drm/amdgpu: fix pm notifier handling

Set the s3/s0ix and s4 flags in the pm notifier so that we can skip
the resource evictions properly in pm prepare based on whether
we are suspending or hibernati

drm/amdgpu: fix pm notifier handling

Set the s3/s0ix and s4 flags in the pm notifier so that we can skip
the resource evictions properly in pm prepare based on whether
we are suspending or hibernating. Drop the eviction as processes
are not frozen at this time, we we can end up getting stuck trying
to evict VRAM while applications continue to submit work which
causes the buffers to get pulled back into VRAM.

v2: Move suspend flags out of pm notifier (Mario)

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4178
Fixes: 2965e6355dcd ("drm/amd: Add Suspend/Hibernate notification callback support")
Cc: Mario Limonciello <[email protected]>
Reviewed-by: Mario Limonciello <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>
(cherry picked from commit 06f2dcc241e7e5c681f81fbc46cacdf4bfd7d6d7)
Cc: [email protected]

show more ...

drm/amd: Forbid suspending into non-default suspend states

On systems that default to 'deep' some userspace software likes
to try to suspend in 'deep' first. If there is a failure for any
reason (s

drm/amd: Forbid suspending into non-default suspend states

On systems that default to 'deep' some userspace software likes
to try to suspend in 'deep' first. If there is a failure for any
reason (such as -ENOMEM) the failure is ignored and then it will
try to use 's2idle' as a fallback. This fails, but more importantly
it leads to graphical problems.

Forbid this behavior and only allow suspending in the last state
supported by the system.

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4093
Acked-by: Alex Deucher <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mario Limonciello <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>
(cherry picked from commit 2aabd44aa8a3c08da3d43264c168370f6da5e81d)

show more ...

drm/amd: Handle being compiled without SI or CIK support better

If compiled without SI or CIK support but amdgpu tries to load it
will run into failures with uninitialized callbacks.

Show a nicer m

drm/amd: Handle being compiled without SI or CIK support better

If compiled without SI or CIK support but amdgpu tries to load it
will run into failures with uninitialized callbacks.

Show a nicer message in this case and fail probe instead.

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4050
Signed-off-by: Mario Limonciello <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>
Cc: [email protected]

show more ...

Documentation/amdgpu: Add debug_mask documentation

Add description for debug_mask bit options.

Signed-off-by: Lijo Lazar <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>
S

Documentation/amdgpu: Add debug_mask documentation

Add description for debug_mask bit options.

Signed-off-by: Lijo Lazar <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amd/pm: Add debug bit for smu pool allocation

In certain cases, it's desirable to avoid PMFW log transactions to
system memory. Add a mask bit to decide whether to allocate smu pool in
device me

drm/amd/pm: Add debug bit for smu pool allocation

In certain cases, it's desirable to avoid PMFW log transactions to
system memory. Add a mask bit to decide whether to allocate smu pool in
device memory or system memory.

Signed-off-by: Lijo Lazar <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: adjust drm_firmware_drivers_only() handling

Move to probe so we can check the PCI device type and
only apply the drm_firmware_drivers_only() check for
PCI DISPLAY classes. Also add a mo

drm/amdgpu: adjust drm_firmware_drivers_only() handling

Move to probe so we can check the PCI device type and
only apply the drm_firmware_drivers_only() check for
PCI DISPLAY classes. Also add a module parameter to
override the nomodeset kernel parameter as a workaround
for platforms that have this hardcoded on their kernel
command lines.

Reviewed-by: Kent Russell <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: drop drm_firmware_drivers_only()

There are a number of systems and cloud providers out there
that have nomodeset hardcoded in their kernel parameters
to block nouveau for the nvidia driv

drm/amdgpu: drop drm_firmware_drivers_only()

There are a number of systems and cloud providers out there
that have nomodeset hardcoded in their kernel parameters
to block nouveau for the nvidia driver. This prevents the
amdgpu driver from loading. Unfortunately the end user cannot
easily change this. The preferred way to block modules from
loading is to use modprobe.blacklist=<driver>. That is what
providers should be using to block specific drivers.

Drop the check to allow the driver to load even when nomodeset
is specified on the kernel command line.

Reviewed-by: Kent Russell <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amd: Keep display off while going into S4

When userspace invokes S4 the flow is:

1) amdgpu_pmops_prepare()
2) amdgpu_pmops_freeze()
3) Create hibernation image
4) amdgpu_pmops_thaw()
5) Write o

drm/amd: Keep display off while going into S4

When userspace invokes S4 the flow is:

1) amdgpu_pmops_prepare()
2) amdgpu_pmops_freeze()
3) Create hibernation image
4) amdgpu_pmops_thaw()
5) Write out image to disk
6) Turn off system

Then on resume amdgpu_pmops_restore() is called.

This flow has a problem that because amdgpu_pmops_thaw() is called
it will call amdgpu_device_resume() which will resume all of the GPU.

This includes turning the display hardware back on and discovering
connectors again.

This is an unexpected experience for the display to turn back on.
Adjust the flow so that during the S4 sequence display hardware is
not turned back on.

Reported-by: Xaver Hugl <[email protected]>
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/2038
Cc: Muhammad Usama Anjum <[email protected]>
Tested-by: Muhammad Usama Anjum <[email protected]>
Acked-by: Alex Deucher <[email protected]>
Acked-by: Harry Wentland <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mario Limonciello <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>
(cherry picked from commit 68bfdc8dc0a1a7fdd9ab61e69907ae71a6fd3d91)

show more ...

drm/amd/display: allow 256B DCC max compressed block sizes on gfx12

The hw supports it.

Signed-off-by: Marek Olšák <[email protected]>
Acked-by: Alex Deucher <[email protected]>
Signed-of

drm/amd/display: allow 256B DCC max compressed block sizes on gfx12

The hw supports it.

Signed-off-by: Marek Olšák <[email protected]>
Acked-by: Alex Deucher <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amd: Keep display off while going into S4

When userspace invokes S4 the flow is:

1) amdgpu_pmops_prepare()
2) amdgpu_pmops_freeze()
3) Create hibernation image
4) amdgpu_pmops_thaw()
5) Write o

drm/amd: Keep display off while going into S4

When userspace invokes S4 the flow is:

1) amdgpu_pmops_prepare()
2) amdgpu_pmops_freeze()
3) Create hibernation image
4) amdgpu_pmops_thaw()
5) Write out image to disk
6) Turn off system

Then on resume amdgpu_pmops_restore() is called.

This flow has a problem that because amdgpu_pmops_thaw() is called
it will call amdgpu_device_resume() which will resume all of the GPU.

This includes turning the display hardware back on and discovering
connectors again.

This is an unexpected experience for the display to turn back on.
Adjust the flow so that during the S4 sequence display hardware is
not turned back on.

Reported-by: Xaver Hugl <[email protected]>
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/2038
Cc: Muhammad Usama Anjum <[email protected]>
Tested-by: Muhammad Usama Anjum <[email protected]>
Acked-by: Alex Deucher <[email protected]>
Acked-by: Harry Wentland <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mario Limonciello <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: Create a debug option to disable ring reset

Prior to the addition of ring reset, the debug option
`debug_disable_soft_recovery` could be used to force a full device
reset. Now that we ha

drm/amdgpu: Create a debug option to disable ring reset

Prior to the addition of ring reset, the debug option
`debug_disable_soft_recovery` could be used to force a full device
reset. Now that we have ring reset, create a debug option to disable
them in amdgpu, forcing the driver to go with the full device
reset path again when both options are combined.

This option is useful for testing and debugging purposes when one wants
to test the full reset from userspace.

Signed-off-by: André Almeida <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: Add flags to distinguish vf/pf/pt mode

Add extra flag definition for ids_flag field to distinguish
between vf/pf/pt modes

v2: Updated kms driver minor version & removed pf check as defa

drm/amdgpu: Add flags to distinguish vf/pf/pt mode

Add extra flag definition for ids_flag field to distinguish
between vf/pf/pt modes

v2: Updated kms driver minor version & removed pf check as default is 0
v3: Fix up version (Alex)
v4: rebase (Alex)

Proposed userspace:
https://github.com/ROCm/amdsmi/commit/e663bed7d6b3df79f5959e73981749b1f22ec698

Signed-off-by: Asad Kamal <[email protected]>
Reviewed-by: Lijo Lazar <[email protected]>
Reviewed-by: Hawking Zhang <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: Update usage for bad page threshold

The driver's behavior varies based on
the configuration of amdgpu_bad_page_threshold setting

Signed-off-by: Hawking Zhang <[email protected]>
Rev

drm/amdgpu: Update usage for bad page threshold

The driver's behavior varies based on
the configuration of amdgpu_bad_page_threshold setting

Signed-off-by: Hawking Zhang <[email protected]>
Reviewed-by: Tao Zhou <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amd: Mark amdgpu.gttsize parameter as deprecated and show warnings on use

When not set `gttsize` module parameter by default will get the
value to use for the GTT pool from the TTM page limit, w

drm/amd: Mark amdgpu.gttsize parameter as deprecated and show warnings on use

When not set `gttsize` module parameter by default will get the
value to use for the GTT pool from the TTM page limit, which is
set by a separate module parameter.

This inevitably leads to people not sure which one to set when they
want more addressable memory for the GPU, and you'll end up seeing
instructions online saying to set both.

Add some messages to try to guide people both who are using or misusing
the parameters and mark the parameter as deprecated with the plan to
drop it after the next LTS kernel release.

Reviewed-by: Alex Deucher <[email protected]>
Signed-off-by: Mario Limonciello <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: bump version for RV/PCO compute fix

Bump the driver version for RV/PCO compute stability fix
so mesa can use this check to enable compute queues on
RV/PCO.

Reviewed-by: Lijo Lazar <lijo

drm/amdgpu: bump version for RV/PCO compute fix

Bump the driver version for RV/PCO compute stability fix
so mesa can use this check to enable compute queues on
RV/PCO.

Reviewed-by: Lijo Lazar <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>
Cc: [email protected] # 6.12.x

show more ...

drm/amdgpu: add a BO metadata flag to disable write compression for Vulkan

Vulkan can't support DCC and Z/S compression on GFX12 without
WRITE_COMPRESS_DISABLE in this commit or a completely differe

drm/amdgpu: add a BO metadata flag to disable write compression for Vulkan

Vulkan can't support DCC and Z/S compression on GFX12 without
WRITE_COMPRESS_DISABLE in this commit or a completely different DCC
interface.

AMDGPU_TILING_GFX12_SCANOUT is added because it's already used by userspace.

Cc: [email protected] # 6.12.x
Signed-off-by: Marek Olšák <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amd: Clarify kdoc for amdgpu.gttsize

Effectively amdgpu.gttsize gets set to ~1/2 of RAM, but that's controlled
by what the TTM page limit is set to. Clarify the kdoc.

Reviewed-by: Alex Deucher

drm/amd: Clarify kdoc for amdgpu.gttsize

Effectively amdgpu.gttsize gets set to ~1/2 of RAM, but that's controlled
by what the TTM page limit is set to. Clarify the kdoc.

Reviewed-by: Alex Deucher <[email protected]>
Signed-off-by: Mario Limonciello <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: Mark debug KFD module params as unsafe

Mark options only meant to be used for debugging as unsafe so that the
kernel is tainted when they are used.

Signed-off-by: Kent Russell <kent.rus

drm/amdgpu: Mark debug KFD module params as unsafe

Mark options only meant to be used for debugging as unsafe so that the
kernel is tainted when they are used.

Signed-off-by: Kent Russell <[email protected]>
Reviewed-by: Alex Deucher <[email protected]>
Reviewed-by: Felix Kuehling <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: mark a bunch of module parameters unsafe

We sometimes have people trying to use debugging options in production
environments.

Mark options only meant to be used for debugging as unsafe

drm/amdgpu: mark a bunch of module parameters unsafe

We sometimes have people trying to use debugging options in production
environments.

Mark options only meant to be used for debugging as unsafe so that the
kernel is tainted when they are used.

Signed-off-by: Christian König <[email protected]>
Acked-by: Felix Kuehling <[email protected]>
Acked-by: Simona Vetter <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amd: Add Suspend/Hibernate notification callback support

As part of the suspend sequence VRAM needs to be evicted on dGPUs.
In order to make suspend/resume more reliable we moved this into
the p

drm/amd: Add Suspend/Hibernate notification callback support

As part of the suspend sequence VRAM needs to be evicted on dGPUs.
In order to make suspend/resume more reliable we moved this into
the pmops prepare() callback so that the suspend sequence would fail
but the system could remain operational under high memory usage suspend.

Another class of issues exist though where due to memory fragementation
there isn't a large enough contiguous space and swap isn't accessible.

Add support for a suspend/hibernate notification callback that could
evict VRAM before tasks are frozen. This should allow paging out to swap
if necessary.

Link: https://github.com/ROCm/ROCK-Kernel-Driver/issues/174
Link: https://gitlab.freedesktop.org/drm/amd/-/issues/3476
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/2362
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3781
Reviewed-by: Lijo Lazar <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mario Limonciello <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm: remove driver date from struct drm_driver and all drivers

We stopped using the driver initialized date in commit 7fb8af6798e8
("drm: deprecate driver date") and (eventually) started returning "

drm: remove driver date from struct drm_driver and all drivers

We stopped using the driver initialized date in commit 7fb8af6798e8
("drm: deprecate driver date") and (eventually) started returning "0"
for drm_version ioctl instead.

Finish the job, and remove the unused date member from struct
drm_driver, its initialization from drivers, along with the common
DRIVER_DATE macros.

v2: Also update drivers/accel (kernel test robot)

Reviewed-by: Javier Martinez Canillas <[email protected]>
Acked-by: Alex Deucher <[email protected]>
Acked-by: Simon Ser <[email protected]>
Acked-by: Jeffrey Hugo <[email protected]>
Acked-by: Lucas De Marchi <[email protected]>
Acked-by: Dmitry Baryshkov <[email protected]> # msm
Reviewed-by: Thomas Zimmermann <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/1f2bf2543aed270a06f6c707fd6ed1b78bf16712.1733322525.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <[email protected]>

show more ...

drm/client: Move public client header to clients/ subdirectory

Move the public header file drm_client_setup.h to the clients/
subdirectory and update all drivers. No functional changes.

Signed-off-

drm/client: Move public client header to clients/ subdirectory

Move the public header file drm_client_setup.h to the clients/
subdirectory and update all drivers. No functional changes.

Signed-off-by: Thomas Zimmermann <[email protected]>
Reviewed-by: Jocelyn Falempe <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/amdgpu: fix comment about amdgpu.abmlevel defaults

Since 040fdcde288a ("drm/amdgpu: respect the abmlevel module parameter value
if it is set"), the default value for amdgpu.abmlevel was set to -

drm/amdgpu: fix comment about amdgpu.abmlevel defaults

Since 040fdcde288a ("drm/amdgpu: respect the abmlevel module parameter value
if it is set"), the default value for amdgpu.abmlevel was set to -1, or auto.
However, the comment explaining the default value was not updated to reflect
the change (-1, or auto; not -1, or disabled).

Clarify that the default value (-1) means auto.

Fixes: 040fdcde288a ("drm/amdgpu: respect the abmlevel module parameter value if it is set")
Reported-by: Ruikai Liu <[email protected]>
Signed-off-by: Mingcong Bai <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: clean up the suspend_complete

To check the status of S3 suspend completion,
use the PM core pm_suspend_global_flags bit(1)
to detect S3 abort events. Therefore, clean up
the AMDGPU drive

drm/amdgpu: clean up the suspend_complete

To check the status of S3 suspend completion,
use the PM core pm_suspend_global_flags bit(1)
to detect S3 abort events. Therefore, clean up
the AMDGPU driver's private flag suspend_complete.

Signed-off-by: Prike Liang <[email protected]>
Reviewed-by: Lijo Lazar <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...

drm/amdgpu: Place NPS mode request on unload

If a user has requested NPS mode switch, place the request through PSP
during unload of the driver. For devices which are part of a hive, all
requests ar

drm/amdgpu: Place NPS mode request on unload

If a user has requested NPS mode switch, place the request through PSP
during unload of the driver. For devices which are part of a hive, all
requests are placed together. If one of them fails, revert back to the
current NPS mode.

Signed-off-by: Lijo Lazar <[email protected]>
Signed-off-by: Rajneesh Bhardwaj <[email protected]>
Reviewed-by: Feifei Xu <[email protected]>
Signed-off-by: Alex Deucher <[email protected]>

show more ...