices.
*/
- info->eu_per_subslice = info->subslice_total ?
- info->eu_total / info->subslice_total :
+ sseu->eu_per_subslice = sseu_subslice_total(sseu) ?
+ sseu->eu_total / sseu_subslice_total(sseu) :
0;
/*
* CHV supports subslice power gating on devices with more than
* one subslice, and supports EU power gating on devices with
* more than one EU pair per subslice.
*/
- info->has_slice_pg = 0;
- info->has_subslice_pg = (info->subslice_total > 1);
- info->has_eu_pg = (info->eu_per_subslice > 2);
+ sseu->has_slice_pg = 0;
+ sseu->has_subslice_pg = sseu_subslice_total(sseu) > 1;
+ sseu->has_eu_pg = (sseu->eu_per_subslice > 2);
}
static void gen9_sseu_info_init(struct drm_i915_private *dev_priv)
{
struct intel_device_info *info = mkwrite_device_info(dev_priv);
+ struct sseu_dev_info *sseu = &info->sseu;
int s_max = 3, ss_max = 4, eu_max = 8;
int s, ss;
- u32 fuse2, s_enable, ss_disable, eu_disable;
+ u32 fuse2, eu_disable;
u8 eu_mask = 0xff;
fuse2 = I915_READ(GEN8_FUSE2);
- s_enable = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT;
- ss_disable = (fuse2 & GEN9_F2_SS_DIS_MASK) >> GEN9_F2_SS_DIS_SHIFT;
+ sseu->slice_mask = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT;
- info->slice_total = hweight32(s_enable);
/*
* The subslice disable field is global, i.e. it applies
* to each of the enabled slices.
*/
- info->subslice_per_slice = ss_max - hweight32(ss_disable);
- info->subslice_total = info->slice_total * info->subslice_per_slice;
+ sseu->subslice_mask = (1 << ss_max) - 1;
+ sseu->subslice_mask &= ~((fuse2 & GEN9_F2_SS_DIS_MASK) >>
+ GEN9_F2_SS_DIS_SHIFT);
/*
* Iterate through enabled slices and subslices to
* count the total enabled EU.
*/
for (s = 0; s < s_max; s++) {
- if (!(s_enable & BIT(s)))
+ if (!(sseu->slice_mask & BIT(s)))
/* skip disabled slice */
continue;
for (ss = 0; ss < ss_max; ss++) {
int eu_per_ss;
- if (ss_disable & BIT(ss))
+ if (!(sseu->subslice_mask & BIT(ss)))
/* skip disabled subslice */
continue;
* subslices if they are unbalanced.
*/
if (eu_per_ss == 7)
- info->subslice_7eu[s] |= BIT(ss);
+ sseu->subslice_7eu[s] |= BIT(ss);
- info->eu_total += eu_per_ss;
+ sseu->eu_total += eu_per_ss;
}
}
* recovery. BXT is expected to be perfectly uniform in EU
* distribution.
*/
- info->eu_per_subslice = info->subslice_total ?
- DIV_ROUND_UP(info->eu_total,
- info->subslice_total) : 0;
+ sseu->eu_per_subslice = sseu_subslice_total(sseu) ?
+ DIV_ROUND_UP(sseu->eu_total,
+ sseu_subslice_total(sseu)) : 0;
/*
* SKL supports slice power gating on devices with more than
* one slice, and supports EU power gating on devices with
* supports EU power gating on devices with more than one EU
* pair per subslice.
*/
- info->has_slice_pg =
+ sseu->has_slice_pg =
(IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) &&
- info->slice_total > 1;
- info->has_subslice_pg =
- IS_BROXTON(dev_priv) && info->subslice_total > 1;
- info->has_eu_pg = info->eu_per_subslice > 2;
+ hweight8(sseu->slice_mask) > 1;
+ sseu->has_subslice_pg =
+ IS_BROXTON(dev_priv) && sseu_subslice_total(sseu) > 1;
+ sseu->has_eu_pg = sseu->eu_per_subslice > 2;
if (IS_BROXTON(dev_priv)) {
-#define IS_SS_DISABLED(_ss_disable, ss) (_ss_disable & BIT(ss))
+#define IS_SS_DISABLED(ss) (!(sseu->subslice_mask & BIT(ss)))
/*
* There is a HW issue in 2x6 fused down parts that requires
* Pooled EU to be enabled as a WA. The pool configuration
* doesn't affect if the device has all 3 subslices enabled.
*/
/* WaEnablePooledEuFor2x6:bxt */
- info->has_pooled_eu = ((info->subslice_per_slice == 3) ||
- (info->subslice_per_slice == 2 &&
+ info->has_pooled_eu = ((hweight8(sseu->subslice_mask) == 3) ||
+ (hweight8(sseu->subslice_mask) == 2 &&
INTEL_REVID(dev_priv) < BXT_REVID_C0));
- info->min_eu_in_pool = 0;
+ sseu->min_eu_in_pool = 0;
if (info->has_pooled_eu) {
- if (IS_SS_DISABLED(ss_disable, 0) ||
- IS_SS_DISABLED(ss_disable, 2))
- info->min_eu_in_pool = 3;
- else if (IS_SS_DISABLED(ss_disable, 1))
- info->min_eu_in_pool = 6;
+ if (IS_SS_DISABLED(2) || IS_SS_DISABLED(0))
+ sseu->min_eu_in_pool = 3;
+ else if (IS_SS_DISABLED(1))
+ sseu->min_eu_in_pool = 6;
else
- info->min_eu_in_pool = 9;
+ sseu->min_eu_in_pool = 9;
}
#undef IS_SS_DISABLED
}
static void broadwell_sseu_info_init(struct drm_i915_private *dev_priv)
{
- struct intel_device_info *info = mkwrite_device_info(dev_priv);
+ struct sseu_dev_info *sseu = &mkwrite_device_info(dev_priv)->sseu;
const int s_max = 3, ss_max = 3, eu_max = 8;
int s, ss;
- u32 fuse2, eu_disable[s_max], s_enable, ss_disable;
+ u32 fuse2, eu_disable[s_max];
fuse2 = I915_READ(GEN8_FUSE2);
- s_enable = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT;
- ss_disable = (fuse2 & GEN8_F2_SS_DIS_MASK) >> GEN8_F2_SS_DIS_SHIFT;
+ sseu->slice_mask = (fuse2 & GEN8_F2_S_ENA_MASK) >> GEN8_F2_S_ENA_SHIFT;
+ /*
+ * The subslice disable field is global, i.e. it applies
+ * to each of the enabled slices.
+ */
+ sseu->subslice_mask = BIT(ss_max) - 1;
+ sseu->subslice_mask &= ~((fuse2 & GEN8_F2_SS_DIS_MASK) >>
+ GEN8_F2_SS_DIS_SHIFT);
eu_disable[0] = I915_READ(GEN8_EU_DISABLE0) & GEN8_EU_DIS0_S0_MASK;
eu_disable[1] = (I915_READ(GEN8_EU_DISABLE0) >> GEN8_EU_DIS0_S1_SHIFT) |
((I915_READ(GEN8_EU_DISABLE2) & GEN8_EU_DIS2_S2_MASK) <<
(32 - GEN8_EU_DIS1_S2_SHIFT));
- info->slice_total = hweight32(s_enable);
-
- /*
- * The subslice disable field is global, i.e. it applies
- * to each of the enabled slices.
- */
- info->subslice_per_slice = ss_max - hweight32(ss_disable);
- info->subslice_total = info->slice_total * info->subslice_per_slice;
-
/*
* Iterate through enabled slices and subslices to
* count the total enabled EU.
*/
for (s = 0; s < s_max; s++) {
- if (!(s_enable & (0x1 << s)))
+ if (!(sseu->slice_mask & BIT(s)))
/* skip disabled slice */
continue;
for (ss = 0; ss < ss_max; ss++) {
u32 n_disabled;
- if (ss_disable & (0x1 << ss))
+ if (!(sseu->subslice_mask & BIT(ss)))
/* skip disabled subslice */
continue;
* Record which subslices have 7 EUs.
*/
if (eu_max - n_disabled == 7)
- info->subslice_7eu[s] |= 1 << ss;
+ sseu->subslice_7eu[s] |= 1 << ss;
- info->eu_total += eu_max - n_disabled;
+ sseu->eu_total += eu_max - n_disabled;
}
}
* subslices with the exception that any one EU in any one subslice may
* be fused off for die recovery.
*/
- info->eu_per_subslice = info->subslice_total ?
- DIV_ROUND_UP(info->eu_total, info->subslice_total) : 0;
+ sseu->eu_per_subslice = sseu_subslice_total(sseu) ?
+ DIV_ROUND_UP(sseu->eu_total,
+ sseu_subslice_total(sseu)) : 0;
/*
* BDW supports slice power gating on devices with more than
* one slice.
*/
- info->has_slice_pg = (info->slice_total > 1);
- info->has_subslice_pg = 0;
- info->has_eu_pg = 0;
+ sseu->has_slice_pg = hweight8(sseu->slice_mask) > 1;
+ sseu->has_subslice_pg = 0;
+ sseu->has_eu_pg = 0;
}
/*
if (IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1))
info->has_snoop = false;
- DRM_DEBUG_DRIVER("slice total: %u\n", info->slice_total);
- DRM_DEBUG_DRIVER("subslice total: %u\n", info->subslice_total);
- DRM_DEBUG_DRIVER("subslice per slice: %u\n", info->subslice_per_slice);
- DRM_DEBUG_DRIVER("EU total: %u\n", info->eu_total);
- DRM_DEBUG_DRIVER("EU per subslice: %u\n", info->eu_per_subslice);
+ DRM_DEBUG_DRIVER("slice mask: %04x\n", info->sseu.slice_mask);
+ DRM_DEBUG_DRIVER("slice total: %u\n", hweight8(info->sseu.slice_mask));
+ DRM_DEBUG_DRIVER("subslice total: %u\n",
+ sseu_subslice_total(&info->sseu));
+ DRM_DEBUG_DRIVER("subslice mask %04x\n", info->sseu.subslice_mask);
+ DRM_DEBUG_DRIVER("subslice per slice: %u\n",
+ hweight8(info->sseu.subslice_mask));
+ DRM_DEBUG_DRIVER("EU total: %u\n", info->sseu.eu_total);
+ DRM_DEBUG_DRIVER("EU per subslice: %u\n", info->sseu.eu_per_subslice);
DRM_DEBUG_DRIVER("has slice power gating: %s\n",
- info->has_slice_pg ? "y" : "n");
+ info->sseu.has_slice_pg ? "y" : "n");
DRM_DEBUG_DRIVER("has subslice power gating: %s\n",
- info->has_subslice_pg ? "y" : "n");
+ info->sseu.has_subslice_pg ? "y" : "n");
DRM_DEBUG_DRIVER("has EU power gating: %s\n",
- info->has_eu_pg ? "y" : "n");
+ info->sseu.has_eu_pg ? "y" : "n");
}
if (HAS_PCH_SPLIT(dev)) {
u32 port_sel;
- pp_reg = PCH_PP_CONTROL;
- port_sel = I915_READ(PCH_PP_ON_DELAYS) & PANEL_PORT_SELECT_MASK;
+ pp_reg = PP_CONTROL(0);
+ port_sel = I915_READ(PP_ON_DELAYS(0)) & PANEL_PORT_SELECT_MASK;
if (port_sel == PANEL_PORT_SELECT_LVDS &&
I915_READ(PCH_LVDS) & LVDS_PIPEB_SELECT)
/* XXX: else fix for eDP */
} else if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) {
/* presumably write lock depends on pipe, not port select */
- pp_reg = VLV_PIPE_PP_CONTROL(pipe);
+ pp_reg = PP_CONTROL(pipe);
panel_pipe = pipe;
} else {
- pp_reg = PP_CONTROL;
+ pp_reg = PP_CONTROL(0);
if (I915_READ(LVDS) & LVDS_PIPEB_SELECT)
panel_pipe = PIPE_B;
}
}
}
-static void lpt_disable_pch_transcoder(struct drm_i915_private *dev_priv)
+void lpt_disable_pch_transcoder(struct drm_i915_private *dev_priv)
{
u32 val;
* a plane. On ILK+ the pipe PLLs are integrated, so we don't
* need the check.
*/
- if (HAS_GMCH_DISPLAY(dev_priv))
+ if (HAS_GMCH_DISPLAY(dev_priv)) {
if (intel_crtc_has_type(crtc->config, INTEL_OUTPUT_DSI))
assert_dsi_pll_enabled(dev_priv);
else
assert_pll_enabled(dev_priv, pipe);
- else {
+ } else {
if (crtc->config->has_pch_encoder) {
/* if driving the PCH, we need FDI enabled */
assert_fdi_rx_pll_enabled(dev_priv, pch_transcoder);
}
}
-static void
-intel_fill_fb_info(struct drm_i915_private *dev_priv,
- struct drm_framebuffer *fb)
-{
- struct intel_rotation_info *info = &to_intel_framebuffer(fb)->rot_info;
- unsigned int tile_size, tile_width, tile_height, cpp;
-
- tile_size = intel_tile_size(dev_priv);
-
- cpp = drm_format_plane_cpp(fb->pixel_format, 0);
- intel_tile_dims(dev_priv, &tile_width, &tile_height,
- fb->modifier[0], cpp);
-
- info->plane[0].width = DIV_ROUND_UP(fb->pitches[0], tile_width * cpp);
- info->plane[0].height = DIV_ROUND_UP(fb->height, tile_height);
-
- if (info->pixel_format == DRM_FORMAT_NV12) {
- cpp = drm_format_plane_cpp(fb->pixel_format, 1);
- intel_tile_dims(dev_priv, &tile_width, &tile_height,
- fb->modifier[1], cpp);
-
- info->uv_offset = fb->offsets[1];
- info->plane[1].width = DIV_ROUND_UP(fb->pitches[1], tile_width * cpp);
- info->plane[1].height = DIV_ROUND_UP(fb->height / 2, tile_height);
- }
-}
-
static unsigned int intel_linear_alignment(const struct drm_i915_private *dev_priv)
{
if (INTEL_INFO(dev_priv)->gen >= 9)
}
}
-int
-intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb,
- unsigned int rotation)
+struct i915_vma *
+intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb, unsigned int rotation)
{
struct drm_device *dev = fb->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct drm_i915_gem_object *obj = intel_fb_obj(fb);
struct i915_ggtt_view view;
+ struct i915_vma *vma;
u32 alignment;
- int ret;
WARN_ON(!mutex_is_locked(&dev->struct_mutex));
*/
intel_runtime_pm_get(dev_priv);
- ret = i915_gem_object_pin_to_display_plane(obj, alignment,
- &view);
- if (ret)
- goto err_pm;
-
- /* Install a fence for tiled scan-out. Pre-i965 always needs a
- * fence, whereas 965+ only requires a fence if using
- * framebuffer compression. For simplicity, we always install
- * a fence as the cost is not that onerous.
- */
- if (view.type == I915_GGTT_VIEW_NORMAL) {
- ret = i915_gem_object_get_fence(obj);
- if (ret == -EDEADLK) {
- /*
- * -EDEADLK means there are no free fences
- * no pending flips.
- *
- * This is propagated to atomic, but it uses
- * -EDEADLK to force a locking recovery, so
- * change the returned error to -EBUSY.
- */
- ret = -EBUSY;
- goto err_unpin;
- } else if (ret)
- goto err_unpin;
+ vma = i915_gem_object_pin_to_display_plane(obj, alignment, &view);
+ if (IS_ERR(vma))
+ goto err;
- i915_gem_object_pin_fence(obj);
+ if (i915_vma_is_map_and_fenceable(vma)) {
+ /* Install a fence for tiled scan-out. Pre-i965 always needs a
+ * fence, whereas 965+ only requires a fence if using
+ * framebuffer compression. For simplicity, we always, when
+ * possible, install a fence as the cost is not that onerous.
+ *
+ * If we fail to fence the tiled scanout, then either the
+ * modeset will reject the change (which is highly unlikely as
+ * the affected systems, all but one, do not have unmappable
+ * space) or we will not be able to enable full powersaving
+ * techniques (also likely not to apply due to various limits
+ * FBC and the like impose on the size of the buffer, which
+ * presumably we violated anyway with this unmappable buffer).
+ * Anyway, it is presumably better to stumble onwards with
+ * something and try to run the system in a "less than optimal"
+ * mode that matches the user configuration.
+ */
+ if (i915_vma_get_fence(vma) == 0)
+ i915_vma_pin_fence(vma);
}
+err:
intel_runtime_pm_put(dev_priv);
- return 0;
-
-err_unpin:
- i915_gem_object_unpin_from_display_plane(obj, &view);
-err_pm:
- intel_runtime_pm_put(dev_priv);
- return ret;
+ return vma;
}
void intel_unpin_fb_obj(struct drm_framebuffer *fb, unsigned int rotation)
{
struct drm_i915_gem_object *obj = intel_fb_obj(fb);
struct i915_ggtt_view view;
+ struct i915_vma *vma;
WARN_ON(!mutex_is_locked(&obj->base.dev->struct_mutex));
intel_fill_fb_ggtt_view(&view, fb, rotation);
+ vma = i915_gem_object_to_ggtt(obj, &view);
- if (view.type == I915_GGTT_VIEW_NORMAL)
- i915_gem_object_unpin_fence(obj);
+ i915_vma_unpin_fence(vma);
+ i915_gem_object_unpin_from_display_plane(vma);
+}
- i915_gem_object_unpin_from_display_plane(obj, &view);
+static int intel_fb_pitch(const struct drm_framebuffer *fb, int plane,
+ unsigned int rotation)
+{
+ if (intel_rotation_90_or_270(rotation))
+ return to_intel_framebuffer(fb)->rotated[plane].pitch;
+ else
+ return fb->pitches[plane];
+}
+
+/*
+ * Convert the x/y offsets into a linear offset.
+ * Only valid with 0/180 degree rotation, which is fine since linear
+ * offset is only used with linear buffers on pre-hsw and tiled buffers
+ * with gen2/3, and 90/270 degree rotations isn't supported on any of them.
+ */
+u32 intel_fb_xy_to_linear(int x, int y,
+ const struct intel_plane_state *state,
+ int plane)
+{
+ const struct drm_framebuffer *fb = state->base.fb;
+ unsigned int cpp = drm_format_plane_cpp(fb->pixel_format, plane);
+ unsigned int pitch = fb->pitches[plane];
+
+ return y * pitch + x * cpp;
+}
+
+/*
+ * Add the x/y offsets derived from fb->offsets[] to the user
+ * specified plane src x/y offsets. The resulting x/y offsets
+ * specify the start of scanout from the beginning of the gtt mapping.
+ */
+void intel_add_fb_offsets(int *x, int *y,
+ const struct intel_plane_state *state,
+ int plane)
+
+{
+ const struct intel_framebuffer *intel_fb = to_intel_framebuffer(state->base.fb);
+ unsigned int rotation = state->base.rotation;
+
+ if (intel_rotation_90_or_270(rotation)) {
+ *x += intel_fb->rotated[plane].x;
+ *y += intel_fb->rotated[plane].y;
+ } else {
+ *x += intel_fb->normal[plane].x;
+ *y += intel_fb->normal[plane].y;
+ }
}
/*
- * Adjust the tile offset by moving the difference into
- * the x/y offsets.
- *
* Input tile dimensions and pitch must already be
* rotated to match x and y, and in pixel units.
*/
-static u32 intel_adjust_tile_offset(int *x, int *y,
- unsigned int tile_width,
- unsigned int tile_height,
- unsigned int tile_size,
- unsigned int pitch_tiles,
- u32 old_offset,
- u32 new_offset)
-{
+static u32 _intel_adjust_tile_offset(int *x, int *y,
+ unsigned int tile_width,
+ unsigned int tile_height,
+ unsigned int tile_size,
+ unsigned int pitch_tiles,
+ u32 old_offset,
+ u32 new_offset)
+{
+ unsigned int pitch_pixels = pitch_tiles * tile_width;
unsigned int tiles;
WARN_ON(old_offset & (tile_size - 1));
*y += tiles / pitch_tiles * tile_height;
*x += tiles % pitch_tiles * tile_width;
+ /* minimize x in case it got needlessly big */
+ *y += *x / pitch_pixels * tile_height;
+ *x %= pitch_pixels;
+
+ return new_offset;
+}
+
+/*
+ * Adjust the tile offset by moving the difference into
+ * the x/y offsets.
+ */
+static u32 intel_adjust_tile_offset(int *x, int *y,
+ const struct intel_plane_state *state, int plane,
+ u32 old_offset, u32 new_offset)
+{
+ const struct drm_i915_private *dev_priv = to_i915(state->base.plane->dev);
+ const struct drm_framebuffer *fb = state->base.fb;
+ unsigned int cpp = drm_format_plane_cpp(fb->pixel_format, plane);
+ unsigned int rotation = state->base.rotation;
+ unsigned int pitch = intel_fb_pitch(fb, plane, rotation);
+
+ WARN_ON(new_offset > old_offset);
+
+ if (fb->modifier[plane] != DRM_FORMAT_MOD_NONE) {
+ unsigned int tile_size, tile_width, tile_height;
+ unsigned int pitch_tiles;
+
+ tile_size = intel_tile_size(dev_priv);
+ intel_tile_dims(dev_priv, &tile_width, &tile_height,
+ fb->modifier[plane], cpp);
+
+ if (intel_rotation_90_or_270(rotation)) {
+ pitch_tiles = pitch / tile_height;
+ swap(tile_width, tile_height);
+ } else {
+ pitch_tiles = pitch / (tile_width * cpp);
+ }
+
+ _intel_adjust_tile_offset(x, y, tile_width, tile_height,
+ tile_size, pitch_tiles,
+ old_offset, new_offset);
+ } else {
+ old_offset += *y * pitch + *x * cpp;
+
+ *y = (old_offset - new_offset) / pitch;
+ *x = ((old_offset - new_offset) - *y * pitch) / cpp;
+ }
+
return new_offset;
}
* In the 90/270 rotated case, x and y are assumed
* to be already rotated to match the rotated GTT view, and
* pitch is the tile_height aligned framebuffer height.
+ *
+ * This function is used when computing the derived information
+ * under intel_framebuffer, so using any of that information
+ * here is not allowed. Anything under drm_framebuffer can be
+ * used. This is why the user has to pass in the pitch since it
+ * is specified in the rotated orientation.
*/
-u32 intel_compute_tile_offset(int *x, int *y,
- const struct drm_framebuffer *fb, int plane,
- unsigned int pitch,
- unsigned int rotation)
+static u32 _intel_compute_tile_offset(const struct drm_i915_private *dev_priv,
+ int *x, int *y,
+ const struct drm_framebuffer *fb, int plane,
+ unsigned int pitch,
+ unsigned int rotation,
+ u32 alignment)
{
- const struct drm_i915_private *dev_priv = to_i915(fb->dev);
uint64_t fb_modifier = fb->modifier[plane];
unsigned int cpp = drm_format_plane_cpp(fb->pixel_format, plane);
- u32 offset, offset_aligned, alignment;
+ u32 offset, offset_aligned;
- alignment = intel_surf_alignment(dev_priv, fb_modifier);
if (alignment)
alignment--;
offset = (tile_rows * pitch_tiles + tiles) * tile_size;
offset_aligned = offset & ~alignment;
- intel_adjust_tile_offset(x, y, tile_width, tile_height,
- tile_size, pitch_tiles,
- offset, offset_aligned);
+ _intel_adjust_tile_offset(x, y, tile_width, tile_height,
+ tile_size, pitch_tiles,
+ offset, offset_aligned);
} else {
offset = *y * pitch + *x * cpp;
offset_aligned = offset & ~alignment;
return offset_aligned;
}
+u32 intel_compute_tile_offset(int *x, int *y,
+ const struct intel_plane_state *state,
+ int plane)
+{
+ const struct drm_i915_private *dev_priv = to_i915(state->base.plane->dev);
+ const struct drm_framebuffer *fb = state->base.fb;
+ unsigned int rotation = state->base.rotation;
+ int pitch = intel_fb_pitch(fb, plane, rotation);
+ u32 alignment;
+
+ /* AUX_DIST needs only 4K alignment */
+ if (fb->pixel_format == DRM_FORMAT_NV12 && plane == 1)
+ alignment = 4096;
+ else
+ alignment = intel_surf_alignment(dev_priv, fb->modifier[plane]);
+
+ return _intel_compute_tile_offset(dev_priv, x, y, fb, plane, pitch,
+ rotation, alignment);
+}
+
+/* Convert the fb->offset[] linear offset into x/y offsets */
+static void intel_fb_offset_to_xy(int *x, int *y,
+ const struct drm_framebuffer *fb, int plane)
+{
+ unsigned int cpp = drm_format_plane_cpp(fb->pixel_format, plane);
+ unsigned int pitch = fb->pitches[plane];
+ u32 linear_offset = fb->offsets[plane];
+
+ *y = linear_offset / pitch;
+ *x = linear_offset % pitch / cpp;
+}
+
+static unsigned int intel_fb_modifier_to_tiling(uint64_t fb_modifier)
+{
+ switch (fb_modifier) {
+ case I915_FORMAT_MOD_X_TILED:
+ return I915_TILING_X;
+ case I915_FORMAT_MOD_Y_TILED:
+ return I915_TILING_Y;
+ default:
+ return I915_TILING_NONE;
+ }
+}
+
+static int
+intel_fill_fb_info(struct drm_i915_private *dev_priv,
+ struct drm_framebuffer *fb)
+{
+ struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
+ struct intel_rotation_info *rot_info = &intel_fb->rot_info;
+ u32 gtt_offset_rotated = 0;
+ unsigned int max_size = 0;
+ uint32_t format = fb->pixel_format;
+ int i, num_planes = drm_format_num_planes(format);
+ unsigned int tile_size = intel_tile_size(dev_priv);
+
+ for (i = 0; i < num_planes; i++) {
+ unsigned int width, height;
+ unsigned int cpp, size;
+ u32 offset;
+ int x, y;
+
+ cpp = drm_format_plane_cpp(format, i);
+ width = drm_format_plane_width(fb->width, format, i);
+ height = drm_format_plane_height(fb->height, format, i);
+
+ intel_fb_offset_to_xy(&x, &y, fb, i);
+
+ /*
+ * The fence (if used) is aligned to the start of the object
+ * so having the framebuffer wrap around across the edge of the
+ * fenced region doesn't really work. We have no API to configure
+ * the fence start offset within the object (nor could we probably
+ * on gen2/3). So it's just easier if we just require that the
+ * fb layout agrees with the fence layout. We already check that the
+ * fb stride matches the fence stride elsewhere.
+ */
+ if (i915_gem_object_is_tiled(intel_fb->obj) &&
+ (x + width) * cpp > fb->pitches[i]) {
+ DRM_DEBUG("bad fb plane %d offset: 0x%x\n",
+ i, fb->offsets[i]);
+ return -EINVAL;
+ }
+
+ /*
+ * First pixel of the framebuffer from
+ * the start of the normal gtt mapping.
+ */
+ intel_fb->normal[i].x = x;
+ intel_fb->normal[i].y = y;
+
+ offset = _intel_compute_tile_offset(dev_priv, &x, &y,
+ fb, 0, fb->pitches[i],
+ DRM_ROTATE_0, tile_size);
+ offset /= tile_size;
+
+ if (fb->modifier[i] != DRM_FORMAT_MOD_NONE) {
+ unsigned int tile_width, tile_height;
+ unsigned int pitch_tiles;
+ struct drm_rect r;
+
+ intel_tile_dims(dev_priv, &tile_width, &tile_height,
+ fb->modifier[i], cpp);
+
+ rot_info->plane[i].offset = offset;
+ rot_info->plane[i].stride = DIV_ROUND_UP(fb->pitches[i], tile_width * cpp);
+ rot_info->plane[i].width = DIV_ROUND_UP(x + width, tile_width);
+ rot_info->plane[i].height = DIV_ROUND_UP(y + height, tile_height);
+
+ intel_fb->rotated[i].pitch =
+ rot_info->plane[i].height * tile_height;
+
+ /* how many tiles does this plane need */
+ size = rot_info->plane[i].stride * rot_info->plane[i].height;
+ /*
+ * If the plane isn't horizontally tile aligned,
+ * we need one more tile.
+ */
+ if (x != 0)
+ size++;
+
+ /* rotate the x/y offsets to match the GTT view */
+ r.x1 = x;
+ r.y1 = y;
+ r.x2 = x + width;
+ r.y2 = y + height;
+ drm_rect_rotate(&r,
+ rot_info->plane[i].width * tile_width,
+ rot_info->plane[i].height * tile_height,
+ DRM_ROTATE_270);
+ x = r.x1;
+ y = r.y1;
+
+ /* rotate the tile dimensions to match the GTT view */
+ pitch_tiles = intel_fb->rotated[i].pitch / tile_height;
+ swap(tile_width, tile_height);
+
+ /*
+ * We only keep the x/y offsets, so push all of the
+ * gtt offset into the x/y offsets.
+ */
+ _intel_adjust_tile_offset(&x, &y, tile_size,
+ tile_width, tile_height, pitch_tiles,
+ gtt_offset_rotated * tile_size, 0);
+
+ gtt_offset_rotated += rot_info->plane[i].width * rot_info->plane[i].height;
+
+ /*
+ * First pixel of the framebuffer from
+ * the start of the rotated gtt mapping.
+ */
+ intel_fb->rotated[i].x = x;
+ intel_fb->rotated[i].y = y;
+ } else {
+ size = DIV_ROUND_UP((y + height) * fb->pitches[i] +
+ x * cpp, tile_size);
+ }
+
+ /* how many tiles in total needed in the bo */
+ max_size = max(max_size, offset + size);
+ }
+
+ if (max_size * tile_size > to_intel_framebuffer(fb)->obj->base.size) {
+ DRM_DEBUG("fb too big for bo (need %u bytes, have %zu bytes)\n",
+ max_size * tile_size, to_intel_framebuffer(fb)->obj->base.size);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
static int i9xx_format_to_fourcc(int format)
{
switch (format) {
continue;
obj = intel_fb_obj(fb);
- if (i915_gem_obj_ggtt_offset(obj) == plane_config->base) {
+ if (i915_gem_object_ggtt_offset(obj, NULL) == plane_config->base) {
drm_framebuffer_reference(fb);
goto valid_fb;
}
&obj->frontbuffer_bits);
}
+static int skl_max_plane_width(const struct drm_framebuffer *fb, int plane,
+ unsigned int rotation)
+{
+ int cpp = drm_format_plane_cpp(fb->pixel_format, plane);
+
+ switch (fb->modifier[plane]) {
+ case DRM_FORMAT_MOD_NONE:
+ case I915_FORMAT_MOD_X_TILED:
+ switch (cpp) {
+ case 8:
+ return 4096;
+ case 4:
+ case 2:
+ case 1:
+ return 8192;
+ default:
+ MISSING_CASE(cpp);
+ break;
+ }
+ break;
+ case I915_FORMAT_MOD_Y_TILED:
+ case I915_FORMAT_MOD_Yf_TILED:
+ switch (cpp) {
+ case 8:
+ return 2048;
+ case 4:
+ return 4096;
+ case 2:
+ case 1:
+ return 8192;
+ default:
+ MISSING_CASE(cpp);
+ break;
+ }
+ break;
+ default:
+ MISSING_CASE(fb->modifier[plane]);
+ }
+
+ return 2048;
+}
+
+static int skl_check_main_surface(struct intel_plane_state *plane_state)
+{
+ const struct drm_i915_private *dev_priv = to_i915(plane_state->base.plane->dev);
+ const struct drm_framebuffer *fb = plane_state->base.fb;
+ unsigned int rotation = plane_state->base.rotation;
+ int x = plane_state->base.src.x1 >> 16;
+ int y = plane_state->base.src.y1 >> 16;
+ int w = drm_rect_width(&plane_state->base.src) >> 16;
+ int h = drm_rect_height(&plane_state->base.src) >> 16;
+ int max_width = skl_max_plane_width(fb, 0, rotation);
+ int max_height = 4096;
+ u32 alignment, offset, aux_offset = plane_state->aux.offset;
+
+ if (w > max_width || h > max_height) {
+ DRM_DEBUG_KMS("requested Y/RGB source size %dx%d too big (limit %dx%d)\n",
+ w, h, max_width, max_height);
+ return -EINVAL;
+ }
+
+ intel_add_fb_offsets(&x, &y, plane_state, 0);
+ offset = intel_compute_tile_offset(&x, &y, plane_state, 0);
+
+ alignment = intel_surf_alignment(dev_priv, fb->modifier[0]);
+
+ /*
+ * AUX surface offset is specified as the distance from the
+ * main surface offset, and it must be non-negative. Make
+ * sure that is what we will get.
+ */
+ if (offset > aux_offset)
+ offset = intel_adjust_tile_offset(&x, &y, plane_state, 0,
+ offset, aux_offset & ~(alignment - 1));
+
+ /*
+ * When using an X-tiled surface, the plane blows up
+ * if the x offset + width exceed the stride.
+ *
+ * TODO: linear and Y-tiled seem fine, Yf untested,
+ */
+ if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED) {
+ int cpp = drm_format_plane_cpp(fb->pixel_format, 0);
+
+ while ((x + w) * cpp > fb->pitches[0]) {
+ if (offset == 0) {
+ DRM_DEBUG_KMS("Unable to find suitable display surface offset\n");
+ return -EINVAL;
+ }
+
+ offset = intel_adjust_tile_offset(&x, &y, plane_state, 0,
+ offset, offset - alignment);
+ }
+ }
+
+ plane_state->main.offset = offset;
+ plane_state->main.x = x;
+ plane_state->main.y = y;
+
+ return 0;
+}
+
+static int skl_check_nv12_aux_surface(struct intel_plane_state *plane_state)
+{
+ const struct drm_framebuffer *fb = plane_state->base.fb;
+ unsigned int rotation = plane_state->base.rotation;
+ int max_width = skl_max_plane_width(fb, 1, rotation);
+ int max_height = 4096;
+ int x = plane_state->base.src.x1 >> 17;
+ int y = plane_state->base.src.y1 >> 17;
+ int w = drm_rect_width(&plane_state->base.src) >> 17;
+ int h = drm_rect_height(&plane_state->base.src) >> 17;
+ u32 offset;
+
+ intel_add_fb_offsets(&x, &y, plane_state, 1);
+ offset = intel_compute_tile_offset(&x, &y, plane_state, 1);
+
+ /* FIXME not quite sure how/if these apply to the chroma plane */
+ if (w > max_width || h > max_height) {
+ DRM_DEBUG_KMS("CbCr source size %dx%d too big (limit %dx%d)\n",
+ w, h, max_width, max_height);
+ return -EINVAL;
+ }
+
+ plane_state->aux.offset = offset;
+ plane_state->aux.x = x;
+ plane_state->aux.y = y;
+
+ return 0;
+}
+
+int skl_check_plane_surface(struct intel_plane_state *plane_state)
+{
+ const struct drm_framebuffer *fb = plane_state->base.fb;
+ unsigned int rotation = plane_state->base.rotation;
+ int ret;
+
+ /* Rotate src coordinates to match rotated GTT view */
+ if (intel_rotation_90_or_270(rotation))
+ drm_rect_rotate(&plane_state->base.src,
+ fb->width, fb->height, DRM_ROTATE_270);
+
+ /*
+ * Handle the AUX surface first since
+ * the main surface setup depends on it.
+ */
+ if (fb->pixel_format == DRM_FORMAT_NV12) {
+ ret = skl_check_nv12_aux_surface(plane_state);
+ if (ret)
+ return ret;
+ } else {
+ plane_state->aux.offset = ~0xfff;
+ plane_state->aux.x = 0;
+ plane_state->aux.y = 0;
+ }
+
+ ret = skl_check_main_surface(plane_state);
+ if (ret)
+ return ret;
+
+ return 0;
+}
+
static void i9xx_update_primary_plane(struct drm_plane *primary,
const struct intel_crtc_state *crtc_state,
const struct intel_plane_state *plane_state)
u32 dspcntr;
i915_reg_t reg = DSPCNTR(plane);
unsigned int rotation = plane_state->base.rotation;
- int cpp = drm_format_plane_cpp(fb->pixel_format, 0);
int x = plane_state->base.src.x1 >> 16;
int y = plane_state->base.src.y1 >> 16;
BUG();
}
- if (INTEL_INFO(dev)->gen >= 4 && i915_gem_object_is_tiled(obj))
+ if (INTEL_GEN(dev_priv) >= 4 &&
+ fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
dspcntr |= DISPPLANE_TILED;
if (IS_G4X(dev))
dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE;
- linear_offset = y * fb->pitches[0] + x * cpp;
+ intel_add_fb_offsets(&x, &y, plane_state, 0);
- if (INTEL_INFO(dev)->gen >= 4) {
+ if (INTEL_INFO(dev)->gen >= 4)
intel_crtc->dspaddr_offset =
- intel_compute_tile_offset(&x, &y, fb, 0,
- fb->pitches[0], rotation);
- linear_offset -= intel_crtc->dspaddr_offset;
- } else {
- intel_crtc->dspaddr_offset = linear_offset;
- }
+ intel_compute_tile_offset(&x, &y, plane_state, 0);
if (rotation == DRM_ROTATE_180) {
dspcntr |= DISPPLANE_ROTATE_180;
x += (crtc_state->pipe_src_w - 1);
y += (crtc_state->pipe_src_h - 1);
-
- /* Finding the last pixel of the last line of the display
- data and adding to linear_offset*/
- linear_offset +=
- (crtc_state->pipe_src_h - 1) * fb->pitches[0] +
- (crtc_state->pipe_src_w - 1) * cpp;
}
+ linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0);
+
+ if (INTEL_INFO(dev)->gen < 4)
+ intel_crtc->dspaddr_offset = linear_offset;
+
intel_crtc->adjusted_x = x;
intel_crtc->adjusted_y = y;
I915_WRITE(DSPSTRIDE(plane), fb->pitches[0]);
if (INTEL_INFO(dev)->gen >= 4) {
I915_WRITE(DSPSURF(plane),
- i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+ intel_fb_gtt_offset(fb, rotation) +
+ intel_crtc->dspaddr_offset);
I915_WRITE(DSPTILEOFF(plane), (y << 16) | x);
I915_WRITE(DSPLINOFF(plane), linear_offset);
} else
- I915_WRITE(DSPADDR(plane), i915_gem_obj_ggtt_offset(obj) + linear_offset);
+ I915_WRITE(DSPADDR(plane), i915_gem_object_ggtt_offset(obj, NULL) + linear_offset);
POSTING_READ(reg);
}
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
struct drm_framebuffer *fb = plane_state->base.fb;
- struct drm_i915_gem_object *obj = intel_fb_obj(fb);
int plane = intel_crtc->plane;
u32 linear_offset;
u32 dspcntr;
i915_reg_t reg = DSPCNTR(plane);
unsigned int rotation = plane_state->base.rotation;
- int cpp = drm_format_plane_cpp(fb->pixel_format, 0);
int x = plane_state->base.src.x1 >> 16;
int y = plane_state->base.src.y1 >> 16;
BUG();
}
- if (i915_gem_object_is_tiled(obj))
+ if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
dspcntr |= DISPPLANE_TILED;
if (!IS_HASWELL(dev) && !IS_BROADWELL(dev))
dspcntr |= DISPPLANE_TRICKLE_FEED_DISABLE;
- linear_offset = y * fb->pitches[0] + x * cpp;
+ intel_add_fb_offsets(&x, &y, plane_state, 0);
+
intel_crtc->dspaddr_offset =
- intel_compute_tile_offset(&x, &y, fb, 0,
- fb->pitches[0], rotation);
- linear_offset -= intel_crtc->dspaddr_offset;
+ intel_compute_tile_offset(&x, &y, plane_state, 0);
+
if (rotation == DRM_ROTATE_180) {
dspcntr |= DISPPLANE_ROTATE_180;
if (!IS_HASWELL(dev) && !IS_BROADWELL(dev)) {
x += (crtc_state->pipe_src_w - 1);
y += (crtc_state->pipe_src_h - 1);
-
- /* Finding the last pixel of the last line of the display
- data and adding to linear_offset*/
- linear_offset +=
- (crtc_state->pipe_src_h - 1) * fb->pitches[0] +
- (crtc_state->pipe_src_w - 1) * cpp;
}
}
+ linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0);
+
intel_crtc->adjusted_x = x;
intel_crtc->adjusted_y = y;
I915_WRITE(DSPSTRIDE(plane), fb->pitches[0]);
I915_WRITE(DSPSURF(plane),
- i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset);
+ intel_fb_gtt_offset(fb, rotation) +
+ intel_crtc->dspaddr_offset);
if (IS_HASWELL(dev) || IS_BROADWELL(dev)) {
I915_WRITE(DSPOFFSET(plane), (y << 16) | x);
} else {
}
}
-u32 intel_plane_obj_offset(struct intel_plane *intel_plane,
- struct drm_i915_gem_object *obj,
- unsigned int plane)
+u32 intel_fb_gtt_offset(struct drm_framebuffer *fb,
+ unsigned int rotation)
{
+ struct drm_i915_gem_object *obj = intel_fb_obj(fb);
struct i915_ggtt_view view;
struct i915_vma *vma;
- u64 offset;
- intel_fill_fb_ggtt_view(&view, intel_plane->base.state->fb,
- intel_plane->base.state->rotation);
+ intel_fill_fb_ggtt_view(&view, fb, rotation);
- vma = i915_gem_obj_to_ggtt_view(obj, &view);
+ vma = i915_gem_object_to_ggtt(obj, &view);
if (WARN(!vma, "ggtt vma for display object not found! (view=%u)\n",
- view.type))
+ view.type))
return -1;
- offset = vma->node.start;
-
- if (plane == 1) {
- offset += vma->ggtt_view.params.rotated.uv_start_page *
- PAGE_SIZE;
- }
-
- WARN_ON(upper_32_bits(offset));
-
- return lower_32_bits(offset);
+ return i915_ggtt_offset(vma);
}
static void skl_detach_scaler(struct intel_crtc *intel_crtc, int id)
}
}
+u32 skl_plane_stride(const struct drm_framebuffer *fb, int plane,
+ unsigned int rotation)
+{
+ const struct drm_i915_private *dev_priv = to_i915(fb->dev);
+ u32 stride = intel_fb_pitch(fb, plane, rotation);
+
+ /*
+ * The stride is either expressed as a multiple of 64 bytes chunks for
+ * linear buffers or in number of tiles for tiled buffers.
+ */
+ if (intel_rotation_90_or_270(rotation)) {
+ int cpp = drm_format_plane_cpp(fb->pixel_format, plane);
+
+ stride /= intel_tile_height(dev_priv, fb->modifier[0], cpp);
+ } else {
+ stride /= intel_fb_stride_alignment(dev_priv, fb->modifier[0],
+ fb->pixel_format);
+ }
+
+ return stride;
+}
+
u32 skl_plane_ctl_format(uint32_t pixel_format)
{
switch (pixel_format) {
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
struct drm_framebuffer *fb = plane_state->base.fb;
- struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+ const struct skl_wm_values *wm = &dev_priv->wm.skl_results;
int pipe = intel_crtc->pipe;
- u32 plane_ctl, stride_div, stride;
- u32 tile_height, plane_offset, plane_size;
+ u32 plane_ctl;
unsigned int rotation = plane_state->base.rotation;
- int x_offset, y_offset;
- u32 surf_addr;
+ u32 stride = skl_plane_stride(fb, 0, rotation);
+ u32 surf_addr = plane_state->main.offset;
int scaler_id = plane_state->scaler_id;
- int src_x = plane_state->base.src.x1 >> 16;
- int src_y = plane_state->base.src.y1 >> 16;
+ int src_x = plane_state->main.x;
+ int src_y = plane_state->main.y;
int src_w = drm_rect_width(&plane_state->base.src) >> 16;
int src_h = drm_rect_height(&plane_state->base.src) >> 16;
int dst_x = plane_state->base.dst.x1;
plane_ctl |= PLANE_CTL_PLANE_GAMMA_DISABLE;
plane_ctl |= skl_plane_ctl_rotation(rotation);
- stride_div = intel_fb_stride_alignment(dev_priv, fb->modifier[0],
- fb->pixel_format);
- surf_addr = intel_plane_obj_offset(to_intel_plane(plane), obj, 0);
+ /* Sizes are 0 based */
+ src_w--;
+ src_h--;
+ dst_w--;
+ dst_h--;
- WARN_ON(drm_rect_width(&plane_state->base.src) == 0);
+ intel_crtc->adjusted_x = src_x;
+ intel_crtc->adjusted_y = src_y;
- if (intel_rotation_90_or_270(rotation)) {
- int cpp = drm_format_plane_cpp(fb->pixel_format, 0);
-
- /* stride = Surface height in tiles */
- tile_height = intel_tile_height(dev_priv, fb->modifier[0], cpp);
- stride = DIV_ROUND_UP(fb->height, tile_height);
- x_offset = stride * tile_height - src_y - src_h;
- y_offset = src_x;
- plane_size = (src_w - 1) << 16 | (src_h - 1);
- } else {
- stride = fb->pitches[0] / stride_div;
- x_offset = src_x;
- y_offset = src_y;
- plane_size = (src_h - 1) << 16 | (src_w - 1);
- }
- plane_offset = y_offset << 16 | x_offset;
-
- intel_crtc->adjusted_x = x_offset;
- intel_crtc->adjusted_y = y_offset;
+ if (wm->dirty_pipes & drm_crtc_mask(&intel_crtc->base))
+ skl_write_plane_wm(intel_crtc, wm, 0);
I915_WRITE(PLANE_CTL(pipe, 0), plane_ctl);
- I915_WRITE(PLANE_OFFSET(pipe, 0), plane_offset);
- I915_WRITE(PLANE_SIZE(pipe, 0), plane_size);
+ I915_WRITE(PLANE_OFFSET(pipe, 0), (src_y << 16) | src_x);
I915_WRITE(PLANE_STRIDE(pipe, 0), stride);
+ I915_WRITE(PLANE_SIZE(pipe, 0), (src_h << 16) | src_w);
if (scaler_id >= 0) {
uint32_t ps_ctrl = 0;
I915_WRITE(PLANE_POS(pipe, 0), (dst_y << 16) | dst_x);
}
- I915_WRITE(PLANE_SURF(pipe, 0), surf_addr);
+ I915_WRITE(PLANE_SURF(pipe, 0),
+ intel_fb_gtt_offset(fb, rotation) + surf_addr);
POSTING_READ(PLANE_SURF(pipe, 0));
}
{
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
- int pipe = to_intel_crtc(crtc)->pipe;
+ struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+ int pipe = intel_crtc->pipe;
+
+ /*
+ * We only populate skl_results on watermark updates, and if the
+ * plane's visiblity isn't actually changing neither is its watermarks.
+ */
+ if (!crtc->primary->state->visible)
+ skl_write_plane_wm(intel_crtc, &dev_priv->wm.skl_results, 0);
I915_WRITE(PLANE_CTL(pipe, 0), 0);
I915_WRITE(PLANE_SURF(pipe, 0), 0);
for_each_crtc(dev, crtc) {
struct intel_plane *plane = to_intel_plane(crtc->primary);
- struct intel_plane_state *plane_state;
-
- drm_modeset_lock_crtc(crtc, &plane->base);
- plane_state = to_intel_plane_state(plane->base.state);
+ struct intel_plane_state *plane_state =
+ to_intel_plane_state(plane->base.state);
if (plane_state->base.visible)
plane->update_plane(&plane->base,
to_intel_crtc_state(crtc->state),
plane_state);
+ }
+}
- drm_modeset_unlock_crtc(crtc);
+static int
+__intel_display_resume(struct drm_device *dev,
+ struct drm_atomic_state *state)
+{
+ struct drm_crtc_state *crtc_state;
+ struct drm_crtc *crtc;
+ int i, ret;
+
+ intel_modeset_setup_hw_state(dev);
+ i915_redisable_vga(dev);
+
+ if (!state)
+ return 0;
+
+ for_each_crtc_in_state(state, crtc, crtc_state, i) {
+ /*
+ * Force recalculation even if we restore
+ * current state. With fast modeset this may not result
+ * in a modeset when the state is compatible.
+ */
+ crtc_state->mode_changed = true;
}
+
+ /* ignore any reset values/BIOS leftovers in the WM registers */
+ to_intel_atomic_state(state)->skip_intermediate_wm = true;
+
+ ret = drm_atomic_commit(state);
+
+ WARN_ON(ret == -EDEADLK);
+ return ret;
+}
+
+static bool gpu_reset_clobbers_display(struct drm_i915_private *dev_priv)
+{
+ return intel_has_gpu_reset(dev_priv) &&
+ INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv);
}
void intel_prepare_reset(struct drm_i915_private *dev_priv)
{
- /* no reset support for gen2 */
- if (IS_GEN2(dev_priv))
- return;
+ struct drm_device *dev = &dev_priv->drm;
+ struct drm_modeset_acquire_ctx *ctx = &dev_priv->reset_ctx;
+ struct drm_atomic_state *state;
+ int ret;
- /* reset doesn't touch the display */
- if (INTEL_GEN(dev_priv) >= 5 || IS_G4X(dev_priv))
+ /*
+ * Need mode_config.mutex so that we don't
+ * trample ongoing ->detect() and whatnot.
+ */
+ mutex_lock(&dev->mode_config.mutex);
+ drm_modeset_acquire_init(ctx, 0);
+ while (1) {
+ ret = drm_modeset_lock_all_ctx(dev, ctx);
+ if (ret != -EDEADLK)
+ break;
+
+ drm_modeset_backoff(ctx);
+ }
+
+ /* reset doesn't touch the display, but flips might get nuked anyway, */
+ if (!i915.force_reset_modeset_test &&
+ !gpu_reset_clobbers_display(dev_priv))
return;
- drm_modeset_lock_all(&dev_priv->drm);
/*
* Disabling the crtcs gracefully seems nicer. Also the
* g33 docs say we should at least disable all the planes.
*/
- intel_display_suspend(&dev_priv->drm);
+ state = drm_atomic_helper_duplicate_state(dev, ctx);
+ if (IS_ERR(state)) {
+ ret = PTR_ERR(state);
+ state = NULL;
+ DRM_ERROR("Duplicating state failed with %i\n", ret);
+ goto err;
+ }
+
+ ret = drm_atomic_helper_disable_all(dev, ctx);
+ if (ret) {
+ DRM_ERROR("Suspending crtc's failed with %i\n", ret);
+ goto err;
+ }
+
+ dev_priv->modeset_restore_state = state;
+ state->acquire_ctx = ctx;
+ return;
+
+err:
+ drm_atomic_state_free(state);
}
void intel_finish_reset(struct drm_i915_private *dev_priv)
{
+ struct drm_device *dev = &dev_priv->drm;
+ struct drm_modeset_acquire_ctx *ctx = &dev_priv->reset_ctx;
+ struct drm_atomic_state *state = dev_priv->modeset_restore_state;
+ int ret;
+
/*
* Flips in the rings will be nuked by the reset,
* so complete all pending flips so that user space
*/
intel_complete_page_flips(dev_priv);
- /* no reset support for gen2 */
- if (IS_GEN2(dev_priv))
- return;
+ dev_priv->modeset_restore_state = NULL;
/* reset doesn't touch the display */
- if (INTEL_GEN(dev_priv) >= 5 || IS_G4X(dev_priv)) {
+ if (!gpu_reset_clobbers_display(dev_priv)) {
+ if (!state) {
+ /*
+ * Flips in the rings have been nuked by the reset,
+ * so update the base address of all primary
+ * planes to the the last fb to make sure we're
+ * showing the correct fb after a reset.
+ *
+ * FIXME: Atomic will make this obsolete since we won't schedule
+ * CS-based flips (which might get lost in gpu resets) any more.
+ */
+ intel_update_primary_planes(dev);
+ } else {
+ ret = __intel_display_resume(dev, state);
+ if (ret)
+ DRM_ERROR("Restoring old state failed with %i\n", ret);
+ }
+ } else {
/*
- * Flips in the rings have been nuked by the reset,
- * so update the base address of all primary
- * planes to the the last fb to make sure we're
- * showing the correct fb after a reset.
- *
- * FIXME: Atomic will make this obsolete since we won't schedule
- * CS-based flips (which might get lost in gpu resets) any more.
+ * The display has been reset as well,
+ * so need a full re-initialization.
*/
- intel_update_primary_planes(&dev_priv->drm);
- return;
- }
+ intel_runtime_pm_disable_interrupts(dev_priv);
+ intel_runtime_pm_enable_interrupts(dev_priv);
- /*
- * The display has been reset as well,
- * so need a full re-initialization.
- */
- intel_runtime_pm_disable_interrupts(dev_priv);
- intel_runtime_pm_enable_interrupts(dev_priv);
+ intel_modeset_init_hw(dev);
- intel_modeset_init_hw(&dev_priv->drm);
+ spin_lock_irq(&dev_priv->irq_lock);
+ if (dev_priv->display.hpd_irq_setup)
+ dev_priv->display.hpd_irq_setup(dev_priv);
+ spin_unlock_irq(&dev_priv->irq_lock);
- spin_lock_irq(&dev_priv->irq_lock);
- if (dev_priv->display.hpd_irq_setup)
- dev_priv->display.hpd_irq_setup(dev_priv);
- spin_unlock_irq(&dev_priv->irq_lock);
+ ret = __intel_display_resume(dev, state);
+ if (ret)
+ DRM_ERROR("Restoring old state failed with %i\n", ret);
- intel_display_resume(&dev_priv->drm);
+ intel_hpd_init(dev_priv);
+ }
+
+ drm_modeset_drop_locks(ctx);
+ drm_modeset_acquire_fini(ctx);
+ mutex_unlock(&dev->mode_config.mutex);
+}
- intel_hpd_init(dev_priv);
+static bool abort_flip_on_reset(struct intel_crtc *crtc)
+{
+ struct i915_gpu_error *error = &to_i915(crtc->base.dev)->gpu_error;
+
+ if (i915_reset_in_progress(error))
+ return true;
- drm_modeset_unlock_all(&dev_priv->drm);
+ if (crtc->reset_count != i915_reset_count(error))
+ return true;
+
+ return false;
}
static bool intel_crtc_has_pending_flip(struct drm_crtc *crtc)
{
struct drm_device *dev = crtc->dev;
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- unsigned reset_counter;
bool pending;
- reset_counter = i915_reset_counter(&to_i915(dev)->gpu_error);
- if (intel_crtc->reset_counter != reset_counter)
+ if (abort_flip_on_reset(intel_crtc))
return false;
spin_lock_irq(&dev->event_lock);
return 0;
}
-static void lpt_disable_iclkip(struct drm_i915_private *dev_priv)
+void lpt_disable_iclkip(struct drm_i915_private *dev_priv)
{
u32 temp;
intel_frontbuffer_flip(to_i915(dev), INTEL_FRONTBUFFER_ALL_MASK(pipe));
}
-static void ironlake_crtc_enable(struct drm_crtc *crtc)
+static void intel_encoders_pre_pll_enable(struct drm_crtc *crtc,
+ struct intel_crtc_state *crtc_state,
+ struct drm_atomic_state *old_state)
+{
+ struct drm_connector_state *old_conn_state;
+ struct drm_connector *conn;
+ int i;
+
+ for_each_connector_in_state(old_state, conn, old_conn_state, i) {
+ struct drm_connector_state *conn_state = conn->state;
+ struct intel_encoder *encoder =
+ to_intel_encoder(conn_state->best_encoder);
+
+ if (conn_state->crtc != crtc)
+ continue;
+
+ if (encoder->pre_pll_enable)
+ encoder->pre_pll_enable(encoder, crtc_state, conn_state);
+ }
+}
+
+static void intel_encoders_pre_enable(struct drm_crtc *crtc,
+ struct intel_crtc_state *crtc_state,
+ struct drm_atomic_state *old_state)
+{
+ struct drm_connector_state *old_conn_state;
+ struct drm_connector *conn;
+ int i;
+
+ for_each_connector_in_state(old_state, conn, old_conn_state, i) {
+ struct drm_connector_state *conn_state = conn->state;
+ struct intel_encoder *encoder =
+ to_intel_encoder(conn_state->best_encoder);
+
+ if (conn_state->crtc != crtc)
+ continue;
+
+ if (encoder->pre_enable)
+ encoder->pre_enable(encoder, crtc_state, conn_state);
+ }
+}
+
+static void intel_encoders_enable(struct drm_crtc *crtc,
+ struct intel_crtc_state *crtc_state,
+ struct drm_atomic_state *old_state)
+{
+ struct drm_connector_state *old_conn_state;
+ struct drm_connector *conn;
+ int i;
+
+ for_each_connector_in_state(old_state, conn, old_conn_state, i) {
+ struct drm_connector_state *conn_state = conn->state;
+ struct intel_encoder *encoder =
+ to_intel_encoder(conn_state->best_encoder);
+
+ if (conn_state->crtc != crtc)
+ continue;
+
+ encoder->enable(encoder, crtc_state, conn_state);
+ intel_opregion_notify_encoder(encoder, true);
+ }
+}
+
+static void intel_encoders_disable(struct drm_crtc *crtc,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_atomic_state *old_state)
+{
+ struct drm_connector_state *old_conn_state;
+ struct drm_connector *conn;
+ int i;
+
+ for_each_connector_in_state(old_state, conn, old_conn_state, i) {
+ struct intel_encoder *encoder =
+ to_intel_encoder(old_conn_state->best_encoder);
+
+ if (old_conn_state->crtc != crtc)
+ continue;
+
+ intel_opregion_notify_encoder(encoder, false);
+ encoder->disable(encoder, old_crtc_state, old_conn_state);
+ }
+}
+
+static void intel_encoders_post_disable(struct drm_crtc *crtc,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_atomic_state *old_state)
+{
+ struct drm_connector_state *old_conn_state;
+ struct drm_connector *conn;
+ int i;
+
+ for_each_connector_in_state(old_state, conn, old_conn_state, i) {
+ struct intel_encoder *encoder =
+ to_intel_encoder(old_conn_state->best_encoder);
+
+ if (old_conn_state->crtc != crtc)
+ continue;
+
+ if (encoder->post_disable)
+ encoder->post_disable(encoder, old_crtc_state, old_conn_state);
+ }
+}
+
+static void intel_encoders_post_pll_disable(struct drm_crtc *crtc,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_atomic_state *old_state)
+{
+ struct drm_connector_state *old_conn_state;
+ struct drm_connector *conn;
+ int i;
+
+ for_each_connector_in_state(old_state, conn, old_conn_state, i) {
+ struct intel_encoder *encoder =
+ to_intel_encoder(old_conn_state->best_encoder);
+
+ if (old_conn_state->crtc != crtc)
+ continue;
+
+ if (encoder->post_pll_disable)
+ encoder->post_pll_disable(encoder, old_crtc_state, old_conn_state);
+ }
+}
+
+static void ironlake_crtc_enable(struct intel_crtc_state *pipe_config,
+ struct drm_atomic_state *old_state)
{
+ struct drm_crtc *crtc = pipe_config->base.crtc;
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_encoder *encoder;
int pipe = intel_crtc->pipe;
- struct intel_crtc_state *pipe_config =
- to_intel_crtc_state(crtc->state);
if (WARN_ON(intel_crtc->active))
return;
intel_crtc->active = true;
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->pre_enable)
- encoder->pre_enable(encoder);
+ intel_encoders_pre_enable(crtc, pipe_config, old_state);
if (intel_crtc->config->has_pch_encoder) {
/* Note: FDI PLL enabling _must_ be done before we enable the
assert_vblank_disabled(crtc);
drm_crtc_vblank_on(crtc);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- encoder->enable(encoder);
+ intel_encoders_enable(crtc, pipe_config, old_state);
if (HAS_PCH_CPT(dev))
cpt_verify_modeset(dev, intel_crtc->pipe);
return HAS_IPS(crtc->base.dev) && crtc->pipe == PIPE_A;
}
-static void haswell_crtc_enable(struct drm_crtc *crtc)
+static void haswell_crtc_enable(struct intel_crtc_state *pipe_config,
+ struct drm_atomic_state *old_state)
{
+ struct drm_crtc *crtc = pipe_config->base.crtc;
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_encoder *encoder;
int pipe = intel_crtc->pipe, hsw_workaround_pipe;
enum transcoder cpu_transcoder = intel_crtc->config->cpu_transcoder;
- struct intel_crtc_state *pipe_config =
- to_intel_crtc_state(crtc->state);
if (WARN_ON(intel_crtc->active))
return;
intel_set_pch_fifo_underrun_reporting(dev_priv, TRANSCODER_A,
false);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->pre_pll_enable)
- encoder->pre_pll_enable(encoder);
+ intel_encoders_pre_pll_enable(crtc, pipe_config, old_state);
if (intel_crtc->config->shared_dpll)
intel_enable_shared_dpll(intel_crtc);
else
intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
- for_each_encoder_on_crtc(dev, crtc, encoder) {
- if (encoder->pre_enable)
- encoder->pre_enable(encoder);
- }
+ intel_encoders_pre_enable(crtc, pipe_config, old_state);
if (intel_crtc->config->has_pch_encoder)
dev_priv->display.fdi_link_train(crtc);
assert_vblank_disabled(crtc);
drm_crtc_vblank_on(crtc);
- for_each_encoder_on_crtc(dev, crtc, encoder) {
- encoder->enable(encoder);
- intel_opregion_notify_encoder(encoder, true);
- }
+ intel_encoders_enable(crtc, pipe_config, old_state);
if (intel_crtc->config->has_pch_encoder) {
intel_wait_for_vblank(dev, pipe);
}
}
-static void ironlake_crtc_disable(struct drm_crtc *crtc)
+static void ironlake_crtc_disable(struct intel_crtc_state *old_crtc_state,
+ struct drm_atomic_state *old_state)
{
+ struct drm_crtc *crtc = old_crtc_state->base.crtc;
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_encoder *encoder;
int pipe = intel_crtc->pipe;
/*
intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, false);
}
- for_each_encoder_on_crtc(dev, crtc, encoder)
- encoder->disable(encoder);
+ intel_encoders_disable(crtc, old_crtc_state, old_state);
drm_crtc_vblank_off(crtc);
assert_vblank_disabled(crtc);
if (intel_crtc->config->has_pch_encoder)
ironlake_fdi_disable(crtc);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->post_disable)
- encoder->post_disable(encoder);
+ intel_encoders_post_disable(crtc, old_crtc_state, old_state);
if (intel_crtc->config->has_pch_encoder) {
ironlake_disable_pch_transcoder(dev_priv, pipe);
intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, true);
}
-static void haswell_crtc_disable(struct drm_crtc *crtc)
+static void haswell_crtc_disable(struct intel_crtc_state *old_crtc_state,
+ struct drm_atomic_state *old_state)
{
+ struct drm_crtc *crtc = old_crtc_state->base.crtc;
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_encoder *encoder;
enum transcoder cpu_transcoder = intel_crtc->config->cpu_transcoder;
if (intel_crtc->config->has_pch_encoder)
intel_set_pch_fifo_underrun_reporting(dev_priv, TRANSCODER_A,
false);
- for_each_encoder_on_crtc(dev, crtc, encoder) {
- intel_opregion_notify_encoder(encoder, false);
- encoder->disable(encoder);
- }
+ intel_encoders_disable(crtc, old_crtc_state, old_state);
drm_crtc_vblank_off(crtc);
assert_vblank_disabled(crtc);
if (!transcoder_is_dsi(cpu_transcoder))
intel_ddi_disable_pipe_clock(intel_crtc);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->post_disable)
- encoder->post_disable(encoder);
-
- if (intel_crtc->config->has_pch_encoder) {
- lpt_disable_pch_transcoder(dev_priv);
- lpt_disable_iclkip(dev_priv);
- intel_ddi_fdi_disable(crtc);
+ intel_encoders_post_disable(crtc, old_crtc_state, old_state);
+ if (old_crtc_state->has_pch_encoder)
intel_set_pch_fifo_underrun_reporting(dev_priv, TRANSCODER_A,
true);
- }
}
static void i9xx_pfit_enable(struct intel_crtc *crtc)
intel_display_power_put(dev_priv, POWER_DOMAIN_PIPE_A);
}
-static void valleyview_crtc_enable(struct drm_crtc *crtc)
+static void valleyview_crtc_enable(struct intel_crtc_state *pipe_config,
+ struct drm_atomic_state *old_state)
{
+ struct drm_crtc *crtc = pipe_config->base.crtc;
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_encoder *encoder;
- struct intel_crtc_state *pipe_config =
- to_intel_crtc_state(crtc->state);
int pipe = intel_crtc->pipe;
if (WARN_ON(intel_crtc->active))
intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->pre_pll_enable)
- encoder->pre_pll_enable(encoder);
+ intel_encoders_pre_pll_enable(crtc, pipe_config, old_state);
if (IS_CHERRYVIEW(dev)) {
chv_prepare_pll(intel_crtc, intel_crtc->config);
vlv_enable_pll(intel_crtc, intel_crtc->config);
}
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->pre_enable)
- encoder->pre_enable(encoder);
+ intel_encoders_pre_enable(crtc, pipe_config, old_state);
i9xx_pfit_enable(intel_crtc);
assert_vblank_disabled(crtc);
drm_crtc_vblank_on(crtc);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- encoder->enable(encoder);
+ intel_encoders_enable(crtc, pipe_config, old_state);
}
static void i9xx_set_pll_dividers(struct intel_crtc *crtc)
I915_WRITE(FP1(crtc->pipe), crtc->config->dpll_hw_state.fp1);
}
-static void i9xx_crtc_enable(struct drm_crtc *crtc)
+static void i9xx_crtc_enable(struct intel_crtc_state *pipe_config,
+ struct drm_atomic_state *old_state)
{
+ struct drm_crtc *crtc = pipe_config->base.crtc;
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_encoder *encoder;
- struct intel_crtc_state *pipe_config =
- to_intel_crtc_state(crtc->state);
enum pipe pipe = intel_crtc->pipe;
if (WARN_ON(intel_crtc->active))
if (!IS_GEN2(dev))
intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->pre_enable)
- encoder->pre_enable(encoder);
+ intel_encoders_pre_enable(crtc, pipe_config, old_state);
i9xx_enable_pll(intel_crtc);
assert_vblank_disabled(crtc);
drm_crtc_vblank_on(crtc);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- encoder->enable(encoder);
+ intel_encoders_enable(crtc, pipe_config, old_state);
}
static void i9xx_pfit_disable(struct intel_crtc *crtc)
I915_WRITE(PFIT_CONTROL, 0);
}
-static void i9xx_crtc_disable(struct drm_crtc *crtc)
+static void i9xx_crtc_disable(struct intel_crtc_state *old_crtc_state,
+ struct drm_atomic_state *old_state)
{
+ struct drm_crtc *crtc = old_crtc_state->base.crtc;
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- struct intel_encoder *encoder;
int pipe = intel_crtc->pipe;
/*
if (IS_GEN2(dev))
intel_wait_for_vblank(dev, pipe);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- encoder->disable(encoder);
+ intel_encoders_disable(crtc, old_crtc_state, old_state);
drm_crtc_vblank_off(crtc);
assert_vblank_disabled(crtc);
i9xx_pfit_disable(intel_crtc);
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->post_disable)
- encoder->post_disable(encoder);
+ intel_encoders_post_disable(crtc, old_crtc_state, old_state);
if (!intel_crtc_has_type(intel_crtc->config, INTEL_OUTPUT_DSI)) {
if (IS_CHERRYVIEW(dev))
i9xx_disable_pll(intel_crtc);
}
- for_each_encoder_on_crtc(dev, crtc, encoder)
- if (encoder->post_pll_disable)
- encoder->post_pll_disable(encoder);
+ intel_encoders_post_pll_disable(crtc, old_crtc_state, old_state);
if (!IS_GEN2(dev))
intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
struct drm_i915_private *dev_priv = to_i915(crtc->dev);
enum intel_display_power_domain domain;
unsigned long domains;
+ struct drm_atomic_state *state;
+ struct intel_crtc_state *crtc_state;
+ int ret;
if (!intel_crtc->active)
return;
to_intel_plane_state(crtc->primary->state)->base.visible = false;
}
- dev_priv->display.crtc_disable(crtc);
+ state = drm_atomic_state_alloc(crtc->dev);
+ state->acquire_ctx = crtc->dev->mode_config.acquire_ctx;
+
+ /* Everything's already locked, -EDEADLK can't happen. */
+ crtc_state = intel_atomic_get_crtc_state(state, intel_crtc);
+ ret = drm_atomic_add_affected_connectors(state, crtc);
+
+ WARN_ON(IS_ERR(crtc_state) || ret);
+
+ dev_priv->display.crtc_disable(crtc_state, state);
+
+ drm_atomic_state_free(state);
DRM_DEBUG_KMS("[CRTC:%d:%s] hw state adjusted, was enabled, now disabled\n",
crtc->base.id, crtc->name);
static int pnv_get_display_clock_speed(struct drm_device *dev)
{
+ struct pci_dev *pdev = dev->pdev;
u16 gcfgc = 0;
- pci_read_config_word(dev->pdev, GCFGC, &gcfgc);
+ pci_read_config_word(pdev, GCFGC, &gcfgc);
switch (gcfgc & GC_DISPLAY_CLOCK_MASK) {
case GC_DISPLAY_CLOCK_267_MHZ_PNV:
static int i915gm_get_display_clock_speed(struct drm_device *dev)
{
+ struct pci_dev *pdev = dev->pdev;
u16 gcfgc = 0;
- pci_read_config_word(dev->pdev, GCFGC, &gcfgc);
+ pci_read_config_word(pdev, GCFGC, &gcfgc);
if (gcfgc & GC_LOW_FREQUENCY_ENABLE)
return 133333;
static int i85x_get_display_clock_speed(struct drm_device *dev)
{
+ struct pci_dev *pdev = dev->pdev;
u16 hpllcc = 0;
/*
* encoding is different :(
* FIXME is this the right way to detect 852GM/852GMV?
*/
- if (dev->pdev->revision == 0x1)
+ if (pdev->revision == 0x1)
return 133333;
- pci_bus_read_config_word(dev->pdev->bus,
+ pci_bus_read_config_word(pdev->bus,
PCI_DEVFN(0, 3), HPLLCC, &hpllcc);
/* Assume that the hardware is in the high speed state. This
static int gm45_get_display_clock_speed(struct drm_device *dev)
{
+ struct pci_dev *pdev = dev->pdev;
unsigned int cdclk_sel, vco = intel_hpll_vco(dev);
uint16_t tmp = 0;
- pci_read_config_word(dev->pdev, GCFGC, &tmp);
+ pci_read_config_word(pdev, GCFGC, &tmp);
cdclk_sel = (tmp >> 12) & 0x1;
static int i965gm_get_display_clock_speed(struct drm_device *dev)
{
+ struct pci_dev *pdev = dev->pdev;
static const uint8_t div_3200[] = { 16, 10, 8 };
static const uint8_t div_4000[] = { 20, 12, 10 };
static const uint8_t div_5333[] = { 24, 16, 14 };
unsigned int cdclk_sel, vco = intel_hpll_vco(dev);
uint16_t tmp = 0;
- pci_read_config_word(dev->pdev, GCFGC, &tmp);
+ pci_read_config_word(pdev, GCFGC, &tmp);
cdclk_sel = ((tmp >> 8) & 0x1f) - 1;
static int g33_get_display_clock_speed(struct drm_device *dev)
{
+ struct pci_dev *pdev = dev->pdev;
static const uint8_t div_3200[] = { 12, 10, 8, 7, 5, 16 };
static const uint8_t div_4000[] = { 14, 12, 10, 8, 6, 20 };
static const uint8_t div_4800[] = { 20, 14, 12, 10, 8, 24 };
unsigned int cdclk_sel, vco = intel_hpll_vco(dev);
uint16_t tmp = 0;
- pci_read_config_word(dev->pdev, GCFGC, &tmp);
+ pci_read_config_word(pdev, GCFGC, &tmp);
cdclk_sel = (tmp >> 4) & 0x7;
return;
error:
- kfree(fb);
+ kfree(intel_fb);
}
static void ironlake_get_pfit_config(struct intel_crtc *crtc,
I915_STATE_WARN(I915_READ(SPLL_CTL) & SPLL_PLL_ENABLE, "SPLL enabled\n");
I915_STATE_WARN(I915_READ(WRPLL_CTL(0)) & WRPLL_PLL_ENABLE, "WRPLL1 enabled\n");
I915_STATE_WARN(I915_READ(WRPLL_CTL(1)) & WRPLL_PLL_ENABLE, "WRPLL2 enabled\n");
- I915_STATE_WARN(I915_READ(PCH_PP_STATUS) & PP_ON, "Panel power on\n");
+ I915_STATE_WARN(I915_READ(PP_STATUS(0)) & PP_ON, "Panel power on\n");
I915_STATE_WARN(I915_READ(BLC_PWM_CPU_CTL2) & BLM_PWM_ENABLE,
"CPU PWM1 enabled\n");
if (IS_HASWELL(dev))
mutex_lock(&dev_priv->rps.hw_lock);
if (sandybridge_pcode_write(dev_priv, GEN6_PCODE_WRITE_D_COMP,
val))
- DRM_ERROR("Failed to write to D_COMP\n");
+ DRM_DEBUG_KMS("Failed to write to D_COMP\n");
mutex_unlock(&dev_priv->rps.hw_lock);
} else {
I915_WRITE(D_COMP_BDW, val);
switch (port) {
case PORT_A:
- pipe_config->ddi_pll_sel = SKL_DPLL0;
id = DPLL_ID_SKL_DPLL0;
break;
case PORT_B:
- pipe_config->ddi_pll_sel = SKL_DPLL1;
id = DPLL_ID_SKL_DPLL1;
break;
case PORT_C:
- pipe_config->ddi_pll_sel = SKL_DPLL2;
id = DPLL_ID_SKL_DPLL2;
break;
default:
u32 temp;
temp = I915_READ(DPLL_CTRL2) & DPLL_CTRL2_DDI_CLK_SEL_MASK(port);
- pipe_config->ddi_pll_sel = temp >> (port * 3 + 1);
+ id = temp >> (port * 3 + 1);
- switch (pipe_config->ddi_pll_sel) {
- case SKL_DPLL0:
- id = DPLL_ID_SKL_DPLL0;
- break;
- case SKL_DPLL1:
- id = DPLL_ID_SKL_DPLL1;
- break;
- case SKL_DPLL2:
- id = DPLL_ID_SKL_DPLL2;
- break;
- case SKL_DPLL3:
- id = DPLL_ID_SKL_DPLL3;
- break;
- default:
- MISSING_CASE(pipe_config->ddi_pll_sel);
+ if (WARN_ON(id < SKL_DPLL0 || id > SKL_DPLL3))
return;
- }
pipe_config->shared_dpll = intel_get_shared_dpll_by_id(dev_priv, id);
}
struct intel_crtc_state *pipe_config)
{
enum intel_dpll_id id;
+ uint32_t ddi_pll_sel = I915_READ(PORT_CLK_SEL(port));
- pipe_config->ddi_pll_sel = I915_READ(PORT_CLK_SEL(port));
-
- switch (pipe_config->ddi_pll_sel) {
+ switch (ddi_pll_sel) {
case PORT_CLK_SEL_WRPLL1:
id = DPLL_ID_WRPLL1;
break;
id = DPLL_ID_LCPLL_2700;
break;
default:
- MISSING_CASE(pipe_config->ddi_pll_sel);
+ MISSING_CASE(ddi_pll_sel);
/* fall through */
case PORT_CLK_SEL_NONE:
return;
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+ const struct skl_wm_values *wm = &dev_priv->wm.skl_results;
int pipe = intel_crtc->pipe;
uint32_t cntl = 0;
+ if (INTEL_GEN(dev_priv) >= 9 && wm->dirty_pipes & drm_crtc_mask(crtc))
+ skl_write_cursor_wm(intel_crtc, wm);
+
if (plane_state && plane_state->base.visible) {
cntl = MCURSOR_GAMMA_ENABLE;
switch (plane_state->base.crtc_w) {
{
struct drm_device *dev = crtc->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
- unsigned reset_counter;
- reset_counter = i915_reset_counter(&dev_priv->gpu_error);
- if (crtc->reset_counter != reset_counter)
+ if (abort_flip_on_reset(crtc))
return true;
/*
MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
intel_ring_emit(ring, fb->pitches[0]);
intel_ring_emit(ring, intel_crtc->flip_work->gtt_offset |
- i915_gem_object_get_tiling(obj));
+ intel_fb_modifier_to_tiling(fb->modifier[0]));
/* XXX Enabling the panel-fitter across page-flip is so far
* untested on non-native modes, so ignore it for now.
intel_ring_emit(ring, MI_DISPLAY_FLIP |
MI_DISPLAY_FLIP_PLANE(intel_crtc->plane));
- intel_ring_emit(ring, fb->pitches[0] | i915_gem_object_get_tiling(obj));
+ intel_ring_emit(ring, fb->pitches[0] |
+ intel_fb_modifier_to_tiling(fb->modifier[0]));
intel_ring_emit(ring, intel_crtc->flip_work->gtt_offset);
/* Contrary to the suggestions in the documentation,
intel_ring_emit(ring, MI_STORE_REGISTER_MEM |
MI_SRM_LRM_GLOBAL_GTT);
intel_ring_emit_reg(ring, DERRMR);
- intel_ring_emit(ring, req->engine->scratch.gtt_offset + 256);
+ intel_ring_emit(ring,
+ i915_ggtt_offset(req->engine->scratch) + 256);
if (IS_GEN8(dev)) {
intel_ring_emit(ring, 0);
intel_ring_emit(ring, MI_NOOP);
}
intel_ring_emit(ring, MI_DISPLAY_FLIP_I915 | plane_bit);
- intel_ring_emit(ring, fb->pitches[0] | i915_gem_object_get_tiling(obj));
+ intel_ring_emit(ring, fb->pitches[0] |
+ intel_fb_modifier_to_tiling(fb->modifier[0]));
intel_ring_emit(ring, intel_crtc->flip_work->gtt_offset);
intel_ring_emit(ring, (MI_NOOP));
struct drm_i915_private *dev_priv = to_i915(dev);
struct drm_framebuffer *fb = intel_crtc->base.primary->fb;
const enum pipe pipe = intel_crtc->pipe;
- u32 ctl, stride, tile_height;
+ u32 ctl, stride = skl_plane_stride(fb, 0, rotation);
ctl = I915_READ(PLANE_CTL(pipe, 0));
ctl &= ~PLANE_CTL_TILED_MASK;
MISSING_CASE(fb->modifier[0]);
}
- /*
- * The stride is either expressed as a multiple of 64 bytes chunks for
- * linear buffers or in number of tiles for tiled buffers.
- */
- if (intel_rotation_90_or_270(rotation)) {
- /* stride = Surface height in tiles */
- tile_height = intel_tile_height(dev_priv, fb->modifier[0], 0);
- stride = DIV_ROUND_UP(fb->height, tile_height);
- } else {
- stride = fb->pitches[0] /
- intel_fb_stride_alignment(dev_priv, fb->modifier[0],
- fb->pixel_format);
- }
-
/*
* Both PLANE_CTL and PLANE_STRIDE are not updated on vblank but on
* PLANE_SURF updates, the update is then guaranteed to be atomic.
{
struct drm_device *dev = intel_crtc->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_framebuffer *intel_fb =
- to_intel_framebuffer(intel_crtc->base.primary->fb);
- struct drm_i915_gem_object *obj = intel_fb->obj;
+ struct drm_framebuffer *fb = intel_crtc->base.primary->fb;
i915_reg_t reg = DSPCNTR(intel_crtc->plane);
u32 dspcntr;
dspcntr = I915_READ(reg);
- if (i915_gem_object_is_tiled(obj))
+ if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
dspcntr |= DISPPLANE_TILED;
else
dspcntr &= ~DISPPLANE_TILED;
if (work->flip_queued_req)
WARN_ON(i915_wait_request(work->flip_queued_req,
- false, NULL,
- NO_WAITBOOST));
+ 0, NULL, NO_WAITBOOST));
/* For framebuffer backed by dmabuf, wait for fence */
resv = i915_gem_object_get_dmabuf_resv(obj);
struct intel_engine_cs *engine;
bool mmio_flip;
struct drm_i915_gem_request *request;
+ struct i915_vma *vma;
int ret;
/*
crtc->primary->fb = fb;
update_state_fb(crtc->primary);
- intel_fbc_pre_update(intel_crtc, intel_crtc->config,
- to_intel_plane_state(primary->state));
-
work->pending_flip_obj = i915_gem_object_get(obj);
ret = i915_mutex_lock_interruptible(dev);
if (ret)
goto cleanup;
- intel_crtc->reset_counter = i915_reset_counter(&dev_priv->gpu_error);
- if (__i915_reset_in_progress_or_wedged(intel_crtc->reset_counter)) {
+ intel_crtc->reset_count = i915_reset_count(&dev_priv->gpu_error);
+ if (i915_reset_in_progress_or_wedged(&dev_priv->gpu_error)) {
ret = -EIO;
goto cleanup;
}
if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) {
engine = &dev_priv->engine[BCS];
- if (i915_gem_object_get_tiling(obj) !=
- i915_gem_object_get_tiling(intel_fb_obj(work->old_fb)))
+ if (fb->modifier[0] != old_fb->modifier[0])
/* vlv: DISPLAY_FLIP fails to change tiling */
engine = NULL;
} else if (IS_IVYBRIDGE(dev) || IS_HASWELL(dev)) {
mmio_flip = use_mmio_flip(engine, obj);
- ret = intel_pin_and_fence_fb_obj(fb, primary->state->rotation);
- if (ret)
+ vma = intel_pin_and_fence_fb_obj(fb, primary->state->rotation);
+ if (IS_ERR(vma)) {
+ ret = PTR_ERR(vma);
goto cleanup_pending;
+ }
- work->gtt_offset = intel_plane_obj_offset(to_intel_plane(primary),
- obj, 0);
+ work->gtt_offset = intel_fb_gtt_offset(fb, primary->state->rotation);
work->gtt_offset += intel_crtc->dspaddr_offset;
work->rotation = crtc->primary->state->rotation;
+ /*
+ * There's the potential that the next frame will not be compatible with
+ * FBC, so we want to call pre_update() before the actual page flip.
+ * The problem is that pre_update() caches some information about the fb
+ * object, so we want to do this only after the object is pinned. Let's
+ * be on the safe side and do this immediately before scheduling the
+ * flip.
+ */
+ intel_fbc_pre_update(intel_crtc, intel_crtc->config,
+ to_intel_plane_state(primary->state));
+
if (mmio_flip) {
INIT_WORK(&work->mmio_work, intel_mmio_flip_work_func);
goto cleanup_unpin;
}
- ret = i915_gem_object_sync(obj, request);
+ ret = i915_gem_request_await_object(request, obj, false);
if (ret)
goto cleanup_request;
DRM_DEBUG_KMS("double wide: %i\n", pipe_config->double_wide);
if (IS_BROXTON(dev)) {
- DRM_DEBUG_KMS("ddi_pll_sel: %u; dpll_hw_state: ebb0: 0x%x, ebb4: 0x%x,"
+ DRM_DEBUG_KMS("dpll_hw_state: ebb0: 0x%x, ebb4: 0x%x,"
"pll0: 0x%x, pll1: 0x%x, pll2: 0x%x, pll3: 0x%x, "
"pll6: 0x%x, pll8: 0x%x, pll9: 0x%x, pll10: 0x%x, pcsdw12: 0x%x\n",
- pipe_config->ddi_pll_sel,
pipe_config->dpll_hw_state.ebb0,
pipe_config->dpll_hw_state.ebb4,
pipe_config->dpll_hw_state.pll0,
pipe_config->dpll_hw_state.pll10,
pipe_config->dpll_hw_state.pcsdw12);
} else if (IS_SKYLAKE(dev) || IS_KABYLAKE(dev)) {
- DRM_DEBUG_KMS("ddi_pll_sel: %u; dpll_hw_state: "
+ DRM_DEBUG_KMS("dpll_hw_state: "
"ctrl1: 0x%x, cfgcr1: 0x%x, cfgcr2: 0x%x\n",
- pipe_config->ddi_pll_sel,
pipe_config->dpll_hw_state.ctrl1,
pipe_config->dpll_hw_state.cfgcr1,
pipe_config->dpll_hw_state.cfgcr2);
} else if (HAS_DDI(dev)) {
- DRM_DEBUG_KMS("ddi_pll_sel: 0x%x; dpll_hw_state: wrpll: 0x%x spll: 0x%x\n",
- pipe_config->ddi_pll_sel,
+ DRM_DEBUG_KMS("dpll_hw_state: wrpll: 0x%x spll: 0x%x\n",
pipe_config->dpll_hw_state.wrpll,
pipe_config->dpll_hw_state.spll);
} else {
struct intel_crtc_scaler_state scaler_state;
struct intel_dpll_hw_state dpll_hw_state;
struct intel_shared_dpll *shared_dpll;
- uint32_t ddi_pll_sel;
bool force_thru;
/* FIXME: before the switch to atomic started, a new pipe_config was
scaler_state = crtc_state->scaler_state;
shared_dpll = crtc_state->shared_dpll;
dpll_hw_state = crtc_state->dpll_hw_state;
- ddi_pll_sel = crtc_state->ddi_pll_sel;
force_thru = crtc_state->pch_pfit.force_thru;
memset(crtc_state, 0, sizeof *crtc_state);
crtc_state->scaler_state = scaler_state;
crtc_state->shared_dpll = shared_dpll;
crtc_state->dpll_hw_state = dpll_hw_state;
- crtc_state->ddi_pll_sel = ddi_pll_sel;
crtc_state->pch_pfit.force_thru = force_thru;
}
encoder = to_intel_encoder(connector_state->best_encoder);
- if (!(encoder->compute_config(encoder, pipe_config))) {
+ if (!(encoder->compute_config(encoder, pipe_config, connector_state))) {
DRM_DEBUG_KMS("Encoder config failure\n");
goto fail;
}
return false;
}
-#define for_each_intel_crtc_masked(dev, mask, intel_crtc) \
- list_for_each_entry((intel_crtc), \
- &(dev)->mode_config.crtc_list, \
- base.head) \
- for_each_if (mask & (1 <<(intel_crtc)->pipe))
-
static bool
intel_compare_m_n(unsigned int m, unsigned int n,
unsigned int m2, unsigned int n2,
PIPE_CONF_CHECK_I(double_wide);
- PIPE_CONF_CHECK_X(ddi_pll_sel);
-
PIPE_CONF_CHECK_P(shared_dpll);
PIPE_CONF_CHECK_X(dpll_hw_state.dpll);
PIPE_CONF_CHECK_X(dpll_hw_state.dpll_md);
hw_entry->start, hw_entry->end);
}
- /* cursor */
- hw_entry = &hw_ddb.plane[pipe][PLANE_CURSOR];
- sw_entry = &sw_ddb->plane[pipe][PLANE_CURSOR];
-
- if (!skl_ddb_entry_equal(hw_entry, sw_entry)) {
- DRM_ERROR("mismatch in DDB state pipe %c cursor "
- "(expected (%u,%u), found (%u,%u))\n",
- pipe_name(pipe),
- sw_entry->start, sw_entry->end,
- hw_entry->start, hw_entry->end);
+ /*
+ * cursor
+ * If the cursor plane isn't active, we may not have updated it's ddb
+ * allocation. In that case since the ddb allocation will be updated
+ * once the plane becomes visible, we can skip this check
+ */
+ if (intel_crtc->cursor_addr) {
+ hw_entry = &hw_ddb.plane[pipe][PLANE_CURSOR];
+ sw_entry = &sw_ddb->plane[pipe][PLANE_CURSOR];
+
+ if (!skl_ddb_entry_equal(hw_entry, sw_entry)) {
+ DRM_ERROR("mismatch in DDB state pipe %c cursor "
+ "(expected (%u,%u), found (%u,%u))\n",
+ pipe_name(pipe),
+ sw_entry->start, sw_entry->end,
+ hw_entry->start, hw_entry->end);
+ }
}
}
continue;
ret = i915_wait_request(intel_plane_state->wait_req,
- true, NULL, NULL);
+ I915_WAIT_INTERRUPTIBLE,
+ NULL, NULL);
if (ret) {
/* Any hang should be swallowed by the wait */
WARN_ON(ret == -EIO);
return false;
}
+static void intel_update_crtc(struct drm_crtc *crtc,
+ struct drm_atomic_state *state,
+ struct drm_crtc_state *old_crtc_state,
+ unsigned int *crtc_vblank_mask)
+{
+ struct drm_device *dev = crtc->dev;
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+ struct intel_crtc_state *pipe_config = to_intel_crtc_state(crtc->state);
+ bool modeset = needs_modeset(crtc->state);
+
+ if (modeset) {
+ update_scanline_offset(intel_crtc);
+ dev_priv->display.crtc_enable(pipe_config, state);
+ } else {
+ intel_pre_plane_update(to_intel_crtc_state(old_crtc_state));
+ }
+
+ if (drm_atomic_get_existing_plane_state(state, crtc->primary)) {
+ intel_fbc_enable(
+ intel_crtc, pipe_config,
+ to_intel_plane_state(crtc->primary->state));
+ }
+
+ drm_atomic_helper_commit_planes_on_crtc(old_crtc_state);
+
+ if (needs_vblank_wait(pipe_config))
+ *crtc_vblank_mask |= drm_crtc_mask(crtc);
+}
+
+static void intel_update_crtcs(struct drm_atomic_state *state,
+ unsigned int *crtc_vblank_mask)
+{
+ struct drm_crtc *crtc;
+ struct drm_crtc_state *old_crtc_state;
+ int i;
+
+ for_each_crtc_in_state(state, crtc, old_crtc_state, i) {
+ if (!crtc->state->active)
+ continue;
+
+ intel_update_crtc(crtc, state, old_crtc_state,
+ crtc_vblank_mask);
+ }
+}
+
+static void skl_update_crtcs(struct drm_atomic_state *state,
+ unsigned int *crtc_vblank_mask)
+{
+ struct drm_device *dev = state->dev;
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
+ struct drm_crtc *crtc;
+ struct drm_crtc_state *old_crtc_state;
+ struct skl_ddb_allocation *new_ddb = &intel_state->wm_results.ddb;
+ struct skl_ddb_allocation *cur_ddb = &dev_priv->wm.skl_hw.ddb;
+ unsigned int updated = 0;
+ bool progress;
+ enum pipe pipe;
+
+ /*
+ * Whenever the number of active pipes changes, we need to make sure we
+ * update the pipes in the right order so that their ddb allocations
+ * never overlap with eachother inbetween CRTC updates. Otherwise we'll
+ * cause pipe underruns and other bad stuff.
+ */
+ do {
+ int i;
+ progress = false;
+
+ for_each_crtc_in_state(state, crtc, old_crtc_state, i) {
+ bool vbl_wait = false;
+ unsigned int cmask = drm_crtc_mask(crtc);
+ pipe = to_intel_crtc(crtc)->pipe;
+
+ if (updated & cmask || !crtc->state->active)
+ continue;
+ if (skl_ddb_allocation_overlaps(state, cur_ddb, new_ddb,
+ pipe))
+ continue;
+
+ updated |= cmask;
+
+ /*
+ * If this is an already active pipe, it's DDB changed,
+ * and this isn't the last pipe that needs updating
+ * then we need to wait for a vblank to pass for the
+ * new ddb allocation to take effect.
+ */
+ if (!skl_ddb_allocation_equals(cur_ddb, new_ddb, pipe) &&
+ !crtc->state->active_changed &&
+ intel_state->wm_results.dirty_pipes != updated)
+ vbl_wait = true;
+
+ intel_update_crtc(crtc, state, old_crtc_state,
+ crtc_vblank_mask);
+
+ if (vbl_wait)
+ intel_wait_for_vblank(dev, pipe);
+
+ progress = true;
+ }
+ } while (progress);
+}
+
static void intel_atomic_commit_tail(struct drm_atomic_state *state)
{
struct drm_device *dev = state->dev;
continue;
ret = i915_wait_request(intel_plane_state->wait_req,
- true, NULL, NULL);
+ 0, NULL, NULL);
/* EIO should be eaten, and we can't get interrupted in the
* worker, and blocking commits have waited already. */
WARN_ON(ret);
if (old_crtc_state->active) {
intel_crtc_disable_planes(crtc, old_crtc_state->plane_mask);
- dev_priv->display.crtc_disable(crtc);
+ dev_priv->display.crtc_disable(to_intel_crtc_state(old_crtc_state), state);
intel_crtc->active = false;
intel_fbc_disable(intel_crtc);
intel_disable_shared_dpll(intel_crtc);
intel_state->cdclk_pll_vco != dev_priv->cdclk_pll.vco))
dev_priv->display.modeset_commit_cdclk(state);
+ /*
+ * SKL workaround: bspec recommends we disable the SAGV when we
+ * have more then one pipe enabled
+ */
+ if (IS_SKYLAKE(dev_priv) && !skl_can_enable_sagv(state))
+ skl_disable_sagv(dev_priv);
+
intel_modeset_verify_disabled(dev);
}
- /* Now enable the clocks, plane, pipe, and connectors that we set up. */
+ /* Complete the events for pipes that have now been disabled */
for_each_crtc_in_state(state, crtc, old_crtc_state, i) {
- struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
bool modeset = needs_modeset(crtc->state);
- struct intel_crtc_state *pipe_config =
- to_intel_crtc_state(crtc->state);
-
- if (modeset && crtc->state->active) {
- update_scanline_offset(to_intel_crtc(crtc));
- dev_priv->display.crtc_enable(crtc);
- }
/* Complete events for now disable pipes here. */
if (modeset && !crtc->state->active && crtc->state->event) {
crtc->state->event = NULL;
}
-
- if (!modeset)
- intel_pre_plane_update(to_intel_crtc_state(old_crtc_state));
-
- if (crtc->state->active &&
- drm_atomic_get_existing_plane_state(state, crtc->primary))
- intel_fbc_enable(intel_crtc, pipe_config, to_intel_plane_state(crtc->primary->state));
-
- if (crtc->state->active)
- drm_atomic_helper_commit_planes_on_crtc(old_crtc_state);
-
- if (pipe_config->base.active && needs_vblank_wait(pipe_config))
- crtc_vblank_mask |= 1 << i;
}
+ /* Now enable the clocks, plane, pipe, and connectors that we set up. */
+ dev_priv->display.update_crtcs(state, &crtc_vblank_mask);
+
/* FIXME: We should call drm_atomic_helper_commit_hw_done() here
* already, but still need the state for the delayed optimization. To
* fix this:
intel_modeset_verify_crtc(crtc, old_crtc_state, crtc->state);
}
+ if (IS_SKYLAKE(dev_priv) && intel_state->modeset &&
+ skl_can_enable_sagv(state))
+ skl_enable_sagv(dev_priv);
+
drm_atomic_helper_commit_hw_done(state);
if (intel_state->modeset)
drm_atomic_state_free(state);
}
-#undef for_each_intel_crtc_masked
-
/*
* FIXME: Remove this once i915 is fully DRIVER_ATOMIC by calling
* drm_atomic_helper_legacy_gamma_set() directly.
if (ret)
DRM_DEBUG_KMS("failed to attach phys object\n");
} else {
- ret = intel_pin_and_fence_fb_obj(fb, new_state->rotation);
+ struct i915_vma *vma;
+
+ vma = intel_pin_and_fence_fb_obj(fb, new_state->rotation);
+ if (IS_ERR(vma))
+ ret = PTR_ERR(vma);
}
if (ret == 0) {
struct intel_crtc_state *crtc_state,
struct intel_plane_state *state)
{
+ struct drm_i915_private *dev_priv = to_i915(plane->dev);
struct drm_crtc *crtc = state->base.crtc;
int min_scale = DRM_PLANE_HELPER_NO_SCALING;
int max_scale = DRM_PLANE_HELPER_NO_SCALING;
bool can_position = false;
+ int ret;
- if (INTEL_INFO(plane->dev)->gen >= 9) {
+ if (INTEL_GEN(dev_priv) >= 9) {
/* use scaler when colorkey is not required */
if (state->ckey.flags == I915_SET_COLORKEY_NONE) {
min_scale = 1;
can_position = true;
}
- return drm_plane_helper_check_state(&state->base,
- &state->clip,
- min_scale, max_scale,
- can_position, true);
+ ret = drm_plane_helper_check_state(&state->base,
+ &state->clip,
+ min_scale, max_scale,
+ can_position, true);
+ if (ret)
+ return ret;
+
+ if (!state->base.fb)
+ return 0;
+
+ if (INTEL_GEN(dev_priv) >= 9) {
+ ret = skl_check_plane_surface(state);
+ if (ret)
+ return ret;
+ }
+
+ return 0;
}
static void intel_begin_crtc_commit(struct drm_crtc *crtc,
struct drm_crtc_state *old_crtc_state)
{
struct drm_device *dev = crtc->dev;
+ struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
struct intel_crtc_state *old_intel_state =
to_intel_crtc_state(old_crtc_state);
bool modeset = needs_modeset(crtc->state);
+ enum pipe pipe = intel_crtc->pipe;
/* Perform vblank evasion around commit operation */
intel_pipe_update_start(intel_crtc);
if (to_intel_crtc_state(crtc->state)->update_pipe)
intel_update_pipe_config(intel_crtc, old_intel_state);
- else if (INTEL_INFO(dev)->gen >= 9)
+ else if (INTEL_GEN(dev_priv) >= 9) {
skl_detach_scalers(intel_crtc);
+
+ I915_WRITE(PIPE_WM_LINETIME(pipe),
+ dev_priv->wm.skl_hw.wm_linetime[pipe]);
+ }
}
static void intel_finish_crtc_commit(struct drm_crtc *crtc,
if (!obj)
addr = 0;
else if (!INTEL_INFO(dev)->cursor_needs_physical)
- addr = i915_gem_obj_ggtt_offset(obj);
+ addr = i915_gem_object_ggtt_offset(obj, NULL);
else
addr = obj->phys_handle->busaddr;
return true;
}
+void intel_pps_unlock_regs_wa(struct drm_i915_private *dev_priv)
+{
+ int pps_num;
+ int pps_idx;
+
+ if (HAS_DDI(dev_priv))
+ return;
+ /*
+ * This w/a is needed at least on CPT/PPT, but to be sure apply it
+ * everywhere where registers can be write protected.
+ */
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+ pps_num = 2;
+ else
+ pps_num = 1;
+
+ for (pps_idx = 0; pps_idx < pps_num; pps_idx++) {
+ u32 val = I915_READ(PP_CONTROL(pps_idx));
+
+ val = (val & ~PANEL_UNLOCK_MASK) | PANEL_UNLOCK_REGS;
+ I915_WRITE(PP_CONTROL(pps_idx), val);
+ }
+}
+
+static void intel_pps_init(struct drm_i915_private *dev_priv)
+{
+ if (HAS_PCH_SPLIT(dev_priv) || IS_BROXTON(dev_priv))
+ dev_priv->pps_mmio_base = PCH_PPS_BASE;
+ else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+ dev_priv->pps_mmio_base = VLV_PPS_BASE;
+ else
+ dev_priv->pps_mmio_base = PPS_BASE;
+
+ intel_pps_unlock_regs_wa(dev_priv);
+}
+
static void intel_setup_outputs(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_encoder *encoder;
bool dpd_is_edp = false;
+ intel_pps_init(dev_priv);
+
/*
* intel_edp_init_connector() depends on this completing first, to
* prevent the registeration of both eDP and LVDS and the incorrect
struct drm_i915_gem_object *obj)
{
struct drm_i915_private *dev_priv = to_i915(dev);
- unsigned int aligned_height;
+ unsigned int tiling = i915_gem_object_get_tiling(obj);
int ret;
u32 pitch_limit, stride_alignment;
char *format_name;
WARN_ON(!mutex_is_locked(&dev->struct_mutex));
if (mode_cmd->flags & DRM_MODE_FB_MODIFIERS) {
- /* Enforce that fb modifier and tiling mode match, but only for
- * X-tiled. This is needed for FBC. */
- if (!!(i915_gem_object_get_tiling(obj) == I915_TILING_X) !=
- !!(mode_cmd->modifier[0] == I915_FORMAT_MOD_X_TILED)) {
+ /*
+ * If there's a fence, enforce that
+ * the fb modifier and tiling mode match.
+ */
+ if (tiling != I915_TILING_NONE &&
+ tiling != intel_fb_modifier_to_tiling(mode_cmd->modifier[0])) {
DRM_DEBUG("tiling_mode doesn't match fb modifier\n");
return -EINVAL;
}
} else {
- if (i915_gem_object_get_tiling(obj) == I915_TILING_X)
+ if (tiling == I915_TILING_X) {
mode_cmd->modifier[0] = I915_FORMAT_MOD_X_TILED;
- else if (i915_gem_object_get_tiling(obj) == I915_TILING_Y) {
+ } else if (tiling == I915_TILING_Y) {
DRM_DEBUG("No Y tiling for legacy addfb\n");
return -EINVAL;
}
return -EINVAL;
}
+ /*
+ * gen2/3 display engine uses the fence if present,
+ * so the tiling mode must match the fb modifier exactly.
+ */
+ if (INTEL_INFO(dev_priv)->gen < 4 &&
+ tiling != intel_fb_modifier_to_tiling(mode_cmd->modifier[0])) {
+ DRM_DEBUG("tiling_mode must match fb modifier exactly on gen2/3\n");
+ return -EINVAL;
+ }
+
stride_alignment = intel_fb_stride_alignment(dev_priv,
mode_cmd->modifier[0],
mode_cmd->pixel_format);
return -EINVAL;
}
- if (mode_cmd->modifier[0] == I915_FORMAT_MOD_X_TILED &&
+ /*
+ * If there's a fence, enforce that
+ * the fb pitch and fence stride match.
+ */
+ if (tiling != I915_TILING_NONE &&
mode_cmd->pitches[0] != i915_gem_object_get_stride(obj)) {
DRM_DEBUG("pitch (%d) must match tiling stride (%d)\n",
mode_cmd->pitches[0],
if (mode_cmd->offsets[0] != 0)
return -EINVAL;
- aligned_height = intel_fb_align_height(dev, mode_cmd->height,
- mode_cmd->pixel_format,
- mode_cmd->modifier[0]);
- /* FIXME drm helper for size checks (especially planar formats)? */
- if (obj->base.size < aligned_height * mode_cmd->pitches[0])
- return -EINVAL;
-
drm_helper_mode_fill_fb_struct(&intel_fb->base, mode_cmd);
intel_fb->obj = obj;
- intel_fill_fb_info(dev_priv, &intel_fb->base);
+ ret = intel_fill_fb_info(dev_priv, &intel_fb->base);
+ if (ret)
+ return ret;
ret = drm_framebuffer_init(dev, &intel_fb->base, &intel_fb_funcs);
if (ret) {
skl_modeset_calc_cdclk;
}
+ if (dev_priv->info.gen >= 9)
+ dev_priv->display.update_crtcs = skl_update_crtcs;
+ else
+ dev_priv->display.update_crtcs = intel_update_crtcs;
+
switch (INTEL_INFO(dev_priv)->gen) {
case 2:
dev_priv->display.queue_flip = intel_gen2_queue_flip;
static void i915_disable_vga(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
u8 sr1;
i915_reg_t vga_reg = i915_vgacntrl_reg(dev);
/* WaEnableVGAAccessThroughIOPort:ctg,elk,ilk,snb,ivb,vlv,hsw */
- vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO);
+ vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
outb(SR01, VGA_SR_INDEX);
sr1 = inb(VGA_SR_DATA);
outb(sr1 | 1<<5, VGA_SR_DATA);
- vga_put(dev->pdev, VGA_RSRC_LEGACY_IO);
+ vga_put(pdev, VGA_RSRC_LEGACY_IO);
udelay(300);
I915_WRITE(vga_reg, VGA_DISP_DISABLE);
return false;
}
-static bool intel_encoder_has_connectors(struct intel_encoder *encoder)
+static struct intel_connector *intel_encoder_find_connector(struct intel_encoder *encoder)
{
struct drm_device *dev = encoder->base.dev;
struct intel_connector *connector;
for_each_connector_on_encoder(dev, &encoder->base, connector)
- return true;
+ return connector;
- return false;
+ return NULL;
+}
+
+static bool has_pch_trancoder(struct drm_i915_private *dev_priv,
+ enum transcoder pch_transcoder)
+{
+ return HAS_PCH_IBX(dev_priv) || HAS_PCH_CPT(dev_priv) ||
+ (HAS_PCH_LPT_H(dev_priv) && pch_transcoder == TRANSCODER_A);
}
static void intel_sanitize_crtc(struct intel_crtc *crtc)
* worst a fifo underrun happens which also sets this to false.
*/
crtc->cpu_fifo_underrun_disabled = true;
- crtc->pch_fifo_underrun_disabled = true;
+ /*
+ * We track the PCH trancoder underrun reporting state
+ * within the crtc. With crtc for pipe A housing the underrun
+ * reporting state for PCH transcoder A, crtc for pipe B housing
+ * it for PCH transcoder B, etc. LPT-H has only PCH transcoder A,
+ * and marking underrun reporting as disabled for the non-existing
+ * PCH transcoders B and C would prevent enabling the south
+ * error interrupt (see cpt_can_enable_serr_int()).
+ */
+ if (has_pch_trancoder(dev_priv, (enum transcoder)crtc->pipe))
+ crtc->pch_fifo_underrun_disabled = true;
}
}
static void intel_sanitize_encoder(struct intel_encoder *encoder)
{
struct intel_connector *connector;
- struct drm_device *dev = encoder->base.dev;
/* We need to check both for a crtc link (meaning that the
* encoder is active and trying to read from a pipe) and the
bool has_active_crtc = encoder->base.crtc &&
to_intel_crtc(encoder->base.crtc)->active;
- if (intel_encoder_has_connectors(encoder) && !has_active_crtc) {
+ connector = intel_encoder_find_connector(encoder);
+ if (connector && !has_active_crtc) {
DRM_DEBUG_KMS("[ENCODER:%d:%s] has active connectors but no active pipe!\n",
encoder->base.base.id,
encoder->base.name);
* fallout from our resume register restoring. Disable
* the encoder manually again. */
if (encoder->base.crtc) {
+ struct drm_crtc_state *crtc_state = encoder->base.crtc->state;
+
DRM_DEBUG_KMS("[ENCODER:%d:%s] manually disabled\n",
encoder->base.base.id,
encoder->base.name);
- encoder->disable(encoder);
+ encoder->disable(encoder, to_intel_crtc_state(crtc_state), connector->base.state);
if (encoder->post_disable)
- encoder->post_disable(encoder);
+ encoder->post_disable(encoder, to_intel_crtc_state(crtc_state), connector->base.state);
}
encoder->base.crtc = NULL;
* a bug in one of the get_hw_state functions. Or someplace else
* in our code, like the register restore mess on resume. Clamp
* things to off as a safer default. */
- for_each_intel_connector(dev, connector) {
- if (connector->encoder != encoder)
- continue;
- connector->base.dpms = DRM_MODE_DPMS_OFF;
- connector->base.encoder = NULL;
- }
+
+ connector->base.dpms = DRM_MODE_DPMS_OFF;
+ connector->base.encoder = NULL;
}
/* Enabled encoders without active connectors will be fixed in
* the crtc fixup. */
struct drm_atomic_state *state = dev_priv->modeset_restore_state;
struct drm_modeset_acquire_ctx ctx;
int ret;
- bool setup = false;
dev_priv->modeset_restore_state = NULL;
+ if (state)
+ state->acquire_ctx = &ctx;
/*
* This is a cludge because with real atomic modeset mode_config.mutex
mutex_lock(&dev->mode_config.mutex);
drm_modeset_acquire_init(&ctx, 0);
-retry:
- ret = drm_modeset_lock_all_ctx(dev, &ctx);
-
- if (ret == 0 && !setup) {
- setup = true;
-
- intel_modeset_setup_hw_state(dev);
- i915_redisable_vga(dev);
- }
-
- if (ret == 0 && state) {
- struct drm_crtc_state *crtc_state;
- struct drm_crtc *crtc;
- int i;
-
- state->acquire_ctx = &ctx;
-
- /* ignore any reset values/BIOS leftovers in the WM registers */
- to_intel_atomic_state(state)->skip_intermediate_wm = true;
-
- for_each_crtc_in_state(state, crtc, crtc_state, i) {
- /*
- * Force recalculation even if we restore
- * current state. With fast modeset this may not result
- * in a modeset when the state is compatible.
- */
- crtc_state->mode_changed = true;
- }
-
- ret = drm_atomic_commit(state);
- }
+ while (1) {
+ ret = drm_modeset_lock_all_ctx(dev, &ctx);
+ if (ret != -EDEADLK)
+ break;
- if (ret == -EDEADLK) {
drm_modeset_backoff(&ctx);
- goto retry;
}
+ if (!ret)
+ ret = __intel_display_resume(dev, state);
+
drm_modeset_drop_locks(&ctx);
drm_modeset_acquire_fini(&ctx);
mutex_unlock(&dev->mode_config.mutex);
struct drm_i915_private *dev_priv = to_i915(dev);
struct drm_crtc *c;
struct drm_i915_gem_object *obj;
- int ret;
intel_init_gt_powersave(dev_priv);
* for this.
*/
for_each_crtc(dev, c) {
+ struct i915_vma *vma;
+
obj = intel_fb_obj(c->primary->fb);
if (obj == NULL)
continue;
mutex_lock(&dev->struct_mutex);
- ret = intel_pin_and_fence_fb_obj(c->primary->fb,
+ vma = intel_pin_and_fence_fb_obj(c->primary->fb,
c->primary->state->rotation);
mutex_unlock(&dev->struct_mutex);
- if (ret) {
+ if (IS_ERR(vma)) {
DRM_ERROR("failed to pin boot fb on pipe %d\n",
to_intel_crtc(c)->pipe);
drm_framebuffer_unreference(c->primary->fb);
return (max_link_clock * max_lanes * 8) / 10;
}
+static int
+intel_dp_downstream_max_dotclock(struct intel_dp *intel_dp)
+{
+ struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+ struct intel_encoder *encoder = &intel_dig_port->base;
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+ int max_dotclk = dev_priv->max_dotclk_freq;
+ int ds_max_dotclk;
+
+ int type = intel_dp->downstream_ports[0] & DP_DS_PORT_TYPE_MASK;
+
+ if (type != DP_DS_PORT_TYPE_VGA)
+ return max_dotclk;
+
+ ds_max_dotclk = drm_dp_downstream_max_clock(intel_dp->dpcd,
+ intel_dp->downstream_ports);
+
+ if (ds_max_dotclk != 0)
+ max_dotclk = min(max_dotclk, ds_max_dotclk);
+
+ return max_dotclk;
+}
+
static enum drm_mode_status
intel_dp_mode_valid(struct drm_connector *connector,
struct drm_display_mode *mode)
struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode;
int target_clock = mode->clock;
int max_rate, mode_rate, max_lanes, max_link_clock;
- int max_dotclk = to_i915(connector->dev)->max_dotclk_freq;
+ int max_dotclk;
+
+ max_dotclk = intel_dp_downstream_max_dotclock(intel_dp);
if (is_edp(intel_dp) && fixed_mode) {
if (mode->hdisplay > fixed_mode->hdisplay)
static void
intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
struct intel_dp *intel_dp);
+static void
+intel_dp_pps_init(struct drm_device *dev, struct intel_dp *intel_dp);
static void pps_lock(struct intel_dp *intel_dp)
{
static bool vlv_pipe_has_pp_on(struct drm_i915_private *dev_priv,
enum pipe pipe)
{
- return I915_READ(VLV_PIPE_PP_STATUS(pipe)) & PP_ON;
+ return I915_READ(PP_STATUS(pipe)) & PP_ON;
}
static bool vlv_pipe_has_vdd_on(struct drm_i915_private *dev_priv,
enum pipe pipe)
{
- return I915_READ(VLV_PIPE_PP_CONTROL(pipe)) & EDP_FORCE_VDD;
+ return I915_READ(PP_CONTROL(pipe)) & EDP_FORCE_VDD;
}
static bool vlv_pipe_any(struct drm_i915_private *dev_priv,
enum pipe pipe;
for (pipe = PIPE_A; pipe <= PIPE_B; pipe++) {
- u32 port_sel = I915_READ(VLV_PIPE_PP_ON_DELAYS(pipe)) &
+ u32 port_sel = I915_READ(PP_ON_DELAYS(pipe)) &
PANEL_PORT_SELECT_MASK;
if (port_sel != PANEL_PORT_SELECT_VLV(port))
struct intel_dp *intel_dp,
struct pps_registers *regs)
{
+ int pps_idx = 0;
+
memset(regs, 0, sizeof(*regs));
- if (IS_BROXTON(dev_priv)) {
- int idx = bxt_power_sequencer_idx(intel_dp);
-
- regs->pp_ctrl = BXT_PP_CONTROL(idx);
- regs->pp_stat = BXT_PP_STATUS(idx);
- regs->pp_on = BXT_PP_ON_DELAYS(idx);
- regs->pp_off = BXT_PP_OFF_DELAYS(idx);
- } else if (HAS_PCH_SPLIT(dev_priv)) {
- regs->pp_ctrl = PCH_PP_CONTROL;
- regs->pp_stat = PCH_PP_STATUS;
- regs->pp_on = PCH_PP_ON_DELAYS;
- regs->pp_off = PCH_PP_OFF_DELAYS;
- regs->pp_div = PCH_PP_DIVISOR;
- } else {
- enum pipe pipe = vlv_power_sequencer_pipe(intel_dp);
+ if (IS_BROXTON(dev_priv))
+ pps_idx = bxt_power_sequencer_idx(intel_dp);
+ else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+ pps_idx = vlv_power_sequencer_pipe(intel_dp);
- regs->pp_ctrl = VLV_PIPE_PP_CONTROL(pipe);
- regs->pp_stat = VLV_PIPE_PP_STATUS(pipe);
- regs->pp_on = VLV_PIPE_PP_ON_DELAYS(pipe);
- regs->pp_off = VLV_PIPE_PP_OFF_DELAYS(pipe);
- regs->pp_div = VLV_PIPE_PP_DIVISOR(pipe);
- }
+ regs->pp_ctrl = PP_CONTROL(pps_idx);
+ regs->pp_stat = PP_STATUS(pps_idx);
+ regs->pp_on = PP_ON_DELAYS(pps_idx);
+ regs->pp_off = PP_OFF_DELAYS(pps_idx);
+ if (!IS_BROXTON(dev_priv))
+ regs->pp_div = PP_DIVISOR(pps_idx);
}
static i915_reg_t
i915_reg_t pp_ctrl_reg, pp_div_reg;
u32 pp_div;
- pp_ctrl_reg = VLV_PIPE_PP_CONTROL(pipe);
- pp_div_reg = VLV_PIPE_PP_DIVISOR(pipe);
+ pp_ctrl_reg = PP_CONTROL(pipe);
+ pp_div_reg = PP_DIVISOR(pipe);
pp_div = I915_READ(pp_div_reg);
pp_div &= PP_REFERENCE_DIVIDER_MASK;
}
static void
-intel_dp_aux_init(struct intel_dp *intel_dp, struct intel_connector *connector)
+intel_dp_aux_init(struct intel_dp *intel_dp)
{
struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
enum port port = intel_dig_port->port;
DRM_DEBUG_KMS("common rates: %s\n", str);
}
+static void intel_dp_print_hw_revision(struct intel_dp *intel_dp)
+{
+ uint8_t rev;
+ int len;
+
+ if ((drm_debug & DRM_UT_KMS) == 0)
+ return;
+
+ if (!(intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] &
+ DP_DWN_STRM_PORT_PRESENT))
+ return;
+
+ len = drm_dp_dpcd_read(&intel_dp->aux, DP_BRANCH_HW_REV, &rev, 1);
+ if (len < 0)
+ return;
+
+ DRM_DEBUG_KMS("sink hw revision: %d.%d\n", (rev & 0xf0) >> 4, rev & 0xf);
+}
+
+static void intel_dp_print_sw_revision(struct intel_dp *intel_dp)
+{
+ uint8_t rev[2];
+ int len;
+
+ if ((drm_debug & DRM_UT_KMS) == 0)
+ return;
+
+ if (!(intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] &
+ DP_DWN_STRM_PORT_PRESENT))
+ return;
+
+ len = drm_dp_dpcd_read(&intel_dp->aux, DP_BRANCH_SW_REV, &rev, 2);
+ if (len < 0)
+ return;
+
+ DRM_DEBUG_KMS("sink sw revision: %d.%d\n", rev[0], rev[1]);
+}
+
static int rate_to_index(int find, const int *rates)
{
int i = 0;
}
}
+static int intel_dp_compute_bpp(struct intel_dp *intel_dp,
+ struct intel_crtc_state *pipe_config)
+{
+ int bpp, bpc;
+
+ bpp = pipe_config->pipe_bpp;
+ bpc = drm_dp_downstream_max_bpc(intel_dp->dpcd, intel_dp->downstream_ports);
+
+ if (bpc > 0)
+ bpp = min(bpp, 3*bpc);
+
+ return bpp;
+}
+
bool
intel_dp_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
/* Walk through all bpp values. Luckily they're all nicely spaced with 2
* bpc in between. */
- bpp = pipe_config->pipe_bpp;
+ bpp = intel_dp_compute_bpp(intel_dp, pipe_config);
if (is_edp(intel_dp)) {
/* Get bpp from vbt only for panels that dont have bpp in edid */
}
void intel_dp_set_link_params(struct intel_dp *intel_dp,
- const struct intel_crtc_state *pipe_config)
+ int link_rate, uint8_t lane_count,
+ bool link_mst)
{
- intel_dp->link_rate = pipe_config->port_clock;
- intel_dp->lane_count = pipe_config->lane_count;
- intel_dp->link_mst = intel_crtc_has_type(pipe_config, INTEL_OUTPUT_DP_MST);
+ intel_dp->link_rate = link_rate;
+ intel_dp->lane_count = lane_count;
+ intel_dp->link_mst = link_mst;
}
-static void intel_dp_prepare(struct intel_encoder *encoder)
+static void intel_dp_prepare(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
enum port port = dp_to_dig_port(intel_dp)->port;
struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
- const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode;
+ const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
- intel_dp_set_link_params(intel_dp, crtc->config);
+ intel_dp_set_link_params(intel_dp, pipe_config->port_clock,
+ pipe_config->lane_count,
+ intel_crtc_has_type(pipe_config,
+ INTEL_OUTPUT_DP_MST));
/*
* There are four kinds of DP registers:
/* Handle DP bits in common between all three register formats */
intel_dp->DP |= DP_VOLTAGE_0_4 | DP_PRE_EMPHASIS_0;
- intel_dp->DP |= DP_PORT_WIDTH(crtc->config->lane_count);
+ intel_dp->DP |= DP_PORT_WIDTH(pipe_config->lane_count);
/* Split out the IBX/CPU vs CPT settings */
I915_WRITE(TRANS_DP_CTL(crtc->pipe), trans_dp);
} else {
if (!HAS_PCH_SPLIT(dev) && !IS_VALLEYVIEW(dev) &&
- !IS_CHERRYVIEW(dev) && crtc->config->limited_color_range)
+ !IS_CHERRYVIEW(dev) && pipe_config->limited_color_range)
intel_dp->DP |= DP_COLOR_RANGE_16_235;
if (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC)
lockdep_assert_held(&dev_priv->pps_mutex);
control = I915_READ(_pp_ctrl_reg(intel_dp));
- if (!IS_BROXTON(dev)) {
+ if (WARN_ON(!HAS_DDI(dev_priv) &&
+ (control & PANEL_UNLOCK_MASK) != PANEL_UNLOCK_REGS)) {
control &= ~PANEL_UNLOCK_MASK;
control |= PANEL_UNLOCK_REGS;
}
DRM_DEBUG_KMS("PP_STATUS: 0x%08x PP_CONTROL: 0x%08x\n",
I915_READ(pp_stat_reg), I915_READ(pp_ctrl_reg));
- if ((pp & POWER_TARGET_ON) == 0)
+ if ((pp & PANEL_POWER_ON) == 0)
intel_dp->panel_power_off_time = ktime_get_boottime();
power_domain = intel_display_port_aux_power_domain(intel_encoder);
POSTING_READ(pp_ctrl_reg);
}
- pp |= POWER_TARGET_ON;
+ pp |= PANEL_POWER_ON;
if (!IS_GEN5(dev))
pp |= PANEL_POWER_RESET;
pp = ironlake_get_pp_control(intel_dp);
/* We need to switch off panel power _and_ force vdd, for otherwise some
* panels get very unhappy and cease to work. */
- pp &= ~(POWER_TARGET_ON | PANEL_POWER_RESET | EDP_FORCE_VDD |
+ pp &= ~(PANEL_POWER_ON | PANEL_POWER_RESET | EDP_FORCE_VDD |
EDP_BLC_ENABLE);
pp_ctrl_reg = _pp_ctrl_reg(intel_dp);
#define assert_edp_pll_enabled(d) assert_edp_pll((d), true)
#define assert_edp_pll_disabled(d) assert_edp_pll((d), false)
-static void ironlake_edp_pll_on(struct intel_dp *intel_dp)
+static void ironlake_edp_pll_on(struct intel_dp *intel_dp,
+ struct intel_crtc_state *pipe_config)
{
- struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
- struct intel_crtc *crtc = to_intel_crtc(intel_dig_port->base.base.crtc);
+ struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
assert_pipe_disabled(dev_priv, crtc->pipe);
assert_edp_pll_disabled(dev_priv);
DRM_DEBUG_KMS("enabling eDP PLL for clock %d\n",
- crtc->config->port_clock);
+ pipe_config->port_clock);
intel_dp->DP &= ~DP_PLL_FREQ_MASK;
- if (crtc->config->port_clock == 162000)
+ if (pipe_config->port_clock == 162000)
intel_dp->DP |= DP_PLL_FREQ_162MHZ;
else
intel_dp->DP |= DP_PLL_FREQ_270MHZ;
}
}
-static void intel_disable_dp(struct intel_encoder *encoder)
+static void intel_disable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
- struct drm_device *dev = encoder->base.dev;
- struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
- if (crtc->config->has_audio)
+ if (old_crtc_state->has_audio)
intel_audio_codec_disable(encoder);
- if (HAS_PSR(dev) && !HAS_DDI(dev))
+ if (HAS_PSR(dev_priv) && !HAS_DDI(dev_priv))
intel_psr_disable(intel_dp);
/* Make sure the panel is off before trying to change the mode. But also
intel_edp_panel_off(intel_dp);
/* disable the port before the pipe on g4x */
- if (INTEL_INFO(dev)->gen < 5)
+ if (INTEL_GEN(dev_priv) < 5)
intel_dp_link_down(intel_dp);
}
-static void ilk_post_disable_dp(struct intel_encoder *encoder)
+static void ilk_post_disable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
enum port port = dp_to_dig_port(intel_dp)->port;
ironlake_edp_pll_off(intel_dp);
}
-static void vlv_post_disable_dp(struct intel_encoder *encoder)
+static void vlv_post_disable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
intel_dp_link_down(intel_dp);
}
-static void chv_post_disable_dp(struct intel_encoder *encoder)
+static void chv_post_disable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
enum port port = intel_dig_port->port;
+ if (dp_train_pat & DP_TRAINING_PATTERN_MASK)
+ DRM_DEBUG_KMS("Using DP training pattern TPS%d\n",
+ dp_train_pat & DP_TRAINING_PATTERN_MASK);
+
if (HAS_DDI(dev)) {
uint32_t temp = I915_READ(DP_TP_CTL(port));
*DP |= DP_LINK_TRAIN_PAT_2_CPT;
break;
case DP_TRAINING_PATTERN_3:
- DRM_ERROR("DP training pattern 3 not supported\n");
+ DRM_DEBUG_KMS("TPS3 not supported, using TPS2 instead\n");
*DP |= DP_LINK_TRAIN_PAT_2_CPT;
break;
}
if (IS_CHERRYVIEW(dev)) {
*DP |= DP_LINK_TRAIN_PAT_3_CHV;
} else {
- DRM_ERROR("DP training pattern 3 not supported\n");
+ DRM_DEBUG_KMS("TPS3 not supported, using TPS2 instead\n");
*DP |= DP_LINK_TRAIN_PAT_2;
}
break;
}
}
-static void intel_dp_enable_port(struct intel_dp *intel_dp)
+static void intel_dp_enable_port(struct intel_dp *intel_dp,
+ struct intel_crtc_state *old_crtc_state)
{
struct drm_device *dev = intel_dp_to_dev(intel_dp);
struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_crtc *crtc =
- to_intel_crtc(dp_to_dig_port(intel_dp)->base.base.crtc);
/* enable with pattern 1 (as per spec) */
- _intel_dp_set_link_train(intel_dp, &intel_dp->DP,
- DP_TRAINING_PATTERN_1);
- I915_WRITE(intel_dp->output_reg, intel_dp->DP);
- POSTING_READ(intel_dp->output_reg);
+ intel_dp_program_link_training_pattern(intel_dp, DP_TRAINING_PATTERN_1);
/*
* Magic for VLV/CHV. We _must_ first set up the register
* fail when the power sequencer is freshly used for this port.
*/
intel_dp->DP |= DP_PORT_EN;
- if (crtc->config->has_audio)
+ if (old_crtc_state->has_audio)
intel_dp->DP |= DP_AUDIO_OUTPUT_ENABLE;
I915_WRITE(intel_dp->output_reg, intel_dp->DP);
POSTING_READ(intel_dp->output_reg);
}
-static void intel_enable_dp(struct intel_encoder *encoder)
+static void intel_enable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config)
{
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
struct drm_device *dev = encoder->base.dev;
if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
vlv_init_panel_power_sequencer(intel_dp);
- intel_dp_enable_port(intel_dp);
+ intel_dp_enable_port(intel_dp, pipe_config);
edp_panel_vdd_on(intel_dp);
edp_panel_on(intel_dp);
unsigned int lane_mask = 0x0;
if (IS_CHERRYVIEW(dev))
- lane_mask = intel_dp_unused_lane_mask(crtc->config->lane_count);
+ lane_mask = intel_dp_unused_lane_mask(pipe_config->lane_count);
vlv_wait_port_ready(dev_priv, dp_to_dig_port(intel_dp),
lane_mask);
intel_dp_start_link_train(intel_dp);
intel_dp_stop_link_train(intel_dp);
- if (crtc->config->has_audio) {
+ if (pipe_config->has_audio) {
DRM_DEBUG_DRIVER("Enabling DP audio on pipe %c\n",
pipe_name(pipe));
intel_audio_codec_enable(encoder);
}
}
-static void g4x_enable_dp(struct intel_encoder *encoder)
+static void g4x_enable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
- intel_enable_dp(encoder);
+ intel_enable_dp(encoder, pipe_config);
intel_edp_backlight_on(intel_dp);
}
-static void vlv_enable_dp(struct intel_encoder *encoder)
+static void vlv_enable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
intel_psr_enable(intel_dp);
}
-static void g4x_pre_enable_dp(struct intel_encoder *encoder)
+static void g4x_pre_enable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
enum port port = dp_to_dig_port(intel_dp)->port;
- intel_dp_prepare(encoder);
+ intel_dp_prepare(encoder, pipe_config);
/* Only ilk+ has port A */
if (port == PORT_A)
- ironlake_edp_pll_on(intel_dp);
+ ironlake_edp_pll_on(intel_dp, pipe_config);
}
static void vlv_detach_power_sequencer(struct intel_dp *intel_dp)
struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
struct drm_i915_private *dev_priv = to_i915(intel_dig_port->base.base.dev);
enum pipe pipe = intel_dp->pps_pipe;
- i915_reg_t pp_on_reg = VLV_PIPE_PP_ON_DELAYS(pipe);
+ i915_reg_t pp_on_reg = PP_ON_DELAYS(pipe);
edp_panel_vdd_off_sync(intel_dp);
intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
}
-static void vlv_pre_enable_dp(struct intel_encoder *encoder)
+static void vlv_pre_enable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
vlv_phy_pre_encoder_enable(encoder);
- intel_enable_dp(encoder);
+ intel_enable_dp(encoder, pipe_config);
}
-static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder)
+static void vlv_dp_pre_pll_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
- intel_dp_prepare(encoder);
+ intel_dp_prepare(encoder, pipe_config);
vlv_phy_pre_pll_enable(encoder);
}
-static void chv_pre_enable_dp(struct intel_encoder *encoder)
+static void chv_pre_enable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
chv_phy_pre_encoder_enable(encoder);
- intel_enable_dp(encoder);
+ intel_enable_dp(encoder, pipe_config);
/* Second common lane will stay alive on its own now */
chv_phy_release_cl2_override(encoder);
}
-static void chv_dp_pre_pll_enable(struct intel_encoder *encoder)
+static void chv_dp_pre_pll_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
- intel_dp_prepare(encoder);
+ intel_dp_prepare(encoder, pipe_config);
chv_phy_pre_pll_enable(encoder);
}
-static void chv_dp_post_pll_disable(struct intel_encoder *encoder)
+static void chv_dp_post_pll_disable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
chv_phy_post_pll_disable(encoder);
}
*
* Return %true if @port is connected, %false otherwise.
*/
-bool intel_digital_port_connected(struct drm_i915_private *dev_priv,
+static bool intel_digital_port_connected(struct drm_i915_private *dev_priv,
struct intel_digital_port *port)
{
if (HAS_PCH_IBX(dev_priv))
intel_dp_probe_oui(intel_dp);
+ intel_dp_print_hw_revision(intel_dp);
+ intel_dp_print_sw_revision(intel_dp);
+
intel_dp_configure_mst(intel_dp);
if (intel_dp->is_mst) {
pps_lock(intel_dp);
- /*
- * Read out the current power sequencer assignment,
- * in case the BIOS did something with it.
- */
- if (IS_VALLEYVIEW(encoder->dev) || IS_CHERRYVIEW(encoder->dev))
- vlv_initial_power_sequencer_setup(intel_dp);
-
+ /* Reinit the power sequencer, in case BIOS did something with it. */
+ intel_dp_pps_init(encoder->dev, intel_dp);
intel_edp_panel_vdd_sanitize(intel_dp);
pps_unlock(intel_dp);
I915_READ(regs.pp_div));
}
+static void intel_dp_pps_init(struct drm_device *dev,
+ struct intel_dp *intel_dp)
+{
+ if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) {
+ vlv_initial_power_sequencer_setup(intel_dp);
+ } else {
+ intel_dp_init_panel_power_sequencer(dev, intel_dp);
+ intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
+ }
+}
+
/**
* intel_dp_set_drrs_state - program registers for RR switch to take effect
- * @dev: DRM device
+ * @dev_priv: i915 device
+ * @crtc_state: a pointer to the active intel_crtc_state
* @refresh_rate: RR to be programmed
*
* This function gets called when refresh rate (RR) has to be changed from
*
* The caller of this function needs to take a lock on dev_priv->drrs.
*/
-static void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate)
+static void intel_dp_set_drrs_state(struct drm_i915_private *dev_priv,
+ struct intel_crtc_state *crtc_state,
+ int refresh_rate)
{
- struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_encoder *encoder;
struct intel_digital_port *dig_port = NULL;
struct intel_dp *intel_dp = dev_priv->drrs.dp;
- struct intel_crtc_state *config = NULL;
- struct intel_crtc *intel_crtc = NULL;
+ struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
enum drrs_refresh_rate_type index = DRRS_HIGH_RR;
if (refresh_rate <= 0) {
return;
}
- config = intel_crtc->config;
-
if (dev_priv->drrs.type < SEAMLESS_DRRS_SUPPORT) {
DRM_DEBUG_KMS("Only Seamless DRRS supported.\n");
return;
return;
}
- if (!intel_crtc->active) {
+ if (!crtc_state->base.active) {
DRM_DEBUG_KMS("eDP encoder disabled. CRTC not Active\n");
return;
}
- if (INTEL_INFO(dev)->gen >= 8 && !IS_CHERRYVIEW(dev)) {
+ if (INTEL_GEN(dev_priv) >= 8 && !IS_CHERRYVIEW(dev_priv)) {
switch (index) {
case DRRS_HIGH_RR:
intel_dp_set_m_n(intel_crtc, M1_N1);
default:
DRM_ERROR("Unsupported refreshrate type\n");
}
- } else if (INTEL_INFO(dev)->gen > 6) {
- i915_reg_t reg = PIPECONF(intel_crtc->config->cpu_transcoder);
+ } else if (INTEL_GEN(dev_priv) > 6) {
+ i915_reg_t reg = PIPECONF(crtc_state->cpu_transcoder);
u32 val;
val = I915_READ(reg);
if (index > DRRS_HIGH_RR) {
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
val |= PIPECONF_EDP_RR_MODE_SWITCH_VLV;
else
val |= PIPECONF_EDP_RR_MODE_SWITCH;
} else {
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev))
+ if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
val &= ~PIPECONF_EDP_RR_MODE_SWITCH_VLV;
else
val &= ~PIPECONF_EDP_RR_MODE_SWITCH;
/**
* intel_edp_drrs_enable - init drrs struct if supported
* @intel_dp: DP struct
+ * @crtc_state: A pointer to the active crtc state.
*
* Initializes frontbuffer_bits and drrs.dp
*/
-void intel_edp_drrs_enable(struct intel_dp *intel_dp)
+void intel_edp_drrs_enable(struct intel_dp *intel_dp,
+ struct intel_crtc_state *crtc_state)
{
struct drm_device *dev = intel_dp_to_dev(intel_dp);
struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
- struct drm_crtc *crtc = dig_port->base.base.crtc;
- struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- if (!intel_crtc->config->has_drrs) {
+ if (!crtc_state->has_drrs) {
DRM_DEBUG_KMS("Panel doesn't support DRRS\n");
return;
}
/**
* intel_edp_drrs_disable - Disable DRRS
* @intel_dp: DP struct
+ * @old_crtc_state: Pointer to old crtc_state.
*
*/
-void intel_edp_drrs_disable(struct intel_dp *intel_dp)
+void intel_edp_drrs_disable(struct intel_dp *intel_dp,
+ struct intel_crtc_state *old_crtc_state)
{
struct drm_device *dev = intel_dp_to_dev(intel_dp);
struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
- struct drm_crtc *crtc = dig_port->base.base.crtc;
- struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
- if (!intel_crtc->config->has_drrs)
+ if (!old_crtc_state->has_drrs)
return;
mutex_lock(&dev_priv->drrs.mutex);
}
if (dev_priv->drrs.refresh_rate_type == DRRS_LOW_RR)
- intel_dp_set_drrs_state(&dev_priv->drm,
- intel_dp->attached_connector->panel.
- fixed_mode->vrefresh);
+ intel_dp_set_drrs_state(dev_priv, old_crtc_state,
+ intel_dp->attached_connector->panel.fixed_mode->vrefresh);
dev_priv->drrs.dp = NULL;
mutex_unlock(&dev_priv->drrs.mutex);
if (dev_priv->drrs.busy_frontbuffer_bits)
goto unlock;
- if (dev_priv->drrs.refresh_rate_type != DRRS_LOW_RR)
- intel_dp_set_drrs_state(&dev_priv->drm,
- intel_dp->attached_connector->panel.
- downclock_mode->vrefresh);
+ if (dev_priv->drrs.refresh_rate_type != DRRS_LOW_RR) {
+ struct drm_crtc *crtc = dp_to_dig_port(intel_dp)->base.base.crtc;
+
+ intel_dp_set_drrs_state(dev_priv, to_intel_crtc(crtc)->config,
+ intel_dp->attached_connector->panel.downclock_mode->vrefresh);
+ }
unlock:
mutex_unlock(&dev_priv->drrs.mutex);
/* invalidate means busy screen hence upclock */
if (frontbuffer_bits && dev_priv->drrs.refresh_rate_type == DRRS_LOW_RR)
- intel_dp_set_drrs_state(&dev_priv->drm,
- dev_priv->drrs.dp->attached_connector->panel.
- fixed_mode->vrefresh);
+ intel_dp_set_drrs_state(dev_priv, to_intel_crtc(crtc)->config,
+ dev_priv->drrs.dp->attached_connector->panel.fixed_mode->vrefresh);
mutex_unlock(&dev_priv->drrs.mutex);
}
/* flush means busy screen hence upclock */
if (frontbuffer_bits && dev_priv->drrs.refresh_rate_type == DRRS_LOW_RR)
- intel_dp_set_drrs_state(&dev_priv->drm,
- dev_priv->drrs.dp->attached_connector->panel.
- fixed_mode->vrefresh);
+ intel_dp_set_drrs_state(dev_priv, to_intel_crtc(crtc)->config,
+ dev_priv->drrs.dp->attached_connector->panel.fixed_mode->vrefresh);
/*
* flush also means no more activity hence schedule downclock, if all
pps_lock(intel_dp);
intel_dp_init_panel_power_timestamps(intel_dp);
-
- if (IS_VALLEYVIEW(dev) || IS_CHERRYVIEW(dev)) {
- vlv_initial_power_sequencer_setup(intel_dp);
- } else {
- intel_dp_init_panel_power_sequencer(dev, intel_dp);
- intel_dp_init_panel_power_sequencer_registers(dev, intel_dp);
- }
-
+ intel_dp_pps_init(dev, intel_dp);
intel_edp_panel_vdd_sanitize(intel_dp);
pps_unlock(intel_dp);
connector->interlace_allowed = true;
connector->doublescan_allowed = 0;
- intel_dp_aux_init(intel_dp, intel_connector);
+ intel_dp_aux_init(intel_dp);
INIT_DELAYED_WORK(&intel_dp->panel_vdd_work,
edp_panel_vdd_work);
#include "intel_drv.h"
+static void
+intel_dp_dump_link_status(const uint8_t link_status[DP_LINK_STATUS_SIZE])
+{
+
+ DRM_DEBUG_KMS("ln0_1:0x%x ln2_3:0x%x align:0x%x sink:0x%x adj_req0_1:0x%x adj_req2_3:0x%x",
+ link_status[0], link_status[1], link_status[2],
+ link_status[3], link_status[4], link_status[5]);
+}
+
static void
intel_get_adjust_train(struct intel_dp *intel_dp,
const uint8_t link_status[DP_LINK_STATUS_SIZE])
return ret == intel_dp->lane_count;
}
+static bool intel_dp_link_max_vswing_reached(struct intel_dp *intel_dp)
+{
+ int lane;
+
+ for (lane = 0; lane < intel_dp->lane_count; lane++)
+ if ((intel_dp->train_set[lane] &
+ DP_TRAIN_MAX_SWING_REACHED) == 0)
+ return false;
+
+ return true;
+}
+
/* Enable corresponding port and start training pattern 1 */
-static void
+static bool
intel_dp_link_training_clock_recovery(struct intel_dp *intel_dp)
{
- int i;
uint8_t voltage;
- int voltage_tries, loop_tries;
+ int voltage_tries, max_vswing_tries;
uint8_t link_config[2];
uint8_t link_bw, rate_select;
if (drm_dp_enhanced_frame_cap(intel_dp->dpcd))
link_config[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_BW_SET, link_config, 2);
+
if (intel_dp->num_sink_rates)
drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_RATE_SET,
&rate_select, 1);
DP_TRAINING_PATTERN_1 |
DP_LINK_SCRAMBLING_DISABLE)) {
DRM_ERROR("failed to enable link training\n");
- return;
+ return false;
}
- voltage = 0xff;
- voltage_tries = 0;
- loop_tries = 0;
+ voltage_tries = 1;
+ max_vswing_tries = 0;
for (;;) {
uint8_t link_status[DP_LINK_STATUS_SIZE];
drm_dp_link_train_clock_recovery_delay(intel_dp->dpcd);
+
if (!intel_dp_get_link_status(intel_dp, link_status)) {
DRM_ERROR("failed to get link status\n");
- break;
+ return false;
}
if (drm_dp_clock_recovery_ok(link_status, intel_dp->lane_count)) {
DRM_DEBUG_KMS("clock recovery OK\n");
- break;
+ return true;
}
- /* Check to see if we've tried the max voltage */
- for (i = 0; i < intel_dp->lane_count; i++)
- if ((intel_dp->train_set[i] & DP_TRAIN_MAX_SWING_REACHED) == 0)
- break;
- if (i == intel_dp->lane_count) {
- ++loop_tries;
- if (loop_tries == 5) {
- DRM_ERROR("too many full retries, give up\n");
- break;
- }
- intel_dp_reset_link_train(intel_dp,
- DP_TRAINING_PATTERN_1 |
- DP_LINK_SCRAMBLING_DISABLE);
- voltage_tries = 0;
- continue;
+ if (voltage_tries == 5) {
+ DRM_DEBUG_KMS("Same voltage tried 5 times\n");
+ return false;
+ }
+
+ if (max_vswing_tries == 1) {
+ DRM_DEBUG_KMS("Max Voltage Swing reached\n");
+ return false;
}
- /* Check to see if we've tried the same voltage 5 times */
- if ((intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK) == voltage) {
- ++voltage_tries;
- if (voltage_tries == 5) {
- DRM_ERROR("too many voltage retries, give up\n");
- break;
- }
- } else
- voltage_tries = 0;
voltage = intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK;
/* Update training set as requested by target */
intel_get_adjust_train(intel_dp, link_status);
if (!intel_dp_update_link_train(intel_dp)) {
DRM_ERROR("failed to update link training\n");
- break;
+ return false;
}
+
+ if ((intel_dp->train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK) ==
+ voltage)
+ ++voltage_tries;
+ else
+ voltage_tries = 1;
+
+ if (intel_dp_link_max_vswing_reached(intel_dp))
+ ++max_vswing_tries;
+
}
}
return training_pattern;
}
-static void
+static bool
intel_dp_link_training_channel_equalization(struct intel_dp *intel_dp)
{
- bool channel_eq = false;
- int tries, cr_tries;
+ int tries;
u32 training_pattern;
+ uint8_t link_status[DP_LINK_STATUS_SIZE];
training_pattern = intel_dp_training_pattern(intel_dp);
training_pattern |
DP_LINK_SCRAMBLING_DISABLE)) {
DRM_ERROR("failed to start channel equalization\n");
- return;
+ return false;
}
- tries = 0;
- cr_tries = 0;
- channel_eq = false;
- for (;;) {
- uint8_t link_status[DP_LINK_STATUS_SIZE];
-
- if (cr_tries > 5) {
- DRM_ERROR("failed to train DP, aborting\n");
- break;
- }
+ intel_dp->channel_eq_status = false;
+ for (tries = 0; tries < 5; tries++) {
drm_dp_link_train_channel_eq_delay(intel_dp->dpcd);
if (!intel_dp_get_link_status(intel_dp, link_status)) {
/* Make sure clock is still ok */
if (!drm_dp_clock_recovery_ok(link_status,
intel_dp->lane_count)) {
- intel_dp_link_training_clock_recovery(intel_dp);
- intel_dp_set_link_train(intel_dp,
- training_pattern |
- DP_LINK_SCRAMBLING_DISABLE);
- cr_tries++;
- continue;
+ intel_dp_dump_link_status(link_status);
+ DRM_DEBUG_KMS("Clock recovery check failed, cannot "
+ "continue channel equalization\n");
+ break;
}
if (drm_dp_channel_eq_ok(link_status,
intel_dp->lane_count)) {
- channel_eq = true;
+ intel_dp->channel_eq_status = true;
+ DRM_DEBUG_KMS("Channel EQ done. DP Training "
+ "successful\n");
break;
}
- /* Try 5 times, then try clock recovery if that fails */
- if (tries > 5) {
- intel_dp_link_training_clock_recovery(intel_dp);
- intel_dp_set_link_train(intel_dp,
- training_pattern |
- DP_LINK_SCRAMBLING_DISABLE);
- tries = 0;
- cr_tries++;
- continue;
- }
-
/* Update training set as requested by target */
intel_get_adjust_train(intel_dp, link_status);
if (!intel_dp_update_link_train(intel_dp)) {
DRM_ERROR("failed to update link training\n");
break;
}
- ++tries;
+ }
+
+ /* Try 5 times, else fail and try at lower BW */
+ if (tries == 5) {
+ intel_dp_dump_link_status(link_status);
+ DRM_DEBUG_KMS("Channel equalization failed 5 times\n");
}
intel_dp_set_idle_link_train(intel_dp);
- if (channel_eq)
- DRM_DEBUG_KMS("Channel EQ done. DP Training successful\n");
+ return intel_dp->channel_eq_status;
+
}
void intel_dp_stop_link_train(struct intel_dp *intel_dp)
#include <drm/drm_edid.h>
static bool intel_dp_mst_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
struct intel_digital_port *intel_dig_port = intel_mst->primary;
struct intel_dp *intel_dp = &intel_dig_port->dp;
struct drm_atomic_state *state;
- int bpp, i;
+ int bpp;
int lane_count, slots;
const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
- struct drm_connector *drm_connector;
- struct intel_connector *connector, *found = NULL;
- struct drm_connector_state *connector_state;
int mst_pbn;
pipe_config->dp_encoder_is_mst = true;
*/
lane_count = drm_dp_max_lane_count(intel_dp->dpcd);
-
pipe_config->lane_count = lane_count;
pipe_config->pipe_bpp = 24;
state = pipe_config->base.state;
- for_each_connector_in_state(state, drm_connector, connector_state, i) {
- connector = to_intel_connector(drm_connector);
-
- if (connector_state->best_encoder == &encoder->base) {
- found = connector;
- break;
- }
- }
-
- if (!found) {
- DRM_ERROR("can't find connector\n");
- return false;
- }
-
mst_pbn = drm_dp_calc_pbn_mode(adjusted_mode->crtc_clock, bpp);
pipe_config->pbn = mst_pbn;
}
-static void intel_mst_disable_dp(struct intel_encoder *encoder)
+static void intel_mst_disable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
struct intel_digital_port *intel_dig_port = intel_mst->primary;
struct intel_dp *intel_dp = &intel_dig_port->dp;
+ struct intel_connector *connector =
+ to_intel_connector(old_conn_state->connector);
int ret;
DRM_DEBUG_KMS("%d\n", intel_dp->active_mst_links);
- drm_dp_mst_reset_vcpi_slots(&intel_dp->mst_mgr, intel_mst->connector->port);
+ drm_dp_mst_reset_vcpi_slots(&intel_dp->mst_mgr, connector->port);
ret = drm_dp_update_payload_part1(&intel_dp->mst_mgr);
if (ret) {
}
}
-static void intel_mst_post_disable_dp(struct intel_encoder *encoder)
+static void intel_mst_post_disable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
struct intel_digital_port *intel_dig_port = intel_mst->primary;
struct intel_dp *intel_dp = &intel_dig_port->dp;
+ struct intel_connector *connector =
+ to_intel_connector(old_conn_state->connector);
DRM_DEBUG_KMS("%d\n", intel_dp->active_mst_links);
/* and this can also fail */
drm_dp_update_payload_part2(&intel_dp->mst_mgr);
- drm_dp_mst_deallocate_vcpi(&intel_dp->mst_mgr, intel_mst->connector->port);
+ drm_dp_mst_deallocate_vcpi(&intel_dp->mst_mgr, connector->port);
intel_dp->active_mst_links--;
intel_mst->connector = NULL;
if (intel_dp->active_mst_links == 0) {
- intel_dig_port->base.post_disable(&intel_dig_port->base);
+ intel_dig_port->base.post_disable(&intel_dig_port->base,
+ NULL, NULL);
+
intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
}
}
-static void intel_mst_pre_enable_dp(struct intel_encoder *encoder)
+static void intel_mst_pre_enable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
struct intel_digital_port *intel_dig_port = intel_mst->primary;
struct intel_dp *intel_dp = &intel_dig_port->dp;
- struct drm_device *dev = encoder->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
enum port port = intel_dig_port->port;
+ struct intel_connector *connector =
+ to_intel_connector(conn_state->connector);
int ret;
uint32_t temp;
- struct intel_connector *found = NULL, *connector;
int slots;
- struct drm_crtc *crtc = encoder->base.crtc;
- struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-
- for_each_intel_connector(dev, connector) {
- if (connector->base.state->best_encoder == &encoder->base) {
- found = connector;
- break;
- }
- }
-
- if (!found) {
- DRM_ERROR("can't find connector\n");
- return;
- }
/* MST encoders are bound to a crtc, not to a connector,
* force the mapping here for get_hw_state.
*/
- found->encoder = encoder;
+ connector->encoder = encoder;
+ intel_mst->connector = connector;
DRM_DEBUG_KMS("%d\n", intel_dp->active_mst_links);
- intel_mst->connector = found;
-
if (intel_dp->active_mst_links == 0) {
- intel_ddi_clk_select(&intel_dig_port->base, intel_crtc->config);
+ intel_ddi_clk_select(&intel_dig_port->base,
+ pipe_config->shared_dpll);
intel_prepare_dp_ddi_buffers(&intel_dig_port->base);
-
- intel_dp_set_link_params(intel_dp, intel_crtc->config);
+ intel_dp_set_link_params(intel_dp,
+ pipe_config->port_clock,
+ pipe_config->lane_count,
+ true);
intel_ddi_init_dp_buf_reg(&intel_dig_port->base);
}
ret = drm_dp_mst_allocate_vcpi(&intel_dp->mst_mgr,
- intel_mst->connector->port,
- intel_crtc->config->pbn, &slots);
+ connector->port,
+ pipe_config->pbn, &slots);
if (ret == false) {
DRM_ERROR("failed to allocate vcpi\n");
return;
ret = drm_dp_update_payload_part1(&intel_dp->mst_mgr);
}
-static void intel_mst_enable_dp(struct intel_encoder *encoder)
+static void intel_mst_enable_dp(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
struct intel_digital_port *intel_dig_port = intel_mst->primary;
struct intel_dp *intel_dp = &intel_dig_port->dp;
- struct drm_device *dev = intel_dig_port->base.base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
enum port port = intel_dig_port->port;
int ret;
{
struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
struct intel_digital_port *intel_dig_port = intel_mst->primary;
- struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
- struct drm_device *dev = encoder->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
enum transcoder cpu_transcoder = pipe_config->cpu_transcoder;
u32 temp, flags = 0;
#include "intel_drv.h"
+struct intel_shared_dpll *
+skl_find_link_pll(struct drm_i915_private *dev_priv, int clock)
+{
+ struct intel_shared_dpll *pll = NULL;
+ struct intel_dpll_hw_state dpll_hw_state;
+ enum intel_dpll_id i;
+ bool found = false;
+
+ if (!skl_ddi_dp_set_dpll_hw_state(clock, &dpll_hw_state))
+ return pll;
+
+ for (i = DPLL_ID_SKL_DPLL1; i <= DPLL_ID_SKL_DPLL3; i++) {
+ pll = &dev_priv->shared_dplls[i];
+
+ /* Only want to check enabled timings first */
+ if (pll->config.crtc_mask == 0)
+ continue;
+
+ if (memcmp(&dpll_hw_state, &pll->config.hw_state,
+ sizeof(pll->config.hw_state)) == 0) {
+ found = true;
+ break;
+ }
+ }
+
+ /* Ok no matching timings, maybe there's a free one? */
+ for (i = DPLL_ID_SKL_DPLL1;
+ ((found == false) && (i <= DPLL_ID_SKL_DPLL3)); i++) {
+ pll = &dev_priv->shared_dplls[i];
+ if (pll->config.crtc_mask == 0) {
+ pll->config.hw_state = dpll_hw_state;
+ break;
+ }
+ }
+
+ return pll;
+}
+
struct intel_shared_dpll *
intel_get_shared_dpll_by_id(struct drm_i915_private *dev_priv,
enum intel_dpll_id id)
return val & SPLL_PLL_ENABLE;
}
-static uint32_t hsw_pll_to_ddi_pll_sel(struct intel_shared_dpll *pll)
-{
- switch (pll->id) {
- case DPLL_ID_WRPLL1:
- return PORT_CLK_SEL_WRPLL1;
- case DPLL_ID_WRPLL2:
- return PORT_CLK_SEL_WRPLL2;
- case DPLL_ID_SPLL:
- return PORT_CLK_SEL_SPLL;
- case DPLL_ID_LCPLL_810:
- return PORT_CLK_SEL_LCPLL_810;
- case DPLL_ID_LCPLL_1350:
- return PORT_CLK_SEL_LCPLL_1350;
- case DPLL_ID_LCPLL_2700:
- return PORT_CLK_SEL_LCPLL_2700;
- default:
- return PORT_CLK_SEL_NONE;
- }
-}
-
#define LC_FREQ 2700
#define LC_FREQ_2K U64_C(LC_FREQ * 2000)
*r2_out = best.r2;
}
+static struct intel_shared_dpll *hsw_ddi_hdmi_get_dpll(int clock,
+ struct intel_crtc *crtc,
+ struct intel_crtc_state *crtc_state)
+{
+ struct intel_shared_dpll *pll;
+ uint32_t val;
+ unsigned int p, n2, r2;
+
+ hsw_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
+
+ val = WRPLL_PLL_ENABLE | WRPLL_PLL_LCPLL |
+ WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) |
+ WRPLL_DIVIDER_POST(p);
+
+ crtc_state->dpll_hw_state.wrpll = val;
+
+ pll = intel_find_shared_dpll(crtc, crtc_state,
+ DPLL_ID_WRPLL1, DPLL_ID_WRPLL2);
+
+ if (!pll)
+ return NULL;
+
+ return pll;
+}
+
+struct intel_shared_dpll *hsw_ddi_dp_get_dpll(struct intel_encoder *encoder,
+ int clock)
+{
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+ struct intel_shared_dpll *pll;
+ enum intel_dpll_id pll_id;
+
+ switch (clock / 2) {
+ case 81000:
+ pll_id = DPLL_ID_LCPLL_810;
+ break;
+ case 135000:
+ pll_id = DPLL_ID_LCPLL_1350;
+ break;
+ case 270000:
+ pll_id = DPLL_ID_LCPLL_2700;
+ break;
+ default:
+ DRM_DEBUG_KMS("Invalid clock for DP: %d\n", clock);
+ return NULL;
+ }
+
+ pll = intel_get_shared_dpll_by_id(dev_priv, pll_id);
+
+ if (!pll)
+ return NULL;
+
+ return pll;
+}
+
static struct intel_shared_dpll *
hsw_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state,
struct intel_encoder *encoder)
{
- struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
struct intel_shared_dpll *pll;
int clock = crtc_state->port_clock;
sizeof(crtc_state->dpll_hw_state));
if (encoder->type == INTEL_OUTPUT_HDMI) {
- uint32_t val;
- unsigned p, n2, r2;
-
- hsw_ddi_calculate_wrpll(clock * 1000, &r2, &n2, &p);
-
- val = WRPLL_PLL_ENABLE | WRPLL_PLL_LCPLL |
- WRPLL_DIVIDER_REFERENCE(r2) | WRPLL_DIVIDER_FEEDBACK(n2) |
- WRPLL_DIVIDER_POST(p);
-
- crtc_state->dpll_hw_state.wrpll = val;
-
- pll = intel_find_shared_dpll(crtc, crtc_state,
- DPLL_ID_WRPLL1, DPLL_ID_WRPLL2);
+ pll = hsw_ddi_hdmi_get_dpll(clock, crtc, crtc_state);
} else if (encoder->type == INTEL_OUTPUT_DP ||
encoder->type == INTEL_OUTPUT_DP_MST ||
encoder->type == INTEL_OUTPUT_EDP) {
- enum intel_dpll_id pll_id;
-
- switch (clock / 2) {
- case 81000:
- pll_id = DPLL_ID_LCPLL_810;
- break;
- case 135000:
- pll_id = DPLL_ID_LCPLL_1350;
- break;
- case 270000:
- pll_id = DPLL_ID_LCPLL_2700;
- break;
- default:
- DRM_DEBUG_KMS("Invalid clock for DP: %d\n", clock);
- return NULL;
- }
-
- pll = intel_get_shared_dpll_by_id(dev_priv, pll_id);
+ pll = hsw_ddi_dp_get_dpll(encoder, clock);
} else if (encoder->type == INTEL_OUTPUT_ANALOG) {
if (WARN_ON(crtc_state->port_clock / 2 != 135000))
if (!pll)
return NULL;
- crtc_state->ddi_pll_sel = hsw_pll_to_ddi_pll_sel(pll);
-
intel_reference_shared_dpll(pll, crtc_state);
return pll;
}
-
static const struct intel_shared_dpll_funcs hsw_ddi_wrpll_funcs = {
.enable = hsw_ddi_wrpll_enable,
.disable = hsw_ddi_wrpll_disable,
return true;
}
-static struct intel_shared_dpll *
-skl_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state,
- struct intel_encoder *encoder)
+static bool skl_ddi_hdmi_pll_dividers(struct intel_crtc *crtc,
+ struct intel_crtc_state *crtc_state,
+ int clock)
{
- struct intel_shared_dpll *pll;
uint32_t ctrl1, cfgcr1, cfgcr2;
- int clock = crtc_state->port_clock;
+ struct skl_wrpll_params wrpll_params = { 0, };
/*
* See comment in intel_dpll_hw_state to understand why we always use 0
* as the DPLL id in this function.
*/
-
ctrl1 = DPLL_CTRL1_OVERRIDE(0);
- if (encoder->type == INTEL_OUTPUT_HDMI) {
- struct skl_wrpll_params wrpll_params = { 0, };
+ ctrl1 |= DPLL_CTRL1_HDMI_MODE(0);
- ctrl1 |= DPLL_CTRL1_HDMI_MODE(0);
+ if (!skl_ddi_calculate_wrpll(clock * 1000, &wrpll_params))
+ return false;
- if (!skl_ddi_calculate_wrpll(clock * 1000, &wrpll_params))
- return NULL;
+ cfgcr1 = DPLL_CFGCR1_FREQ_ENABLE |
+ DPLL_CFGCR1_DCO_FRACTION(wrpll_params.dco_fraction) |
+ wrpll_params.dco_integer;
+
+ cfgcr2 = DPLL_CFGCR2_QDIV_RATIO(wrpll_params.qdiv_ratio) |
+ DPLL_CFGCR2_QDIV_MODE(wrpll_params.qdiv_mode) |
+ DPLL_CFGCR2_KDIV(wrpll_params.kdiv) |
+ DPLL_CFGCR2_PDIV(wrpll_params.pdiv) |
+ wrpll_params.central_freq;
+
+ memset(&crtc_state->dpll_hw_state, 0,
+ sizeof(crtc_state->dpll_hw_state));
+
+ crtc_state->dpll_hw_state.ctrl1 = ctrl1;
+ crtc_state->dpll_hw_state.cfgcr1 = cfgcr1;
+ crtc_state->dpll_hw_state.cfgcr2 = cfgcr2;
+ return true;
+}
- cfgcr1 = DPLL_CFGCR1_FREQ_ENABLE |
- DPLL_CFGCR1_DCO_FRACTION(wrpll_params.dco_fraction) |
- wrpll_params.dco_integer;
- cfgcr2 = DPLL_CFGCR2_QDIV_RATIO(wrpll_params.qdiv_ratio) |
- DPLL_CFGCR2_QDIV_MODE(wrpll_params.qdiv_mode) |
- DPLL_CFGCR2_KDIV(wrpll_params.kdiv) |
- DPLL_CFGCR2_PDIV(wrpll_params.pdiv) |
- wrpll_params.central_freq;
+bool skl_ddi_dp_set_dpll_hw_state(int clock,
+ struct intel_dpll_hw_state *dpll_hw_state)
+{
+ uint32_t ctrl1;
+
+ /*
+ * See comment in intel_dpll_hw_state to understand why we always use 0
+ * as the DPLL id in this function.
+ */
+ ctrl1 = DPLL_CTRL1_OVERRIDE(0);
+ switch (clock / 2) {
+ case 81000:
+ ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_810, 0);
+ break;
+ case 135000:
+ ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1350, 0);
+ break;
+ case 270000:
+ ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_2700, 0);
+ break;
+ /* eDP 1.4 rates */
+ case 162000:
+ ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1620, 0);
+ break;
+ case 108000:
+ ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1080, 0);
+ break;
+ case 216000:
+ ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_2160, 0);
+ break;
+ }
+
+ dpll_hw_state->ctrl1 = ctrl1;
+ return true;
+}
+
+static struct intel_shared_dpll *
+skl_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state,
+ struct intel_encoder *encoder)
+{
+ struct intel_shared_dpll *pll;
+ int clock = crtc_state->port_clock;
+ bool bret;
+ struct intel_dpll_hw_state dpll_hw_state;
+
+ memset(&dpll_hw_state, 0, sizeof(dpll_hw_state));
+
+ if (encoder->type == INTEL_OUTPUT_HDMI) {
+ bret = skl_ddi_hdmi_pll_dividers(crtc, crtc_state, clock);
+ if (!bret) {
+ DRM_DEBUG_KMS("Could not get HDMI pll dividers.\n");
+ return NULL;
+ }
} else if (encoder->type == INTEL_OUTPUT_DP ||
encoder->type == INTEL_OUTPUT_DP_MST ||
encoder->type == INTEL_OUTPUT_EDP) {
- switch (crtc_state->port_clock / 2) {
- case 81000:
- ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_810, 0);
- break;
- case 135000:
- ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1350, 0);
- break;
- case 270000:
- ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_2700, 0);
- break;
- /* eDP 1.4 rates */
- case 162000:
- ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1620, 0);
- break;
- case 108000:
- ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_1080, 0);
- break;
- case 216000:
- ctrl1 |= DPLL_CTRL1_LINK_RATE(DPLL_CTRL1_LINK_RATE_2160, 0);
- break;
+ bret = skl_ddi_dp_set_dpll_hw_state(clock, &dpll_hw_state);
+ if (!bret) {
+ DRM_DEBUG_KMS("Could not set DP dpll HW state.\n");
+ return NULL;
}
-
- cfgcr1 = cfgcr2 = 0;
+ crtc_state->dpll_hw_state = dpll_hw_state;
} else {
return NULL;
}
- memset(&crtc_state->dpll_hw_state, 0,
- sizeof(crtc_state->dpll_hw_state));
-
- crtc_state->dpll_hw_state.ctrl1 = ctrl1;
- crtc_state->dpll_hw_state.cfgcr1 = cfgcr1;
- crtc_state->dpll_hw_state.cfgcr2 = cfgcr2;
-
if (encoder->type == INTEL_OUTPUT_EDP)
pll = intel_find_shared_dpll(crtc, crtc_state,
DPLL_ID_SKL_DPLL0,
if (!pll)
return NULL;
- crtc_state->ddi_pll_sel = pll->id;
-
intel_reference_shared_dpll(pll, crtc_state);
return pll;
uint32_t m2_frac;
bool m2_frac_en;
uint32_t n;
+
+ int vco;
};
/* pre-calculated values for DP linkrates */
{432000, 3, 1, 32, 1677722, 1, 1}
};
-static struct intel_shared_dpll *
-bxt_get_dpll(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state,
- struct intel_encoder *encoder)
+static bool
+bxt_ddi_hdmi_pll_dividers(struct intel_crtc *intel_crtc,
+ struct intel_crtc_state *crtc_state, int clock,
+ struct bxt_clk_div *clk_div)
{
- struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
- struct intel_shared_dpll *pll;
- enum intel_dpll_id i;
- struct intel_digital_port *intel_dig_port;
- struct bxt_clk_div clk_div = {0};
- int vco = 0;
- uint32_t prop_coef, int_coef, gain_ctl, targ_cnt;
- uint32_t lanestagger;
- int clock = crtc_state->port_clock;
+ struct dpll best_clock;
- if (encoder->type == INTEL_OUTPUT_HDMI) {
- struct dpll best_clock;
+ /* Calculate HDMI div */
+ /*
+ * FIXME: tie the following calculation into
+ * i9xx_crtc_compute_clock
+ */
+ if (!bxt_find_best_dpll(crtc_state, clock, &best_clock)) {
+ DRM_DEBUG_DRIVER("no PLL dividers found for clock %d pipe %c\n",
+ clock, pipe_name(intel_crtc->pipe));
+ return false;
+ }
- /* Calculate HDMI div */
- /*
- * FIXME: tie the following calculation into
- * i9xx_crtc_compute_clock
- */
- if (!bxt_find_best_dpll(crtc_state, clock, &best_clock)) {
- DRM_DEBUG_DRIVER("no PLL dividers found for clock %d pipe %c\n",
- clock, pipe_name(crtc->pipe));
- return NULL;
- }
+ clk_div->p1 = best_clock.p1;
+ clk_div->p2 = best_clock.p2;
+ WARN_ON(best_clock.m1 != 2);
+ clk_div->n = best_clock.n;
+ clk_div->m2_int = best_clock.m2 >> 22;
+ clk_div->m2_frac = best_clock.m2 & ((1 << 22) - 1);
+ clk_div->m2_frac_en = clk_div->m2_frac != 0;
- clk_div.p1 = best_clock.p1;
- clk_div.p2 = best_clock.p2;
- WARN_ON(best_clock.m1 != 2);
- clk_div.n = best_clock.n;
- clk_div.m2_int = best_clock.m2 >> 22;
- clk_div.m2_frac = best_clock.m2 & ((1 << 22) - 1);
- clk_div.m2_frac_en = clk_div.m2_frac != 0;
+ clk_div->vco = best_clock.vco;
- vco = best_clock.vco;
- } else if (encoder->type == INTEL_OUTPUT_DP ||
- encoder->type == INTEL_OUTPUT_EDP) {
- int i;
+ return true;
+}
- clk_div = bxt_dp_clk_val[0];
- for (i = 0; i < ARRAY_SIZE(bxt_dp_clk_val); ++i) {
- if (bxt_dp_clk_val[i].clock == clock) {
- clk_div = bxt_dp_clk_val[i];
- break;
- }
+static void bxt_ddi_dp_pll_dividers(int clock, struct bxt_clk_div *clk_div)
+{
+ int i;
+
+ *clk_div = bxt_dp_clk_val[0];
+ for (i = 0; i < ARRAY_SIZE(bxt_dp_clk_val); ++i) {
+ if (bxt_dp_clk_val[i].clock == clock) {
+ *clk_div = bxt_dp_clk_val[i];
+ break;
}
- vco = clock * 10 / 2 * clk_div.p1 * clk_div.p2;
}
+ clk_div->vco = clock * 10 / 2 * clk_div->p1 * clk_div->p2;
+}
+
+static bool bxt_ddi_set_dpll_hw_state(int clock,
+ struct bxt_clk_div *clk_div,
+ struct intel_dpll_hw_state *dpll_hw_state)
+{
+ int vco = clk_div->vco;
+ uint32_t prop_coef, int_coef, gain_ctl, targ_cnt;
+ uint32_t lanestagger;
+
if (vco >= 6200000 && vco <= 6700000) {
prop_coef = 4;
int_coef = 9;
targ_cnt = 9;
} else {
DRM_ERROR("Invalid VCO\n");
- return NULL;
+ return false;
}
- memset(&crtc_state->dpll_hw_state, 0,
- sizeof(crtc_state->dpll_hw_state));
-
if (clock > 270000)
lanestagger = 0x18;
else if (clock > 135000)
else
lanestagger = 0x02;
- crtc_state->dpll_hw_state.ebb0 =
- PORT_PLL_P1(clk_div.p1) | PORT_PLL_P2(clk_div.p2);
- crtc_state->dpll_hw_state.pll0 = clk_div.m2_int;
- crtc_state->dpll_hw_state.pll1 = PORT_PLL_N(clk_div.n);
- crtc_state->dpll_hw_state.pll2 = clk_div.m2_frac;
+ dpll_hw_state->ebb0 = PORT_PLL_P1(clk_div->p1) | PORT_PLL_P2(clk_div->p2);
+ dpll_hw_state->pll0 = clk_div->m2_int;
+ dpll_hw_state->pll1 = PORT_PLL_N(clk_div->n);
+ dpll_hw_state->pll2 = clk_div->m2_frac;
- if (clk_div.m2_frac_en)
- crtc_state->dpll_hw_state.pll3 =
- PORT_PLL_M2_FRAC_ENABLE;
+ if (clk_div->m2_frac_en)
+ dpll_hw_state->pll3 = PORT_PLL_M2_FRAC_ENABLE;
- crtc_state->dpll_hw_state.pll6 =
- prop_coef | PORT_PLL_INT_COEFF(int_coef);
- crtc_state->dpll_hw_state.pll6 |=
- PORT_PLL_GAIN_CTL(gain_ctl);
+ dpll_hw_state->pll6 = prop_coef | PORT_PLL_INT_COEFF(int_coef);
+ dpll_hw_state->pll6 |= PORT_PLL_GAIN_CTL(gain_ctl);
- crtc_state->dpll_hw_state.pll8 = targ_cnt;
+ dpll_hw_state->pll8 = targ_cnt;
- crtc_state->dpll_hw_state.pll9 = 5 << PORT_PLL_LOCK_THRESHOLD_SHIFT;
+ dpll_hw_state->pll9 = 5 << PORT_PLL_LOCK_THRESHOLD_SHIFT;
- crtc_state->dpll_hw_state.pll10 =
+ dpll_hw_state->pll10 =
PORT_PLL_DCO_AMP(PORT_PLL_DCO_AMP_DEFAULT)
| PORT_PLL_DCO_AMP_OVR_EN_H;
- crtc_state->dpll_hw_state.ebb4 = PORT_PLL_10BIT_CLK_ENABLE;
+ dpll_hw_state->ebb4 = PORT_PLL_10BIT_CLK_ENABLE;
- crtc_state->dpll_hw_state.pcsdw12 =
- LANESTAGGER_STRAP_OVRD | lanestagger;
+ dpll_hw_state->pcsdw12 = LANESTAGGER_STRAP_OVRD | lanestagger;
- intel_dig_port = enc_to_dig_port(&encoder->base);
+ return true;
+}
+
+bool bxt_ddi_dp_set_dpll_hw_state(int clock,
+ struct intel_dpll_hw_state *dpll_hw_state)
+{
+ struct bxt_clk_div clk_div = {0};
+
+ bxt_ddi_dp_pll_dividers(clock, &clk_div);
+
+ return bxt_ddi_set_dpll_hw_state(clock, &clk_div, dpll_hw_state);
+}
+
+static struct intel_shared_dpll *
+bxt_get_dpll(struct intel_crtc *crtc,
+ struct intel_crtc_state *crtc_state,
+ struct intel_encoder *encoder)
+{
+ struct bxt_clk_div clk_div = {0};
+ struct intel_dpll_hw_state dpll_hw_state = {0};
+ struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+ struct intel_digital_port *intel_dig_port;
+ struct intel_shared_dpll *pll;
+ int i, clock = crtc_state->port_clock;
+
+ if (encoder->type == INTEL_OUTPUT_HDMI
+ && !bxt_ddi_hdmi_pll_dividers(crtc, crtc_state,
+ clock, &clk_div))
+ return NULL;
+
+ if ((encoder->type == INTEL_OUTPUT_DP ||
+ encoder->type == INTEL_OUTPUT_EDP) &&
+ !bxt_ddi_dp_set_dpll_hw_state(clock, &dpll_hw_state))
+ return NULL;
+
+ memset(&crtc_state->dpll_hw_state, 0,
+ sizeof(crtc_state->dpll_hw_state));
+
+ crtc_state->dpll_hw_state = dpll_hw_state;
+
+ if (encoder->type == INTEL_OUTPUT_DP_MST) {
+ struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
+
+ intel_dig_port = intel_mst->primary;
+ } else
+ intel_dig_port = enc_to_dig_port(&encoder->base);
/* 1:1 mapping between ports and PLLs */
i = (enum intel_dpll_id) intel_dig_port->port;
intel_reference_shared_dpll(pll, crtc_state);
- /* shared DPLL id 0 is DPLL A */
- crtc_state->ddi_pll_sel = pll->id;
-
return pll;
}
void intel_shared_dpll_commit(struct drm_atomic_state *state);
void intel_shared_dpll_init(struct drm_device *dev);
+/* BXT dpll related functions */
+bool bxt_ddi_dp_set_dpll_hw_state(int clock,
+ struct intel_dpll_hw_state *dpll_hw_state);
+
+
+/* SKL dpll related functions */
+bool skl_ddi_dp_set_dpll_hw_state(int clock,
+ struct intel_dpll_hw_state *dpll_hw_state);
+struct intel_shared_dpll *skl_find_link_pll(struct drm_i915_private *dev_priv,
+ int clock);
+
+
+/* HSW dpll related functions */
+struct intel_shared_dpll *hsw_ddi_dp_get_dpll(struct intel_encoder *encoder,
+ int clock);
#endif /* _INTEL_DPLL_MGR_H_ */
*/
#define _wait_for(COND, US, W) ({ \
unsigned long timeout__ = jiffies + usecs_to_jiffies(US) + 1; \
- int ret__ = 0; \
- while (!(COND)) { \
- if (time_after(jiffies, timeout__)) { \
- if (!(COND)) \
- ret__ = -ETIMEDOUT; \
+ int ret__; \
+ for (;;) { \
+ bool expired__ = time_after(jiffies, timeout__); \
+ if (COND) { \
+ ret__ = 0; \
+ break; \
+ } \
+ if (expired__) { \
+ ret__ = -ETIMEDOUT; \
break; \
} \
if ((W) && drm_can_sleep()) { \
struct drm_framebuffer base;
struct drm_i915_gem_object *obj;
struct intel_rotation_info rot_info;
+
+ /* for each plane in the normal GTT view */
+ struct {
+ unsigned int x, y;
+ } normal[2];
+ /* for each plane in the rotated GTT view */
+ struct {
+ unsigned int x, y;
+ unsigned int pitch; /* pixels */
+ } rotated[2];
};
struct intel_fbdev {
struct drm_fb_helper helper;
struct intel_framebuffer *fb;
+ struct i915_vma *vma;
async_cookie_t cookie;
int preferred_bpp;
};
unsigned int cloneable;
void (*hot_plug)(struct intel_encoder *);
bool (*compute_config)(struct intel_encoder *,
- struct intel_crtc_state *);
- void (*pre_pll_enable)(struct intel_encoder *);
- void (*pre_enable)(struct intel_encoder *);
- void (*enable)(struct intel_encoder *);
- void (*mode_set)(struct intel_encoder *intel_encoder);
- void (*disable)(struct intel_encoder *);
- void (*post_disable)(struct intel_encoder *);
- void (*post_pll_disable)(struct intel_encoder *);
+ struct intel_crtc_state *,
+ struct drm_connector_state *);
+ void (*pre_pll_enable)(struct intel_encoder *,
+ struct intel_crtc_state *,
+ struct drm_connector_state *);
+ void (*pre_enable)(struct intel_encoder *,
+ struct intel_crtc_state *,
+ struct drm_connector_state *);
+ void (*enable)(struct intel_encoder *,
+ struct intel_crtc_state *,
+ struct drm_connector_state *);
+ void (*disable)(struct intel_encoder *,
+ struct intel_crtc_state *,
+ struct drm_connector_state *);
+ void (*post_disable)(struct intel_encoder *,
+ struct intel_crtc_state *,
+ struct drm_connector_state *);
+ void (*post_pll_disable)(struct intel_encoder *,
+ struct intel_crtc_state *,
+ struct drm_connector_state *);
/* Read out the current hw state of this connector, returning true if
* the encoder is active. If the encoder is enabled it also set the pipe
* it is connected to in the pipe parameter. */
struct drm_plane_state base;
struct drm_rect clip;
+ struct {
+ u32 offset;
+ int x, y;
+ } main;
+ struct {
+ u32 offset;
+ int x, y;
+ } aux;
+
/*
* scaler_id
* = -1 : not using a scaler
/* Selected dpll when shared or NULL. */
struct intel_shared_dpll *shared_dpll;
- /*
- * - PORT_CLK_SEL for DDI ports on HSW/BDW.
- * - enum skl_dpll on SKL
- */
- uint32_t ddi_pll_sel;
-
/* Actual register state of the dpll, for shared dpll cross-checking. */
struct intel_dpll_hw_state dpll_hw_state;
struct intel_crtc_state *config;
- /* reset counter value when the last flip was submitted */
- unsigned int reset_counter;
+ /* global reset count when the last flip was submitted */
+ unsigned int reset_count;
/* Access to these should be protected by dev_priv->irq_lock. */
bool cpu_fifo_underrun_disabled;
bool link_mst;
bool has_audio;
bool detect_done;
+ bool channel_eq_status;
enum hdmi_force_audio force_audio;
bool limited_color_range;
bool color_range_auto;
/* intel_ddi.c */
void intel_ddi_clk_select(struct intel_encoder *encoder,
- const struct intel_crtc_state *pipe_config);
+ struct intel_shared_dpll *pll);
+void intel_ddi_fdi_post_disable(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state);
void intel_prepare_dp_ddi_buffers(struct intel_encoder *encoder);
void hsw_fdi_link_train(struct drm_crtc *crtc);
void intel_ddi_init(struct drm_device *dev, enum port port);
void intel_ddi_set_pipe_settings(struct drm_crtc *crtc);
void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp);
bool intel_ddi_connector_get_hw_state(struct intel_connector *intel_connector);
-void intel_ddi_fdi_disable(struct drm_crtc *crtc);
void intel_ddi_get_config(struct intel_encoder *encoder,
struct intel_crtc_state *pipe_config);
struct intel_encoder *
struct intel_crtc_state *pipe_config);
void intel_ddi_set_vc_payload_alloc(struct drm_crtc *crtc, bool state);
uint32_t ddi_signal_levels(struct intel_dp *intel_dp);
-
+struct intel_shared_dpll *intel_ddi_get_link_dpll(struct intel_dp *intel_dp,
+ int clock);
unsigned int intel_fb_align_height(struct drm_device *dev,
unsigned int height,
uint32_t pixel_format,
void intel_update_rawclk(struct drm_i915_private *dev_priv);
int vlv_get_cck_clock(struct drm_i915_private *dev_priv,
const char *name, u32 reg, int ref_freq);
+void lpt_disable_pch_transcoder(struct drm_i915_private *dev_priv);
+void lpt_disable_iclkip(struct drm_i915_private *dev_priv);
extern const struct drm_plane_funcs intel_plane_funcs;
void intel_init_display_hooks(struct drm_i915_private *dev_priv);
+unsigned int intel_fb_xy_to_linear(int x, int y,
+ const struct intel_plane_state *state,
+ int plane);
+void intel_add_fb_offsets(int *x, int *y,
+ const struct intel_plane_state *state, int plane);
unsigned int intel_rotation_info_size(const struct intel_rotation_info *rot_info);
bool intel_has_pending_fb_unpin(struct drm_device *dev);
void intel_mark_busy(struct drm_i915_private *dev_priv);
void intel_mark_idle(struct drm_i915_private *dev_priv);
void intel_crtc_restore_mode(struct drm_crtc *crtc);
int intel_display_suspend(struct drm_device *dev);
+void intel_pps_unlock_regs_wa(struct drm_i915_private *dev_priv);
void intel_encoder_destroy(struct drm_encoder *encoder);
int intel_connector_init(struct intel_connector *);
struct intel_connector *intel_connector_alloc(void);
void intel_release_load_detect_pipe(struct drm_connector *connector,
struct intel_load_detect_pipe *old,
struct drm_modeset_acquire_ctx *ctx);
-int intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb,
- unsigned int rotation);
+struct i915_vma *
+intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb, unsigned int rotation);
void intel_unpin_fb_obj(struct drm_framebuffer *fb, unsigned int rotation);
struct drm_framebuffer *
__intel_framebuffer_create(struct drm_device *dev,
#define assert_pipe_enabled(d, p) assert_pipe(d, p, true)
#define assert_pipe_disabled(d, p) assert_pipe(d, p, false)
u32 intel_compute_tile_offset(int *x, int *y,
- const struct drm_framebuffer *fb, int plane,
- unsigned int pitch,
- unsigned int rotation);
+ const struct intel_plane_state *state, int plane);
void intel_prepare_reset(struct drm_i915_private *dev_priv);
void intel_finish_reset(struct drm_i915_private *dev_priv);
void hsw_enable_pc8(struct drm_i915_private *dev_priv);
int skl_update_scaler_crtc(struct intel_crtc_state *crtc_state);
int skl_max_scale(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state);
-u32 intel_plane_obj_offset(struct intel_plane *intel_plane,
- struct drm_i915_gem_object *obj,
- unsigned int plane);
+u32 intel_fb_gtt_offset(struct drm_framebuffer *fb, unsigned int rotation);
u32 skl_plane_ctl_format(uint32_t pixel_format);
u32 skl_plane_ctl_tiling(uint64_t fb_modifier);
u32 skl_plane_ctl_rotation(unsigned int rotation);
+u32 skl_plane_stride(const struct drm_framebuffer *fb, int plane,
+ unsigned int rotation);
+int skl_check_plane_surface(struct intel_plane_state *plane_state);
/* intel_csr.c */
void intel_csr_ucode_init(struct drm_i915_private *);
bool intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
struct intel_connector *intel_connector);
void intel_dp_set_link_params(struct intel_dp *intel_dp,
- const struct intel_crtc_state *pipe_config);
+ int link_rate, uint8_t lane_count,
+ bool link_mst);
void intel_dp_start_link_train(struct intel_dp *intel_dp);
void intel_dp_stop_link_train(struct intel_dp *intel_dp);
void intel_dp_sink_dpms(struct intel_dp *intel_dp, int mode);
void intel_dp_encoder_destroy(struct drm_encoder *encoder);
int intel_dp_sink_crc(struct intel_dp *intel_dp, u8 *crc);
bool intel_dp_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config);
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state);
bool intel_dp_is_edp(struct drm_device *dev, enum port port);
enum irqreturn intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port,
bool long_hpd);
void intel_power_sequencer_reset(struct drm_i915_private *dev_priv);
uint32_t intel_dp_pack_aux(const uint8_t *src, int src_bytes);
void intel_plane_destroy(struct drm_plane *plane);
-void intel_edp_drrs_enable(struct intel_dp *intel_dp);
-void intel_edp_drrs_disable(struct intel_dp *intel_dp);
+void intel_edp_drrs_enable(struct intel_dp *intel_dp,
+ struct intel_crtc_state *crtc_state);
+void intel_edp_drrs_disable(struct intel_dp *intel_dp,
+ struct intel_crtc_state *crtc_state);
void intel_edp_drrs_invalidate(struct drm_i915_private *dev_priv,
unsigned int frontbuffer_bits);
void intel_edp_drrs_flush(struct drm_i915_private *dev_priv,
unsigned int frontbuffer_bits);
-bool intel_digital_port_connected(struct drm_i915_private *dev_priv,
- struct intel_digital_port *port);
void
intel_dp_program_link_training_pattern(struct intel_dp *intel_dp,
struct intel_connector *intel_connector);
struct intel_hdmi *enc_to_intel_hdmi(struct drm_encoder *encoder);
bool intel_hdmi_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config);
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state);
void intel_dp_dual_mode_set_tmds_output(struct intel_hdmi *hdmi, bool enable);
void skl_wm_get_hw_state(struct drm_device *dev);
void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv,
struct skl_ddb_allocation *ddb /* out */);
+bool skl_can_enable_sagv(struct drm_atomic_state *state);
+int skl_enable_sagv(struct drm_i915_private *dev_priv);
+int skl_disable_sagv(struct drm_i915_private *dev_priv);
+bool skl_ddb_allocation_equals(const struct skl_ddb_allocation *old,
+ const struct skl_ddb_allocation *new,
+ enum pipe pipe);
+bool skl_ddb_allocation_overlaps(struct drm_atomic_state *state,
+ const struct skl_ddb_allocation *old,
+ const struct skl_ddb_allocation *new,
+ enum pipe pipe);
+void skl_write_cursor_wm(struct intel_crtc *intel_crtc,
+ const struct skl_wm_values *wm);
+void skl_write_plane_wm(struct intel_crtc *intel_crtc,
+ const struct skl_wm_values *wm,
+ int plane);
uint32_t ilk_pipe_pixel_rate(const struct intel_crtc_state *pipe_config);
bool ilk_disable_lp_wm(struct drm_device *dev);
int sanitize_rc6_option(struct drm_i915_private *dev_priv, int enable_rc6);
}
static bool intel_dsi_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
struct intel_dsi *intel_dsi = container_of(encoder, struct intel_dsi,
intel_panel_enable_backlight(intel_dsi->attached_connector);
}
-static void intel_dsi_prepare(struct intel_encoder *intel_encoder);
+static void intel_dsi_prepare(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *pipe_config);
-static void intel_dsi_pre_enable(struct intel_encoder *encoder)
+static void intel_dsi_pre_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
- struct drm_device *dev = encoder->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
- struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
enum port port;
DRM_DEBUG_KMS("\n");
* lock. It needs to be fully powered down to fix it.
*/
intel_disable_dsi_pll(encoder);
- intel_enable_dsi_pll(encoder, crtc->config);
+ intel_enable_dsi_pll(encoder, pipe_config);
- intel_dsi_prepare(encoder);
+ intel_dsi_prepare(encoder, pipe_config);
/* Panel Enable over CRC PMIC */
if (intel_dsi->gpio_panel)
intel_dsi_enable(encoder);
}
-static void intel_dsi_enable_nop(struct intel_encoder *encoder)
+static void intel_dsi_enable_nop(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
DRM_DEBUG_KMS("\n");
*/
}
-static void intel_dsi_pre_disable(struct intel_encoder *encoder)
+static void intel_dsi_pre_disable(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
enum port port;
intel_disable_dsi_pll(encoder);
}
-static void intel_dsi_post_disable(struct intel_encoder *encoder)
+static void intel_dsi_post_disable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
u16 crtc_htotal_sw, crtc_hsync_start_sw, crtc_hsync_end_sw,
crtc_hblank_start_sw, crtc_hblank_end_sw;
+ /* FIXME: hw readout should not depend on SW state */
intel_crtc = to_intel_crtc(encoder->base.crtc);
adjusted_mode_sw = &intel_crtc->config->base.adjusted_mode;
}
}
-static void intel_dsi_prepare(struct intel_encoder *intel_encoder)
+static void intel_dsi_prepare(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *pipe_config)
{
struct drm_encoder *encoder = &intel_encoder->base;
struct drm_device *dev = encoder->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc);
+ struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
struct intel_dsi *intel_dsi = enc_to_intel_dsi(encoder);
- const struct drm_display_mode *adjusted_mode = &intel_crtc->config->base.adjusted_mode;
+ const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
enum port port;
unsigned int bpp = mipi_dsi_pixel_format_to_bpp(intel_dsi->pixel_format);
u32 val, tmp;
intel_connector->panel.fitting_mode = val;
}
- crtc = intel_attached_encoder(connector)->base.crtc;
+ crtc = connector->state->crtc;
if (crtc && crtc->state->enable) {
/*
* If the CRTC is enabled, the display will be changed
pipe_config->base.adjusted_mode.crtc_clock = pipe_config->port_clock;
}
-static void intel_disable_dvo(struct intel_encoder *encoder)
+static void intel_disable_dvo(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
struct intel_dvo *intel_dvo = enc_to_dvo(encoder);
I915_READ(dvo_reg);
}
-static void intel_enable_dvo(struct intel_encoder *encoder)
+static void intel_enable_dvo(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
struct intel_dvo *intel_dvo = enc_to_dvo(encoder);
- struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
i915_reg_t dvo_reg = intel_dvo->dev.dvo_reg;
u32 temp = I915_READ(dvo_reg);
intel_dvo->dev.dev_ops->mode_set(&intel_dvo->dev,
- &crtc->config->base.mode,
- &crtc->config->base.adjusted_mode);
+ &pipe_config->base.mode,
+ &pipe_config->base.adjusted_mode);
I915_WRITE(dvo_reg, temp | DVO_ENABLE);
I915_READ(dvo_reg);
}
static bool intel_dvo_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_dvo *intel_dvo = enc_to_dvo(encoder);
const struct drm_display_mode *fixed_mode =
return true;
}
-static void intel_dvo_pre_enable(struct intel_encoder *encoder)
+static void intel_dvo_pre_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
- struct drm_device *dev = encoder->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
- const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode;
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+ struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+ const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
struct intel_dvo *intel_dvo = enc_to_dvo(encoder);
int pipe = crtc->pipe;
u32 dvo_val;
return;
}
- drm_encoder_cleanup(&intel_encoder->base);
kfree(intel_dvo);
kfree(intel_connector);
}
static const struct engine_info {
const char *name;
unsigned exec_id;
- unsigned guc_id;
+ enum intel_engine_hw_id hw_id;
u32 mmio_base;
unsigned irq_shift;
int (*init_legacy)(struct intel_engine_cs *engine);
[RCS] = {
.name = "render ring",
.exec_id = I915_EXEC_RENDER,
- .guc_id = GUC_RENDER_ENGINE,
+ .hw_id = RCS_HW,
.mmio_base = RENDER_RING_BASE,
.irq_shift = GEN8_RCS_IRQ_SHIFT,
.init_execlists = logical_render_ring_init,
[BCS] = {
.name = "blitter ring",
.exec_id = I915_EXEC_BLT,
- .guc_id = GUC_BLITTER_ENGINE,
+ .hw_id = BCS_HW,
.mmio_base = BLT_RING_BASE,
.irq_shift = GEN8_BCS_IRQ_SHIFT,
.init_execlists = logical_xcs_ring_init,
[VCS] = {
.name = "bsd ring",
.exec_id = I915_EXEC_BSD,
- .guc_id = GUC_VIDEO_ENGINE,
+ .hw_id = VCS_HW,
.mmio_base = GEN6_BSD_RING_BASE,
.irq_shift = GEN8_VCS1_IRQ_SHIFT,
.init_execlists = logical_xcs_ring_init,
[VCS2] = {
.name = "bsd2 ring",
.exec_id = I915_EXEC_BSD,
- .guc_id = GUC_VIDEO_ENGINE2,
+ .hw_id = VCS2_HW,
.mmio_base = GEN8_BSD2_RING_BASE,
.irq_shift = GEN8_VCS2_IRQ_SHIFT,
.init_execlists = logical_xcs_ring_init,
[VECS] = {
.name = "video enhancement ring",
.exec_id = I915_EXEC_VEBOX,
- .guc_id = GUC_VIDEOENHANCE_ENGINE,
+ .hw_id = VECS_HW,
.mmio_base = VEBOX_RING_BASE,
.irq_shift = GEN8_VECS_IRQ_SHIFT,
.init_execlists = logical_xcs_ring_init,
engine->i915 = dev_priv;
engine->name = info->name;
engine->exec_id = info->exec_id;
- engine->hw_id = engine->guc_id = info->guc_id;
+ engine->hw_id = engine->guc_id = info->hw_id;
engine->mmio_base = info->mmio_base;
engine->irq_shift = info->irq_shift;
int intel_engines_init(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct intel_device_info *device_info = mkwrite_device_info(dev_priv);
unsigned int mask = 0;
int (*init)(struct intel_engine_cs *engine);
unsigned int i;
* are added to the driver by a warning and disabling the forgotten
* engines.
*/
- if (WARN_ON(mask != INTEL_INFO(dev_priv)->ring_mask)) {
- struct intel_device_info *info =
- (struct intel_device_info *)&dev_priv->info;
- info->ring_mask = mask;
- }
+ if (WARN_ON(mask != INTEL_INFO(dev_priv)->ring_mask))
+ device_info->ring_mask = mask;
+
+ device_info->num_rings = hweight32(mask);
return 0;
return ret;
}
+void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno)
+{
+ struct drm_i915_private *dev_priv = engine->i915;
+
+ /* Our semaphore implementation is strictly monotonic (i.e. we proceed
+ * so long as the semaphore value in the register/page is greater
+ * than the sync value), so whenever we reset the seqno,
+ * so long as we reset the tracking semaphore value to 0, it will
+ * always be before the next request's seqno. If we don't reset
+ * the semaphore value, then when the seqno moves backwards all
+ * future waits will complete instantly (causing rendering corruption).
+ */
+ if (IS_GEN6(dev_priv) || IS_GEN7(dev_priv)) {
+ I915_WRITE(RING_SYNC_0(engine->mmio_base), 0);
+ I915_WRITE(RING_SYNC_1(engine->mmio_base), 0);
+ if (HAS_VEBOX(dev_priv))
+ I915_WRITE(RING_SYNC_2(engine->mmio_base), 0);
+ }
+ if (dev_priv->semaphore) {
+ struct page *page = i915_vma_first_page(dev_priv->semaphore);
+ void *semaphores;
+
+ /* Semaphores are in noncoherent memory, flush to be safe */
+ semaphores = kmap(page);
+ memset(semaphores + GEN8_SEMAPHORE_OFFSET(engine->id, 0),
+ 0, I915_NUM_ENGINES * gen8_semaphore_seqno_size);
+ drm_clflush_virt_range(semaphores + GEN8_SEMAPHORE_OFFSET(engine->id, 0),
+ I915_NUM_ENGINES * gen8_semaphore_seqno_size);
+ kunmap(page);
+ }
+ memset(engine->semaphore.sync_seqno, 0,
+ sizeof(engine->semaphore.sync_seqno));
+
+ intel_write_status_page(engine, I915_GEM_HWS_INDEX, seqno);
+ if (engine->irq_seqno_barrier)
+ engine->irq_seqno_barrier(engine);
+ engine->last_submitted_seqno = seqno;
+
+ engine->hangcheck.seqno = seqno;
+
+ /* After manually advancing the seqno, fake the interrupt in case
+ * there are any waiters for that seqno.
+ */
+ intel_engine_wakeup(engine);
+}
+
void intel_engine_init_hangcheck(struct intel_engine_cs *engine)
{
memset(&engine->hangcheck, 0, sizeof(engine->hangcheck));
+ clear_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings);
+ if (intel_engine_has_waiter(engine))
+ i915_queue_hangcheck(engine->i915);
}
static void intel_engine_init_requests(struct intel_engine_cs *engine)
*/
void intel_engine_setup_common(struct intel_engine_cs *engine)
{
- INIT_LIST_HEAD(&engine->buffers);
INIT_LIST_HEAD(&engine->execlist_queue);
spin_lock_init(&engine->execlist_lock);
intel_engine_init_requests(engine);
intel_engine_init_hangcheck(engine);
i915_gem_batch_pool_init(engine, &engine->batch_pool);
+
+ intel_engine_init_cmd_parser(engine);
+}
+
+int intel_engine_create_scratch(struct intel_engine_cs *engine, int size)
+{
+ struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
+ int ret;
+
+ WARN_ON(engine->scratch);
+
+ obj = i915_gem_object_create_stolen(&engine->i915->drm, size);
+ if (!obj)
+ obj = i915_gem_object_create(&engine->i915->drm, size);
+ if (IS_ERR(obj)) {
+ DRM_ERROR("Failed to allocate scratch page\n");
+ return PTR_ERR(obj);
+ }
+
+ vma = i915_vma_create(obj, &engine->i915->ggtt.base, NULL);
+ if (IS_ERR(vma)) {
+ ret = PTR_ERR(vma);
+ goto err_unref;
+ }
+
+ ret = i915_vma_pin(vma, 0, 4096, PIN_GLOBAL | PIN_HIGH);
+ if (ret)
+ goto err_unref;
+
+ engine->scratch = vma;
+ DRM_DEBUG_DRIVER("%s pipe control offset: 0x%08x\n",
+ engine->name, i915_ggtt_offset(vma));
+ return 0;
+
+err_unref:
+ i915_gem_object_put(obj);
+ return ret;
+}
+
+static void intel_engine_cleanup_scratch(struct intel_engine_cs *engine)
+{
+ i915_vma_unpin_and_release(&engine->scratch);
}
/**
if (ret)
return ret;
- return intel_engine_init_cmd_parser(engine);
+ return 0;
+}
+
+void intel_engine_reset_irq(struct intel_engine_cs *engine)
+{
+ struct drm_i915_private *dev_priv = engine->i915;
+
+ spin_lock_irq(&dev_priv->irq_lock);
+ if (intel_engine_has_waiter(engine))
+ engine->irq_enable(engine);
+ else
+ engine->irq_disable(engine);
+ spin_unlock_irq(&dev_priv->irq_lock);
}
/**
*/
void intel_engine_cleanup_common(struct intel_engine_cs *engine)
{
- intel_engine_cleanup_cmd_parser(engine);
+ intel_engine_cleanup_scratch(engine);
+
intel_engine_fini_breadcrumbs(engine);
+ intel_engine_cleanup_cmd_parser(engine);
i915_gem_batch_pool_fini(&engine->batch_pool);
}
dpfc_ctl |= DPFC_CTL_LIMIT_2X;
else
dpfc_ctl |= DPFC_CTL_LIMIT_1X;
- dpfc_ctl |= DPFC_CTL_FENCE_EN | params->fb.fence_reg;
- I915_WRITE(DPFC_FENCE_YOFF, params->crtc.fence_y_offset);
+ if (params->fb.fence_reg != I915_FENCE_REG_NONE) {
+ dpfc_ctl |= DPFC_CTL_FENCE_EN | params->fb.fence_reg;
+ I915_WRITE(DPFC_FENCE_YOFF, params->crtc.fence_y_offset);
+ } else {
+ I915_WRITE(DPFC_FENCE_YOFF, 0);
+ }
/* enable it... */
I915_WRITE(DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
dpfc_ctl |= DPFC_CTL_LIMIT_1X;
break;
}
- dpfc_ctl |= DPFC_CTL_FENCE_EN;
- if (IS_GEN5(dev_priv))
- dpfc_ctl |= params->fb.fence_reg;
+
+ if (params->fb.fence_reg != I915_FENCE_REG_NONE) {
+ dpfc_ctl |= DPFC_CTL_FENCE_EN;
+ if (IS_GEN5(dev_priv))
+ dpfc_ctl |= params->fb.fence_reg;
+ if (IS_GEN6(dev_priv)) {
+ I915_WRITE(SNB_DPFC_CTL_SA,
+ SNB_CPU_FENCE_ENABLE | params->fb.fence_reg);
+ I915_WRITE(DPFC_CPU_FENCE_OFFSET,
+ params->crtc.fence_y_offset);
+ }
+ } else {
+ if (IS_GEN6(dev_priv)) {
+ I915_WRITE(SNB_DPFC_CTL_SA, 0);
+ I915_WRITE(DPFC_CPU_FENCE_OFFSET, 0);
+ }
+ }
I915_WRITE(ILK_DPFC_FENCE_YOFF, params->crtc.fence_y_offset);
I915_WRITE(ILK_FBC_RT_BASE, params->fb.ggtt_offset | ILK_FBC_RT_VALID);
/* enable it... */
I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
- if (IS_GEN6(dev_priv)) {
- I915_WRITE(SNB_DPFC_CTL_SA,
- SNB_CPU_FENCE_ENABLE | params->fb.fence_reg);
- I915_WRITE(DPFC_CPU_FENCE_OFFSET, params->crtc.fence_y_offset);
- }
-
intel_fbc_recompress(dev_priv);
}
break;
}
- dpfc_ctl |= IVB_DPFC_CTL_FENCE_EN;
+ if (params->fb.fence_reg != I915_FENCE_REG_NONE) {
+ dpfc_ctl |= IVB_DPFC_CTL_FENCE_EN;
+ I915_WRITE(SNB_DPFC_CTL_SA,
+ SNB_CPU_FENCE_ENABLE | params->fb.fence_reg);
+ I915_WRITE(DPFC_CPU_FENCE_OFFSET, params->crtc.fence_y_offset);
+ } else {
+ I915_WRITE(SNB_DPFC_CTL_SA,0);
+ I915_WRITE(DPFC_CPU_FENCE_OFFSET, 0);
+ }
if (dev_priv->fbc.false_color)
dpfc_ctl |= FBC_CTL_FALSE_COLOR;
I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
- I915_WRITE(SNB_DPFC_CTL_SA,
- SNB_CPU_FENCE_ENABLE | params->fb.fence_reg);
- I915_WRITE(DPFC_CPU_FENCE_OFFSET, params->crtc.fence_y_offset);
-
intel_fbc_recompress(dev_priv);
}
return effective_w <= max_w && effective_h <= max_h;
}
+/* XXX replace me when we have VMA tracking for intel_plane_state */
+static int get_fence_id(struct drm_framebuffer *fb)
+{
+ struct i915_vma *vma = i915_gem_object_to_ggtt(intel_fb_obj(fb), NULL);
+
+ return vma && vma->fence ? vma->fence->id : I915_FENCE_REG_NONE;
+}
+
static void intel_fbc_update_state_cache(struct intel_crtc *crtc,
struct intel_crtc_state *crtc_state,
struct intel_plane_state *plane_state)
/* FIXME: We lack the proper locking here, so only run this on the
* platforms that need. */
if (IS_GEN(dev_priv, 5, 6))
- cache->fb.ilk_ggtt_offset = i915_gem_obj_ggtt_offset(obj);
+ cache->fb.ilk_ggtt_offset = i915_gem_object_ggtt_offset(obj, NULL);
cache->fb.pixel_format = fb->pixel_format;
cache->fb.stride = fb->pitches[0];
- cache->fb.fence_reg = obj->fence_reg;
+ cache->fb.fence_reg = get_fence_id(fb);
cache->fb.tiling_mode = i915_gem_object_get_tiling(obj);
}
/* The use of a CPU fence is mandatory in order to detect writes
* by the CPU to the scanout and trigger updates to the FBC.
+ *
+ * Note that is possible for a tiled surface to be unmappable (and
+ * so have no fence associated with it) due to aperture constaints
+ * at the time of pinning.
*/
if (cache->fb.tiling_mode != I915_TILING_X ||
cache->fb.fence_reg == I915_FENCE_REG_NONE) {
struct intel_framebuffer *intel_fb = ifbdev->fb;
struct drm_device *dev = helper->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
struct i915_ggtt *ggtt = &dev_priv->ggtt;
struct fb_info *info;
struct drm_framebuffer *fb;
struct i915_vma *vma;
- struct drm_i915_gem_object *obj;
bool prealloc = false;
void __iomem *vaddr;
int ret;
sizes->fb_height = intel_fb->base.height;
}
- obj = intel_fb->obj;
-
mutex_lock(&dev->struct_mutex);
/* Pin the GGTT vma for our access via info->screen_base.
* This also validates that any existing fb inherited from the
* BIOS is suitable for own access.
*/
- ret = intel_pin_and_fence_fb_obj(&ifbdev->fb->base, DRM_ROTATE_0);
- if (ret)
+ vma = intel_pin_and_fence_fb_obj(&ifbdev->fb->base, DRM_ROTATE_0);
+ if (IS_ERR(vma)) {
+ ret = PTR_ERR(vma);
goto out_unlock;
+ }
info = drm_fb_helper_alloc_fbi(helper);
if (IS_ERR(info)) {
info->flags = FBINFO_DEFAULT | FBINFO_CAN_FORCE_OUTPUT;
info->fbops = &intelfb_ops;
- vma = i915_gem_obj_to_ggtt(obj);
-
/* setup aperture base/size for vesafb takeover */
info->apertures->ranges[0].base = dev->mode_config.fb_base;
info->apertures->ranges[0].size = ggtt->mappable_end;
- info->fix.smem_start = dev->mode_config.fb_base + vma->node.start;
+ info->fix.smem_start = dev->mode_config.fb_base + i915_ggtt_offset(vma);
info->fix.smem_len = vma->node.size;
vaddr = i915_vma_pin_iomap(vma);
* If the object is stolen however, it will be full of whatever
* garbage was left in there.
*/
- if (ifbdev->fb->obj->stolen && !prealloc)
+ if (intel_fb->obj->stolen && !prealloc)
memset_io(info->screen_base, 0, info->screen_size);
/* Use default scratch pixmap (info->pixmap.flags = FB_PIXMAP_SYSTEM) */
- DRM_DEBUG_KMS("allocated %dx%d fb: 0x%08llx, bo %p\n",
- fb->width, fb->height,
- i915_gem_obj_ggtt_offset(obj), obj);
+ DRM_DEBUG_KMS("allocated %dx%d fb: 0x%08x\n",
+ fb->width, fb->height, i915_ggtt_offset(vma));
+ ifbdev->vma = vma;
mutex_unlock(&dev->struct_mutex);
- vga_switcheroo_client_fb_set(dev->pdev, info);
+ vga_switcheroo_client_fb_set(pdev, info);
return 0;
out_destroy_fbi:
* retcode: errno from last guc_submit()
*/
struct i915_guc_client {
- struct drm_i915_gem_object *client_obj;
+ struct i915_vma *vma;
void *client_base; /* first page (only) of above */
struct i915_gem_context *owner;
struct intel_guc *guc;
+
+ uint32_t engines; /* bitmap of (host) engine ids */
uint32_t priority;
uint32_t ctx_index;
-
uint32_t proc_desc_offset;
+
uint32_t doorbell_offset;
uint32_t cookie;
uint16_t doorbell_id;
- uint16_t padding; /* Maintain alignment */
+ uint16_t padding[3]; /* Maintain alignment */
+ spinlock_t wq_lock;
uint32_t wq_offset;
uint32_t wq_size;
uint32_t wq_tail;
- uint32_t unused; /* Was 'wq_head' */
-
+ uint32_t wq_rsvd;
uint32_t no_wq_space;
- uint32_t q_fail; /* No longer used */
uint32_t b_fail;
int retcode;
struct intel_guc {
struct intel_guc_fw guc_fw;
uint32_t log_flags;
- struct drm_i915_gem_object *log_obj;
-
- struct drm_i915_gem_object *ads_obj;
+ struct i915_vma *log_vma;
- struct drm_i915_gem_object *ctx_pool_obj;
+ struct i915_vma *ads_vma;
+ struct i915_vma *ctx_pool_vma;
struct ida ctx_ids;
struct i915_guc_client *execbuf_client;
/* i915_guc_submission.c */
int i915_guc_submission_init(struct drm_i915_private *dev_priv);
int i915_guc_submission_enable(struct drm_i915_private *dev_priv);
-int i915_guc_wq_check_space(struct drm_i915_gem_request *rq);
+int i915_guc_wq_reserve(struct drm_i915_gem_request *rq);
void i915_guc_submission_disable(struct drm_i915_private *dev_priv);
void i915_guc_submission_fini(struct drm_i915_private *dev_priv);
*
*/
-#define I915_SKL_GUC_UCODE "i915/skl_guc_ver6_1.bin"
+#define SKL_FW_MAJOR 6
+#define SKL_FW_MINOR 1
+
+#define BXT_FW_MAJOR 8
+#define BXT_FW_MINOR 7
+
+#define KBL_FW_MAJOR 9
+#define KBL_FW_MINOR 14
+
+#define GUC_FW_PATH(platform, major, minor) \
+ "i915/" __stringify(platform) "_guc_ver" __stringify(major) "_" __stringify(minor) ".bin"
+
+#define I915_SKL_GUC_UCODE GUC_FW_PATH(skl, SKL_FW_MAJOR, SKL_FW_MINOR)
MODULE_FIRMWARE(I915_SKL_GUC_UCODE);
-#define I915_BXT_GUC_UCODE "i915/bxt_guc_ver8_7.bin"
+#define I915_BXT_GUC_UCODE GUC_FW_PATH(bxt, BXT_FW_MAJOR, BXT_FW_MINOR)
MODULE_FIRMWARE(I915_BXT_GUC_UCODE);
-#define I915_KBL_GUC_UCODE "i915/kbl_guc_ver9_14.bin"
+#define I915_KBL_GUC_UCODE GUC_FW_PATH(kbl, KBL_FW_MAJOR, KBL_FW_MINOR)
MODULE_FIRMWARE(I915_KBL_GUC_UCODE);
/* User-friendly representation of an enum */
}
};
-static void direct_interrupts_to_host(struct drm_i915_private *dev_priv)
+static void guc_interrupts_release(struct drm_i915_private *dev_priv)
{
struct intel_engine_cs *engine;
int irqs;
I915_WRITE(GUC_WD_VECS_IER, 0);
}
-static void direct_interrupts_to_guc(struct drm_i915_private *dev_priv)
+static void guc_interrupts_capture(struct drm_i915_private *dev_priv)
{
struct intel_engine_cs *engine;
int irqs;
I915_WRITE(GUC_WD_VECS_IER, ~irqs);
/*
- * If GuC has routed PM interrupts to itself, don't keep it.
- * and keep other interrupts those are unmasked by GuC.
- */
+ * The REDIRECT_TO_GUC bit of the PMINTRMSK register directs all
+ * (unmasked) PM interrupts to the GuC. All other bits of this
+ * register *disable* generation of a specific interrupt.
+ *
+ * 'pm_intr_keep' indicates bits that are NOT to be set when
+ * writing to the PM interrupt mask register, i.e. interrupts
+ * that must not be disabled.
+ *
+ * If the GuC is handling these interrupts, then we must not let
+ * the PM code disable ANY interrupt that the GuC is expecting.
+ * So for each ENABLED (0) bit in this register, we must SET the
+ * bit in pm_intr_keep so that it's left enabled for the GuC.
+ *
+ * OTOH the REDIRECT_TO_GUC bit is initially SET in pm_intr_keep
+ * (so interrupts go to the DISPLAY unit at first); but here we
+ * need to CLEAR that bit, which will result in the register bit
+ * being left SET!
+ */
tmp = I915_READ(GEN6_PMINTRMSK);
- if (tmp & GEN8_PMINTR_REDIRECT_TO_NON_DISP) {
- dev_priv->rps.pm_intr_keep |= ~(tmp & ~GEN8_PMINTR_REDIRECT_TO_NON_DISP);
- dev_priv->rps.pm_intr_keep &= ~GEN8_PMINTR_REDIRECT_TO_NON_DISP;
+ if (tmp & GEN8_PMINTR_REDIRECT_TO_GUC) {
+ dev_priv->rps.pm_intr_keep |= ~tmp;
+ dev_priv->rps.pm_intr_keep &= ~GEN8_PMINTR_REDIRECT_TO_GUC;
}
}
static u32 get_core_family(struct drm_i915_private *dev_priv)
{
- switch (INTEL_INFO(dev_priv)->gen) {
+ u32 gen = INTEL_GEN(dev_priv);
+
+ switch (gen) {
case 9:
return GFXCORE_FAMILY_GEN9;
default:
- DRM_ERROR("GUC: unsupported core family\n");
+ WARN(1, "GEN%d does not support GuC operation!\n", gen);
return GFXCORE_FAMILY_UNKNOWN;
}
}
-static void set_guc_init_params(struct drm_i915_private *dev_priv)
+/*
+ * Initialise the GuC parameter block before starting the firmware
+ * transfer. These parameters are read by the firmware on startup
+ * and cannot be changed thereafter.
+ */
+static void guc_params_init(struct drm_i915_private *dev_priv)
{
struct intel_guc *guc = &dev_priv->guc;
u32 params[GUC_CTL_MAX_DWORDS];
i915.guc_log_level << GUC_LOG_VERBOSITY_SHIFT;
}
- if (guc->ads_obj) {
- u32 ads = (u32)i915_gem_obj_ggtt_offset(guc->ads_obj)
- >> PAGE_SHIFT;
+ if (guc->ads_vma) {
+ u32 ads = i915_ggtt_offset(guc->ads_vma) >> PAGE_SHIFT;
params[GUC_CTL_DEBUG] |= ads << GUC_ADS_ADDR_SHIFT;
params[GUC_CTL_DEBUG] |= GUC_ADS_ENABLED;
}
/* If GuC submission is enabled, set up additional parameters here */
if (i915.enable_guc_submission) {
- u32 pgs = i915_gem_obj_ggtt_offset(dev_priv->guc.ctx_pool_obj);
+ u32 pgs = i915_ggtt_offset(dev_priv->guc.ctx_pool_vma);
u32 ctx_in_16 = GUC_MAX_GPU_CONTEXTS / 16;
pgs >>= PAGE_SHIFT;
* Note that GuC needs the CSS header plus uKernel code to be copied by the
* DMA engine in one operation, whereas the RSA signature is loaded via MMIO.
*/
-static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv)
+static int guc_ucode_xfer_dma(struct drm_i915_private *dev_priv,
+ struct i915_vma *vma)
{
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
- struct drm_i915_gem_object *fw_obj = guc_fw->guc_fw_obj;
unsigned long offset;
- struct sg_table *sg = fw_obj->pages;
+ struct sg_table *sg = vma->pages;
u32 status, rsa[UOS_RSA_SCRATCH_MAX_COUNT];
int i, ret = 0;
I915_WRITE(DMA_COPY_SIZE, guc_fw->header_size + guc_fw->ucode_size);
/* Set the source address for the new blob */
- offset = i915_gem_obj_ggtt_offset(fw_obj) + guc_fw->header_offset;
+ offset = i915_ggtt_offset(vma) + guc_fw->header_offset;
I915_WRITE(DMA_ADDR_0_LOW, lower_32_bits(offset));
I915_WRITE(DMA_ADDR_0_HIGH, upper_32_bits(offset) & 0xFFFF);
{
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
struct drm_device *dev = &dev_priv->drm;
+ struct i915_vma *vma;
int ret;
ret = i915_gem_object_set_to_gtt_domain(guc_fw->guc_fw_obj, false);
return ret;
}
- ret = i915_gem_object_ggtt_pin(guc_fw->guc_fw_obj, NULL, 0, 0, 0);
- if (ret) {
- DRM_DEBUG_DRIVER("pin failed %d\n", ret);
- return ret;
+ vma = i915_gem_object_ggtt_pin(guc_fw->guc_fw_obj, NULL, 0, 0, 0);
+ if (IS_ERR(vma)) {
+ DRM_DEBUG_DRIVER("pin failed %d\n", (int)PTR_ERR(vma));
+ return PTR_ERR(vma);
}
/* Invalidate GuC TLB to let GuC take the latest updates to GTT. */
I915_WRITE(GEN7_MISCCPCTL, (GEN8_DOP_CLOCK_GATE_GUC_ENABLE |
I915_READ(GEN7_MISCCPCTL)));
- /* allows for 5us before GT can go to RC6 */
+ /* allows for 5us (in 10ns units) before GT can go to RC6 */
I915_WRITE(GUC_ARAT_C6DIS, 0x1FF);
}
- set_guc_init_params(dev_priv);
+ guc_params_init(dev_priv);
- ret = guc_ucode_xfer_dma(dev_priv);
+ ret = guc_ucode_xfer_dma(dev_priv, vma);
intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
* We keep the object pages for reuse during resume. But we can unpin it
* now that DMA has completed, so it doesn't continue to take up space.
*/
- i915_gem_object_ggtt_unpin(guc_fw->guc_fw_obj);
+ i915_vma_unpin(vma);
return ret;
}
-static int i915_reset_guc(struct drm_i915_private *dev_priv)
+static int guc_hw_reset(struct drm_i915_private *dev_priv)
{
int ret;
u32 guc_status;
goto fail;
} else if (*fw_path == '\0') {
/* Device has a GuC but we don't know what f/w to load? */
- DRM_INFO("No GuC firmware known for this platform\n");
+ WARN(1, "No GuC firmware known for this platform!\n");
err = -ENODEV;
goto fail;
}
goto fail;
}
- direct_interrupts_to_host(dev_priv);
+ guc_interrupts_release(dev_priv);
guc_fw->guc_fw_load_status = GUC_FIRMWARE_PENDING;
* Always reset the GuC just before (re)loading, so
* that the state and timing are fairly predictable
*/
- err = i915_reset_guc(dev_priv);
- if (err) {
- DRM_ERROR("GuC reset failed: %d\n", err);
+ err = guc_hw_reset(dev_priv);
+ if (err)
goto fail;
- }
err = guc_ucode_xfer(dev_priv);
if (!err)
err = i915_guc_submission_enable(dev_priv);
if (err)
goto fail;
- direct_interrupts_to_guc(dev_priv);
+ guc_interrupts_capture(dev_priv);
}
return 0;
if (guc_fw->guc_fw_load_status == GUC_FIRMWARE_PENDING)
guc_fw->guc_fw_load_status = GUC_FIRMWARE_FAIL;
- direct_interrupts_to_host(dev_priv);
+ guc_interrupts_release(dev_priv);
i915_guc_submission_disable(dev_priv);
i915_guc_submission_fini(dev_priv);
else if (err == 0)
DRM_INFO("GuC firmware load skipped\n");
else if (ret != -EIO)
- DRM_INFO("GuC firmware load failed: %d\n", err);
+ DRM_NOTE("GuC firmware load failed: %d\n", err);
else
- DRM_ERROR("GuC firmware load failed: %d\n", err);
+ DRM_WARN("GuC firmware load failed: %d\n", err);
if (i915.enable_guc_submission) {
if (fw_path == NULL)
DRM_INFO("GuC submission without firmware not supported\n");
if (ret == 0)
- DRM_INFO("Falling back from GuC submission to execlist mode\n");
+ DRM_NOTE("Falling back from GuC submission to execlist mode\n");
else
DRM_ERROR("GuC init failed: %d\n", ret);
}
static void guc_fw_fetch(struct drm_device *dev, struct intel_guc_fw *guc_fw)
{
+ struct pci_dev *pdev = dev->pdev;
struct drm_i915_gem_object *obj;
const struct firmware *fw;
struct guc_css_header *css;
DRM_DEBUG_DRIVER("before requesting firmware: GuC fw fetch status %s\n",
intel_guc_fw_status_repr(guc_fw->guc_fw_fetch_status));
- err = request_firmware(&fw, guc_fw->guc_fw_path, &dev->pdev->dev);
+ err = request_firmware(&fw, guc_fw->guc_fw_path, &pdev->dev);
if (err)
goto fail;
if (!fw)
/* Check the size of the blob before examining buffer contents */
if (fw->size < sizeof(struct guc_css_header)) {
- DRM_ERROR("Firmware header is missing\n");
+ DRM_NOTE("Firmware header is missing\n");
goto fail;
}
css->key_size_dw - css->exponent_size_dw) * sizeof(u32);
if (guc_fw->header_size != sizeof(struct guc_css_header)) {
- DRM_ERROR("CSS header definition mismatch\n");
+ DRM_NOTE("CSS header definition mismatch\n");
goto fail;
}
/* now RSA */
if (css->key_size_dw != UOS_RSA_SCRATCH_MAX_COUNT) {
- DRM_ERROR("RSA key size is bad\n");
+ DRM_NOTE("RSA key size is bad\n");
goto fail;
}
guc_fw->rsa_offset = guc_fw->ucode_offset + guc_fw->ucode_size;
/* At least, it should have header, uCode and RSA. Size of all three. */
size = guc_fw->header_size + guc_fw->ucode_size + guc_fw->rsa_size;
if (fw->size < size) {
- DRM_ERROR("Missing firmware components\n");
+ DRM_NOTE("Missing firmware components\n");
goto fail;
}
/* Header and uCode will be loaded to WOPCM. Size of the two. */
size = guc_fw->header_size + guc_fw->ucode_size;
if (size > guc_wopcm_size(to_i915(dev))) {
- DRM_ERROR("Firmware is too large to fit in WOPCM\n");
+ DRM_NOTE("Firmware is too large to fit in WOPCM\n");
goto fail;
}
if (guc_fw->guc_fw_major_found != guc_fw->guc_fw_major_wanted ||
guc_fw->guc_fw_minor_found < guc_fw->guc_fw_minor_wanted) {
- DRM_ERROR("GuC firmware version %d.%d, required %d.%d\n",
+ DRM_NOTE("GuC firmware version %d.%d, required %d.%d\n",
guc_fw->guc_fw_major_found, guc_fw->guc_fw_minor_found,
guc_fw->guc_fw_major_wanted, guc_fw->guc_fw_minor_wanted);
err = -ENOEXEC;
return;
fail:
+ DRM_WARN("Failed to fetch valid GuC firmware from %s (error %d)\n",
+ guc_fw->guc_fw_path, err);
DRM_DEBUG_DRIVER("GuC fw fetch status FAIL; err %d, fw %p, obj %p\n",
err, fw, guc_fw->guc_fw_obj);
- DRM_ERROR("Failed to fetch GuC firmware from %s (error %d)\n",
- guc_fw->guc_fw_path, err);
mutex_lock(&dev->struct_mutex);
obj = guc_fw->guc_fw_obj;
fw_path = NULL;
} else if (IS_SKYLAKE(dev)) {
fw_path = I915_SKL_GUC_UCODE;
- guc_fw->guc_fw_major_wanted = 6;
- guc_fw->guc_fw_minor_wanted = 1;
+ guc_fw->guc_fw_major_wanted = SKL_FW_MAJOR;
+ guc_fw->guc_fw_minor_wanted = SKL_FW_MINOR;
} else if (IS_BROXTON(dev)) {
fw_path = I915_BXT_GUC_UCODE;
- guc_fw->guc_fw_major_wanted = 8;
- guc_fw->guc_fw_minor_wanted = 7;
+ guc_fw->guc_fw_major_wanted = BXT_FW_MAJOR;
+ guc_fw->guc_fw_minor_wanted = BXT_FW_MINOR;
} else if (IS_KABYLAKE(dev)) {
fw_path = I915_KBL_GUC_UCODE;
- guc_fw->guc_fw_major_wanted = 9;
- guc_fw->guc_fw_minor_wanted = 14;
+ guc_fw->guc_fw_major_wanted = KBL_FW_MAJOR;
+ guc_fw->guc_fw_minor_wanted = KBL_FW_MINOR;
} else {
fw_path = ""; /* unknown device */
}
struct intel_guc_fw *guc_fw = &dev_priv->guc.guc_fw;
mutex_lock(&dev->struct_mutex);
- direct_interrupts_to_host(dev_priv);
+ guc_interrupts_release(dev_priv);
i915_guc_submission_disable(dev_priv);
i915_guc_submission_fini(dev_priv);
intel_audio_codec_enable(encoder);
}
-static void g4x_enable_hdmi(struct intel_encoder *encoder)
+static void g4x_enable_hdmi(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
intel_enable_hdmi_audio(encoder);
}
-static void ibx_enable_hdmi(struct intel_encoder *encoder)
+static void ibx_enable_hdmi(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
intel_enable_hdmi_audio(encoder);
}
-static void cpt_enable_hdmi(struct intel_encoder *encoder)
+static void cpt_enable_hdmi(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
intel_enable_hdmi_audio(encoder);
}
-static void vlv_enable_hdmi(struct intel_encoder *encoder)
+static void vlv_enable_hdmi(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
}
-static void intel_disable_hdmi(struct intel_encoder *encoder)
+static void intel_disable_hdmi(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
intel_dp_dual_mode_set_tmds_output(intel_hdmi, false);
}
-static void g4x_disable_hdmi(struct intel_encoder *encoder)
+static void g4x_disable_hdmi(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
if (crtc->config->has_audio)
intel_audio_codec_disable(encoder);
- intel_disable_hdmi(encoder);
+ intel_disable_hdmi(encoder, old_crtc_state, old_conn_state);
}
-static void pch_disable_hdmi(struct intel_encoder *encoder)
+static void pch_disable_hdmi(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
intel_audio_codec_disable(encoder);
}
-static void pch_post_disable_hdmi(struct intel_encoder *encoder)
+static void pch_post_disable_hdmi(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
- intel_disable_hdmi(encoder);
+ intel_disable_hdmi(encoder, old_crtc_state, old_conn_state);
}
static int intel_hdmi_source_max_tmds_clock(struct drm_i915_private *dev_priv)
}
bool intel_hdmi_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
struct drm_device *dev = encoder->base.dev;
}
static bool
-intel_hdmi_set_edid(struct drm_connector *connector, bool force)
+intel_hdmi_set_edid(struct drm_connector *connector)
{
struct drm_i915_private *dev_priv = to_i915(connector->dev);
struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
- struct edid *edid = NULL;
+ struct edid *edid;
bool connected = false;
- if (force) {
- intel_display_power_get(dev_priv, POWER_DOMAIN_GMBUS);
+ intel_display_power_get(dev_priv, POWER_DOMAIN_GMBUS);
- edid = drm_get_edid(connector,
- intel_gmbus_get_adapter(dev_priv,
- intel_hdmi->ddc_bus));
+ edid = drm_get_edid(connector,
+ intel_gmbus_get_adapter(dev_priv,
+ intel_hdmi->ddc_bus));
- intel_hdmi_dp_dual_mode_detect(connector, edid != NULL);
+ intel_hdmi_dp_dual_mode_detect(connector, edid != NULL);
- intel_display_power_put(dev_priv, POWER_DOMAIN_GMBUS);
- }
+ intel_display_power_put(dev_priv, POWER_DOMAIN_GMBUS);
to_intel_connector(connector)->detect_edid = edid;
if (edid && edid->input & DRM_EDID_INPUT_DIGITAL) {
intel_hdmi_detect(struct drm_connector *connector, bool force)
{
enum drm_connector_status status;
- struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
struct drm_i915_private *dev_priv = to_i915(connector->dev);
- bool live_status = false;
- unsigned int try;
DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
connector->base.id, connector->name);
intel_display_power_get(dev_priv, POWER_DOMAIN_GMBUS);
- for (try = 0; !live_status && try < 9; try++) {
- if (try)
- msleep(10);
- live_status = intel_digital_port_connected(dev_priv,
- hdmi_to_dig_port(intel_hdmi));
- }
-
- if (!live_status) {
- DRM_DEBUG_KMS("HDMI live status down\n");
- /*
- * Live status register is not reliable on all intel platforms.
- * So consider live_status only for certain platforms, for
- * others, read EDID to determine presence of sink.
- */
- if (INTEL_INFO(dev_priv)->gen < 7 || IS_IVYBRIDGE(dev_priv))
- live_status = true;
- }
-
intel_hdmi_unset_edid(connector);
- if (intel_hdmi_set_edid(connector, live_status)) {
+ if (intel_hdmi_set_edid(connector)) {
struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
hdmi_to_dig_port(intel_hdmi)->base.type = INTEL_OUTPUT_HDMI;
if (connector->status != connector_status_connected)
return;
- intel_hdmi_set_edid(connector, true);
+ intel_hdmi_set_edid(connector);
hdmi_to_dig_port(intel_hdmi)->base.type = INTEL_OUTPUT_HDMI;
}
return 0;
}
-static void intel_hdmi_pre_enable(struct intel_encoder *encoder)
+static void intel_hdmi_pre_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
adjusted_mode);
}
-static void vlv_hdmi_pre_enable(struct intel_encoder *encoder)
+static void vlv_hdmi_pre_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
struct intel_hdmi *intel_hdmi = &dport->hdmi;
intel_crtc->config->has_hdmi_sink,
adjusted_mode);
- g4x_enable_hdmi(encoder);
+ g4x_enable_hdmi(encoder, pipe_config, conn_state);
vlv_wait_port_ready(dev_priv, dport, 0x0);
}
-static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
+static void vlv_hdmi_pre_pll_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
intel_hdmi_prepare(encoder);
vlv_phy_pre_pll_enable(encoder);
}
-static void chv_hdmi_pre_pll_enable(struct intel_encoder *encoder)
+static void chv_hdmi_pre_pll_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
intel_hdmi_prepare(encoder);
chv_phy_pre_pll_enable(encoder);
}
-static void chv_hdmi_post_pll_disable(struct intel_encoder *encoder)
+static void chv_hdmi_post_pll_disable(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
chv_phy_post_pll_disable(encoder);
}
-static void vlv_hdmi_post_disable(struct intel_encoder *encoder)
+static void vlv_hdmi_post_disable(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
/* Reset lanes to avoid HDMI flicker (VLV w/a) */
vlv_phy_reset_lanes(encoder);
}
-static void chv_hdmi_post_disable(struct intel_encoder *encoder)
+static void chv_hdmi_post_disable(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
mutex_unlock(&dev_priv->sb_lock);
}
-static void chv_hdmi_pre_enable(struct intel_encoder *encoder)
+static void chv_hdmi_pre_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
struct intel_hdmi *intel_hdmi = &dport->hdmi;
intel_crtc->config->has_hdmi_sink,
adjusted_mode);
- g4x_enable_hdmi(encoder);
+ g4x_enable_hdmi(encoder, pipe_config, conn_state);
vlv_wait_port_ready(dev_priv, dport, 0x0);
spin_unlock_irq(&dev_priv->irq_lock);
}
-void i915_hpd_poll_init_work(struct work_struct *work) {
+static void i915_hpd_poll_init_work(struct work_struct *work)
+{
struct drm_i915_private *dev_priv =
container_of(work, struct drm_i915_private,
hotplug.poll_init_work);
algo->data = bus;
}
-static int
-gmbus_wait_hw_status(struct drm_i915_private *dev_priv,
- u32 gmbus2_status,
- u32 gmbus4_irq_en)
+static int gmbus_wait(struct drm_i915_private *dev_priv, u32 status, u32 irq_en)
{
- int i;
- u32 gmbus2 = 0;
DEFINE_WAIT(wait);
-
- if (!HAS_GMBUS_IRQ(dev_priv))
- gmbus4_irq_en = 0;
+ u32 gmbus2;
+ int ret;
/* Important: The hw handles only the first bit, so set only one! Since
* we also need to check for NAKs besides the hw ready/idle signal, we
- * need to wake up periodically and check that ourselves. */
- I915_WRITE(GMBUS4, gmbus4_irq_en);
-
- for (i = 0; i < msecs_to_jiffies_timeout(50); i++) {
- prepare_to_wait(&dev_priv->gmbus_wait_queue, &wait,
- TASK_UNINTERRUPTIBLE);
+ * need to wake up periodically and check that ourselves.
+ */
+ if (!HAS_GMBUS_IRQ(dev_priv))
+ irq_en = 0;
- gmbus2 = I915_READ_NOTRACE(GMBUS2);
- if (gmbus2 & (GMBUS_SATOER | gmbus2_status))
- break;
+ add_wait_queue(&dev_priv->gmbus_wait_queue, &wait);
+ I915_WRITE_FW(GMBUS4, irq_en);
- schedule_timeout(1);
- }
- finish_wait(&dev_priv->gmbus_wait_queue, &wait);
+ status |= GMBUS_SATOER;
+ ret = wait_for_us((gmbus2 = I915_READ_FW(GMBUS2)) & status, 2);
+ if (ret)
+ ret = wait_for((gmbus2 = I915_READ_FW(GMBUS2)) & status, 50);
- I915_WRITE(GMBUS4, 0);
+ I915_WRITE_FW(GMBUS4, 0);
+ remove_wait_queue(&dev_priv->gmbus_wait_queue, &wait);
if (gmbus2 & GMBUS_SATOER)
return -ENXIO;
- if (gmbus2 & gmbus2_status)
- return 0;
- return -ETIMEDOUT;
+
+ return ret;
}
static int
gmbus_wait_idle(struct drm_i915_private *dev_priv)
{
+ DEFINE_WAIT(wait);
+ u32 irq_enable;
int ret;
- if (!HAS_GMBUS_IRQ(dev_priv))
- return intel_wait_for_register(dev_priv,
- GMBUS2, GMBUS_ACTIVE, 0,
- 10);
-
/* Important: The hw handles only the first bit, so set only one! */
- I915_WRITE(GMBUS4, GMBUS_IDLE_EN);
+ irq_enable = 0;
+ if (HAS_GMBUS_IRQ(dev_priv))
+ irq_enable = GMBUS_IDLE_EN;
- ret = wait_event_timeout(dev_priv->gmbus_wait_queue,
- (I915_READ_NOTRACE(GMBUS2) & GMBUS_ACTIVE) == 0,
- msecs_to_jiffies_timeout(10));
+ add_wait_queue(&dev_priv->gmbus_wait_queue, &wait);
+ I915_WRITE_FW(GMBUS4, irq_enable);
- I915_WRITE(GMBUS4, 0);
+ ret = intel_wait_for_register_fw(dev_priv,
+ GMBUS2, GMBUS_ACTIVE, 0,
+ 10);
- if (ret)
- return 0;
- else
- return -ETIMEDOUT;
+ I915_WRITE_FW(GMBUS4, 0);
+ remove_wait_queue(&dev_priv->gmbus_wait_queue, &wait);
+
+ return ret;
}
static int
unsigned short addr, u8 *buf, unsigned int len,
u32 gmbus1_index)
{
- I915_WRITE(GMBUS1,
- gmbus1_index |
- GMBUS_CYCLE_WAIT |
- (len << GMBUS_BYTE_COUNT_SHIFT) |
- (addr << GMBUS_SLAVE_ADDR_SHIFT) |
- GMBUS_SLAVE_READ | GMBUS_SW_RDY);
+ I915_WRITE_FW(GMBUS1,
+ gmbus1_index |
+ GMBUS_CYCLE_WAIT |
+ (len << GMBUS_BYTE_COUNT_SHIFT) |
+ (addr << GMBUS_SLAVE_ADDR_SHIFT) |
+ GMBUS_SLAVE_READ | GMBUS_SW_RDY);
while (len) {
int ret;
u32 val, loop = 0;
- ret = gmbus_wait_hw_status(dev_priv, GMBUS_HW_RDY,
- GMBUS_HW_RDY_EN);
+ ret = gmbus_wait(dev_priv, GMBUS_HW_RDY, GMBUS_HW_RDY_EN);
if (ret)
return ret;
- val = I915_READ(GMBUS3);
+ val = I915_READ_FW(GMBUS3);
do {
*buf++ = val & 0xff;
val >>= 8;
len -= 1;
}
- I915_WRITE(GMBUS3, val);
- I915_WRITE(GMBUS1,
- GMBUS_CYCLE_WAIT |
- (chunk_size << GMBUS_BYTE_COUNT_SHIFT) |
- (addr << GMBUS_SLAVE_ADDR_SHIFT) |
- GMBUS_SLAVE_WRITE | GMBUS_SW_RDY);
+ I915_WRITE_FW(GMBUS3, val);
+ I915_WRITE_FW(GMBUS1,
+ GMBUS_CYCLE_WAIT |
+ (chunk_size << GMBUS_BYTE_COUNT_SHIFT) |
+ (addr << GMBUS_SLAVE_ADDR_SHIFT) |
+ GMBUS_SLAVE_WRITE | GMBUS_SW_RDY);
while (len) {
int ret;
val |= *buf++ << (8 * loop);
} while (--len && ++loop < 4);
- I915_WRITE(GMBUS3, val);
+ I915_WRITE_FW(GMBUS3, val);
- ret = gmbus_wait_hw_status(dev_priv, GMBUS_HW_RDY,
- GMBUS_HW_RDY_EN);
+ ret = gmbus_wait(dev_priv, GMBUS_HW_RDY, GMBUS_HW_RDY_EN);
if (ret)
return ret;
}
/* GMBUS5 holds 16-bit index */
if (gmbus5)
- I915_WRITE(GMBUS5, gmbus5);
+ I915_WRITE_FW(GMBUS5, gmbus5);
ret = gmbus_xfer_read(dev_priv, &msgs[1], gmbus1_index);
/* Clear GMBUS5 after each index transfer */
if (gmbus5)
- I915_WRITE(GMBUS5, 0);
+ I915_WRITE_FW(GMBUS5, 0);
return ret;
}
struct intel_gmbus,
adapter);
struct drm_i915_private *dev_priv = bus->dev_priv;
+ const unsigned int fw =
+ intel_uncore_forcewake_for_reg(dev_priv, GMBUS0,
+ FW_REG_READ | FW_REG_WRITE);
int i = 0, inc, try = 0;
int ret = 0;
+ intel_uncore_forcewake_get(dev_priv, fw);
retry:
- I915_WRITE(GMBUS0, bus->reg0);
+ I915_WRITE_FW(GMBUS0, bus->reg0);
for (; i < num; i += inc) {
inc = 1;
}
if (!ret)
- ret = gmbus_wait_hw_status(dev_priv, GMBUS_HW_WAIT_PHASE,
- GMBUS_HW_WAIT_EN);
+ ret = gmbus_wait(dev_priv,
+ GMBUS_HW_WAIT_PHASE, GMBUS_HW_WAIT_EN);
if (ret == -ETIMEDOUT)
goto timeout;
else if (ret)
* a STOP on the very first cycle. To simplify the code we
* unconditionally generate the STOP condition with an additional gmbus
* cycle. */
- I915_WRITE(GMBUS1, GMBUS_CYCLE_STOP | GMBUS_SW_RDY);
+ I915_WRITE_FW(GMBUS1, GMBUS_CYCLE_STOP | GMBUS_SW_RDY);
/* Mark the GMBUS interface as disabled after waiting for idle.
* We will re-enable it at the start of the next xfer,
adapter->name);
ret = -ETIMEDOUT;
}
- I915_WRITE(GMBUS0, 0);
+ I915_WRITE_FW(GMBUS0, 0);
ret = ret ?: i;
goto out;
* of resetting the GMBUS controller and so clearing the
* BUS_ERROR raised by the slave's NAK.
*/
- I915_WRITE(GMBUS1, GMBUS_SW_CLR_INT);
- I915_WRITE(GMBUS1, 0);
- I915_WRITE(GMBUS0, 0);
+ I915_WRITE_FW(GMBUS1, GMBUS_SW_CLR_INT);
+ I915_WRITE_FW(GMBUS1, 0);
+ I915_WRITE_FW(GMBUS0, 0);
DRM_DEBUG_KMS("GMBUS [%s] NAK for addr: %04x %c(%d)\n",
adapter->name, msgs[i].addr,
timeout:
DRM_DEBUG_KMS("GMBUS [%s] timed out, falling back to bit banging on pin %d\n",
bus->adapter.name, bus->reg0 & 0xff);
- I915_WRITE(GMBUS0, 0);
+ I915_WRITE_FW(GMBUS0, 0);
/*
* Hardware may not support GMBUS over these pins? Try GPIO bitbanging
ret = -EAGAIN;
out:
+ intel_uncore_forcewake_put(dev_priv, fw);
return ret;
}
int intel_setup_gmbus(struct drm_device *dev)
{
struct drm_i915_private *dev_priv = to_i915(dev);
+ struct pci_dev *pdev = dev_priv->drm.pdev;
struct intel_gmbus *bus;
unsigned int pin;
int ret;
"i915 gmbus %s",
get_gmbus_pin(dev_priv, pin)->name);
- bus->adapter.dev.parent = &dev->pdev->dev;
+ bus->adapter.dev.parent = &pdev->dev;
bus->dev_priv = dev_priv;
bus->adapter.algo = &gmbus_algorithm;
#define GEN8_CTX_STATUS_COMPLETE (1 << 4)
#define GEN8_CTX_STATUS_LITE_RESTORE (1 << 15)
+#define GEN8_CTX_STATUS_COMPLETED_MASK \
+ (GEN8_CTX_STATUS_ACTIVE_IDLE | \
+ GEN8_CTX_STATUS_PREEMPTED | \
+ GEN8_CTX_STATUS_ELEMENT_SWITCH)
+
#define CTX_LRI_HEADER_0 0x01
#define CTX_CONTEXT_CONTROL 0x02
#define CTX_RING_HEAD 0x04
{
struct drm_i915_private *dev_priv = engine->i915;
- if (IS_GEN8(dev_priv) || IS_GEN9(dev_priv))
- engine->idle_lite_restore_wa = ~0;
-
- engine->disable_lite_restore_wa = (IS_SKL_REVID(dev_priv, 0, SKL_REVID_B0) ||
- IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) &&
- (engine->id == VCS || engine->id == VCS2);
+ engine->disable_lite_restore_wa =
+ (IS_SKL_REVID(dev_priv, 0, SKL_REVID_B0) ||
+ IS_BXT_REVID(dev_priv, 0, BXT_REVID_A1)) &&
+ (engine->id == VCS || engine->id == VCS2);
engine->ctx_desc_template = GEN8_CTX_VALID;
if (IS_GEN8(dev_priv))
desc = ctx->desc_template; /* bits 3-4 */
desc |= engine->ctx_desc_template; /* bits 0-11 */
- desc |= ce->lrc_vma->node.start + LRC_PPHWSP_PN * PAGE_SIZE;
+ desc |= i915_ggtt_offset(ce->state) + LRC_PPHWSP_PN * PAGE_SIZE;
/* bits 12-31 */
desc |= (u64)ctx->hw_id << GEN8_CTX_ID_SHIFT; /* bits 32-52 */
return ctx->engine[engine->id].lrc_desc;
}
-static void execlists_elsp_write(struct drm_i915_gem_request *rq0,
- struct drm_i915_gem_request *rq1)
+static inline void
+execlists_context_status_change(struct drm_i915_gem_request *rq,
+ unsigned long status)
{
+ /*
+ * Only used when GVT-g is enabled now. When GVT-g is disabled,
+ * The compiler should eliminate this function as dead-code.
+ */
+ if (!IS_ENABLED(CONFIG_DRM_I915_GVT))
+ return;
- struct intel_engine_cs *engine = rq0->engine;
- struct drm_i915_private *dev_priv = rq0->i915;
- uint64_t desc[2];
-
- if (rq1) {
- desc[1] = intel_lr_context_descriptor(rq1->ctx, rq1->engine);
- rq1->elsp_submitted++;
- } else {
- desc[1] = 0;
- }
-
- desc[0] = intel_lr_context_descriptor(rq0->ctx, rq0->engine);
- rq0->elsp_submitted++;
-
- /* You must always write both descriptors in the order below. */
- I915_WRITE_FW(RING_ELSP(engine), upper_32_bits(desc[1]));
- I915_WRITE_FW(RING_ELSP(engine), lower_32_bits(desc[1]));
-
- I915_WRITE_FW(RING_ELSP(engine), upper_32_bits(desc[0]));
- /* The context is automatically loaded after the following */
- I915_WRITE_FW(RING_ELSP(engine), lower_32_bits(desc[0]));
-
- /* ELSP is a wo register, use another nearby reg for posting */
- POSTING_READ_FW(RING_EXECLIST_STATUS_LO(engine));
+ atomic_notifier_call_chain(&rq->ctx->status_notifier, status, rq);
}
static void
ASSIGN_CTX_PDP(ppgtt, reg_state, 0);
}
-static void execlists_update_context(struct drm_i915_gem_request *rq)
+static u64 execlists_update_context(struct drm_i915_gem_request *rq)
{
- struct intel_engine_cs *engine = rq->engine;
+ struct intel_context *ce = &rq->ctx->engine[rq->engine->id];
struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
- uint32_t *reg_state = rq->ctx->engine[engine->id].lrc_reg_state;
+ u32 *reg_state = ce->lrc_reg_state;
reg_state[CTX_RING_TAIL+1] = intel_ring_offset(rq->ring, rq->tail);
*/
if (ppgtt && !USES_FULL_48BIT_PPGTT(ppgtt->base.dev))
execlists_update_context_pdps(ppgtt, reg_state);
+
+ return ce->lrc_desc;
}
-static void execlists_elsp_submit_contexts(struct drm_i915_gem_request *rq0,
- struct drm_i915_gem_request *rq1)
+static void execlists_submit_ports(struct intel_engine_cs *engine)
{
- struct drm_i915_private *dev_priv = rq0->i915;
- unsigned int fw_domains = rq0->engine->fw_domains;
-
- execlists_update_context(rq0);
+ struct drm_i915_private *dev_priv = engine->i915;
+ struct execlist_port *port = engine->execlist_port;
+ u32 __iomem *elsp =
+ dev_priv->regs + i915_mmio_reg_offset(RING_ELSP(engine));
+ u64 desc[2];
- if (rq1)
- execlists_update_context(rq1);
+ if (!port[0].count)
+ execlists_context_status_change(port[0].request,
+ INTEL_CONTEXT_SCHEDULE_IN);
+ desc[0] = execlists_update_context(port[0].request);
+ engine->preempt_wa = port[0].count++; /* bdw only? fixed on skl? */
- spin_lock_irq(&dev_priv->uncore.lock);
- intel_uncore_forcewake_get__locked(dev_priv, fw_domains);
+ if (port[1].request) {
+ GEM_BUG_ON(port[1].count);
+ execlists_context_status_change(port[1].request,
+ INTEL_CONTEXT_SCHEDULE_IN);
+ desc[1] = execlists_update_context(port[1].request);
+ port[1].count = 1;
+ } else {
+ desc[1] = 0;
+ }
+ GEM_BUG_ON(desc[0] == desc[1]);
- execlists_elsp_write(rq0, rq1);
+ /* You must always write both descriptors in the order below. */
+ writel(upper_32_bits(desc[1]), elsp);
+ writel(lower_32_bits(desc[1]), elsp);
- intel_uncore_forcewake_put__locked(dev_priv, fw_domains);
- spin_unlock_irq(&dev_priv->uncore.lock);
+ writel(upper_32_bits(desc[0]), elsp);
+ /* The context is automatically loaded after the following */
+ writel(lower_32_bits(desc[0]), elsp);
}
-static inline void execlists_context_status_change(
- struct drm_i915_gem_request *rq,
- unsigned long status)
+static bool ctx_single_port_submission(const struct i915_gem_context *ctx)
{
- /*
- * Only used when GVT-g is enabled now. When GVT-g is disabled,
- * The compiler should eliminate this function as dead-code.
- */
- if (!IS_ENABLED(CONFIG_DRM_I915_GVT))
- return;
-
- atomic_notifier_call_chain(&rq->ctx->status_notifier, status, rq);
+ return (IS_ENABLED(CONFIG_DRM_I915_GVT) &&
+ ctx->execlists_force_single_submission);
}
-static void execlists_unqueue(struct intel_engine_cs *engine)
+static bool can_merge_ctx(const struct i915_gem_context *prev,
+ const struct i915_gem_context *next)
{
- struct drm_i915_gem_request *req0 = NULL, *req1 = NULL;
- struct drm_i915_gem_request *cursor, *tmp;
+ if (prev != next)
+ return false;
- assert_spin_locked(&engine->execlist_lock);
+ if (ctx_single_port_submission(prev))
+ return false;
- /*
- * If irqs are not active generate a warning as batches that finish
- * without the irqs may get lost and a GPU Hang may occur.
- */
- WARN_ON(!intel_irqs_enabled(engine->i915));
-
- /* Try to read in pairs */
- list_for_each_entry_safe(cursor, tmp, &engine->execlist_queue,
- execlist_link) {
- if (!req0) {
- req0 = cursor;
- } else if (req0->ctx == cursor->ctx) {
- /* Same ctx: ignore first request, as second request
- * will update tail past first request's workload */
- cursor->elsp_submitted = req0->elsp_submitted;
- list_del(&req0->execlist_link);
- i915_gem_request_put(req0);
- req0 = cursor;
- } else {
- if (IS_ENABLED(CONFIG_DRM_I915_GVT)) {
- /*
- * req0 (after merged) ctx requires single
- * submission, stop picking
- */
- if (req0->ctx->execlists_force_single_submission)
- break;
- /*
- * req0 ctx doesn't require single submission,
- * but next req ctx requires, stop picking
- */
- if (cursor->ctx->execlists_force_single_submission)
- break;
- }
- req1 = cursor;
- WARN_ON(req1->elsp_submitted);
- break;
- }
- }
+ return true;
+}
- if (unlikely(!req0))
- return;
+static void execlists_dequeue(struct intel_engine_cs *engine)
+{
+ struct drm_i915_gem_request *cursor, *last;
+ struct execlist_port *port = engine->execlist_port;
+ bool submit = false;
+
+ last = port->request;
+ if (last)
+ /* WaIdleLiteRestore:bdw,skl
+ * Apply the wa NOOPs to prevent ring:HEAD == req:TAIL
+ * as we resubmit the request. See gen8_emit_request()
+ * for where we prepare the padding after the end of the
+ * request.
+ */
+ last->tail = last->wa_tail;
- execlists_context_status_change(req0, INTEL_CONTEXT_SCHEDULE_IN);
+ GEM_BUG_ON(port[1].request);
- if (req1)
- execlists_context_status_change(req1,
- INTEL_CONTEXT_SCHEDULE_IN);
+ /* Hardware submission is through 2 ports. Conceptually each port
+ * has a (RING_START, RING_HEAD, RING_TAIL) tuple. RING_START is
+ * static for a context, and unique to each, so we only execute
+ * requests belonging to a single context from each ring. RING_HEAD
+ * is maintained by the CS in the context image, it marks the place
+ * where it got up to last time, and through RING_TAIL we tell the CS
+ * where we want to execute up to this time.
+ *
+ * In this list the requests are in order of execution. Consecutive
+ * requests from the same context are adjacent in the ringbuffer. We
+ * can combine these requests into a single RING_TAIL update:
+ *
+ * RING_HEAD...req1...req2
+ * ^- RING_TAIL
+ * since to execute req2 the CS must first execute req1.
+ *
+ * Our goal then is to point each port to the end of a consecutive
+ * sequence of requests as being the most optimal (fewest wake ups
+ * and context switches) submission.
+ */
- if (req0->elsp_submitted & engine->idle_lite_restore_wa) {
- /*
- * WaIdleLiteRestore: make sure we never cause a lite restore
- * with HEAD==TAIL.
+ spin_lock(&engine->execlist_lock);
+ list_for_each_entry(cursor, &engine->execlist_queue, execlist_link) {
+ /* Can we combine this request with the current port? It has to
+ * be the same context/ringbuffer and not have any exceptions
+ * (e.g. GVT saying never to combine contexts).
*
- * Apply the wa NOOPS to prevent ring:HEAD == req:TAIL as we
- * resubmit the request. See gen8_emit_request() for where we
- * prepare the padding after the end of the request.
+ * If we can combine the requests, we can execute both by
+ * updating the RING_TAIL to point to the end of the second
+ * request, and so we never need to tell the hardware about
+ * the first.
*/
- req0->tail += 8;
- req0->tail &= req0->ring->size - 1;
+ if (last && !can_merge_ctx(cursor->ctx, last->ctx)) {
+ /* If we are on the second port and cannot combine
+ * this request with the last, then we are done.
+ */
+ if (port != engine->execlist_port)
+ break;
+
+ /* If GVT overrides us we only ever submit port[0],
+ * leaving port[1] empty. Note that we also have
+ * to be careful that we don't queue the same
+ * context (even though a different request) to
+ * the second port.
+ */
+ if (ctx_single_port_submission(cursor->ctx))
+ break;
+
+ GEM_BUG_ON(last->ctx == cursor->ctx);
+
+ i915_gem_request_assign(&port->request, last);
+ port++;
+ }
+ last = cursor;
+ submit = true;
+ }
+ if (submit) {
+ /* Decouple all the requests submitted from the queue */
+ engine->execlist_queue.next = &cursor->execlist_link;
+ cursor->execlist_link.prev = &engine->execlist_queue;
+
+ i915_gem_request_assign(&port->request, last);
}
+ spin_unlock(&engine->execlist_lock);
- execlists_elsp_submit_contexts(req0, req1);
+ if (submit)
+ execlists_submit_ports(engine);
}
-static unsigned int
-execlists_check_remove_request(struct intel_engine_cs *engine, u32 ctx_id)
+static bool execlists_elsp_idle(struct intel_engine_cs *engine)
{
- struct drm_i915_gem_request *head_req;
-
- assert_spin_locked(&engine->execlist_lock);
-
- head_req = list_first_entry_or_null(&engine->execlist_queue,
- struct drm_i915_gem_request,
- execlist_link);
-
- if (WARN_ON(!head_req || (head_req->ctx_hw_id != ctx_id)))
- return 0;
-
- WARN(head_req->elsp_submitted == 0, "Never submitted head request\n");
-
- if (--head_req->elsp_submitted > 0)
- return 0;
-
- execlists_context_status_change(head_req, INTEL_CONTEXT_SCHEDULE_OUT);
-
- list_del(&head_req->execlist_link);
- i915_gem_request_put(head_req);
-
- return 1;
+ return !engine->execlist_port[0].request;
}
-static u32
-get_context_status(struct intel_engine_cs *engine, unsigned int read_pointer,
- u32 *context_id)
+static bool execlists_elsp_ready(struct intel_engine_cs *engine)
{
- struct drm_i915_private *dev_priv = engine->i915;
- u32 status;
-
- read_pointer %= GEN8_CSB_ENTRIES;
-
- status = I915_READ_FW(RING_CONTEXT_STATUS_BUF_LO(engine, read_pointer));
-
- if (status & GEN8_CTX_STATUS_IDLE_ACTIVE)
- return 0;
+ int port;
- *context_id = I915_READ_FW(RING_CONTEXT_STATUS_BUF_HI(engine,
- read_pointer));
+ port = 1; /* wait for a free slot */
+ if (engine->disable_lite_restore_wa || engine->preempt_wa)
+ port = 0; /* wait for GPU to be idle before continuing */
- return status;
+ return !engine->execlist_port[port].request;
}
/*
static void intel_lrc_irq_handler(unsigned long data)
{
struct intel_engine_cs *engine = (struct intel_engine_cs *)data;
+ struct execlist_port *port = engine->execlist_port;
struct drm_i915_private *dev_priv = engine->i915;
- u32 status_pointer;
- unsigned int read_pointer, write_pointer;
- u32 csb[GEN8_CSB_ENTRIES][2];
- unsigned int csb_read = 0, i;
- unsigned int submit_contexts = 0;
intel_uncore_forcewake_get(dev_priv, engine->fw_domains);
- status_pointer = I915_READ_FW(RING_CONTEXT_STATUS_PTR(engine));
-
- read_pointer = engine->next_context_status_buffer;
- write_pointer = GEN8_CSB_WRITE_PTR(status_pointer);
- if (read_pointer > write_pointer)
- write_pointer += GEN8_CSB_ENTRIES;
-
- while (read_pointer < write_pointer) {
- if (WARN_ON_ONCE(csb_read == GEN8_CSB_ENTRIES))
- break;
- csb[csb_read][0] = get_context_status(engine, ++read_pointer,
- &csb[csb_read][1]);
- csb_read++;
- }
-
- engine->next_context_status_buffer = write_pointer % GEN8_CSB_ENTRIES;
-
- /* Update the read pointer to the old write pointer. Manual ringbuffer
- * management ftw </sarcasm> */
- I915_WRITE_FW(RING_CONTEXT_STATUS_PTR(engine),
- _MASKED_FIELD(GEN8_CSB_READ_PTR_MASK,
- engine->next_context_status_buffer << 8));
-
- intel_uncore_forcewake_put(dev_priv, engine->fw_domains);
-
- spin_lock(&engine->execlist_lock);
+ if (!execlists_elsp_idle(engine)) {
+ u32 __iomem *csb_mmio =
+ dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_PTR(engine));
+ u32 __iomem *buf =
+ dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_BUF_LO(engine, 0));
+ unsigned int csb, head, tail;
+
+ csb = readl(csb_mmio);
+ head = GEN8_CSB_READ_PTR(csb);
+ tail = GEN8_CSB_WRITE_PTR(csb);
+ if (tail < head)
+ tail += GEN8_CSB_ENTRIES;
+ while (head < tail) {
+ unsigned int idx = ++head % GEN8_CSB_ENTRIES;
+ unsigned int status = readl(buf + 2 * idx);
+
+ if (!(status & GEN8_CTX_STATUS_COMPLETED_MASK))
+ continue;
+
+ GEM_BUG_ON(port[0].count == 0);
+ if (--port[0].count == 0) {
+ GEM_BUG_ON(status & GEN8_CTX_STATUS_PREEMPTED);
+ execlists_context_status_change(port[0].request,
+ INTEL_CONTEXT_SCHEDULE_OUT);
+
+ i915_gem_request_put(port[0].request);
+ port[0] = port[1];
+ memset(&port[1], 0, sizeof(port[1]));
+
+ engine->preempt_wa = false;
+ }
- for (i = 0; i < csb_read; i++) {
- if (unlikely(csb[i][0] & GEN8_CTX_STATUS_PREEMPTED)) {
- if (csb[i][0] & GEN8_CTX_STATUS_LITE_RESTORE) {
- if (execlists_check_remove_request(engine, csb[i][1]))
- WARN(1, "Lite Restored request removed from queue\n");
- } else
- WARN(1, "Preemption without Lite Restore\n");
+ GEM_BUG_ON(port[0].count == 0 &&
+ !(status & GEN8_CTX_STATUS_ACTIVE_IDLE));
}
- if (csb[i][0] & (GEN8_CTX_STATUS_ACTIVE_IDLE |
- GEN8_CTX_STATUS_ELEMENT_SWITCH))
- submit_contexts +=
- execlists_check_remove_request(engine, csb[i][1]);
+ writel(_MASKED_FIELD(GEN8_CSB_READ_PTR_MASK,
+ GEN8_CSB_WRITE_PTR(csb) << 8),
+ csb_mmio);
}
- if (submit_contexts) {
- if (!engine->disable_lite_restore_wa ||
- (csb[i][0] & GEN8_CTX_STATUS_ACTIVE_IDLE))
- execlists_unqueue(engine);
- }
+ if (execlists_elsp_ready(engine))
+ execlists_dequeue(engine);
- spin_unlock(&engine->execlist_lock);
-
- if (unlikely(submit_contexts > 2))
- DRM_ERROR("More than two context complete events?\n");
+ intel_uncore_forcewake_put(dev_priv, engine->fw_domains);
}
static void execlists_submit_request(struct drm_i915_gem_request *request)
{
struct intel_engine_cs *engine = request->engine;
- struct drm_i915_gem_request *cursor;
- int num_elements = 0;
+ unsigned long flags;
- spin_lock_bh(&engine->execlist_lock);
+ spin_lock_irqsave(&engine->execlist_lock, flags);
- list_for_each_entry(cursor, &engine->execlist_queue, execlist_link)
- if (++num_elements > 2)
- break;
-
- if (num_elements > 2) {
- struct drm_i915_gem_request *tail_req;
-
- tail_req = list_last_entry(&engine->execlist_queue,
- struct drm_i915_gem_request,
- execlist_link);
-
- if (request->ctx == tail_req->ctx) {
- WARN(tail_req->elsp_submitted != 0,
- "More than 2 already-submitted reqs queued\n");
- list_del(&tail_req->execlist_link);
- i915_gem_request_put(tail_req);
- }
- }
-
- i915_gem_request_get(request);
list_add_tail(&request->execlist_link, &engine->execlist_queue);
- request->ctx_hw_id = request->ctx->hw_id;
- if (num_elements == 0)
- execlists_unqueue(engine);
+ if (execlists_elsp_idle(engine))
+ tasklet_hi_schedule(&engine->irq_tasklet);
- spin_unlock_bh(&engine->execlist_lock);
+ spin_unlock_irqrestore(&engine->execlist_lock, flags);
}
int intel_logical_ring_alloc_request_extras(struct drm_i915_gem_request *request)
* going any further, as the i915_add_request() call
* later on mustn't fail ...
*/
- ret = i915_guc_wq_check_space(request);
+ ret = i915_guc_wq_reserve(request);
if (ret)
return ret;
}
intel_ring_emit(ring, MI_NOOP);
intel_ring_emit(ring, MI_NOOP);
intel_ring_advance(ring);
+ request->wa_tail = ring->tail;
/* We keep the previous context alive until we retire the following
* request. This ensures that any the context object is still pinned
return 0;
}
-void intel_execlists_cancel_requests(struct intel_engine_cs *engine)
-{
- struct drm_i915_gem_request *req, *tmp;
- LIST_HEAD(cancel_list);
-
- WARN_ON(!mutex_is_locked(&engine->i915->drm.struct_mutex));
-
- spin_lock_bh(&engine->execlist_lock);
- list_replace_init(&engine->execlist_queue, &cancel_list);
- spin_unlock_bh(&engine->execlist_lock);
-
- list_for_each_entry_safe(req, tmp, &cancel_list, execlist_link) {
- list_del(&req->execlist_link);
- i915_gem_request_put(req);
- }
-}
-
static int intel_lr_context_pin(struct i915_gem_context *ctx,
struct intel_engine_cs *engine)
{
- struct drm_i915_private *dev_priv = ctx->i915;
struct intel_context *ce = &ctx->engine[engine->id];
void *vaddr;
u32 *lrc_reg_state;
if (ce->pin_count++)
return 0;
- ret = i915_gem_object_ggtt_pin(ce->state, NULL,
- 0, GEN8_LR_CONTEXT_ALIGN,
- PIN_OFFSET_BIAS | GUC_WOPCM_TOP);
+ ret = i915_vma_pin(ce->state, 0, GEN8_LR_CONTEXT_ALIGN,
+ PIN_OFFSET_BIAS | GUC_WOPCM_TOP | PIN_GLOBAL);
if (ret)
goto err;
- vaddr = i915_gem_object_pin_map(ce->state);
+ vaddr = i915_gem_object_pin_map(ce->state->obj, I915_MAP_WB);
if (IS_ERR(vaddr)) {
ret = PTR_ERR(vaddr);
- goto unpin_ctx_obj;
+ goto unpin_vma;
}
lrc_reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE;
if (ret)
goto unpin_map;
- ce->lrc_vma = i915_gem_obj_to_ggtt(ce->state);
intel_lr_context_descriptor_update(ctx, engine);
- lrc_reg_state[CTX_RING_BUFFER_START+1] = ce->ring->vma->node.start;
+ lrc_reg_state[CTX_RING_BUFFER_START+1] =
+ i915_ggtt_offset(ce->ring->vma);
ce->lrc_reg_state = lrc_reg_state;
- ce->state->dirty = true;
+ ce->state->obj->dirty = true;
/* Invalidate GuC TLB. */
- if (i915.enable_guc_submission)
+ if (i915.enable_guc_submission) {
+ struct drm_i915_private *dev_priv = ctx->i915;
I915_WRITE(GEN8_GTCR, GEN8_GTCR_INVALIDATE);
+ }
i915_gem_context_get(ctx);
return 0;
unpin_map:
- i915_gem_object_unpin_map(ce->state);
-unpin_ctx_obj:
- i915_gem_object_ggtt_unpin(ce->state);
+ i915_gem_object_unpin_map(ce->state->obj);
+unpin_vma:
+ __i915_vma_unpin(ce->state);
err:
ce->pin_count = 0;
return ret;
intel_ring_unpin(ce->ring);
- i915_gem_object_unpin_map(ce->state);
- i915_gem_object_ggtt_unpin(ce->state);
-
- ce->lrc_vma = NULL;
- ce->lrc_desc = 0;
- ce->lrc_reg_state = NULL;
+ i915_gem_object_unpin_map(ce->state->obj);
+ i915_vma_unpin(ce->state);
i915_gem_context_put(ctx);
}
wa_ctx_emit(batch, index, (MI_STORE_REGISTER_MEM_GEN8 |
MI_SRM_LRM_GLOBAL_GTT));
wa_ctx_emit_reg(batch, index, GEN8_L3SQCREG4);
- wa_ctx_emit(batch, index, engine->scratch.gtt_offset + 256);
+ wa_ctx_emit(batch, index, i915_ggtt_offset(engine->scratch) + 256);
wa_ctx_emit(batch, index, 0);
wa_ctx_emit(batch, index, MI_LOAD_REGISTER_IMM(1));
wa_ctx_emit(batch, index, (MI_LOAD_REGISTER_MEM_GEN8 |
MI_SRM_LRM_GLOBAL_GTT));
wa_ctx_emit_reg(batch, index, GEN8_L3SQCREG4);
- wa_ctx_emit(batch, index, engine->scratch.gtt_offset + 256);
+ wa_ctx_emit(batch, index, i915_ggtt_offset(engine->scratch) + 256);
wa_ctx_emit(batch, index, 0);
return index;
/* WaClearSlmSpaceAtContextSwitch:bdw,chv */
/* Actual scratch location is at 128 bytes offset */
- scratch_addr = engine->scratch.gtt_offset + 2*CACHELINE_BYTES;
+ scratch_addr = i915_ggtt_offset(engine->scratch) + 2 * CACHELINE_BYTES;
wa_ctx_emit(batch, index, GFX_OP_PIPE_CONTROL(6));
wa_ctx_emit(batch, index, (PIPE_CONTROL_FLUSH_L3 |
/* WaClearSlmSpaceAtContextSwitch:kbl */
/* Actual scratch location is at 128 bytes offset */
if (IS_KBL_REVID(dev_priv, 0, KBL_REVID_A0)) {
- uint32_t scratch_addr
- = engine->scratch.gtt_offset + 2*CACHELINE_BYTES;
+ u32 scratch_addr =
+ i915_ggtt_offset(engine->scratch) + 2 * CACHELINE_BYTES;
wa_ctx_emit(batch, index, GFX_OP_PIPE_CONTROL(6));
wa_ctx_emit(batch, index, (PIPE_CONTROL_FLUSH_L3 |
static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size)
{
- int ret;
+ struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
+ int err;
- engine->wa_ctx.obj = i915_gem_object_create(&engine->i915->drm,
- PAGE_ALIGN(size));
- if (IS_ERR(engine->wa_ctx.obj)) {
- DRM_DEBUG_DRIVER("alloc LRC WA ctx backing obj failed.\n");
- ret = PTR_ERR(engine->wa_ctx.obj);
- engine->wa_ctx.obj = NULL;
- return ret;
- }
+ obj = i915_gem_object_create(&engine->i915->drm, PAGE_ALIGN(size));
+ if (IS_ERR(obj))
+ return PTR_ERR(obj);
- ret = i915_gem_object_ggtt_pin(engine->wa_ctx.obj, NULL,
- 0, PAGE_SIZE, 0);
- if (ret) {
- DRM_DEBUG_DRIVER("pin LRC WA ctx backing obj failed: %d\n",
- ret);
- i915_gem_object_put(engine->wa_ctx.obj);
- return ret;
+ vma = i915_vma_create(obj, &engine->i915->ggtt.base, NULL);
+ if (IS_ERR(vma)) {
+ err = PTR_ERR(vma);
+ goto err;
}
+ err = i915_vma_pin(vma, 0, PAGE_SIZE, PIN_GLOBAL | PIN_HIGH);
+ if (err)
+ goto err;
+
+ engine->wa_ctx.vma = vma;
return 0;
+
+err:
+ i915_gem_object_put(obj);
+ return err;
}
static void lrc_destroy_wa_ctx_obj(struct intel_engine_cs *engine)
{
- if (engine->wa_ctx.obj) {
- i915_gem_object_ggtt_unpin(engine->wa_ctx.obj);
- i915_gem_object_put(engine->wa_ctx.obj);
- engine->wa_ctx.obj = NULL;
- }
+ i915_vma_unpin_and_release(&engine->wa_ctx.vma);
}
static int intel_init_workaround_bb(struct intel_engine_cs *engine)
{
- int ret;
+ struct i915_ctx_workarounds *wa_ctx = &engine->wa_ctx;
uint32_t *batch;
uint32_t offset;
struct page *page;
- struct i915_ctx_workarounds *wa_ctx = &engine->wa_ctx;
+ int ret;
WARN_ON(engine->id != RCS);
}
/* some WA perform writes to scratch page, ensure it is valid */
- if (engine->scratch.obj == NULL) {
+ if (!engine->scratch) {
DRM_ERROR("scratch page not allocated for %s\n", engine->name);
return -EINVAL;
}
return ret;
}
- page = i915_gem_object_get_dirty_page(wa_ctx->obj, 0);
+ page = i915_gem_object_get_dirty_page(wa_ctx->vma->obj, 0);
batch = kmap_atomic(page);
offset = 0;
struct drm_i915_private *dev_priv = engine->i915;
I915_WRITE(RING_HWS_PGA(engine->mmio_base),
- (u32)engine->status_page.gfx_addr);
+ engine->status_page.ggtt_offset);
POSTING_READ(RING_HWS_PGA(engine->mmio_base));
}
static int gen8_init_common_ring(struct intel_engine_cs *engine)
{
struct drm_i915_private *dev_priv = engine->i915;
- unsigned int next_context_status_buffer_hw;
+ int ret;
+
+ ret = intel_mocs_init_engine(engine);
+ if (ret)
+ return ret;
lrc_init_hws(engine);
- I915_WRITE_IMR(engine,
- ~(engine->irq_enable_mask | engine->irq_keep_mask));
+ intel_engine_reset_irq(engine);
+
I915_WRITE(RING_HWSTAM(engine->mmio_base), 0xffffffff);
I915_WRITE(RING_MODE_GEN7(engine),
_MASKED_BIT_DISABLE(GFX_REPLAY_MODE) |
_MASKED_BIT_ENABLE(GFX_RUN_LIST_ENABLE));
- POSTING_READ(RING_MODE_GEN7(engine));
-
- /*
- * Instead of resetting the Context Status Buffer (CSB) read pointer to
- * zero, we need to read the write pointer from hardware and use its
- * value because "this register is power context save restored".
- * Effectively, these states have been observed:
- *
- * | Suspend-to-idle (freeze) | Suspend-to-RAM (mem) |
- * BDW | CSB regs not reset | CSB regs reset |
- * CHT | CSB regs not reset | CSB regs not reset |
- * SKL | ? | ? |
- * BXT | ? | ? |
- */
- next_context_status_buffer_hw =
- GEN8_CSB_WRITE_PTR(I915_READ(RING_CONTEXT_STATUS_PTR(engine)));
-
- /*
- * When the CSB registers are reset (also after power-up / gpu reset),
- * CSB write pointer is set to all 1's, which is not valid, use '5' in
- * this special case, so the first element read is CSB[0].
- */
- if (next_context_status_buffer_hw == GEN8_CSB_PTR_MASK)
- next_context_status_buffer_hw = (GEN8_CSB_ENTRIES - 1);
- engine->next_context_status_buffer = next_context_status_buffer_hw;
DRM_DEBUG_DRIVER("Execlists enabled for %s\n", engine->name);
intel_engine_init_hangcheck(engine);
- return intel_mocs_init_engine(engine);
+ if (!execlists_elsp_idle(engine))
+ execlists_submit_ports(engine);
+
+ return 0;
}
static int gen8_init_render_ring(struct intel_engine_cs *engine)
return init_workarounds_ring(engine);
}
+static void reset_common_ring(struct intel_engine_cs *engine,
+ struct drm_i915_gem_request *request)
+{
+ struct drm_i915_private *dev_priv = engine->i915;
+ struct execlist_port *port = engine->execlist_port;
+ struct intel_context *ce = &request->ctx->engine[engine->id];
+
+ /* Move the RING_HEAD onto the breadcrumb, past the hanging batch */
+ ce->lrc_reg_state[CTX_RING_HEAD+1] = request->postfix;
+ request->ring->head = request->postfix;
+ request->ring->last_retired_head = -1;
+ intel_ring_update_space(request->ring);
+
+ if (i915.enable_guc_submission)
+ return;
+
+ /* Catch up with any missed context-switch interrupts */
+ I915_WRITE(RING_CONTEXT_STATUS_PTR(engine), _MASKED_FIELD(0xffff, 0));
+ if (request->ctx != port[0].request->ctx) {
+ i915_gem_request_put(port[0].request);
+ port[0] = port[1];
+ memset(&port[1], 0, sizeof(port[1]));
+ }
+
+ /* CS is stopped, and we will resubmit both ports on resume */
+ GEM_BUG_ON(request->ctx != port[0].request->ctx);
+ port[0].count = 0;
+ port[1].count = 0;
+}
+
static int intel_logical_ring_emit_pdps(struct drm_i915_gem_request *req)
{
struct i915_hw_ppgtt *ppgtt = req->ctx->ppgtt;
{
struct intel_ring *ring = request->ring;
struct intel_engine_cs *engine = request->engine;
- u32 scratch_addr = engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+ u32 scratch_addr =
+ i915_ggtt_offset(engine->scratch) + 2 * CACHELINE_BYTES;
bool vf_flush_wa = false, dc_flush_wa = false;
u32 flags = 0;
int ret;
intel_engine_cleanup_common(engine);
- if (engine->status_page.obj) {
- i915_gem_object_unpin_map(engine->status_page.obj);
- engine->status_page.obj = NULL;
+ if (engine->status_page.vma) {
+ i915_gem_object_unpin_map(engine->status_page.vma->obj);
+ engine->status_page.vma = NULL;
}
intel_lr_context_unpin(dev_priv->kernel_context, engine);
- engine->idle_lite_restore_wa = 0;
- engine->disable_lite_restore_wa = false;
- engine->ctx_desc_template = 0;
-
lrc_destroy_wa_ctx_obj(engine);
engine->i915 = NULL;
}
{
/* Default vfuncs which can be overriden by each engine. */
engine->init_hw = gen8_init_common_ring;
+ engine->reset_hw = reset_common_ring;
engine->emit_flush = gen8_emit_flush;
engine->emit_request = gen8_emit_request;
engine->submit_request = execlists_submit_request;
}
static int
-lrc_setup_hws(struct intel_engine_cs *engine,
- struct drm_i915_gem_object *dctx_obj)
+lrc_setup_hws(struct intel_engine_cs *engine, struct i915_vma *vma)
{
+ const int hws_offset = LRC_PPHWSP_PN * PAGE_SIZE;
void *hws;
/* The HWSP is part of the default context object in LRC mode. */
- engine->status_page.gfx_addr = i915_gem_obj_ggtt_offset(dctx_obj) +
- LRC_PPHWSP_PN * PAGE_SIZE;
- hws = i915_gem_object_pin_map(dctx_obj);
+ hws = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
if (IS_ERR(hws))
return PTR_ERR(hws);
- engine->status_page.page_addr = hws + LRC_PPHWSP_PN * PAGE_SIZE;
- engine->status_page.obj = dctx_obj;
+
+ engine->status_page.page_addr = hws + hws_offset;
+ engine->status_page.ggtt_offset = i915_ggtt_offset(vma) + hws_offset;
+ engine->status_page.vma = vma;
return 0;
}
else
engine->init_hw = gen8_init_render_ring;
engine->init_context = gen8_init_rcs_context;
- engine->cleanup = intel_fini_pipe_control;
engine->emit_flush = gen8_emit_flush_render;
engine->emit_request = gen8_emit_request_render;
- ret = intel_init_pipe_control(engine, 4096);
+ ret = intel_engine_create_scratch(engine, 4096);
if (ret)
return ret;
* must make an explicit request through RPCS for full
* enablement.
*/
- if (INTEL_INFO(dev_priv)->has_slice_pg) {
+ if (INTEL_INFO(dev_priv)->sseu.has_slice_pg) {
rpcs |= GEN8_RPCS_S_CNT_ENABLE;
- rpcs |= INTEL_INFO(dev_priv)->slice_total <<
+ rpcs |= hweight8(INTEL_INFO(dev_priv)->sseu.slice_mask) <<
GEN8_RPCS_S_CNT_SHIFT;
rpcs |= GEN8_RPCS_ENABLE;
}
- if (INTEL_INFO(dev_priv)->has_subslice_pg) {
+ if (INTEL_INFO(dev_priv)->sseu.has_subslice_pg) {
rpcs |= GEN8_RPCS_SS_CNT_ENABLE;
- rpcs |= INTEL_INFO(dev_priv)->subslice_per_slice <<
+ rpcs |= hweight8(INTEL_INFO(dev_priv)->sseu.subslice_mask) <<
GEN8_RPCS_SS_CNT_SHIFT;
rpcs |= GEN8_RPCS_ENABLE;
}
- if (INTEL_INFO(dev_priv)->has_eu_pg) {
- rpcs |= INTEL_INFO(dev_priv)->eu_per_subslice <<
+ if (INTEL_INFO(dev_priv)->sseu.has_eu_pg) {
+ rpcs |= INTEL_INFO(dev_priv)->sseu.eu_per_subslice <<
GEN8_RPCS_EU_MIN_SHIFT;
- rpcs |= INTEL_INFO(dev_priv)->eu_per_subslice <<
+ rpcs |= INTEL_INFO(dev_priv)->sseu.eu_per_subslice <<
GEN8_RPCS_EU_MAX_SHIFT;
rpcs |= GEN8_RPCS_ENABLE;
}
return ret;
}
- vaddr = i915_gem_object_pin_map(ctx_obj);
+ vaddr = i915_gem_object_pin_map(ctx_obj, I915_MAP_WB);
if (IS_ERR(vaddr)) {
ret = PTR_ERR(vaddr);
DRM_DEBUG_DRIVER("Could not map object pages! (%d)\n", ret);
RING_INDIRECT_CTX(engine->mmio_base), 0);
ASSIGN_CTX_REG(reg_state, CTX_RCS_INDIRECT_CTX_OFFSET,
RING_INDIRECT_CTX_OFFSET(engine->mmio_base), 0);
- if (engine->wa_ctx.obj) {
+ if (engine->wa_ctx.vma) {
struct i915_ctx_workarounds *wa_ctx = &engine->wa_ctx;
- uint32_t ggtt_offset = i915_gem_obj_ggtt_offset(wa_ctx->obj);
+ u32 ggtt_offset = i915_ggtt_offset(wa_ctx->vma);
reg_state[CTX_RCS_INDIRECT_CTX+1] =
(ggtt_offset + wa_ctx->indirect_ctx.offset * sizeof(uint32_t)) |
{
struct drm_i915_gem_object *ctx_obj;
struct intel_context *ce = &ctx->engine[engine->id];
+ struct i915_vma *vma;
uint32_t context_size;
struct intel_ring *ring;
int ret;
return PTR_ERR(ctx_obj);
}
+ vma = i915_vma_create(ctx_obj, &ctx->i915->ggtt.base, NULL);
+ if (IS_ERR(vma)) {
+ ret = PTR_ERR(vma);
+ goto error_deref_obj;
+ }
+
ring = intel_engine_create_ring(engine, ctx->ring_size);
if (IS_ERR(ring)) {
ret = PTR_ERR(ring);
}
ce->ring = ring;
- ce->state = ctx_obj;
+ ce->state = vma;
ce->initialised = engine->init_context == NULL;
return 0;
intel_ring_free(ring);
error_deref_obj:
i915_gem_object_put(ctx_obj);
- ce->ring = NULL;
- ce->state = NULL;
return ret;
}
-void intel_lr_context_reset(struct drm_i915_private *dev_priv,
- struct i915_gem_context *ctx)
+void intel_lr_context_resume(struct drm_i915_private *dev_priv)
{
+ struct i915_gem_context *ctx = dev_priv->kernel_context;
struct intel_engine_cs *engine;
for_each_engine(engine, dev_priv) {
struct intel_context *ce = &ctx->engine[engine->id];
- struct drm_i915_gem_object *ctx_obj = ce->state;
void *vaddr;
uint32_t *reg_state;
- if (!ctx_obj)
+ if (!ce->state)
continue;
- vaddr = i915_gem_object_pin_map(ctx_obj);
+ vaddr = i915_gem_object_pin_map(ce->state->obj, I915_MAP_WB);
if (WARN_ON(IS_ERR(vaddr)))
continue;
reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE;
- ctx_obj->dirty = true;
reg_state[CTX_RING_HEAD+1] = 0;
reg_state[CTX_RING_TAIL+1] = 0;
- i915_gem_object_unpin_map(ctx_obj);
+ ce->state->obj->dirty = true;
+ i915_gem_object_unpin_map(ce->state->obj);
ce->ring->head = 0;
ce->ring->tail = 0;
struct drm_i915_private;
-void intel_lr_context_reset(struct drm_i915_private *dev_priv,
- struct i915_gem_context *ctx);
+void intel_lr_context_resume(struct drm_i915_private *dev_priv);
uint64_t intel_lr_context_descriptor(struct i915_gem_context *ctx,
struct intel_engine_cs *engine);
int enable_execlists);
void intel_execlists_enable_submission(struct drm_i915_private *dev_priv);
-void intel_execlists_cancel_requests(struct intel_engine_cs *engine);
-
#endif /* _INTEL_LRC_H_ */
struct notifier_block lid_notifier;
};
+struct intel_lvds_pps {
+ /* 100us units */
+ int t1_t2;
+ int t3;
+ int t4;
+ int t5;
+ int tx;
+
+ int divider;
+
+ int port;
+ bool powerdown_on_reset;
+};
+
struct intel_lvds_encoder {
struct intel_encoder base;
i915_reg_t reg;
u32 a3_power;
+ struct intel_lvds_pps init_pps;
+ u32 init_lvds_val;
+
struct intel_lvds_connector *attached_connector;
};
pipe_config->base.adjusted_mode.crtc_clock = pipe_config->port_clock;
}
-static void intel_pre_enable_lvds(struct intel_encoder *encoder)
+static void intel_lvds_pps_get_hw_state(struct drm_i915_private *dev_priv,
+ struct intel_lvds_pps *pps)
+{
+ u32 val;
+
+ pps->powerdown_on_reset = I915_READ(PP_CONTROL(0)) & PANEL_POWER_RESET;
+
+ val = I915_READ(PP_ON_DELAYS(0));
+ pps->port = (val & PANEL_PORT_SELECT_MASK) >>
+ PANEL_PORT_SELECT_SHIFT;
+ pps->t1_t2 = (val & PANEL_POWER_UP_DELAY_MASK) >>
+ PANEL_POWER_UP_DELAY_SHIFT;
+ pps->t5 = (val & PANEL_LIGHT_ON_DELAY_MASK) >>
+ PANEL_LIGHT_ON_DELAY_SHIFT;
+
+ val = I915_READ(PP_OFF_DELAYS(0));
+ pps->t3 = (val & PANEL_POWER_DOWN_DELAY_MASK) >>
+ PANEL_POWER_DOWN_DELAY_SHIFT;
+ pps->tx = (val & PANEL_LIGHT_OFF_DELAY_MASK) >>
+ PANEL_LIGHT_OFF_DELAY_SHIFT;
+
+ val = I915_READ(PP_DIVISOR(0));
+ pps->divider = (val & PP_REFERENCE_DIVIDER_MASK) >>
+ PP_REFERENCE_DIVIDER_SHIFT;
+ val = (val & PANEL_POWER_CYCLE_DELAY_MASK) >>
+ PANEL_POWER_CYCLE_DELAY_SHIFT;
+ /*
+ * Remove the BSpec specified +1 (100ms) offset that accounts for a
+ * too short power-cycle delay due to the asynchronous programming of
+ * the register.
+ */
+ if (val)
+ val--;
+ /* Convert from 100ms to 100us units */
+ pps->t4 = val * 1000;
+
+ if (INTEL_INFO(dev_priv)->gen <= 4 &&
+ pps->t1_t2 == 0 && pps->t5 == 0 && pps->t3 == 0 && pps->tx == 0) {
+ DRM_DEBUG_KMS("Panel power timings uninitialized, "
+ "setting defaults\n");
+ /* Set T2 to 40ms and T5 to 200ms in 100 usec units */
+ pps->t1_t2 = 40 * 10;
+ pps->t5 = 200 * 10;
+ /* Set T3 to 35ms and Tx to 200ms in 100 usec units */
+ pps->t3 = 35 * 10;
+ pps->tx = 200 * 10;
+ }
+
+ DRM_DEBUG_DRIVER("LVDS PPS:t1+t2 %d t3 %d t4 %d t5 %d tx %d "
+ "divider %d port %d powerdown_on_reset %d\n",
+ pps->t1_t2, pps->t3, pps->t4, pps->t5, pps->tx,
+ pps->divider, pps->port, pps->powerdown_on_reset);
+}
+
+static void intel_lvds_pps_init_hw(struct drm_i915_private *dev_priv,
+ struct intel_lvds_pps *pps)
+{
+ u32 val;
+
+ val = I915_READ(PP_CONTROL(0));
+ WARN_ON((val & PANEL_UNLOCK_MASK) != PANEL_UNLOCK_REGS);
+ if (pps->powerdown_on_reset)
+ val |= PANEL_POWER_RESET;
+ I915_WRITE(PP_CONTROL(0), val);
+
+ I915_WRITE(PP_ON_DELAYS(0), (pps->port << PANEL_PORT_SELECT_SHIFT) |
+ (pps->t1_t2 << PANEL_POWER_UP_DELAY_SHIFT) |
+ (pps->t5 << PANEL_LIGHT_ON_DELAY_SHIFT));
+ I915_WRITE(PP_OFF_DELAYS(0), (pps->t3 << PANEL_POWER_DOWN_DELAY_SHIFT) |
+ (pps->tx << PANEL_LIGHT_OFF_DELAY_SHIFT));
+
+ val = pps->divider << PP_REFERENCE_DIVIDER_SHIFT;
+ val |= (DIV_ROUND_UP(pps->t4, 1000) + 1) <<
+ PANEL_POWER_CYCLE_DELAY_SHIFT;
+ I915_WRITE(PP_DIVISOR(0), val);
+}
+
+static void intel_pre_enable_lvds(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
- struct drm_device *dev = encoder->base.dev;
- struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_crtc *crtc = to_intel_crtc(encoder->base.crtc);
- const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode;
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+ struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+ const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
int pipe = crtc->pipe;
u32 temp;
- if (HAS_PCH_SPLIT(dev)) {
+ if (HAS_PCH_SPLIT(dev_priv)) {
assert_fdi_rx_pll_disabled(dev_priv, pipe);
assert_shared_dpll_disabled(dev_priv,
- crtc->config->shared_dpll);
+ pipe_config->shared_dpll);
} else {
assert_pll_disabled(dev_priv, pipe);
}
- temp = I915_READ(lvds_encoder->reg);
+ intel_lvds_pps_init_hw(dev_priv, &lvds_encoder->init_pps);
+
+ temp = lvds_encoder->init_lvds_val;
temp |= LVDS_PORT_EN | LVDS_A0A2_CLKA_POWER_UP;
- if (HAS_PCH_CPT(dev)) {
+ if (HAS_PCH_CPT(dev_priv)) {
temp &= ~PORT_TRANS_SEL_MASK;
temp |= PORT_TRANS_SEL_CPT(pipe);
} else {
/* set the corresponsding LVDS_BORDER bit */
temp &= ~LVDS_BORDER_ENABLE;
- temp |= crtc->config->gmch_pfit.lvds_border_bits;
+ temp |= pipe_config->gmch_pfit.lvds_border_bits;
/* Set the B0-B3 data pairs corresponding to whether we're going to
* set the DPLLs for dual-channel mode or not.
*/
if (IS_GEN4(dev_priv)) {
/* Bspec wording suggests that LVDS port dithering only exists
* for 18bpp panels. */
- if (crtc->config->dither && crtc->config->pipe_bpp == 18)
+ if (pipe_config->dither && pipe_config->pipe_bpp == 18)
temp |= LVDS_ENABLE_DITHER;
else
temp &= ~LVDS_ENABLE_DITHER;
/**
* Sets the power state for the panel.
*/
-static void intel_enable_lvds(struct intel_encoder *encoder)
+static void intel_enable_lvds(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
struct intel_connector *intel_connector =
&lvds_encoder->attached_connector->base;
struct drm_i915_private *dev_priv = to_i915(dev);
- i915_reg_t ctl_reg, stat_reg;
-
- if (HAS_PCH_SPLIT(dev)) {
- ctl_reg = PCH_PP_CONTROL;
- stat_reg = PCH_PP_STATUS;
- } else {
- ctl_reg = PP_CONTROL;
- stat_reg = PP_STATUS;
- }
I915_WRITE(lvds_encoder->reg, I915_READ(lvds_encoder->reg) | LVDS_PORT_EN);
- I915_WRITE(ctl_reg, I915_READ(ctl_reg) | POWER_TARGET_ON);
+ I915_WRITE(PP_CONTROL(0), I915_READ(PP_CONTROL(0)) | PANEL_POWER_ON);
POSTING_READ(lvds_encoder->reg);
- if (intel_wait_for_register(dev_priv, stat_reg, PP_ON, PP_ON, 1000))
+ if (intel_wait_for_register(dev_priv, PP_STATUS(0), PP_ON, PP_ON, 1000))
DRM_ERROR("timed out waiting for panel to power on\n");
intel_panel_enable_backlight(intel_connector);
}
-static void intel_disable_lvds(struct intel_encoder *encoder)
+static void intel_disable_lvds(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
- struct drm_device *dev = encoder->base.dev;
struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
- struct drm_i915_private *dev_priv = to_i915(dev);
- i915_reg_t ctl_reg, stat_reg;
-
- if (HAS_PCH_SPLIT(dev)) {
- ctl_reg = PCH_PP_CONTROL;
- stat_reg = PCH_PP_STATUS;
- } else {
- ctl_reg = PP_CONTROL;
- stat_reg = PP_STATUS;
- }
+ struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
- I915_WRITE(ctl_reg, I915_READ(ctl_reg) & ~POWER_TARGET_ON);
- if (intel_wait_for_register(dev_priv, stat_reg, PP_ON, 0, 1000))
+ I915_WRITE(PP_CONTROL(0), I915_READ(PP_CONTROL(0)) & ~PANEL_POWER_ON);
+ if (intel_wait_for_register(dev_priv, PP_STATUS(0), PP_ON, 0, 1000))
DRM_ERROR("timed out waiting for panel to power off\n");
I915_WRITE(lvds_encoder->reg, I915_READ(lvds_encoder->reg) & ~LVDS_PORT_EN);
POSTING_READ(lvds_encoder->reg);
}
-static void gmch_disable_lvds(struct intel_encoder *encoder)
+static void gmch_disable_lvds(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
+
{
struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
struct intel_connector *intel_connector =
intel_panel_disable_backlight(intel_connector);
- intel_disable_lvds(encoder);
+ intel_disable_lvds(encoder, old_crtc_state, old_conn_state);
}
-static void pch_disable_lvds(struct intel_encoder *encoder)
+static void pch_disable_lvds(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
struct intel_connector *intel_connector =
intel_panel_disable_backlight(intel_connector);
}
-static void pch_post_disable_lvds(struct intel_encoder *encoder)
+static void pch_post_disable_lvds(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
- intel_disable_lvds(encoder);
+ intel_disable_lvds(encoder, old_crtc_state, old_conn_state);
}
static enum drm_mode_status
}
static bool intel_lvds_compute_config(struct intel_encoder *intel_encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = intel_encoder->base.dev;
struct intel_lvds_encoder *lvds_encoder =
int pipe;
u8 pin;
- /*
- * Unlock registers and just leave them unlocked. Do this before
- * checking quirk lists to avoid bogus WARNINGs.
- */
- if (HAS_PCH_SPLIT(dev)) {
- I915_WRITE(PCH_PP_CONTROL,
- I915_READ(PCH_PP_CONTROL) | PANEL_UNLOCK_REGS);
- } else if (INTEL_INFO(dev_priv)->gen < 5) {
- I915_WRITE(PP_CONTROL,
- I915_READ(PP_CONTROL) | PANEL_UNLOCK_REGS);
- }
if (!intel_lvds_supported(dev))
return;
DRM_DEBUG_KMS("LVDS is not present in VBT, but enabled anyway\n");
}
- /* Set the Panel Power On/Off timings if uninitialized. */
- if (INTEL_INFO(dev_priv)->gen < 5 &&
- I915_READ(PP_ON_DELAYS) == 0 && I915_READ(PP_OFF_DELAYS) == 0) {
- /* Set T2 to 40ms and T5 to 200ms */
- I915_WRITE(PP_ON_DELAYS, 0x019007d0);
-
- /* Set T3 to 35ms and Tx to 200ms */
- I915_WRITE(PP_OFF_DELAYS, 0x015e07d0);
-
- DRM_DEBUG_KMS("Panel power timings uninitialized, setting defaults\n");
- }
-
lvds_encoder = kzalloc(sizeof(*lvds_encoder), GFP_KERNEL);
if (!lvds_encoder)
return;
dev->mode_config.scaling_mode_property,
DRM_MODE_SCALE_ASPECT);
intel_connector->panel.fitting_mode = DRM_MODE_SCALE_ASPECT;
+
+ intel_lvds_pps_get_hw_state(dev_priv, &lvds_encoder->init_pps);
+ lvds_encoder->init_lvds_val = lvds;
+
/*
* LVDS discovery:
* 1) check for EDID on DDC
return err;
}
+static int intel_use_opregion_panel_type_callback(const struct dmi_system_id *id)
+{
+ DRM_INFO("Using panel type from OpRegion on %s\n", id->ident);
+ return 1;
+}
+
+static const struct dmi_system_id intel_use_opregion_panel_type[] = {
+ {
+ .callback = intel_use_opregion_panel_type_callback,
+ .ident = "Conrac GmbH IX45GM2",
+ .matches = {DMI_MATCH(DMI_SYS_VENDOR, "Conrac GmbH"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "IX45GM2"),
+ },
+ },
+ { }
+};
+
int
intel_opregion_get_panel_type(struct drm_i915_private *dev_priv)
{
return -ENODEV;
}
+ /*
+ * So far we know that some machined must use it, others must not use it.
+ * There doesn't seem to be any way to determine which way to go, except
+ * via a quirk list :(
+ */
+ if (!dmi_check_system(intel_use_opregion_panel_type)) {
+ DRM_DEBUG_KMS("Ignoring OpRegion panel type (%d)\n", ret - 1);
+ return -ENODEV;
+ }
+
/*
* FIXME On Dell XPS 13 9350 the OpRegion panel type (0) gives us
* low vswing for eDP, whereas the VBT panel type (2) gives us normal
struct intel_overlay {
struct drm_i915_private *i915;
struct intel_crtc *crtc;
- struct drm_i915_gem_object *vid_bo;
- struct drm_i915_gem_object *old_vid_bo;
+ struct i915_vma *vma;
+ struct i915_vma *old_vma;
bool active;
bool pfit_active;
u32 pfit_vscale_ratio; /* shifted-point number, (1<<12) == 1.0 */
if (OVERLAY_NEEDS_PHYSICAL(dev_priv))
regs = (struct overlay_registers __iomem *)overlay->reg_bo->phys_handle->vaddr;
else
- regs = io_mapping_map_wc(dev_priv->ggtt.mappable,
+ regs = io_mapping_map_wc(&dev_priv->ggtt.mappable,
overlay->flip_addr,
PAGE_SIZE);
{
struct intel_overlay *overlay =
container_of(active, typeof(*overlay), last_flip);
- struct drm_i915_gem_object *obj = overlay->old_vid_bo;
+ struct i915_vma *vma;
- i915_gem_track_fb(obj, NULL,
- INTEL_FRONTBUFFER_OVERLAY(overlay->crtc->pipe));
+ vma = fetch_and_zero(&overlay->old_vma);
+ if (WARN_ON(!vma))
+ return;
- i915_gem_object_ggtt_unpin(obj);
- i915_gem_object_put(obj);
+ i915_gem_track_fb(vma->obj, NULL,
+ INTEL_FRONTBUFFER_OVERLAY(overlay->crtc->pipe));
- overlay->old_vid_bo = NULL;
+ i915_gem_object_unpin_from_display_plane(vma);
+ i915_vma_put(vma);
}
static void intel_overlay_off_tail(struct i915_gem_active *active,
{
struct intel_overlay *overlay =
container_of(active, typeof(*overlay), last_flip);
- struct drm_i915_gem_object *obj = overlay->vid_bo;
+ struct i915_vma *vma;
/* never have the overlay hw on without showing a frame */
- if (WARN_ON(!obj))
+ vma = fetch_and_zero(&overlay->vma);
+ if (WARN_ON(!vma))
return;
- i915_gem_object_ggtt_unpin(obj);
- i915_gem_object_put(obj);
- overlay->vid_bo = NULL;
+ i915_gem_object_unpin_from_display_plane(vma);
+ i915_vma_put(vma);
overlay->crtc->overlay = NULL;
overlay->crtc = NULL;
/* Only wait if there is actually an old frame to release to
* guarantee forward progress.
*/
- if (!overlay->old_vid_bo)
+ if (!overlay->old_vma)
return 0;
if (I915_READ(ISR) & I915_OVERLAY_PLANE_FLIP_PENDING_INTERRUPT) {
struct drm_i915_private *dev_priv = overlay->i915;
u32 swidth, swidthsw, sheight, ostride;
enum pipe pipe = overlay->crtc->pipe;
+ struct i915_vma *vma;
lockdep_assert_held(&dev_priv->drm.struct_mutex);
WARN_ON(!drm_modeset_is_locked(&dev_priv->drm.mode_config.connection_mutex));
if (ret != 0)
return ret;
- ret = i915_gem_object_pin_to_display_plane(new_bo, 0,
+ vma = i915_gem_object_pin_to_display_plane(new_bo, 0,
&i915_ggtt_view_normal);
- if (ret != 0)
- return ret;
+ if (IS_ERR(vma))
+ return PTR_ERR(vma);
- ret = i915_gem_object_put_fence(new_bo);
+ ret = i915_vma_put_fence(vma);
if (ret)
goto out_unpin;
swidth = params->src_w;
swidthsw = calc_swidthsw(dev_priv, params->offset_Y, tmp_width);
sheight = params->src_h;
- iowrite32(i915_gem_obj_ggtt_offset(new_bo) + params->offset_Y, ®s->OBUF_0Y);
+ iowrite32(i915_ggtt_offset(vma) + params->offset_Y, ®s->OBUF_0Y);
ostride = params->stride_Y;
if (params->format & I915_OVERLAY_YUV_PLANAR) {
params->src_w/uv_hscale);
swidthsw |= max_t(u32, tmp_U, tmp_V) << 16;
sheight |= (params->src_h/uv_vscale) << 16;
- iowrite32(i915_gem_obj_ggtt_offset(new_bo) + params->offset_U, ®s->OBUF_0U);
- iowrite32(i915_gem_obj_ggtt_offset(new_bo) + params->offset_V, ®s->OBUF_0V);
+ iowrite32(i915_ggtt_offset(vma) + params->offset_U,
+ ®s->OBUF_0U);
+ iowrite32(i915_ggtt_offset(vma) + params->offset_V,
+ ®s->OBUF_0V);
ostride |= params->stride_UV << 16;
}
if (ret)
goto out_unpin;
- i915_gem_track_fb(overlay->vid_bo, new_bo,
+ i915_gem_track_fb(overlay->vma->obj, new_bo,
INTEL_FRONTBUFFER_OVERLAY(pipe));
- overlay->old_vid_bo = overlay->vid_bo;
- overlay->vid_bo = new_bo;
+ overlay->old_vma = overlay->vma;
+ overlay->vma = vma;
intel_frontbuffer_flip(dev_priv, INTEL_FRONTBUFFER_OVERLAY(pipe));
return 0;
out_unpin:
- i915_gem_object_ggtt_unpin(new_bo);
+ i915_gem_object_unpin_from_display_plane(vma);
return ret;
}
struct intel_overlay *overlay;
struct drm_i915_gem_object *reg_bo;
struct overlay_registers __iomem *regs;
+ struct i915_vma *vma = NULL;
int ret;
if (!HAS_OVERLAY(dev_priv))
}
overlay->flip_addr = reg_bo->phys_handle->busaddr;
} else {
- ret = i915_gem_object_ggtt_pin(reg_bo, NULL,
+ vma = i915_gem_object_ggtt_pin(reg_bo, NULL,
0, PAGE_SIZE, PIN_MAPPABLE);
- if (ret) {
+ if (IS_ERR(vma)) {
DRM_ERROR("failed to pin overlay register bo\n");
+ ret = PTR_ERR(vma);
goto out_free_bo;
}
- overlay->flip_addr = i915_gem_obj_ggtt_offset(reg_bo);
+ overlay->flip_addr = i915_ggtt_offset(vma);
ret = i915_gem_object_set_to_gtt_domain(reg_bo, true);
if (ret) {
return;
out_unpin_bo:
- if (!OVERLAY_NEEDS_PHYSICAL(dev_priv))
- i915_gem_object_ggtt_unpin(reg_bo);
+ if (vma)
+ i915_vma_unpin(vma);
out_free_bo:
i915_gem_object_put(reg_bo);
out_free:
regs = (struct overlay_registers __iomem *)
overlay->reg_bo->phys_handle->vaddr;
else
- regs = io_mapping_map_atomic_wc(dev_priv->ggtt.mappable,
+ regs = io_mapping_map_atomic_wc(&dev_priv->ggtt.mappable,
overlay->flip_addr);
return regs;
panel->backlight.min = get_backlight_min_vbt(connector);
val = lpt_get_backlight(connector);
- panel->backlight.level = intel_panel_compute_brightness(connector, val);
+ val = intel_panel_compute_brightness(connector, val);
+ panel->backlight.level = clamp(val, panel->backlight.min,
+ panel->backlight.max);
- panel->backlight.enabled = (pch_ctl1 & BLM_PCH_PWM_ENABLE) &&
- panel->backlight.level != 0;
+ panel->backlight.enabled = pch_ctl1 & BLM_PCH_PWM_ENABLE;
return 0;
}
panel->backlight.min = get_backlight_min_vbt(connector);
val = pch_get_backlight(connector);
- panel->backlight.level = intel_panel_compute_brightness(connector, val);
+ val = intel_panel_compute_brightness(connector, val);
+ panel->backlight.level = clamp(val, panel->backlight.min,
+ panel->backlight.max);
cpu_ctl2 = I915_READ(BLC_PWM_CPU_CTL2);
panel->backlight.enabled = (cpu_ctl2 & BLM_PWM_ENABLE) &&
- (pch_ctl1 & BLM_PCH_PWM_ENABLE) && panel->backlight.level != 0;
+ (pch_ctl1 & BLM_PCH_PWM_ENABLE);
return 0;
}
panel->backlight.min = get_backlight_min_vbt(connector);
val = i9xx_get_backlight(connector);
- panel->backlight.level = intel_panel_compute_brightness(connector, val);
+ val = intel_panel_compute_brightness(connector, val);
+ panel->backlight.level = clamp(val, panel->backlight.min,
+ panel->backlight.max);
- panel->backlight.enabled = panel->backlight.level != 0;
+ panel->backlight.enabled = val != 0;
return 0;
}
panel->backlight.min = get_backlight_min_vbt(connector);
val = i9xx_get_backlight(connector);
- panel->backlight.level = intel_panel_compute_brightness(connector, val);
+ val = intel_panel_compute_brightness(connector, val);
+ panel->backlight.level = clamp(val, panel->backlight.min,
+ panel->backlight.max);
- panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) &&
- panel->backlight.level != 0;
+ panel->backlight.enabled = ctl2 & BLM_PWM_ENABLE;
return 0;
}
panel->backlight.min = get_backlight_min_vbt(connector);
val = _vlv_get_backlight(dev_priv, pipe);
- panel->backlight.level = intel_panel_compute_brightness(connector, val);
+ val = intel_panel_compute_brightness(connector, val);
+ panel->backlight.level = clamp(val, panel->backlight.min,
+ panel->backlight.max);
- panel->backlight.enabled = (ctl2 & BLM_PWM_ENABLE) &&
- panel->backlight.level != 0;
+ panel->backlight.enabled = ctl2 & BLM_PWM_ENABLE;
return 0;
}
return -ENODEV;
val = bxt_get_backlight(connector);
- panel->backlight.level = intel_panel_compute_brightness(connector, val);
+ val = intel_panel_compute_brightness(connector, val);
+ panel->backlight.level = clamp(val, panel->backlight.min,
+ panel->backlight.max);
- panel->backlight.enabled = (pwm_ctl & BXT_BLC_PWM_ENABLE) &&
- panel->backlight.level != 0;
+ panel->backlight.enabled = pwm_ctl & BXT_BLC_PWM_ENABLE;
return 0;
}
return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL);
}
-/*
- * On gen9, we need to allocate Display Data Buffer (DDB) portions to the
- * different active planes.
- */
-
-#define SKL_DDB_SIZE 896 /* in blocks */
-#define BXT_DDB_SIZE 512
+#define SKL_SAGV_BLOCK_TIME 30 /* µs */
/*
* Return the index of a plane in the SKL DDB and wm result arrays. Primary
}
}
+/*
+ * SAGV dynamically adjusts the system agent voltage and clock frequencies
+ * depending on power and performance requirements. The display engine access
+ * to system memory is blocked during the adjustment time. Because of the
+ * blocking time, having this enabled can cause full system hangs and/or pipe
+ * underruns if we don't meet all of the following requirements:
+ *
+ * - <= 1 pipe enabled
+ * - All planes can enable watermarks for latencies >= SAGV engine block time
+ * - We're not using an interlaced display configuration
+ */
+int
+skl_enable_sagv(struct drm_i915_private *dev_priv)
+{
+ int ret;
+
+ if (dev_priv->skl_sagv_status == I915_SKL_SAGV_NOT_CONTROLLED ||
+ dev_priv->skl_sagv_status == I915_SKL_SAGV_ENABLED)
+ return 0;
+
+ DRM_DEBUG_KMS("Enabling the SAGV\n");
+ mutex_lock(&dev_priv->rps.hw_lock);
+
+ ret = sandybridge_pcode_write(dev_priv, GEN9_PCODE_SAGV_CONTROL,
+ GEN9_SAGV_ENABLE);
+
+ /* We don't need to wait for the SAGV when enabling */
+ mutex_unlock(&dev_priv->rps.hw_lock);
+
+ /*
+ * Some skl systems, pre-release machines in particular,
+ * don't actually have an SAGV.
+ */
+ if (ret == -ENXIO) {
+ DRM_DEBUG_DRIVER("No SAGV found on system, ignoring\n");
+ dev_priv->skl_sagv_status = I915_SKL_SAGV_NOT_CONTROLLED;
+ return 0;
+ } else if (ret < 0) {
+ DRM_ERROR("Failed to enable the SAGV\n");
+ return ret;
+ }
+
+ dev_priv->skl_sagv_status = I915_SKL_SAGV_ENABLED;
+ return 0;
+}
+
+static int
+skl_do_sagv_disable(struct drm_i915_private *dev_priv)
+{
+ int ret;
+ uint32_t temp = GEN9_SAGV_DISABLE;
+
+ ret = sandybridge_pcode_read(dev_priv, GEN9_PCODE_SAGV_CONTROL,
+ &temp);
+ if (ret)
+ return ret;
+ else
+ return temp & GEN9_SAGV_IS_DISABLED;
+}
+
+int
+skl_disable_sagv(struct drm_i915_private *dev_priv)
+{
+ int ret, result;
+
+ if (dev_priv->skl_sagv_status == I915_SKL_SAGV_NOT_CONTROLLED ||
+ dev_priv->skl_sagv_status == I915_SKL_SAGV_DISABLED)
+ return 0;
+
+ DRM_DEBUG_KMS("Disabling the SAGV\n");
+ mutex_lock(&dev_priv->rps.hw_lock);
+
+ /* bspec says to keep retrying for at least 1 ms */
+ ret = wait_for(result = skl_do_sagv_disable(dev_priv), 1);
+ mutex_unlock(&dev_priv->rps.hw_lock);
+
+ if (ret == -ETIMEDOUT) {
+ DRM_ERROR("Request to disable SAGV timed out\n");
+ return -ETIMEDOUT;
+ }
+
+ /*
+ * Some skl systems, pre-release machines in particular,
+ * don't actually have an SAGV.
+ */
+ if (result == -ENXIO) {
+ DRM_DEBUG_DRIVER("No SAGV found on system, ignoring\n");
+ dev_priv->skl_sagv_status = I915_SKL_SAGV_NOT_CONTROLLED;
+ return 0;
+ } else if (result < 0) {
+ DRM_ERROR("Failed to disable the SAGV\n");
+ return result;
+ }
+
+ dev_priv->skl_sagv_status = I915_SKL_SAGV_DISABLED;
+ return 0;
+}
+
+bool skl_can_enable_sagv(struct drm_atomic_state *state)
+{
+ struct drm_device *dev = state->dev;
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
+ struct drm_crtc *crtc;
+ enum pipe pipe;
+ int level, plane;
+
+ /*
+ * SKL workaround: bspec recommends we disable the SAGV when we have
+ * more then one pipe enabled
+ *
+ * If there are no active CRTCs, no additional checks need be performed
+ */
+ if (hweight32(intel_state->active_crtcs) == 0)
+ return true;
+ else if (hweight32(intel_state->active_crtcs) > 1)
+ return false;
+
+ /* Since we're now guaranteed to only have one active CRTC... */
+ pipe = ffs(intel_state->active_crtcs) - 1;
+ crtc = dev_priv->pipe_to_crtc_mapping[pipe];
+
+ if (crtc->state->mode.flags & DRM_MODE_FLAG_INTERLACE)
+ return false;
+
+ for_each_plane(dev_priv, pipe, plane) {
+ /* Skip this plane if it's not enabled */
+ if (intel_state->wm_results.plane[pipe][plane][0] == 0)
+ continue;
+
+ /* Find the highest enabled wm level for this plane */
+ for (level = ilk_wm_max_level(dev);
+ intel_state->wm_results.plane[pipe][plane][level] == 0; --level)
+ { }
+
+ /*
+ * If any of the planes on this pipe don't enable wm levels
+ * that incur memory latencies higher then 30µs we can't enable
+ * the SAGV
+ */
+ if (dev_priv->wm.skl_latency[level] < SKL_SAGV_BLOCK_TIME)
+ return false;
+ }
+
+ return true;
+}
+
static void
skl_ddb_get_pipe_allocation_limits(struct drm_device *dev,
const struct intel_crtc_state *cstate,
else
*num_active = hweight32(dev_priv->active_crtcs);
- if (IS_BROXTON(dev))
- ddb_size = BXT_DDB_SIZE;
- else
- ddb_size = SKL_DDB_SIZE;
+ ddb_size = INTEL_INFO(dev_priv)->ddb_size;
+ WARN_ON(ddb_size == 0);
ddb_size -= 4; /* 4 blocks for bypass path allocation */
total_data_rate += intel_cstate->wm.skl.plane_y_data_rate[id];
}
- WARN_ON(cstate->plane_mask && total_data_rate == 0);
-
return total_data_rate;
}
I915_WRITE(reg, 0);
}
-static void skl_write_wm_values(struct drm_i915_private *dev_priv,
- const struct skl_wm_values *new)
+void skl_write_plane_wm(struct intel_crtc *intel_crtc,
+ const struct skl_wm_values *wm,
+ int plane)
{
- struct drm_device *dev = &dev_priv->drm;
- struct intel_crtc *crtc;
-
- for_each_intel_crtc(dev, crtc) {
- int i, level, max_level = ilk_wm_max_level(dev);
- enum pipe pipe = crtc->pipe;
-
- if ((new->dirty_pipes & drm_crtc_mask(&crtc->base)) == 0)
- continue;
- if (!crtc->active)
- continue;
-
- I915_WRITE(PIPE_WM_LINETIME(pipe), new->wm_linetime[pipe]);
-
- for (level = 0; level <= max_level; level++) {
- for (i = 0; i < intel_num_planes(crtc); i++)
- I915_WRITE(PLANE_WM(pipe, i, level),
- new->plane[pipe][i][level]);
- I915_WRITE(CUR_WM(pipe, level),
- new->plane[pipe][PLANE_CURSOR][level]);
- }
- for (i = 0; i < intel_num_planes(crtc); i++)
- I915_WRITE(PLANE_WM_TRANS(pipe, i),
- new->plane_trans[pipe][i]);
- I915_WRITE(CUR_WM_TRANS(pipe),
- new->plane_trans[pipe][PLANE_CURSOR]);
-
- for (i = 0; i < intel_num_planes(crtc); i++) {
- skl_ddb_entry_write(dev_priv,
- PLANE_BUF_CFG(pipe, i),
- &new->ddb.plane[pipe][i]);
- skl_ddb_entry_write(dev_priv,
- PLANE_NV12_BUF_CFG(pipe, i),
- &new->ddb.y_plane[pipe][i]);
- }
+ struct drm_crtc *crtc = &intel_crtc->base;
+ struct drm_device *dev = crtc->dev;
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ int level, max_level = ilk_wm_max_level(dev);
+ enum pipe pipe = intel_crtc->pipe;
- skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe),
- &new->ddb.plane[pipe][PLANE_CURSOR]);
+ for (level = 0; level <= max_level; level++) {
+ I915_WRITE(PLANE_WM(pipe, plane, level),
+ wm->plane[pipe][plane][level]);
}
-}
+ I915_WRITE(PLANE_WM_TRANS(pipe, plane), wm->plane_trans[pipe][plane]);
-/*
- * When setting up a new DDB allocation arrangement, we need to correctly
- * sequence the times at which the new allocations for the pipes are taken into
- * account or we'll have pipes fetching from space previously allocated to
- * another pipe.
- *
- * Roughly the sequence looks like:
- * 1. re-allocate the pipe(s) with the allocation being reduced and not
- * overlapping with a previous light-up pipe (another way to put it is:
- * pipes with their new allocation strickly included into their old ones).
- * 2. re-allocate the other pipes that get their allocation reduced
- * 3. allocate the pipes having their allocation increased
- *
- * Steps 1. and 2. are here to take care of the following case:
- * - Initially DDB looks like this:
- * | B | C |
- * - enable pipe A.
- * - pipe B has a reduced DDB allocation that overlaps with the old pipe C
- * allocation
- * | A | B | C |
- *
- * We need to sequence the re-allocation: C, B, A (and not B, C, A).
- */
+ skl_ddb_entry_write(dev_priv, PLANE_BUF_CFG(pipe, plane),
+ &wm->ddb.plane[pipe][plane]);
+ skl_ddb_entry_write(dev_priv, PLANE_NV12_BUF_CFG(pipe, plane),
+ &wm->ddb.y_plane[pipe][plane]);
+}
-static void
-skl_wm_flush_pipe(struct drm_i915_private *dev_priv, enum pipe pipe, int pass)
+void skl_write_cursor_wm(struct intel_crtc *intel_crtc,
+ const struct skl_wm_values *wm)
{
- int plane;
-
- DRM_DEBUG_KMS("flush pipe %c (pass %d)\n", pipe_name(pipe), pass);
+ struct drm_crtc *crtc = &intel_crtc->base;
+ struct drm_device *dev = crtc->dev;
+ struct drm_i915_private *dev_priv = to_i915(dev);
+ int level, max_level = ilk_wm_max_level(dev);
+ enum pipe pipe = intel_crtc->pipe;
- for_each_plane(dev_priv, pipe, plane) {
- I915_WRITE(PLANE_SURF(pipe, plane),
- I915_READ(PLANE_SURF(pipe, plane)));
+ for (level = 0; level <= max_level; level++) {
+ I915_WRITE(CUR_WM(pipe, level),
+ wm->plane[pipe][PLANE_CURSOR][level]);
}
- I915_WRITE(CURBASE(pipe), I915_READ(CURBASE(pipe)));
+ I915_WRITE(CUR_WM_TRANS(pipe), wm->plane_trans[pipe][PLANE_CURSOR]);
+
+ skl_ddb_entry_write(dev_priv, CUR_BUF_CFG(pipe),
+ &wm->ddb.plane[pipe][PLANE_CURSOR]);
}
-static bool
-skl_ddb_allocation_included(const struct skl_ddb_allocation *old,
- const struct skl_ddb_allocation *new,
- enum pipe pipe)
+bool skl_ddb_allocation_equals(const struct skl_ddb_allocation *old,
+ const struct skl_ddb_allocation *new,
+ enum pipe pipe)
{
- uint16_t old_size, new_size;
-
- old_size = skl_ddb_entry_size(&old->pipe[pipe]);
- new_size = skl_ddb_entry_size(&new->pipe[pipe]);
-
- return old_size != new_size &&
- new->pipe[pipe].start >= old->pipe[pipe].start &&
- new->pipe[pipe].end <= old->pipe[pipe].end;
+ return new->pipe[pipe].start == old->pipe[pipe].start &&
+ new->pipe[pipe].end == old->pipe[pipe].end;
}
-static void skl_flush_wm_values(struct drm_i915_private *dev_priv,
- struct skl_wm_values *new_values)
+static inline bool skl_ddb_entries_overlap(const struct skl_ddb_entry *a,
+ const struct skl_ddb_entry *b)
{
- struct drm_device *dev = &dev_priv->drm;
- struct skl_ddb_allocation *cur_ddb, *new_ddb;
- bool reallocated[I915_MAX_PIPES] = {};
- struct intel_crtc *crtc;
- enum pipe pipe;
-
- new_ddb = &new_values->ddb;
- cur_ddb = &dev_priv->wm.skl_hw.ddb;
-
- /*
- * First pass: flush the pipes with the new allocation contained into
- * the old space.
- *
- * We'll wait for the vblank on those pipes to ensure we can safely
- * re-allocate the freed space without this pipe fetching from it.
- */
- for_each_intel_crtc(dev, crtc) {
- if (!crtc->active)
- continue;
-
- pipe = crtc->pipe;
-
- if (!skl_ddb_allocation_included(cur_ddb, new_ddb, pipe))
- continue;
-
- skl_wm_flush_pipe(dev_priv, pipe, 1);
- intel_wait_for_vblank(dev, pipe);
-
- reallocated[pipe] = true;
- }
-
+ return a->start < b->end && b->start < a->end;
+}
- /*
- * Second pass: flush the pipes that are having their allocation
- * reduced, but overlapping with a previous allocation.
- *
- * Here as well we need to wait for the vblank to make sure the freed
- * space is not used anymore.
- */
- for_each_intel_crtc(dev, crtc) {
- if (!crtc->active)
- continue;
+bool skl_ddb_allocation_overlaps(struct drm_atomic_state *state,
+ const struct skl_ddb_allocation *old,
+ const struct skl_ddb_allocation *new,
+ enum pipe pipe)
+{
+ struct drm_device *dev = state->dev;
+ struct intel_crtc *intel_crtc;
+ enum pipe otherp;
- pipe = crtc->pipe;
+ for_each_intel_crtc(dev, intel_crtc) {
+ otherp = intel_crtc->pipe;
- if (reallocated[pipe])
+ if (otherp == pipe)
continue;
- if (skl_ddb_entry_size(&new_ddb->pipe[pipe]) <
- skl_ddb_entry_size(&cur_ddb->pipe[pipe])) {
- skl_wm_flush_pipe(dev_priv, pipe, 2);
- intel_wait_for_vblank(dev, pipe);
- reallocated[pipe] = true;
- }
+ if (skl_ddb_entries_overlap(&new->pipe[pipe],
+ &old->pipe[otherp]))
+ return true;
}
- /*
- * Third pass: flush the pipes that got more space allocated.
- *
- * We don't need to actively wait for the update here, next vblank
- * will just get more DDB space with the correct WM values.
- */
- for_each_intel_crtc(dev, crtc) {
- if (!crtc->active)
- continue;
-
- pipe = crtc->pipe;
-
- /*
- * At this point, only the pipes more space than before are
- * left to re-allocate.
- */
- if (reallocated[pipe])
- continue;
-
- skl_wm_flush_pipe(dev_priv, pipe, 3);
- }
+ return false;
}
static int skl_update_pipe_wm(struct drm_crtc_state *cstate,
* pretend that all pipes switched active status so that we'll
* ensure a full DDB recompute.
*/
- if (dev_priv->wm.distrust_bios_wm)
+ if (dev_priv->wm.distrust_bios_wm) {
+ ret = drm_modeset_lock(&dev->mode_config.connection_mutex,
+ state->acquire_ctx);
+ if (ret)
+ return ret;
+
intel_state->active_pipe_changes = ~0;
+ /*
+ * We usually only initialize intel_state->active_crtcs if we
+ * we're doing a modeset; make sure this field is always
+ * initialized during the sanitization process that happens
+ * on the first commit too.
+ */
+ if (!intel_state->modeset)
+ intel_state->active_crtcs = dev_priv->active_crtcs;
+ }
+
/*
* If the modeset changes which CRTC's are active, we need to
* recompute the DDB allocation for *all* active pipes, even
ret = skl_allocate_pipe_ddb(cstate, ddb);
if (ret)
return ret;
+
+ ret = drm_atomic_add_affected_planes(state, &intel_crtc->base);
+ if (ret)
+ return ret;
}
return 0;
}
+static void
+skl_copy_wm_for_pipe(struct skl_wm_values *dst,
+ struct skl_wm_values *src,
+ enum pipe pipe)
+{
+ dst->wm_linetime[pipe] = src->wm_linetime[pipe];
+ memcpy(dst->plane[pipe], src->plane[pipe],
+ sizeof(dst->plane[pipe]));
+ memcpy(dst->plane_trans[pipe], src->plane_trans[pipe],
+ sizeof(dst->plane_trans[pipe]));
+
+ dst->ddb.pipe[pipe] = src->ddb.pipe[pipe];
+ memcpy(dst->ddb.y_plane[pipe], src->ddb.y_plane[pipe],
+ sizeof(dst->ddb.y_plane[pipe]));
+ memcpy(dst->ddb.plane[pipe], src->ddb.plane[pipe],
+ sizeof(dst->ddb.plane[pipe]));
+}
+
static int
skl_compute_wm(struct drm_atomic_state *state)
{
struct drm_device *dev = crtc->dev;
struct drm_i915_private *dev_priv = to_i915(dev);
struct skl_wm_values *results = &dev_priv->wm.skl_results;
+ struct skl_wm_values *hw_vals = &dev_priv->wm.skl_hw;
struct intel_crtc_state *cstate = to_intel_crtc_state(crtc->state);
struct skl_pipe_wm *pipe_wm = &cstate->wm.skl.optimal;
+ enum pipe pipe = intel_crtc->pipe;
if ((results->dirty_pipes & drm_crtc_mask(crtc)) == 0)
return;
mutex_lock(&dev_priv->wm.wm_mutex);
- skl_write_wm_values(dev_priv, results);
- skl_flush_wm_values(dev_priv, results);
+ /*
+ * If this pipe isn't active already, we're going to be enabling it
+ * very soon. Since it's safe to update a pipe's ddb allocation while
+ * the pipe's shut off, just do so here. Already active pipes will have
+ * their watermarks updated once we update their planes.
+ */
+ if (crtc->state->active_changed) {
+ int plane;
- /* store the new configuration */
- dev_priv->wm.skl_hw = *results;
+ for (plane = 0; plane < intel_num_planes(intel_crtc); plane++)
+ skl_write_plane_wm(intel_crtc, results, plane);
+
+ skl_write_cursor_wm(intel_crtc, results);
+ }
+
+ skl_copy_wm_for_pipe(hw_vals, results, pipe);
mutex_unlock(&dev_priv->wm.wm_mutex);
}
val = vlv_punit_read(dev_priv, FB_GFX_FMAX_AT_VMAX_FUSE);
- switch (INTEL_INFO(dev_priv)->eu_total) {
+ switch (INTEL_INFO(dev_priv)->sseu.eu_total) {
case 8:
/* (2 * 4) config */
rp0 = (val >> FB_GFX_FMAX_AT_VMAX_2SS4EU_FUSE_SHIFT);
u32 pcbr;
int pctx_size = 24*1024;
- mutex_lock(&dev_priv->drm.struct_mutex);
-
pcbr = I915_READ(VLV_PCBR);
if (pcbr) {
/* BIOS set it up already, grab the pre-alloc'd space */
out:
DRM_DEBUG_DRIVER("PCBR: 0x%08x\n", I915_READ(VLV_PCBR));
dev_priv->vlv_pctx = pctx;
- mutex_unlock(&dev_priv->drm.struct_mutex);
}
static void valleyview_cleanup_pctx(struct drm_i915_private *dev_priv)
intel_runtime_pm_get(dev_priv);
}
+ mutex_lock(&dev_priv->drm.struct_mutex);
mutex_lock(&dev_priv->rps.hw_lock);
/* Initialize RPS limits (for userspace) */
dev_priv->rps.boost_freq = dev_priv->rps.max_freq;
mutex_unlock(&dev_priv->rps.hw_lock);
+ mutex_unlock(&dev_priv->drm.struct_mutex);
intel_autoenable_gt_powersave(dev_priv);
}
if (IS_IRONLAKE_M(dev_priv)) {
ironlake_enable_drps(dev_priv);
- mutex_lock(&dev_priv->drm.struct_mutex);
intel_init_emon(dev_priv);
- mutex_unlock(&dev_priv->drm.struct_mutex);
} else if (INTEL_INFO(dev_priv)->gen >= 6) {
/*
* PCU communication is slow and this doesn't need to be
}
}
+static inline int gen6_check_mailbox_status(struct drm_i915_private *dev_priv)
+{
+ uint32_t flags =
+ I915_READ_FW(GEN6_PCODE_MAILBOX) & GEN6_PCODE_ERROR_MASK;
+
+ switch (flags) {
+ case GEN6_PCODE_SUCCESS:
+ return 0;
+ case GEN6_PCODE_UNIMPLEMENTED_CMD:
+ case GEN6_PCODE_ILLEGAL_CMD:
+ return -ENXIO;
+ case GEN6_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE:
+ case GEN7_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE:
+ return -EOVERFLOW;
+ case GEN6_PCODE_TIMEOUT:
+ return -ETIMEDOUT;
+ default:
+ MISSING_CASE(flags)
+ return 0;
+ }
+}
+
+static inline int gen7_check_mailbox_status(struct drm_i915_private *dev_priv)
+{
+ uint32_t flags =
+ I915_READ_FW(GEN6_PCODE_MAILBOX) & GEN6_PCODE_ERROR_MASK;
+
+ switch (flags) {
+ case GEN6_PCODE_SUCCESS:
+ return 0;
+ case GEN6_PCODE_ILLEGAL_CMD:
+ return -ENXIO;
+ case GEN7_PCODE_TIMEOUT:
+ return -ETIMEDOUT;
+ case GEN7_PCODE_ILLEGAL_DATA:
+ return -EINVAL;
+ case GEN7_PCODE_MIN_FREQ_TABLE_GT_RATIO_OUT_OF_RANGE:
+ return -EOVERFLOW;
+ default:
+ MISSING_CASE(flags);
+ return 0;
+ }
+}
+
int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u32 mbox, u32 *val)
{
+ int status;
+
WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
/* GEN6_PCODE_* are outside of the forcewake domain, we can
*val = I915_READ_FW(GEN6_PCODE_DATA);
I915_WRITE_FW(GEN6_PCODE_DATA, 0);
+ if (INTEL_GEN(dev_priv) > 6)
+ status = gen7_check_mailbox_status(dev_priv);
+ else
+ status = gen6_check_mailbox_status(dev_priv);
+
+ if (status) {
+ DRM_DEBUG_DRIVER("warning: pcode (read) mailbox access failed: %d\n",
+ status);
+ return status;
+ }
+
return 0;
}
int sandybridge_pcode_write(struct drm_i915_private *dev_priv,
- u32 mbox, u32 val)
+ u32 mbox, u32 val)
{
+ int status;
+
WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
/* GEN6_PCODE_* are outside of the forcewake domain, we can
I915_WRITE_FW(GEN6_PCODE_DATA, 0);
+ if (INTEL_GEN(dev_priv) > 6)
+ status = gen7_check_mailbox_status(dev_priv);
+ else
+ status = gen6_check_mailbox_status(dev_priv);
+
+ if (status) {
+ DRM_DEBUG_DRIVER("warning: pcode (write) mailbox access failed: %d\n",
+ status);
+ return status;
+ }
+
return 0;
}
struct drm_i915_private *dev_priv = to_i915(dev);
uint32_t max_sleep_time = 0x1f;
- /* Lately it was identified that depending on panel idle frame count
- * calculated at HW can be off by 1. So let's use what came
- * from VBT + 1.
- * There are also other cases where panel demands at least 4
- * but VBT is not being set. To cover these 2 cases lets use
- * at least 5 when VBT isn't set to be on the safest side.
+ /*
+ * Let's respect VBT in case VBT asks a higher idle_frame value.
+ * Let's use 6 as the minimum to cover all known cases including
+ * the off-by-one issue that HW has in some cases. Also there are
+ * cases where sink should be able to train
+ * with the 5 or 6 idle patterns.
*/
- uint32_t idle_frames = dev_priv->vbt.psr.idle_frames + 1;
+ uint32_t idle_frames = max(6, dev_priv->vbt.psr.idle_frames);
uint32_t val = EDP_PSR_ENABLE;
val |= max_sleep_time << EDP_PSR_MAX_SLEEP_TIME_SHIFT;
{
struct intel_ring *ring = req->ring;
u32 scratch_addr =
- req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+ i915_ggtt_offset(req->engine->scratch) + 2 * CACHELINE_BYTES;
int ret;
ret = intel_ring_begin(req, 6);
{
struct intel_ring *ring = req->ring;
u32 scratch_addr =
- req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+ i915_ggtt_offset(req->engine->scratch) + 2 * CACHELINE_BYTES;
u32 flags = 0;
int ret;
{
struct intel_ring *ring = req->ring;
u32 scratch_addr =
- req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+ i915_ggtt_offset(req->engine->scratch) + 2 * CACHELINE_BYTES;
u32 flags = 0;
int ret;
static int
gen8_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
{
- u32 scratch_addr = req->engine->scratch.gtt_offset + 2 * CACHELINE_BYTES;
+ u32 scratch_addr =
+ i915_ggtt_offset(req->engine->scratch) + 2 * CACHELINE_BYTES;
u32 flags = 0;
int ret;
mmio = RING_HWS_PGA(engine->mmio_base);
}
- I915_WRITE(mmio, (u32)engine->status_page.gfx_addr);
+ I915_WRITE(mmio, engine->status_page.ggtt_offset);
POSTING_READ(mmio);
/*
{
struct drm_i915_private *dev_priv = engine->i915;
- if (!IS_GEN2(dev_priv)) {
+ if (INTEL_GEN(dev_priv) > 2) {
I915_WRITE_MODE(engine, _MASKED_BIT_ENABLE(STOP_RING));
if (intel_wait_for_register(dev_priv,
RING_MI_MODE(engine->mmio_base),
I915_WRITE_HEAD(engine, 0);
I915_WRITE_TAIL(engine, 0);
- if (!IS_GEN2(dev_priv)) {
+ if (INTEL_GEN(dev_priv) > 2) {
(void)I915_READ_CTL(engine);
I915_WRITE_MODE(engine, _MASKED_BIT_DISABLE(STOP_RING));
}
{
struct drm_i915_private *dev_priv = engine->i915;
struct intel_ring *ring = engine->buffer;
- struct drm_i915_gem_object *obj = ring->obj;
int ret = 0;
intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
}
}
- if (I915_NEED_GFX_HWS(dev_priv))
- intel_ring_setup_status_page(engine);
- else
+ if (HWS_NEEDS_PHYSICAL(dev_priv))
ring_setup_phys_status_page(engine);
+ else
+ intel_ring_setup_status_page(engine);
+
+ intel_engine_reset_irq(engine);
/* Enforce ordering by reading HEAD register back */
I915_READ_HEAD(engine);
* registers with the above sequence (the readback of the HEAD registers
* also enforces ordering), otherwise the hw might lose the new ring
* register values. */
- I915_WRITE_START(engine, i915_gem_obj_ggtt_offset(obj));
+ I915_WRITE_START(engine, i915_ggtt_offset(ring->vma));
/* WaClearRingBufHeadRegAtInit:ctg,elk */
if (I915_READ_HEAD(engine))
DRM_DEBUG("%s initialization failed [head=%08x], fudging\n",
engine->name, I915_READ_HEAD(engine));
- I915_WRITE_HEAD(engine, 0);
- (void)I915_READ_HEAD(engine);
+
+ intel_ring_update_space(ring);
+ I915_WRITE_HEAD(engine, ring->head);
+ I915_WRITE_TAIL(engine, ring->tail);
+ (void)I915_READ_TAIL(engine);
I915_WRITE_CTL(engine,
((ring->size - PAGE_SIZE) & RING_NR_PAGES)
| RING_VALID);
/* If the head is still not zero, the ring is dead */
- if (wait_for((I915_READ_CTL(engine) & RING_VALID) != 0 &&
- I915_READ_START(engine) == i915_gem_obj_ggtt_offset(obj) &&
- (I915_READ_HEAD(engine) & HEAD_ADDR) == 0, 50)) {
+ if (intel_wait_for_register_fw(dev_priv, RING_CTL(engine->mmio_base),
+ RING_VALID, RING_VALID,
+ 50)) {
DRM_ERROR("%s initialization failed "
- "ctl %08x (valid? %d) head %08x tail %08x start %08x [expected %08lx]\n",
+ "ctl %08x (valid? %d) head %08x [%08x] tail %08x [%08x] start %08x [expected %08x]\n",
engine->name,
I915_READ_CTL(engine),
I915_READ_CTL(engine) & RING_VALID,
- I915_READ_HEAD(engine), I915_READ_TAIL(engine),
+ I915_READ_HEAD(engine), ring->head,
+ I915_READ_TAIL(engine), ring->tail,
I915_READ_START(engine),
- (unsigned long)i915_gem_obj_ggtt_offset(obj));
+ i915_ggtt_offset(ring->vma));
ret = -EIO;
goto out;
}
- ring->last_retired_head = -1;
- ring->head = I915_READ_HEAD(engine);
- ring->tail = I915_READ_TAIL(engine) & TAIL_ADDR;
- intel_ring_update_space(ring);
-
intel_engine_init_hangcheck(engine);
out:
return ret;
}
-void intel_fini_pipe_control(struct intel_engine_cs *engine)
+static void reset_ring_common(struct intel_engine_cs *engine,
+ struct drm_i915_gem_request *request)
{
- if (engine->scratch.obj == NULL)
- return;
-
- i915_gem_object_ggtt_unpin(engine->scratch.obj);
- i915_gem_object_put(engine->scratch.obj);
- engine->scratch.obj = NULL;
-}
-
-int intel_init_pipe_control(struct intel_engine_cs *engine, int size)
-{
- struct drm_i915_gem_object *obj;
- int ret;
+ struct intel_ring *ring = request->ring;
- WARN_ON(engine->scratch.obj);
-
- obj = i915_gem_object_create_stolen(&engine->i915->drm, size);
- if (!obj)
- obj = i915_gem_object_create(&engine->i915->drm, size);
- if (IS_ERR(obj)) {
- DRM_ERROR("Failed to allocate scratch page\n");
- ret = PTR_ERR(obj);
- goto err;
- }
-
- ret = i915_gem_object_ggtt_pin(obj, NULL, 0, 4096, PIN_HIGH);
- if (ret)
- goto err_unref;
-
- engine->scratch.obj = obj;
- engine->scratch.gtt_offset = i915_gem_obj_ggtt_offset(obj);
- DRM_DEBUG_DRIVER("%s pipe control offset: 0x%08x\n",
- engine->name, engine->scratch.gtt_offset);
- return 0;
-
-err_unref:
- i915_gem_object_put(engine->scratch.obj);
-err:
- return ret;
+ ring->head = request->postfix;
+ ring->last_retired_head = -1;
}
static int intel_ring_workarounds_emit(struct drm_i915_gem_request *req)
* Only consider slices where one, and only one, subslice has 7
* EUs
*/
- if (!is_power_of_2(dev_priv->info.subslice_7eu[i]))
+ if (!is_power_of_2(INTEL_INFO(dev_priv)->sseu.subslice_7eu[i]))
continue;
/*
*
* -> 0 <= ss <= 3;
*/
- ss = ffs(dev_priv->info.subslice_7eu[i]) - 1;
+ ss = ffs(INTEL_INFO(dev_priv)->sseu.subslice_7eu[i]) - 1;
vals[i] = 3 - ss;
}
{
struct drm_i915_private *dev_priv = engine->i915;
- if (dev_priv->semaphore_obj) {
- i915_gem_object_ggtt_unpin(dev_priv->semaphore_obj);
- i915_gem_object_put(dev_priv->semaphore_obj);
- dev_priv->semaphore_obj = NULL;
- }
-
- intel_fini_pipe_control(engine);
+ i915_vma_unpin_and_release(&dev_priv->semaphore);
}
static int gen8_rcs_signal(struct drm_i915_gem_request *req)
enum intel_engine_id id;
int ret, num_rings;
- num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
+ num_rings = INTEL_INFO(dev_priv)->num_rings;
ret = intel_ring_begin(req, (num_rings-1) * 8);
if (ret)
return ret;
enum intel_engine_id id;
int ret, num_rings;
- num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
+ num_rings = INTEL_INFO(dev_priv)->num_rings;
ret = intel_ring_begin(req, (num_rings-1) * 6);
if (ret)
return ret;
{
struct intel_ring *ring = req->ring;
struct drm_i915_private *dev_priv = req->i915;
- struct intel_engine_cs *useless;
- enum intel_engine_id id;
+ struct intel_engine_cs *engine;
int ret, num_rings;
- num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask);
+ num_rings = INTEL_INFO(dev_priv)->num_rings;
ret = intel_ring_begin(req, round_up((num_rings-1) * 3, 2));
if (ret)
return ret;
- for_each_engine_id(useless, dev_priv, id) {
- i915_reg_t mbox_reg = req->engine->semaphore.mbox.signal[id];
+ for_each_engine(engine, dev_priv) {
+ i915_reg_t mbox_reg;
+
+ if (!(BIT(engine->hw_id) & GEN6_SEMAPHORES_MASK))
+ continue;
+ mbox_reg = req->engine->semaphore.mbox.signal[engine->hw_id];
if (i915_mmio_reg_valid(mbox_reg)) {
intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(1));
intel_ring_emit_reg(ring, mbox_reg);
u32 dw1 = MI_SEMAPHORE_MBOX |
MI_SEMAPHORE_COMPARE |
MI_SEMAPHORE_REGISTER;
- u32 wait_mbox = signal->engine->semaphore.mbox.wait[req->engine->id];
+ u32 wait_mbox = signal->engine->semaphore.mbox.wait[req->engine->hw_id];
int ret;
WARN_ON(wait_mbox == MI_SEMAPHORE_SYNC_INVALID);
unsigned int dispatch_flags)
{
struct intel_ring *ring = req->ring;
- u32 cs_offset = req->engine->scratch.gtt_offset;
+ u32 cs_offset = i915_ggtt_offset(req->engine->scratch);
int ret;
ret = intel_ring_begin(req, 6);
static void cleanup_status_page(struct intel_engine_cs *engine)
{
- struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
- obj = engine->status_page.obj;
- if (obj == NULL)
+ vma = fetch_and_zero(&engine->status_page.vma);
+ if (!vma)
return;
- kunmap(sg_page(obj->pages->sgl));
- i915_gem_object_ggtt_unpin(obj);
- i915_gem_object_put(obj);
- engine->status_page.obj = NULL;
+ i915_vma_unpin(vma);
+ i915_gem_object_unpin_map(vma->obj);
+ i915_vma_put(vma);
}
static int init_status_page(struct intel_engine_cs *engine)
{
- struct drm_i915_gem_object *obj = engine->status_page.obj;
-
- if (obj == NULL) {
- unsigned flags;
- int ret;
+ struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
+ unsigned int flags;
+ int ret;
- obj = i915_gem_object_create(&engine->i915->drm, 4096);
- if (IS_ERR(obj)) {
- DRM_ERROR("Failed to allocate status page\n");
- return PTR_ERR(obj);
- }
+ obj = i915_gem_object_create(&engine->i915->drm, 4096);
+ if (IS_ERR(obj)) {
+ DRM_ERROR("Failed to allocate status page\n");
+ return PTR_ERR(obj);
+ }
- ret = i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
- if (ret)
- goto err_unref;
-
- flags = 0;
- if (!HAS_LLC(engine->i915))
- /* On g33, we cannot place HWS above 256MiB, so
- * restrict its pinning to the low mappable arena.
- * Though this restriction is not documented for
- * gen4, gen5, or byt, they also behave similarly
- * and hang if the HWS is placed at the top of the
- * GTT. To generalise, it appears that all !llc
- * platforms have issues with us placing the HWS
- * above the mappable region (even though we never
- * actualy map it).
- */
- flags |= PIN_MAPPABLE;
- ret = i915_gem_object_ggtt_pin(obj, NULL, 0, 4096, flags);
- if (ret) {
-err_unref:
- i915_gem_object_put(obj);
- return ret;
- }
+ ret = i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
+ if (ret)
+ goto err;
- engine->status_page.obj = obj;
+ vma = i915_vma_create(obj, &engine->i915->ggtt.base, NULL);
+ if (IS_ERR(vma)) {
+ ret = PTR_ERR(vma);
+ goto err;
}
- engine->status_page.gfx_addr = i915_gem_obj_ggtt_offset(obj);
- engine->status_page.page_addr = kmap(sg_page(obj->pages->sgl));
- memset(engine->status_page.page_addr, 0, PAGE_SIZE);
+ flags = PIN_GLOBAL;
+ if (!HAS_LLC(engine->i915))
+ /* On g33, we cannot place HWS above 256MiB, so
+ * restrict its pinning to the low mappable arena.
+ * Though this restriction is not documented for
+ * gen4, gen5, or byt, they also behave similarly
+ * and hang if the HWS is placed at the top of the
+ * GTT. To generalise, it appears that all !llc
+ * platforms have issues with us placing the HWS
+ * above the mappable region (even though we never
+ * actualy map it).
+ */
+ flags |= PIN_MAPPABLE;
+ ret = i915_vma_pin(vma, 0, 4096, flags);
+ if (ret)
+ goto err;
- DRM_DEBUG_DRIVER("%s hws offset: 0x%08x\n",
- engine->name, engine->status_page.gfx_addr);
+ engine->status_page.vma = vma;
+ engine->status_page.ggtt_offset = i915_ggtt_offset(vma);
+ engine->status_page.page_addr =
+ i915_gem_object_pin_map(obj, I915_MAP_WB);
+ DRM_DEBUG_DRIVER("%s hws offset: 0x%08x\n",
+ engine->name, i915_ggtt_offset(vma));
return 0;
+
+err:
+ i915_gem_object_put(obj);
+ return ret;
}
static int init_phys_status_page(struct intel_engine_cs *engine)
{
struct drm_i915_private *dev_priv = engine->i915;
- if (!dev_priv->status_page_dmah) {
- dev_priv->status_page_dmah =
- drm_pci_alloc(&dev_priv->drm, PAGE_SIZE, PAGE_SIZE);
- if (!dev_priv->status_page_dmah)
- return -ENOMEM;
- }
+ dev_priv->status_page_dmah =
+ drm_pci_alloc(&dev_priv->drm, PAGE_SIZE, PAGE_SIZE);
+ if (!dev_priv->status_page_dmah)
+ return -ENOMEM;
engine->status_page.page_addr = dev_priv->status_page_dmah->vaddr;
memset(engine->status_page.page_addr, 0, PAGE_SIZE);
int intel_ring_pin(struct intel_ring *ring)
{
- struct drm_i915_private *dev_priv = ring->engine->i915;
- struct drm_i915_gem_object *obj = ring->obj;
/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
- unsigned flags = PIN_OFFSET_BIAS | 4096;
+ unsigned int flags = PIN_GLOBAL | PIN_OFFSET_BIAS | 4096;
+ enum i915_map_type map;
+ struct i915_vma *vma = ring->vma;
void *addr;
int ret;
- if (HAS_LLC(dev_priv) && !obj->stolen) {
- ret = i915_gem_object_ggtt_pin(obj, NULL, 0, PAGE_SIZE, flags);
- if (ret)
- return ret;
+ GEM_BUG_ON(ring->vaddr);
- ret = i915_gem_object_set_to_cpu_domain(obj, true);
- if (ret)
- goto err_unpin;
+ map = HAS_LLC(ring->engine->i915) ? I915_MAP_WB : I915_MAP_WC;
- addr = i915_gem_object_pin_map(obj);
- if (IS_ERR(addr)) {
- ret = PTR_ERR(addr);
- goto err_unpin;
- }
- } else {
- ret = i915_gem_object_ggtt_pin(obj, NULL, 0, PAGE_SIZE,
- flags | PIN_MAPPABLE);
- if (ret)
- return ret;
+ if (vma->obj->stolen)
+ flags |= PIN_MAPPABLE;
- ret = i915_gem_object_set_to_gtt_domain(obj, true);
- if (ret)
- goto err_unpin;
+ if (!(vma->flags & I915_VMA_GLOBAL_BIND)) {
+ if (flags & PIN_MAPPABLE || map == I915_MAP_WC)
+ ret = i915_gem_object_set_to_gtt_domain(vma->obj, true);
+ else
+ ret = i915_gem_object_set_to_cpu_domain(vma->obj, true);
+ if (unlikely(ret))
+ return ret;
+ }
- /* Access through the GTT requires the device to be awake. */
- assert_rpm_wakelock_held(dev_priv);
+ ret = i915_vma_pin(vma, 0, PAGE_SIZE, flags);
+ if (unlikely(ret))
+ return ret;
- addr = (void __force *)
- i915_vma_pin_iomap(i915_gem_obj_to_ggtt(obj));
- if (IS_ERR(addr)) {
- ret = PTR_ERR(addr);
- goto err_unpin;
- }
- }
+ if (i915_vma_is_map_and_fenceable(vma))
+ addr = (void __force *)i915_vma_pin_iomap(vma);
+ else
+ addr = i915_gem_object_pin_map(vma->obj, map);
+ if (IS_ERR(addr))
+ goto err;
ring->vaddr = addr;
- ring->vma = i915_gem_obj_to_ggtt(obj);
return 0;
-err_unpin:
- i915_gem_object_ggtt_unpin(obj);
- return ret;
+err:
+ i915_vma_unpin(vma);
+ return PTR_ERR(addr);
}
void intel_ring_unpin(struct intel_ring *ring)
GEM_BUG_ON(!ring->vma);
GEM_BUG_ON(!ring->vaddr);
- if (HAS_LLC(ring->engine->i915) && !ring->obj->stolen)
- i915_gem_object_unpin_map(ring->obj);
- else
+ if (i915_vma_is_map_and_fenceable(ring->vma))
i915_vma_unpin_iomap(ring->vma);
+ else
+ i915_gem_object_unpin_map(ring->vma->obj);
ring->vaddr = NULL;
- i915_gem_object_ggtt_unpin(ring->obj);
- ring->vma = NULL;
+ i915_vma_unpin(ring->vma);
}
-static void intel_destroy_ringbuffer_obj(struct intel_ring *ring)
-{
- i915_gem_object_put(ring->obj);
- ring->obj = NULL;
-}
-
-static int intel_alloc_ringbuffer_obj(struct drm_device *dev,
- struct intel_ring *ring)
+static struct i915_vma *
+intel_ring_create_vma(struct drm_i915_private *dev_priv, int size)
{
struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
- obj = NULL;
- if (!HAS_LLC(dev))
- obj = i915_gem_object_create_stolen(dev, ring->size);
- if (obj == NULL)
- obj = i915_gem_object_create(dev, ring->size);
+ obj = i915_gem_object_create_stolen(&dev_priv->drm, size);
+ if (!obj)
+ obj = i915_gem_object_create(&dev_priv->drm, size);
if (IS_ERR(obj))
- return PTR_ERR(obj);
+ return ERR_CAST(obj);
/* mark ring buffers as read-only from GPU side by default */
obj->gt_ro = 1;
- ring->obj = obj;
+ vma = i915_vma_create(obj, &dev_priv->ggtt.base, NULL);
+ if (IS_ERR(vma))
+ goto err;
- return 0;
+ return vma;
+
+err:
+ i915_gem_object_put(obj);
+ return vma;
}
struct intel_ring *
intel_engine_create_ring(struct intel_engine_cs *engine, int size)
{
struct intel_ring *ring;
- int ret;
+ struct i915_vma *vma;
GEM_BUG_ON(!is_power_of_2(size));
ring = kzalloc(sizeof(*ring), GFP_KERNEL);
- if (ring == NULL) {
- DRM_DEBUG_DRIVER("Failed to allocate ringbuffer %s\n",
- engine->name);
+ if (!ring)
return ERR_PTR(-ENOMEM);
- }
ring->engine = engine;
- list_add(&ring->link, &engine->buffers);
INIT_LIST_HEAD(&ring->request_list);
ring->last_retired_head = -1;
intel_ring_update_space(ring);
- ret = intel_alloc_ringbuffer_obj(&engine->i915->drm, ring);
- if (ret) {
- DRM_DEBUG_DRIVER("Failed to allocate ringbuffer %s: %d\n",
- engine->name, ret);
- list_del(&ring->link);
+ vma = intel_ring_create_vma(engine->i915, size);
+ if (IS_ERR(vma)) {
kfree(ring);
- return ERR_PTR(ret);
+ return ERR_CAST(vma);
}
+ ring->vma = vma;
return ring;
}
void
intel_ring_free(struct intel_ring *ring)
{
- intel_destroy_ringbuffer_obj(ring);
- list_del(&ring->link);
+ i915_vma_put(ring->vma);
kfree(ring);
}
return 0;
if (ce->state) {
- ret = i915_gem_object_ggtt_pin(ce->state, NULL, 0,
- ctx->ggtt_alignment, 0);
+ ret = i915_gem_object_set_to_gtt_domain(ce->state->obj, false);
+ if (ret)
+ goto error;
+
+ ret = i915_vma_pin(ce->state, 0, ctx->ggtt_alignment,
+ PIN_GLOBAL | PIN_HIGH);
if (ret)
goto error;
}
return;
if (ce->state)
- i915_gem_object_ggtt_unpin(ce->state);
+ i915_vma_unpin(ce->state);
i915_gem_context_put(ctx);
}
ret = PTR_ERR(ring);
goto error;
}
- engine->buffer = ring;
- if (I915_NEED_GFX_HWS(dev_priv)) {
- ret = init_status_page(engine);
+ if (HWS_NEEDS_PHYSICAL(dev_priv)) {
+ WARN_ON(engine->id != RCS);
+ ret = init_phys_status_page(engine);
if (ret)
goto error;
} else {
- WARN_ON(engine->id != RCS);
- ret = init_phys_status_page(engine);
+ ret = init_status_page(engine);
if (ret)
goto error;
}
ret = intel_ring_pin(ring);
if (ret) {
- DRM_ERROR("Failed to pin and map ringbuffer %s: %d\n",
- engine->name, ret);
- intel_destroy_ringbuffer_obj(ring);
+ intel_ring_free(ring);
goto error;
}
+ engine->buffer = ring;
return 0;
dev_priv = engine->i915;
if (engine->buffer) {
- WARN_ON(!IS_GEN2(dev_priv) && (I915_READ_MODE(engine) & MODE_IDLE) == 0);
+ WARN_ON(INTEL_GEN(dev_priv) > 2 &&
+ (I915_READ_MODE(engine) & MODE_IDLE) == 0);
intel_ring_unpin(engine->buffer);
intel_ring_free(engine->buffer);
if (engine->cleanup)
engine->cleanup(engine);
- if (I915_NEED_GFX_HWS(dev_priv)) {
- cleanup_status_page(engine);
- } else {
+ if (HWS_NEEDS_PHYSICAL(dev_priv)) {
WARN_ON(engine->id != RCS);
cleanup_phys_status_page(engine);
+ } else {
+ cleanup_status_page(engine);
}
intel_engine_cleanup_common(engine);
engine->i915 = NULL;
}
+void intel_legacy_submission_resume(struct drm_i915_private *dev_priv)
+{
+ struct intel_engine_cs *engine;
+
+ for_each_engine(engine, dev_priv) {
+ engine->buffer->head = engine->buffer->tail;
+ engine->buffer->last_retired_head = -1;
+ }
+}
+
int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request)
{
int ret;
if (WARN_ON(&target->ring_link == &ring->request_list))
return -ENOSPC;
- ret = i915_wait_request(target, true, NULL, NO_WAITBOOST);
+ ret = i915_wait_request(target,
+ I915_WAIT_INTERRUPTIBLE | I915_WAIT_LOCKED,
+ NULL, NO_WAITBOOST);
if (ret)
return ret;
- if (i915_reset_in_progress(&target->i915->gpu_error))
- return -EAGAIN;
-
i915_gem_request_retire_upto(target);
intel_ring_update_space(ring);
return 0;
}
-void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno)
-{
- struct drm_i915_private *dev_priv = engine->i915;
-
- /* Our semaphore implementation is strictly monotonic (i.e. we proceed
- * so long as the semaphore value in the register/page is greater
- * than the sync value), so whenever we reset the seqno,
- * so long as we reset the tracking semaphore value to 0, it will
- * always be before the next request's seqno. If we don't reset
- * the semaphore value, then when the seqno moves backwards all
- * future waits will complete instantly (causing rendering corruption).
- */
- if (IS_GEN6(dev_priv) || IS_GEN7(dev_priv)) {
- I915_WRITE(RING_SYNC_0(engine->mmio_base), 0);
- I915_WRITE(RING_SYNC_1(engine->mmio_base), 0);
- if (HAS_VEBOX(dev_priv))
- I915_WRITE(RING_SYNC_2(engine->mmio_base), 0);
- }
- if (dev_priv->semaphore_obj) {
- struct drm_i915_gem_object *obj = dev_priv->semaphore_obj;
- struct page *page = i915_gem_object_get_dirty_page(obj, 0);
- void *semaphores = kmap(page);
- memset(semaphores + GEN8_SEMAPHORE_OFFSET(engine->id, 0),
- 0, I915_NUM_ENGINES * gen8_semaphore_seqno_size);
- kunmap(page);
- }
- memset(engine->semaphore.sync_seqno, 0,
- sizeof(engine->semaphore.sync_seqno));
-
- intel_write_status_page(engine, I915_GEM_HWS_INDEX, seqno);
- if (engine->irq_seqno_barrier)
- engine->irq_seqno_barrier(engine);
- engine->last_submitted_seqno = seqno;
-
- engine->hangcheck.seqno = seqno;
-
- /* After manually advancing the seqno, fake the interrupt in case
- * there are any waiters for that seqno.
- */
- rcu_read_lock();
- intel_engine_wakeup(engine);
- rcu_read_unlock();
-}
-
static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
{
struct drm_i915_private *dev_priv = request->i915;
if (!i915.semaphores)
return;
- if (INTEL_GEN(dev_priv) >= 8 && !dev_priv->semaphore_obj) {
+ if (INTEL_GEN(dev_priv) >= 8 && !dev_priv->semaphore) {
+ struct i915_vma *vma;
+
obj = i915_gem_object_create(&dev_priv->drm, 4096);
- if (IS_ERR(obj)) {
- DRM_ERROR("Failed to allocate semaphore bo. Disabling semaphores\n");
- i915.semaphores = 0;
- } else {
- i915_gem_object_set_cache_level(obj, I915_CACHE_LLC);
- ret = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, 0);
- if (ret != 0) {
- i915_gem_object_put(obj);
- DRM_ERROR("Failed to pin semaphore bo. Disabling semaphores\n");
- i915.semaphores = 0;
- } else {
- dev_priv->semaphore_obj = obj;
- }
- }
- }
+ if (IS_ERR(obj))
+ goto err;
- if (!i915.semaphores)
- return;
+ vma = i915_vma_create(obj, &dev_priv->ggtt.base, NULL);
+ if (IS_ERR(vma))
+ goto err_obj;
+
+ ret = i915_gem_object_set_to_gtt_domain(obj, false);
+ if (ret)
+ goto err_obj;
+
+ ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
+ if (ret)
+ goto err_obj;
+
+ dev_priv->semaphore = vma;
+ }
if (INTEL_GEN(dev_priv) >= 8) {
- u64 offset = i915_gem_obj_ggtt_offset(dev_priv->semaphore_obj);
+ u32 offset = i915_ggtt_offset(dev_priv->semaphore);
engine->semaphore.sync_to = gen8_ring_sync_to;
engine->semaphore.signal = gen8_xcs_signal;
for (i = 0; i < I915_NUM_ENGINES; i++) {
- u64 ring_offset;
+ u32 ring_offset;
if (i != engine->id)
ring_offset = offset + GEN8_SEMAPHORE_OFFSET(engine->id, i);
* initialized as INVALID. Gen8 will initialize the
* sema between VCS2 and RCS later.
*/
- for (i = 0; i < I915_NUM_ENGINES; i++) {
+ for (i = 0; i < GEN6_NUM_SEMAPHORES; i++) {
static const struct {
u32 wait_mbox;
i915_reg_t mbox_reg;
- } sem_data[I915_NUM_ENGINES][I915_NUM_ENGINES] = {
- [RCS] = {
- [VCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_RV, .mbox_reg = GEN6_VRSYNC },
- [BCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_RB, .mbox_reg = GEN6_BRSYNC },
- [VECS] = { .wait_mbox = MI_SEMAPHORE_SYNC_RVE, .mbox_reg = GEN6_VERSYNC },
+ } sem_data[GEN6_NUM_SEMAPHORES][GEN6_NUM_SEMAPHORES] = {
+ [RCS_HW] = {
+ [VCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_RV, .mbox_reg = GEN6_VRSYNC },
+ [BCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_RB, .mbox_reg = GEN6_BRSYNC },
+ [VECS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_RVE, .mbox_reg = GEN6_VERSYNC },
},
- [VCS] = {
- [RCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_VR, .mbox_reg = GEN6_RVSYNC },
- [BCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_VB, .mbox_reg = GEN6_BVSYNC },
- [VECS] = { .wait_mbox = MI_SEMAPHORE_SYNC_VVE, .mbox_reg = GEN6_VEVSYNC },
+ [VCS_HW] = {
+ [RCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_VR, .mbox_reg = GEN6_RVSYNC },
+ [BCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_VB, .mbox_reg = GEN6_BVSYNC },
+ [VECS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_VVE, .mbox_reg = GEN6_VEVSYNC },
},
- [BCS] = {
- [RCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_BR, .mbox_reg = GEN6_RBSYNC },
- [VCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_BV, .mbox_reg = GEN6_VBSYNC },
- [VECS] = { .wait_mbox = MI_SEMAPHORE_SYNC_BVE, .mbox_reg = GEN6_VEBSYNC },
+ [BCS_HW] = {
+ [RCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_BR, .mbox_reg = GEN6_RBSYNC },
+ [VCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_BV, .mbox_reg = GEN6_VBSYNC },
+ [VECS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_BVE, .mbox_reg = GEN6_VEBSYNC },
},
- [VECS] = {
- [RCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_VER, .mbox_reg = GEN6_RVESYNC },
- [VCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_VEV, .mbox_reg = GEN6_VVESYNC },
- [BCS] = { .wait_mbox = MI_SEMAPHORE_SYNC_VEB, .mbox_reg = GEN6_BVESYNC },
+ [VECS_HW] = {
+ [RCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_VER, .mbox_reg = GEN6_RVESYNC },
+ [VCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_VEV, .mbox_reg = GEN6_VVESYNC },
+ [BCS_HW] = { .wait_mbox = MI_SEMAPHORE_SYNC_VEB, .mbox_reg = GEN6_BVESYNC },
},
};
u32 wait_mbox;
i915_reg_t mbox_reg;
- if (i == engine->id || i == VCS2) {
+ if (i == engine->hw_id) {
wait_mbox = MI_SEMAPHORE_SYNC_INVALID;
mbox_reg = GEN6_NOSYNC;
} else {
- wait_mbox = sem_data[engine->id][i].wait_mbox;
- mbox_reg = sem_data[engine->id][i].mbox_reg;
+ wait_mbox = sem_data[engine->hw_id][i].wait_mbox;
+ mbox_reg = sem_data[engine->hw_id][i].mbox_reg;
}
engine->semaphore.mbox.wait[i] = wait_mbox;
engine->semaphore.mbox.signal[i] = mbox_reg;
}
}
+
+ return;
+
+err_obj:
+ i915_gem_object_put(obj);
+err:
+ DRM_DEBUG_DRIVER("Failed to allocate space for semaphores, disabling\n");
+ i915.semaphores = 0;
}
static void intel_ring_init_irq(struct drm_i915_private *dev_priv,
intel_ring_init_semaphores(dev_priv, engine);
engine->init_hw = init_ring_common;
+ engine->reset_hw = reset_ring_common;
engine->emit_request = i9xx_emit_request;
if (i915.semaphores)
return ret;
if (INTEL_GEN(dev_priv) >= 6) {
- ret = intel_init_pipe_control(engine, 4096);
+ ret = intel_engine_create_scratch(engine, 4096);
if (ret)
return ret;
} else if (HAS_BROKEN_CS_TLB(dev_priv)) {
- ret = intel_init_pipe_control(engine, I830_WA_SIZE);
+ ret = intel_engine_create_scratch(engine, I830_WA_SIZE);
if (ret)
return ret;
}
*/
#define I915_RING_FREE_SPACE 64
-struct intel_hw_status_page {
- u32 *page_addr;
- unsigned int gfx_addr;
- struct drm_i915_gem_object *obj;
+struct intel_hw_status_page {
+ struct i915_vma *vma;
+ u32 *page_addr;
+ u32 ggtt_offset;
};
#define I915_READ_TAIL(engine) I915_READ(RING_TAIL((engine)->mmio_base))
#define GEN8_SEMAPHORE_OFFSET(__from, __to) \
(((__from) * I915_NUM_ENGINES + (__to)) * gen8_semaphore_seqno_size)
#define GEN8_SIGNAL_OFFSET(__ring, to) \
- (i915_gem_obj_ggtt_offset(dev_priv->semaphore_obj) + \
+ (dev_priv->semaphore->node.start + \
GEN8_SEMAPHORE_OFFSET((__ring)->id, (to)))
#define GEN8_WAIT_OFFSET(__ring, from) \
- (i915_gem_obj_ggtt_offset(dev_priv->semaphore_obj) + \
+ (dev_priv->semaphore->node.start + \
GEN8_SEMAPHORE_OFFSET(from, (__ring)->id))
enum intel_engine_hangcheck_action {
struct intel_engine_hangcheck {
u64 acthd;
- unsigned long user_interrupts;
u32 seqno;
int score;
enum intel_engine_hangcheck_action action;
};
struct intel_ring {
- struct drm_i915_gem_object *obj;
- void *vaddr;
struct i915_vma *vma;
+ void *vaddr;
struct intel_engine_cs *engine;
- struct list_head link;
struct list_head request_list;
* an option for future use.
* size: size of the batch in DWORDS
*/
-struct i915_ctx_workarounds {
+struct i915_ctx_workarounds {
struct i915_wa_ctx_bb {
u32 offset;
u32 size;
} indirect_ctx, per_ctx;
- struct drm_i915_gem_object *obj;
+ struct i915_vma *vma;
};
struct drm_i915_gem_request;
#define I915_NUM_ENGINES 5
#define _VCS(n) (VCS + (n))
unsigned int exec_id;
- unsigned int hw_id;
- unsigned int guc_id; /* XXX same as hw_id? */
+ enum intel_engine_hw_id {
+ RCS_HW = 0,
+ VCS_HW,
+ BCS_HW,
+ VECS_HW,
+ VCS2_HW
+ } hw_id;
+ enum intel_engine_hw_id guc_id; /* XXX same as hw_id? */
u64 fence_context;
u32 mmio_base;
unsigned int irq_shift;
struct intel_ring *buffer;
- struct list_head buffers;
/* Rather than have every client wait upon all user interrupts,
* with the herd waking after every interrupt and each doing the
* the overhead of waking that client is much preferred.
*/
struct intel_breadcrumbs {
- struct task_struct *irq_seqno_bh; /* bh for user interrupts */
- unsigned long irq_wakeups;
+ struct task_struct __rcu *irq_seqno_bh; /* bh for interrupts */
bool irq_posted;
spinlock_t lock; /* protects the lists of requests */
struct task_struct *signaler; /* used for fence signalling */
struct drm_i915_gem_request *first_signal;
struct timer_list fake_irq; /* used after a missed interrupt */
+ struct timer_list hangcheck; /* detect missed interrupts */
+
+ unsigned long timeout;
bool irq_enabled : 1;
bool rpm_wakelock : 1;
struct intel_hw_status_page status_page;
struct i915_ctx_workarounds wa_ctx;
+ struct i915_vma *scratch;
u32 irq_keep_mask; /* always keep these interrupts */
u32 irq_enable_mask; /* bitmask to enable ring interrupt */
void (*irq_disable)(struct intel_engine_cs *engine);
int (*init_hw)(struct intel_engine_cs *engine);
+ void (*reset_hw)(struct intel_engine_cs *engine,
+ struct drm_i915_gem_request *req);
int (*init_context)(struct drm_i915_gem_request *req);
#define I915_DISPATCH_PINNED BIT(1)
#define I915_DISPATCH_RS BIT(2)
int (*emit_request)(struct drm_i915_gem_request *req);
+
+ /* Pass the request to the hardware queue (e.g. directly into
+ * the legacy ringbuffer or to the end of an execlist).
+ *
+ * This is called from an atomic context with irqs disabled; must
+ * be irq safe.
+ */
void (*submit_request)(struct drm_i915_gem_request *req);
+
/* Some chipsets are not quite as coherent as advertised and need
* an expensive kick to force a true read of the up-to-date seqno.
* However, the up-to-date seqno is not always required and the last
u32 sync_seqno[I915_NUM_ENGINES-1];
union {
+#define GEN6_SEMAPHORE_LAST VECS_HW
+#define GEN6_NUM_SEMAPHORES (GEN6_SEMAPHORE_LAST + 1)
+#define GEN6_SEMAPHORES_MASK GENMASK(GEN6_SEMAPHORE_LAST, 0)
struct {
/* our mbox written by others */
- u32 wait[I915_NUM_ENGINES];
+ u32 wait[GEN6_NUM_SEMAPHORES];
/* mboxes this ring signals to */
- i915_reg_t signal[I915_NUM_ENGINES];
+ i915_reg_t signal[GEN6_NUM_SEMAPHORES];
} mbox;
u64 signal_ggtt[I915_NUM_ENGINES];
};
/* Execlists */
struct tasklet_struct irq_tasklet;
spinlock_t execlist_lock; /* used inside tasklet, use spin_lock_bh */
+ struct execlist_port {
+ struct drm_i915_gem_request *request;
+ unsigned int count;
+ } execlist_port[2];
struct list_head execlist_queue;
unsigned int fw_domains;
- unsigned int next_context_status_buffer;
- unsigned int idle_lite_restore_wa;
bool disable_lite_restore_wa;
+ bool preempt_wa;
u32 ctx_desc_template;
/**
/* An RCU guarded pointer to the last request. No reference is
* held to the request, users must carefully acquire a reference to
- * the request using i915_gem_active_get_request_rcu(), or hold the
+ * the request using i915_gem_active_get_rcu(), or hold the
* struct_mutex.
*/
struct i915_gem_active last_request;
struct intel_engine_hangcheck hangcheck;
- struct {
- struct drm_i915_gem_object *obj;
- u32 gtt_offset;
- } scratch;
-
bool needs_cmd_parser;
/*
void intel_engine_stop(struct intel_engine_cs *engine);
void intel_engine_cleanup(struct intel_engine_cs *engine);
+void intel_legacy_submission_resume(struct drm_i915_private *dev_priv);
+
int intel_ring_alloc_request_extras(struct drm_i915_gem_request *request);
int __must_check intel_ring_begin(struct drm_i915_gem_request *req, int n);
void intel_ring_update_space(struct intel_ring *ring);
void intel_engine_init_seqno(struct intel_engine_cs *engine, u32 seqno);
-
-int intel_init_pipe_control(struct intel_engine_cs *engine, int size);
-void intel_fini_pipe_control(struct intel_engine_cs *engine);
+void intel_engine_reset_irq(struct intel_engine_cs *engine);
void intel_engine_setup_common(struct intel_engine_cs *engine);
int intel_engine_init_common(struct intel_engine_cs *engine);
+int intel_engine_create_scratch(struct intel_engine_cs *engine, int size);
void intel_engine_cleanup_common(struct intel_engine_cs *engine);
static inline int intel_engine_idle(struct intel_engine_cs *engine,
- bool interruptible)
+ unsigned int flags)
{
/* Wait upon the last request to be completed */
return i915_gem_active_wait_unlocked(&engine->last_request,
- interruptible, NULL, NULL);
+ flags, NULL, NULL);
}
int intel_init_render_ring_buffer(struct intel_engine_cs *engine);
static inline u32 intel_hws_seqno_address(struct intel_engine_cs *engine)
{
- return engine->status_page.gfx_addr + I915_GEM_HWS_INDEX_ADDR;
+ return engine->status_page.ggtt_offset + I915_GEM_HWS_INDEX_ADDR;
}
/* intel_breadcrumbs.c -- user interrupt bottom-half for waiters */
struct intel_wait *wait);
void intel_engine_enable_signaling(struct drm_i915_gem_request *request);
-static inline bool intel_engine_has_waiter(struct intel_engine_cs *engine)
+static inline bool intel_engine_has_waiter(const struct intel_engine_cs *engine)
{
- return READ_ONCE(engine->breadcrumbs.irq_seqno_bh);
+ return rcu_access_pointer(engine->breadcrumbs.irq_seqno_bh);
}
-static inline bool intel_engine_wakeup(struct intel_engine_cs *engine)
+static inline bool intel_engine_wakeup(const struct intel_engine_cs *engine)
{
bool wakeup = false;
- struct task_struct *tsk = READ_ONCE(engine->breadcrumbs.irq_seqno_bh);
+
/* Note that for this not to dangerously chase a dangling pointer,
- * the caller is responsible for ensure that the task remain valid for
- * wake_up_process() i.e. that the RCU grace period cannot expire.
+ * we must hold the rcu_read_lock here.
*
* Also note that tsk is likely to be in !TASK_RUNNING state so an
* early test for tsk->state != TASK_RUNNING before wake_up_process()
* is unlikely to be beneficial.
*/
- if (tsk)
- wakeup = wake_up_process(tsk);
+ if (intel_engine_has_waiter(engine)) {
+ struct task_struct *tsk;
+
+ rcu_read_lock();
+ tsk = rcu_dereference(engine->breadcrumbs.irq_seqno_bh);
+ if (tsk)
+ wakeup = wake_up_process(tsk);
+ rcu_read_unlock();
+ }
+
return wakeup;
}
-void intel_engine_enable_fake_irq(struct intel_engine_cs *engine);
void intel_engine_fini_breadcrumbs(struct intel_engine_cs *engine);
unsigned int intel_kick_waiters(struct drm_i915_private *i915);
unsigned int intel_kick_signalers(struct drm_i915_private *i915);
*/
static void hsw_power_well_post_enable(struct drm_i915_private *dev_priv)
{
+ struct pci_dev *pdev = dev_priv->drm.pdev;
struct drm_device *dev = &dev_priv->drm;
/*
* sure vgacon can keep working normally without triggering interrupts
* and error messages.
*/
- vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO);
+ vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
outb(inb(VGA_MSR_READ), VGA_MSR_WRITE);
- vga_put(dev->pdev, VGA_RSRC_LEGACY_IO);
+ vga_put(pdev, VGA_RSRC_LEGACY_IO);
if (IS_BROADWELL(dev))
gen8_irq_power_well_post_enable(dev_priv,
static void skl_power_well_post_enable(struct drm_i915_private *dev_priv,
struct i915_power_well *power_well)
{
- struct drm_device *dev = &dev_priv->drm;
+ struct pci_dev *pdev = dev_priv->drm.pdev;
/*
* After we re-enable the power well, if we touch VGA register 0x3d5
* and error messages.
*/
if (power_well->data == SKL_DISP_PW_2) {
- vga_get_uninterruptible(dev->pdev, VGA_RSRC_LEGACY_IO);
+ vga_get_uninterruptible(pdev, VGA_RSRC_LEGACY_IO);
outb(inb(VGA_MSR_READ), VGA_MSR_WRITE);
- vga_put(dev->pdev, VGA_RSRC_LEGACY_IO);
+ vga_put(pdev, VGA_RSRC_LEGACY_IO);
gen8_irq_power_well_post_enable(dev_priv,
1 << PIPE_C | 1 << PIPE_B);
DRM_DEBUG_KMS("Disabling DC9\n");
gen9_set_dc_state(dev_priv, DC_STATE_DISABLE);
+
+ intel_pps_unlock_regs_wa(dev_priv);
}
static void assert_csr_loaded(struct drm_i915_private *dev_priv)
struct i915_power_well *power_well)
{
enum skl_disp_power_wells power_well_id = power_well->data;
- struct i915_power_well *cmn_a_well;
+ struct i915_power_well *cmn_a_well = NULL;
if (power_well_id == BXT_DPIO_CMN_BC) {
/*
bxt_ddi_phy_init(dev_priv, bxt_power_well_to_phy(power_well));
- if (power_well_id == BXT_DPIO_CMN_BC)
+ if (cmn_a_well)
intel_power_well_put(dev_priv, cmn_a_well);
}
}
i915_redisable_vga_power_on(&dev_priv->drm);
+
+ intel_pps_unlock_regs_wa(dev_priv);
}
static void vlv_display_power_well_deinit(struct drm_i915_private *dev_priv)
*/
void intel_power_domains_fini(struct drm_i915_private *dev_priv)
{
- struct device *device = &dev_priv->drm.pdev->dev;
+ struct device *kdev = &dev_priv->drm.pdev->dev;
/*
* The i915.ko module is still not prepared to be loaded when
* the platform doesn't support runtime PM.
*/
if (!HAS_RUNTIME_PM(dev_priv))
- pm_runtime_put(device);
+ pm_runtime_put(kdev);
}
static void intel_power_domains_sync_hw(struct drm_i915_private *dev_priv)
*/
void intel_runtime_pm_get(struct drm_i915_private *dev_priv)
{
- struct drm_device *dev = &dev_priv->drm;
- struct device *device = &dev->pdev->dev;
+ struct pci_dev *pdev = dev_priv->drm.pdev;
+ struct device *kdev = &pdev->dev;
- pm_runtime_get_sync(device);
+ pm_runtime_get_sync(kdev);
atomic_inc(&dev_priv->pm.wakeref_count);
assert_rpm_wakelock_held(dev_priv);
*/
bool intel_runtime_pm_get_if_in_use(struct drm_i915_private *dev_priv)
{
- struct drm_device *dev = &dev_priv->drm;
- struct device *device = &dev->pdev->dev;
+ struct pci_dev *pdev = dev_priv->drm.pdev;
+ struct device *kdev = &pdev->dev;
if (IS_ENABLED(CONFIG_PM)) {
- int ret = pm_runtime_get_if_in_use(device);
+ int ret = pm_runtime_get_if_in_use(kdev);
/*
* In cases runtime PM is disabled by the RPM core and we get
*/
void intel_runtime_pm_get_noresume(struct drm_i915_private *dev_priv)
{
- struct drm_device *dev = &dev_priv->drm;
- struct device *device = &dev->pdev->dev;
+ struct pci_dev *pdev = dev_priv->drm.pdev;
+ struct device *kdev = &pdev->dev;
assert_rpm_wakelock_held(dev_priv);
- pm_runtime_get_noresume(device);
+ pm_runtime_get_noresume(kdev);
atomic_inc(&dev_priv->pm.wakeref_count);
}
*/
void intel_runtime_pm_put(struct drm_i915_private *dev_priv)
{
- struct drm_device *dev = &dev_priv->drm;
- struct device *device = &dev->pdev->dev;
+ struct pci_dev *pdev = dev_priv->drm.pdev;
+ struct device *kdev = &pdev->dev;
assert_rpm_wakelock_held(dev_priv);
if (atomic_dec_and_test(&dev_priv->pm.wakeref_count))
atomic_inc(&dev_priv->pm.atomic_seq);
- pm_runtime_mark_last_busy(device);
- pm_runtime_put_autosuspend(device);
+ pm_runtime_mark_last_busy(kdev);
+ pm_runtime_put_autosuspend(kdev);
}
/**
*/
void intel_runtime_pm_enable(struct drm_i915_private *dev_priv)
{
+ struct pci_dev *pdev = dev_priv->drm.pdev;
struct drm_device *dev = &dev_priv->drm;
- struct device *device = &dev->pdev->dev;
+ struct device *kdev = &pdev->dev;
- pm_runtime_set_autosuspend_delay(device, 10000); /* 10s */
- pm_runtime_mark_last_busy(device);
+ pm_runtime_set_autosuspend_delay(kdev, 10000); /* 10s */
+ pm_runtime_mark_last_busy(kdev);
/*
* Take a permanent reference to disable the RPM functionality and drop
* platforms without RPM support.
*/
if (!HAS_RUNTIME_PM(dev)) {
- pm_runtime_dont_use_autosuspend(device);
- pm_runtime_get_sync(device);
+ pm_runtime_dont_use_autosuspend(kdev);
+ pm_runtime_get_sync(kdev);
} else {
- pm_runtime_use_autosuspend(device);
+ pm_runtime_use_autosuspend(kdev);
}
/*
* We drop that here and will reacquire it during unloading in
* intel_power_domains_fini().
*/
- pm_runtime_put_autosuspend(device);
+ pm_runtime_put_autosuspend(kdev);
}
-
}
static bool intel_sdvo_set_avi_infoframe(struct intel_sdvo *intel_sdvo,
- const struct drm_display_mode *adjusted_mode)
+ struct intel_crtc_state *pipe_config)
{
uint8_t sdvo_data[HDMI_INFOFRAME_SIZE(AVI)];
- struct drm_crtc *crtc = intel_sdvo->base.base.crtc;
- struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
union hdmi_infoframe frame;
int ret;
ssize_t len;
ret = drm_hdmi_avi_infoframe_from_display_mode(&frame.avi,
- adjusted_mode);
+ &pipe_config->base.adjusted_mode);
if (ret < 0) {
DRM_ERROR("couldn't fill AVI infoframe\n");
return false;
}
if (intel_sdvo->rgb_quant_range_selectable) {
- if (intel_crtc->config->limited_color_range)
+ if (pipe_config->limited_color_range)
frame.avi.quantization_range =
HDMI_QUANTIZATION_RANGE_LIMITED;
else
}
static bool intel_sdvo_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_sdvo *intel_sdvo = to_sdvo(encoder);
struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
return true;
}
-static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder)
+static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder,
+ struct intel_crtc_state *crtc_state,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = intel_encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
- struct intel_crtc *crtc = to_intel_crtc(intel_encoder->base.crtc);
- const struct drm_display_mode *adjusted_mode = &crtc->config->base.adjusted_mode;
- struct drm_display_mode *mode = &crtc->config->base.mode;
+ struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+ const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode;
+ struct drm_display_mode *mode = &crtc_state->base.mode;
struct intel_sdvo *intel_sdvo = to_sdvo(intel_encoder);
u32 sdvox;
struct intel_sdvo_in_out_map in_out;
struct intel_sdvo_dtd input_dtd, output_dtd;
int rate;
- if (!mode)
- return;
-
/* First, set the input mapping for the first input to our controlled
* output. This is only correct if we're a single-input device, in
* which case the first input is the output from the appropriate SDVO
if (!intel_sdvo_set_target_input(intel_sdvo))
return;
- if (crtc->config->has_hdmi_sink) {
+ if (crtc_state->has_hdmi_sink) {
intel_sdvo_set_encode(intel_sdvo, SDVO_ENCODE_HDMI);
intel_sdvo_set_colorimetry(intel_sdvo,
SDVO_COLORIMETRY_RGB256);
- intel_sdvo_set_avi_infoframe(intel_sdvo, adjusted_mode);
+ intel_sdvo_set_avi_infoframe(intel_sdvo, crtc_state);
} else
intel_sdvo_set_encode(intel_sdvo, SDVO_ENCODE_DVI);
DRM_INFO("Setting input timings on %s failed\n",
SDVO_NAME(intel_sdvo));
- switch (crtc->config->pixel_multiplier) {
+ switch (crtc_state->pixel_multiplier) {
default:
WARN(1, "unknown pixel multiplier specified\n");
case 1: rate = SDVO_CLOCK_RATE_MULT_1X; break;
/* The real mode polarity is set by the SDVO commands, using
* struct intel_sdvo_dtd. */
sdvox = SDVO_VSYNC_ACTIVE_HIGH | SDVO_HSYNC_ACTIVE_HIGH;
- if (!HAS_PCH_SPLIT(dev) && crtc->config->limited_color_range)
+ if (!HAS_PCH_SPLIT(dev) && crtc_state->limited_color_range)
sdvox |= HDMI_COLOR_RANGE_16_235;
if (INTEL_INFO(dev)->gen < 5)
sdvox |= SDVO_BORDER_ENABLE;
} else if (IS_I945G(dev) || IS_I945GM(dev) || IS_G33(dev)) {
/* done in crtc_mode_set as it lives inside the dpll register */
} else {
- sdvox |= (crtc->config->pixel_multiplier - 1)
+ sdvox |= (crtc_state->pixel_multiplier - 1)
<< SDVO_PORT_MULTIPLY_SHIFT;
}
pipe_config->pixel_multiplier, encoder_pixel_multiplier);
}
-static void intel_disable_sdvo(struct intel_encoder *encoder)
+static void intel_disable_sdvo(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *conn_state)
{
struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
struct intel_sdvo *intel_sdvo = to_sdvo(encoder);
}
}
-static void pch_disable_sdvo(struct intel_encoder *encoder)
+static void pch_disable_sdvo(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
}
-static void pch_post_disable_sdvo(struct intel_encoder *encoder)
+static void pch_post_disable_sdvo(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
- intel_disable_sdvo(encoder);
+ intel_disable_sdvo(encoder, old_crtc_state, old_conn_state);
}
-static void intel_enable_sdvo(struct intel_encoder *encoder)
+static void intel_enable_sdvo(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
intel_sdvo_init_ddc_proxy(struct intel_sdvo *sdvo,
struct drm_device *dev)
{
+ struct pci_dev *pdev = dev->pdev;
+
sdvo->ddc.owner = THIS_MODULE;
sdvo->ddc.class = I2C_CLASS_DDC;
snprintf(sdvo->ddc.name, I2C_NAME_SIZE, "SDVO DDC proxy");
- sdvo->ddc.dev.parent = &dev->pdev->dev;
+ sdvo->ddc.dev.parent = &pdev->dev;
sdvo->ddc.algo_data = sdvo;
sdvo->ddc.algo = &intel_sdvo_ddc_proxy;
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_plane *intel_plane = to_intel_plane(drm_plane);
struct drm_framebuffer *fb = plane_state->base.fb;
- struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+ const struct skl_wm_values *wm = &dev_priv->wm.skl_results;
+ struct drm_crtc *crtc = crtc_state->base.crtc;
+ struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
const int pipe = intel_plane->pipe;
const int plane = intel_plane->plane + 1;
- u32 plane_ctl, stride_div, stride;
+ u32 plane_ctl;
const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
- u32 surf_addr;
- u32 tile_height, plane_offset, plane_size;
+ u32 surf_addr = plane_state->main.offset;
unsigned int rotation = plane_state->base.rotation;
- int x_offset, y_offset;
+ u32 stride = skl_plane_stride(fb, 0, rotation);
int crtc_x = plane_state->base.dst.x1;
int crtc_y = plane_state->base.dst.y1;
uint32_t crtc_w = drm_rect_width(&plane_state->base.dst);
uint32_t crtc_h = drm_rect_height(&plane_state->base.dst);
- uint32_t x = plane_state->base.src.x1 >> 16;
- uint32_t y = plane_state->base.src.y1 >> 16;
+ uint32_t x = plane_state->main.x;
+ uint32_t y = plane_state->main.y;
uint32_t src_w = drm_rect_width(&plane_state->base.src) >> 16;
uint32_t src_h = drm_rect_height(&plane_state->base.src) >> 16;
plane_ctl |= skl_plane_ctl_rotation(rotation);
- stride_div = intel_fb_stride_alignment(dev_priv, fb->modifier[0],
- fb->pixel_format);
-
- /* Sizes are 0 based */
- src_w--;
- src_h--;
- crtc_w--;
- crtc_h--;
+ if (wm->dirty_pipes & drm_crtc_mask(crtc))
+ skl_write_plane_wm(intel_crtc, wm, plane);
if (key->flags) {
I915_WRITE(PLANE_KEYVAL(pipe, plane), key->min_value);
else if (key->flags & I915_SET_COLORKEY_SOURCE)
plane_ctl |= PLANE_CTL_KEY_ENABLE_SOURCE;
- surf_addr = intel_plane_obj_offset(intel_plane, obj, 0);
-
- if (intel_rotation_90_or_270(rotation)) {
- int cpp = drm_format_plane_cpp(fb->pixel_format, 0);
-
- /* stride: Surface height in tiles */
- tile_height = intel_tile_height(dev_priv, fb->modifier[0], cpp);
- stride = DIV_ROUND_UP(fb->height, tile_height);
- plane_size = (src_w << 16) | src_h;
- x_offset = stride * tile_height - y - (src_h + 1);
- y_offset = x;
- } else {
- stride = fb->pitches[0] / stride_div;
- plane_size = (src_h << 16) | src_w;
- x_offset = x;
- y_offset = y;
- }
- plane_offset = y_offset << 16 | x_offset;
+ /* Sizes are 0 based */
+ src_w--;
+ src_h--;
+ crtc_w--;
+ crtc_h--;
- I915_WRITE(PLANE_OFFSET(pipe, plane), plane_offset);
+ I915_WRITE(PLANE_OFFSET(pipe, plane), (y << 16) | x);
I915_WRITE(PLANE_STRIDE(pipe, plane), stride);
- I915_WRITE(PLANE_SIZE(pipe, plane), plane_size);
+ I915_WRITE(PLANE_SIZE(pipe, plane), (src_h << 16) | src_w);
/* program plane scaler */
if (plane_state->scaler_id >= 0) {
}
I915_WRITE(PLANE_CTL(pipe, plane), plane_ctl);
- I915_WRITE(PLANE_SURF(pipe, plane), surf_addr);
+ I915_WRITE(PLANE_SURF(pipe, plane),
+ intel_fb_gtt_offset(fb, rotation) + surf_addr);
POSTING_READ(PLANE_SURF(pipe, plane));
}
const int pipe = intel_plane->pipe;
const int plane = intel_plane->plane + 1;
+ /*
+ * We only populate skl_results on watermark updates, and if the
+ * plane's visiblity isn't actually changing neither is its watermarks.
+ */
+ if (!dplane->state->visible)
+ skl_write_plane_wm(to_intel_crtc(crtc),
+ &dev_priv->wm.skl_results, plane);
+
I915_WRITE(PLANE_CTL(pipe, plane), 0);
I915_WRITE(PLANE_SURF(pipe, plane), 0);
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_plane *intel_plane = to_intel_plane(dplane);
struct drm_framebuffer *fb = plane_state->base.fb;
- struct drm_i915_gem_object *obj = intel_fb_obj(fb);
int pipe = intel_plane->pipe;
int plane = intel_plane->plane;
u32 sprctl;
u32 sprsurf_offset, linear_offset;
unsigned int rotation = dplane->state->rotation;
- int cpp = drm_format_plane_cpp(fb->pixel_format, 0);
const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
int crtc_x = plane_state->base.dst.x1;
int crtc_y = plane_state->base.dst.y1;
*/
sprctl |= SP_GAMMA_ENABLE;
- if (i915_gem_object_is_tiled(obj))
+ if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
sprctl |= SP_TILED;
/* Sizes are 0 based */
crtc_w--;
crtc_h--;
- linear_offset = y * fb->pitches[0] + x * cpp;
- sprsurf_offset = intel_compute_tile_offset(&x, &y, fb, 0,
- fb->pitches[0], rotation);
- linear_offset -= sprsurf_offset;
+ intel_add_fb_offsets(&x, &y, plane_state, 0);
+ sprsurf_offset = intel_compute_tile_offset(&x, &y, plane_state, 0);
if (rotation == DRM_ROTATE_180) {
sprctl |= SP_ROTATE_180;
x += src_w;
y += src_h;
- linear_offset += src_h * fb->pitches[0] + src_w * cpp;
}
+ linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0);
+
if (key->flags) {
I915_WRITE(SPKEYMINVAL(pipe, plane), key->min_value);
I915_WRITE(SPKEYMAXVAL(pipe, plane), key->max_value);
I915_WRITE(SPSTRIDE(pipe, plane), fb->pitches[0]);
I915_WRITE(SPPOS(pipe, plane), (crtc_y << 16) | crtc_x);
- if (i915_gem_object_is_tiled(obj))
+ if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
I915_WRITE(SPTILEOFF(pipe, plane), (y << 16) | x);
else
I915_WRITE(SPLINOFF(pipe, plane), linear_offset);
I915_WRITE(SPSIZE(pipe, plane), (crtc_h << 16) | crtc_w);
I915_WRITE(SPCNTR(pipe, plane), sprctl);
- I915_WRITE(SPSURF(pipe, plane), i915_gem_obj_ggtt_offset(obj) +
- sprsurf_offset);
+ I915_WRITE(SPSURF(pipe, plane),
+ intel_fb_gtt_offset(fb, rotation) + sprsurf_offset);
POSTING_READ(SPSURF(pipe, plane));
}
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_plane *intel_plane = to_intel_plane(plane);
struct drm_framebuffer *fb = plane_state->base.fb;
- struct drm_i915_gem_object *obj = intel_fb_obj(fb);
enum pipe pipe = intel_plane->pipe;
u32 sprctl, sprscale = 0;
u32 sprsurf_offset, linear_offset;
unsigned int rotation = plane_state->base.rotation;
- int cpp = drm_format_plane_cpp(fb->pixel_format, 0);
const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
int crtc_x = plane_state->base.dst.x1;
int crtc_y = plane_state->base.dst.y1;
*/
sprctl |= SPRITE_GAMMA_ENABLE;
- if (i915_gem_object_is_tiled(obj))
+ if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
sprctl |= SPRITE_TILED;
if (IS_HASWELL(dev) || IS_BROADWELL(dev))
if (crtc_w != src_w || crtc_h != src_h)
sprscale = SPRITE_SCALE_ENABLE | (src_w << 16) | src_h;
- linear_offset = y * fb->pitches[0] + x * cpp;
- sprsurf_offset = intel_compute_tile_offset(&x, &y, fb, 0,
- fb->pitches[0], rotation);
- linear_offset -= sprsurf_offset;
+ intel_add_fb_offsets(&x, &y, plane_state, 0);
+ sprsurf_offset = intel_compute_tile_offset(&x, &y, plane_state, 0);
if (rotation == DRM_ROTATE_180) {
sprctl |= SPRITE_ROTATE_180;
if (!IS_HASWELL(dev) && !IS_BROADWELL(dev)) {
x += src_w;
y += src_h;
- linear_offset += src_h * fb->pitches[0] + src_w * cpp;
}
}
+ linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0);
+
if (key->flags) {
I915_WRITE(SPRKEYVAL(pipe), key->min_value);
I915_WRITE(SPRKEYMAX(pipe), key->max_value);
* register */
if (IS_HASWELL(dev) || IS_BROADWELL(dev))
I915_WRITE(SPROFFSET(pipe), (y << 16) | x);
- else if (i915_gem_object_is_tiled(obj))
+ else if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
I915_WRITE(SPRTILEOFF(pipe), (y << 16) | x);
else
I915_WRITE(SPRLINOFF(pipe), linear_offset);
I915_WRITE(SPRSCALE(pipe), sprscale);
I915_WRITE(SPRCTL(pipe), sprctl);
I915_WRITE(SPRSURF(pipe),
- i915_gem_obj_ggtt_offset(obj) + sprsurf_offset);
+ intel_fb_gtt_offset(fb, rotation) + sprsurf_offset);
POSTING_READ(SPRSURF(pipe));
}
struct drm_i915_private *dev_priv = to_i915(dev);
struct intel_plane *intel_plane = to_intel_plane(plane);
struct drm_framebuffer *fb = plane_state->base.fb;
- struct drm_i915_gem_object *obj = intel_fb_obj(fb);
int pipe = intel_plane->pipe;
u32 dvscntr, dvsscale;
u32 dvssurf_offset, linear_offset;
unsigned int rotation = plane_state->base.rotation;
- int cpp = drm_format_plane_cpp(fb->pixel_format, 0);
const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
int crtc_x = plane_state->base.dst.x1;
int crtc_y = plane_state->base.dst.y1;
*/
dvscntr |= DVS_GAMMA_ENABLE;
- if (i915_gem_object_is_tiled(obj))
+ if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
dvscntr |= DVS_TILED;
if (IS_GEN6(dev))
if (crtc_w != src_w || crtc_h != src_h)
dvsscale = DVS_SCALE_ENABLE | (src_w << 16) | src_h;
- linear_offset = y * fb->pitches[0] + x * cpp;
- dvssurf_offset = intel_compute_tile_offset(&x, &y, fb, 0,
- fb->pitches[0], rotation);
- linear_offset -= dvssurf_offset;
+ intel_add_fb_offsets(&x, &y, plane_state, 0);
+ dvssurf_offset = intel_compute_tile_offset(&x, &y, plane_state, 0);
if (rotation == DRM_ROTATE_180) {
dvscntr |= DVS_ROTATE_180;
x += src_w;
y += src_h;
- linear_offset += src_h * fb->pitches[0] + src_w * cpp;
}
+ linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0);
+
if (key->flags) {
I915_WRITE(DVSKEYVAL(pipe), key->min_value);
I915_WRITE(DVSKEYMAX(pipe), key->max_value);
I915_WRITE(DVSSTRIDE(pipe), fb->pitches[0]);
I915_WRITE(DVSPOS(pipe), (crtc_y << 16) | crtc_x);
- if (i915_gem_object_is_tiled(obj))
+ if (fb->modifier[0] == I915_FORMAT_MOD_X_TILED)
I915_WRITE(DVSTILEOFF(pipe), (y << 16) | x);
else
I915_WRITE(DVSLINOFF(pipe), linear_offset);
I915_WRITE(DVSSCALE(pipe), dvsscale);
I915_WRITE(DVSCNTR(pipe), dvscntr);
I915_WRITE(DVSSURF(pipe),
- i915_gem_obj_ggtt_offset(obj) + dvssurf_offset);
+ intel_fb_gtt_offset(fb, rotation) + dvssurf_offset);
POSTING_READ(DVSSURF(pipe));
}
int hscale, vscale;
int max_scale, min_scale;
bool can_scale;
+ int ret;
src->x1 = state->base.src_x;
src->y1 = state->base.src_y;
dst->y1 = crtc_y;
dst->y2 = crtc_y + crtc_h;
+ if (INTEL_GEN(dev) >= 9) {
+ ret = skl_check_plane_surface(state);
+ if (ret)
+ return ret;
+ }
+
return 0;
}
}
static void
-intel_enable_tv(struct intel_encoder *encoder)
+intel_enable_tv(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
}
static void
-intel_disable_tv(struct intel_encoder *encoder)
+intel_disable_tv(struct intel_encoder *encoder,
+ struct intel_crtc_state *old_crtc_state,
+ struct drm_connector_state *old_conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
static bool
intel_tv_compute_config(struct intel_encoder *encoder,
- struct intel_crtc_state *pipe_config)
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct intel_tv *intel_tv = enc_to_tv(encoder);
const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv);
color_conversion->av);
}
-static void intel_tv_pre_enable(struct intel_encoder *encoder)
+static void intel_tv_pre_enable(struct intel_encoder *encoder,
+ struct intel_crtc_state *pipe_config,
+ struct drm_connector_state *conn_state)
{
struct drm_device *dev = encoder->base.dev;
struct drm_i915_private *dev_priv = to_i915(dev);
__gen5_write(8)
__gen5_write(16)
__gen5_write(32)
-__gen5_write(64)
__gen2_write(8)
__gen2_write(16)
__gen2_write(32)
-__gen2_write(64)
#undef __gen5_write
#undef __gen2_write
__gen9_write(8)
__gen9_write(16)
__gen9_write(32)
-__gen9_write(64)
__chv_write(8)
__chv_write(16)
__chv_write(32)
-__chv_write(64)
__gen8_write(8)
__gen8_write(16)
__gen8_write(32)
-__gen8_write(64)
__hsw_write(8)
__hsw_write(16)
__hsw_write(32)
-__hsw_write(64)
__gen6_write(8)
__gen6_write(16)
__gen6_write(32)
-__gen6_write(64)
#undef __gen9_write
#undef __chv_write
__vgpu_write(8)
__vgpu_write(16)
__vgpu_write(32)
-__vgpu_write(64)
#undef __vgpu_write
#undef VGPU_WRITE_FOOTER
dev_priv->uncore.funcs.mmio_writeb = x##_write8; \
dev_priv->uncore.funcs.mmio_writew = x##_write16; \
dev_priv->uncore.funcs.mmio_writel = x##_write32; \
- dev_priv->uncore.funcs.mmio_writeq = x##_write64; \
} while (0)
#define ASSIGN_READ_MMIO_VFUNCS(x) \
if (engine_mask == ALL_ENGINES) {
hw_mask = GEN6_GRDOM_FULL;
} else {
+ unsigned int tmp;
+
hw_mask = 0;
- for_each_engine_masked(engine, dev_priv, engine_mask)
+ for_each_engine_masked(engine, dev_priv, engine_mask, tmp)
hw_mask |= hw_engine_mask[engine->id];
}
unsigned engine_mask)
{
struct intel_engine_cs *engine;
+ unsigned int tmp;
- for_each_engine_masked(engine, dev_priv, engine_mask)
+ for_each_engine_masked(engine, dev_priv, engine_mask, tmp)
if (gen8_request_engine_reset(engine))
goto not_ready;
return gen6_reset_engines(dev_priv, engine_mask);
not_ready:
- for_each_engine_masked(engine, dev_priv, engine_mask)
+ for_each_engine_masked(engine, dev_priv, engine_mask, tmp)
gen8_unrequest_engine_reset(engine);
return -EIO;
struct msm_drm_private *priv = dev->dev_private;
struct msm_kms *kms = priv->kms;
- drm_atomic_helper_wait_for_fences(dev, state);
+ drm_atomic_helper_wait_for_fences(dev, state, false);
kms->funcs->prepare_commit(kms, state);
if (ret)
goto out;
- ret = ttm_bo_move_ttm(bo, true, intr, no_wait_gpu, new_mem);
+ ret = ttm_bo_move_ttm(bo, intr, no_wait_gpu, new_mem);
out:
ttm_bo_mem_put(bo, &tmp_mem);
return ret;
if (ret)
return ret;
- ret = ttm_bo_move_ttm(bo, true, intr, no_wait_gpu, &tmp_mem);
+ ret = ttm_bo_move_ttm(bo, intr, no_wait_gpu, &tmp_mem);
if (ret)
goto out;
/* Fallback to software copy. */
ret = ttm_bo_wait(bo, intr, no_wait_gpu);
if (ret == 0)
- ret = ttm_bo_move_memcpy(bo, evict, intr, no_wait_gpu, new_mem);
+ ret = ttm_bo_move_memcpy(bo, intr, no_wait_gpu, new_mem);
out:
if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) {
static int
alloc_drawable(struct qxl_device *qdev, struct qxl_release **release)
{
- int ret;
- ret = qxl_alloc_release_reserved(qdev, sizeof(struct qxl_drawable),
- QXL_RELEASE_DRAWABLE, release,
- NULL);
- return ret;
+ return qxl_alloc_release_reserved(qdev, sizeof(struct qxl_drawable),
+ QXL_RELEASE_DRAWABLE, release, NULL);
}
static void
if (domain == QXL_GEM_DOMAIN_VRAM)
qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_VRAM | pflag;
if (domain == QXL_GEM_DOMAIN_SURFACE)
- qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_PRIV0 | pflag;
+ qbo->placements[c++].flags = TTM_PL_FLAG_CACHED | TTM_PL_FLAG_PRIV | pflag;
if (domain == QXL_GEM_DOMAIN_CPU)
qbo->placements[c++].flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM | pflag;
if (!c)
if (bo->tbo.mem.mem_type == TTM_PL_VRAM)
map = qdev->vram_mapping;
- else if (bo->tbo.mem.mem_type == TTM_PL_PRIV0)
+ else if (bo->tbo.mem.mem_type == TTM_PL_PRIV)
map = qdev->surface_mapping;
else
goto fallback;
if (bo->tbo.mem.mem_type == TTM_PL_VRAM)
map = qdev->vram_mapping;
- else if (bo->tbo.mem.mem_type == TTM_PL_PRIV0)
+ else if (bo->tbo.mem.mem_type == TTM_PL_PRIV)
map = qdev->surface_mapping;
else
goto fallback;
int qxl_surf_evict(struct qxl_device *qdev)
{
- return ttm_bo_evict_mm(&qdev->mman.bdev, TTM_PL_PRIV0);
+ return ttm_bo_evict_mm(&qdev->mman.bdev, TTM_PL_PRIV);
}
int qxl_vram_evict(struct qxl_device *qdev)
static int qxl_release_bo_alloc(struct qxl_device *qdev,
struct qxl_bo **bo)
{
- int ret;
/* pin releases bo's they are too messy to evict */
- ret = qxl_bo_create(qdev, PAGE_SIZE, false, true,
- QXL_GEM_DOMAIN_VRAM, NULL,
- bo);
- return ret;
+ return qxl_bo_create(qdev, PAGE_SIZE, false, true,
+ QXL_GEM_DOMAIN_VRAM, NULL, bo);
}
int qxl_release_list_add(struct qxl_release *release, struct qxl_bo *bo)
man->default_caching = TTM_PL_FLAG_CACHED;
break;
case TTM_PL_VRAM:
- case TTM_PL_PRIV0:
+ case TTM_PL_PRIV:
/* "On-card" video ram */
man->func = &ttm_bo_manager_func;
man->gpu_offset = 0;
mem->bus.base = qdev->vram_base;
mem->bus.offset = mem->start << PAGE_SHIFT;
break;
- case TTM_PL_PRIV0:
+ case TTM_PL_PRIV:
mem->bus.is_iomem = true;
mem->bus.base = qdev->surfaceram_base;
mem->bus.offset = mem->start << PAGE_SHIFT;
qxl_move_null(bo, new_mem);
return 0;
}
- return ttm_bo_move_memcpy(bo, evict, interruptible,
- no_wait_gpu, new_mem);
+ return ttm_bo_move_memcpy(bo, interruptible, no_wait_gpu,
+ new_mem);
}
static void qxl_bo_move_notify(struct ttm_buffer_object *bo,
qbo = to_qxl_bo(bo);
qdev = qbo->gem_base.dev->dev_private;
- if (bo->mem.mem_type == TTM_PL_PRIV0 && qbo->surface_id)
+ if (bo->mem.mem_type == TTM_PL_PRIV && qbo->surface_id)
qxl_surface_evict(qdev, qbo, new_mem ? true : false);
}
DRM_ERROR("Failed initializing VRAM heap.\n");
return r;
}
- r = ttm_bo_init_mm(&qdev->mman.bdev, TTM_PL_PRIV0,
+ r = ttm_bo_init_mm(&qdev->mman.bdev, TTM_PL_PRIV,
qdev->surfaceram_size / PAGE_SIZE);
if (r) {
DRM_ERROR("Failed initializing Surfaces heap.\n");
void qxl_ttm_fini(struct qxl_device *qdev)
{
ttm_bo_clean_mm(&qdev->mman.bdev, TTM_PL_VRAM);
- ttm_bo_clean_mm(&qdev->mman.bdev, TTM_PL_PRIV0);
+ ttm_bo_clean_mm(&qdev->mman.bdev, TTM_PL_PRIV);
ttm_bo_device_release(&qdev->mman.bdev);
qxl_ttm_global_fini(qdev);
DRM_INFO("qxl: ttm finalized\n");
if (i == 0)
qxl_mem_types_list[i].data = qdev->mman.bdev.man[TTM_PL_VRAM].priv;
else
- qxl_mem_types_list[i].data = qdev->mman.bdev.man[TTM_PL_PRIV0].priv;
+ qxl_mem_types_list[i].data = qdev->mman.bdev.man[TTM_PL_PRIV].priv;
}
return qxl_debugfs_add_files(qdev, qxl_mem_types_list, i);
WREG32(EVERGREEN_VIEWPORT_SIZE + radeon_crtc->crtc_offset,
(viewport_w << 16) | viewport_h);
- /* set pageflip to happen only at start of vblank interval (front porch) */
- WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3);
+ /* set pageflip to happen anywhere in vblank interval */
+ WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0);
if (!atomic && fb && fb != crtc->primary->fb) {
radeon_fb = to_radeon_framebuffer(fb);
WREG32(AVIVO_D1MODE_VIEWPORT_SIZE + radeon_crtc->crtc_offset,
(viewport_w << 16) | viewport_h);
- /* set pageflip to happen only at start of vblank interval (front porch) */
- WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3);
+ /* set pageflip to happen anywhere in vblank interval */
+ WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0);
if (!atomic && fb && fb != crtc->primary->fb) {
radeon_fb = to_radeon_framebuffer(fb);
{
struct radeon_connector_atom_dig *dig_connector = radeon_connector->con_priv;
u8 msg[DP_DPCD_SIZE];
- int ret, i;
+ int ret;
- for (i = 0; i < 7; i++) {
- ret = drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_DPCD_REV, msg,
- DP_DPCD_SIZE);
- if (ret == DP_DPCD_SIZE) {
- memcpy(dig_connector->dpcd, msg, DP_DPCD_SIZE);
+ ret = drm_dp_dpcd_read(&radeon_connector->ddc_bus->aux, DP_DPCD_REV, msg,
+ DP_DPCD_SIZE);
+ if (ret == DP_DPCD_SIZE) {
+ memcpy(dig_connector->dpcd, msg, DP_DPCD_SIZE);
- DRM_DEBUG_KMS("DPCD: %*ph\n", (int)sizeof(dig_connector->dpcd),
- dig_connector->dpcd);
+ DRM_DEBUG_KMS("DPCD: %*ph\n", (int)sizeof(dig_connector->dpcd),
+ dig_connector->dpcd);
- radeon_dp_probe_oui(radeon_connector);
+ radeon_dp_probe_oui(radeon_connector);
- return true;
- }
+ return true;
}
+
dig_connector->dpcd[0] = 0;
return false;
}
{
const __be32 *fw_data = NULL;
const __le32 *new_fw_data = NULL;
- u32 running, blackout = 0, tmp;
+ u32 running, tmp;
u32 *io_mc_regs = NULL;
const __le32 *new_io_mc_regs = NULL;
int i, regs_size, ucode_size;
running = RREG32(MC_SEQ_SUP_CNTL) & RUN_MASK;
if (running == 0) {
- if (running) {
- blackout = RREG32(MC_SHARED_BLACKOUT_CNTL);
- WREG32(MC_SHARED_BLACKOUT_CNTL, blackout | 1);
- }
-
/* reset the engine and set to writable */
WREG32(MC_SEQ_SUP_CNTL, 0x00000008);
WREG32(MC_SEQ_SUP_CNTL, 0x00000010);
break;
udelay(1);
}
-
- if (running)
- WREG32(MC_SHARED_BLACKOUT_CNTL, blackout);
}
return 0;
u32 cik_gfx_get_wptr(struct radeon_device *rdev,
struct radeon_ring *ring)
{
- u32 wptr;
-
- wptr = RREG32(CP_RB0_WPTR);
-
- return wptr;
+ return RREG32(CP_RB0_WPTR);
}
void cik_gfx_set_wptr(struct radeon_device *rdev,
return;
ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
- r = radeon_ring_init(rdev, ring, ring->ring_size, 0, RADEON_CP_PACKET2);
+ r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
if (r) {
dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
return;
#define UVD_UDEC_ADDR_CONFIG 0xef4c
#define UVD_UDEC_DB_ADDR_CONFIG 0xef50
#define UVD_UDEC_DBW_ADDR_CONFIG 0xef54
+#define UVD_NO_OP 0xeffc
#define UVD_LMI_EXT40_ADDR 0xf498
#define UVD_GP_SCRATCH4 0xf4e0
for (i = 0; i < rdev->num_crtc; i++) {
if (save->crtc_enabled[i]) {
tmp = RREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i]);
- if ((tmp & 0x7) != 3) {
+ if ((tmp & 0x7) != 0) {
tmp &= ~0x7;
- tmp |= 0x3;
WREG32(EVERGREEN_MASTER_UPDATE_MODE + crtc_offsets[i], tmp);
}
tmp = RREG32(EVERGREEN_GRPH_UPDATE + crtc_offsets[i]);
return;
ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
- r = radeon_ring_init(rdev, ring, ring->ring_size, 0, RADEON_CP_PACKET2);
+ r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
if (r) {
dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
return;
#define UVD_UDEC_ADDR_CONFIG 0xef4c
#define UVD_UDEC_DB_ADDR_CONFIG 0xef50
#define UVD_UDEC_DBW_ADDR_CONFIG 0xef54
+#define UVD_NO_OP 0xeffc
#define UVD_RBC_RB_RPTR 0xf690
#define UVD_RBC_RB_WPTR 0xf694
#define UVD_STATUS 0xf6bc
return;
ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
- r = radeon_ring_init(rdev, ring, ring->ring_size, 0, RADEON_CP_PACKET2);
+ r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
if (r) {
dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
return;
#define UVD_UDEC_ADDR_CONFIG 0xEF4C
#define UVD_UDEC_DB_ADDR_CONFIG 0xEF50
#define UVD_UDEC_DBW_ADDR_CONFIG 0xEF54
+#define UVD_NO_OP 0xEFFC
#define UVD_RBC_RB_RPTR 0xF690
#define UVD_RBC_RB_WPTR 0xF694
#define UVD_STATUS 0xf6bc
u32 r100_gfx_get_wptr(struct radeon_device *rdev,
struct radeon_ring *ring)
{
- u32 wptr;
-
- wptr = RREG32(RADEON_CP_RB_WPTR);
-
- return wptr;
+ return RREG32(RADEON_CP_RB_WPTR);
}
void r100_gfx_set_wptr(struct radeon_device *rdev,
u32 r600_gfx_get_wptr(struct radeon_device *rdev,
struct radeon_ring *ring)
{
- u32 wptr;
-
- wptr = RREG32(R600_CP_RB_WPTR);
-
- return wptr;
+ return RREG32(R600_CP_RB_WPTR);
}
void r600_gfx_set_wptr(struct radeon_device *rdev,
return;
ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
- r = radeon_ring_init(rdev, ring, ring->ring_size, 0, RADEON_CP_PACKET2);
+ r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
if (r) {
dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
return;
#define UVD_GPCOM_VCPU_DATA0 0xef10
#define UVD_GPCOM_VCPU_DATA1 0xef14
#define UVD_ENGINE_CNTL 0xef18
+#define UVD_NO_OP 0xeffc
#define UVD_SEMA_CNTL 0xf400
#define UVD_RB_ARB_CTRL 0xf480
struct work_struct unpin_work;
struct radeon_device *rdev;
int crtc_id;
+ u32 target_vblank;
uint64_t base;
struct drm_pending_vblank_event *event;
struct radeon_bo *old_rbo;
* Used at driver startup.
* Returns true if virtual or false if not.
*/
-static bool radeon_device_is_virtual(void)
+bool radeon_device_is_virtual(void)
{
#ifdef CONFIG_X86
return boot_cpu_has(X86_FEATURE_HYPERVISOR);
rdev = dev->dev_private;
- if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF ||
+ dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF)
return 0;
drm_kms_helper_poll_disable(dev);
struct drm_crtc *crtc;
int r;
- if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
+ if (dev->switch_power_state == DRM_SWITCH_POWER_OFF ||
+ dev->switch_power_state == DRM_SWITCH_POWER_DYNAMIC_OFF)
return 0;
if (fbcon) {
}
#endif
}
-
-#if defined(CONFIG_DEBUG_FS)
-int radeon_debugfs_init(struct drm_minor *minor)
-{
- return 0;
-}
-
-void radeon_debugfs_cleanup(struct drm_minor *minor)
-{
-}
-#endif
struct radeon_flip_work *work =
container_of(__work, struct radeon_flip_work, flip_work);
struct radeon_device *rdev = work->rdev;
+ struct drm_device *dev = rdev->ddev;
struct radeon_crtc *radeon_crtc = rdev->mode_info.crtcs[work->crtc_id];
struct drm_crtc *crtc = &radeon_crtc->base;
unsigned long flags;
int r;
- int vpos, hpos, stat, min_udelay = 0;
- unsigned repcnt = 4;
- struct drm_vblank_crtc *vblank = &crtc->dev->vblank[work->crtc_id];
+ int vpos, hpos;
down_read(&rdev->exclusive_lock);
if (work->fence) {
work->fence = NULL;
}
+ /* Wait until we're out of the vertical blank period before the one
+ * targeted by the flip
+ */
+ while (radeon_crtc->enabled &&
+ (radeon_get_crtc_scanoutpos(dev, work->crtc_id, 0,
+ &vpos, &hpos, NULL, NULL,
+ &crtc->hwmode)
+ & (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK)) ==
+ (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK) &&
+ (int)(work->target_vblank -
+ dev->driver->get_vblank_counter(dev, work->crtc_id)) > 0)
+ usleep_range(1000, 2000);
+
/* We borrow the event spin lock for protecting flip_status */
spin_lock_irqsave(&crtc->dev->event_lock, flags);
/* set the proper interrupt */
radeon_irq_kms_pflip_irq_get(rdev, radeon_crtc->crtc_id);
- /* If this happens to execute within the "virtually extended" vblank
- * interval before the start of the real vblank interval then it needs
- * to delay programming the mmio flip until the real vblank is entered.
- * This prevents completing a flip too early due to the way we fudge
- * our vblank counter and vblank timestamps in order to work around the
- * problem that the hw fires vblank interrupts before actual start of
- * vblank (when line buffer refilling is done for a frame). It
- * complements the fudging logic in radeon_get_crtc_scanoutpos() for
- * timestamping and radeon_get_vblank_counter_kms() for vblank counts.
- *
- * In practice this won't execute very often unless on very fast
- * machines because the time window for this to happen is very small.
- */
- while (radeon_crtc->enabled && --repcnt) {
- /* GET_DISTANCE_TO_VBLANKSTART returns distance to real vblank
- * start in hpos, and to the "fudged earlier" vblank start in
- * vpos.
- */
- stat = radeon_get_crtc_scanoutpos(rdev->ddev, work->crtc_id,
- GET_DISTANCE_TO_VBLANKSTART,
- &vpos, &hpos, NULL, NULL,
- &crtc->hwmode);
-
- if ((stat & (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE)) !=
- (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE) ||
- !(vpos >= 0 && hpos <= 0))
- break;
-
- /* Sleep at least until estimated real start of hw vblank */
- min_udelay = (-hpos + 1) * max(vblank->linedur_ns / 1000, 5);
- if (min_udelay > vblank->framedur_ns / 2000) {
- /* Don't wait ridiculously long - something is wrong */
- repcnt = 0;
- break;
- }
- spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
- usleep_range(min_udelay, 2 * min_udelay);
- spin_lock_irqsave(&crtc->dev->event_lock, flags);
- };
-
- if (!repcnt)
- DRM_DEBUG_DRIVER("Delay problem on crtc %d: min_udelay %d, "
- "framedur %d, linedur %d, stat %d, vpos %d, "
- "hpos %d\n", work->crtc_id, min_udelay,
- vblank->framedur_ns / 1000,
- vblank->linedur_ns / 1000, stat, vpos, hpos);
-
/* do the flip (mmio) */
radeon_page_flip(rdev, radeon_crtc->crtc_id, work->base, work->async);
up_read(&rdev->exclusive_lock);
}
-static int radeon_crtc_page_flip(struct drm_crtc *crtc,
- struct drm_framebuffer *fb,
- struct drm_pending_vblank_event *event,
- uint32_t page_flip_flags)
+static int radeon_crtc_page_flip_target(struct drm_crtc *crtc,
+ struct drm_framebuffer *fb,
+ struct drm_pending_vblank_event *event,
+ uint32_t page_flip_flags,
+ uint32_t target)
{
struct drm_device *dev = crtc->dev;
struct radeon_device *rdev = dev->dev_private;
base &= ~7;
}
work->base = base;
-
- r = drm_crtc_vblank_get(crtc);
- if (r) {
- DRM_ERROR("failed to get vblank before flip\n");
- goto pflip_cleanup;
- }
+ work->target_vblank = target - drm_crtc_vblank_count(crtc) +
+ dev->driver->get_vblank_counter(dev, work->crtc_id);
/* We borrow the event spin lock for protecting flip_work */
spin_lock_irqsave(&crtc->dev->event_lock, flags);
DRM_DEBUG_DRIVER("flip queue: crtc already busy\n");
spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
r = -EBUSY;
- goto vblank_cleanup;
+ goto pflip_cleanup;
}
radeon_crtc->flip_status = RADEON_FLIP_PENDING;
radeon_crtc->flip_work = work;
queue_work(radeon_crtc->flip_queue, &work->flip_work);
return 0;
-vblank_cleanup:
- drm_crtc_vblank_put(crtc);
-
pflip_cleanup:
if (unlikely(radeon_bo_reserve(new_rbo, false) != 0)) {
DRM_ERROR("failed to reserve new rbo in error path\n");
.gamma_set = radeon_crtc_gamma_set,
.set_config = radeon_crtc_set_config,
.destroy = radeon_crtc_destroy,
- .page_flip = radeon_crtc_page_flip,
+ .page_flip_target = radeon_crtc_page_flip_target,
};
static void radeon_crtc_init(struct drm_device *dev, int index)
}
if (tmp & AUX_SW_RX_TIMEOUT) {
- DRM_DEBUG_KMS("dp_aux_ch timed out\n");
ret = -ETIMEDOUT;
goto done;
}
* 2.44.0 - SET_APPEND_CNT packet3 support
* 2.45.0 - Allow setting shader registers using DMA/COPY packet3 on SI
* 2.46.0 - Add PFP_SYNC_ME support on evergreen
+ * 2.47.0 - Add UVD_NO_OP register support
*/
#define KMS_DRIVER_MAJOR 2
-#define KMS_DRIVER_MINOR 46
+#define KMS_DRIVER_MINOR 47
#define KMS_DRIVER_PATCHLEVEL 0
int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags);
int radeon_driver_unload_kms(struct drm_device *dev);
extern long radeon_kms_compat_ioctl(struct file *filp, unsigned int cmd,
unsigned long arg);
-#if defined(CONFIG_DEBUG_FS)
-int radeon_debugfs_init(struct drm_minor *minor);
-void radeon_debugfs_cleanup(struct drm_minor *minor);
-#endif
-
/* atpx handler */
#if defined(CONFIG_VGA_SWITCHEROO)
void radeon_register_atpx_handler(void);
static struct drm_driver kms_driver;
+bool radeon_device_is_virtual(void);
+
static int radeon_kick_out_firmware_fb(struct pci_dev *pdev)
{
struct apertures_struct *ap;
drm_put_dev(dev);
}
+static void
+radeon_pci_shutdown(struct pci_dev *pdev)
+{
+ /* if we are running in a VM, make sure the device
+ * torn down properly on reboot/shutdown
+ */
+ if (radeon_device_is_virtual())
+ radeon_pci_remove(pdev);
+}
+
static int radeon_pmops_suspend(struct device *dev)
{
struct pci_dev *pdev = to_pci_dev(dev);
{
struct pci_dev *pdev = to_pci_dev(dev);
struct drm_device *drm_dev = pci_get_drvdata(pdev);
+
+ /* GPU comes up enabled by the bios on resume */
+ if (radeon_is_px(drm_dev)) {
+ pm_runtime_disable(dev);
+ pm_runtime_set_active(dev);
+ pm_runtime_enable(dev);
+ }
+
return radeon_resume_kms(drm_dev, true, true);
}
.disable_vblank = radeon_disable_vblank_kms,
.get_vblank_timestamp = radeon_get_vblank_timestamp_kms,
.get_scanout_position = radeon_get_crtc_scanoutpos,
-#if defined(CONFIG_DEBUG_FS)
- .debugfs_init = radeon_debugfs_init,
- .debugfs_cleanup = radeon_debugfs_cleanup,
-#endif
.irq_preinstall = radeon_driver_irq_preinstall_kms,
.irq_postinstall = radeon_driver_irq_postinstall_kms,
.irq_uninstall = radeon_driver_irq_uninstall_kms,
.id_table = pciidlist,
.probe = radeon_pci_probe,
.remove = radeon_pci_remove,
+ .shutdown = radeon_pci_shutdown,
.driver.pm = &radeon_pm_ops,
};
*/
#include <linux/module.h>
#include <linux/slab.h>
+#include <linux/pm_runtime.h>
#include <drm/drmP.h>
#include <drm/drm_crtc.h>
struct radeon_device *rdev;
};
+static int
+radeonfb_open(struct fb_info *info, int user)
+{
+ struct radeon_fbdev *rfbdev = info->par;
+ struct radeon_device *rdev = rfbdev->rdev;
+ int ret = pm_runtime_get_sync(rdev->ddev->dev);
+ if (ret < 0 && ret != -EACCES) {
+ pm_runtime_mark_last_busy(rdev->ddev->dev);
+ pm_runtime_put_autosuspend(rdev->ddev->dev);
+ return ret;
+ }
+ return 0;
+}
+
+static int
+radeonfb_release(struct fb_info *info, int user)
+{
+ struct radeon_fbdev *rfbdev = info->par;
+ struct radeon_device *rdev = rfbdev->rdev;
+
+ pm_runtime_mark_last_busy(rdev->ddev->dev);
+ pm_runtime_put_autosuspend(rdev->ddev->dev);
+ return 0;
+}
+
static struct fb_ops radeonfb_ops = {
.owner = THIS_MODULE,
+ .fb_open = radeonfb_open,
+ .fb_release = radeonfb_release,
.fb_check_var = drm_fb_helper_check_var,
.fb_set_par = drm_fb_helper_set_par,
.fb_fillrect = drm_fb_helper_cfb_fillrect,
"Radeon i2c hw bus %s", name);
i2c->adapter.algo = &radeon_i2c_algo;
ret = i2c_add_adapter(&i2c->adapter);
- if (ret) {
- DRM_ERROR("Failed to register hw i2c %s\n", name);
+ if (ret)
goto out_free;
- }
} else if (rec->hw_capable &&
radeon_hw_i2c &&
ASIC_IS_DCE3(rdev)) {
"Radeon i2c hw bus %s", name);
i2c->adapter.algo = &radeon_atom_i2c_algo;
ret = i2c_add_adapter(&i2c->adapter);
- if (ret) {
- DRM_ERROR("Failed to register hw i2c %s\n", name);
+ if (ret)
goto out_free;
- }
} else {
/* set the radeon bit adapter */
snprintf(i2c->adapter.name, sizeof(i2c->adapter.name),
if (rdev->family >= CHIP_CAYMAN) {
struct radeon_fpriv *fpriv;
struct radeon_vm *vm;
- int r;
fpriv = kzalloc(sizeof(*fpriv), GFP_KERNEL);
if (unlikely(!fpriv)) {
- return -ENOMEM;
+ r = -ENOMEM;
+ goto out_suspend;
}
if (rdev->accel_working) {
r = radeon_vm_init(rdev, vm);
if (r) {
kfree(fpriv);
- return r;
+ goto out_suspend;
}
r = radeon_bo_reserve(rdev->ring_tmp_bo.bo, false);
if (r) {
radeon_vm_fini(rdev, vm);
kfree(fpriv);
- return r;
+ goto out_suspend;
}
/* map the ib pool buffer read only into
if (r) {
radeon_vm_fini(rdev, vm);
kfree(fpriv);
- return r;
+ goto out_suspend;
}
}
file_priv->driver_priv = fpriv;
}
+out_suspend:
pm_runtime_mark_last_busy(dev->dev);
pm_runtime_put_autosuspend(dev->dev);
- return 0;
+ return r;
}
/**
kfree(fpriv);
file_priv->driver_priv = NULL;
}
+ pm_runtime_mark_last_busy(dev->dev);
+ pm_runtime_put_autosuspend(dev->dev);
}
/**
{
struct radeon_device *rdev = dev->dev_private;
+ pm_runtime_get_sync(dev->dev);
+
mutex_lock(&rdev->gem.mutex);
if (rdev->hyperz_filp == file_priv)
rdev->hyperz_filp = NULL;
if (unlikely(r)) {
goto out_cleanup;
}
- r = ttm_bo_move_ttm(bo, true, interruptible, no_wait_gpu, new_mem);
+ r = ttm_bo_move_ttm(bo, interruptible, no_wait_gpu, new_mem);
out_cleanup:
ttm_bo_mem_put(bo, &tmp_mem);
return r;
if (unlikely(r)) {
return r;
}
- r = ttm_bo_move_ttm(bo, true, interruptible, no_wait_gpu, &tmp_mem);
+ r = ttm_bo_move_ttm(bo, interruptible, no_wait_gpu, &tmp_mem);
if (unlikely(r)) {
goto out_cleanup;
}
if (r) {
memcpy:
- r = ttm_bo_move_memcpy(bo, evict, interruptible,
- no_wait_gpu, new_mem);
+ r = ttm_bo_move_memcpy(bo, interruptible, no_wait_gpu, new_mem);
if (r) {
return r;
}
return r;
break;
case UVD_ENGINE_CNTL:
+ case UVD_NO_OP:
break;
default:
DRM_ERROR("Invalid reg 0x%X!\n",
ib.ptr[3] = addr >> 32;
ib.ptr[4] = PACKET0(UVD_GPCOM_VCPU_CMD, 0);
ib.ptr[5] = 0;
- for (i = 6; i < 16; ++i)
- ib.ptr[i] = PACKET2(0);
+ for (i = 6; i < 16; i += 2) {
+ ib.ptr[i] = PACKET0(UVD_NO_OP, 0);
+ ib.ptr[i+1] = 0;
+ }
ib.length_dw = 16;
r = radeon_ib_schedule(rdev, &ib, NULL, false);
for (i = 0; i < rdev->num_crtc; i++) {
if (save->crtc_enabled[i]) {
tmp = RREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + crtc_offsets[i]);
- if ((tmp & 0x7) != 3) {
+ if ((tmp & 0x7) != 0) {
tmp &= ~0x7;
- tmp |= 0x3;
WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + crtc_offsets[i], tmp);
}
tmp = RREG32(AVIVO_D1GRPH_UPDATE + crtc_offsets[i]);
return;
ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
- r = radeon_ring_init(rdev, ring, ring->ring_size, 0, RADEON_CP_PACKET2);
+ r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
if (r) {
dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
return;
#define UVD_UDEC_TILING_CONFIG 0xef40
#define UVD_UDEC_DB_TILING_CONFIG 0xef44
#define UVD_UDEC_DBW_TILING_CONFIG 0xef48
+#define UVD_NO_OP 0xeffc
#define GC_USER_SHADER_PIPE_CONFIG 0x8954
#define INACTIVE_QD_PIPES(x) ((x) << 8)
{
const __be32 *fw_data = NULL;
const __le32 *new_fw_data = NULL;
- u32 running, blackout = 0;
+ u32 running;
u32 *io_mc_regs = NULL;
const __le32 *new_io_mc_regs = NULL;
int i, regs_size, ucode_size;
running = RREG32(MC_SEQ_SUP_CNTL) & RUN_MASK;
if (running == 0) {
- if (running) {
- blackout = RREG32(MC_SHARED_BLACKOUT_CNTL);
- WREG32(MC_SHARED_BLACKOUT_CNTL, blackout | 1);
- }
-
/* reset the engine and set to writable */
WREG32(MC_SEQ_SUP_CNTL, 0x00000008);
WREG32(MC_SEQ_SUP_CNTL, 0x00000010);
break;
udelay(1);
}
-
- if (running)
- WREG32(MC_SHARED_BLACKOUT_CNTL, blackout);
}
return 0;
return;
ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
- r = radeon_ring_init(rdev, ring, ring->ring_size, 0, RADEON_CP_PACKET2);
+ r = radeon_ring_init(rdev, ring, ring->ring_size, 0, PACKET0(UVD_NO_OP, 0));
if (r) {
dev_err(rdev->dev, "failed initializing UVD ring (%d).\n", r);
return;
#define UVD_UDEC_ADDR_CONFIG 0xEF4C
#define UVD_UDEC_DB_ADDR_CONFIG 0xEF50
#define UVD_UDEC_DBW_ADDR_CONFIG 0xEF54
+#define UVD_NO_OP 0xEFFC
#define UVD_RBC_RB_RPTR 0xF690
#define UVD_RBC_RB_WPTR 0xF694
#define UVD_STATUS 0xf6bc
# Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
rockchipdrm-y := rockchip_drm_drv.o rockchip_drm_fb.o \
- rockchip_drm_gem.o rockchip_drm_vop.o
+ rockchip_drm_gem.o rockchip_drm_psr.o rockchip_drm_vop.o
rockchipdrm-$(CONFIG_DRM_FBDEV_EMULATION) += rockchip_drm_fbdev.o
obj-$(CONFIG_ROCKCHIP_ANALOGIX_DP) += analogix_dp-rockchip.o
#include <drm/bridge/analogix_dp.h>
#include "rockchip_drm_drv.h"
+#include "rockchip_drm_psr.h"
#include "rockchip_drm_vop.h"
#define RK3288_GRF_SOC_CON6 0x25c
#define HIWORD_UPDATE(val, mask) (val | (mask) << 16)
+#define PSR_WAIT_LINE_FLAG_TIMEOUT_MS 100
+
#define to_dp(nm) container_of(nm, struct rockchip_dp_device, nm)
/**
struct regmap *grf;
struct reset_control *rst;
+ struct work_struct psr_work;
+ spinlock_t psr_lock;
+ unsigned int psr_state;
+
const struct rockchip_dp_chip_data *data;
struct analogix_dp_plat_data plat_data;
};
+static void analogix_dp_psr_set(struct drm_encoder *encoder, bool enabled)
+{
+ struct rockchip_dp_device *dp = to_dp(encoder);
+ unsigned long flags;
+
+ dev_dbg(dp->dev, "%s PSR...\n", enabled ? "Entry" : "Exit");
+
+ spin_lock_irqsave(&dp->psr_lock, flags);
+ if (enabled)
+ dp->psr_state = EDP_VSC_PSR_STATE_ACTIVE;
+ else
+ dp->psr_state = ~EDP_VSC_PSR_STATE_ACTIVE;
+
+ schedule_work(&dp->psr_work);
+ spin_unlock_irqrestore(&dp->psr_lock, flags);
+}
+
+static void analogix_dp_psr_work(struct work_struct *work)
+{
+ struct rockchip_dp_device *dp =
+ container_of(work, typeof(*dp), psr_work);
+ struct drm_crtc *crtc = dp->encoder.crtc;
+ int psr_state = dp->psr_state;
+ int vact_end;
+ int ret;
+ unsigned long flags;
+
+ if (!crtc)
+ return;
+
+ vact_end = crtc->mode.vtotal - crtc->mode.vsync_start + crtc->mode.vdisplay;
+
+ ret = rockchip_drm_wait_line_flag(dp->encoder.crtc, vact_end,
+ PSR_WAIT_LINE_FLAG_TIMEOUT_MS);
+ if (ret) {
+ dev_err(dp->dev, "line flag interrupt did not arrive\n");
+ return;
+ }
+
+ spin_lock_irqsave(&dp->psr_lock, flags);
+ if (psr_state == EDP_VSC_PSR_STATE_ACTIVE)
+ analogix_dp_enable_psr(dp->dev);
+ else
+ analogix_dp_disable_psr(dp->dev);
+ spin_unlock_irqrestore(&dp->psr_lock, flags);
+}
+
static int rockchip_dp_pre_init(struct rockchip_dp_device *dp)
{
reset_control_assert(dp->rst);
struct rockchip_dp_device *dp = to_dp(plat_data);
int ret;
+ cancel_work_sync(&dp->psr_work);
+
ret = clk_prepare_enable(dp->pclk);
if (ret < 0) {
dev_err(dp->dev, "failed to enable pclk %d\n", ret);
dp->plat_data.power_off = rockchip_dp_powerdown;
dp->plat_data.get_modes = rockchip_dp_get_modes;
+ spin_lock_init(&dp->psr_lock);
+ dp->psr_state = ~EDP_VSC_PSR_STATE_ACTIVE;
+ INIT_WORK(&dp->psr_work, analogix_dp_psr_work);
+
+ rockchip_drm_psr_register(&dp->encoder, analogix_dp_psr_set);
+
return analogix_dp_bind(dev, dp->drm_dev, &dp->plat_data);
}
static void rockchip_dp_unbind(struct device *dev, struct device *master,
void *data)
{
+ struct rockchip_dp_device *dp = dev_get_drvdata(dev);
+
+ rockchip_drm_psr_unregister(&dp->encoder);
+
return analogix_dp_unbind(dev, master, data);
}
panel = of_drm_find_panel(panel_node);
of_node_put(panel_node);
- if (!panel) {
- DRM_ERROR("failed to find panel\n");
+ if (!panel)
return -EPROBE_DEFER;
- }
}
dp = devm_kzalloc(dev, sizeof(*dp), GFP_KERNEL);
drm_dev->dev_private = private;
+ INIT_LIST_HEAD(&private->psr_list);
+ spin_lock_init(&private->psr_list_lock);
+
drm_mode_config_init(drm_dev);
rockchip_drm_mode_config_init(drm_dev);
struct drm_gem_object *fbdev_bo;
const struct rockchip_crtc_funcs *crtc_funcs[ROCKCHIP_MAX_CRTC];
struct drm_atomic_state *state;
+
+ struct list_head psr_list;
+ spinlock_t psr_list_lock;
};
int rockchip_register_crtc_funcs(struct drm_crtc *crtc,
struct device *dev);
void rockchip_drm_dma_detach_device(struct drm_device *drm_dev,
struct device *dev);
+int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num,
+ unsigned int mstimeout);
+
#endif /* _ROCKCHIP_DRM_DRV_H_ */
#include "rockchip_drm_drv.h"
#include "rockchip_drm_fb.h"
#include "rockchip_drm_gem.h"
+#include "rockchip_drm_psr.h"
#define to_rockchip_fb(x) container_of(x, struct rockchip_drm_fb, fb)
rockchip_fb->obj[0], handle);
}
+static int rockchip_drm_fb_dirty(struct drm_framebuffer *fb,
+ struct drm_file *file,
+ unsigned int flags, unsigned int color,
+ struct drm_clip_rect *clips,
+ unsigned int num_clips)
+{
+ rockchip_drm_psr_flush(fb->dev);
+ return 0;
+}
+
static const struct drm_framebuffer_funcs rockchip_drm_fb_funcs = {
.destroy = rockchip_drm_fb_destroy,
.create_handle = rockchip_drm_fb_create_handle,
+ .dirty = rockchip_drm_fb_dirty,
};
static struct rockchip_drm_fb *
--- /dev/null
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author: Yakir Yang <ykk@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+
+#include "rockchip_drm_drv.h"
+#include "rockchip_drm_psr.h"
+
+#define PSR_FLUSH_TIMEOUT msecs_to_jiffies(3000) /* 3 seconds */
+
+enum psr_state {
+ PSR_FLUSH,
+ PSR_ENABLE,
+ PSR_DISABLE,
+};
+
+struct psr_drv {
+ struct list_head list;
+ struct drm_encoder *encoder;
+
+ spinlock_t lock;
+ enum psr_state state;
+
+ struct timer_list flush_timer;
+
+ void (*set)(struct drm_encoder *encoder, bool enable);
+};
+
+static struct psr_drv *find_psr_by_crtc(struct drm_crtc *crtc)
+{
+ struct rockchip_drm_private *drm_drv = crtc->dev->dev_private;
+ struct psr_drv *psr;
+ unsigned long flags;
+
+ spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
+ list_for_each_entry(psr, &drm_drv->psr_list, list) {
+ if (psr->encoder->crtc == crtc)
+ goto out;
+ }
+ psr = ERR_PTR(-ENODEV);
+
+out:
+ spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
+ return psr;
+}
+
+static void psr_set_state_locked(struct psr_drv *psr, enum psr_state state)
+{
+ /*
+ * Allowed finite state machine:
+ *
+ * PSR_ENABLE < = = = = = > PSR_FLUSH
+ * | ^ |
+ * | | |
+ * v | |
+ * PSR_DISABLE < - - - - - - - - -
+ */
+ if (state == psr->state)
+ return;
+
+ /* Requesting a flush when disabled is a noop */
+ if (state == PSR_FLUSH && psr->state == PSR_DISABLE)
+ return;
+
+ psr->state = state;
+
+ /* Already disabled in flush, change the state, but not the hardware */
+ if (state == PSR_DISABLE && psr->state == PSR_FLUSH)
+ return;
+
+ /* Actually commit the state change to hardware */
+ switch (psr->state) {
+ case PSR_ENABLE:
+ psr->set(psr->encoder, true);
+ break;
+
+ case PSR_DISABLE:
+ case PSR_FLUSH:
+ psr->set(psr->encoder, false);
+ break;
+ }
+}
+
+static void psr_set_state(struct psr_drv *psr, enum psr_state state)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&psr->lock, flags);
+ psr_set_state_locked(psr, state);
+ spin_unlock_irqrestore(&psr->lock, flags);
+}
+
+static void psr_flush_handler(unsigned long data)
+{
+ struct psr_drv *psr = (struct psr_drv *)data;
+ unsigned long flags;
+
+ /* If the state has changed since we initiated the flush, do nothing */
+ spin_lock_irqsave(&psr->lock, flags);
+ if (psr->state == PSR_FLUSH)
+ psr_set_state_locked(psr, PSR_ENABLE);
+ spin_unlock_irqrestore(&psr->lock, flags);
+}
+
+/**
+ * rockchip_drm_psr_enable - enable the encoder PSR which bind to given CRTC
+ * @crtc: CRTC to obtain the PSR encoder
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int rockchip_drm_psr_enable(struct drm_crtc *crtc)
+{
+ struct psr_drv *psr = find_psr_by_crtc(crtc);
+
+ if (IS_ERR(psr))
+ return PTR_ERR(psr);
+
+ psr_set_state(psr, PSR_ENABLE);
+ return 0;
+}
+EXPORT_SYMBOL(rockchip_drm_psr_enable);
+
+/**
+ * rockchip_drm_psr_disable - disable the encoder PSR which bind to given CRTC
+ * @crtc: CRTC to obtain the PSR encoder
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int rockchip_drm_psr_disable(struct drm_crtc *crtc)
+{
+ struct psr_drv *psr = find_psr_by_crtc(crtc);
+
+ if (IS_ERR(psr))
+ return PTR_ERR(psr);
+
+ psr_set_state(psr, PSR_DISABLE);
+ return 0;
+}
+EXPORT_SYMBOL(rockchip_drm_psr_disable);
+
+/**
+ * rockchip_drm_psr_flush - force to flush all registered PSR encoders
+ * @dev: drm device
+ *
+ * Disable the PSR function for all registered encoders, and then enable the
+ * PSR function back after PSR_FLUSH_TIMEOUT. If encoder PSR state have been
+ * changed during flush time, then keep the state no change after flush
+ * timeout.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+void rockchip_drm_psr_flush(struct drm_device *dev)
+{
+ struct rockchip_drm_private *drm_drv = dev->dev_private;
+ struct psr_drv *psr;
+ unsigned long flags;
+
+ spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
+ list_for_each_entry(psr, &drm_drv->psr_list, list) {
+ mod_timer(&psr->flush_timer,
+ round_jiffies_up(jiffies + PSR_FLUSH_TIMEOUT));
+
+ psr_set_state(psr, PSR_FLUSH);
+ }
+ spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
+}
+EXPORT_SYMBOL(rockchip_drm_psr_flush);
+
+/**
+ * rockchip_drm_psr_register - register encoder to psr driver
+ * @encoder: encoder that obtain the PSR function
+ * @psr_set: call back to set PSR state
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int rockchip_drm_psr_register(struct drm_encoder *encoder,
+ void (*psr_set)(struct drm_encoder *, bool enable))
+{
+ struct rockchip_drm_private *drm_drv = encoder->dev->dev_private;
+ struct psr_drv *psr;
+ unsigned long flags;
+
+ if (!encoder || !psr_set)
+ return -EINVAL;
+
+ psr = kzalloc(sizeof(struct psr_drv), GFP_KERNEL);
+ if (!psr)
+ return -ENOMEM;
+
+ setup_timer(&psr->flush_timer, psr_flush_handler, (unsigned long)psr);
+ spin_lock_init(&psr->lock);
+
+ psr->state = PSR_DISABLE;
+ psr->encoder = encoder;
+ psr->set = psr_set;
+
+ spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
+ list_add_tail(&psr->list, &drm_drv->psr_list);
+ spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
+
+ return 0;
+}
+EXPORT_SYMBOL(rockchip_drm_psr_register);
+
+/**
+ * rockchip_drm_psr_unregister - unregister encoder to psr driver
+ * @encoder: encoder that obtain the PSR function
+ * @psr_set: call back to set PSR state
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+void rockchip_drm_psr_unregister(struct drm_encoder *encoder)
+{
+ struct rockchip_drm_private *drm_drv = encoder->dev->dev_private;
+ struct psr_drv *psr, *n;
+ unsigned long flags;
+
+ spin_lock_irqsave(&drm_drv->psr_list_lock, flags);
+ list_for_each_entry_safe(psr, n, &drm_drv->psr_list, list) {
+ if (psr->encoder == encoder) {
+ del_timer(&psr->flush_timer);
+ list_del(&psr->list);
+ kfree(psr);
+ }
+ }
+ spin_unlock_irqrestore(&drm_drv->psr_list_lock, flags);
+}
+EXPORT_SYMBOL(rockchip_drm_psr_unregister);
--- /dev/null
+/*
+ * Copyright (C) Fuzhou Rockchip Electronics Co.Ltd
+ * Author: Yakir Yang <ykk@rock-chips.com>
+ *
+ * This software is licensed under the terms of the GNU General Public
+ * License version 2, as published by the Free Software Foundation, and
+ * may be copied, distributed, and modified under those terms.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ */
+
+#ifndef __ROCKCHIP_DRM_PSR___
+#define __ROCKCHIP_DRM_PSR___
+
+void rockchip_drm_psr_flush(struct drm_device *dev);
+int rockchip_drm_psr_enable(struct drm_crtc *crtc);
+int rockchip_drm_psr_disable(struct drm_crtc *crtc);
+
+int rockchip_drm_psr_register(struct drm_encoder *encoder,
+ void (*psr_set)(struct drm_encoder *, bool enable));
+void rockchip_drm_psr_unregister(struct drm_encoder *encoder);
+
+#endif /* __ROCKCHIP_DRM_PSR__ */
#include "rockchip_drm_drv.h"
#include "rockchip_drm_gem.h"
#include "rockchip_drm_fb.h"
+#include "rockchip_drm_psr.h"
#include "rockchip_drm_vop.h"
-#define __REG_SET_RELAXED(x, off, mask, shift, v) \
- vop_mask_write_relaxed(x, off, (mask) << shift, (v) << shift)
-#define __REG_SET_NORMAL(x, off, mask, shift, v) \
- vop_mask_write(x, off, (mask) << shift, (v) << shift)
+#define __REG_SET_RELAXED(x, off, mask, shift, v, write_mask) \
+ vop_mask_write(x, off, mask, shift, v, write_mask, true)
+
+#define __REG_SET_NORMAL(x, off, mask, shift, v, write_mask) \
+ vop_mask_write(x, off, mask, shift, v, write_mask, false)
#define REG_SET(x, base, reg, v, mode) \
- __REG_SET_##mode(x, base + reg.offset, reg.mask, reg.shift, v)
+ __REG_SET_##mode(x, base + reg.offset, \
+ reg.mask, reg.shift, v, reg.write_mask)
#define REG_SET_MASK(x, base, reg, mask, v, mode) \
- __REG_SET_##mode(x, base + reg.offset, mask, reg.shift, v)
+ __REG_SET_##mode(x, base + reg.offset, \
+ mask, reg.shift, v, reg.write_mask)
#define VOP_WIN_SET(x, win, name, v) \
REG_SET(x, win->base, win->phy->name, v, RELAXED)
struct device *dev;
struct drm_device *drm_dev;
bool is_enabled;
+ bool vblank_active;
/* mutex vsync_ work */
struct mutex vsync_mutex;
/* protected by dev->event_lock */
struct drm_pending_vblank_event *event;
+ struct completion line_flag_completion;
+
const struct vop_data *data;
uint32_t *regsbak;
}
static inline void vop_mask_write(struct vop *vop, uint32_t offset,
- uint32_t mask, uint32_t v)
+ uint32_t mask, uint32_t shift, uint32_t v,
+ bool write_mask, bool relaxed)
{
- if (mask) {
- uint32_t cached_val = vop->regsbak[offset >> 2];
-
- cached_val = (cached_val & ~mask) | v;
- writel(cached_val, vop->regs + offset);
- vop->regsbak[offset >> 2] = cached_val;
- }
-}
+ if (!mask)
+ return;
-static inline void vop_mask_write_relaxed(struct vop *vop, uint32_t offset,
- uint32_t mask, uint32_t v)
-{
- if (mask) {
+ if (write_mask) {
+ v = ((v << shift) & 0xffff) | (mask << (shift + 16));
+ } else {
uint32_t cached_val = vop->regsbak[offset >> 2];
- cached_val = (cached_val & ~mask) | v;
- writel_relaxed(cached_val, vop->regs + offset);
- vop->regsbak[offset >> 2] = cached_val;
+ v = (cached_val & ~(mask << shift)) | ((v & mask) << shift);
+ vop->regsbak[offset >> 2] = v;
}
+
+ if (relaxed)
+ writel_relaxed(v, vop->regs + offset);
+ else
+ writel(v, vop->regs + offset);
}
static inline uint32_t vop_get_intr_type(struct vop *vop,
spin_unlock_irqrestore(&vop->irq_lock, flags);
}
+/*
+ * (1) each frame starts at the start of the Vsync pulse which is signaled by
+ * the "FRAME_SYNC" interrupt.
+ * (2) the active data region of each frame ends at dsp_vact_end
+ * (3) we should program this same number (dsp_vact_end) into dsp_line_frag_num,
+ * to get "LINE_FLAG" interrupt at the end of the active on screen data.
+ *
+ * VOP_INTR_CTRL0.dsp_line_frag_num = VOP_DSP_VACT_ST_END.dsp_vact_end
+ * Interrupts
+ * LINE_FLAG -------------------------------+
+ * FRAME_SYNC ----+ |
+ * | |
+ * v v
+ * | Vsync | Vbp | Vactive | Vfp |
+ * ^ ^ ^ ^
+ * | | | |
+ * | | | |
+ * dsp_vs_end ------------+ | | | VOP_DSP_VTOTAL_VS_END
+ * dsp_vact_start --------------+ | | VOP_DSP_VACT_ST_END
+ * dsp_vact_end ----------------------------+ | VOP_DSP_VACT_ST_END
+ * dsp_total -------------------------------------+ VOP_DSP_VTOTAL_VS_END
+ */
+static bool vop_line_flag_irq_is_enabled(struct vop *vop)
+{
+ uint32_t line_flag_irq;
+ unsigned long flags;
+
+ spin_lock_irqsave(&vop->irq_lock, flags);
+
+ line_flag_irq = VOP_INTR_GET_TYPE(vop, enable, LINE_FLAG_INTR);
+
+ spin_unlock_irqrestore(&vop->irq_lock, flags);
+
+ return !!line_flag_irq;
+}
+
+static void vop_line_flag_irq_enable(struct vop *vop, int line_num)
+{
+ unsigned long flags;
+
+ if (WARN_ON(!vop->is_enabled))
+ return;
+
+ spin_lock_irqsave(&vop->irq_lock, flags);
+
+ VOP_CTRL_SET(vop, line_flag_num[0], line_num);
+ VOP_INTR_SET_TYPE(vop, enable, LINE_FLAG_INTR, 1);
+
+ spin_unlock_irqrestore(&vop->irq_lock, flags);
+}
+
+static void vop_line_flag_irq_disable(struct vop *vop)
+{
+ unsigned long flags;
+
+ if (WARN_ON(!vop->is_enabled))
+ return;
+
+ spin_lock_irqsave(&vop->irq_lock, flags);
+
+ VOP_INTR_SET_TYPE(vop, enable, LINE_FLAG_INTR, 0);
+
+ spin_unlock_irqrestore(&vop->irq_lock, flags);
+}
+
static int vop_enable(struct drm_crtc *crtc)
{
struct vop *vop = to_vop(crtc);
spin_unlock_irqrestore(&vop->irq_lock, flags);
+ rockchip_drm_psr_disable(&vop->crtc);
+
return 0;
}
VOP_INTR_SET_TYPE(vop, enable, FS_INTR, 0);
spin_unlock_irqrestore(&vop->irq_lock, flags);
+
+ rockchip_drm_psr_enable(&vop->crtc);
}
static void vop_crtc_wait_for_update(struct drm_crtc *crtc)
u16 vsync_len = adjusted_mode->vsync_end - adjusted_mode->vsync_start;
u16 vact_st = adjusted_mode->vtotal - adjusted_mode->vsync_start;
u16 vact_end = vact_st + vdisplay;
- uint32_t val;
+ uint32_t pin_pol, val;
int ret;
WARN_ON(vop->event);
vop_dsp_hold_valid_irq_disable(vop);
}
- val = 0x8;
- val |= (adjusted_mode->flags & DRM_MODE_FLAG_NHSYNC) ? 0 : 1;
- val |= (adjusted_mode->flags & DRM_MODE_FLAG_NVSYNC) ? 0 : (1 << 1);
- VOP_CTRL_SET(vop, pin_pol, val);
+ pin_pol = 0x8;
+ pin_pol |= (adjusted_mode->flags & DRM_MODE_FLAG_NHSYNC) ? 0 : 1;
+ pin_pol |= (adjusted_mode->flags & DRM_MODE_FLAG_NVSYNC) ? 0 : (1 << 1);
+ VOP_CTRL_SET(vop, pin_pol, pin_pol);
+
switch (s->output_type) {
case DRM_MODE_CONNECTOR_LVDS:
VOP_CTRL_SET(vop, rgb_en, 1);
+ VOP_CTRL_SET(vop, rgb_pin_pol, pin_pol);
break;
case DRM_MODE_CONNECTOR_eDP:
+ VOP_CTRL_SET(vop, edp_pin_pol, pin_pol);
VOP_CTRL_SET(vop, edp_en, 1);
break;
case DRM_MODE_CONNECTOR_HDMIA:
+ VOP_CTRL_SET(vop, hdmi_pin_pol, pin_pol);
VOP_CTRL_SET(vop, hdmi_en, 1);
break;
case DRM_MODE_CONNECTOR_DSI:
+ VOP_CTRL_SET(vop, mipi_pin_pol, pin_pol);
VOP_CTRL_SET(vop, mipi_en, 1);
break;
default:
struct vop *vop = to_vop(crtc);
spin_lock_irq(&crtc->dev->event_lock);
- if (crtc->state->event) {
- WARN_ON(drm_crtc_vblank_get(crtc) != 0);
- WARN_ON(vop->event);
+ vop->vblank_active = true;
+ WARN_ON(drm_crtc_vblank_get(crtc) != 0);
+ WARN_ON(vop->event);
+ if (crtc->state->event) {
vop->event = crtc->state->event;
crtc->state->event = NULL;
}
spin_lock_irqsave(&drm->event_lock, flags);
if (vop->event) {
-
drm_crtc_send_vblank_event(crtc, vop->event);
- drm_crtc_vblank_put(crtc);
vop->event = NULL;
}
+ if (vop->vblank_active) {
+ vop->vblank_active = false;
+ drm_crtc_vblank_put(crtc);
+ }
spin_unlock_irqrestore(&drm->event_lock, flags);
if (!completion_done(&vop->wait_update_complete))
ret = IRQ_HANDLED;
}
+ if (active_irqs & LINE_FLAG_INTR) {
+ complete(&vop->line_flag_completion);
+ active_irqs &= ~LINE_FLAG_INTR;
+ ret = IRQ_HANDLED;
+ }
+
if (active_irqs & FS_INTR) {
drm_crtc_handle_vblank(crtc);
vop_handle_vblank(vop);
init_completion(&vop->dsp_hold_completion);
init_completion(&vop->wait_update_complete);
+ init_completion(&vop->line_flag_completion);
crtc->port = port;
rockchip_register_crtc_funcs(crtc, &private_crtc_funcs);
clk_disable(vop->aclk);
vop->is_enabled = false;
+ vop->vblank_active = false;
return 0;
}
}
+/**
+ * rockchip_drm_wait_line_flag - acqiure the give line flag event
+ * @crtc: CRTC to enable line flag
+ * @line_num: interested line number
+ * @mstimeout: millisecond for timeout
+ *
+ * Driver would hold here until the interested line flag interrupt have
+ * happened or timeout to wait.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num,
+ unsigned int mstimeout)
+{
+ struct vop *vop = to_vop(crtc);
+ unsigned long jiffies_left;
+
+ if (!crtc || !vop->is_enabled)
+ return -ENODEV;
+
+ if (line_num > crtc->mode.vtotal || mstimeout <= 0)
+ return -EINVAL;
+
+ if (vop_line_flag_irq_is_enabled(vop))
+ return -EBUSY;
+
+ reinit_completion(&vop->line_flag_completion);
+ vop_line_flag_irq_enable(vop, line_num);
+
+ jiffies_left = wait_for_completion_timeout(&vop->line_flag_completion,
+ msecs_to_jiffies(mstimeout));
+ vop_line_flag_irq_disable(vop);
+
+ if (jiffies_left == 0) {
+ dev_err(vop->dev, "Timeout waiting for IRQ\n");
+ return -ETIMEDOUT;
+ }
+
+ return 0;
+}
+EXPORT_SYMBOL(rockchip_drm_wait_line_flag);
+
static int vop_bind(struct device *dev, struct device *master, void *data)
{
struct platform_device *pdev = to_platform_device(dev);
return ret;
pm_runtime_enable(&pdev->dev);
+
return 0;
}
uint32_t offset;
uint32_t shift;
uint32_t mask;
+ bool write_mask;
};
struct vop_ctrl {
struct vop_reg dither_down;
struct vop_reg dither_up;
struct vop_reg pin_pol;
+ struct vop_reg rgb_pin_pol;
+ struct vop_reg hdmi_pin_pol;
+ struct vop_reg edp_pin_pol;
+ struct vop_reg mipi_pin_pol;
struct vop_reg htotal_pw;
struct vop_reg hact_st_end;
struct vop_reg hpost_st_end;
struct vop_reg vpost_st_end;
+ struct vop_reg line_flag_num[2];
+
struct vop_reg cfg_done;
};
#define VOP_REG(off, _mask, s) \
{.offset = off, \
.mask = _mask, \
- .shift = s,}
+ .shift = s, \
+ .write_mask = false,}
+
+#define VOP_REG_MASK(off, _mask, s) \
+ {.offset = off, \
+ .mask = _mask, \
+ .shift = s, \
+ .write_mask = true,}
static const uint32_t formats_win_full[] = {
DRM_FORMAT_XRGB8888,
DRM_FORMAT_BGR565,
};
+static const struct vop_scl_regs rk3036_win_scl = {
+ .scale_yrgb_x = VOP_REG(RK3036_WIN0_SCL_FACTOR_YRGB, 0xffff, 0x0),
+ .scale_yrgb_y = VOP_REG(RK3036_WIN0_SCL_FACTOR_YRGB, 0xffff, 16),
+ .scale_cbcr_x = VOP_REG(RK3036_WIN0_SCL_FACTOR_CBR, 0xffff, 0x0),
+ .scale_cbcr_y = VOP_REG(RK3036_WIN0_SCL_FACTOR_CBR, 0xffff, 16),
+};
+
+static const struct vop_win_phy rk3036_win0_data = {
+ .scl = &rk3036_win_scl,
+ .data_formats = formats_win_full,
+ .nformats = ARRAY_SIZE(formats_win_full),
+ .enable = VOP_REG(RK3036_SYS_CTRL, 0x1, 0),
+ .format = VOP_REG(RK3036_SYS_CTRL, 0x7, 3),
+ .rb_swap = VOP_REG(RK3036_SYS_CTRL, 0x1, 15),
+ .act_info = VOP_REG(RK3036_WIN0_ACT_INFO, 0x1fff1fff, 0),
+ .dsp_info = VOP_REG(RK3036_WIN0_DSP_INFO, 0x0fff0fff, 0),
+ .dsp_st = VOP_REG(RK3036_WIN0_DSP_ST, 0x1fff1fff, 0),
+ .yrgb_mst = VOP_REG(RK3036_WIN0_YRGB_MST, 0xffffffff, 0),
+ .uv_mst = VOP_REG(RK3036_WIN0_CBR_MST, 0xffffffff, 0),
+ .yrgb_vir = VOP_REG(RK3036_WIN0_VIR, 0xffff, 0),
+ .uv_vir = VOP_REG(RK3036_WIN0_VIR, 0x1fff, 16),
+};
+
+static const struct vop_win_phy rk3036_win1_data = {
+ .data_formats = formats_win_lite,
+ .nformats = ARRAY_SIZE(formats_win_lite),
+ .enable = VOP_REG(RK3036_SYS_CTRL, 0x1, 1),
+ .format = VOP_REG(RK3036_SYS_CTRL, 0x7, 6),
+ .rb_swap = VOP_REG(RK3036_SYS_CTRL, 0x1, 19),
+ .act_info = VOP_REG(RK3036_WIN1_ACT_INFO, 0x1fff1fff, 0),
+ .dsp_info = VOP_REG(RK3036_WIN1_DSP_INFO, 0x0fff0fff, 0),
+ .dsp_st = VOP_REG(RK3036_WIN1_DSP_ST, 0x1fff1fff, 0),
+ .yrgb_mst = VOP_REG(RK3036_WIN1_MST, 0xffffffff, 0),
+ .yrgb_vir = VOP_REG(RK3036_WIN1_VIR, 0xffff, 0),
+};
+
+static const struct vop_win_data rk3036_vop_win_data[] = {
+ { .base = 0x00, .phy = &rk3036_win0_data,
+ .type = DRM_PLANE_TYPE_PRIMARY },
+ { .base = 0x00, .phy = &rk3036_win1_data,
+ .type = DRM_PLANE_TYPE_CURSOR },
+};
+
+static const int rk3036_vop_intrs[] = {
+ DSP_HOLD_VALID_INTR,
+ FS_INTR,
+ LINE_FLAG_INTR,
+ BUS_ERROR_INTR,
+};
+
+static const struct vop_intr rk3036_intr = {
+ .intrs = rk3036_vop_intrs,
+ .nintrs = ARRAY_SIZE(rk3036_vop_intrs),
+ .status = VOP_REG(RK3036_INT_STATUS, 0xf, 0),
+ .enable = VOP_REG(RK3036_INT_STATUS, 0xf, 4),
+ .clear = VOP_REG(RK3036_INT_STATUS, 0xf, 8),
+};
+
+static const struct vop_ctrl rk3036_ctrl_data = {
+ .standby = VOP_REG(RK3036_SYS_CTRL, 0x1, 30),
+ .out_mode = VOP_REG(RK3036_DSP_CTRL0, 0xf, 0),
+ .pin_pol = VOP_REG(RK3036_DSP_CTRL0, 0xf, 4),
+ .htotal_pw = VOP_REG(RK3036_DSP_HTOTAL_HS_END, 0x1fff1fff, 0),
+ .hact_st_end = VOP_REG(RK3036_DSP_HACT_ST_END, 0x1fff1fff, 0),
+ .vtotal_pw = VOP_REG(RK3036_DSP_VTOTAL_VS_END, 0x1fff1fff, 0),
+ .vact_st_end = VOP_REG(RK3036_DSP_VACT_ST_END, 0x1fff1fff, 0),
+ .line_flag_num[0] = VOP_REG(RK3036_INT_STATUS, 0xfff, 12),
+ .cfg_done = VOP_REG(RK3036_REG_CFG_DONE, 0x1, 0),
+};
+
+static const struct vop_reg_data rk3036_vop_init_reg_table[] = {
+ {RK3036_DSP_CTRL1, 0x00000000},
+};
+
+static const struct vop_data rk3036_vop = {
+ .init_table = rk3036_vop_init_reg_table,
+ .table_size = ARRAY_SIZE(rk3036_vop_init_reg_table),
+ .ctrl = &rk3036_ctrl_data,
+ .intr = &rk3036_intr,
+ .win = rk3036_vop_win_data,
+ .win_size = ARRAY_SIZE(rk3036_vop_win_data),
+};
+
static const struct vop_scl_extension rk3288_win_full_scl_ext = {
.cbcr_vsd_mode = VOP_REG(RK3288_WIN0_CTRL1, 0x1, 31),
.cbcr_vsu_mode = VOP_REG(RK3288_WIN0_CTRL1, 0x1, 30),
.vact_st_end = VOP_REG(RK3288_DSP_VACT_ST_END, 0x1fff1fff, 0),
.hpost_st_end = VOP_REG(RK3288_POST_DSP_HACT_INFO, 0x1fff1fff, 0),
.vpost_st_end = VOP_REG(RK3288_POST_DSP_VACT_INFO, 0x1fff1fff, 0),
+ .line_flag_num[0] = VOP_REG(RK3288_INTR_CTRL0, 0x1fff, 12),
.cfg_done = VOP_REG(RK3288_REG_CFG_DONE, 0x1, 0),
};
.win_size = ARRAY_SIZE(rk3288_vop_win_data),
};
-static const struct vop_scl_regs rk3036_win_scl = {
- .scale_yrgb_x = VOP_REG(RK3036_WIN0_SCL_FACTOR_YRGB, 0xffff, 0x0),
- .scale_yrgb_y = VOP_REG(RK3036_WIN0_SCL_FACTOR_YRGB, 0xffff, 16),
- .scale_cbcr_x = VOP_REG(RK3036_WIN0_SCL_FACTOR_CBR, 0xffff, 0x0),
- .scale_cbcr_y = VOP_REG(RK3036_WIN0_SCL_FACTOR_CBR, 0xffff, 16),
-};
-
-static const struct vop_win_phy rk3036_win0_data = {
- .scl = &rk3036_win_scl,
- .data_formats = formats_win_full,
- .nformats = ARRAY_SIZE(formats_win_full),
- .enable = VOP_REG(RK3036_SYS_CTRL, 0x1, 0),
- .format = VOP_REG(RK3036_SYS_CTRL, 0x7, 3),
- .rb_swap = VOP_REG(RK3036_SYS_CTRL, 0x1, 15),
- .act_info = VOP_REG(RK3036_WIN0_ACT_INFO, 0x1fff1fff, 0),
- .dsp_info = VOP_REG(RK3036_WIN0_DSP_INFO, 0x0fff0fff, 0),
- .dsp_st = VOP_REG(RK3036_WIN0_DSP_ST, 0x1fff1fff, 0),
- .yrgb_mst = VOP_REG(RK3036_WIN0_YRGB_MST, 0xffffffff, 0),
- .uv_mst = VOP_REG(RK3036_WIN0_CBR_MST, 0xffffffff, 0),
- .yrgb_vir = VOP_REG(RK3036_WIN0_VIR, 0xffff, 0),
- .uv_vir = VOP_REG(RK3036_WIN0_VIR, 0x1fff, 16),
+static const struct vop_ctrl rk3399_ctrl_data = {
+ .standby = VOP_REG(RK3399_SYS_CTRL, 0x1, 22),
+ .gate_en = VOP_REG(RK3399_SYS_CTRL, 0x1, 23),
+ .rgb_en = VOP_REG(RK3399_SYS_CTRL, 0x1, 12),
+ .hdmi_en = VOP_REG(RK3399_SYS_CTRL, 0x1, 13),
+ .edp_en = VOP_REG(RK3399_SYS_CTRL, 0x1, 14),
+ .mipi_en = VOP_REG(RK3399_SYS_CTRL, 0x1, 15),
+ .dither_down = VOP_REG(RK3399_DSP_CTRL1, 0xf, 1),
+ .dither_up = VOP_REG(RK3399_DSP_CTRL1, 0x1, 6),
+ .data_blank = VOP_REG(RK3399_DSP_CTRL0, 0x1, 19),
+ .out_mode = VOP_REG(RK3399_DSP_CTRL0, 0xf, 0),
+ .rgb_pin_pol = VOP_REG(RK3399_DSP_CTRL1, 0xf, 16),
+ .hdmi_pin_pol = VOP_REG(RK3399_DSP_CTRL1, 0xf, 20),
+ .edp_pin_pol = VOP_REG(RK3399_DSP_CTRL1, 0xf, 24),
+ .mipi_pin_pol = VOP_REG(RK3399_DSP_CTRL1, 0xf, 28),
+ .htotal_pw = VOP_REG(RK3399_DSP_HTOTAL_HS_END, 0x1fff1fff, 0),
+ .hact_st_end = VOP_REG(RK3399_DSP_HACT_ST_END, 0x1fff1fff, 0),
+ .vtotal_pw = VOP_REG(RK3399_DSP_VTOTAL_VS_END, 0x1fff1fff, 0),
+ .vact_st_end = VOP_REG(RK3399_DSP_VACT_ST_END, 0x1fff1fff, 0),
+ .hpost_st_end = VOP_REG(RK3399_POST_DSP_HACT_INFO, 0x1fff1fff, 0),
+ .vpost_st_end = VOP_REG(RK3399_POST_DSP_VACT_INFO, 0x1fff1fff, 0),
+ .line_flag_num[0] = VOP_REG(RK3399_LINE_FLAG, 0xffff, 0),
+ .line_flag_num[1] = VOP_REG(RK3399_LINE_FLAG, 0xffff, 16),
+ .cfg_done = VOP_REG_MASK(RK3399_REG_CFG_DONE, 0x1, 0),
};
-static const struct vop_win_phy rk3036_win1_data = {
- .data_formats = formats_win_lite,
- .nformats = ARRAY_SIZE(formats_win_lite),
- .enable = VOP_REG(RK3036_SYS_CTRL, 0x1, 1),
- .format = VOP_REG(RK3036_SYS_CTRL, 0x7, 6),
- .rb_swap = VOP_REG(RK3036_SYS_CTRL, 0x1, 19),
- .act_info = VOP_REG(RK3036_WIN1_ACT_INFO, 0x1fff1fff, 0),
- .dsp_info = VOP_REG(RK3036_WIN1_DSP_INFO, 0x0fff0fff, 0),
- .dsp_st = VOP_REG(RK3036_WIN1_DSP_ST, 0x1fff1fff, 0),
- .yrgb_mst = VOP_REG(RK3036_WIN1_MST, 0xffffffff, 0),
- .yrgb_vir = VOP_REG(RK3036_WIN1_VIR, 0xffff, 0),
-};
-
-static const struct vop_win_data rk3036_vop_win_data[] = {
- { .base = 0x00, .phy = &rk3036_win0_data,
- .type = DRM_PLANE_TYPE_PRIMARY },
- { .base = 0x00, .phy = &rk3036_win1_data,
- .type = DRM_PLANE_TYPE_CURSOR },
-};
-
-static const int rk3036_vop_intrs[] = {
- DSP_HOLD_VALID_INTR,
+static const int rk3399_vop_intrs[] = {
FS_INTR,
+ 0, 0,
LINE_FLAG_INTR,
+ 0,
BUS_ERROR_INTR,
+ 0, 0, 0, 0, 0, 0, 0,
+ DSP_HOLD_VALID_INTR,
};
-static const struct vop_intr rk3036_intr = {
- .intrs = rk3036_vop_intrs,
- .nintrs = ARRAY_SIZE(rk3036_vop_intrs),
- .status = VOP_REG(RK3036_INT_STATUS, 0xf, 0),
- .enable = VOP_REG(RK3036_INT_STATUS, 0xf, 4),
- .clear = VOP_REG(RK3036_INT_STATUS, 0xf, 8),
+static const struct vop_intr rk3399_vop_intr = {
+ .intrs = rk3399_vop_intrs,
+ .nintrs = ARRAY_SIZE(rk3399_vop_intrs),
+ .status = VOP_REG_MASK(RK3399_INTR_STATUS0, 0xffff, 0),
+ .enable = VOP_REG_MASK(RK3399_INTR_EN0, 0xffff, 0),
+ .clear = VOP_REG_MASK(RK3399_INTR_CLEAR0, 0xffff, 0),
};
-static const struct vop_ctrl rk3036_ctrl_data = {
- .standby = VOP_REG(RK3036_SYS_CTRL, 0x1, 30),
- .out_mode = VOP_REG(RK3036_DSP_CTRL0, 0xf, 0),
- .pin_pol = VOP_REG(RK3036_DSP_CTRL0, 0xf, 4),
- .htotal_pw = VOP_REG(RK3036_DSP_HTOTAL_HS_END, 0x1fff1fff, 0),
- .hact_st_end = VOP_REG(RK3036_DSP_HACT_ST_END, 0x1fff1fff, 0),
- .vtotal_pw = VOP_REG(RK3036_DSP_VTOTAL_VS_END, 0x1fff1fff, 0),
- .vact_st_end = VOP_REG(RK3036_DSP_VACT_ST_END, 0x1fff1fff, 0),
- .cfg_done = VOP_REG(RK3036_REG_CFG_DONE, 0x1, 0),
+static const struct vop_reg_data rk3399_init_reg_table[] = {
+ {RK3399_SYS_CTRL, 0x2000f800},
+ {RK3399_DSP_CTRL0, 0x00000000},
+ {RK3399_WIN0_CTRL0, 0x00000080},
+ {RK3399_WIN1_CTRL0, 0x00000080},
+ /* TODO: Win2/3 support multiple area function, but we haven't found
+ * a suitable way to use it yet, so let's just use them as other windows
+ * with only area 0 enabled.
+ */
+ {RK3399_WIN2_CTRL0, 0x00000010},
+ {RK3399_WIN3_CTRL0, 0x00000010},
};
-static const struct vop_reg_data rk3036_vop_init_reg_table[] = {
- {RK3036_DSP_CTRL1, 0x00000000},
+static const struct vop_data rk3399_vop_big = {
+ .init_table = rk3399_init_reg_table,
+ .table_size = ARRAY_SIZE(rk3399_init_reg_table),
+ .intr = &rk3399_vop_intr,
+ .ctrl = &rk3399_ctrl_data,
+ /*
+ * rk3399 vop big windows register layout is same as rk3288.
+ */
+ .win = rk3288_vop_win_data,
+ .win_size = ARRAY_SIZE(rk3288_vop_win_data),
};
-static const struct vop_data rk3036_vop = {
- .init_table = rk3036_vop_init_reg_table,
- .table_size = ARRAY_SIZE(rk3036_vop_init_reg_table),
- .ctrl = &rk3036_ctrl_data,
- .intr = &rk3036_intr,
- .win = rk3036_vop_win_data,
- .win_size = ARRAY_SIZE(rk3036_vop_win_data),
+static const struct vop_win_data rk3399_vop_lit_win_data[] = {
+ { .base = 0x00, .phy = &rk3288_win01_data,
+ .type = DRM_PLANE_TYPE_PRIMARY },
+ { .base = 0x00, .phy = &rk3288_win23_data,
+ .type = DRM_PLANE_TYPE_CURSOR},
+};
+
+static const struct vop_data rk3399_vop_lit = {
+ .init_table = rk3399_init_reg_table,
+ .table_size = ARRAY_SIZE(rk3399_init_reg_table),
+ .intr = &rk3399_vop_intr,
+ .ctrl = &rk3399_ctrl_data,
+ /*
+ * rk3399 vop lit windows register layout is same as rk3288,
+ * but cut off the win1 and win3 windows.
+ */
+ .win = rk3399_vop_lit_win_data,
+ .win_size = ARRAY_SIZE(rk3399_vop_lit_win_data),
};
static const struct of_device_id vop_driver_dt_match[] = {
- { .compatible = "rockchip,rk3288-vop",
- .data = &rk3288_vop },
{ .compatible = "rockchip,rk3036-vop",
.data = &rk3036_vop },
+ { .compatible = "rockchip,rk3288-vop",
+ .data = &rk3288_vop },
+ { .compatible = "rockchip,rk3399-vop-big",
+ .data = &rk3399_vop_big },
+ { .compatible = "rockchip,rk3399-vop-lit",
+ .data = &rk3399_vop_lit },
{},
};
MODULE_DEVICE_TABLE(of, vop_driver_dt_match);
#define RK3036_HWC_LUT_ADDR 0x800
/* rk3036 register definition end */
+/* rk3399 register definition */
+#define RK3399_REG_CFG_DONE 0x00000
+#define RK3399_VERSION_INFO 0x00004
+#define RK3399_SYS_CTRL 0x00008
+#define RK3399_SYS_CTRL1 0x0000c
+#define RK3399_DSP_CTRL0 0x00010
+#define RK3399_DSP_CTRL1 0x00014
+#define RK3399_DSP_BG 0x00018
+#define RK3399_MCU_CTRL 0x0001c
+#define RK3399_WB_CTRL0 0x00020
+#define RK3399_WB_CTRL1 0x00024
+#define RK3399_WB_YRGB_MST 0x00028
+#define RK3399_WB_CBR_MST 0x0002c
+#define RK3399_WIN0_CTRL0 0x00030
+#define RK3399_WIN0_CTRL1 0x00034
+#define RK3399_WIN0_COLOR_KEY 0x00038
+#define RK3399_WIN0_VIR 0x0003c
+#define RK3399_WIN0_YRGB_MST 0x00040
+#define RK3399_WIN0_CBR_MST 0x00044
+#define RK3399_WIN0_ACT_INFO 0x00048
+#define RK3399_WIN0_DSP_INFO 0x0004c
+#define RK3399_WIN0_DSP_ST 0x00050
+#define RK3399_WIN0_SCL_FACTOR_YRGB 0x00054
+#define RK3399_WIN0_SCL_FACTOR_CBR 0x00058
+#define RK3399_WIN0_SCL_OFFSET 0x0005c
+#define RK3399_WIN0_SRC_ALPHA_CTRL 0x00060
+#define RK3399_WIN0_DST_ALPHA_CTRL 0x00064
+#define RK3399_WIN0_FADING_CTRL 0x00068
+#define RK3399_WIN0_CTRL2 0x0006c
+#define RK3399_WIN1_CTRL0 0x00070
+#define RK3399_WIN1_CTRL1 0x00074
+#define RK3399_WIN1_COLOR_KEY 0x00078
+#define RK3399_WIN1_VIR 0x0007c
+#define RK3399_WIN1_YRGB_MST 0x00080
+#define RK3399_WIN1_CBR_MST 0x00084
+#define RK3399_WIN1_ACT_INFO 0x00088
+#define RK3399_WIN1_DSP_INFO 0x0008c
+#define RK3399_WIN1_DSP_ST 0x00090
+#define RK3399_WIN1_SCL_FACTOR_YRGB 0x00094
+#define RK3399_WIN1_SCL_FACTOR_CBR 0x00098
+#define RK3399_WIN1_SCL_OFFSET 0x0009c
+#define RK3399_WIN1_SRC_ALPHA_CTRL 0x000a0
+#define RK3399_WIN1_DST_ALPHA_CTRL 0x000a4
+#define RK3399_WIN1_FADING_CTRL 0x000a8
+#define RK3399_WIN1_CTRL2 0x000ac
+#define RK3399_WIN2_CTRL0 0x000b0
+#define RK3399_WIN2_CTRL1 0x000b4
+#define RK3399_WIN2_VIR0_1 0x000b8
+#define RK3399_WIN2_VIR2_3 0x000bc
+#define RK3399_WIN2_MST0 0x000c0
+#define RK3399_WIN2_DSP_INFO0 0x000c4
+#define RK3399_WIN2_DSP_ST0 0x000c8
+#define RK3399_WIN2_COLOR_KEY 0x000cc
+#define RK3399_WIN2_MST1 0x000d0
+#define RK3399_WIN2_DSP_INFO1 0x000d4
+#define RK3399_WIN2_DSP_ST1 0x000d8
+#define RK3399_WIN2_SRC_ALPHA_CTRL 0x000dc
+#define RK3399_WIN2_MST2 0x000e0
+#define RK3399_WIN2_DSP_INFO2 0x000e4
+#define RK3399_WIN2_DSP_ST2 0x000e8
+#define RK3399_WIN2_DST_ALPHA_CTRL 0x000ec
+#define RK3399_WIN2_MST3 0x000f0
+#define RK3399_WIN2_DSP_INFO3 0x000f4
+#define RK3399_WIN2_DSP_ST3 0x000f8
+#define RK3399_WIN2_FADING_CTRL 0x000fc
+#define RK3399_WIN3_CTRL0 0x00100
+#define RK3399_WIN3_CTRL1 0x00104
+#define RK3399_WIN3_VIR0_1 0x00108
+#define RK3399_WIN3_VIR2_3 0x0010c
+#define RK3399_WIN3_MST0 0x00110
+#define RK3399_WIN3_DSP_INFO0 0x00114
+#define RK3399_WIN3_DSP_ST0 0x00118
+#define RK3399_WIN3_COLOR_KEY 0x0011c
+#define RK3399_WIN3_MST1 0x00120
+#define RK3399_WIN3_DSP_INFO1 0x00124
+#define RK3399_WIN3_DSP_ST1 0x00128
+#define RK3399_WIN3_SRC_ALPHA_CTRL 0x0012c
+#define RK3399_WIN3_MST2 0x00130
+#define RK3399_WIN3_DSP_INFO2 0x00134
+#define RK3399_WIN3_DSP_ST2 0x00138
+#define RK3399_WIN3_DST_ALPHA_CTRL 0x0013c
+#define RK3399_WIN3_MST3 0x00140
+#define RK3399_WIN3_DSP_INFO3 0x00144
+#define RK3399_WIN3_DSP_ST3 0x00148
+#define RK3399_WIN3_FADING_CTRL 0x0014c
+#define RK3399_HWC_CTRL0 0x00150
+#define RK3399_HWC_CTRL1 0x00154
+#define RK3399_HWC_MST 0x00158
+#define RK3399_HWC_DSP_ST 0x0015c
+#define RK3399_HWC_SRC_ALPHA_CTRL 0x00160
+#define RK3399_HWC_DST_ALPHA_CTRL 0x00164
+#define RK3399_HWC_FADING_CTRL 0x00168
+#define RK3399_HWC_RESERVED1 0x0016c
+#define RK3399_POST_DSP_HACT_INFO 0x00170
+#define RK3399_POST_DSP_VACT_INFO 0x00174
+#define RK3399_POST_SCL_FACTOR_YRGB 0x00178
+#define RK3399_POST_RESERVED 0x0017c
+#define RK3399_POST_SCL_CTRL 0x00180
+#define RK3399_POST_DSP_VACT_INFO_F1 0x00184
+#define RK3399_DSP_HTOTAL_HS_END 0x00188
+#define RK3399_DSP_HACT_ST_END 0x0018c
+#define RK3399_DSP_VTOTAL_VS_END 0x00190
+#define RK3399_DSP_VACT_ST_END 0x00194
+#define RK3399_DSP_VS_ST_END_F1 0x00198
+#define RK3399_DSP_VACT_ST_END_F1 0x0019c
+#define RK3399_PWM_CTRL 0x001a0
+#define RK3399_PWM_PERIOD_HPR 0x001a4
+#define RK3399_PWM_DUTY_LPR 0x001a8
+#define RK3399_PWM_CNT 0x001ac
+#define RK3399_BCSH_COLOR_BAR 0x001b0
+#define RK3399_BCSH_BCS 0x001b4
+#define RK3399_BCSH_H 0x001b8
+#define RK3399_BCSH_CTRL 0x001bc
+#define RK3399_CABC_CTRL0 0x001c0
+#define RK3399_CABC_CTRL1 0x001c4
+#define RK3399_CABC_CTRL2 0x001c8
+#define RK3399_CABC_CTRL3 0x001cc
+#define RK3399_CABC_GAUSS_LINE0_0 0x001d0
+#define RK3399_CABC_GAUSS_LINE0_1 0x001d4
+#define RK3399_CABC_GAUSS_LINE1_0 0x001d8
+#define RK3399_CABC_GAUSS_LINE1_1 0x001dc
+#define RK3399_CABC_GAUSS_LINE2_0 0x001e0
+#define RK3399_CABC_GAUSS_LINE2_1 0x001e4
+#define RK3399_FRC_LOWER01_0 0x001e8
+#define RK3399_FRC_LOWER01_1 0x001ec
+#define RK3399_FRC_LOWER10_0 0x001f0
+#define RK3399_FRC_LOWER10_1 0x001f4
+#define RK3399_FRC_LOWER11_0 0x001f8
+#define RK3399_FRC_LOWER11_1 0x001fc
+#define RK3399_AFBCD0_CTRL 0x00200
+#define RK3399_AFBCD0_HDR_PTR 0x00204
+#define RK3399_AFBCD0_PIC_SIZE 0x00208
+#define RK3399_AFBCD0_STATUS 0x0020c
+#define RK3399_AFBCD1_CTRL 0x00220
+#define RK3399_AFBCD1_HDR_PTR 0x00224
+#define RK3399_AFBCD1_PIC_SIZE 0x00228
+#define RK3399_AFBCD1_STATUS 0x0022c
+#define RK3399_AFBCD2_CTRL 0x00240
+#define RK3399_AFBCD2_HDR_PTR 0x00244
+#define RK3399_AFBCD2_PIC_SIZE 0x00248
+#define RK3399_AFBCD2_STATUS 0x0024c
+#define RK3399_AFBCD3_CTRL 0x00260
+#define RK3399_AFBCD3_HDR_PTR 0x00264
+#define RK3399_AFBCD3_PIC_SIZE 0x00268
+#define RK3399_AFBCD3_STATUS 0x0026c
+#define RK3399_INTR_EN0 0x00280
+#define RK3399_INTR_CLEAR0 0x00284
+#define RK3399_INTR_STATUS0 0x00288
+#define RK3399_INTR_RAW_STATUS0 0x0028c
+#define RK3399_INTR_EN1 0x00290
+#define RK3399_INTR_CLEAR1 0x00294
+#define RK3399_INTR_STATUS1 0x00298
+#define RK3399_INTR_RAW_STATUS1 0x0029c
+#define RK3399_LINE_FLAG 0x002a0
+#define RK3399_VOP_STATUS 0x002a4
+#define RK3399_BLANKING_VALUE 0x002a8
+#define RK3399_MCU_BYPASS_PORT 0x002ac
+#define RK3399_WIN0_DSP_BG 0x002b0
+#define RK3399_WIN1_DSP_BG 0x002b4
+#define RK3399_WIN2_DSP_BG 0x002b8
+#define RK3399_WIN3_DSP_BG 0x002bc
+#define RK3399_YUV2YUV_WIN 0x002c0
+#define RK3399_YUV2YUV_POST 0x002c4
+#define RK3399_AUTO_GATING_EN 0x002cc
+#define RK3399_WIN0_CSC_COE 0x003a0
+#define RK3399_WIN1_CSC_COE 0x003c0
+#define RK3399_WIN2_CSC_COE 0x003e0
+#define RK3399_WIN3_CSC_COE 0x00400
+#define RK3399_HWC_CSC_COE 0x00420
+#define RK3399_BCSH_R2Y_CSC_COE 0x00440
+#define RK3399_BCSH_Y2R_CSC_COE 0x00460
+#define RK3399_POST_YUV2YUV_Y2R_COE 0x00480
+#define RK3399_POST_YUV2YUV_3X3_COE 0x004a0
+#define RK3399_POST_YUV2YUV_R2Y_COE 0x004c0
+#define RK3399_WIN0_YUV2YUV_Y2R 0x004e0
+#define RK3399_WIN0_YUV2YUV_3X3 0x00500
+#define RK3399_WIN0_YUV2YUV_R2Y 0x00520
+#define RK3399_WIN1_YUV2YUV_Y2R 0x00540
+#define RK3399_WIN1_YUV2YUV_3X3 0x00560
+#define RK3399_WIN1_YUV2YUV_R2Y 0x00580
+#define RK3399_WIN2_YUV2YUV_Y2R 0x005a0
+#define RK3399_WIN2_YUV2YUV_3X3 0x005c0
+#define RK3399_WIN2_YUV2YUV_R2Y 0x005e0
+#define RK3399_WIN3_YUV2YUV_Y2R 0x00600
+#define RK3399_WIN3_YUV2YUV_3X3 0x00620
+#define RK3399_WIN3_YUV2YUV_R2Y 0x00640
+#define RK3399_WIN2_LUT_ADDR 0x01000
+#define RK3399_WIN3_LUT_ADDR 0x01400
+#define RK3399_HWC_LUT_ADDR 0x01800
+#define RK3399_CABC_GAMMA_LUT_ADDR 0x01c00
+#define RK3399_GAMMA_LUT_ADDR 0x02000
+/* rk3399 register definition end */
+
#endif /* _ROCKCHIP_VOP_REG_H */
obj-$(CONFIG_DRM_SUN4I) += sun4i-drm.o sun4i-tcon.o
obj-$(CONFIG_DRM_SUN4I) += sun4i_backend.o
-
+obj-$(CONFIG_DRM_SUN4I) += sun6i_drc.o
obj-$(CONFIG_DRM_SUN4I) += sun4i_tv.o
}
EXPORT_SYMBOL(sun4i_backend_update_layer_buffer);
+static int sun4i_backend_init_sat(struct device *dev) {
+ struct sun4i_backend *backend = dev_get_drvdata(dev);
+ int ret;
+
+ backend->sat_reset = devm_reset_control_get(dev, "sat");
+ if (IS_ERR(backend->sat_reset)) {
+ dev_err(dev, "Couldn't get the SAT reset line\n");
+ return PTR_ERR(backend->sat_reset);
+ }
+
+ ret = reset_control_deassert(backend->sat_reset);
+ if (ret) {
+ dev_err(dev, "Couldn't deassert the SAT reset line\n");
+ return ret;
+ }
+
+ backend->sat_clk = devm_clk_get(dev, "sat");
+ if (IS_ERR(backend->sat_clk)) {
+ dev_err(dev, "Couldn't get our SAT clock\n");
+ ret = PTR_ERR(backend->sat_clk);
+ goto err_assert_reset;
+ }
+
+ ret = clk_prepare_enable(backend->sat_clk);
+ if (ret) {
+ dev_err(dev, "Couldn't enable the SAT clock\n");
+ return ret;
+ }
+
+ return 0;
+
+err_assert_reset:
+ reset_control_assert(backend->sat_reset);
+ return ret;
+}
+
+static int sun4i_backend_free_sat(struct device *dev) {
+ struct sun4i_backend *backend = dev_get_drvdata(dev);
+
+ clk_disable_unprepare(backend->sat_clk);
+ reset_control_assert(backend->sat_reset);
+
+ return 0;
+}
+
static struct regmap_config sun4i_backend_regmap_config = {
.reg_bits = 32,
.val_bits = 32,
}
clk_prepare_enable(backend->ram_clk);
+ if (of_device_is_compatible(dev->of_node,
+ "allwinner,sun8i-a33-display-backend")) {
+ ret = sun4i_backend_init_sat(dev);
+ if (ret) {
+ dev_err(dev, "Couldn't init SAT resources\n");
+ goto err_disable_ram_clk;
+ }
+ }
+
/* Reset the registers */
for (i = 0x800; i < 0x1000; i += 4)
regmap_write(backend->regs, i, 0);
return 0;
+err_disable_ram_clk:
+ clk_disable_unprepare(backend->ram_clk);
err_disable_mod_clk:
clk_disable_unprepare(backend->mod_clk);
err_disable_bus_clk:
{
struct sun4i_backend *backend = dev_get_drvdata(dev);
+ if (of_device_is_compatible(dev->of_node,
+ "allwinner,sun8i-a33-display-backend"))
+ sun4i_backend_free_sat(dev);
+
clk_disable_unprepare(backend->ram_clk);
clk_disable_unprepare(backend->mod_clk);
clk_disable_unprepare(backend->bus_clk);
static const struct of_device_id sun4i_backend_of_table[] = {
{ .compatible = "allwinner,sun5i-a13-display-backend" },
+ { .compatible = "allwinner,sun8i-a33-display-backend" },
{ }
};
MODULE_DEVICE_TABLE(of, sun4i_backend_of_table);
struct clk *bus_clk;
struct clk *mod_clk;
struct clk *ram_clk;
+
+ struct clk *sat_clk;
+ struct reset_control *sat_reset;
};
void sun4i_backend_apply_color_correction(struct sun4i_backend *backend);
#include <linux/regmap.h>
#include "sun4i_tcon.h"
+#include "sun4i_dotclock.h"
struct sun4i_dclk {
struct clk_hw hw;
static bool sun4i_drv_node_is_frontend(struct device_node *node)
{
- return of_device_is_compatible(node,
- "allwinner,sun5i-a13-display-frontend");
+ return of_device_is_compatible(node, "allwinner,sun5i-a13-display-frontend") ||
+ of_device_is_compatible(node, "allwinner,sun8i-a33-display-frontend");
}
static bool sun4i_drv_node_is_tcon(struct device_node *node)
{
- return of_device_is_compatible(node, "allwinner,sun5i-a13-tcon");
+ return of_device_is_compatible(node, "allwinner,sun5i-a13-tcon") ||
+ of_device_is_compatible(node, "allwinner,sun8i-a33-tcon");
}
static int compare_of(struct device *dev, void *data)
}
/*
- * If the node is our TCON, the first port is used for our
- * panel, and will not be part of the
+ * If the node is our TCON, the first port is used for
+ * panel or bridges, and will not be part of the
* component framework.
*/
if (sun4i_drv_node_is_tcon(node)) {
static const struct of_device_id sun4i_drv_of_table[] = {
{ .compatible = "allwinner,sun5i-a13-display-engine" },
+ { .compatible = "allwinner,sun8i-a33-display-engine" },
{ }
};
MODULE_DEVICE_TABLE(of, sun4i_drv_of_table);
#include <drm/drmP.h>
#include "sun4i_drv.h"
+#include "sun4i_framebuffer.h"
static void sun4i_de_output_poll_changed(struct drm_device *drm)
{
#include "sun4i_drv.h"
#include "sun4i_tcon.h"
+#include "sun4i_rgb.h"
struct sun4i_rgb {
struct drm_connector connector;
DRM_DEBUG_DRIVER("Enabling RGB output\n");
- drm_panel_enable(tcon->panel);
+ if (!IS_ERR(tcon->panel)) {
+ drm_panel_prepare(tcon->panel);
+ drm_panel_enable(tcon->panel);
+ }
+
+ /* encoder->bridge can be NULL; drm_bridge_enable checks for it */
+ drm_bridge_enable(encoder->bridge);
+
sun4i_tcon_channel_enable(tcon, 0);
}
DRM_DEBUG_DRIVER("Disabling RGB output\n");
sun4i_tcon_channel_disable(tcon, 0);
- drm_panel_disable(tcon->panel);
+
+ /* encoder->bridge can be NULL; drm_bridge_disable checks for it */
+ drm_bridge_disable(encoder->bridge);
+
+ if (!IS_ERR(tcon->panel)) {
+ drm_panel_disable(tcon->panel);
+ drm_panel_unprepare(tcon->panel);
+ }
}
static void sun4i_rgb_encoder_mode_set(struct drm_encoder *encoder,
{
struct sun4i_drv *drv = drm->dev_private;
struct sun4i_tcon *tcon = drv->tcon;
+ struct drm_encoder *encoder;
struct sun4i_rgb *rgb;
int ret;
- /* If we don't have a panel, there's no point in going on */
- if (IS_ERR(tcon->panel))
- return -ENODEV;
-
rgb = devm_kzalloc(drm->dev, sizeof(*rgb), GFP_KERNEL);
if (!rgb)
return -ENOMEM;
rgb->drv = drv;
+ encoder = &rgb->encoder;
+
+ tcon->panel = sun4i_tcon_find_panel(tcon->dev->of_node);
+ encoder->bridge = sun4i_tcon_find_bridge(tcon->dev->of_node);
+ if (IS_ERR(tcon->panel) && IS_ERR(encoder->bridge)) {
+ dev_info(drm->dev, "No panel or bridge found... RGB output disabled\n");
+ return 0;
+ }
drm_encoder_helper_add(&rgb->encoder,
&sun4i_rgb_enc_helper_funcs);
/* The RGB encoder can only work with the TCON channel 0 */
rgb->encoder.possible_crtcs = BIT(0);
- drm_connector_helper_add(&rgb->connector,
- &sun4i_rgb_con_helper_funcs);
- ret = drm_connector_init(drm, &rgb->connector,
- &sun4i_rgb_con_funcs,
- DRM_MODE_CONNECTOR_Unknown);
- if (ret) {
- dev_err(drm->dev, "Couldn't initialise the rgb connector\n");
- goto err_cleanup_connector;
+ if (!IS_ERR(tcon->panel)) {
+ drm_connector_helper_add(&rgb->connector,
+ &sun4i_rgb_con_helper_funcs);
+ ret = drm_connector_init(drm, &rgb->connector,
+ &sun4i_rgb_con_funcs,
+ DRM_MODE_CONNECTOR_Unknown);
+ if (ret) {
+ dev_err(drm->dev, "Couldn't initialise the rgb connector\n");
+ goto err_cleanup_connector;
+ }
+
+ drm_mode_connector_attach_encoder(&rgb->connector,
+ &rgb->encoder);
+
+ ret = drm_panel_attach(tcon->panel, &rgb->connector);
+ if (ret) {
+ dev_err(drm->dev, "Couldn't attach our panel\n");
+ goto err_cleanup_connector;
+ }
}
- drm_mode_connector_attach_encoder(&rgb->connector, &rgb->encoder);
+ if (!IS_ERR(encoder->bridge)) {
+ encoder->bridge->encoder = &rgb->encoder;
- drm_panel_attach(tcon->panel, &rgb->connector);
+ ret = drm_bridge_attach(drm, encoder->bridge);
+ if (ret) {
+ dev_err(drm->dev, "Couldn't attach our bridge\n");
+ goto err_cleanup_connector;
+ }
+ } else {
+ encoder->bridge = NULL;
+ }
return 0;
regmap_update_bits(tcon->regs, SUN4I_TCON0_CTL_REG,
SUN4I_TCON0_CTL_TCON_ENABLE, 0);
clk_disable_unprepare(tcon->dclk);
- } else if (channel == 1) {
- regmap_update_bits(tcon->regs, SUN4I_TCON1_CTL_REG,
- SUN4I_TCON1_CTL_TCON_ENABLE, 0);
- clk_disable_unprepare(tcon->sclk1);
+ return;
}
+
+ WARN_ON(!tcon->has_channel_1);
+ regmap_update_bits(tcon->regs, SUN4I_TCON1_CTL_REG,
+ SUN4I_TCON1_CTL_TCON_ENABLE, 0);
+ clk_disable_unprepare(tcon->sclk1);
}
EXPORT_SYMBOL(sun4i_tcon_channel_disable);
SUN4I_TCON0_CTL_TCON_ENABLE,
SUN4I_TCON0_CTL_TCON_ENABLE);
clk_prepare_enable(tcon->dclk);
- } else if (channel == 1) {
- regmap_update_bits(tcon->regs, SUN4I_TCON1_CTL_REG,
- SUN4I_TCON1_CTL_TCON_ENABLE,
- SUN4I_TCON1_CTL_TCON_ENABLE);
- clk_prepare_enable(tcon->sclk1);
+ return;
}
+
+ WARN_ON(!tcon->has_channel_1);
+ regmap_update_bits(tcon->regs, SUN4I_TCON1_CTL_REG,
+ SUN4I_TCON1_CTL_TCON_ENABLE,
+ SUN4I_TCON1_CTL_TCON_ENABLE);
+ clk_prepare_enable(tcon->sclk1);
}
EXPORT_SYMBOL(sun4i_tcon_channel_enable);
u8 clk_delay;
u32 val;
+ WARN_ON(!tcon->has_channel_1);
+
/* Adjust clock delay */
clk_delay = sun4i_tcon_get_clk_delay(mode, 1);
regmap_update_bits(tcon->regs, SUN4I_TCON1_CTL_REG,
return PTR_ERR(tcon->sclk0);
}
- tcon->sclk1 = devm_clk_get(dev, "tcon-ch1");
- if (IS_ERR(tcon->sclk1)) {
- dev_err(dev, "Couldn't get the TCON channel 1 clock\n");
- return PTR_ERR(tcon->sclk1);
+ if (tcon->has_channel_1) {
+ tcon->sclk1 = devm_clk_get(dev, "tcon-ch1");
+ if (IS_ERR(tcon->sclk1)) {
+ dev_err(dev, "Couldn't get the TCON channel 1 clock\n");
+ return PTR_ERR(tcon->sclk1);
+ }
}
return sun4i_dclk_create(dev, tcon);
res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
regs = devm_ioremap_resource(dev, res);
- if (IS_ERR(regs)) {
- dev_err(dev, "Couldn't map the TCON registers\n");
+ if (IS_ERR(regs))
return PTR_ERR(regs);
- }
tcon->regs = devm_regmap_init_mmio(dev, regs,
&sun4i_tcon_regmap_config);
return 0;
}
-static struct drm_panel *sun4i_tcon_find_panel(struct device_node *node)
+struct drm_panel *sun4i_tcon_find_panel(struct device_node *node)
{
struct device_node *port, *remote, *child;
struct device_node *end_node = NULL;
return of_drm_find_panel(remote) ?: ERR_PTR(-EPROBE_DEFER);
}
+struct drm_bridge *sun4i_tcon_find_bridge(struct device_node *node)
+{
+ struct device_node *port, *remote, *child;
+ struct device_node *end_node = NULL;
+
+ /* Inputs are listed first, then outputs */
+ port = of_graph_get_port_by_id(node, 1);
+
+ /*
+ * Our first output is the RGB interface where the panel will
+ * be connected.
+ */
+ for_each_child_of_node(port, child) {
+ u32 reg;
+
+ of_property_read_u32(child, "reg", ®);
+ if (reg == 0)
+ end_node = child;
+ }
+
+ if (!end_node) {
+ DRM_DEBUG_DRIVER("Missing bridge endpoint\n");
+ return ERR_PTR(-ENODEV);
+ }
+
+ remote = of_graph_get_remote_port_parent(end_node);
+ if (!remote) {
+ DRM_DEBUG_DRIVER("Enable to parse remote node\n");
+ return ERR_PTR(-EINVAL);
+ }
+
+ return of_drm_find_bridge(remote) ?: ERR_PTR(-EPROBE_DEFER);
+}
+
static int sun4i_tcon_bind(struct device *dev, struct device *master,
void *data)
{
dev_set_drvdata(dev, tcon);
drv->tcon = tcon;
tcon->drm = drm;
+ tcon->dev = dev;
- if (of_device_is_compatible(dev->of_node, "allwinner,sun5i-a13-tcon"))
+ if (of_device_is_compatible(dev->of_node, "allwinner,sun5i-a13-tcon")) {
tcon->has_mux = true;
+ tcon->has_channel_1 = true;
+ } else {
+ tcon->has_mux = false;
+ tcon->has_channel_1 = false;
+ }
tcon->lcd_rst = devm_reset_control_get(dev, "lcd");
if (IS_ERR(tcon->lcd_rst)) {
goto err_free_clocks;
}
- tcon->panel = sun4i_tcon_find_panel(dev->of_node);
- if (IS_ERR(tcon->panel)) {
- dev_info(dev, "No panel found... RGB output disabled\n");
- return 0;
- }
-
ret = sun4i_rgb_init(drm);
if (ret < 0)
goto err_free_clocks;
static int sun4i_tcon_probe(struct platform_device *pdev)
{
struct device_node *node = pdev->dev.of_node;
+ struct drm_bridge *bridge;
struct drm_panel *panel;
/*
- * The panel is not ready.
+ * Neither the bridge or the panel is ready.
* Defer the probe.
*/
panel = sun4i_tcon_find_panel(node);
+ bridge = sun4i_tcon_find_bridge(node);
/*
* If we don't have a panel endpoint, just go on
*/
- if (PTR_ERR(panel) == -EPROBE_DEFER) {
- DRM_DEBUG_DRIVER("Still waiting for our panel. Deferring...\n");
+ if ((PTR_ERR(panel) == -EPROBE_DEFER) &&
+ (PTR_ERR(bridge) == -EPROBE_DEFER)) {
+ DRM_DEBUG_DRIVER("Still waiting for our panel/bridge. Deferring...\n");
return -EPROBE_DEFER;
}
static const struct of_device_id sun4i_tcon_of_table[] = {
{ .compatible = "allwinner,sun5i-a13-tcon" },
+ { .compatible = "allwinner,sun8i-a33-tcon" },
{ }
};
MODULE_DEVICE_TABLE(of, sun4i_tcon_of_table);
#define SUN4I_TCON_MAX_CHANNELS 2
struct sun4i_tcon {
+ struct device *dev;
struct drm_device *drm;
struct regmap *regs;
bool has_mux;
struct drm_panel *panel;
+
+ bool has_channel_1;
};
+struct drm_bridge *sun4i_tcon_find_bridge(struct device_node *node);
+struct drm_panel *sun4i_tcon_find_panel(struct device_node *node);
+
/* Global Control */
void sun4i_tcon_disable(struct sun4i_tcon *tcon);
void sun4i_tcon_enable(struct sun4i_tcon *tcon);
--- /dev/null
+/*
+ * Copyright (C) 2016 Free Electrons
+ *
+ * Maxime Ripard <maxime.ripard@free-electrons.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ */
+
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/reset.h>
+
+struct sun6i_drc {
+ struct clk *bus_clk;
+ struct clk *mod_clk;
+ struct reset_control *reset;
+};
+
+static int sun6i_drc_bind(struct device *dev, struct device *master,
+ void *data)
+{
+ struct sun6i_drc *drc;
+ int ret;
+
+ drc = devm_kzalloc(dev, sizeof(*drc), GFP_KERNEL);
+ if (!drc)
+ return -ENOMEM;
+ dev_set_drvdata(dev, drc);
+
+ drc->reset = devm_reset_control_get(dev, NULL);
+ if (IS_ERR(drc->reset)) {
+ dev_err(dev, "Couldn't get our reset line\n");
+ return PTR_ERR(drc->reset);
+ }
+
+ ret = reset_control_deassert(drc->reset);
+ if (ret) {
+ dev_err(dev, "Couldn't deassert our reset line\n");
+ return ret;
+ }
+
+ drc->bus_clk = devm_clk_get(dev, "ahb");
+ if (IS_ERR(drc->bus_clk)) {
+ dev_err(dev, "Couldn't get our bus clock\n");
+ ret = PTR_ERR(drc->bus_clk);
+ goto err_assert_reset;
+ }
+ clk_prepare_enable(drc->bus_clk);
+
+ drc->mod_clk = devm_clk_get(dev, "mod");
+ if (IS_ERR(drc->mod_clk)) {
+ dev_err(dev, "Couldn't get our mod clock\n");
+ ret = PTR_ERR(drc->mod_clk);
+ goto err_disable_bus_clk;
+ }
+ clk_prepare_enable(drc->mod_clk);
+
+ return 0;
+
+err_disable_bus_clk:
+ clk_disable_unprepare(drc->bus_clk);
+err_assert_reset:
+ reset_control_assert(drc->reset);
+ return ret;
+}
+
+static void sun6i_drc_unbind(struct device *dev, struct device *master,
+ void *data)
+{
+ struct sun6i_drc *drc = dev_get_drvdata(dev);
+
+ clk_disable_unprepare(drc->mod_clk);
+ clk_disable_unprepare(drc->bus_clk);
+ reset_control_assert(drc->reset);
+}
+
+static struct component_ops sun6i_drc_ops = {
+ .bind = sun6i_drc_bind,
+ .unbind = sun6i_drc_unbind,
+};
+
+static int sun6i_drc_probe(struct platform_device *pdev)
+{
+ return component_add(&pdev->dev, &sun6i_drc_ops);
+}
+
+static int sun6i_drc_remove(struct platform_device *pdev)
+{
+ component_del(&pdev->dev, &sun6i_drc_ops);
+
+ return 0;
+}
+
+static const struct of_device_id sun6i_drc_of_table[] = {
+ { .compatible = "allwinner,sun8i-a33-drc" },
+ { }
+};
+MODULE_DEVICE_TABLE(of, sun6i_drc_of_table);
+
+static struct platform_driver sun6i_drc_platform_driver = {
+ .probe = sun6i_drc_probe,
+ .remove = sun6i_drc_remove,
+ .driver = {
+ .name = "sun6i-drc",
+ .of_match_table = sun6i_drc_of_table,
+ },
+};
+module_platform_driver(sun6i_drc_platform_driver);
+
+MODULE_AUTHOR("Maxime Ripard <maxime.ripard@free-electrons.com>");
+MODULE_DESCRIPTION("Allwinner A31 Dynamic Range Control (DRC) Driver");
+MODULE_LICENSE("GPL");
struct drm_gem_cma_object *gem;
unsigned int depth, bpp;
dma_addr_t start, end;
+ u64 dma_base_and_ceiling;
drm_fb_get_bpp_depth(fb->pixel_format, &depth, &bpp);
gem = drm_fb_cma_get_gem_obj(fb, 0);
end = start + (crtc->mode.vdisplay * fb->pitches[0]);
- tilcdc_write(dev, LCDC_DMA_FB_BASE_ADDR_0_REG, start);
- tilcdc_write(dev, LCDC_DMA_FB_CEILING_ADDR_0_REG, end);
+ /* Write LCDC_DMA_FB_BASE_ADDR_0_REG and LCDC_DMA_FB_CEILING_ADDR_0_REG
+ * with a single insruction, if available. This should make it more
+ * unlikely that LCDC would fetch the DMA addresses in the middle of
+ * an update.
+ */
+ dma_base_and_ceiling = (u64)(end - 1) << 32 | start;
+ tilcdc_write64(dev, LCDC_DMA_FB_BASE_ADDR_0_REG, dma_base_and_ceiling);
if (tilcdc_crtc->curr_fb)
drm_flip_work_queue(&tilcdc_crtc->unref_work,
if (priv->rev == 1) {
tilcdc_set(dev, LCDC_RASTER_CTRL_REG,
LCDC_V1_UNDERFLOW_INT_ENA);
+ tilcdc_set(dev, LCDC_DMA_CTRL_REG,
+ LCDC_V1_END_OF_FRAME_INT_ENA);
} else {
tilcdc_write(dev, LCDC_INT_ENABLE_SET_REG,
LCDC_V2_UNDERFLOW_INT_ENA |
static LIST_HEAD(module_list);
+static const u32 tilcdc_rev1_formats[] = { DRM_FORMAT_RGB565 };
+
+static const u32 tilcdc_straight_formats[] = { DRM_FORMAT_RGB565,
+ DRM_FORMAT_BGR888,
+ DRM_FORMAT_XBGR8888 };
+
+static const u32 tilcdc_crossed_formats[] = { DRM_FORMAT_BGR565,
+ DRM_FORMAT_RGB888,
+ DRM_FORMAT_XRGB8888 };
+
+static const u32 tilcdc_legacy_formats[] = { DRM_FORMAT_RGB565,
+ DRM_FORMAT_RGB888,
+ DRM_FORMAT_XRGB8888 };
+
void tilcdc_module_init(struct tilcdc_module *mod, const char *name,
const struct tilcdc_module_ops *funcs)
{
struct platform_device *pdev = dev->platformdev;
struct device_node *node = pdev->dev.of_node;
struct tilcdc_drm_private *priv;
- struct tilcdc_module *mod;
struct resource *res;
u32 bpp = 0;
int ret;
pm_runtime_put_sync(dev->dev);
+ if (priv->rev == 1) {
+ DBG("Revision 1 LCDC supports only RGB565 format");
+ priv->pixelformats = tilcdc_rev1_formats;
+ priv->num_pixelformats = ARRAY_SIZE(tilcdc_rev1_formats);
+ bpp = 16;
+ } else {
+ const char *str = "\0";
+
+ of_property_read_string(node, "blue-and-red-wiring", &str);
+ if (0 == strcmp(str, "crossed")) {
+ DBG("Configured for crossed blue and red wires");
+ priv->pixelformats = tilcdc_crossed_formats;
+ priv->num_pixelformats =
+ ARRAY_SIZE(tilcdc_crossed_formats);
+ bpp = 32; /* Choose bpp with RGB support for fbdef */
+ } else if (0 == strcmp(str, "straight")) {
+ DBG("Configured for straight blue and red wires");
+ priv->pixelformats = tilcdc_straight_formats;
+ priv->num_pixelformats =
+ ARRAY_SIZE(tilcdc_straight_formats);
+ bpp = 16; /* Choose bpp with RGB support for fbdef */
+ } else {
+ DBG("Blue and red wiring '%s' unknown, use legacy mode",
+ str);
+ priv->pixelformats = tilcdc_legacy_formats;
+ priv->num_pixelformats =
+ ARRAY_SIZE(tilcdc_legacy_formats);
+ bpp = 16; /* This is just a guess */
+ }
+ }
+
ret = modeset_init(dev);
if (ret < 0) {
dev_err(dev->dev, "failed to initialize mode setting\n");
if (ret < 0)
goto fail_mode_config_cleanup;
- ret = tilcdc_add_external_encoders(dev, &bpp);
+ ret = tilcdc_add_external_encoders(dev);
if (ret < 0)
goto fail_component_cleanup;
}
goto fail_vblank_cleanup;
}
- list_for_each_entry(mod, &module_list, list) {
- DBG("%s: preferred_bpp: %d", mod->name, mod->preferred_bpp);
- bpp = mod->preferred_bpp;
- if (bpp > 0)
- break;
- }
-
- drm_helper_disable_unused_functions(dev);
-
drm_mode_config_reset(dev);
priv->fbdev = drm_fbdev_cma_init(dev, bpp,
*/
uint32_t max_width;
+ /* Supported pixel formats */
+ const uint32_t *pixelformats;
+ uint32_t num_pixelformats;
+
/* The context for pm susped/resume cycle is stored here */
struct drm_atomic_state *saved_state;
const char *name;
struct list_head list;
const struct tilcdc_module_ops *funcs;
- unsigned int preferred_bpp;
};
void tilcdc_module_init(struct tilcdc_module *mod, const char *name,
return MODE_OK;
}
-static int tilcdc_add_external_encoder(struct drm_device *dev, int *bpp,
+static int tilcdc_add_external_encoder(struct drm_device *dev,
struct drm_connector *connector)
{
struct tilcdc_drm_private *priv = dev->dev_private;
/* Only tda998x is supported at the moment. */
tilcdc_crtc_set_simulate_vesa_sync(priv->crtc, true);
tilcdc_crtc_set_panel_info(priv->crtc, &panel_info_tda998x);
- *bpp = panel_info_tda998x.bpp;
connector_funcs = devm_kzalloc(dev->dev, sizeof(*connector_funcs),
GFP_KERNEL);
return 0;
}
-int tilcdc_add_external_encoders(struct drm_device *dev, int *bpp)
+int tilcdc_add_external_encoders(struct drm_device *dev)
{
struct tilcdc_drm_private *priv = dev->dev_private;
struct drm_connector *connector;
if (connector == priv->connectors[i])
found = true;
if (!found) {
- ret = tilcdc_add_external_encoder(dev, bpp, connector);
+ ret = tilcdc_add_external_encoder(dev, connector);
if (ret)
return ret;
}
while ((ep = of_graph_get_next_endpoint(dev->of_node, ep))) {
node = of_graph_get_remote_port_parent(ep);
- if (!node && !of_device_is_available(node)) {
+ if (!node || !of_device_is_available(node)) {
of_node_put(node);
continue;
}
#ifndef __TILCDC_EXTERNAL_H__
#define __TILCDC_EXTERNAL_H__
-int tilcdc_add_external_encoders(struct drm_device *dev, int *bpp);
+int tilcdc_add_external_encoders(struct drm_device *dev);
void tilcdc_remove_external_encoders(struct drm_device *dev);
int tilcdc_get_external_components(struct device *dev,
struct component_match **match);
goto fail_timings;
}
- mod->preferred_bpp = panel_mod->info->bpp;
-
return 0;
fail_timings:
#include "tilcdc_drv.h"
-static const u32 tilcdc_formats[] = { DRM_FORMAT_RGB565,
- DRM_FORMAT_RGB888,
- DRM_FORMAT_XRGB8888 };
-
static struct drm_plane_funcs tilcdc_plane_funcs = {
.update_plane = drm_atomic_helper_update_plane,
.disable_plane = drm_atomic_helper_disable_plane,
int tilcdc_plane_init(struct drm_device *dev,
struct drm_plane *plane)
{
+ struct tilcdc_drm_private *priv = dev->dev_private;
int ret;
ret = drm_plane_init(dev, plane, 1,
&tilcdc_plane_funcs,
- tilcdc_formats,
- ARRAY_SIZE(tilcdc_formats),
+ priv->pixelformats,
+ priv->num_pixelformats,
true);
if (ret) {
dev_err(dev->dev, "Failed to initialize plane: %d\n", ret);
iowrite32(data, priv->mmio + reg);
}
+static inline void tilcdc_write64(struct drm_device *dev, u32 reg, u64 data)
+{
+ struct tilcdc_drm_private *priv = dev->dev_private;
+ volatile void __iomem *addr = priv->mmio + reg;
+
+#ifdef iowrite64
+ iowrite64(data, addr);
+#else
+ __iowmb();
+ /* This compiles to strd (=64-bit write) on ARM7 */
+ *(volatile u64 __force *)addr = __cpu_to_le64(data);
+#endif
+}
+
static inline u32 tilcdc_read(struct drm_device *dev, u32 reg)
{
struct tilcdc_drm_private *priv = dev->dev_private;
goto fail;
}
- mod->preferred_bpp = dvi_info.bpp;
-
i2c_node = of_find_node_by_phandle(i2c_phandle);
if (!i2c_node) {
dev_err(&pdev->dev, "could not get i2c bus node\n");
static inline int ttm_mem_type_from_place(const struct ttm_place *place,
uint32_t *mem_type)
{
- int i;
+ int pos;
- for (i = 0; i <= TTM_PL_PRIV5; i++)
- if (place->flags & (1 << i)) {
- *mem_type = i;
- return 0;
- }
- return -EINVAL;
+ pos = ffs(place->flags & TTM_PL_MASK_MEM);
+ if (unlikely(!pos))
+ return -EINVAL;
+
+ *mem_type = pos - 1;
+ return 0;
}
static void ttm_mem_type_debug(struct ttm_bo_device *bdev, int mem_type)
if (!(old_man->flags & TTM_MEMTYPE_FLAG_FIXED) &&
!(new_man->flags & TTM_MEMTYPE_FLAG_FIXED))
- ret = ttm_bo_move_ttm(bo, evict, interruptible, no_wait_gpu,
- mem);
+ ret = ttm_bo_move_ttm(bo, interruptible, no_wait_gpu, mem);
else if (bdev->driver->move)
ret = bdev->driver->move(bo, evict, interruptible,
no_wait_gpu, mem);
else
- ret = ttm_bo_move_memcpy(bo, evict, interruptible,
- no_wait_gpu, mem);
+ ret = ttm_bo_move_memcpy(bo, interruptible, no_wait_gpu, mem);
if (ret) {
if (bdev->driver->move_notify) {
}
int ttm_bo_move_ttm(struct ttm_buffer_object *bo,
- bool evict, bool interruptible,
- bool no_wait_gpu, struct ttm_mem_reg *new_mem)
+ bool interruptible, bool no_wait_gpu,
+ struct ttm_mem_reg *new_mem)
{
struct ttm_tt *ttm = bo->ttm;
struct ttm_mem_reg *old_mem = &bo->mem;
}
int ttm_bo_move_memcpy(struct ttm_buffer_object *bo,
- bool evict, bool interruptible,
- bool no_wait_gpu,
+ bool interruptible, bool no_wait_gpu,
struct ttm_mem_reg *new_mem)
{
struct ttm_bo_device *bdev = bo->bdev;
return 0;
}
EXPORT_SYMBOL(ttm_round_pot);
+
+uint64_t ttm_get_kernel_zone_memory_size(struct ttm_mem_global *glob)
+{
+ return glob->zone_kernel->max_mem;
+}
+EXPORT_SYMBOL(ttm_get_kernel_zone_memory_size);
if (count) {
d_page = list_first_entry(&pool->free_list, struct dma_page, page_list);
ttm->pages[index] = d_page->p;
- ttm_dma->cpu_address[index] = d_page->vaddr;
ttm_dma->dma_address[index] = d_page->dma;
list_move_tail(&d_page->page_list, &ttm_dma->pages_list);
r = 0;
INIT_LIST_HEAD(&ttm_dma->pages_list);
for (i = 0; i < ttm->num_pages; i++) {
ttm->pages[i] = NULL;
- ttm_dma->cpu_address[i] = 0;
ttm_dma->dma_address[i] = 0;
}
{
ttm->ttm.pages = drm_calloc_large(ttm->ttm.num_pages,
sizeof(*ttm->ttm.pages) +
- sizeof(*ttm->dma_address) +
- sizeof(*ttm->cpu_address));
- ttm->cpu_address = (void *) (ttm->ttm.pages + ttm->ttm.num_pages);
- ttm->dma_address = (void *) (ttm->cpu_address + ttm->ttm.num_pages);
+ sizeof(*ttm->dma_address));
+ ttm->dma_address = (void *) (ttm->ttm.pages + ttm->ttm.num_pages);
}
#ifdef CONFIG_X86
drm_free_large(ttm->pages);
ttm->pages = NULL;
- ttm_dma->cpu_address = NULL;
ttm_dma->dma_address = NULL;
}
EXPORT_SYMBOL(ttm_dma_tt_fini);
return 0;
}
+static int udl_usb_suspend(struct usb_interface *interface,
+ pm_message_t message)
+{
+ return 0;
+}
+
+static int udl_usb_resume(struct usb_interface *interface)
+{
+ struct drm_device *dev = usb_get_intfdata(interface);
+
+ udl_modeset_restore(dev);
+ return 0;
+}
+
static const struct vm_operations_struct udl_gem_vm_ops = {
.fault = udl_gem_fault,
.open = drm_gem_vm_open,
.name = "udl",
.probe = udl_usb_probe,
.disconnect = udl_usb_disconnect,
+ .suspend = udl_usb_suspend,
+ .resume = udl_usb_resume,
.id_table = id_table,
};
module_usb_driver(udl_driver);
struct device *dev;
struct drm_device *ddev;
struct usb_device *udev;
+ struct drm_crtc *crtc;
int sku_pixel_limit;
/* modeset */
int udl_modeset_init(struct drm_device *dev);
+void udl_modeset_restore(struct drm_device *dev);
void udl_modeset_cleanup(struct drm_device *dev);
int udl_connector_init(struct drm_device *dev, struct drm_encoder *encoder);
char *wrptr;
int color_depth = 0;
+ udl->crtc = crtc;
+
buf = (char *)udl->mode_buf;
/* for now we just clip 24 -> 16 - if we fix that fix this */
return 0;
}
+void udl_modeset_restore(struct drm_device *dev)
+{
+ struct udl_device *udl = dev->dev_private;
+ struct udl_framebuffer *ufb;
+
+ if (!udl->crtc || !udl->crtc->primary->fb)
+ return;
+ udl_crtc_commit(udl->crtc);
+ ufb = to_udl_fb(udl->crtc->primary->fb);
+ udl_handle_damage(ufb, 0, 0, ufb->base.width, ufb->base.height);
+}
+
void udl_modeset_cleanup(struct drm_device *dev)
{
drm_mode_config_cleanup(dev);
int vblank_lines;
int ret = 0;
- /*
- * XXX Doesn't work well in interlaced mode yet, partially due
- * to problems in vc4 kms or drm core interlaced mode handling,
- * so disable for now in interlaced mode.
- */
- if (mode->flags & DRM_MODE_FLAG_INTERLACE)
- return ret;
-
/* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */
/* Get optional system timestamp before query. */
/* Vertical position of hvs composed scanline. */
*vpos = VC4_GET_FIELD(val, SCALER_DISPSTATX_LINE);
+ *hpos = 0;
- /* No hpos info available. */
- if (hpos)
- *hpos = 0;
+ if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
+ *vpos /= 2;
+
+ /* Use hpos to correct for field offset in interlaced mode. */
+ if (VC4_GET_FIELD(val, SCALER_DISPSTATX_FRAME_COUNT) % 2)
+ *hpos += mode->crtc_htotal / 2;
+ }
/* This is the offset we need for translating hvs -> pv scanout pos. */
fifo_lines = vc4_crtc->cob_size / mode->crtc_hdisplay;
* position of the PV.
*/
*vpos -= fifo_lines + 1;
- if (mode->flags & DRM_MODE_FLAG_INTERLACE)
- *vpos /= 2;
ret |= DRM_SCANOUTPOS_ACCURATE;
return ret;
int ret;
require_hvs_enabled(dev);
+ /* Disable vblank irq handling before crtc is disabled. */
+ drm_crtc_vblank_off(crtc);
+
CRTC_WRITE(PV_V_CONTROL,
CRTC_READ(PV_V_CONTROL) & ~PV_VCONTROL_VIDEN);
ret = wait_for(!(CRTC_READ(PV_V_CONTROL) & PV_VCONTROL_VIDEN), 1);
/* Turn on the pixel valve, which will emit the vstart signal. */
CRTC_WRITE(PV_V_CONTROL,
CRTC_READ(PV_V_CONTROL) | PV_VCONTROL_VIDEN);
+
+ /* Enable vblank irq handling after crtc is started. */
+ drm_crtc_vblank_on(crtc);
+}
+
+static bool vc4_crtc_mode_fixup(struct drm_crtc *crtc,
+ const struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+ /* Do not allow doublescan modes from user space */
+ if (adjusted_mode->flags & DRM_MODE_FLAG_DBLSCAN) {
+ DRM_DEBUG_KMS("[CRTC:%d] Doublescan mode rejected.\n",
+ crtc->base.id);
+ return false;
+ }
+
+ /*
+ * Interlaced video modes got CRTC_INTERLACE_HALVE_V applied when
+ * coming from user space. We don't want this, as it screws up
+ * vblank timestamping, so fix it up.
+ */
+ drm_mode_set_crtcinfo(adjusted_mode, 0);
+
+ DRM_DEBUG_KMS("[CRTC:%d] adjusted_mode :\n", crtc->base.id);
+ drm_mode_debug_printmodeline(adjusted_mode);
+
+ return true;
}
static int vc4_crtc_atomic_check(struct drm_crtc *crtc,
.mode_set_nofb = vc4_crtc_mode_set_nofb,
.disable = vc4_crtc_disable,
.enable = vc4_crtc_enable,
+ .mode_fixup = vc4_crtc_mode_fixup,
.atomic_check = vc4_crtc_atomic_check,
.atomic_flush = vc4_crtc_atomic_flush,
};
}
}
+static bool vc4_dpi_encoder_mode_fixup(struct drm_encoder *encoder,
+ const struct drm_display_mode *mode,
+ struct drm_display_mode *adjusted_mode)
+{
+ if (adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE)
+ return false;
+
+ return true;
+}
+
static const struct drm_encoder_helper_funcs vc4_dpi_encoder_helper_funcs = {
.disable = vc4_dpi_encoder_disable,
.enable = vc4_dpi_encoder_enable,
+ .mode_fixup = vc4_dpi_encoder_mode_fixup,
};
static const struct of_device_id vc4_dpi_dt_match[] = {
vc4_flush_caches(dev);
- /* Disable the binner's pre-loaded overflow memory address */
- V3D_WRITE(V3D_BPOA, 0);
- V3D_WRITE(V3D_BPOS, 0);
-
/* Either put the job in the binner if it uses the binner, or
* immediately move it to the to-be-rendered queue.
*/
return ret;
}
+/*
+ * drm_helper_probe_single_connector_modes() applies drm_mode_set_crtcinfo to
+ * all modes with flag CRTC_INTERLACE_HALVE_V. We don't want this, as it
+ * screws up vblank timestamping for interlaced modes, so fix it up.
+ */
+static int vc4_hdmi_connector_probe_modes(struct drm_connector *connector,
+ uint32_t maxX, uint32_t maxY)
+{
+ struct drm_display_mode *mode;
+ int count;
+
+ count = drm_helper_probe_single_connector_modes(connector, maxX, maxY);
+ if (count == 0)
+ return 0;
+
+ DRM_DEBUG_KMS("[CONNECTOR:%d:%s] probed adapted modes :\n",
+ connector->base.id, connector->name);
+ list_for_each_entry(mode, &connector->modes, head) {
+ drm_mode_set_crtcinfo(mode, 0);
+ drm_mode_debug_printmodeline(mode);
+ }
+
+ return count;
+}
+
static const struct drm_connector_funcs vc4_hdmi_connector_funcs = {
.dpms = drm_atomic_helper_connector_dpms,
.detect = vc4_hdmi_connector_detect,
- .fill_modes = drm_helper_probe_single_connector_modes,
+ .fill_modes = vc4_hdmi_connector_probe_modes,
.destroy = vc4_hdmi_connector_destroy,
.reset = drm_atomic_helper_connector_reset,
.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
connector->polled = (DRM_CONNECTOR_POLL_CONNECT |
DRM_CONNECTOR_POLL_DISCONNECT);
- connector->interlace_allowed = 0;
+ connector->interlace_allowed = 1;
connector->doublescan_allowed = 0;
drm_mode_connector_attach_encoder(connector, encoder);
VMWGFX_NUM_GB_SURFACE +\
VMWGFX_NUM_GB_SCREEN_TARGET)
-#define VMW_PL_GMR TTM_PL_PRIV0
-#define VMW_PL_FLAG_GMR TTM_PL_FLAG_PRIV0
-#define VMW_PL_MOB TTM_PL_PRIV1
-#define VMW_PL_FLAG_MOB TTM_PL_FLAG_PRIV1
+#define VMW_PL_GMR (TTM_PL_PRIV + 0)
+#define VMW_PL_FLAG_GMR (TTM_PL_FLAG_PRIV << 0)
+#define VMW_PL_MOB (TTM_PL_PRIV + 1)
+#define VMW_PL_FLAG_MOB (TTM_PL_FLAG_PRIV << 1)
#define VMW_RES_CONTEXT ttm_driver_type0
#define VMW_RES_SURFACE ttm_driver_type1
struct drm_connector *);
};
+int analogix_dp_enable_psr(struct device *dev);
+int analogix_dp_disable_psr(struct device *dev);
+
int analogix_dp_resume(struct device *dev);
int analogix_dp_suspend(struct device *dev);
/** \name Macros to make printk easier */
/*@{*/
+#define _DRM_PRINTK(once, level, fmt, ...) \
+ do { \
+ printk##once(KERN_##level "[" DRM_NAME "] " fmt, \
+ ##__VA_ARGS__); \
+ } while (0)
+
+#define DRM_INFO(fmt, ...) \
+ _DRM_PRINTK(, INFO, fmt, ##__VA_ARGS__)
+#define DRM_NOTE(fmt, ...) \
+ _DRM_PRINTK(, NOTICE, fmt, ##__VA_ARGS__)
+#define DRM_WARN(fmt, ...) \
+ _DRM_PRINTK(, WARNING, fmt, ##__VA_ARGS__)
+
+#define DRM_INFO_ONCE(fmt, ...) \
+ _DRM_PRINTK(_once, INFO, fmt, ##__VA_ARGS__)
+#define DRM_NOTE_ONCE(fmt, ...) \
+ _DRM_PRINTK(_once, NOTICE, fmt, ##__VA_ARGS__)
+#define DRM_WARN_ONCE(fmt, ...) \
+ _DRM_PRINTK(_once, WARNING, fmt, ##__VA_ARGS__)
+
/**
* Error output.
*
#define DRM_DEV_INFO(dev, fmt, ...) \
drm_dev_printk(dev, KERN_INFO, DRM_UT_NONE, __func__, "", fmt, \
##__VA_ARGS__)
-#define DRM_INFO(fmt, ...) \
- drm_printk(KERN_INFO, DRM_UT_NONE, __func__, "", fmt, ##__VA_ARGS__)
#define DRM_DEV_INFO_ONCE(dev, fmt, ...) \
({ \
DRM_DEV_INFO(dev, fmt, ##__VA_ARGS__); \
} \
})
-#define DRM_INFO_ONCE(fmt, ...) DRM_DEV_INFO_ONCE(NULL, fmt, ##__VA_ARGS__)
/**
* Debug output.
struct drm_atomic_state *state,
bool nonblock);
-void drm_atomic_helper_wait_for_fences(struct drm_device *dev,
- struct drm_atomic_state *state);
+int drm_atomic_helper_wait_for_fences(struct drm_device *dev,
+ struct drm_atomic_state *state,
+ bool pre_swap);
bool drm_atomic_helper_framebuffer_changed(struct drm_device *dev,
struct drm_atomic_state *old_state,
struct drm_crtc *crtc);
#include <linux/list.h>
#include <linux/ctype.h>
-#include <drm/drm_modeset.h>
+#include <drm/drm_mode_object.h>
struct drm_connector_helper_funcs;
struct drm_device;
#include <uapi/drm/drm_fourcc.h>
#include <drm/drm_modeset_lock.h>
#include <drm/drm_rect.h>
-#include <drm/drm_modeset.h>
+#include <drm/drm_mode_object.h>
#include <drm/drm_framebuffer.h>
#include <drm/drm_modes.h>
#include <drm/drm_connector.h>
+#include <drm/drm_encoder.h>
+#include <drm/drm_property.h>
struct drm_device;
struct drm_mode_set;
u8 group_data[8];
};
-struct drm_property_blob {
- struct drm_mode_object base;
- struct drm_device *dev;
- struct list_head head_global;
- struct list_head head_file;
- size_t length;
- unsigned char data[];
-};
-
-struct drm_property_enum {
- uint64_t value;
- struct list_head head;
- char name[DRM_PROP_NAME_LEN];
-};
-
-struct drm_property {
- struct list_head head;
- struct drm_mode_object base;
- uint32_t flags;
- char name[DRM_PROP_NAME_LEN];
- uint32_t num_values;
- uint64_t *values;
- struct drm_device *dev;
-
- struct list_head enum_list;
-};
-
struct drm_crtc;
struct drm_encoder;
struct drm_pending_vblank_event;
struct drm_pending_vblank_event *event,
uint32_t flags);
+ /**
+ * @page_flip_target:
+ *
+ * Same as @page_flip but with an additional parameter specifying the
+ * absolute target vertical blank period (as reported by
+ * drm_crtc_vblank_count()) when the flip should take effect.
+ *
+ * Note that the core code calls drm_crtc_vblank_get before this entry
+ * point, and will call drm_crtc_vblank_put if this entry point returns
+ * any non-0 error code. It's the driver's responsibility to call
+ * drm_crtc_vblank_put after this entry point returns 0, typically when
+ * the flip completes.
+ */
+ int (*page_flip_target)(struct drm_crtc *crtc,
+ struct drm_framebuffer *fb,
+ struct drm_pending_vblank_event *event,
+ uint32_t flags, uint32_t target);
+
/**
* @set_property:
*
struct drm_modeset_acquire_ctx *acquire_ctx;
};
-/**
- * struct drm_encoder_funcs - encoder controls
- *
- * Encoders sit between CRTCs and connectors.
- */
-struct drm_encoder_funcs {
- /**
- * @reset:
- *
- * Reset encoder hardware and software state to off. This function isn't
- * called by the core directly, only through drm_mode_config_reset().
- * It's not a helper hook only for historical reasons.
- */
- void (*reset)(struct drm_encoder *encoder);
-
- /**
- * @destroy:
- *
- * Clean up encoder resources. This is only called at driver unload time
- * through drm_mode_config_cleanup() since an encoder cannot be
- * hotplugged in DRM.
- */
- void (*destroy)(struct drm_encoder *encoder);
-
- /**
- * @late_register:
- *
- * This optional hook can be used to register additional userspace
- * interfaces attached to the encoder like debugfs interfaces.
- * It is called late in the driver load sequence from drm_dev_register().
- * Everything added from this callback should be unregistered in
- * the early_unregister callback.
- *
- * Returns:
- *
- * 0 on success, or a negative error code on failure.
- */
- int (*late_register)(struct drm_encoder *encoder);
-
- /**
- * @early_unregister:
- *
- * This optional hook should be used to unregister the additional
- * userspace interfaces attached to the encoder from
- * late_unregister(). It is called from drm_dev_unregister(),
- * early in the driver unload sequence to disable userspace access
- * before data structures are torndown.
- */
- void (*early_unregister)(struct drm_encoder *encoder);
-};
-
-/**
- * struct drm_encoder - central DRM encoder structure
- * @dev: parent DRM device
- * @head: list management
- * @base: base KMS object
- * @name: human readable name, can be overwritten by the driver
- * @encoder_type: one of the DRM_MODE_ENCODER_<foo> types in drm_mode.h
- * @possible_crtcs: bitmask of potential CRTC bindings
- * @possible_clones: bitmask of potential sibling encoders for cloning
- * @crtc: currently bound CRTC
- * @bridge: bridge associated to the encoder
- * @funcs: control functions
- * @helper_private: mid-layer private data
- *
- * CRTCs drive pixels to encoders, which convert them into signals
- * appropriate for a given connector or set of connectors.
- */
-struct drm_encoder {
- struct drm_device *dev;
- struct list_head head;
-
- struct drm_mode_object base;
- char *name;
- int encoder_type;
-
- /**
- * @index: Position inside the mode_config.list, can be used as an array
- * index. It is invariant over the lifetime of the encoder.
- */
- unsigned index;
-
- uint32_t possible_crtcs;
- uint32_t possible_clones;
-
- struct drm_crtc *crtc;
- struct drm_bridge *bridge;
- const struct drm_encoder_funcs *funcs;
- const struct drm_encoder_helper_funcs *helper_private;
-};
-
/**
* struct drm_plane_state - mutable plane state
* @plane: backpointer to the plane
for_each_if ((encoder_mask) & (1 << drm_encoder_index(encoder)))
#define obj_to_crtc(x) container_of(x, struct drm_crtc, base)
-#define obj_to_encoder(x) container_of(x, struct drm_encoder, base)
#define obj_to_mode(x) container_of(x, struct drm_display_mode, base)
#define obj_to_fb(x) container_of(x, struct drm_framebuffer, base)
-#define obj_to_property(x) container_of(x, struct drm_property, base)
#define obj_to_blob(x) container_of(x, struct drm_property_blob, base)
#define obj_to_plane(x) container_of(x, struct drm_plane, base)
-struct drm_prop_enum_list {
- int type;
- char *name;
-};
-
extern __printf(6, 7)
int drm_crtc_init_with_planes(struct drm_device *dev,
struct drm_crtc *crtc,
return 1 << drm_crtc_index(crtc);
}
-extern __printf(5, 6)
-int drm_encoder_init(struct drm_device *dev,
- struct drm_encoder *encoder,
- const struct drm_encoder_funcs *funcs,
- int encoder_type, const char *name, ...);
-
-/**
- * drm_encoder_index - find the index of a registered encoder
- * @encoder: encoder to find index for
- *
- * Given a registered encoder, return the index of that encoder within a DRM
- * device's list of encoders.
- */
-static inline unsigned int drm_encoder_index(struct drm_encoder *encoder)
-{
- return encoder->index;
-}
-
-/**
- * drm_encoder_crtc_ok - can a given crtc drive a given encoder?
- * @encoder: encoder to test
- * @crtc: crtc to test
- *
- * Return false if @encoder can't be driven by @crtc, true otherwise.
- */
-static inline bool drm_encoder_crtc_ok(struct drm_encoder *encoder,
- struct drm_crtc *crtc)
-{
- return !!(encoder->possible_crtcs & drm_crtc_mask(crtc));
-}
-
extern __printf(8, 9)
int drm_universal_plane_init(struct drm_device *dev,
struct drm_plane *plane,
extern int drm_crtc_force_disable(struct drm_crtc *crtc);
extern int drm_crtc_force_disable_all(struct drm_device *dev);
-extern void drm_encoder_cleanup(struct drm_encoder *encoder);
-
extern void drm_mode_config_init(struct drm_device *dev);
extern void drm_mode_config_reset(struct drm_device *dev);
extern void drm_mode_config_cleanup(struct drm_device *dev);
-static inline bool drm_property_type_is(struct drm_property *property,
- uint32_t type)
-{
- /* instanceof for props.. handles extended type vs original types: */
- if (property->flags & DRM_MODE_PROP_EXTENDED_TYPE)
- return (property->flags & DRM_MODE_PROP_EXTENDED_TYPE) == type;
- return property->flags & type;
-}
-
-extern int drm_object_property_set_value(struct drm_mode_object *obj,
- struct drm_property *property,
- uint64_t val);
-extern int drm_object_property_get_value(struct drm_mode_object *obj,
- struct drm_property *property,
- uint64_t *value);
-
-extern void drm_object_attach_property(struct drm_mode_object *obj,
- struct drm_property *property,
- uint64_t init_val);
-extern struct drm_property *drm_property_create(struct drm_device *dev, int flags,
- const char *name, int num_values);
-extern struct drm_property *drm_property_create_enum(struct drm_device *dev, int flags,
- const char *name,
- const struct drm_prop_enum_list *props,
- int num_values);
-struct drm_property *drm_property_create_bitmask(struct drm_device *dev,
- int flags, const char *name,
- const struct drm_prop_enum_list *props,
- int num_props,
- uint64_t supported_bits);
-struct drm_property *drm_property_create_range(struct drm_device *dev, int flags,
- const char *name,
- uint64_t min, uint64_t max);
-struct drm_property *drm_property_create_signed_range(struct drm_device *dev,
- int flags, const char *name,
- int64_t min, int64_t max);
-struct drm_property *drm_property_create_object(struct drm_device *dev,
- int flags, const char *name, uint32_t type);
-struct drm_property *drm_property_create_bool(struct drm_device *dev, int flags,
- const char *name);
-struct drm_property_blob *drm_property_create_blob(struct drm_device *dev,
- size_t length,
- const void *data);
-struct drm_property_blob *drm_property_lookup_blob(struct drm_device *dev,
- uint32_t id);
-int drm_property_replace_global_blob(struct drm_device *dev,
- struct drm_property_blob **replace,
- size_t length,
- const void *data,
- struct drm_mode_object *obj_holds_id,
- struct drm_property *prop_holds_id);
-struct drm_property_blob *drm_property_reference_blob(struct drm_property_blob *blob);
-void drm_property_unreference_blob(struct drm_property_blob *blob);
-extern void drm_property_destroy(struct drm_device *dev, struct drm_property *property);
-extern int drm_property_add_enum(struct drm_property *property, int index,
- uint64_t value, const char *name);
extern int drm_mode_crtc_set_gamma_size(struct drm_crtc *crtc,
int gamma_size);
extern int drm_mode_set_config_internal(struct drm_mode_set *set);
-extern uint32_t drm_mode_legacy_fb_format(uint32_t bpp, uint32_t depth);
-
extern struct drm_tile_group *drm_mode_create_tile_group(struct drm_device *dev,
char topology[8]);
extern struct drm_tile_group *drm_mode_get_tile_group(struct drm_device *dev,
return mo ? obj_to_crtc(mo) : NULL;
}
-static inline struct drm_encoder *drm_encoder_find(struct drm_device *dev,
- uint32_t id)
-{
- struct drm_mode_object *mo;
- mo = drm_mode_object_find(dev, id, DRM_MODE_OBJECT_ENCODER);
- return mo ? obj_to_encoder(mo) : NULL;
-}
-
-static inline struct drm_property *drm_property_find(struct drm_device *dev,
- uint32_t id)
-{
- struct drm_mode_object *mo;
- mo = drm_mode_object_find(dev, id, DRM_MODE_OBJECT_PROPERTY);
- return mo ? obj_to_property(mo) : NULL;
-}
-
/*
* Extract a degamma/gamma LUT value provided by user and round it to the
* precision supported by the hardware.
# define DP_DS_PORT_TYPE_DVI 2
# define DP_DS_PORT_TYPE_HDMI 3
# define DP_DS_PORT_TYPE_NON_EDID 4
+# define DP_DS_PORT_TYPE_DP_DUALMODE 5
+# define DP_DS_PORT_TYPE_WIRELESS 6
# define DP_DS_PORT_HPD (1 << 3)
/* offset 1 for VGA is maximum megapixels per second / 8 */
/* offset 2 */
-# define DP_DS_VGA_MAX_BPC_MASK (3 << 0)
-# define DP_DS_VGA_8BPC 0
-# define DP_DS_VGA_10BPC 1
-# define DP_DS_VGA_12BPC 2
-# define DP_DS_VGA_16BPC 3
+# define DP_DS_MAX_BPC_MASK (3 << 0)
+# define DP_DS_8BPC 0
+# define DP_DS_10BPC 1
+# define DP_DS_12BPC 2
+# define DP_DS_16BPC 3
/* link configuration */
#define DP_LINK_BW_SET 0x100
#define DP_SOURCE_OUI 0x300
#define DP_SINK_OUI 0x400
#define DP_BRANCH_OUI 0x500
+#define DP_BRANCH_ID 0x503
+#define DP_BRANCH_HW_REV 0x509
+#define DP_BRANCH_SW_REV 0x50A
#define DP_SET_POWER 0x600
# define DP_SET_POWER_D0 0x1
int drm_dp_link_power_up(struct drm_dp_aux *aux, struct drm_dp_link *link);
int drm_dp_link_power_down(struct drm_dp_aux *aux, struct drm_dp_link *link);
int drm_dp_link_configure(struct drm_dp_aux *aux, struct drm_dp_link *link);
+int drm_dp_downstream_max_clock(const u8 dpcd[DP_RECEIVER_CAP_SIZE],
+ const u8 port_cap[4]);
+int drm_dp_downstream_max_bpc(const u8 dpcd[DP_RECEIVER_CAP_SIZE],
+ const u8 port_cap[4]);
+int drm_dp_downstream_id(struct drm_dp_aux *aux, char id[6]);
+void drm_dp_downstream_debug(struct seq_file *m, const u8 dpcd[DP_RECEIVER_CAP_SIZE],
+ const u8 port_cap[4], struct drm_dp_aux *aux);
void drm_dp_aux_init(struct drm_dp_aux *aux);
int drm_dp_aux_register(struct drm_dp_aux *aux);
--- /dev/null
+/*
+ * Copyright (c) 2016 Intel Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission. The copyright holders make no representations
+ * about the suitability of this software for any purpose. It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#ifndef __DRM_ENCODER_H__
+#define __DRM_ENCODER_H__
+
+#include <linux/list.h>
+#include <linux/ctype.h>
+#include <drm/drm_mode_object.h>
+
+/**
+ * struct drm_encoder_funcs - encoder controls
+ *
+ * Encoders sit between CRTCs and connectors.
+ */
+struct drm_encoder_funcs {
+ /**
+ * @reset:
+ *
+ * Reset encoder hardware and software state to off. This function isn't
+ * called by the core directly, only through drm_mode_config_reset().
+ * It's not a helper hook only for historical reasons.
+ */
+ void (*reset)(struct drm_encoder *encoder);
+
+ /**
+ * @destroy:
+ *
+ * Clean up encoder resources. This is only called at driver unload time
+ * through drm_mode_config_cleanup() since an encoder cannot be
+ * hotplugged in DRM.
+ */
+ void (*destroy)(struct drm_encoder *encoder);
+
+ /**
+ * @late_register:
+ *
+ * This optional hook can be used to register additional userspace
+ * interfaces attached to the encoder like debugfs interfaces.
+ * It is called late in the driver load sequence from drm_dev_register().
+ * Everything added from this callback should be unregistered in
+ * the early_unregister callback.
+ *
+ * Returns:
+ *
+ * 0 on success, or a negative error code on failure.
+ */
+ int (*late_register)(struct drm_encoder *encoder);
+
+ /**
+ * @early_unregister:
+ *
+ * This optional hook should be used to unregister the additional
+ * userspace interfaces attached to the encoder from
+ * late_unregister(). It is called from drm_dev_unregister(),
+ * early in the driver unload sequence to disable userspace access
+ * before data structures are torndown.
+ */
+ void (*early_unregister)(struct drm_encoder *encoder);
+};
+
+/**
+ * struct drm_encoder - central DRM encoder structure
+ * @dev: parent DRM device
+ * @head: list management
+ * @base: base KMS object
+ * @name: human readable name, can be overwritten by the driver
+ * @crtc: currently bound CRTC
+ * @bridge: bridge associated to the encoder
+ * @funcs: control functions
+ * @helper_private: mid-layer private data
+ *
+ * CRTCs drive pixels to encoders, which convert them into signals
+ * appropriate for a given connector or set of connectors.
+ */
+struct drm_encoder {
+ struct drm_device *dev;
+ struct list_head head;
+
+ struct drm_mode_object base;
+ char *name;
+ /**
+ * @encoder_type:
+ *
+ * One of the DRM_MODE_ENCODER_<foo> types in drm_mode.h. The following
+ * encoder types are defined thus far:
+ *
+ * - DRM_MODE_ENCODER_DAC for VGA and analog on DVI-I/DVI-A.
+ *
+ * - DRM_MODE_ENCODER_TMDS for DVI, HDMI and (embedded) DisplayPort.
+ *
+ * - DRM_MODE_ENCODER_LVDS for display panels, or in general any panel
+ * with a proprietary parallel connector.
+ *
+ * - DRM_MODE_ENCODER_TVDAC for TV output (Composite, S-Video,
+ * Component, SCART).
+ *
+ * - DRM_MODE_ENCODER_VIRTUAL for virtual machine displays
+ *
+ * - DRM_MODE_ENCODER_DSI for panels connected using the DSI serial bus.
+ *
+ * - DRM_MODE_ENCODER_DPI for panels connected using the DPI parallel
+ * bus.
+ *
+ * - DRM_MODE_ENCODER_DPMST for special fake encoders used to allow
+ * mutliple DP MST streams to share one physical encoder.
+ */
+ int encoder_type;
+
+ /**
+ * @index: Position inside the mode_config.list, can be used as an array
+ * index. It is invariant over the lifetime of the encoder.
+ */
+ unsigned index;
+
+ /**
+ * @possible_crtcs: Bitmask of potential CRTC bindings, using
+ * drm_crtc_index() as the index into the bitfield. The driver must set
+ * the bits for all &drm_crtc objects this encoder can be connected to
+ * before calling drm_encoder_init().
+ *
+ * In reality almost every driver gets this wrong.
+ *
+ * Note that since CRTC objects can't be hotplugged the assigned indices
+ * are stable and hence known before registering all objects.
+ */
+ uint32_t possible_crtcs;
+
+ /**
+ * @possible_clones: Bitmask of potential sibling encoders for cloning,
+ * using drm_encoder_index() as the index into the bitfield. The driver
+ * must set the bits for all &drm_encoder objects which can clone a
+ * &drm_crtc together with this encoder before calling
+ * drm_encoder_init(). Drivers should set the bit representing the
+ * encoder itself, too. Cloning bits should be set such that when two
+ * encoders can be used in a cloned configuration, they both should have
+ * each another bits set.
+ *
+ * In reality almost every driver gets this wrong.
+ *
+ * Note that since encoder objects can't be hotplugged the assigned indices
+ * are stable and hence known before registering all objects.
+ */
+ uint32_t possible_clones;
+
+ struct drm_crtc *crtc;
+ struct drm_bridge *bridge;
+ const struct drm_encoder_funcs *funcs;
+ const struct drm_encoder_helper_funcs *helper_private;
+};
+
+#define obj_to_encoder(x) container_of(x, struct drm_encoder, base)
+
+__printf(5, 6)
+int drm_encoder_init(struct drm_device *dev,
+ struct drm_encoder *encoder,
+ const struct drm_encoder_funcs *funcs,
+ int encoder_type, const char *name, ...);
+
+/**
+ * drm_encoder_index - find the index of a registered encoder
+ * @encoder: encoder to find index for
+ *
+ * Given a registered encoder, return the index of that encoder within a DRM
+ * device's list of encoders.
+ */
+static inline unsigned int drm_encoder_index(struct drm_encoder *encoder)
+{
+ return encoder->index;
+}
+
+/* FIXME: We have an include file mess still, drm_crtc.h needs untangling. */
+static inline uint32_t drm_crtc_mask(struct drm_crtc *crtc);
+
+/**
+ * drm_encoder_crtc_ok - can a given crtc drive a given encoder?
+ * @encoder: encoder to test
+ * @crtc: crtc to test
+ *
+ * Returns false if @encoder can't be driven by @crtc, true otherwise.
+ */
+static inline bool drm_encoder_crtc_ok(struct drm_encoder *encoder,
+ struct drm_crtc *crtc)
+{
+ return !!(encoder->possible_crtcs & drm_crtc_mask(crtc));
+}
+
+/**
+ * drm_encoder_find - find a &drm_encoder
+ * @dev: DRM device
+ * @id: encoder id
+ *
+ * Returns the encoder with @id, NULL if it doesn't exist. Simple wrapper around
+ * drm_mode_object_find().
+ */
+static inline struct drm_encoder *drm_encoder_find(struct drm_device *dev,
+ uint32_t id)
+{
+ struct drm_mode_object *mo;
+
+ mo = drm_mode_object_find(dev, id, DRM_MODE_OBJECT_ENCODER);
+
+ return mo ? obj_to_encoder(mo) : NULL;
+}
+
+void drm_encoder_cleanup(struct drm_encoder *encoder);
+
+#endif
drm_fb_helper_remove_conflicting_framebuffers(struct apertures_struct *a,
const char *name, bool primary)
{
-#if IS_ENABLED(CONFIG_FB)
+#if IS_REACHABLE(CONFIG_FB)
return remove_conflicting_framebuffers(a, name, primary);
#else
return 0;
#include <linux/types.h>
#include <uapi/drm/drm_fourcc.h>
+uint32_t drm_mode_legacy_fb_format(uint32_t bpp, uint32_t depth);
void drm_fb_get_bpp_depth(uint32_t format, unsigned int *depth, int *bpp);
int drm_format_num_planes(uint32_t format);
int drm_format_plane_cpp(uint32_t format, int plane);
#include <linux/list.h>
#include <linux/ctype.h>
-#include <drm/drm_modeset.h>
+#include <drm/drm_mode_object.h>
struct drm_framebuffer;
struct drm_file;
--- /dev/null
+/*
+ * Copyright (c) 2016 Intel Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission. The copyright holders make no representations
+ * about the suitability of this software for any purpose. It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#ifndef __DRM_MODESET_H__
+#define __DRM_MODESET_H__
+
+#include <linux/kref.h>
+struct drm_object_properties;
+struct drm_property;
+
+/**
+ * struct drm_mode_object - base structure for modeset objects
+ * @id: userspace visible identifier
+ * @type: type of the object, one of DRM_MODE_OBJECT\_\*
+ * @properties: properties attached to this object, including values
+ * @refcount: reference count for objects which with dynamic lifetime
+ * @free_cb: free function callback, only set for objects with dynamic lifetime
+ *
+ * Base structure for modeset objects visible to userspace. Objects can be
+ * looked up using drm_mode_object_find(). Besides basic uapi interface
+ * properties like @id and @type it provides two services:
+ *
+ * - It tracks attached properties and their values. This is used by &drm_crtc,
+ * &drm_plane and &drm_connector. Properties are attached by calling
+ * drm_object_attach_property() before the object is visible to userspace.
+ *
+ * - For objects with dynamic lifetimes (as indicated by a non-NULL @free_cb) it
+ * provides reference counting through drm_mode_object_reference() and
+ * drm_mode_object_unreference(). This is used by &drm_framebuffer,
+ * &drm_connector and &drm_property_blob. These objects provide specialized
+ * reference counting wrappers.
+ */
+struct drm_mode_object {
+ uint32_t id;
+ uint32_t type;
+ struct drm_object_properties *properties;
+ struct kref refcount;
+ void (*free_cb)(struct kref *kref);
+};
+
+#define DRM_OBJECT_MAX_PROPERTY 24
+/**
+ * struct drm_object_properties - property tracking for &drm_mode_object
+ */
+struct drm_object_properties {
+ /**
+ * @count: number of valid properties, must be less than or equal to
+ * DRM_OBJECT_MAX_PROPERTY.
+ */
+
+ int count;
+ /**
+ * @properties: Array of pointers to &drm_property.
+ *
+ * NOTE: if we ever start dynamically destroying properties (ie.
+ * not at drm_mode_config_cleanup() time), then we'd have to do
+ * a better job of detaching property from mode objects to avoid
+ * dangling property pointers:
+ */
+ struct drm_property *properties[DRM_OBJECT_MAX_PROPERTY];
+
+ /**
+ * @values: Array to store the property values, matching @properties. Do
+ * not read/write values directly, but use
+ * drm_object_property_get_value() and drm_object_property_set_value().
+ *
+ * Note that atomic drivers do not store mutable properties in this
+ * array, but only the decoded values in the corresponding state
+ * structure. The decoding is done using the ->atomic_get_property and
+ * ->atomic_set_property hooks of the corresponding object. Hence atomic
+ * drivers should not use drm_object_property_set_value() and
+ * drm_object_property_get_value() on mutable objects, i.e. those
+ * without the DRM_MODE_PROP_IMMUTABLE flag set.
+ */
+ uint64_t values[DRM_OBJECT_MAX_PROPERTY];
+};
+
+/* Avoid boilerplate. I'm tired of typing. */
+#define DRM_ENUM_NAME_FN(fnname, list) \
+ const char *fnname(int val) \
+ { \
+ int i; \
+ for (i = 0; i < ARRAY_SIZE(list); i++) { \
+ if (list[i].type == val) \
+ return list[i].name; \
+ } \
+ return "(unknown)"; \
+ }
+
+struct drm_mode_object *drm_mode_object_find(struct drm_device *dev,
+ uint32_t id, uint32_t type);
+void drm_mode_object_reference(struct drm_mode_object *obj);
+void drm_mode_object_unreference(struct drm_mode_object *obj);
+
+int drm_object_property_set_value(struct drm_mode_object *obj,
+ struct drm_property *property,
+ uint64_t val);
+int drm_object_property_get_value(struct drm_mode_object *obj,
+ struct drm_property *property,
+ uint64_t *value);
+
+void drm_object_attach_property(struct drm_mode_object *obj,
+ struct drm_property *property,
+ uint64_t init_val);
+#endif
#ifndef __DRM_MODES_H__
#define __DRM_MODES_H__
-#include <drm/drm_modeset.h>
+#include <drm/drm_mode_object.h>
#include <drm/drm_connector.h>
/*
+++ /dev/null
-/*
- * Copyright (c) 2016 Intel Corporation
- *
- * Permission to use, copy, modify, distribute, and sell this software and its
- * documentation for any purpose is hereby granted without fee, provided that
- * the above copyright notice appear in all copies and that both that copyright
- * notice and this permission notice appear in supporting documentation, and
- * that the name of the copyright holders not be used in advertising or
- * publicity pertaining to distribution of the software without specific,
- * written prior permission. The copyright holders make no representations
- * about the suitability of this software for any purpose. It is provided "as
- * is" without express or implied warranty.
- *
- * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
- * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
- * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
- * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
- * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
- * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
- * OF THIS SOFTWARE.
- */
-
-#ifndef __DRM_MODESET_H__
-#define __DRM_MODESET_H__
-
-#include <linux/kref.h>
-struct drm_object_properties;
-struct drm_property;
-
-struct drm_mode_object {
- uint32_t id;
- uint32_t type;
- struct drm_object_properties *properties;
- struct kref refcount;
- void (*free_cb)(struct kref *kref);
-};
-
-#define DRM_OBJECT_MAX_PROPERTY 24
-struct drm_object_properties {
- int count, atomic_count;
- /* NOTE: if we ever start dynamically destroying properties (ie.
- * not at drm_mode_config_cleanup() time), then we'd have to do
- * a better job of detaching property from mode objects to avoid
- * dangling property pointers:
- */
- struct drm_property *properties[DRM_OBJECT_MAX_PROPERTY];
- /* do not read/write values directly, but use drm_object_property_get_value()
- * and drm_object_property_set_value():
- */
- uint64_t values[DRM_OBJECT_MAX_PROPERTY];
-};
-
-/* Avoid boilerplate. I'm tired of typing. */
-#define DRM_ENUM_NAME_FN(fnname, list) \
- const char *fnname(int val) \
- { \
- int i; \
- for (i = 0; i < ARRAY_SIZE(list); i++) { \
- if (list[i].type == val) \
- return list[i].name; \
- } \
- return "(unknown)"; \
- }
-
-struct drm_mode_object *drm_mode_object_find(struct drm_device *dev,
- uint32_t id, uint32_t type);
-void drm_mode_object_reference(struct drm_mode_object *obj);
-void drm_mode_object_unreference(struct drm_mode_object *obj);
-
-#endif
--- /dev/null
+/*
+ * Copyright (c) 2016 Intel Corporation
+ *
+ * Permission to use, copy, modify, distribute, and sell this software and its
+ * documentation for any purpose is hereby granted without fee, provided that
+ * the above copyright notice appear in all copies and that both that copyright
+ * notice and this permission notice appear in supporting documentation, and
+ * that the name of the copyright holders not be used in advertising or
+ * publicity pertaining to distribution of the software without specific,
+ * written prior permission. The copyright holders make no representations
+ * about the suitability of this software for any purpose. It is provided "as
+ * is" without express or implied warranty.
+ *
+ * THE COPYRIGHT HOLDERS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
+ * EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
+ * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
+ * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
+ * OF THIS SOFTWARE.
+ */
+
+#ifndef __DRM_PROPERTY_H__
+#define __DRM_PROPERTY_H__
+
+#include <linux/list.h>
+#include <linux/ctype.h>
+#include <drm/drm_mode_object.h>
+
+/**
+ * struct drm_property_enum - symbolic values for enumerations
+ * @value: numeric property value for this enum entry
+ * @head: list of enum values, linked to enum_list in &drm_property
+ * @name: symbolic name for the enum
+ *
+ * For enumeration and bitmask properties this structure stores the symbolic
+ * decoding for each value. This is used for example for the rotation property.
+ */
+struct drm_property_enum {
+ uint64_t value;
+ struct list_head head;
+ char name[DRM_PROP_NAME_LEN];
+};
+
+/**
+ * struct drm_property - modeset object property
+ *
+ * This structure represent a modeset object property. It combines both the name
+ * of the property with the set of permissible values. This means that when a
+ * driver wants to use a property with the same name on different objects, but
+ * with different value ranges, then it must create property for each one. An
+ * example would be rotation of &drm_plane, when e.g. the primary plane cannot
+ * be rotated. But if both the name and the value range match, then the same
+ * property structure can be instantiated multiple times for the same object.
+ * Userspace must be able to cope with this and cannot assume that the same
+ * symbolic property will have the same modeset object ID on all modeset
+ * objects.
+ *
+ * Properties are created by one of the special functions, as explained in
+ * detail in the @flags structure member.
+ *
+ * To actually expose a property it must be attached to each object using
+ * drm_object_attach_property(). Currently properties can only be attached to
+ * &drm_connector, &drm_crtc and &drm_plane.
+ *
+ * Properties are also used as the generic metadatatransport for the atomic
+ * IOCTL. Everything that was set directly in structures in the legacy modeset
+ * IOCTLs (like the plane source or destination windows, or e.g. the links to
+ * the CRTC) is exposed as a property with the DRM_MODE_PROP_ATOMIC flag set.
+ */
+struct drm_property {
+ /**
+ * @head: per-device list of properties, for cleanup.
+ */
+ struct list_head head;
+
+ /**
+ * @base: base KMS object
+ */
+ struct drm_mode_object base;
+
+ /**
+ * @flags:
+ *
+ * Property flags and type. A property needs to be one of the following
+ * types:
+ *
+ * DRM_MODE_PROP_RANGE
+ * Range properties report their minimum and maximum admissible unsigned values.
+ * The KMS core verifies that values set by application fit in that
+ * range. The range is unsigned. Range properties are created using
+ * drm_property_create_range().
+ *
+ * DRM_MODE_PROP_SIGNED_RANGE
+ * Range properties report their minimum and maximum admissible unsigned values.
+ * The KMS core verifies that values set by application fit in that
+ * range. The range is signed. Range properties are created using
+ * drm_property_create_signed_range().
+ *
+ * DRM_MODE_PROP_ENUM
+ * Enumerated properties take a numerical value that ranges from 0 to
+ * the number of enumerated values defined by the property minus one,
+ * and associate a free-formed string name to each value. Applications
+ * can retrieve the list of defined value-name pairs and use the
+ * numerical value to get and set property instance values. Enum
+ * properties are created using drm_property_create_enum().
+ *
+ * DRM_MODE_PROP_BITMASK
+ * Bitmask properties are enumeration properties that additionally
+ * restrict all enumerated values to the 0..63 range. Bitmask property
+ * instance values combine one or more of the enumerated bits defined
+ * by the property. Bitmask properties are created using
+ * drm_property_create_bitmask().
+ *
+ * DRM_MODE_PROB_OBJECT
+ * Object properties are used to link modeset objects. This is used
+ * extensively in the atomic support to create the display pipeline,
+ * by linking &drm_framebuffer to &drm_plane, &drm_plane to
+ * &drm_crtc and &drm_connector to &drm_crtc. An object property can
+ * only link to a specific type of &drm_mode_object, this limit is
+ * enforced by the core. Object properties are created using
+ * drm_property_create_object().
+ *
+ * Object properties work like blob properties, but in a more
+ * general fashion. They are limited to atomic drivers and must have
+ * the DRM_MODE_PROP_ATOMIC flag set.
+ *
+ * DRM_MODE_PROP_BLOB
+ * Blob properties store a binary blob without any format restriction.
+ * The binary blobs are created as KMS standalone objects, and blob
+ * property instance values store the ID of their associated blob
+ * object. Blob properties are created by calling
+ * drm_property_create() with DRM_MODE_PROP_BLOB as the type.
+ *
+ * Actual blob objects to contain blob data are created using
+ * drm_property_create_blob(), or through the corresponding IOCTL.
+ *
+ * Besides the built-in limit to only accept blob objects blob
+ * properties work exactly like object properties. The only reasons
+ * blob properties exist is backwards compatibility with existing
+ * userspace.
+ *
+ * In addition a property can have any combination of the below flags:
+ *
+ * DRM_MODE_PROP_ATOMIC
+ * Set for properties which encode atomic modeset state. Such
+ * properties are not exposed to legacy userspace.
+ *
+ * DRM_MODE_PROP_IMMUTABLE
+ * Set for properties where userspace cannot be changed by
+ * userspace. The kernel is allowed to update the value of these
+ * properties. This is generally used to expose probe state to
+ * usersapce, e.g. the EDID, or the connector path property on DP
+ * MST sinks.
+ */
+ uint32_t flags;
+
+ /**
+ * @name: symbolic name of the properties
+ */
+ char name[DRM_PROP_NAME_LEN];
+
+ /**
+ * @num_values: size of the @values array.
+ */
+ uint32_t num_values;
+
+ /**
+ * @values:
+ *
+ * Array with limits and values for the property. The
+ * interpretation of these limits is dependent upon the type per @flags.
+ */
+ uint64_t *values;
+
+ /**
+ * @dev: DRM device
+ */
+ struct drm_device *dev;
+
+ /**
+ * @enum_list:
+ *
+ * List of &drm_prop_enum_list structures with the symbolic names for
+ * enum and bitmask values.
+ */
+ struct list_head enum_list;
+};
+
+/**
+ * struct drm_property_blob - Blob data for &drm_property
+ * @base: base KMS object
+ * @dev: DRM device
+ * @head_global: entry on the global blob list in &drm_mode_config
+ * property_blob_list.
+ * @head_file: entry on the per-file blob list in &drm_file blobs list.
+ * @length: size of the blob in bytes, invariant over the lifetime of the object
+ * @data: actual data, embedded at the end of this structure
+ *
+ * Blobs are used to store bigger values than what fits directly into the 64
+ * bits available for a &drm_property.
+ *
+ * Blobs are reference counted using drm_property_reference_blob() and
+ * drm_property_unreference_blob(). They are created using
+ * drm_property_create_blob().
+ */
+struct drm_property_blob {
+ struct drm_mode_object base;
+ struct drm_device *dev;
+ struct list_head head_global;
+ struct list_head head_file;
+ size_t length;
+ unsigned char data[];
+};
+
+struct drm_prop_enum_list {
+ int type;
+ char *name;
+};
+
+#define obj_to_property(x) container_of(x, struct drm_property, base)
+
+/**
+ * drm_property_type_is - check the type of a property
+ * @property: property to check
+ * @type: property type to compare with
+ *
+ * This is a helper function becauase the uapi encoding of property types is
+ * a bit special for historical reasons.
+ */
+static inline bool drm_property_type_is(struct drm_property *property,
+ uint32_t type)
+{
+ /* instanceof for props.. handles extended type vs original types: */
+ if (property->flags & DRM_MODE_PROP_EXTENDED_TYPE)
+ return (property->flags & DRM_MODE_PROP_EXTENDED_TYPE) == type;
+ return property->flags & type;
+}
+
+struct drm_property *drm_property_create(struct drm_device *dev, int flags,
+ const char *name, int num_values);
+struct drm_property *drm_property_create_enum(struct drm_device *dev, int flags,
+ const char *name,
+ const struct drm_prop_enum_list *props,
+ int num_values);
+struct drm_property *drm_property_create_bitmask(struct drm_device *dev,
+ int flags, const char *name,
+ const struct drm_prop_enum_list *props,
+ int num_props,
+ uint64_t supported_bits);
+struct drm_property *drm_property_create_range(struct drm_device *dev, int flags,
+ const char *name,
+ uint64_t min, uint64_t max);
+struct drm_property *drm_property_create_signed_range(struct drm_device *dev,
+ int flags, const char *name,
+ int64_t min, int64_t max);
+struct drm_property *drm_property_create_object(struct drm_device *dev,
+ int flags, const char *name, uint32_t type);
+struct drm_property *drm_property_create_bool(struct drm_device *dev, int flags,
+ const char *name);
+int drm_property_add_enum(struct drm_property *property, int index,
+ uint64_t value, const char *name);
+void drm_property_destroy(struct drm_device *dev, struct drm_property *property);
+
+struct drm_property_blob *drm_property_create_blob(struct drm_device *dev,
+ size_t length,
+ const void *data);
+struct drm_property_blob *drm_property_lookup_blob(struct drm_device *dev,
+ uint32_t id);
+int drm_property_replace_global_blob(struct drm_device *dev,
+ struct drm_property_blob **replace,
+ size_t length,
+ const void *data,
+ struct drm_mode_object *obj_holds_id,
+ struct drm_property *prop_holds_id);
+struct drm_property_blob *drm_property_reference_blob(struct drm_property_blob *blob);
+void drm_property_unreference_blob(struct drm_property_blob *blob);
+
+/**
+ * drm_connector_find - find property object
+ * @dev: DRM device
+ * @id: property object id
+ *
+ * This function looks up the property object specified by id and returns it.
+ */
+static inline struct drm_property *drm_property_find(struct drm_device *dev,
+ uint32_t id)
+{
+ struct drm_mode_object *mo;
+ mo = drm_mode_object_find(dev, id, DRM_MODE_OBJECT_PROPERTY);
+ return mo ? obj_to_property(mo) : NULL;
+}
+
+#endif
#ifndef __DRM_I2C_TDA998X_H__
#define __DRM_I2C_TDA998X_H__
+#include <linux/hdmi.h>
+#include <dt-bindings/display/tda998x.h>
+
+enum {
+ AFMT_UNUSED = 0,
+ AFMT_SPDIF = TDA998x_SPDIF,
+ AFMT_I2S = TDA998x_I2S,
+};
+
+struct tda998x_audio_params {
+ u8 config;
+ u8 format;
+ unsigned sample_width;
+ unsigned sample_rate;
+ struct hdmi_audio_infoframe cea;
+ u8 status[5];
+};
+
struct tda998x_encoder_params {
u8 swap_b:3;
u8 mirr_b:1;
u8 swap_e:3;
u8 mirr_e:1;
- u8 audio_cfg;
- u8 audio_clk_cfg;
- u8 audio_frame[6];
-
- enum {
- AFMT_SPDIF,
- AFMT_I2S
- } audio_format;
-
- unsigned audio_sample_rate;
+ struct tda998x_audio_params audio_params;
};
#endif
#define I845_TSEG_SIZE_1M (3 << 1)
#define INTEL_BSM 0x5c
-#define INTEL_BSM_MASK (0xFFFF << 20)
+#define INTEL_BSM_MASK (-(1u << 20))
#endif /* _I915_DRM_H_ */
#define INTEL_IVB_Q_IDS(info) \
INTEL_QUANTA_VGA_DEVICE(info) /* Quanta transcode */
-#define INTEL_HSW_D_IDS(info) \
+#define INTEL_HSW_IDS(info) \
INTEL_VGA_DEVICE(0x0402, info), /* GT1 desktop */ \
INTEL_VGA_DEVICE(0x0412, info), /* GT2 desktop */ \
INTEL_VGA_DEVICE(0x0422, info), /* GT3 desktop */ \
INTEL_VGA_DEVICE(0x0D2B, info), /* CRW GT3 reserved */ \
INTEL_VGA_DEVICE(0x0D0E, info), /* CRW GT1 reserved */ \
INTEL_VGA_DEVICE(0x0D1E, info), /* CRW GT2 reserved */ \
- INTEL_VGA_DEVICE(0x0D2E, info) /* CRW GT3 reserved */ \
-
-#define INTEL_HSW_M_IDS(info) \
+ INTEL_VGA_DEVICE(0x0D2E, info), /* CRW GT3 reserved */ \
INTEL_VGA_DEVICE(0x0406, info), /* GT1 mobile */ \
INTEL_VGA_DEVICE(0x0416, info), /* GT2 mobile */ \
INTEL_VGA_DEVICE(0x0426, info), /* GT2 mobile */ \
INTEL_VGA_DEVICE(0x0D16, info), /* CRW GT2 mobile */ \
INTEL_VGA_DEVICE(0x0D26, info) /* CRW GT3 mobile */
-#define INTEL_VLV_M_IDS(info) \
+#define INTEL_VLV_IDS(info) \
INTEL_VGA_DEVICE(0x0f30, info), \
INTEL_VGA_DEVICE(0x0f31, info), \
INTEL_VGA_DEVICE(0x0f32, info), \
INTEL_VGA_DEVICE(0x0f33, info), \
- INTEL_VGA_DEVICE(0x0157, info)
-
-#define INTEL_VLV_D_IDS(info) \
+ INTEL_VGA_DEVICE(0x0157, info), \
INTEL_VGA_DEVICE(0x0155, info)
-#define INTEL_BDW_GT12M_IDS(info) \
+#define INTEL_BDW_GT12_IDS(info) \
INTEL_VGA_DEVICE(0x1602, info), /* GT1 ULT */ \
INTEL_VGA_DEVICE(0x1606, info), /* GT1 ULT */ \
INTEL_VGA_DEVICE(0x160B, info), /* GT1 Iris */ \
INTEL_VGA_DEVICE(0x1612, info), /* GT2 Halo */ \
INTEL_VGA_DEVICE(0x1616, info), /* GT2 ULT */ \
INTEL_VGA_DEVICE(0x161B, info), /* GT2 ULT */ \
- INTEL_VGA_DEVICE(0x161E, info) /* GT2 ULX */
-
-#define INTEL_BDW_GT12D_IDS(info) \
+ INTEL_VGA_DEVICE(0x161E, info), /* GT2 ULX */ \
INTEL_VGA_DEVICE(0x160A, info), /* GT1 Server */ \
INTEL_VGA_DEVICE(0x160D, info), /* GT1 Workstation */ \
INTEL_VGA_DEVICE(0x161A, info), /* GT2 Server */ \
INTEL_VGA_DEVICE(0x161D, info) /* GT2 Workstation */
-#define INTEL_BDW_GT3M_IDS(info) \
+#define INTEL_BDW_GT3_IDS(info) \
INTEL_VGA_DEVICE(0x1622, info), /* ULT */ \
INTEL_VGA_DEVICE(0x1626, info), /* ULT */ \
INTEL_VGA_DEVICE(0x162B, info), /* Iris */ \
- INTEL_VGA_DEVICE(0x162E, info) /* ULX */
-
-#define INTEL_BDW_GT3D_IDS(info) \
+ INTEL_VGA_DEVICE(0x162E, info), /* ULX */\
INTEL_VGA_DEVICE(0x162A, info), /* Server */ \
INTEL_VGA_DEVICE(0x162D, info) /* Workstation */
INTEL_VGA_DEVICE(0x163A, info), /* Server */ \
INTEL_VGA_DEVICE(0x163D, info) /* Workstation */
-#define INTEL_BDW_M_IDS(info) \
- INTEL_BDW_GT12M_IDS(info), \
- INTEL_BDW_GT3M_IDS(info), \
- INTEL_BDW_RSVDM_IDS(info)
-
-#define INTEL_BDW_D_IDS(info) \
- INTEL_BDW_GT12D_IDS(info), \
- INTEL_BDW_GT3D_IDS(info), \
+#define INTEL_BDW_IDS(info) \
+ INTEL_BDW_GT12_IDS(info), \
+ INTEL_BDW_GT3_IDS(info), \
+ INTEL_BDW_RSVDM_IDS(info), \
+ INTEL_BDW_GT12_IDS(info), \
+ INTEL_BDW_GT3_IDS(info), \
INTEL_BDW_RSVDD_IDS(info)
#define INTEL_CHV_IDS(info) \
struct drm_mm_node;
-/**
- * struct ttm_place
- *
- * @fpfn: first valid page frame number to put the object
- * @lpfn: last valid page frame number to put the object
- * @flags: memory domain and caching flags for the object
- *
- * Structure indicating a possible place to put an object.
- */
-struct ttm_place {
- unsigned fpfn;
- unsigned lpfn;
- uint32_t flags;
-};
-
-/**
- * struct ttm_placement
- *
- * @num_placement: number of preferred placements
- * @placement: preferred placements
- * @num_busy_placement: number of preferred placements when need to evict buffer
- * @busy_placement: preferred placements when need to evict buffer
- *
- * Structure indicating the placement you request for an object.
- */
-struct ttm_placement {
- unsigned num_placement;
- const struct ttm_place *placement;
- unsigned num_busy_placement;
- const struct ttm_place *busy_placement;
-};
+struct ttm_placement;
/**
* struct ttm_bus_placement
* struct ttm_dma_tt
*
* @ttm: Base ttm_tt struct.
- * @cpu_address: The CPU address of the pages
* @dma_address: The DMA (bus) addresses of the pages
* @pages_list: used by some page allocation backend
*
*/
struct ttm_dma_tt {
struct ttm_tt ttm;
- void **cpu_address;
dma_addr_t *dma_address;
struct list_head pages_list;
};
* ttm_bo_move_ttm
*
* @bo: A pointer to a struct ttm_buffer_object.
- * @evict: 1: This is an eviction. Don't try to pipeline.
* @interruptible: Sleep interruptible if waiting.
* @no_wait_gpu: Return immediately if the GPU is busy.
* @new_mem: struct ttm_mem_reg indicating where to move.
*/
extern int ttm_bo_move_ttm(struct ttm_buffer_object *bo,
- bool evict, bool interruptible, bool no_wait_gpu,
+ bool interruptible, bool no_wait_gpu,
struct ttm_mem_reg *new_mem);
/**
* ttm_bo_move_memcpy
*
* @bo: A pointer to a struct ttm_buffer_object.
- * @evict: 1: This is an eviction. Don't try to pipeline.
* @interruptible: Sleep interruptible if waiting.
* @no_wait_gpu: Return immediately if the GPU is busy.
* @new_mem: struct ttm_mem_reg indicating where to move.
*/
extern int ttm_bo_move_memcpy(struct ttm_buffer_object *bo,
- bool evict, bool interruptible,
- bool no_wait_gpu,
+ bool interruptible, bool no_wait_gpu,
struct ttm_mem_reg *new_mem);
/**
extern void ttm_mem_global_free_page(struct ttm_mem_global *glob,
struct page *page);
extern size_t ttm_round_pot(size_t size);
+extern uint64_t ttm_get_kernel_zone_memory_size(struct ttm_mem_global *glob);
#endif
#ifndef _TTM_PLACEMENT_H_
#define _TTM_PLACEMENT_H_
+
+#include <linux/types.h>
+
/*
* Memory regions for data placement.
*/
#define TTM_PL_SYSTEM 0
#define TTM_PL_TT 1
#define TTM_PL_VRAM 2
-#define TTM_PL_PRIV0 3
-#define TTM_PL_PRIV1 4
-#define TTM_PL_PRIV2 5
-#define TTM_PL_PRIV3 6
-#define TTM_PL_PRIV4 7
-#define TTM_PL_PRIV5 8
-#define TTM_PL_SWAPPED 15
+#define TTM_PL_PRIV 3
#define TTM_PL_FLAG_SYSTEM (1 << TTM_PL_SYSTEM)
#define TTM_PL_FLAG_TT (1 << TTM_PL_TT)
#define TTM_PL_FLAG_VRAM (1 << TTM_PL_VRAM)
-#define TTM_PL_FLAG_PRIV0 (1 << TTM_PL_PRIV0)
-#define TTM_PL_FLAG_PRIV1 (1 << TTM_PL_PRIV1)
-#define TTM_PL_FLAG_PRIV2 (1 << TTM_PL_PRIV2)
-#define TTM_PL_FLAG_PRIV3 (1 << TTM_PL_PRIV3)
-#define TTM_PL_FLAG_PRIV4 (1 << TTM_PL_PRIV4)
-#define TTM_PL_FLAG_PRIV5 (1 << TTM_PL_PRIV5)
-#define TTM_PL_FLAG_SWAPPED (1 << TTM_PL_SWAPPED)
+#define TTM_PL_FLAG_PRIV (1 << TTM_PL_PRIV)
#define TTM_PL_MASK_MEM 0x0000FFFF
/*
#define TTM_PL_FLAG_CACHED (1 << 16)
#define TTM_PL_FLAG_UNCACHED (1 << 17)
#define TTM_PL_FLAG_WC (1 << 18)
-#define TTM_PL_FLAG_SHARED (1 << 20)
#define TTM_PL_FLAG_NO_EVICT (1 << 21)
#define TTM_PL_FLAG_TOPDOWN (1 << 22)
#define TTM_PL_MASK_MEMTYPE (TTM_PL_MASK_MEM | TTM_PL_MASK_CACHING)
-/*
- * Access flags to be used for CPU- and GPU- mappings.
- * The idea is that the TTM synchronization mechanism will
- * allow concurrent READ access and exclusive write access.
- * Currently GPU- and CPU accesses are exclusive.
+/**
+ * struct ttm_place
+ *
+ * @fpfn: first valid page frame number to put the object
+ * @lpfn: last valid page frame number to put the object
+ * @flags: memory domain and caching flags for the object
+ *
+ * Structure indicating a possible place to put an object.
*/
+struct ttm_place {
+ unsigned fpfn;
+ unsigned lpfn;
+ uint32_t flags;
+};
-#define TTM_ACCESS_READ (1 << 0)
-#define TTM_ACCESS_WRITE (1 << 1)
+/**
+ * struct ttm_placement
+ *
+ * @num_placement: number of preferred placements
+ * @placement: preferred placements
+ * @num_busy_placement: number of preferred placements when need to evict buffer
+ * @busy_placement: preferred placements when need to evict buffer
+ *
+ * Structure indicating the placement you request for an object.
+ */
+struct ttm_placement {
+ unsigned num_placement;
+ const struct ttm_place *placement;
+ unsigned num_busy_placement;
+ const struct ttm_place *busy_placement;
+};
#endif
--- /dev/null
+#ifndef _DT_BINDINGS_TDA998X_H
+#define _DT_BINDINGS_TDA998X_H
+
+#define TDA998x_SPDIF 1
+#define TDA998x_I2S 2
+
+#endif /*_DT_BINDINGS_TDA998X_H */
* See Documentation/io-mapping.txt
*/
-#ifdef CONFIG_HAVE_ATOMIC_IOMAP
-
-#include <asm/iomap.h>
-
struct io_mapping {
resource_size_t base;
unsigned long size;
pgprot_t prot;
+ void __iomem *iomem;
};
+#ifdef CONFIG_HAVE_ATOMIC_IOMAP
+
+#include <asm/iomap.h>
/*
* For small address space machines, mapping large objects
* into the kernel virtual space isn't practical. Where
*/
static inline struct io_mapping *
-io_mapping_create_wc(resource_size_t base, unsigned long size)
+io_mapping_init_wc(struct io_mapping *iomap,
+ resource_size_t base,
+ unsigned long size)
{
- struct io_mapping *iomap;
pgprot_t prot;
- iomap = kmalloc(sizeof(*iomap), GFP_KERNEL);
- if (!iomap)
- goto out_err;
-
if (iomap_create_wc(base, size, &prot))
- goto out_free;
+ return NULL;
iomap->base = base;
iomap->size = size;
iomap->prot = prot;
return iomap;
-
-out_free:
- kfree(iomap);
-out_err:
- return NULL;
}
static inline void
-io_mapping_free(struct io_mapping *mapping)
+io_mapping_fini(struct io_mapping *mapping)
{
iomap_free(mapping->base, mapping->size);
- kfree(mapping);
}
/* Atomic map/unmap */
#else
#include <linux/uaccess.h>
-
-/* this struct isn't actually defined anywhere */
-struct io_mapping;
+#include <asm/pgtable.h>
/* Create the io_mapping object*/
static inline struct io_mapping *
-io_mapping_create_wc(resource_size_t base, unsigned long size)
+io_mapping_init_wc(struct io_mapping *iomap,
+ resource_size_t base,
+ unsigned long size)
{
- return (struct io_mapping __force *) ioremap_wc(base, size);
+ iomap->base = base;
+ iomap->size = size;
+ iomap->iomem = ioremap_wc(base, size);
+#if defined(pgprot_noncached_wc) /* archs can't agree on a name ... */
+ iomap->prot = pgprot_noncached_wc(PAGE_KERNEL);
+#elif defined(pgprot_writecombine)
+ iomap->prot = pgprot_writecombine(PAGE_KERNEL);
+#else
+ iomap->prot = pgprot_noncached(PAGE_KERNEL);
+#endif
+
+ return iomap;
}
static inline void
-io_mapping_free(struct io_mapping *mapping)
+io_mapping_fini(struct io_mapping *mapping)
+{
+ iounmap(mapping->iomem);
+}
+
+/* Non-atomic map/unmap */
+static inline void __iomem *
+io_mapping_map_wc(struct io_mapping *mapping,
+ unsigned long offset,
+ unsigned long size)
+{
+ return mapping->iomem + offset;
+}
+
+static inline void
+io_mapping_unmap(void __iomem *vaddr)
{
- iounmap((void __force __iomem *) mapping);
}
/* Atomic map/unmap */
{
preempt_disable();
pagefault_disable();
- return ((char __force __iomem *) mapping) + offset;
+ return io_mapping_map_wc(mapping, offset, PAGE_SIZE);
}
static inline void
io_mapping_unmap_atomic(void __iomem *vaddr)
{
+ io_mapping_unmap(vaddr);
pagefault_enable();
preempt_enable();
}
-/* Non-atomic map/unmap */
-static inline void __iomem *
-io_mapping_map_wc(struct io_mapping *mapping,
- unsigned long offset,
- unsigned long size)
+#endif /* HAVE_ATOMIC_IOMAP */
+
+static inline struct io_mapping *
+io_mapping_create_wc(resource_size_t base,
+ unsigned long size)
{
- return ((char __force __iomem *) mapping) + offset;
+ struct io_mapping *iomap;
+
+ iomap = kmalloc(sizeof(*iomap), GFP_KERNEL);
+ if (!iomap)
+ return NULL;
+
+ if (!io_mapping_init_wc(iomap, base, size)) {
+ kfree(iomap);
+ return NULL;
+ }
+
+ return iomap;
}
static inline void
-io_mapping_unmap(void __iomem *vaddr)
+io_mapping_free(struct io_mapping *iomap)
{
+ io_mapping_fini(iomap);
+ kfree(iomap);
}
-#endif /* HAVE_ATOMIC_IOMAP */
-
#endif /* _LINUX_IO_MAPPING_H */
#define AMDGPU_GEM_CREATE_NO_CPU_ACCESS (1 << 1)
/* Flag that USWC attributes should be used for GTT */
#define AMDGPU_GEM_CREATE_CPU_GTT_USWC (1 << 2)
+/* Flag that the memory should be in VRAM and cleared */
+#define AMDGPU_GEM_CREATE_VRAM_CLEARED (1 << 3)
+/* Flag that create shadow bo(GTT) while allocating vram bo */
+#define AMDGPU_GEM_CREATE_SHADOW (1 << 4)
struct drm_amdgpu_gem_create_in {
/** the requested memory size */
#define AMDGPU_INFO_DEV_INFO 0x16
/* visible vram usage */
#define AMDGPU_INFO_VIS_VRAM_USAGE 0x17
+/* number of TTM buffer evictions */
+#define AMDGPU_INFO_NUM_EVICTIONS 0x18
#define AMDGPU_INFO_MMR_SE_INDEX_SHIFT 0
#define AMDGPU_INFO_MMR_SE_INDEX_MASK 0xff
* Supported GPU families
*/
#define AMDGPU_FAMILY_UNKNOWN 0
+#define AMDGPU_FAMILY_SI 110 /* Hainan, Oland, Verde, Pitcairn, Tahiti */
#define AMDGPU_FAMILY_CI 120 /* Bonaire, Hawaii */
#define AMDGPU_FAMILY_KV 125 /* Kaveri, Kabini, Mullins */
#define AMDGPU_FAMILY_VI 130 /* Iceland, Tonga */
#define DRM_CAP_CURSOR_WIDTH 0x8
#define DRM_CAP_CURSOR_HEIGHT 0x9
#define DRM_CAP_ADDFB2_MODIFIERS 0x10
+#define DRM_CAP_PAGE_FLIP_TARGET 0x11
/** DRM_IOCTL_GET_CAP ioctl argument type */
struct drm_get_cap {
#define DRM_MODE_PAGE_FLIP_EVENT 0x01
#define DRM_MODE_PAGE_FLIP_ASYNC 0x02
-#define DRM_MODE_PAGE_FLIP_FLAGS (DRM_MODE_PAGE_FLIP_EVENT|DRM_MODE_PAGE_FLIP_ASYNC)
+#define DRM_MODE_PAGE_FLIP_TARGET_ABSOLUTE 0x4
+#define DRM_MODE_PAGE_FLIP_TARGET_RELATIVE 0x8
+#define DRM_MODE_PAGE_FLIP_TARGET (DRM_MODE_PAGE_FLIP_TARGET_ABSOLUTE | \
+ DRM_MODE_PAGE_FLIP_TARGET_RELATIVE)
+#define DRM_MODE_PAGE_FLIP_FLAGS (DRM_MODE_PAGE_FLIP_EVENT | \
+ DRM_MODE_PAGE_FLIP_ASYNC | \
+ DRM_MODE_PAGE_FLIP_TARGET)
/*
* Request a page flip on the specified crtc.
* 'as soon as possible', meaning that it not delay waiting for vblank.
* This may cause tearing on the screen.
*
- * The reserved field must be zero until we figure out something
- * clever to use it for.
+ * The reserved field must be zero.
*/
struct drm_mode_crtc_page_flip {
__u64 user_data;
};
+/*
+ * Request a page flip on the specified crtc.
+ *
+ * Same as struct drm_mode_crtc_page_flip, but supports new flags and
+ * re-purposes the reserved field:
+ *
+ * The sequence field must be zero unless either of the
+ * DRM_MODE_PAGE_FLIP_TARGET_ABSOLUTE/RELATIVE flags is specified. When
+ * the ABSOLUTE flag is specified, the sequence field denotes the absolute
+ * vblank sequence when the flip should take effect. When the RELATIVE
+ * flag is specified, the sequence field denotes the relative (to the
+ * current one when the ioctl is called) vblank sequence when the flip
+ * should take effect. NOTE: DRM_IOCTL_WAIT_VBLANK must still be used to
+ * make sure the vblank sequence before the target one has passed before
+ * calling this ioctl. The purpose of the
+ * DRM_MODE_PAGE_FLIP_TARGET_ABSOLUTE/RELATIVE flags is merely to clarify
+ * the target for when code dealing with a page flip runs during a
+ * vertical blank period.
+ */
+
+struct drm_mode_crtc_page_flip_target {
+ __u32 crtc_id;
+ __u32 fb_id;
+ __u32 flags;
+ __u32 sequence;
+ __u64 user_data;
+};
+
/* create a dumb scanout buffer */
struct drm_mode_create_dumb {
__u32 height;
#define I915_PARAM_HAS_EXEC_SOFTPIN 37
#define I915_PARAM_HAS_POOLED_EU 38
#define I915_PARAM_MIN_EU_IN_POOL 39
+#define I915_PARAM_MMAP_GTT_VERSION 40
typedef struct drm_i915_getparam {
__s32 param;
* having flushed any pending activity), and a non-zero return that
* the object is still in-flight on the GPU. (The GPU has not yet
* signaled completion for all pending requests that reference the
- * object.)
+ * object.) An object is guaranteed to become idle eventually (so
+ * long as no new GPU commands are executed upon it). Due to the
+ * asynchronous nature of the hardware, an object reported
+ * as busy may become idle before the ioctl is completed.
+ *
+ * Furthermore, if the object is busy, which engine is busy is only
+ * provided as a guide. There are race conditions which prevent the
+ * report of which engines are busy from being always accurate.
+ * However, the converse is not true. If the object is idle, the
+ * result of the ioctl, that all engines are idle, is accurate.
*
* The returned dword is split into two fields to indicate both
* the engines on which the object is being read, and the
* execution engines, e.g. multiple media engines, which are
* mapped to the same identifier in the EXECBUFFER2 ioctl and
* so are not separately reported for busyness.
+ *
+ * Caveat emptor:
+ * Only the boolean result of this query is reliable; that is whether
+ * the object is idle or busy. The report of which engines are busy
+ * should be only used as a heuristic.
*/
__u32 busy;
};
projects / linux-2.6-block.git / commitdiff