linux-block.git
2 years agoMerge branch 'pci/misc'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:47 +0000 (11:28 -0500)]
Merge branch 'pci/misc'

- Tidy setup-irq.c comments (Pranay Sanghai)

- Fix misspellings (Krzysztof Wilczyński)

- Fix sprintf(), sscanf() format mismatches (Krzysztof Wilczyński)

- Tidy cpqphp code formatting (Krzysztof Wilczyński)

- Remove unused pci_pool wrappers, which have been replaced by dma_pool
  (Cai Huoqing)

- Remove a redundant initialization in __pci_reset_function_locked() (Colin
  Ian King)

- Use 'unsigned int' instead of 'unsigned' (Krzysztof Wilczyński)

- Update PCI subsystem information in MAINTAINERS (Krzysztof Wilczyński)

- Include generic <linux/> headers instead of <asm/> for cpqphp and vmd
  (Krzysztof Wilczyński)

* pci/misc:
  PCI: vmd: Drop redundant includes of <asm/device.h>, <asm/msi.h>
  PCI: cpqphp: Use <linux/io.h> instead of <asm/io.h>
  MAINTAINERS: Update PCI subsystem information
  PCI: Prefer 'unsigned int' over bare 'unsigned'
  PCI: Remove redundant 'rc' initialization
  PCI: Remove unused pci_pool wrappers
  PCI: cpqphp: Format if-statement code block correctly
  PCI: Use unsigned to match sscanf("%x") in pci_dev_str_match_path()
  PCI: hv: Remove unnecessary use of %hx
  PCI: Correct misspelled and remove duplicated words
  PCI: Tidy comments

2 years agoMerge branch 'pci/vpd'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:47 +0000 (11:28 -0500)]
Merge branch 'pci/vpd'

- Add pci_read_vpd_any(), pci_write_vpd_any() to access VPD at arbitrary
  offsets (Heiner Kallweit)

- Use VPD API to replace custom code in cxgb3 driver (Heiner Kallweit)

* pci/vpd:
  cxgb3: Remove seeprom_write and use VPD API
  cxgb3: Use VPD API in t3_seeprom_wp()
  cxgb3: Remove t3_seeprom_read and use VPD API
  PCI/VPD: Use pci_read_vpd_any() in pci_vpd_size()
  PCI/VPD: Add pci_read/write_vpd_any()

2 years agoMerge branch 'pci/virtualization'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:46 +0000 (11:28 -0500)]
Merge branch 'pci/virtualization'

- Avoid bus resets on Atheros QCA6174, since they hang (Ingmar Klein)

- Use store and forward mode on Pericom PI7C9X2G switches to avoid ACS
  erratum with ACS P2P Request Redirect (Nathan Rossi)

* pci/virtualization:
  PCI: Add ACS quirk for Pericom PI7C9X2G switches
  PCI: Mark Atheros QCA6174 to avoid bus reset

2 years agoMerge branch 'pci/sysfs'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:46 +0000 (11:28 -0500)]
Merge branch 'pci/sysfs'

- Check for CAP_SYS_ADMIN before validating sysfs user input, not after
  (Krzysztof Wilczyński)

- Always return -EINVAL from sysfs "store" functions for invalid user input
  instead of -EINVAL sometimes and -ERANGE others (Krzysztof Wilczyński)

- Use kstrtobool() directly instead of the strtobool() wrapper (Krzysztof
  Wilczyński)

* pci/sysfs:
  PCI: Use kstrtobool() directly, sans strtobool() wrapper
  PCI/sysfs: Return -EINVAL consistently from "store" functions
  PCI/sysfs: Check CAP_SYS_ADMIN before parsing user input

# Conflicts:
# drivers/pci/iov.c

2 years agoMerge branch 'pci/switchtec'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:45 +0000 (11:28 -0500)]
Merge branch 'pci/switchtec'

- Return error to application when command execution fails because an
  out-of-band reset has cleared the device BARs, Memory Space Enable, etc
  (Kelvin Cao)

- Fix MRPC error status handling issue (Kelvin Cao)

- Mask out other bits when reading of management VEP instance ID (Kelvin
  Cao)

- Return EOPNOTSUPP instead of ENOTSUPP from sysfs show functions (Kelvin
  Cao)

- Add check of event support (Logan Gunthorpe)

* pci/switchtec:
  PCI/switchtec: Add check of event support
  PCI/switchtec: Replace ENOTSUPP with EOPNOTSUPP
  PCI/switchtec: Update the way of getting management VEP instance ID
  PCI/switchtec: Fix a MRPC error status handling issue
  PCI/switchtec: Error out MRPC execution when MMIO reads fail

2 years agoMerge branch 'pci/resource'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:45 +0000 (11:28 -0500)]
Merge branch 'pci/resource'

- Coalesce host bridge contiguous apertures to allow P2P bridge windows
  that span several contiguous host bridge apertures (Kai-Heng Feng)

* pci/resource:
  PCI: Coalesce host bridge contiguous apertures

2 years agoMerge branch 'pci/portdrv'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:45 +0000 (11:28 -0500)]
Merge branch 'pci/portdrv'

- Don't setup portdrv IRQs if there are no port drivers that use them, to
  conserve vectors and avoid spurious events (Jan Kiszka)

* pci/portdrv:
  PCI/portdrv: Do not setup up IRQs if there are no users

2 years agoMerge branch 'pci/p2pdma'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:44 +0000 (11:28 -0500)]
Merge branch 'pci/p2pdma'

- Apply bus offset correctly in DMA address calculation, which used the
  wrong sign before (Wang Lu)

* pci/p2pdma:
  PCI/P2PDMA: Apply bus offset correctly in DMA address calculation

2 years agoMerge branch 'pci/msi'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:44 +0000 (11:28 -0500)]
Merge branch 'pci/msi'

- Document sysfs "irq" attribute, which contains either the INTx IRQ (the
  intended behavior) or the first MSI IRQ (historical mistake retained for
  backwards compatibility) (Barry Song)

- Rework "irq" sysfs show function to explicitly fetch first MSI IRQ
  instead of depending on core to put it in dev->irq, to enable future core
  cleanup (Barry Song)

* pci/msi:
  PCI/sysfs: Explicitly show first MSI IRQ for 'irq'
  PCI: Document /sys/bus/pci/devices/.../irq

2 years agoMerge branch 'pci/hotplug'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:43 +0000 (11:28 -0500)]
Merge branch 'pci/hotplug'

- Ignore Link Down/Up caused by error-induced Hot Reset so endpoint driver
  can remain bound to device during error recovery (Lukas Wunner)

- Remove unused resume err_handler (Lukas Wunner)

- Remove unused pcie_port_bus_{,un}register() declarations (Lukas Wunner)

- Skip compiling err.c when CONFIG_PCIEAER not set (Lukas Wunner)

* pci/hotplug:
  PCI/ERR: Reduce compile time for CONFIG_PCIEAER=n
  PCI/portdrv: Remove unused pcie_port_bus_{,un}register() declarations
  PCI/portdrv: Remove unused resume err_handler
  PCI: pciehp: Ignore Link Down/Up caused by error-induced Hot Reset
  PCI/portdrv: Rename pm_iter() to pcie_port_device_iter()

2 years agoMerge branch 'pci/driver'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:43 +0000 (11:28 -0500)]
Merge branch 'pci/driver'

- Drop the struct pci_dev.driver pointer, which is redundant with the
  struct device.driver pointer (Uwe Kleine-König)

* pci/driver:
  PCI: Remove struct pci_dev->driver
  PCI: Use to_pci_driver() instead of pci_dev->driver
  x86/pci/probe_roms: Use to_pci_driver() instead of pci_dev->driver
  perf/x86/intel/uncore: Use to_pci_driver() instead of pci_dev->driver
  powerpc/eeh: Use to_pci_driver() instead of pci_dev->driver
  usb: xhci: Use to_pci_driver() instead of pci_dev->driver
  cxl: Use to_pci_driver() instead of pci_dev->driver
  cxl: Factor out common dev->driver expressions
  xen/pcifront: Use to_pci_driver() instead of pci_dev->driver
  xen/pcifront: Drop pcifront_common_process() tests of pcidev, pdrv
  nfp: use dev_driver_string() instead of pci_dev->driver->name
  mlxsw: pci: Use dev_driver_string() instead of pci_dev->driver->name
  net: marvell: prestera: use dev_driver_string() instead of pci_dev->driver->name
  net: hns3: use dev_driver_string() instead of pci_dev->driver->name
  crypto: hisilicon - use dev_driver_string() instead of pci_dev->driver->name
  powerpc/eeh: Use dev_driver_string() instead of struct pci_dev->driver->name
  ssb: Use dev_driver_string() instead of pci_dev->driver->name
  bcma: simplify reference to driver name
  crypto: qat - simplify adf_enable_aer()
  scsi: message: fusion: Remove unused mpt_pci driver .probe() 'id' parameter
  PCI/ERR: Factor out common dev->driver expressions
  PCI: Drop pci_device_probe() test of !pci_dev->driver
  PCI: Drop pci_device_remove() test of pci_dev->driver
  PCI: Return NULL for to_pci_driver(NULL)

2 years agoMerge branch 'pci/enumeration'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:42 +0000 (11:28 -0500)]
Merge branch 'pci/enumeration'

- Rename pcibios_add_device() to pcibios_device_add() since it's called
  from pci_device_add() (Oliver O'Halloran)

- Don't try to enable AtomicOps on VFs, since they can only be enabled on
  the PF (Selvin Xavier)

* pci/enumeration:
  PCI: Do not enable AtomicOps on VFs
  PCI: Rename pcibios_add_device() to pcibios_device_add()

2 years agoMerge branch 'pci/aspm'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:42 +0000 (11:28 -0500)]
Merge branch 'pci/aspm'

- Re-enable LTR in Downstream Ports after it has been disabled by reset or
  hotplug to allow use of ASPM L1.2 again and prevent Unsupported Request
  errors when Endpoint sends LTR messages (Mingchuang Qiao)

* pci/aspm:
  PCI: Re-enable Downstream Port LTR after reset or hotplug

2 years agoMerge branch 'pci/acpi'
Bjorn Helgaas [Fri, 5 Nov 2021 16:28:42 +0000 (11:28 -0500)]
Merge branch 'pci/acpi'

- Simplify _OSC negotiation with platform for control of PCIe features
  (Joerg Roedel)

* pci/acpi:
  PCI/ACPI: Check for _OSC support in acpi_pci_osc_control_set()
  PCI/ACPI: Move _OSC query checks to separate function
  PCI/ACPI: Move supported and control calculations to separate functions
  PCI/ACPI: Remove OSC_PCI_SUPPORT_MASKS and OSC_PCI_CONTROL_MASKS

2 years agoPCI: Add ACS quirk for Pericom PI7C9X2G switches
Nathan Rossi [Fri, 10 Sep 2021 02:58:23 +0000 (02:58 +0000)]
PCI: Add ACS quirk for Pericom PI7C9X2G switches

The Pericom PI7C9X2G404/PI7C9X2G304/PI7C9X2G303 PCIe switches have an
erratum for ACS P2P Request Redirect behaviour when used in the cut-through
forwarding mode. The recommended work around for this issue is to use the
switch in store and forward mode. The erratum results in packets being
queued and not being delivered upstream, which can be observed as very poor
downstream device performance and/or dropped device-generated
data/interrupts.

Add a fixup so that when enabling or resuming the downstream port we check
if it has enabled ACS P2P Request Redirect, and if so, change the device
(via the upstream port) to use the store and forward operating mode.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=177471
Link: https://lore.kernel.org/r/20210910025823.196508-1-nathan@nathanrossi.com
Tested-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Nathan Rossi <nathan.rossi@digi.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: Do not enable AtomicOps on VFs
Selvin Xavier [Sat, 11 Sep 2021 10:03:05 +0000 (03:03 -0700)]
PCI: Do not enable AtomicOps on VFs

Host crashes when pci_enable_atomic_ops_to_root() is called for VFs with
virtual buses. The virtual buses added to SR-IOV have bus->self set to NULL
and host crashes due to this.

  PID: 4481   TASK: ffff89c6941b0000  CPU: 53  COMMAND: "bash"
  ...
   #3 [ffff9a9481713808] oops_end at ffffffffb9025cd6
   #4 [ffff9a9481713828] page_fault_oops at ffffffffb906e417
   #5 [ffff9a9481713888] exc_page_fault at ffffffffb9a0ad14
   #6 [ffff9a94817138b0] asm_exc_page_fault at ffffffffb9c00ace
      [exception RIP: pcie_capability_read_dword+28]
      RIP: ffffffffb952fd5c  RSP: ffff9a9481713960  RFLAGS: 00010246
      RAX: 0000000000000001  RBX: ffff89c6b1096000  RCX: 0000000000000000
      RDX: ffff9a9481713990  RSI: 0000000000000024  RDI: 0000000000000000
      RBP: 0000000000000080   R8: 0000000000000008   R9: ffff89c64341a2f8
      R10: 0000000000000002  R11: 0000000000000000  R12: ffff89c648bab000
      R13: 0000000000000000  R14: 0000000000000000  R15: ffff89c648bab0c8
      ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018
   #7 [ffff9a9481713988] pci_enable_atomic_ops_to_root at ffffffffb95359a6
   #8 [ffff9a94817139c0] bnxt_qplib_determine_atomics at ffffffffc08c1a33 [bnxt_re]
   #9 [ffff9a94817139d0] bnxt_re_dev_init at ffffffffc08ba2d1 [bnxt_re]

Per PCIe r5.0, sec 9.3.5.10, the AtomicOp Requester Enable bit in Device
Control 2 is reserved for VFs.  The PF value applies to all associated VFs.

Return -EINVAL if pci_enable_atomic_ops_to_root() is called for a VF.

Link: https://lore.kernel.org/r/1631354585-16597-1-git-send-email-selvin.xavier@broadcom.com
Fixes: 35f5ace5dea4 ("RDMA/bnxt_re: Enable global atomic ops if platform supports")
Fixes: 430a23689dea ("PCI: Add pci_enable_atomic_ops_to_root()")
Signed-off-by: Selvin Xavier <selvin.xavier@broadcom.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Andy Gospodarek <gospo@broadcom.com>
2 years agoPCI: vmd: Drop redundant includes of <asm/device.h>, <asm/msi.h>
Krzysztof Wilczyński [Wed, 13 Oct 2021 00:31:44 +0000 (00:31 +0000)]
PCI: vmd: Drop redundant includes of <asm/device.h>, <asm/msi.h>

We already include <linux/device.h> and <linux/msi.h>, which
include <asm/device.h> and <asm/msi.h>.

Drop the redundant includes of <asm/device.h> and <asm/msi.h>.

[bhelgaas: squash in fix from Wan Jiabing <wanjiabing@vivo.com>:
https://lore.kernel.org/r/20211104063720.29375-1-wanjiabing@vivo.com]
Link: https://lore.kernel.org/r/20211013003145.1107148-1-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Jonathan Derrick <jonathan.derrick@linux.dev>
2 years agoPCI: cpqphp: Use <linux/io.h> instead of <asm/io.h>
Krzysztof Wilczyński [Wed, 13 Oct 2021 00:31:45 +0000 (00:31 +0000)]
PCI: cpqphp: Use <linux/io.h> instead of <asm/io.h>

Use the preferred generic header file linux/io.h that already includes the
corresponding asm/io.h file.

Link: https://lore.kernel.org/r/20211013003145.1107148-2-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Jonathan Derrick <jonathan.derrick@linux.dev>
2 years agoMAINTAINERS: Update PCI subsystem information
Krzysztof Wilczyński [Wed, 27 Oct 2021 10:50:41 +0000 (10:50 +0000)]
MAINTAINERS: Update PCI subsystem information

Update the following information related to the PCI subsystem which
includes the PCI drivers, PCI native host bridge and endpoint drivers,
and the PCI endpoint sub-system:

 - Sort fields as per preferred order
 - Sort files in the alphabetical order
 - Update old Patchwork URLs
 - Update Git repository for the PCI endpoint subsystem
 - Add Bugzilla link
 - Add link to the official IRC channel
 - Add files "drivers/pci/pci-bridge-emul.{c,h}" to the right
   section so that proper ownership is returned for both files
   from the get_maintainer.pl script

Link: https://lore.kernel.org/r/20211027105041.24087-1-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: Prefer 'unsigned int' over bare 'unsigned'
Krzysztof Wilczyński [Wed, 13 Oct 2021 01:41:36 +0000 (01:41 +0000)]
PCI: Prefer 'unsigned int' over bare 'unsigned'

The bare "unsigned" type implicitly means "unsigned int", but the preferred
coding style is to use the complete type name.

Update the bare use of "unsigned" to the preferred "unsigned int".

No change to functionality intended.

See a1ce18e4f941 ("checkpatch: warn on bare unsigned or signed declarations
without int").

Link: https://lore.kernel.org/r/20211013014136.1117543-1-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: Remove redundant 'rc' initialization
Colin Ian King [Fri, 10 Sep 2021 16:14:17 +0000 (17:14 +0100)]
PCI: Remove redundant 'rc' initialization

The variable 'rc' is being initialized with a value that is never read.
Remove the redundant assignment.

Addresses-Coverity: ("Unused value")
Link: https://lore.kernel.org/r/20210910161417.91001-1-colin.king@canonical.com
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Krzysztof Wilczyński <kw@linux.com>
2 years agocxgb3: Remove seeprom_write and use VPD API
Heiner Kallweit [Fri, 10 Sep 2021 06:27:08 +0000 (08:27 +0200)]
cxgb3: Remove seeprom_write and use VPD API

Using the VPD API allows to simplify the code and completely get
rid of t3_seeprom_write().

Link: https://lore.kernel.org/r/a0291004-dda3-ea08-4d6c-a2f8826c8527@gmail.com
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
2 years agocxgb3: Use VPD API in t3_seeprom_wp()
Heiner Kallweit [Fri, 10 Sep 2021 06:25:49 +0000 (08:25 +0200)]
cxgb3: Use VPD API in t3_seeprom_wp()

Use standard VPD API to replace t3_seeprom_write(), this prepares for
removing this function. Chelsio T3 maps the EEPROM write protect flag
to an arbitrary place in VPD address space, therefore we have to use
pci_write_vpd_any().

Link: https://lore.kernel.org/r/f768fdbe-3a16-d539-57d2-c7c908294336@gmail.com
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
2 years agocxgb3: Remove t3_seeprom_read and use VPD API
Heiner Kallweit [Fri, 10 Sep 2021 06:24:07 +0000 (08:24 +0200)]
cxgb3: Remove t3_seeprom_read and use VPD API

Using the VPD API allows to simplify the code and completely get rid
of t3_seeprom_read(). Note that we don't have to use pci_read_vpd_any()
here because a VPD quirk sets dev->vpd.len to the full EEPROM size.

Tested with a T320 card.

Link: https://lore.kernel.org/r/68ef15bb-b6bf-40ad-160c-aaa72c4a70f8@gmail.com
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
2 years agoPCI/VPD: Use pci_read_vpd_any() in pci_vpd_size()
Heiner Kallweit [Fri, 10 Sep 2021 06:22:49 +0000 (08:22 +0200)]
PCI/VPD: Use pci_read_vpd_any() in pci_vpd_size()

Use new function pci_read_vpd_any() to simplify the code.

[bhelgaas: squash in fix for stack overflow reported & tested by
Qian [1] and Kunihiko [2]:
[1] https://lore.kernel.org/netdev/e89087c5-c495-c5ca-feb1-54cf3a8775c5@quicinc.com/
[2] https://lore.kernel.org/r/2f7e3770-ab47-42b5-719c-f7c661c07d28@socionext.com
Link: https://lore.kernel.org/r/6211be8a-5d10-8f3a-6d33-af695dc35caf@gmail.com
Reported-by: Qian Cai <quic_qiancai@quicinc.com>
Tested-by: Qian Cai <quic_qiancai@quicinc.com>
Reported-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
Tested-by: Kunihiko Hayashi <hayashi.kunihiko@socionext.com>
]

Link: https://lore.kernel.org/r/049fa71c-c7af-9c69-51c0-05c1bc2bf660@gmail.com
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
2 years agoPCI: Re-enable Downstream Port LTR after reset or hotplug
Mingchuang Qiao [Tue, 12 Oct 2021 07:56:14 +0000 (15:56 +0800)]
PCI: Re-enable Downstream Port LTR after reset or hotplug

Per PCIe r5.0, sec 7.5.3.16, Downstream Ports must disable LTR if the link
goes down (the Port goes DL_Down status).  This is a problem because the
Downstream Port's dev->ltr_path is still set, so we think LTR is still
enabled, and we enable LTR in the Endpoint.  When it sends LTR messages,
they cause Unsupported Request errors at the Downstream Port.

This happens in the reset path, where we may enable LTR in
pci_restore_pcie_state() even though the Downstream Port disabled LTR
because the reset caused a link down event.

It also happens in the hot-remove and hot-add path, where we may enable LTR
in pci_configure_ltr() even though the Downstream Port disabled LTR when
the hot-remove took the link down.

In these two scenarios, check the upstream bridge and restore its LTR
enable if appropriate.

The Unsupported Request may be logged by AER as follows:

  pcieport 0000:00:1d.0: AER: Uncorrected (Non-Fatal) error received: id=00e8
  pcieport 0000:00:1d.0: PCIe Bus Error: severity=Uncorrected (Non-Fatal), type=Transaction Layer, id=00e8(Requester ID)
  pcieport 0000:00:1d.0:   device [8086:9d18] error status/mask=00100000/00010000
  pcieport 0000:00:1d.0:    [20] Unsupported Request    (First)

In addition, if LTR is not configured correctly, the link cannot enter the
L1.2 state, which prevents some machines from entering the S0ix low power
state.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20211012075614.54576-1-mingchuang.qiao@mediatek.com
Reported-by: Utkarsh H Patel <utkarsh.h.patel@intel.com>
Signed-off-by: Mingchuang Qiao <mingchuang.qiao@mediatek.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
2 years agoPCI/sysfs: Explicitly show first MSI IRQ for 'irq'
Barry Song [Wed, 25 Aug 2021 10:26:35 +0000 (18:26 +0800)]
PCI/sysfs: Explicitly show first MSI IRQ for 'irq'

The sysfs "irq" file contains the legacy INTx IRQ.  Or, if the device has
MSI enabled, it contains the first MSI IRQ instead.

Previously this file showed the pci_dev.irq value directly.  But we'd
prefer to use pci_dev.irq only for the INTx IRQ and decouple that from any
MSI or MSI-X IRQs.

If the device has MSI enabled, explicitly look up and show the first MSI
IRQ in the sysfs "irq" file.  Otherwise, show the INTx IRQ.

This removes the requirement that msi_capability_init() set pci_dev.irq to
the first MSI IRQ when enabling MSI and pci_msi_shutdown() restore the INTx
IRQ when disabling MSI.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20210825102636.52757-3-21cnbao@gmail.com
Signed-off-by: Barry Song <song.bao.hua@hisilicon.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: Document /sys/bus/pci/devices/.../irq
Barry Song [Wed, 25 Aug 2021 10:26:34 +0000 (18:26 +0800)]
PCI: Document /sys/bus/pci/devices/.../irq

Document /sys/bus/pci/devices/.../irq.

This file contains the IRQ of the INTx interrupt (or zero if the device
doesn't support INTx interrupts).

If the device has enabled MSI (not MSI-X), it contains the first MSI IRQ
instead.  This is a historical mistake because devices may support several
MSI or MSI-X vectors, and this file can't contain them all.  But we
preserve this behavior to avoid breaking userspace.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20210825102636.52757-2-21cnbao@gmail.com
Signed-off-by: Barry Song <song.bao.hua@hisilicon.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2 years agoPCI: Remove unused pci_pool wrappers
Cai Huoqing [Mon, 18 Oct 2021 12:41:09 +0000 (20:41 +0800)]
PCI: Remove unused pci_pool wrappers

The pci_pool users have been converted to dma_pool.  Remove the unused
pci_pool wrappers.

Link: https://lore.kernel.org/r/20211018124110.214-1-caihuoqing@baidu.com
Signed-off-by: Cai Huoqing <caihuoqing@baidu.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: Remove struct pci_dev->driver
Uwe Kleine-König [Mon, 4 Oct 2021 12:59:35 +0000 (14:59 +0200)]
PCI: Remove struct pci_dev->driver

There are no remaining uses of the struct pci_dev->driver pointer, so
remove it.

Link: https://lore.kernel.org/r/20211004125935.2300113-12-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: Use to_pci_driver() instead of pci_dev->driver
Uwe Kleine-König [Tue, 12 Oct 2021 21:11:24 +0000 (16:11 -0500)]
PCI: Use to_pci_driver() instead of pci_dev->driver

Struct pci_driver contains a struct device_driver, so for PCI devices, it's
easy to convert a device_driver * to a pci_driver * with to_pci_driver().
The device_driver * is in struct device, so we don't need to also keep
track of the pci_driver * in struct pci_dev.

Replace pci_dev->driver with to_pci_driver().  This is a step toward
removing pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-11-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agox86/pci/probe_roms: Use to_pci_driver() instead of pci_dev->driver
Uwe Kleine-König [Tue, 12 Oct 2021 21:05:47 +0000 (16:05 -0500)]
x86/pci/probe_roms: Use to_pci_driver() instead of pci_dev->driver

Struct pci_driver contains a struct device_driver, so for PCI devices, it's
easy to convert a device_driver * to a pci_driver * with to_pci_driver().
The device_driver * is in struct device, so we don't need to also keep
track of the pci_driver * in struct pci_dev.

Replace pdev->driver with to_pci_driver().  This is a step toward removing
pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-11-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoperf/x86/intel/uncore: Use to_pci_driver() instead of pci_dev->driver
Uwe Kleine-König [Tue, 12 Oct 2021 21:38:32 +0000 (16:38 -0500)]
perf/x86/intel/uncore: Use to_pci_driver() instead of pci_dev->driver

Struct pci_driver contains a struct device_driver, so for PCI devices, it's
easy to convert a device_driver * to a pci_driver * with to_pci_driver().
The device_driver * is in struct device, so we don't need to also keep
track of the pci_driver * in struct pci_dev.

Replace pdev->driver with to_pci_driver().  This is a step toward removing
pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-11-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agopowerpc/eeh: Use to_pci_driver() instead of pci_dev->driver
Uwe Kleine-König [Tue, 12 Oct 2021 21:37:55 +0000 (16:37 -0500)]
powerpc/eeh: Use to_pci_driver() instead of pci_dev->driver

Struct pci_driver contains a struct device_driver, so for PCI devices, it's
easy to convert a device_driver * to a pci_driver * with to_pci_driver().
The device_driver * is in struct device, so we don't need to also keep
track of the pci_driver * in struct pci_dev.

Replace pdev->driver with to_pci_driver().  This is a step toward removing
pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-11-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agousb: xhci: Use to_pci_driver() instead of pci_dev->driver
Uwe Kleine-König [Tue, 12 Oct 2021 21:36:05 +0000 (16:36 -0500)]
usb: xhci: Use to_pci_driver() instead of pci_dev->driver

Struct pci_driver contains a struct device_driver, so for PCI devices, it's
easy to convert a device_driver * to a pci_driver * with to_pci_driver().
The device_driver * is in struct device, so we don't need to also keep
track of the pci_driver * in struct pci_dev.

Replace pdev->driver with to_pci_driver().  This is a step toward removing
pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-11-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agocxl: Use to_pci_driver() instead of pci_dev->driver
Uwe Kleine-König [Tue, 12 Oct 2021 21:03:39 +0000 (16:03 -0500)]
cxl: Use to_pci_driver() instead of pci_dev->driver

Struct pci_driver contains a struct device_driver, so for PCI devices, it's
easy to convert a device_driver * to a pci_driver * with to_pci_driver().
The device_driver * is in struct device, so we don't need to also keep
track of the pci_driver * in struct pci_dev.

Replace pdev->driver with to_pci_driver().  This is a step toward removing
pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-11-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agocxl: Factor out common dev->driver expressions
Bjorn Helgaas [Tue, 12 Oct 2021 20:51:32 +0000 (15:51 -0500)]
cxl: Factor out common dev->driver expressions

Save the struct pci_driver and struct pci_error_handlers pointers from
pdev->driver instead of chasing the pointers several times.  No functional
change intended.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI/ERR: Reduce compile time for CONFIG_PCIEAER=n
Lukas Wunner [Sat, 31 Jul 2021 12:39:04 +0000 (14:39 +0200)]
PCI/ERR: Reduce compile time for CONFIG_PCIEAER=n

The sole non-static function in err.c, pcie_do_recovery(), is only
called from:

* aer.c (if CONFIG_PCIEAER=y)
* dpc.c (if CONFIG_PCIE_DPC=y, which depends on CONFIG_PCIEAER)
* edr.c (if CONFIG_PCIE_EDR=y, which depends on CONFIG_PCIE_DPC)

Thus, err.c need not be compiled if CONFIG_PCIEAER=n.

Also, pci_uevent_ers() and pcie_clear_device_status(), which are called
from err.c, can be #ifdef'ed away unless CONFIG_PCIEAER=y.

Since x86_64_defconfig doesn't enable CONFIG_PCIEAER, this change may
slightly reduce compile time for anyone doing a test build with that
config.

Link: https://lore.kernel.org/r/98f9041151268c1c035ab64cca320ad86803f64a.1627638184.git.lukas@wunner.de
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI/portdrv: Remove unused pcie_port_bus_{,un}register() declarations
Lukas Wunner [Sat, 31 Jul 2021 12:39:03 +0000 (14:39 +0200)]
PCI/portdrv: Remove unused pcie_port_bus_{,un}register() declarations

Commit c6c889d932bb ("PCI/portdrv: Remove pcie_port_bus_type link order
dependency") removed pcie_port_bus_{,un}register() but erroneously
retained their declarations in portdrv.h.  Remove them as well.

Link: https://lore.kernel.org/r/7fd76b0591c37287ab94d911d8fd9ab9a2afcd16.1627638184.git.lukas@wunner.de
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI/portdrv: Remove unused resume err_handler
Lukas Wunner [Sat, 31 Jul 2021 12:39:02 +0000 (14:39 +0200)]
PCI/portdrv: Remove unused resume err_handler

Commit 3e41a317ae45 ("PCI/AER: Remove unused aer_error_resume()")
removed the resume err_handler from AER.  Since no other port service
implements the callback, support for it can be removed from portdrv.
It can be revived later if need be, preferably by re-using the
pcie_port_device_iter() iterator.

Link: https://lore.kernel.org/r/25334149b604e005058aeb0fdf51e01f991d5d74.1627638184.git.lukas@wunner.de
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: Keith Busch <kbusch@kernel.org>
2 years agoPCI: pciehp: Ignore Link Down/Up caused by error-induced Hot Reset
Lukas Wunner [Sat, 31 Jul 2021 12:39:01 +0000 (14:39 +0200)]
PCI: pciehp: Ignore Link Down/Up caused by error-induced Hot Reset

Stuart Hayes reports that an error handled by DPC at a Root Port results
in pciehp gratuitously bringing down a subordinate hotplug port:

  RP -- UP -- DP -- UP -- DP (hotplug) -- EP

pciehp brings the slot down because the Link to the Endpoint goes down.
That is caused by a Hot Reset being propagated as a result of DPC.
Per PCIe Base Spec 5.0, section 6.6.1 "Conventional Reset":

  For a Switch, the following must cause a hot reset to be sent on all
  Downstream Ports: [...]

  * The Data Link Layer of the Upstream Port reporting DL_Down status.
    In Switches that support Link speeds greater than 5.0 GT/s, the
    Upstream Port must direct the LTSSM of each Downstream Port to the
    Hot Reset state, but not hold the LTSSMs in that state. This permits
    each Downstream Port to begin Link training immediately after its
    hot reset completes. This behavior is recommended for all Switches.

  * Receiving a hot reset on the Upstream Port.

Once DPC recovers, pcie_do_recovery() walks down the hierarchy and
invokes pcie_portdrv_slot_reset() to restore each port's config space.
At that point, a hotplug interrupt is signaled per PCIe Base Spec r5.0,
section 6.7.3.4 "Software Notification of Hot-Plug Events":

  If the Port is enabled for edge-triggered interrupt signaling using
  MSI or MSI-X, an interrupt message must be sent every time the logical
  AND of the following conditions transitions from FALSE to TRUE: [...]

  * The Hot-Plug Interrupt Enable bit in the Slot Control register is
    set to 1b.

  * At least one hot-plug event status bit in the Slot Status register
    and its associated enable bit in the Slot Control register are both
    set to 1b.

Prevent pciehp from gratuitously bringing down the slot by clearing the
error-induced Data Link Layer State Changed event before restoring
config space.  Afterwards, check whether the link has unexpectedly
failed to retrain and synthesize a DLLSC event if so.

Allow each pcie_port_service_driver (one of them being pciehp) to define
a slot_reset callback and re-use the existing pm_iter() function to
iterate over the callbacks.

Thereby, the Endpoint driver remains bound throughout error recovery and
may restore the device to working state.

Surprise removal during error recovery is detected through a Presence
Detect Changed event.  The hotplug port is expected to not signal that
event as a result of a Hot Reset.

The issue isn't DPC-specific, it also occurs when an error is handled by
AER through aer_root_reset().  So while the issue was noticed only now,
it's been around since 2006 when AER support was first introduced.

[bhelgaas: drop PCI_ERROR_RECOVERY Kconfig, split pm_iter() rename to
preparatory patch]
Link: https://lore.kernel.org/linux-pci/08c046b0-c9f2-3489-eeef-7e7aca435bb9@gmail.com/
Fixes: 6c2b374d7485 ("PCI-Express AER implemetation: AER core and aerdriver")
Link: https://lore.kernel.org/r/251f4edcc04c14f873ff1c967bc686169cd07d2d.1627638184.git.lukas@wunner.de
Reported-by: Stuart Hayes <stuart.w.hayes@gmail.com>
Tested-by: Stuart Hayes <stuart.w.hayes@gmail.com>
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@vger.kernel.org # v2.6.19+: ba952824e6c1: PCI/portdrv: Report reset for frozen channel
Cc: Keith Busch <kbusch@kernel.org>
2 years agoPCI/portdrv: Rename pm_iter() to pcie_port_device_iter()
Lukas Wunner [Fri, 15 Oct 2021 18:58:40 +0000 (13:58 -0500)]
PCI/portdrv: Rename pm_iter() to pcie_port_device_iter()

Rename pm_iter() to pcie_port_device_iter() and make it visible outside
CONFIG_PM and portdrv_core.c so it can be used for pciehp slot reset
recovery.

[bhelgaas: split into its own patch]
Link: https://lore.kernel.org/linux-pci/08c046b0-c9f2-3489-eeef-7e7aca435bb9@gmail.com/
Link: https://lore.kernel.org/r/251f4edcc04c14f873ff1c967bc686169cd07d2d.1627638184.git.lukas@wunner.de
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoxen/pcifront: Use to_pci_driver() instead of pci_dev->driver
Uwe Kleine-König [Tue, 12 Oct 2021 21:35:34 +0000 (16:35 -0500)]
xen/pcifront: Use to_pci_driver() instead of pci_dev->driver

Struct pci_driver contains a struct device_driver, so for PCI devices, it's
easy to convert a device_driver * to a pci_driver * with to_pci_driver().
The device_driver * is in struct device, so we don't need to also keep
track of the pci_driver * in struct pci_dev.

Replace pdev->driver with to_pci_driver().  This is a step toward removing
pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-11-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
2 years agoxen/pcifront: Drop pcifront_common_process() tests of pcidev, pdrv
Uwe Kleine-König [Fri, 8 Oct 2021 23:22:08 +0000 (18:22 -0500)]
xen/pcifront: Drop pcifront_common_process() tests of pcidev, pdrv

pcifront_common_process() exits early if pcidev or pcidev->driver are NULL,
so simplify it by not checking them again.

[bhelgaas: split flag change to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-4-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoPCI/switchtec: Add check of event support
Logan Gunthorpe [Thu, 14 Oct 2021 14:18:59 +0000 (14:18 +0000)]
PCI/switchtec: Add check of event support

Not all events are supported by every gen/variant of the Switchtec
firmware. To solve this, since Gen4, a new bit in each event header
is introduced to indicate if an event is supported by the firmware.

Link: https://lore.kernel.org/r/20211014141859.11444-6-kelvin.cao@microchip.com
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Kelvin Cao <kelvin.cao@microchip.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI/switchtec: Replace ENOTSUPP with EOPNOTSUPP
Kelvin Cao [Thu, 14 Oct 2021 14:18:58 +0000 (14:18 +0000)]
PCI/switchtec: Replace ENOTSUPP with EOPNOTSUPP

ENOTSUPP is not a SUSV4 error code, and the following checkpatch.pl
warning will be given for new patches which still use ENOTSUPP.

    WARNING: ENOTSUPP is not a SUSV4 error code, prefer EOPNOTSUPP

See the link below for the discussion.

  https://lore.kernel.org/netdev/20200511165319.2251678-1-kuba@kernel.org/

Replace ENOTSUPP with EOPNOTSUPP to align with future patches which will
be using EOPNOTSUPP.

Link: https://lore.kernel.org/r/20211014141859.11444-5-kelvin.cao@microchip.com
Signed-off-by: Kelvin Cao <kelvin.cao@microchip.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI/switchtec: Update the way of getting management VEP instance ID
Kelvin Cao [Thu, 14 Oct 2021 14:18:57 +0000 (14:18 +0000)]
PCI/switchtec: Update the way of getting management VEP instance ID

Gen4 firmware adds DMA VEP and NVMe VEP support in VEP (virtual EP)
instance ID register in addtion to management EP, update the way of
getting management VEP instance ID.

Link: https://lore.kernel.org/r/20211014141859.11444-4-kelvin.cao@microchip.com
Signed-off-by: Kelvin Cao <kelvin.cao@microchip.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI/switchtec: Fix a MRPC error status handling issue
Kelvin Cao [Thu, 14 Oct 2021 14:18:56 +0000 (14:18 +0000)]
PCI/switchtec: Fix a MRPC error status handling issue

If an error is encountered when executing a MRPC command, the firmware
will set the status register to SWITCHTEC_MRPC_STATUS_ERROR and return
the error code in the return value register.

Add handling of SWITCHTEC_MRPC_STATUS_ERROR on status register when
completing a MRPC command.

Link: https://lore.kernel.org/r/20211014141859.11444-3-kelvin.cao@microchip.com
Signed-off-by: Kelvin Cao <kelvin.cao@microchip.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI/switchtec: Error out MRPC execution when MMIO reads fail
Kelvin Cao [Thu, 14 Oct 2021 14:18:55 +0000 (14:18 +0000)]
PCI/switchtec: Error out MRPC execution when MMIO reads fail

A firmware hard reset may be initiated by various mechanisms including a
UART interface, TWI sideband interface from BMC, MRPC command from
userspace, etc. The switchtec management driver is unaware of these
resets.

The reset clears PCI state including the BARs and Memory Space Enable
bits, so the device no longer responds to the MMIO accesses the driver
uses to operate it.

MMIO reads to the device will fail with a PCIe error. When the root
complex handles that error, it typically fabricates ~0 data to complete
the CPU read.

Check for this sort of error by reading the device ID from MMIO space.
This ID can never be ~0, so if we see that value, it probably means the
PCIe Memory Read failed and we should return an error indication to the
application using the switchtec driver.

Link: https://lore.kernel.org/r/20211014141859.11444-2-kelvin.cao@microchip.com
Signed-off-by: Kelvin Cao <kelvin.cao@microchip.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: cpqphp: Format if-statement code block correctly
Krzysztof Wilczyński [Wed, 13 Oct 2021 01:14:12 +0000 (01:14 +0000)]
PCI: cpqphp: Format if-statement code block correctly

The code block related to the if-statement in cpqhp_set_irq() is
somewhat awkwardly formatted making the code hard to read.

Make it readable.  No change to functionality intended.

Link: https://lore.kernel.org/r/20211013011412.1110829-1-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agonfp: use dev_driver_string() instead of pci_dev->driver->name
Uwe Kleine-König [Tue, 12 Oct 2021 22:55:03 +0000 (17:55 -0500)]
nfp: use dev_driver_string() instead of pci_dev->driver->name

Replace dev->driver_name() by dev_driver_string() for the corresponding
struct device.  This is a step toward removing pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-8-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Simon Horman <simon.horman@corigine.com>
2 years agomlxsw: pci: Use dev_driver_string() instead of pci_dev->driver->name
Uwe Kleine-König [Tue, 12 Oct 2021 22:54:37 +0000 (17:54 -0500)]
mlxsw: pci: Use dev_driver_string() instead of pci_dev->driver->name

Replace dev->driver_name() by dev_driver_string() for the corresponding
struct device.  This is a step toward removing pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-8-u.kleine-koenig@pengutronix.de
Tested-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
2 years agonet: marvell: prestera: use dev_driver_string() instead of pci_dev->driver->name
Uwe Kleine-König [Tue, 12 Oct 2021 22:53:45 +0000 (17:53 -0500)]
net: marvell: prestera: use dev_driver_string() instead of pci_dev->driver->name

Replace dev->driver_name() by dev_driver_string() for the corresponding
struct device.  This is a step toward removing pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-8-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agonet: hns3: use dev_driver_string() instead of pci_dev->driver->name
Uwe Kleine-König [Tue, 12 Oct 2021 22:52:59 +0000 (17:52 -0500)]
net: hns3: use dev_driver_string() instead of pci_dev->driver->name

Replace dev->driver_name() by dev_driver_string() for the corresponding
struct device.  This is a step toward removing pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-8-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agocrypto: hisilicon - use dev_driver_string() instead of pci_dev->driver->name
Uwe Kleine-König [Tue, 12 Oct 2021 22:51:53 +0000 (17:51 -0500)]
crypto: hisilicon - use dev_driver_string() instead of pci_dev->driver->name

Replace dev->driver_name() by dev_driver_string() for the corresponding
struct device.  This is a step toward removing pci_dev->driver.

[bhelgaas: split to separate patch]
Link: https://lore.kernel.org/r/20211004125935.2300113-8-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agopowerpc/eeh: Use dev_driver_string() instead of struct pci_dev->driver->name
Uwe Kleine-König [Mon, 4 Oct 2021 12:59:29 +0000 (14:59 +0200)]
powerpc/eeh: Use dev_driver_string() instead of struct pci_dev->driver->name

Replace pdev->driver->name by dev_driver_string() for the corresponding
struct device.  This is a step toward removing pci_dev->driver.

Move the function nearer its only user and instead of the ?: operator use a
normal "if" which is more readable.

Link: https://lore.kernel.org/r/20211004125935.2300113-6-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agossb: Use dev_driver_string() instead of pci_dev->driver->name
Uwe Kleine-König [Mon, 4 Oct 2021 12:59:30 +0000 (14:59 +0200)]
ssb: Use dev_driver_string() instead of pci_dev->driver->name

All drivers that use ssb_pcihost_probe(), i.e., b43_pci_bridge_driver and
b44_pci_driver, set the pci_driver.name, and __pci_register_driver() sets
the struct driver.name member to the same value.

Replace dev->driver_name() by dev_driver_string() for the corresponding
struct device.  This is a step toward removing pci_dev->driver.

Link: https://lore.kernel.org/r/20211004125935.2300113-7-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Michael Büsch <m@bues.ch>
2 years agobcma: simplify reference to driver name
Uwe Kleine-König [Mon, 4 Oct 2021 12:59:28 +0000 (14:59 +0200)]
bcma: simplify reference to driver name

When bcma_host_pci_probe() is called, the PCI driver core has already
assigned the device's driver in local_pci_probe(). So dev->driver is always
true, and here it points to bcma_pci_bridge_driver, which has .name set to
"bcma-pci-bridge". Simplify accordingly.

Link: https://lore.kernel.org/r/20211004125935.2300113-5-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agocrypto: qat - simplify adf_enable_aer()
Uwe Kleine-König [Mon, 4 Oct 2021 12:59:33 +0000 (14:59 +0200)]
crypto: qat - simplify adf_enable_aer()

A struct pci_driver is shared across all device instances, so assigning
pci_driver.err_handler once per device isn't really sensible.

Set adf_driver.err_handler statically instead of in adf_enable_aer().
This removes a use of pci_dev->driver, which is a step toward removing
pci_dev->driver altogether.

Since adf_enable_aer() returns zero unconditionally, make it a void
function.

Link: https://lore.kernel.org/r/20211004125935.2300113-10-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoscsi: message: fusion: Remove unused mpt_pci driver .probe() 'id' parameter
Uwe Kleine-König [Mon, 4 Oct 2021 12:59:32 +0000 (14:59 +0200)]
scsi: message: fusion: Remove unused mpt_pci driver .probe() 'id' parameter

The only two drivers don't make use of the id parameter, so drop it.  This
is a step toward removing pci_dev->driver.

Link: https://lore.kernel.org/r/20211004125935.2300113-9-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Acked-by: Martin K. Petersen <martin.petersen@oracle.com>
2 years agoPCI/ERR: Factor out common dev->driver expressions
Bjorn Helgaas [Tue, 12 Oct 2021 22:20:06 +0000 (17:20 -0500)]
PCI/ERR: Factor out common dev->driver expressions

Save the struct pci_driver pointer from pdev->driver instead of repeating
it several times.  No functional change.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: Drop pci_device_probe() test of !pci_dev->driver
Uwe Kleine-König [Mon, 4 Oct 2021 12:59:26 +0000 (14:59 +0200)]
PCI: Drop pci_device_probe() test of !pci_dev->driver

When the device core calls the .probe() callback for a device, the device
is never bound, so pci_dev->driver is always NULL.

Remove the unnecessary test of !pci_dev->driver.

Link: https://lore.kernel.org/r/20211004125935.2300113-3-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoPCI: Drop pci_device_remove() test of pci_dev->driver
Uwe Kleine-König [Mon, 4 Oct 2021 12:59:25 +0000 (14:59 +0200)]
PCI: Drop pci_device_remove() test of pci_dev->driver

When the driver core calls pci_device_remove(), there is a driver bound
to the device, so pci_dev->driver is never NULL.

Remove the unnecessary test of pci_dev->driver.

Link: https://lore.kernel.org/r/20211004125935.2300113-2-u.kleine-koenig@pengutronix.de
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
2 years agoPCI: Return NULL for to_pci_driver(NULL)
Bjorn Helgaas [Tue, 12 Oct 2021 20:42:59 +0000 (15:42 -0500)]
PCI: Return NULL for to_pci_driver(NULL)

to_pci_driver() takes a pointer to a struct device_driver and uses
container_of() to find the struct pci_driver that contains it.

If given a NULL pointer to a struct device_driver, return a NULL pci_driver
pointer instead of applying container_of() to NULL.

This simplifies callers that would otherwise have to check for a NULL
pointer first.

Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: Use unsigned to match sscanf("%x") in pci_dev_str_match_path()
Krzysztof Wilczyński [Fri, 8 Oct 2021 22:27:32 +0000 (22:27 +0000)]
PCI: Use unsigned to match sscanf("%x") in pci_dev_str_match_path()

Cppcheck warns that pci_dev_str_match_path() passes pointers to signed ints
to sscanf("%x"), which expects pointers to *unsigned* ints:

  invalidScanfArgType_int drivers/pci/pci.c:312 %x in format string (no. 2) requires 'unsigned int *' but the argument type is 'signed int *'.

Declare the variables as unsigned to avoid this issue.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/20211008222732.2868493-3-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI: hv: Remove unnecessary use of %hx
Krzysztof Wilczyński [Fri, 8 Oct 2021 22:27:30 +0000 (22:27 +0000)]
PCI: hv: Remove unnecessary use of %hx

"dom_req" is a u16 but varargs automatically promotes it to int, so there's
no point in using the %h modifier.  Drop it.

See cbacb5ab0aa0 ("docs: printk-formats: Stop encouraging use of
unnecessary %h[xudi] and %hh[xudi]") and 70eb2275ff8e ("checkpatch: add
warning for unnecessary use of %h[xudi] and %hh[xudi]").

Link: https://lore.kernel.org/r/20211008222732.2868493-1-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
2 years agoPCI/VPD: Add pci_read/write_vpd_any()
Heiner Kallweit [Fri, 10 Sep 2021 06:22:06 +0000 (08:22 +0200)]
PCI/VPD: Add pci_read/write_vpd_any()

In certain cases we need a variant of pci_read_vpd()/pci_write_vpd() that
does not check against dev->vpd.len. Such cases are:

  - Reading VPD if dev->vpd.len isn't set yet (in pci_vpd_size())

  - Devices that map non-VPD information to arbitrary places in VPD address
    space (example: Chelsio T3 EEPROM write-protect flag)

Therefore add pci_read_vpd_any() and pci_write_vpd_any() that check against
PCI_VPD_MAX_SIZE only.

Link: https://lore.kernel.org/r/93ecce28-a158-f02a-d134-8afcaced8efe@gmail.com
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
2 years agoPCI: Correct misspelled and remove duplicated words
Krzysztof Wilczyński [Wed, 6 Oct 2021 23:38:27 +0000 (23:38 +0000)]
PCI: Correct misspelled and remove duplicated words

Correct a number of misspelled words and remove any words that were
duplicated in the PCI tree.  No change to functionality intended.

Link: https://lore.kernel.org/r/20211006233827.147328-1-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
3 years agoPCI: Coalesce host bridge contiguous apertures
Kai-Heng Feng [Tue, 13 Jul 2021 12:50:07 +0000 (20:50 +0800)]
PCI: Coalesce host bridge contiguous apertures

Built-in graphics at 07:00.0 on HP EliteDesk 805 G6 doesn't work because
graphics can't get the BAR it needs.  The BIOS configuration is
correct: BARs 0 and 2 both fit in the 00:08.1 bridge window.

But that 00:08.1 window covers two host bridge apertures from _CRS.
Previously we assumed this was illegal, so we clipped the window to fit
into one aperture (see 0f7e7aee2f37 ("PCI: Add pci_bus_clip_resource() to
clip to fit upstream window")).

  pci_bus 0000:00: root bus resource [mem 0x10020200000-0x100303fffff window]
  pci_bus 0000:00: root bus resource [mem 0x10030400000-0x100401fffff window]

  pci 0000:00:08.1:   bridge window [mem 0x10030000000-0x100401fffff 64bit pref]
  pci 0000:07:00.0: reg 0x10: [mem 0x10030000000-0x1003fffffff 64bit pref]
  pci 0000:07:00.0: reg 0x18: [mem 0x10040000000-0x100401fffff 64bit pref]

  pci 0000:00:08.1: can't claim BAR 15 [mem 0x10030000000-0x100401fffff 64bit pref]: no compatible bridge window
  pci 0000:00:08.1: [mem 0x10030000000-0x100401fffff 64bit pref] clipped to [mem 0x10030000000-0x100303fffff 64bit pref]
  pci 0000:00:08.1:   bridge window [mem 0x10030000000-0x100303fffff 64bit pref]

  pci 0000:07:00.0: can't claim BAR 0 [mem 0x10030000000-0x1003fffffff 64bit pref]: no compatible bridge window
  pci 0000:07:00.0: can't claim BAR 2 [mem 0x10040000000-0x100401fffff 64bit pref]: no compatible bridge window

However, the host bridge apertures are contiguous, so there's no need to
clip in this case.  Coalesce contiguous apertures so we can allocate from
the entire contiguous region.

Previous commit 65db04053efe ("PCI: Coalesce host bridge contiguous
apertures") was similar but sorted the apertures, and Guenter Roeck
reported a regression in ppc:sam460ex qemu emulation from nvme; see
https://lore.kernel.org/all/20210709231529.GA3270116@roeck-us.net/

[bhelgaas: commit log]
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=212013
Suggested-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://lore.kernel.org/r/20210713125007.1260304-1-kai.heng.feng@canonical.com
Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: Guenter Roeck <linux@roeck-us.net>
3 years agoPCI: Use kstrtobool() directly, sans strtobool() wrapper
Krzysztof Wilczyński [Wed, 15 Sep 2021 23:01:27 +0000 (23:01 +0000)]
PCI: Use kstrtobool() directly, sans strtobool() wrapper

strtobool() is a wrapper around kstrtobool() that has been added for
backward compatibility.

There is no reason to use the old API, so use kstrtobool() directly.

Related: ef951599074b ("lib: move strtobool() to kstrtobool()")

Link: https://lore.kernel.org/r/20210915230127.2495723-3-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
3 years agoPCI/sysfs: Return -EINVAL consistently from "store" functions
Krzysztof Wilczyński [Wed, 15 Sep 2021 23:01:26 +0000 (23:01 +0000)]
PCI/sysfs: Return -EINVAL consistently from "store" functions

Most of the "store" functions that handle userspace input via sysfs return
-EINVAL should the value fail validation and/or type conversion.  This
error code is a clear message to userspace that the value is not a valid
input.

However, some of the "show" functions return input parsing error codes
as-is, which may be either -EINVAL or -ERANGE.  The former would often be
from kstrtobool(), and the latter typically from other kstr*() functions
such as kstrtou8(), kstrtou32(), kstrtoint(), etc.

-EINVAL is commonly returned as the error code to indicate that the value
provided is invalid, but -ERANGE is not very useful in userspace.

Therefore, normalize the return error code to be -EINVAL for when the
validation and/or type conversion fails.

Link: https://lore.kernel.org/r/20210915230127.2495723-2-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
3 years agoPCI/sysfs: Check CAP_SYS_ADMIN before parsing user input
Krzysztof Wilczyński [Wed, 15 Sep 2021 23:01:25 +0000 (23:01 +0000)]
PCI/sysfs: Check CAP_SYS_ADMIN before parsing user input

Check if the "CAP_SYS_ADMIN" capability flag is set before parsing user
input as it makes more sense to first check whether the current user
actually has the right permissions before accepting any input from such
user.

This will also make order in which enable_store() and msi_bus_store()
perform the "CAP_SYS_ADMIN" capability check consistent with other
PCI-related sysfs objects that first verify whether user has this
capability set.

Link: https://lore.kernel.org/r/20210915230127.2495723-1-kw@linux.com
Signed-off-by: Krzysztof Wilczyński <kw@linux.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
3 years agoPCI/ACPI: Check for _OSC support in acpi_pci_osc_control_set()
Joerg Roedel [Tue, 24 Aug 2021 12:20:54 +0000 (14:20 +0200)]
PCI/ACPI: Check for _OSC support in acpi_pci_osc_control_set()

Get rid of acpi_pci_osc_support() and check for _OSC supported features
directly in acpi_pci_osc_control_set(). There is no point in doing an
unconditional _OSC query with control=0 even when the kernel later wants to
take control over more features.

This saves one _OSC query and simplifies the code by getting rid of the
acpi_pci_osc_support() function. As a side effect, the !control checks in
acpi_pci_query_osc() can also be removed.

Link: https://lore.kernel.org/r/20210824122054.29481-5-joro@8bytes.org
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki <rafael@kernel.org>
3 years agoPCI/ACPI: Move _OSC query checks to separate function
Joerg Roedel [Tue, 24 Aug 2021 12:20:53 +0000 (14:20 +0200)]
PCI/ACPI: Move _OSC query checks to separate function

Move the checks about whether the _OSC controls are requested from the
firmware to a separate function.

Link: https://lore.kernel.org/r/20210824122054.29481-4-joro@8bytes.org
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki <rafael@kernel.org>
3 years agoPCI/ACPI: Move supported and control calculations to separate functions
Joerg Roedel [Tue, 24 Aug 2021 12:20:52 +0000 (14:20 +0200)]
PCI/ACPI: Move supported and control calculations to separate functions

Move the calculations of supported and controlled _OSC features out of
negotiate_os_control() into separate functions.

Link: https://lore.kernel.org/r/20210824122054.29481-3-joro@8bytes.org
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki <rafael@kernel.org>
3 years agoPCI/ACPI: Remove OSC_PCI_SUPPORT_MASKS and OSC_PCI_CONTROL_MASKS
Joerg Roedel [Tue, 24 Aug 2021 12:20:51 +0000 (14:20 +0200)]
PCI/ACPI: Remove OSC_PCI_SUPPORT_MASKS and OSC_PCI_CONTROL_MASKS

These masks are only used internally in the PCI Host Bridge _OSC
negotiation code, which already makes sure nothing outside of these masks
is set. Remove the masks and simplify the code.

Suggested-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://lore.kernel.org/r/20210824122054.29481-2-joro@8bytes.org
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki <rafael@kernel.org>
3 years agoPCI: Tidy comments
Pranay Sanghai [Sat, 18 Sep 2021 20:18:02 +0000 (13:18 -0700)]
PCI: Tidy comments

Make comments follow multi-line comment conventions. No functional change
intended.

Link: https://lore.kernel.org/r/YUZJenW2UCA4Qu0O@pranay-desktop
Signed-off-by: Pranay Sanghai <pranaysanghai@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
3 years agoPCI: Rename pcibios_add_device() to pcibios_device_add()
Oliver O'Halloran [Mon, 13 Sep 2021 15:27:08 +0000 (01:27 +1000)]
PCI: Rename pcibios_add_device() to pcibios_device_add()

The general convention for pcibios_* hooks is that they're named after the
corresponding pci_* function they provide a hook for. The exception is
pcibios_add_device() which provides a hook for pci_device_add().

Rename pcibios_add_device() to pcibios_device_add() so it matches
pci_device_add().

Also, remove the export of the microblaze version. The only caller must be
compiled as a built-in so there's no reason for the export.

Link: https://lore.kernel.org/r/20210913152709.48013-1-oohall@gmail.com
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Niklas Schnelle <schnelle@linux.ibm.com> # s390
3 years agoPCI/portdrv: Do not setup up IRQs if there are no users
Jan Kiszka [Mon, 30 Aug 2021 08:08:10 +0000 (10:08 +0200)]
PCI/portdrv: Do not setup up IRQs if there are no users

Avoid registering service IRQs if there is no service that offers them
or no driver to register a handler against them. This saves IRQ vectors
when they are limited (e.g. on x86) and also avoids that spurious events
could hit a missing handler. Such spurious events need to be generated
by the Jailhouse hypervisor for active MSI vectors when enabling or
disabling itself.

Link: https://lore.kernel.org/r/8f9a13ac-8ab1-15ac-06cb-c131b488a36f@siemens.com
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
3 years agoPCI: Mark Atheros QCA6174 to avoid bus reset
Ingmar Klein [Fri, 9 Apr 2021 09:26:33 +0000 (11:26 +0200)]
PCI: Mark Atheros QCA6174 to avoid bus reset

When passing the Atheros QCA6174 through to a virtual machine, the VM hangs
at the point where the ath10k driver loads.

Add a quirk to avoid bus resets on this device, which avoids the hang.

[bhelgaas: commit log]
Link: https://lore.kernel.org/r/08982e05-b6e8-5a8d-24ab-da1488ee50a8@web.de
Signed-off-by: Ingmar Klein <ingmar_klein@web.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Pali Rohár <pali@kernel.org>
Cc: stable@vger.kernel.org
3 years agoPCI/P2PDMA: Apply bus offset correctly in DMA address calculation
Wang Lu [Thu, 9 Sep 2021 03:25:28 +0000 (11:25 +0800)]
PCI/P2PDMA: Apply bus offset correctly in DMA address calculation

The bus offset is bus address - physical address, so the calculation in
__pci_p2pdma_map_sg should be: bus address = physical address + bus offset.

Correct the dma_address computation in __pci_p2pdma_map_sg().

Link: https://lore.kernel.org/r/20210909032528.24517-1-wanglu@dapustor.com
Signed-off-by: Wang Lu <wanglu@dapustor.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
3 years agoLinux 5.15-rc2 v5.15-rc2
Linus Torvalds [Mon, 20 Sep 2021 00:28:22 +0000 (17:28 -0700)]
Linux 5.15-rc2

3 years agopci_iounmap'2: Electric Boogaloo: try to make sense of it all
Linus Torvalds [Mon, 20 Sep 2021 00:13:35 +0000 (17:13 -0700)]
pci_iounmap'2: Electric Boogaloo: try to make sense of it all

Nathan Chancellor reports that the recent change to pci_iounmap in
commit 9caea0007601 ("parisc: Declare pci_iounmap() parisc version only
when CONFIG_PCI enabled") causes build errors on arm64.

It took me about two hours to convince myself that I think I know what
the logic of that mess of #ifdef's in the <asm-generic/io.h> header file
really aim to do, and rewrite it to be easier to follow.

Famous last words.

Anyway, the code has now been lifted from that grotty header file into
lib/pci_iomap.c, and has fairly extensive comments about what the logic
is.  It also avoids indirecting through another confusing (and badly
named) helper function that has other preprocessor config conditionals.

Let's see what odd architecture did something else strange in this area
to break things.  But my arm64 cross build is clean.

Fixes: 9caea0007601 ("parisc: Declare pci_iounmap() parisc version only when CONFIG_PCI enabled")
Reported-by: Nathan Chancellor <nathan@kernel.org>
Cc: Helge Deller <deller@gmx.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Ulrich Teichert <krypton@ulrich-teichert.org>
Cc: James Bottomley <James.Bottomley@hansenpartnership.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoMerge tag 'x86_urgent_for_v5.15_rc2' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 19 Sep 2021 20:29:36 +0000 (13:29 -0700)]
Merge tag 'x86_urgent_for_v5.15_rc2' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Borislav Petkov:

 - Prevent a infinite loop in the MCE recovery on return to user space,
   which was caused by a second MCE queueing work for the same page and
   thereby creating a circular work list.

 - Make kern_addr_valid() handle existing PMD entries, which are marked
   not present in the higher level page table, correctly instead of
   blindly dereferencing them.

 - Pass a valid address to sanitize_phys(). This was caused by the
   mixture of inclusive and exclusive ranges. memtype_reserve() expect
   'end' being exclusive, but sanitize_phys() wants it inclusive. This
   worked so far, but with end being the end of the physical address
   space the fail is exposed.

 - Increase the maximum supported GPIO numbers for 64bit. Newer SoCs
   exceed the previous maximum.

* tag 'x86_urgent_for_v5.15_rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mce: Avoid infinite loop for copy from user recovery
  x86/mm: Fix kern_addr_valid() to cope with existing but not present entries
  x86/platform: Increase maximum GPIO number for X86_64
  x86/pat: Pass valid address to sanitize_phys()

3 years agoMerge tag 'perf-urgent-2021-09-19' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 19 Sep 2021 20:22:40 +0000 (13:22 -0700)]
Merge tag 'perf-urgent-2021-09-19' of git://git./linux/kernel/git/tip/tip

Pull perf event fix from Thomas Gleixner:
 "A single fix for the perf core where a value read with READ_ONCE() was
  checked and then reread which makes all the checks invalid. Reuse the
  already read value instead"

* tag 'perf-urgent-2021-09-19' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  events: Reuse value read using READ_ONCE instead of re-reading it

3 years agoMerge tag 'locking-urgent-2021-09-19' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 19 Sep 2021 20:11:19 +0000 (13:11 -0700)]
Merge tag 'locking-urgent-2021-09-19' of git://git./linux/kernel/git/tip/tip

Pull locking fixes from Thomas Gleixner:
 "A set of updates for the RT specific reader/writer locking base code:

   - Make the fast path reader ordering guarantees correct.

   - Code reshuffling to make the fix simpler"

[ This plays ugly games with atomic_add_return_release() because we
  don't have a plain atomic_add_release(), and should really be cleaned
  up, I think    - Linus ]

* tag 'locking-urgent-2021-09-19' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  locking/rwbase: Take care of ordering guarantee for fastpath reader
  locking/rwbase: Extract __rwbase_write_trylock()
  locking/rwbase: Properly match set_and_save_state() to restore_state()

3 years agoMerge tag 'powerpc-5.15-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc...
Linus Torvalds [Sun, 19 Sep 2021 20:00:23 +0000 (13:00 -0700)]
Merge tag 'powerpc-5.15-2' of git://git./linux/kernel/git/powerpc/linux

Pull powerpc fixes from Michael Ellerman:

 - Fix crashes when scv (System Call Vectored) is used to make a syscall
   when a transaction is active, on Power9 or later.

 - Fix bad interactions between rfscv (Return-from scv) and Power9
   fake-suspend mode.

 - Fix crashes when handling machine checks in LPARs using the Hash MMU.

 - Partly revert a recent change to our XICS interrupt controller code,
   which broke the recently added Microwatt support.

Thanks to Cédric Le Goater, Eirik Fuller, Ganesh Goudar, Gustavo Romero,
Joel Stanley, Nicholas Piggin.

* tag 'powerpc-5.15-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  powerpc/xics: Set the IRQ chip data for the ICS native backend
  powerpc/mce: Fix access error in mce handler
  KVM: PPC: Book3S HV: Tolerate treclaim. in fake-suspend mode changing registers
  powerpc/64s: system call rfscv workaround for TM bugs
  selftests/powerpc: Add scv versions of the basic TM syscall tests
  powerpc/64s: system call scv tabort fix for corrupt irq soft-mask state

3 years agoMerge tag 'kbuild-fixes-v5.15' of git://git.kernel.org/pub/scm/linux/kernel/git/masah...
Linus Torvalds [Sun, 19 Sep 2021 19:55:12 +0000 (12:55 -0700)]
Merge tag 'kbuild-fixes-v5.15' of git://git./linux/kernel/git/masahiroy/linux-kbuild

Pull Kbuild fixes from Masahiro Yamada:

 - Fix bugs in checkkconfigsymbols.py

 - Fix missing sys import in gen_compile_commands.py

 - Fix missing FORCE warning for ARCH=sh builds

 - Fix -Wignored-optimization-argument warnings for Clang builds

 - Turn -Wignored-optimization-argument into an error in order to stop
   building instead of sprinkling warnings

* tag 'kbuild-fixes-v5.15' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
  kbuild: Add -Werror=ignored-optimization-argument to CLANG_FLAGS
  x86/build: Do not add -falign flags unconditionally for clang
  kbuild: Fix comment typo in scripts/Makefile.modpost
  sh: Add missing FORCE prerequisites in Makefile
  gen_compile_commands: fix missing 'sys' package
  checkkconfigsymbols.py: Remove skipping of help lines in parse_kconfig_file
  checkkconfigsymbols.py: Forbid passing 'HEAD' to --commit

3 years agoMerge tag 'perf-tools-fixes-for-v5.15-2021-09-18' of git://git.kernel.org/pub/scm...
Linus Torvalds [Sun, 19 Sep 2021 19:49:17 +0000 (12:49 -0700)]
Merge tag 'perf-tools-fixes-for-v5.15-2021-09-18' of git://git./linux/kernel/git/acme/linux

Pull perf tools fixes from Arnaldo Carvalho de Melo:

 - Fix ip display in 'perf script' when output type != attr->type.

 - Ignore deprecation warning when using libbpf'sg btf__get_from_id(),
   fixing the build with libbpf v0.6+.

 - Make use of FD() robust in libperf, fixing a segfault with 'perf stat
   --iostat list'.

 - Initialize addr_location:srcline pointer to NULL when resolving
   callchain addresses.

 - Fix fused instruction logic for assembly functions in 'perf
   annotate'.

* tag 'perf-tools-fixes-for-v5.15-2021-09-18' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
  perf bpf: Ignore deprecation warning when using libbpf's btf__get_from_id()
  libperf evsel: Make use of FD robust.
  perf machine: Initialize srcline string member in add_location struct
  perf script: Fix ip display when type != attr->type
  perf annotate: Fix fused instr logic for assembly functions

3 years agodmascc: use proper 'virt_to_bus()' rather than casting to 'int'
Linus Torvalds [Sun, 19 Sep 2021 17:49:42 +0000 (10:49 -0700)]
dmascc: use proper 'virt_to_bus()' rather than casting to 'int'

The old dmascc driver depends on the legacy ISA_DMA_API, and blindly
just casts the kernel virtual address to 'int' for set_dma_addr().

That works only incidentally, and because the high bits of the address
will be ignored anyway. And on 64-bit architectures it causes warnings.

Admittedly, 64-bit architectures with ISA are basically dead - I think
the only example of this is alpha, and nobody would ever use the dmascc
driver there.  But hey, the fix is easy enough, the end result is
cleaner, and it's yet another configuration that now builds without
warnings.

If somebody actually uses this driver on an alpha and this fixes it for
you, please email me.  Because that is just incredibly bizarre.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoalpha: enable GENERIC_PCI_IOMAP unconditionally
Linus Torvalds [Sun, 19 Sep 2021 17:37:00 +0000 (10:37 -0700)]
alpha: enable GENERIC_PCI_IOMAP unconditionally

With the previous commit (9caea0007601: "parisc: Declare pci_iounmap()
parisc version only when CONFIG_PCI enabled") we can now enable
GENERIC_PCI_IOMAP unconditionally on alpha, and if PCI is not enabled we
will just get the nice empty helper functions that allow mixed-bus
drivers to build.

Example driver: the old 3com/3c59x.c driver works with either the PCI or
the EISA version of the 3x59x card, but wouldn't build in an EISA-only
configuration because of missing pci_iomap() and pci_iounmap() dummy
wrappers.

Most of the other PCI infrastructure just becomes empty wrappers even
without GENERIC_PCI_IOMAP, and it's not obvious that the pci_iomap
functionality shouldn't do the same, but this works.

Cc: Ulrich Teichert <krypton@ulrich-teichert.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoparisc: Declare pci_iounmap() parisc version only when CONFIG_PCI enabled
Helge Deller [Sun, 19 Sep 2021 17:36:09 +0000 (10:36 -0700)]
parisc: Declare pci_iounmap() parisc version only when CONFIG_PCI enabled

Linus noticed odd declaration rules for pci_iounmap() in iomap.h and
pci_iomap.h, where it dependend on either NO_GENERIC_PCI_IOPORT_MAP or
GENERIC_IOMAP when CONFIG_PCI was disabled.

Testing on parisc seems to indicate that we need pci_iounmap() only when
CONFIG_PCI is enabled, so the declaration of pci_iounmap() can be moved
cleanly into pci_iomap.h in sync with the declarations of pci_iomap().

Link: https://lore.kernel.org/all/CAHk-=wjRrh98pZoQ+AzfWmsTZacWxTJKXZ9eKU2X_0+jM=O8nw@mail.gmail.com/
Signed-off-by: Helge Deller <deller@gmx.de>
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Fixes: 97a29d59fc22 ("[PARISC] fix compile break caused by iomap: make IOPORT/PCI mapping functions conditional")
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Ulrich Teichert <krypton@ulrich-teichert.org>
Cc: James Bottomley <James.Bottomley@hansenpartnership.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoRevert "drm/vc4: hdmi: Remove drm_encoder->crtc usage"
Linus Torvalds [Sun, 19 Sep 2021 17:11:53 +0000 (10:11 -0700)]
Revert "drm/vc4: hdmi: Remove drm_encoder->crtc usage"

This reverts commit 27da370e0fb343a0baf308f503bb3e5dcdfe3362.

Sudip Mukherjee reports that this broke pulseaudio with a NULL pointer
dereference in vc4_hdmi_audio_prepare(), bisected it to this commit, and
confirmed that a revert fixed the problem.

Revert the problematic commit until fixed.

Link: https://lore.kernel.org/all/CADVatmPB9-oKd=ypvj25UYysVo6EZhQ6bCM7EvztQBMyiZfAyw@mail.gmail.com/
Link: https://lore.kernel.org/all/CADVatmN5EpRshGEPS_JozbFQRXg5w_8LFB3OMP1Ai-ghxd3w4g@mail.gmail.com/
Reported-and-tested-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com>
Cc: Maxime Ripard <maxime@cerno.tech>
Cc: Emma Anholt <emma@anholt.net>
Cc: Dave Airlie <airlied@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agoRevert drm/vc4 hdmi runtime PM changes
Linus Torvalds [Sun, 19 Sep 2021 17:06:46 +0000 (10:06 -0700)]
Revert drm/vc4 hdmi runtime PM changes

This reverts commits

  9984d6664ce9 ("drm/vc4: hdmi: Make sure the controller is powered in detect")
  411efa18e4b0 ("drm/vc4: hdmi: Move the HSM clock enable to runtime_pm")

as Michael Stapelberg reports that the new runtime PM changes cause his
Raspberry Pi 3 to hang on boot, probably due to interactions with other
changes in the DRM tree (because a bisect points to the merge in commit
e058a84bfddc: "Merge tag 'drm-next-2021-07-01' of git://.../drm").

Revert these two commits until it's been resolved.

Link: https://lore.kernel.org/all/871r5mp7h2.fsf@midna.i-did-not-set--mail-host-address--so-tickle-me/
Reported-and-tested-by: Michael Stapelberg <michael@stapelberg.ch>
Cc: Maxime Ripard <maxime@cerno.tech>
Cc: Dave Stevenson <dave.stevenson@raspberrypi.com>
Cc: Dave Airlie <airlied@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
3 years agokbuild: Add -Werror=ignored-optimization-argument to CLANG_FLAGS
Nathan Chancellor [Thu, 16 Sep 2021 18:40:17 +0000 (11:40 -0700)]
kbuild: Add -Werror=ignored-optimization-argument to CLANG_FLAGS

Similar to commit 589834b3a009 ("kbuild: Add
-Werror=unknown-warning-option to CLANG_FLAGS").

Clang ignores certain GCC flags that it has not implemented, only
emitting a warning:

$ echo | clang -fsyntax-only -falign-jumps -x c -
clang-14: warning: optimization flag '-falign-jumps' is not supported
[-Wignored-optimization-argument]

When one of these flags gets added to KBUILD_CFLAGS unconditionally, all
subsequent cc-{disable-warning,option} calls fail because -Werror was
added to these invocations to turn the above warning and the equivalent
-W flag warning into errors.

To catch the presence of these flags earlier, turn
-Wignored-optimization-argument into an error so that the flags can
either be implemented or ignored via cc-option and there are no more
weird errors.

Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
3 years agox86/build: Do not add -falign flags unconditionally for clang
Nathan Chancellor [Thu, 16 Sep 2021 18:40:16 +0000 (11:40 -0700)]
x86/build: Do not add -falign flags unconditionally for clang

clang does not support -falign-jumps and only recently gained support
for -falign-loops. When one of the configuration options that adds these
flags is enabled, clang warns and all cc-{disable-warning,option} that
follow fail because -Werror gets added to test for the presence of this
warning:

clang-14: warning: optimization flag '-falign-jumps=0' is not supported
[-Wignored-optimization-argument]

To resolve this, add a couple of cc-option calls when building with
clang; gcc has supported these options since 3.2 so there is no point in
testing for their support. -falign-functions was implemented in clang-7,
-falign-loops was implemented in clang-14, and -falign-jumps has not
been implemented yet.

Link: https://lore.kernel.org/r/YSQE2f5teuvKLkON@Ryzen-9-3900X.localdomain/
Link: https://lore.kernel.org/r/20210824022640.2170859-2-nathan@kernel.org/
Reported-by: kernel test robot <lkp@intel.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Acked-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
3 years agokbuild: Fix comment typo in scripts/Makefile.modpost
Ramji Jiyani [Thu, 16 Sep 2021 09:21:22 +0000 (09:21 +0000)]
kbuild: Fix comment typo in scripts/Makefile.modpost

Change comment "create one <module>.mod.c file pr. module"
to "create one <module>.mod.c file per module"

Signed-off-by: Ramji Jiyani <ramjiyani@google.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
3 years agosh: Add missing FORCE prerequisites in Makefile
Geert Uytterhoeven [Thu, 16 Sep 2021 08:43:53 +0000 (10:43 +0200)]
sh: Add missing FORCE prerequisites in Makefile

make:

    arch/sh/boot/Makefile:87: FORCE prerequisite is missing

Add the missing FORCE prerequisites for all build targets identified by
"make help".

Fixes: e1f86d7b4b2a5213 ("kbuild: warn if FORCE is missing for if_changed(_dep,_rule) and filechk")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
3 years agogen_compile_commands: fix missing 'sys' package
Kortan [Wed, 8 Sep 2021 03:28:48 +0000 (11:28 +0800)]
gen_compile_commands: fix missing 'sys' package

We need to import the 'sys' package since the script has called
sys.exit() method.

Fixes: 6ad7cbc01527 ("Makefile: Add clang-tidy and static analyzer support to makefile")
Signed-off-by: Kortan <kortanzh@gmail.com>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
3 years agocheckkconfigsymbols.py: Remove skipping of help lines in parse_kconfig_file
Ariel Marcovitch [Wed, 1 Sep 2021 16:49:52 +0000 (19:49 +0300)]
checkkconfigsymbols.py: Remove skipping of help lines in parse_kconfig_file

When parsing Kconfig files to find symbol definitions and references,
lines after a 'help' line are skipped until a new config definition
starts.

However, Kconfig statements can actually be after a help section, as
long as these have shallower indentation. These are skipped by the
parser.

This means that symbols referenced in this kind of statements are
ignored by this function and thus are not considered undefined
references in case the symbol is not defined.

Remove the 'skip' logic entirely, as it is not needed if we just use the
STMT regex to find the end of help lines.

However, this means that keywords that appear as part of the help
message (i.e. with the same indentation as the help lines) it will be
considered as a reference/definition. This can happen now as well, but
only with REGEX_KCONFIG_DEF lines. Also, the keyword must have a SYMBOL
after it, which probably means that someone referenced a config in the
help so it seems like a bonus :)

The real solution is to keep track of the indentation when a the first
help line in encountered and then handle DEF and STMT lines only if the
indentation is shallower.

Signed-off-by: Ariel Marcovitch <arielmarcovitch@gmail.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>