drm/bridge: Introduce drm_bridge_is_atomic() helper

We test for whether the bridge is atomic in several places in the source
code, so let's consolidate them.

Suggested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-11-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/bridge: ti-sn65dsi83: Switch to drm_bridge_helper_reset_crtc

Now that we have a helper for bridge drivers to call to reset the output
pipeline, let's use it.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Herve Codina <herve.codina@bootlin.com>
Tested-by: Herve Codina <herve.codina@bootlin.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-10-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/tests: bridge: Provide tests for drm_bridge_helper_reset_crtc

Let's provide a bunch of kunit tests to make sure
drm_bridge_helper_reset_crtc() works as expected.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-9-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/bridge: Add helper to reset bridge pipeline

Let's provide an helper to make it easier for bridge drivers to
power-cycle their bridge.

In order to avoid a circular dependency between that new helper and
drm_atomic_helper_reset_crtc(), this new helper will be in a
drm_bridge_helper.c file to follow the pattern we have for other
objects.

Reviewed-by: Herve Codina <herve.codina@bootlin.com>
Reviewed-by: Simona Vetter <simona.vetter@ffwll.ch>
Tested-by: Herve Codina <herve.codina@bootlin.com>
Co-developed-by: Simona Vetter <simona.vetter@intel.com>
Signed-off-by: Simona Vetter <simona.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-8-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/tests: Create tests for drm_atomic

We don't have a set of kunit tests for the functions under drm_atomic.h.
Let's use the introduction of drm_atomic_get_connector_for_encoder() to
create some tests for it and thus create that set.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-7-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/tests: hdmi_state_helpers: Switch to new helper

We introduced a new helper that supersedes the light_up_connector()
function in drm_hdmi_state_helper_test, so let's convert all our tests
to it.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-6-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/tests: helpers: Create new helper to enable output

We'll need the HDMI state tests light_up_connector() function in more
tests, so let's promote it to a helper.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-5-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/atomic: Introduce helper to lookup connector by encoder

With the bridges switching over to drm_bridge_connector, the direct
association between a bridge driver and its connector was lost.

This is mitigated for atomic bridge drivers by the fact you can access
the encoder, and then call drm_atomic_get_old_connector_for_encoder() or
drm_atomic_get_new_connector_for_encoder() with drm_atomic_state.

This was also made easier by providing drm_atomic_state directly to all
atomic hooks bridges can implement.

However, bridge drivers don't have a way to access drm_atomic_state
outside of the modeset path, like from the hotplug interrupt path or any
interrupt handler.

Let's introduce a function to retrieve the connector currently assigned
to an encoder, without using drm_atomic_state, to make these drivers'
life easier.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Simona Vetter <simona.vetter@ffwll.ch>
Tested-by: Herve Codina <herve.codina@bootlin.com>
Co-developed-by: Simona Vetter <simona.vetter@intel.com>
Signed-off-by: Simona Vetter <simona.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-4-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/tests: Add kunit tests for bridges

None of the drm_bridge function have kunit tests so far. Let's change
that, starting with drm_bridge_get_current_state().

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-3-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/bridge: Provide a helper to retrieve current bridge state

The current bridge state is accessible from the drm_bridge structure,
but since it's fairly indirect it's not easy to figure out.

Provide a helper to retrieve it.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-2-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/bridge: Add encoder parameter to drm_bridge_funcs.attach

The drm_bridge structure contains an encoder pointer that is widely used
by bridge drivers. This pattern is largely documented as deprecated in
other KMS entities for atomic drivers.

However, one of the main use of that pointer is done in attach to just
call drm_bridge_attach on the next bridge to add it to the bridge list.
While this dereferences the bridge->encoder pointer, it's effectively
the same encoder the bridge was being attached to.

We can make it more explicit by adding the encoder the bridge is
attached to to the list of attach parameters. This also removes the need
to dereference bridge->encoder in most drivers.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Douglas Anderson <dianders@chromium.org>
Tested-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Reviewed-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313-bridge-connector-v6-1-511c54a604fb@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

MAINTAINERS: update Microchip's Atmel-HLCDC driver maintainers

Drop Sam Ravnborg and Boris Brezillon as they are no longer interested in
maintaining the drivers. Add myself and Dharma Balasubiramani as the
Maintainer and co-maintainer for Microchip's Atmel-HLCDC driver.
Thanks for their work.

Signed-off-by: Manikandan Muralidharan <manikandan.m@microchip.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Acked-by: Boris Brezillon <boris.brezillon@collabora.com>
Acked-by: Nicolas Ferre <nicolas.ferre@microchip.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240620095856.777390-1-manikandan.m@microchip.com
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>

Documentation: vgaarbiter: Fix grammar

Correct grammar issues:

- Fix "co-exist" subject-verb agreement
- Correct plural form of "server" in context of more than one legacy
  devices
- Use passive mood for intro sentence of libpciaccess section

Signed-off-by: Pranav Tyagi <pranav.tyagi03@gmail.com>
Reviewed-by: Bagas Sanjaya <bagasdotme@gmail.com>
Link: https://lore.kernel.org/r/20250220164946.18007-1-pranav.tyagi03@gmail.com
[Bagas: massage commit message]
Signed-off-by: Bagas Sanjaya <bagasdotme@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250318041249.20786-2-bagasdotme@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vmwgfx: Switch to exclusively using GEM references

Currently we use a combination of TTM and GEM reference counting which is
cumbersome. TTM references are used for kernel internal BOs and operations
like validation. Simply switching the ttm_bo_(get|put) calls to their
GEM equivalents is insufficient as not all BOs are GEM BOs so we must set
the GEM vtable for all BOs even if they are not exposed to userspace.

Suggested-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Ian Forbes <ian.forbes@broadcom.com>
Reviewed-by: Zack Rusin <zack.rusin@broadcom.com>
Signed-off-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250131200321.193939-1-ian.forbes@broadcom.com

drm/vmwgfx: Fix dumb buffer leak

Dumb buffers were not being freed because the GEM reference that was
acquired in gb_surface_define was not dropped like it is in the 2D case.
Dropping this ref uncovered a few additional issues with freeing the
resources associated with dirty tracking in vmw_bo_free/release.

Additionally the TTM object associated with the surface were also leaking
which meant that when the ttm_object_file was closed at process exit the
destructor unreferenced an already destroyed surface.

The solution is to remove the destructor from the vmw_user_surface
associated with the dumb_buffer and immediately unreferencing the TTM
object which his removes it from the ttm_object_file.

This also allows the early return in vmw_user_surface_base_release for the
dumb buffer case to be removed as it should no longer occur.

The chain of references now has the GEM handle(s) owning the dumb buffer.
The GEM handles have a singular GEM reference to the vmw_bo which is
dropped when all handles are closed. When the GEM reference count hits
zero the vmw_bo is freed which then unreferences the surface via
vmw_resource_release in vmw_bo_release.

Fixes: d6667f0ddf46 ("drm/vmwgfx: Fix handling of dumb buffers")
Signed-off-by: Ian Forbes <ian.forbes@broadcom.com>
Reviewed-by: Zack Rusin <zack.rusin@broadcom.com>
Signed-off-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250123204424.836896-1-ian.forbes@broadcom.com

drm: panel-orientation-quirks: Add ZOTAC Gaming Zone

Add a panel orientation quirk for the ZOTAC Gaming Zone handheld gaming device.

Signed-off-by: Vicki Pfau <vi@endrift.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313211643.860786-2-vi@endrift.com
Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>

accel/qaic: Remove redundant 'flush_workqueue()' calls

'destroy_workqueue()' already drains the queue before destroying it, so
there is no need to flush it explicitly.

Remove the redundant 'flush_workqueue()' calls.

This was generated with coccinelle:

@@
expression E;
@@
- flush_workqueue(E);
  destroy_workqueue(E);

Signed-off-by: Chen Ni <nichen@iscas.ac.cn>
Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Signed-off-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250312073404.1429992-1-nichen@iscas.ac.cn

drm: add modifiers for Apple GPU layouts

Apple GPUs support non-linear "GPU-tiled" image layouts. Add modifiers
for these layouts. Mesa requires these modifiers to share non-linear
buffers across processes, but no other userspace or kernel support is
required/expected.

These layouts are notably not used for interchange across hardware
blocks (e.g. with the display controller). There are other layouts for
that but we don't support them either in userspace or kernelspace yet
(even downstream), so we don't add modifiers here.

Acked-by: Faith Ekstrand <faith.ekstrand@collabora.com>
Reviewed-by: Sven Peter <sven@svenpeter.dev>
Link: https://patchwork.freedesktop.org/patch/msgid/20250310-apple-twiddled-modifiers-v4-1-1ccac9544808@rosenzweig.io
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>

drm/sched: Clarify docu concerning drm_sched_job_arm()

The documentation for drm_sched_job_arm() and especially
drm_sched_job_cleanup() does not make it very clear why
drm_sched_job_arm() is a point of no return, which it indeed is.

Make the nature of drm_sched_job_arm() in the docu as clear as possible.

Suggested-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Philipp Stanner <phasta@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313093053.65001-2-phasta@kernel.org

drm/display: hdmi: Mention Infoframes testing with edid-decode

edid-decode gained recently support to check that infoframes are
compliant and match the EDID the monitor exposes.

Since the HDMI helpers provide those infoframes in debugfs, it makes it
easy to check from userspace that the drivers (and helpers) behave
properly.

Let's document it.

Cc: Hans Verkuil <hverkuil@xs4all.nl>
Reviewed-by: Hans Verkuil <hverkuil@xs4all.nl>
Link: https://patchwork.freedesktop.org/patch/msgid/20250312-drm-hdmi-state-docs-v2-2-6352a5d68d5b@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/display: hdmi: Create documentation section

We have had documentation for the public functions in the HDMI helpers,
but those were never referenced anywhere and thus not compiled as part
of the doc.

Let's add a section.

Reviewed-by: Dmitry Baryshkov <lumag@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250312-drm-hdmi-state-docs-v2-1-6352a5d68d5b@kernel.org
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/ttm/tests: fix incorrect assert in ttm_bo_unreserve_bulk()

In the ttm_bo_unreserve_bulk() test function, resv is allocated using
kunit_kzalloc(), but the subsequent assertion mistakenly verifies the
ttm_dev pointer instead of the resv pointer.

Fix the assertion to properly verify the resv pointer.

Signed-off-by: Qasim Ijaz <qasdev00@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250313161424.10688-1-qasdev00@gmail.com
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>

drm/client: s/unsigned int i/int i/

Replace the 'unsigned int i' footguns with plain old signed
int. Avoids accidents if/when someone decides they need
to iterate backwards.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228211454.8138-9-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>

drm/client: Move variables to tighter scope

Bunch of variables are only needed inside loops and whatnot.
Move them to a tighter scope to make the code less confusing.

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228211454.8138-8-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>

drm/client: s/new_crtc/crtc/

Rename the 'new_crtc' variable to just 'crtc' in
drm_client_firmware_config(). We don't call any of the other
stuff in here new or old so this feels out of place.

v2: Rebase

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228211454.8138-7-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>

drm/client: Stop using the legacy crtc->mode

crtc->mode is legacy junk and shouldn't really be used with
atomic drivers.

Most (all?) atomic drivers do end up still calling
drm_atomic_helper_update_legacy_modeset_state() at some
point, so crtc->mode does still get populated, and this
does work for now. But now that the modes[] lifetime issues
have been sorted out we can just switch over to the
proper crtc->state->mode.

v2: Rebase

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228211454.8138-6-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>

drm/client: Make copies of modes

drm_client_firmware_config() is currently picking up the current
mode of the crtc via the legacy crtc->mode, which is not supposed
to be used by atomic drivers at all. We can't simply switch over
to the proper crtc->state->mode because we drop the crtc->mutex
(which protects crtc->state) before the mode gets used.

The most straightforward solution to extend the lifetime of
modes[] seem to be to make full copies of the modes.

And with this we can undo also commit 3eadd887dbac
("drm/client:Fully protect modes[] with dev->mode_config.mutex")
as the lifetime of modes[] no longer has anything to do with
that lock.

v2: Don't try to copy NULL modes
v3: Keep storing pointers and use drm_mode_{duplicate,destroy}()

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228211454.8138-5-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>

drm/client: Streamline mode selection debugs

Get rid of all the redundant debugs and just wait until the end
to print which mode (and of which type) we picked.

Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228211454.8138-4-ville.syrjala@linux.intel.com

drm/client: Use array notation for function arguments

Use the array notation rather that the pointer notation for
function arguments. This makes it clear to the reader that
we are in fact dealing with an array rather than a single
pointer. Functionally the two are equivalent.

Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228211454.8138-3-ville.syrjala@linux.intel.com

drm/client: Constify modes

The modes used by the client code live on the connectors' mode
lists, which are not owned by the client code, and thus it has
no business modifying the modes. Mark the modes const to make
that fact abundantly clear.

v2: Fix up the kunit test

Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250303093847.7698-1-ville.syrjala@linux.intel.com

drm/ci: fix merge request rules

Merge request pipelines were only created when changes
were made to drivers/gpu/drm/ci/, causing MRs that
didn't touch this path to break. Fix MR pipeline rules
to trigger jobs for all changes.

Run jobs automatically for marge-bot and scheduled
pipelines, but in all other cases run manually. Also
remove CI_PROJECT_NAMESPACE checks specific to mesa.

Fixes: df54f04f2020 ("drm/ci: update gitlab rules")
Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Signed-off-by: Helen Koike <helen.fornazier@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228132620.556079-1-vignesh.raman@collabora.com

drm: xlnx: zynqmp_dpsub: fix Kconfig dependencies for ASoC

The new audio code fails to build when sounds support is in a loadable
module but the GPU driver is built-in:

x86_64-linux-ld: zynqmp_dp_audio.c:(.text+0x6a8): undefined reference to `devm_snd_soc_register_card'
x86_64-linux-ld: drivers/gpu/drm/xlnx/zynqmp_dp_audio.o:(.rodata+0x1bc): undefined reference to `snd_soc_info_volsw'
x86_64-linux-ld: drivers/gpu/drm/xlnx/zynqmp_dp_audio.o:(.rodata+0x1f0): undefined reference to `snd_soc_get_volsw'
x86_64-linux-ld: drivers/gpu/drm/xlnx/zynqmp_dp_audio.o:(.rodata+0x1f4): undefined reference to `snd_soc_put_volsw'

Change the Kconfig dependency to disallow the sound support in this
configuration.

Fixes: 3ec5c1579305 ("drm: xlnx: zynqmp_dpsub: Add DP audio support")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250227132036.1136600-1-arnd@kernel.org

drm/sched: revert "drm_sched_job_cleanup(): correct false doc"

This reverts commit 44d2f310f008613c1dbe5e234c2cf2be90cbbfab.

The function drm_sched_job_arm() is indeed the point of no return. The
background is that it is nearly impossible for the driver to correctly
retract the fence and signal it in the order enforced by the dma_fence
framework.

The code in drm_sched_job_cleanup() is for the purpose to cleanup after
the job was armed through drm_sched_job_arm() *and* processed by the
scheduler.

We can certainly improve the documentation, but removing the warning is
clearly not a good idea.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Philipp Stanner <phasta@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250312134400.2176393-1-christian.koenig@amd.com

drm/debugfs: add top-level 'bridges' file showing all added bridges

The global bridges_list holding all the bridges between drm_bridge_add()
and drm_bridge_remove() cannot be inspected via debugfs. Add a file showing
it.

To avoid code duplication, move the code printing a bridge info to a common
function.

Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250226-drm-debugfs-show-all-bridges-v8-2-bb511cc49d83@bootlin.com
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>

drm/bridge: move bridges_show logic from drm_debugfs.c

In preparation to expose more info about bridges in debugfs, which will
require more insight into drm_bridge data structures, move the bridges_show
code to drm_bridge.c.

Suggested-by: Jani Nikula <jani.nikula@linux.intel.com>
Suggested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Luca Ceresoli <luca.ceresoli@bootlin.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250226-drm-debugfs-show-all-bridges-v8-1-bb511cc49d83@bootlin.com
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>

drm/panel: simple: Add POWERTIP PH128800T004-ZZA01 panel entry

Add support for the POWERTIP PH128800T004-ZZA01 10.1" (1280x800)
LCD-TFT panel. Its panel description is very much like the POWERTIP
PH128800T006-ZHC01 configured below this one, only its timings are
different.

Signed-off-by: Antonin Godard <antonin.godard@bootlin.com>
Reviewed-by: Dmitry Baryshkov <lumag@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250311-add-powertip-ph128800t004-v1-2-7f95e6984cea@bootlin.com
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>

dt-bindings: display: simple: Add POWERTIP PH128800T004-ZZA01 panel

Add POWERTIP PH128800T004-ZZA01 10.1" LCD-TFT LVDS panel compatible
string.

Signed-off-by: Antonin Godard <antonin.godard@bootlin.com>
Acked-by: "Rob Herring (Arm)" <robh@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250311-add-powertip-ph128800t004-v1-1-7f95e6984cea@bootlin.com
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>

drm/mxsfb: Remove generic DRM drivers in probe function

Use aperture helpers to remove all generic graphics drivers before
loading mxsfb. Makes mxsfb compatible with simpledrm.

Co-developed-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Reviewed-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305140929.174398-1-dario.binacchi@amarulasolutions.com

drm/udl: Set struct drm_device.dma_dev

Set the dma_dev field provided by the DRM device. Required for PRIME
dma-buf import. Remove the driver's implementation.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250307080836.42848-6-tzimmermann@suse.de

drm/gud: Set struct drm_device.dma_dev

Set the dma_dev field provided by the DRM device. Required for PRIME
dma-buf import. Remove the driver's implementation.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250307080836.42848-5-tzimmermann@suse.de

drm/gm12u320: Set struct drm_device.dma_dev

Set the dma_dev field provided by the DRM device. Required for PRIME
dma-buf import. Remove the driver's implementation.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250307080836.42848-4-tzimmermann@suse.de

drm/appletbdrm: Set struct drm_device.dma_dev

Set the dma_dev field provided by the DRM device. Required for PRIME
dma-buf import. Remove the driver's implementation.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Tested-by: Aditya Garg <gargaditya08@live.com>
Reviewed-by: Aditya Garg <gargaditya08@live.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250307080836.42848-3-tzimmermann@suse.de

drm/prime: Support dedicated DMA device for dma-buf imports

Importing dma-bufs via PRIME requires a DMA-capable device. Devices on
peripheral busses, such as USB, often cannot perform DMA by themselves.
Without DMA-capable device PRIME import fails. DRM drivers for USB
devices already use a separate DMA device for dma-buf imports. Make the
mechanism generally available.

Besides the case of USB, there are embedded DRM devices without DMA
capability. DMA is performed by a separate controller. DRM drivers should
set this accordingly.

Add the field dma_dev to struct drm_device to refer to the device's DMA
device. For USB this should be the USB controller. Use dma_dev in the
PRIME import helpers, if set.

v2:
- acquire internal reference on dma_dev (Jani)
- add DMA-controller usecase to docs (Maxime)

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250307080836.42848-2-tzimmermann@suse.de

drm/ast: cursor: Drop page alignment

The cursor scanout address requires alignment to a multiple of 8,
but does not require page alignment. Change the offset calculation
accordingly. Frees up a few more bytes for the primary framebuffer.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jocelyn Falempe <jfalempe@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305163207.267650-8-tzimmermann@suse.de

drm/ast: Remove vram_fb_available from struct ast_device

Helpers compute the offset and size of the available framebuffer
memory. Remove the obsolete field vram_fb_available from struct
ast_device. Also define the cursor-signature size next to its only
user.

v2:
- initialize plane size

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jocelyn Falempe <jfalempe@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305163207.267650-7-tzimmermann@suse.de

drm/ast: Add helper for computing framebuffer location in video memory

The ast driver stores the primary plane's image in the framebuffer
memory up to where the cursor is located. Add helpers to calculate
the offset and size.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jocelyn Falempe <jfalempe@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305163207.267650-6-tzimmermann@suse.de

drm/ast: cursor: Add helpers for computing location in video memory

The ast drivers stores the cursor image at the end of the video memory.
Add helpers to calculate the offset and size.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jocelyn Falempe <jfalempe@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305163207.267650-5-tzimmermann@suse.de

drm/ast: Add VGACR99 register constants

Add register constants for VGACR99 and use them when detecting the
size of the VGA memory. Aligns the code with the programming manual.
Also replace literal size values with Linux' SZ_ size constants.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jocelyn Falempe <jfalempe@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305163207.267650-4-tzimmermann@suse.de

drm/ast: Add VGACRAA register constants

Add register constants for VGACRAA and use them when detecting the
size of the VGA memory. Aligns the code with the programming manual.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jocelyn Falempe <jfalempe@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305163207.267650-3-tzimmermann@suse.de

drm/ast: Replace AST_VIDMEM_SIZE_ with Linux SZ_ constants

Ast's AST_VIDMEM_SIZE_ constants enumerate supported video-memory
sizes from 8 MiB to 128 MiB. Replace them with Linux' SZ_ constants
of the same value. When expanded, the literal values remain the same.

The size constant for 128 MiB is unused and the default size is not
necessary. Remove both of them.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Jocelyn Falempe <jfalempe@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305163207.267650-2-tzimmermann@suse.de

drm/ci: enable lockdep detection

We have enabled PROVE_LOCKING (which enables LOCKDEP) in drm-ci.
This will output warnings when kernel locking errors are encountered
and will continue executing tests. To detect if lockdep has been
triggered, check the debug_locks value in /proc/lockdep_stats after
the tests have run. When debug_locks is 0, it indicates that lockdep
has detected issues and turned itself off. Check this value, and if
lockdep is detected, exit with an error and configure it as a warning
in GitLab CI.

GitLab CI ignores exit codes other than 1 by default. Pass the correct
exit code with variable FF_USE_NEW_BASH_EVAL_STRATEGY set to true or
exit on failure.

Also update the documentation.

Acked-by: Helen Koike <helen.fornazier@gmail.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250217053719.442644-4-vignesh.raman@collabora.com
Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>

drm/ci: enable CONFIG_DEBUG_WW_MUTEX_SLOWPATH

Enable CONFIG_DEBUG_WW_MUTEX_SLOWPATH for mutex
slowpath debugging.

Acked-by: Helen Koike <helen.fornazier@gmail.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250217053719.442644-3-vignesh.raman@collabora.com
Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>

drm/ci: refactor software-driver stage jobs

Move common job configuration for software-driver
stage jobs to separate job.

Acked-by: Helen Koike <helen.fornazier@gmail.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250217053719.442644-2-vignesh.raman@collabora.com
Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>

Documentation: Update the todo

Update the Documentation to be more precise.

v2: Update for clarity
v3: Further details in Todo

Cc: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Anusha Srivatsa <asrivats@redhat.com>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/640856/?series=144073&rev=5

drm/sti: move to devm_platform_ioremap_resource() usage

Replace platform_get_resource/_byname + devm_ioremap
with just devm_platform_ioremap_resource()

Used Coccinelle to do this change. SmPl patch:

@rule@
identifier res;
expression ioremap;
identifier pdev;
constant mem;
expression name;
@@
-struct resource *res;
...
-res = platform_get_resource_byname(pdev,mem,name);
<...
-if (!res) {
-...
-}
...>
-ioremap = devm_ioremap(...);
+ioremap = devm_platform_ioremap_resource_byname(pdev,name);

and
@rule_2@
identifier res;
expression ioremap;
identifier pdev;
@@
-struct resource *res;
...
-res = platform_get_resource(pdev,...);
<...
-if (!res) {
-...
-}
...>
-ioremap = devm_ioremap(...);
+ioremap = devm_platform_ioremap_resource(pdev,0);

v2: Fix compilation error.
v3: Handle returns properly since the new API return error pointers
and not NULL

Cc: Raphael Gallais-Pou <raphael.gallais-pou@foss.st.com>
Cc: Alain Volmat <alain.volmat@foss.st.com>
Reviewed-by: Maxime Ripard <mripard@kernel.org>(v2)
Acked-by: Raphael Gallais-Pou <rgallaispou@gmail.com>(v2)
Signed-off-by: Anusha Srivatsa <asrivats@redhat.com>
Link: https://patchwork.freedesktop.org/patch/640854/?series=144073&rev=5

drm/sprd: move to devm_platform_ioremap_resource() usage

Replace platform_get_resource + devm_ioremap
with just devm_platform_ioremap_resource()

Used Coccinelle to do this change. SmPl patch:
@rule_2@
identifier res;
expression ioremap;
identifier pdev;
@@
-struct resource *res;
...
-res = platform_get_resource(pdev,...);
<...
-if (!res) {
-...
-}
...>
-ioremap = devm_ioremap(...);
+ioremap = devm_platform_ioremap_resource(pdev,0);

v2: Address the return handling properly since
the new API returns error pointers and not NULL.

Cc: Chunyan Zhang <zhang.lyra@gmail.com>
Cc: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Anusha Srivatsa <asrivats@redhat.com>
Reviewed-by: Chunyan Zhang <zhang.lyra@gmail.com>(v1)
Reviewed-by: Maxime Ripard <mripard@kernel.org>(v1)
Link: https://patchwork.freedesktop.org/patch/640854/?series=144073&rev=5

drm/vmwgfx: Add error path for xa_store in vmw_bo_add_detached_resource

The xa_store() may fail due to memory allocation failure because there
is no guarantee that the index is already used. This fix introduces new
paths to handle the error.

This patch also aligns the order of function calls by calling
vmw_bo_add_detached_resource() before ttm_prime_object_init() in order
to allow consistent error handling.

Fixes: d6667f0ddf46 ("drm/vmwgfx: Fix handling of dumb buffers")
Signed-off-by: Keisuke Nishimura <keisuke.nishimura@inria.fr>
Signed-off-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250225145223.34773-1-keisuke.nishimura@inria.fr

drm/vmwgfx: Add seqno waiter for sync_files

Because sync_files are passive waiters they do not participate in
the processing of fences like the traditional vmw_fence_wait IOCTL.
If userspace exclusively uses sync_files for synchronization then
nothing in the kernel actually processes fence updates as interrupts
for fences are masked and ignored if the kernel does not indicate to the
SVGA device that there are active waiters.

This oversight results in a bug where the entire GUI can freeze waiting
on a sync_file that will never be signalled as we've masked the interrupts
to signal its completion. This bug is incredibly racy as any process which
interacts with the fencing code via the 3D stack can process the stuck
fences on behalf of the stuck process causing it to run again. Even a
simple app like eglinfo is enough to resume the stuck process. Usually
this bug is seen at a login screen like GDM because there are no other
3D apps running.

By adding a seqno waiter we re-enable interrupt based processing of the
dma_fences associated with the sync_file which is signalled as part of a
dma_fence_callback.

This has likely been broken since it was initially added to the kernel in
2017 but has gone unnoticed until mutter recently started using sync_files
heavily over the course of 2024 as part of their explicit sync support.

Fixes: c906965dee22 ("drm/vmwgfx: Add export fence to file descriptor support")
Signed-off-by: Ian Forbes <ian.forbes@broadcom.com>
Signed-off-by: Zack Rusin <zack.rusin@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250228200633.642417-1-ian.forbes@broadcom.com

drm/vmwgfx: Bump the minor version

Bump the minor version of vmwgfx in order to detect releases where the
cursor issues have been fixed.

Cursors created with dumb buffer were broken on vmwgfx. Userspace (e.g.
kwin) has workarounds for those issues and often disables hardware
cursors on vmwgfx. This allows enabling hardware cursors on vmwgfx
again.

Signed-off-by: Zack Rusin <zack.rusin@broadcom.com>
Reviewed-by: Maaz Mombasawala <maaz.mombasawala@broadcom.com>
Reviewed-by: Martin Krastev <martin.krastev@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250307125836.3877138-3-zack.rusin@broadcom.com

drm/vmwgfx: Refactor cursor handling

Refactor cursor handling to make the code maintainable again. Over the
last 12 years the svga device improved support for virtualized cursors
and at the same time the drm interfaces evolved quite a bit from
pre-atomic to current atomic ones. vmwgfx only added new code over
the years, instead of adjusting/refactoring the paths.

Export the cursor plane handling to its own file. Remove special
handling of the legacy cursor support to make it fit within the global
cursor plane mechanism.

Finally redo dirty tracking because memcmp never worked correctly
resulting in the cursor not being properly updated in the guest.

Signed-off-by: Zack Rusin <zack.rusin@broadcom.com>
Reviewed-by: Maaz Mombasawala <maaz.mombasawala@broadcom.com>
Reviewed-by: Martin Krastev <martin.krastev@broadcom.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250307125836.3877138-2-zack.rusin@broadcom.com

MAINTAINERS: Update drm/ci maintainers

Update drm/ci maintainer entries:

* Add myself as drm/ci maintainer.
* Update Helen's email address.

Signed-off-by: Vignesh Raman <vignesh.raman@collabora.com>
Acked-by: Helen Koike <helen.fornazier@gmail.com>
Acked-by: Daniel Stone <daniels@collabora.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250205134811.2002718-1-vignesh.raman@collabora.com

drm/panic: clean Clippy warning

Clippy warns:

    error: manual implementation of an assign operation
       --> drivers/gpu/drm/drm_panic_qr.rs:418:25
        |
    418 |                         self.carry = self.carry % pow;
        |                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ help: replace it with: `self.carry %= pow`
        |
        = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#assign_op_pattern

Thus clean it up.

Fixes: dbed4a797e00 ("drm/panic: Better binary encoding in QR code")
Signed-off-by: Miguel Ojeda <ojeda@kernel.org>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Reviewed-by: Jocelyn Falempe <jfalempe@redhat.com>
Signed-off-by: Jocelyn Falempe <jfalempe@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250303093242.1011790-1-ojeda@kernel.org

drm/gma500: Remove unused psb_mmu_virtual_to_pfn

psb_mmu_virtual_to_pfn() was added in 2011 by
commit 8c8f1c958ab5 ("gma500: introduce the GTT and MMU handling logic")
but hasn't been used.

Remove it.

Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250308234428.255164-1-linux@treblig.org

drm/gma500/psb_intel_modes: Remove unused psb_intel_ddc_probe

psb_intel_ddc_probe() was added in 2011 by
commit 89c78134cc54 ("gma500: Add Poulsbo support")
but has remained unused (probably because drm_get_edid is used
instead).

Remove it.

Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250308234356.255114-1-linux@treblig.org

drm/vc4: plane: fix inconsistent indenting warning

Fix below inconsistent indenting smatch warning.
smatch warnings:
drivers/gpu/drm/vc4/vc4_plane.c:2083 vc6_plane_mode_set() warn: inconsistent indenting

Signed-off-by: Charles Han <hanchunchao@inspur.com>
Signed-off-by: Maíra Canal <mcanal@igalia.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305102107.2595-1-hanchunchao@inspur.com

drm/nouveau/nvkm: introduce new GSP reply policy NVKM_GSP_RPC_REPLY_POLL

Some GSP RPC commands need a new reply policy: "caller don't care about
the message content but want to make sure a reply is received". To
support this case, a new reply policy is introduced.

NV_VGPU_MSG_FUNCTION_ALLOC_MEMORY is a large GSP RPC command. The actual
required policy is NVKM_GSP_RPC_REPLY_POLL. This can be observed from the
dump of the GSP message queue. After the large GSP RPC command is issued,
GSP will write only an empty RPC header in the queue as the reply.

Without this change, the policy "receiving the entire message" is used
for NV_VGPU_MSG_FUNCTION_ALLOC_MEMORY. This causes the timeout of receiving
the returned GSP message in the suspend/resume path.

Introduce the new reply policy NVKM_GSP_RPC_REPLY_POLL, which waits for
the returned GSP message but discards it for the caller. Use the new policy
NVKM_GSP_RPC_REPLY_POLL on the GSP RPC command
NV_VGPU_MSG_FUNCTION_ALLOC_MEMORY.

Fixes: 50f290053d79 ("drm/nouveau: support handling the return of large GSP message")
Cc: Danilo Krummrich <dakr@kernel.org>
Cc: Alexandre Courbot <acourbot@nvidia.com>
Tested-by: Ben Skeggs <bskeggs@nvidia.com>
Signed-off-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250227013554.8269-3-zhiw@nvidia.com

drm/nouveau/nvkm: factor out current GSP RPC command policies

There can be multiple cases of handling the GSP RPC messages, which are
the reply of GSP RPC commands according to the requirement of the
callers and the nature of the GSP RPC commands.

The current supported reply policies are "callers don't care" and "receive
the entire message" according to the requirement of the callers. To
introduce a new policy, factor out the current RPC command reply polices.
Also, centralize the handling of the reply in a single function.

Factor out NVKM_GSP_RPC_REPLY_NOWAIT as "callers don't care" and
NVKM_GSP_RPC_REPLY_RECV as "receive the entire message". Introduce a
kernel doc to document the policies. Factor out
r535_gsp_rpc_handle_reply().

No functional change is intended for small GSP RPC commands. For large GSP
commands, the caller decides the policy of how to handle the returned GSP
RPC message.

Cc: Ben Skeggs <bskeggs@nvidia.com>
Cc: Alexandre Courbot <acourbot@nvidia.com>
Signed-off-by: Zhi Wang <zhiw@nvidia.com>
Signed-off-by: Danilo Krummrich <dakr@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250227013554.8269-2-zhiw@nvidia.com

drm/vkms: Allow to attach connectors and encoders

Add a list of possible encoders to the connector configuration and
helpers to attach and detach them.

Now that the default configuration has its connector and encoder
correctly, configure the output following the configuration.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-15-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Allow to configure multiple connectors

Add a list of connectors to vkms_config and helper functions to add and
remove as many connectors as wanted.

For backwards compatibility, add one enabled connector to the default
configuration.

A future patch will allow to attach connectors and encoders, but for the
moment there are no changes in the way the output is configured.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-14-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Allow to attach encoders and CRTCs

Add a list of possible CRTCs to the encoder configuration and helpers to
attach and detach them.

Now that the default configuration has its encoder and CRTC correctly
attached, configure the output following the configuration.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-13-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Allow to configure multiple encoders

Add a list of encoders to vkms_config and helper functions to add and
remove as many encoders as wanted.

For backwards compatibility, add one encoder to the default
configuration.

A future patch will allow to attach encoders and CRTCs, but for the
moment there are no changes in the way the output is configured.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-12-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Allow to attach planes and CRTCs

Add a list of possible CRTCs to the plane configuration and helpers to
attach, detach and get the primary and cursor planes attached to a CRTC.

Now that the default configuration has its planes and CRTC correctly
attached, configure the output following the configuration.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-11-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Allow to configure multiple CRTCs

Add a list of CRTCs to vkms_config and helper functions to add and
remove as many CRTCs as wanted.

For backwards compatibility, add one CRTC to the default configuration.

A future patch will allow to attach planes and CRTCs, but for the
moment there are no changes in the way the output is configured.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-10-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Allow to configure multiple planes

Add a list of planes to vkms_config and create as many planes as
configured during output initialization.

For backwards compatibility, add one primary plane and, if configured,
one cursor plane and NUM_OVERLAY_PLANES planes to the default
configuration.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-9-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Add a validation function for VKMS configuration

As the configuration will be used by userspace, add a validator to avoid
creating a broken DRM device.

For the moment, the function always returns true, but rules will be
added in future patches.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: José Expósito <jose.exposito89@gmail.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-8-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Set device name from vkms_config

In order to be able to create multiple devices, the device name needs to
be unique.

Allow to set it in the VKMS configuration.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-7-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Move default_config creation to its own function

Extract the initialization of the default configuration to a function.
Refactor, no functional changes.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-6-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Extract vkms_config header

Creating a new vkms_config structure will be more complex once we
start adding more options.

Extract the vkms_config structure to its own header and source files
and add functions to create and delete a vkms_config and to initialize
debugfs.

Refactor, no functional changes.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-5-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Add KUnit test scaffolding

Add the required boilerplate to start creating KUnit test.

To run the tests:

    $ ./tools/testing/kunit/kunit.py run \
      --kunitconfig=drivers/gpu/drm/vkms/tests

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Co-developed-by: Arthur Grillo <arthurgrillo@riseup.net>
Signed-off-by: Arthur Grillo <arthurgrillo@riseup.net>
Co-developed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-4-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Create vkms_connector struct

Create a structure wrapping the drm_connector.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-3-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm/vkms: Extract vkms_connector header

Up until now, the logic to manage connectors was in vkms_output.c.

Since more options will be added to connectors in the future, extract
the code to its own file.

Refactor, no functional changes.

Reviewed-by: Louis Chauvet <louis.chauvet@bootlin.com>
Signed-off-by: José Expósito <jose.exposito89@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250218101214.5790-2-jose.exposito89@gmail.com
Signed-off-by: Maxime Ripard <mripard@kernel.org>

drm: pl111: fix inconsistent indenting warning

Fix below inconsistent indenting smatch warning.
smatch warnings:
drivers/gpu/drm/pl111/pl111_versatile.c:504 pl111_versatile_init() warn: inconsistent indenting

Signed-off-by: Charles Han <hanchunchao@inspur.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305102540.2815-1-hanchunchao@inspur.com

drm/gma500: Remove unused mrst_clock_funcs

The mrst_clock_funcs const was added in 2013 by
commit ac6113ebb70d ("drm/gma500/mrst: Add SDVO clock calculation")
and commented as 'Not used yet'.

It's not been used since, so remove it.
The helper functions it points to are still used elsewhere.

Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Signed-off-by: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250306155155.212599-1-linux@treblig.org

drm/imx: legacy-bridge: fix inconsistent indenting warning

Fix below inconsistent indenting smatch warning.
smatch warnings:
drivers/gpu/drm/bridge/imx/imx-legacy-bridge.c:79 devm_imx_drm_legacy_bridge() warn: inconsistent indenting

Signed-off-by: Charles Han <hanchunchao@inspur.com>
Reviewed-by: Liu Ying <victor.liu@nxp.com>
Signed-off-by: Liu Ying <victor.liu@nxp.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305103042.3017-1-hanchunchao@inspur.com

drm/tegra: move to devm_platform_ioremap_resource() usage

Replace platform_get_resource + devm_ioremap_resource
with just devm_platform_ioremap_resource()

Used Coccinelle to do this change. SmPl patch:
@rule_1@
identifier res;
expression ioremap_res;
identifier pdev;
@@
-struct resource *res;
...
-res = platform_get_resource(pdev,...);
-ioremap_res = devm_ioremap_resource(...);
+ioremap_res = devm_platform_ioremap_resource(pdev,0);

Cc: Thierry Reding <thierry.reding@gmail.com>
Cc: Mikko Perttunen <mperttunen@nvidia.com>
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Anusha Srivatsa <asrivats@redhat.com>
Link: https://patchwork.freedesktop.org/patch/640855/?series=144073&rev=5

drm/mxsfb: move to devm_platform_ioremap_resource() usage

Replace platform_get_resource + devm_ioremap_resource
with just devm_platform_ioremap_resource()

Used Coccinelle to do this change. SmPl patch:
@rule_1@
identifier res;
expression ioremap_res;
identifier pdev;
@@
-struct resource *res;
...
-res = platform_get_resource(pdev,...);
-ioremap_res = devm_ioremap_resource(...);
+ioremap_res = devm_platform_ioremap_resource(pdev,0);

Cc: Marek Vasut <marex@denx.de>
Cc: Stefan Agner <stefan@agner.ch>
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Anusha Srivatsa <asrivats@redhat.com>
Link: https://patchwork.freedesktop.org/patch/640852/?series=144073&rev=5

drm/hisilicon: move to devm_platform_ioremap_resource() usage

Replace platform_get_resource + devm_ioremap_resource
with just devm_platform_ioremap_resource()

Used Coccinelle to do this change. SmPl patch:
@rule_1@
identifier res;
expression ioremap_res;
identifier pdev;
@@
-struct resource *res;
...
-res = platform_get_resource(pdev,...);
-ioremap_res = devm_ioremap_resource(...);
+ioremap_res = devm_platform_ioremap_resource(pdev,0);

Cc: Xinliang Liu <xinliang.liu@linaro.org>
Cc: Tian Tao <tiantao6@hisilicon.com>
Cc: Xinwei Kong <kong.kongxinwei@hisilicon.com>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Yongqin Liu <yongqin.liu@linaro.org>
Cc: John Stultz <jstultz@google.com>
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Anusha Srivatsa <asrivats@redhat.com>
Link: https://patchwork.freedesktop.org/patch/640850/?series=144073&rev=5

drm/fsl-dcu: move to devm_platform_ioremap_resource() usage

Replace platform_get_resource + devm_ioremap_resource
with just devm_platform_ioremap_resource()

Used Coccinelle to do this change. SmPl patch:
@rule_1@
identifier res;
expression ioremap_res;
identifier pdev;
@@
-struct resource *res;
...
-res = platform_get_resource(pdev,...);
-ioremap_res = devm_ioremap_resource(...);
+ioremap_res = devm_platform_ioremap_resource(pdev,0);

Cc: Stefan Agner <stefan@agner.ch>
Cc: Alison Wang <alison.wang@nxp.com>
Reviewed-by: Maxime Ripard <mripard@kernel.org>
Signed-off-by: Anusha Srivatsa <asrivats@redhat.com>
Link: https://patchwork.freedesktop.org/patch/640851/?series=144073&rev=5

drm/ttm: test private resv obj on release/destroy

Test the fences in the private dma_resv object instead of the pointer to
a potentially shared dma_resv object.

This only matters for imported BOs with an SG table since those don't
get their dma_resv pointer replaced on release.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: James Zhu <James.Zhu@amd.com>
Tested-by: James Zhu <James.Zhu@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250129152849.15777-1-christian.koenig@amd.com

drm/sched: Update timedout_job()'s documentation

drm_sched_backend_ops.timedout_job()'s documentation is outdated. It
mentions the deprecated function drm_sched_resubmit_jobs(). Furthermore,
it does not point out the important distinction between hardware and
firmware schedulers.

Since firmware schedulers typically only use one entity per scheduler,
timeout handling is significantly more simple because the entity the
faulted job came from can just be killed without affecting innocent
processes.

Update the documentation with that distinction and other details.

Reformat the docstring to work to a unified style with the other
handles.

Acked-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Philipp Stanner <phasta@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305130551.136682-5-phasta@kernel.org

drm/sched: Document run_job() refcount hazard

drm_sched_backend_ops.run_job() returns a dma_fence for the scheduler.
That fence is signalled by the driver once the hardware completed the
associated job. The scheduler does not increment the reference count on
that fence, but implicitly expects to inherit this fence from run_job().

This is relatively subtle and prone to misunderstandings.

This implies that, to keep a reference for itself, a driver needs to
call dma_fence_get() in addition to dma_fence_init() in that callback.

It's further complicated by the fact that the scheduler even decrements
the refcount in drm_sched_run_job_work() since it created a new
reference in drm_sched_fence_scheduled(). It does, however, still use
its pointer to the fence after calling dma_fence_put() - which is safe
because of the aforementioned new reference, but actually still violates
the refcounting rules.

Move the call to dma_fence_put() to the position behind the last usage
of the fence.

Suggested-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Philipp Stanner <pstanner@redhat.com>
Reviewed-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Philipp Stanner <phasta@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305130551.136682-4-phasta@kernel.org

drm/sched: Adjust outdated docu for run_job()

The documentation for drm_sched_backend_ops.run_job() mentions a certain
function called drm_sched_job_recovery(). This function does not exist.
What's actually meant is drm_sched_resubmit_jobs(), which is by now also
deprecated.

Furthermore, the scheduler expects to "inherit" a reference on the fence
from the run_job() callback. This, so far, is also not documented.

Remove the mention of the removed function.

Discourage the behavior of drm_sched_backend_ops.run_job() being called
multiple times for the same job.

Document the necessity of incrementing the refcount in run_job().

Acked-by: Danilo Krummrich <dakr@kernel.org>
Signed-off-by: Philipp Stanner <phasta@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305130551.136682-3-phasta@kernel.org

dma-buf: drop caching of sg_tables

That was purely for the transition from static to dynamic dma-buf
handling and can be removed again now.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Simona Vetter <simona.vetter@ffwll.ch>
Reviewed-by: Dmitry Osipenko <dmitry.osipenko@collabora.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250211163109.12200-5-christian.koenig@amd.com

dma-buf: dma-buf: stop mapping sg_tables on attach v2

As a workaround to smoothly transit from static to dynamic DMA-buf
handling we cached the sg_table on attach if dynamic handling mismatched
between exporter and importer.

Since Dmitry and Thomas cleaned that up and also documented the lock
handling we can drop this workaround now.

V2: implement Sima's comments

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Simona Vetter <simona.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20250211163109.12200-4-christian.koenig@amd.com

dma-buf/dma-fence: remove unnecessary callbacks

The fence_value_str and timeline_value_str callbacks were just an
unnecessary abstraction in the SW sync implementation.

The only caller of those callbacks already knew that the fence in
questions is a timeline_fence. So print the values directly instead
of using a redirection.

Additional to that remove the implementations from virtgpu and vgem.
As far as I can see those were never used in the first place.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Simona Vetter <simona.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20250211163109.12200-3-christian.koenig@amd.com

dma-buf: fix incorrect dma-fence documentation v2

There isn't much worse than documentation giving an incorrect advise.
Grabbing a spinlock while interrupts are disabled usually means that you
must also disable interrupts for all other uses of this spinlock.

Otherwise really hard to debug issues can occur. So fix that invalid
documentation.

v2: use Dmitry's suggestion on the documentation

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Simona Vetter <simona.vetter@ffwll.ch> (v1)
Link: https://patchwork.freedesktop.org/patch/msgid/20250211163109.12200-2-christian.koenig@amd.com

drm/gma500: fix inconsistent indenting warning

Fix below inconsistent indenting smatch warning.
smatch warnings:
drivers/gpu/drm/gma500/cdv_device.c:218 cdv_errata() warn: inconsistent indenting

Signed-off-by: Charles Han <hanchunchao@inspur.com>
Signed-off-by: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250305084911.6394-1-hanchunchao@inspur.com

drm/gma500: Replace deprecated strncpy() with strscpy()

strncpy() is deprecated for NUL-terminated destination buffers. Use
strscpy() instead and remove the manual NUL-termination.

Compile-tested only.

Link: https://github.com/KSPP/linux/issues/90
Cc: linux-hardening@vger.kernel.org
Signed-off-by: Thorsten Blum <thorsten.blum@linux.dev>
Reviewed-by: Kees Cook <kees@kernel.org>
Signed-off-by: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250225203932.334123-1-thorsten.blum@linux.dev

drm/prime: Use dma_buf from GEM object instance

Avoid dereferencing struct drm_gem_object.import_attach for the
imported dma-buf. The dma_buf field in the GEM object instance refers
to the same buffer. Prepares to make import_attach optional.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Anusha Srivatsa <asrivats@redhat.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250226172457.217725-11-tzimmermann@suse.de

drm/mipi-dbi: Test for imported buffers with drm_gem_is_imported()

Instead of testing import_attach for imported GEM buffers, invoke
drm_gem_is_imported() to do the test.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Anusha Srivatsa <asrivats@redhat.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250226172457.217725-10-tzimmermann@suse.de