drm/amdkfd: Cleanup kfd_process if init_cwsr_apu fails

If init_cwsr_apu fails, we currently leave the kfd_process structure in
place anyway. The next kfd_open will then succeed, using the existing
kfd_process structure. Fix that by cleaning up the kfd_process after a
failure in init_cwsr_apu.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Philip Yang <philip.yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Use mmu_notifier_get

We use mmu_notifier_put to free the MMU notifier. That needs to be
paired with mmu_notifier_get to work correctly. Othewrise the next patch
would cause a kernel oops.

Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Philip Yang <philip.yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: reuse current context instead of recreating one"

This reverts commit 8866a67ab86cc0812e65c04f1ef02bcc41e24d68.

Reason for revert: This breaks hotplug of HDMI on some systems,
resulting in a blank screen. Caused general hangs on boot/hotplugs.

Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1487
Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1492
Bug: https://bugzilla.kernel.org/show_bug.cgi?id=211649
Signed-off-by: Anson Jacob <Anson.Jacob@amd.com>
Reviewed-by: Bhawanpreet Lakha <bhawanpreet.lakha@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Simplify bool conversion

Fix the following coccicheck warning:
./drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c:8142:16-21: WARNING:
conversion to bool not needed here

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: do not issue message while write "r" into pp_od_clk_voltage

We should commit the value after restore them back to default as well.

$ echo "r" > pp_od_clk_voltage
$ echo "c" > pp_od_clk_voltage

Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Old sequence for HUBP blank

New proposed sequence for HUBP blanking causes regressions where the
hardware would fail to enter blank which triggers an assert in the new
sequence. This change brings back the old sequence.

Fixes: 985faf2c4ecb60 ("drm/amd/display: New sequence for HUBP blank")
Signed-off-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Bhawanpreet Lakha <bhawanpreet.lakha@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: mark local function as static

Mark amdgpu_ras_debugfs_create_ctrl_node() as static.

Fixes: eb14235668777b ("drm/amdgpu: do not keep debugfs dentry")
Reported-by: kernel test robot <lkp@intel.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.123

DC version 3.2.123 brings improvements in multiple areas.
In summary, we highlight:

- Firmware release 0.0.52
- Bug fixes on MPC OGAM sequence, display experience, etc.
- Improvements on timing, transmitter control, etc.

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: [FW Promotion] Release 0.0.52

[How]
Add new aux cmd fields for acquire and release

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Implement transmitter control v1.7

[Why]
Moving definition of transmitter control from atomfirmware
to internal header.

[How]
Update the command table code to call 1.7 and make it the
default fallback path.

Signed-off-by: Eric Bernstein <eric.bernstein@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: AVMUTE simplification

[Why]
Simplify AVMUTE logic in coding

[How]
Avoid multiple calls on TMDS AVMUTE
as updated logic now demands it.

Signed-off-by: Chris Park <Chris.Park@amd.com>
Reviewed-by: Charlene Liu <Charlene.Liu@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Copy over soc values before bounding box creation

[Why]
With certain fclock overclocks, state 1 may be chosen
as the closest clock level. This may result in this state
being empty if not populated beforehand, resulting in
black screens and screen corruption.

[How]
Copy over all soc states to clock_limits before bounding
box creation to avoid any cases with empty states.

Signed-off-by: Sung Lee <sung.lee@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Reviewed-by: Yongqiang Sun <yongqiang.sun@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add Freesync HDMI support to DM

[Why]
Add necessary support for Freesync HDMI in Linux DM

[How]
- Support Freesync HDMI by calling DC interace
- Report Freesync capability to vrr_range debugfs from DRM
- Depends on coming DMCU/DMUB firmware to enable feature

Signed-off-by: Stylon Wang <stylon.wang@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Populate dcn2.1 bounding box before state duplication

[Why]
If system is overclocked, only 1 bounding box state will
be sent by SMU. This results in an empty state being copied
for DML calculations causing black screens and corruption.

[How]
Fully populate bounding box before duplicating last state.

Signed-off-by: Sung Lee <sung.lee@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix MPC OGAM power on/off sequence

[Why]
Color corruption can occur on bootup into a login
manager that applies a non-linear gamma LUT because
the LUT may not actually be powered on before writing.

It's cleared on the next full pipe reprogramming as
we switch to LUTB from LUTA and the pipe accessing
the LUT has taken it out of light sleep mode.

[How]
The MPCC_OGAM_MEM_PWR_FORCE register does not force
the current power mode when set to 0. It only forces
when set light sleep, deep sleep or shutdown.

The register to actually force power on and ignore
sleep modes is MPCC_OGAM_MEM_PWR_DIS - a value of 0
will enable power requests and a value of 1 will
disable them.

When PWR_FORCE!=0 is combined with PWR_DIS=0 then
MPCC OGAM memory is forced into the state specified
by the force bits.

If PWR_FORCE is 0 then it respects the mode specified
by MPCC_OGAM_MEM_LOW_PWR_MODE if the RAM LUT is not
in use.

We set that bit to shutdown on low power, but otherwise
it inherits from bootup defaults.

So for the fix:

1. Update the sequence to "force" power on when needed

We can use MPCC_OGAM_MEM_PWR_DIS for this to turn on the
memory even when the block is in bypass and pending to be
enabled for the next frame.

We need this for both low power enabled or disabled.

If we don't set this then we can run into issues when we
first program the LUT from bootup.

2. Don't apply FORCE_SEL

Once we enable power requests with DIS=0 we run into the
issue of the RAM being forced into light sleep and being
unusable for display output. Leave this 0 like we used to
for DCN20.

3. Rely on MPCC OGAM init to determine light sleep/deep sleep

MPC low power debug mode isn't enabled on any ASIC currently
but we'll respect the setting determined during init if it
is.

Lightly tested as working with IGT tests and desktop color
adjustment.

4. Change the MPC resource default for DCN30

It was interleaving the dcn20 and dcn30 versions before
depending on the sequence.

5. REG_WAIT for it to be on whenever we're powering up the
memory

Otherwise we can write register values too early and we'll
get corruption.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Eric Yang <eric.yang2@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add dc_dmub_srv helpers for in/out DMCUB commands

[Why]
We added these in DMCUB for runtime feature detection
but we didn't have helpers to call these with DC error
handling/logging.

[How]
Add helpers.

Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Eric Yang <eric.yang2@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: changing sr exit latency

[Why]
Hardware team remeasured, need to update timings
to increase latency slightly and avoid intermittent
underflows.

[How]
sr exit latency update.

Signed-off-by: Martin Leung <martin.leung@amd.com>
Reviewed-by: Alvin Lee <Alvin.Lee2@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: Unblank hubp based on plane visibility"

This reverts commit fd1c85d3ac2ccfec33b007399e6677b41899a888

[Why]
We are not implementing the planned new HW sequence
to disable HUBP.

[How]
Revert most related changes to minimize possibility
of regression.

Signed-off-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Reviewed-by: Martin Leung <Martin.Leung@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: New path for enabling DPG"

This reverts commit f8e792dc5c45d306a542b9ee991ae80e1c78e4c5

[Why]
We are not implementing the planned new HW
sequence for HUBP disable.

[How]
Revert most related changes to minimize regressions.

Signed-off-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Reviewed-by: Martin Leung <Martin.Leung@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: remove global optimize seamless boot stream count

[Why]
In following sequence driver will add counter twice on
same edp stream.
1. Boot into OS.
2. Set timing with edp only.
3. Set timing with edp and external monitor.
4. Set visibility on for edp.

Step 2 and 3 will add seamless boot counter twice and
subtract it once in step 4.

[How]
Remove global counter and calculate it is used.

Signed-off-by: Lewis Huang <Lewis.Huang@amd.com>
Reviewed-by: Martin Leung <Martin.Leung@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Change ABM sample rate

[Why]
To get the pixel statistics on every frame, change ABM
sample rate from 2 to 1.

[How]
Change LS, HS and BL sample rate to 1.

Signed-off-by: Po-Ting Chen <robin.chen@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon: do not use drm middle layer for debugfs (v2)

Use debugfs API directly instead of drm middle layer.

v2: squash in build fix (Alex)

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon: add rdev in ring struct

Retrieving radeon device struct from ring struct will be
used in next patch where debugfs's show function can only pass
one private data pointer.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon: Remove unused function pointer typedef radeon_packet3_check_t

Remove the 'radeon_packet3_check_t' typedef as it is not used.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Chen Lin <chen.lin5@zte.com.cn>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon/nislands_smc.h: Replace one-element array with flexible-array member in struct NISLANDS_SMC_SWSTATE

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Use flexible-array member in struct NISLANDS_SMC_SWSTATE, instead of
one-element array.

Also, this helps with the ongoing efforts to enable -Warray-bounds by
fixing the following warnings:

drivers/gpu/drm/radeon/ni_dpm.c: In function ‘ni_convert_power_state_to_smc’:
drivers/gpu/drm/radeon/ni_dpm.c:2521:20: warning: array subscript 1 is above array bounds of ‘NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
 2521 |   smc_state->levels[i].dpm2.MaxPS =
      |   ~~~~~~~~~~~~~~~~~^~~
drivers/gpu/drm/radeon/ni_dpm.c:2523:20: warning: array subscript 1 is above array bounds of ‘NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
 2523 |   smc_state->levels[i].dpm2.NearTDPDec = NISLANDS_DPM2_NEAR_TDP_DEC;
      |   ~~~~~~~~~~~~~~~~~^~~
drivers/gpu/drm/radeon/ni_dpm.c:2524:20: warning: array subscript 1 is above array bounds of ‘NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
 2524 |   smc_state->levels[i].dpm2.AboveSafeInc = NISLANDS_DPM2_ABOVE_SAFE_INC;
      |   ~~~~~~~~~~~~~~~~~^~~
drivers/gpu/drm/radeon/ni_dpm.c:2525:20: warning: array subscript 1 is above array bounds of ‘NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
 2525 |   smc_state->levels[i].dpm2.BelowSafeInc = NISLANDS_DPM2_BELOW_SAFE_INC;
      |   ~~~~~~~~~~~~~~~~~^~~
drivers/gpu/drm/radeon/ni_dpm.c:2526:35: warning: array subscript 1 is above array bounds of ‘NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
 2526 |   smc_state->levels[i].stateFlags |=
      |                                   ^~
drivers/gpu/drm/radeon/ni_dpm.c:2526:35: warning: array subscript 1 is above array bounds of ‘NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
 2526 |   smc_state->levels[i].stateFlags |=
      |   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~
 2527 |    ((i != (state->performance_level_count - 1)) && power_boost_limit) ?
      |    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 2528 |    PPSMC_STATEFLAG_POWERBOOST : 0;
      |    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/radeon/ni_dpm.c:2442:20: warning: array subscript 1 is above array bounds of ‘NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’ {aka ‘struct NISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
 2442 |   smc_state->levels[i + 1].aT = cpu_to_be32(a_t);

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9/process/deprecated.html#zero-length-and-one-element-arrays

Link: https://github.com/KSPP/linux/issues/79
Link: https://github.com/KSPP/linux/issues/109
Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/6023ed54.BfIY+9Uz81I6nq19%25lkp@intel.com/
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct _ATOM_Vega10_GFXCLK_Dependency_Table

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Use flexible-array member in struct _ATOM_Vega10_GFXCLK_Dependency_Table,
instead of one-element array.

Also, this helps with the ongoing efforts to enable -Warray-bounds and
fix the following warning:

drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/vega10_hwmgr.c: In function ‘vega10_get_pp_table_entry_callback_func’:
drivers/gpu/drm/amd/amdgpu/../pm/powerplay/hwmgr/vega10_hwmgr.c:3113:30: warning: array subscript 4 is above array bounds of ‘ATOM_Vega10_GFXCLK_Dependency_Record[1]’ {aka ‘struct _ATOM_Vega10_GFXCLK_Dependency_Record[1]’} [-Warray-bounds]
 3113 |     gfxclk_dep_table->entries[4].ulClk;
      |     ~~~~~~~~~~~~~~~~~~~~~~~~~^~~

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9/process/deprecated.html#zero-length-and-one-element-arrays

Link: https://github.com/KSPP/linux/issues/79
Link: https://github.com/KSPP/linux/issues/109
Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/6023ff3d.WY3sSCkGRQPdPlVo%25lkp@intel.com/
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Replace one-element array with flexible-array in struct SISLANDS_SMC_SWSTATE

There is a regular need in the kernel to provide a way to declare having
a dynamically sized set of trailing elements in a structure. Kernel code
should always use “flexible array members”[1] for these cases. The older
style of one-element or zero-length arrays should no longer be used[2].

Refactor the code according to the use of a flexible-array member in
struct SISLANDS_SMC_SWSTATE, instead of a one-element array, and use
the struct_size() helper to calculate the size for the allocation.

Also, this helps with the ongoing efforts to enable -Warray-bounds and
fix the following warnings:

drivers/gpu/drm/amd/amdgpu/../pm/powerplay/si_dpm.c:2448:20: warning: array
subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’
{aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
drivers/gpu/drm/amd/amdgpu/../pm/powerplay/si_dpm.c:2449:20: warning: array
subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’
{aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
drivers/gpu/drm/amd/amdgpu/../pm/powerplay/si_dpm.c:2450:20: warning: array
subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’
{aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
drivers/gpu/drm/amd/amdgpu/../pm/powerplay/si_dpm.c:2451:20: warning: array
subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’
{aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
drivers/gpu/drm/amd/amdgpu/../pm/powerplay/si_dpm.c:2452:20: warning: array
subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’
{aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]
drivers/gpu/drm/amd/amdgpu/../pm/powerplay/si_dpm.c:5570:20: warning: array
subscript 1 is above array bounds of ‘SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’
{aka ‘struct SISLANDS_SMC_HW_PERFORMANCE_LEVEL[1]’} [-Warray-bounds]

[1] https://en.wikipedia.org/wiki/Flexible_array_member
[2] https://www.kernel.org/doc/html/v5.9/process/deprecated.html#zero-length-and-one-element-arrays

Link: https://github.com/KSPP/linux/issues/79
Link: https://github.com/KSPP/linux/issues/109
Build-tested-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/lkml/6023be58.sk66L%2FV4vuSJI5mI%25lkp@intel.com/
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: fix spelling mistake in various messages "power_dpm_force_perfomance_level"

There are spelling mistakes in error and warning messages, the text
power_dpm_force_perfomance_level is missing a letter r and should be
power_dpm_force_performance_level.  Fix them.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: do not use drm middle layer for debugfs

Use debugfs API directly instead of drm middle layer.

This also includes following debugfs file output changes:
1 amdgpu_evict_vram/amdgpu_evict_gtt output will not contain any braces.
  e.g. (0) --> 0
2 amdgpu_gpu_recover output will print return value of
  amdgpu_device_gpu_recover() instead of not so important "gpu recover"
  message.

v2: * checkpatch.pl: use '0444' instead of S_IRUGO.
    * remove S_IFREG from mode.
    * remove mode variable.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: do not use drm middle layer for debugfs

Use debugfs API directly instead of drm middle layer.

v2: * checkpatch.pl: use '0444' instead of S_IRUGO.
    * remove S_IFREG from mode.
    * remove mode variable.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: do not use drm middle layer for debugfs

Use debugfs API directly instead of drm middle layer.

v2: * checkpatch.pl: use '0444' instead of S_IRUGO.
    * remove S_IFREG from mode.
    * remove mode variable.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: remove CONFIG_DRM_AMDGPU_GART_DEBUGFS

Removed unused CONFIG_DRM_AMDGPU_GART_DEBUGFS code.
We can use umr instead of this gart debugfs.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon: cleanup debugfs

Remove unnecessary debugfs dentries and also radeon_ttm_debugfs_fini()
as drm_debugfs_cleanup() will recursively remove debugfs files.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: do not keep debugfs dentry

Cleanup unnecessary debugfs dentries and surrounding functions.

v3: remove return value check for debugfs_create_file()
v2: remove ttm_debugfs_entries array.
    do not init variables.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Set reference clock to 100Mhz on Renoir (v2)

Fixes the rlc reference clock used for GPU timestamps.
Value is 100Mhz.  Confirmed with hardware team.

v2: reword commit message.

Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1480
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/radeon: OLAND boards don't have VCE

Disable it on those boards.  No functional change, this just
removes the message about VCE failing to initialize.

Bug: https://bugzilla.kernel.org/show_bug.cgi?id=197327
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Fix recursive lock warnings

memalloc_nofs_save/restore are no longer sufficient to prevent recursive
lock warnings when holding locks that can be taken in MMU notifiers. Use
memalloc_noreclaim_save/restore instead.

Fixes: f920e413ff9c ("mm: track mmu notifiers in fs_reclaim_acquire/release")
CC: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org # 5.10.x

drm/amd/display: Add FPU wrappers to dcn21_validate_bandwidth()

dcn21_validate_bandwidth() calls functions that use floating point math.
On my machine this sometimes results in simd exceptions when there are
other FPU users such as KVM virtual machines running. The screen freezes
completely in this case.

Wrapping the function with DC_FP_START()/DC_FP_END() seems to solve the
problem. This mirrors the approach used for dcn20_validate_bandwidth.

Tested on a AMD Ryzen 7 PRO 4750U (Renoir).

Bug: https://bugzilla.kernel.org/show_bug.cgi?id=206987
Signed-off-by: Jan Kokemüller <jan.kokemueller@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/amd/display: Fix potential integer overflow

Fix potential integer overflow by casting actual_calculated_clock_100hz
to u64, in order to give the compiler complete information about the
proper arithmetic to use.

Notice that such variable is used in a context that expects
an expression of type u64 (64 bits, unsigned) and the following
expression is currently being evaluated using 32-bit arithmetic:

actual_calculated_clock_100hz * post_divider

Fixes: 7a03fdf628af ("drm/amd/display: fix 64bit division issue on 32bit OS")
Addresses-Coverity-ID: 1501691 ("Unintentional integer overflow")
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/display: remove hdcp_srm sysfs on device removal

Fixes: 9037246bb2da5 ("drm/amd/display: Add sysfs interface for set/get srm")

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: fix CGTS_TCC_DISABLE register offset on gfx10.3

This fixes incorrect TCC harvesting info reported to userspace.
The impact was a very very tiny performance degradation (unnecessary
GL2 cache flushes).

Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/radeon: Simplify bool comparison

Fix the following coccicheck warning:

./drivers/gpu/drm/radeon/rs690.c:190:6-35: WARNING: Comparison to bool.

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Simplify bool comparison

Fix the following coccicheck warning:

./drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h:319:11-23:
WARNING: Comparison to bool.

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Jiapeng Chong <jiapeng.chong@linux.alibaba.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: Update NV1x SR latency values"

This reverts commit 4a3dea8932d3b1199680d2056dd91d31d94d70b7.

This causes blank screens for some users.

Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1388
Cc: Alvin Lee <alvin.lee2@amd.com>
Cc: Jun Lei <Jun.Lei@amd.com>
Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

drm/amdgpu: fix unnecessary NULL check warnings

Remove NULL checks before vfree() to fix these warnings:
drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c:102:2-8: WARNING: NULL
check before some freeing functions is not needed.

Signed-off-by: Tian Tao <tiantao6@hisilicon.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: make the error log more clear for fine grain tuning function

This patch is to make the error log more clear for fine grain tuning
function, it covers Raven/Raven2/Picasso/Renoir/Vangogh.
The fine grain tuning function uses the sysfs file -- pp_od_clk_voltage,
but only when another sysfs file -- power_dpm_force_performance_level is
switched to "manual" mode, it is allowed to access "pp_od_clk_voltage".

Signed-off-by: Xiaojian Du <Xiaojian.Du@amd.com>
Acked-by: Huang Rui <ray.huang@amd.com>
Acked-by: Kevin Wang <kevin1.wang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: extend MAX_KIQ_REG_TRY to 1000

Extend retry times of KIQ to avoid starvation situation caused by
long time full access of GPU by other VFs.

Signed-off-by: Jiawei Gu <Jiawei.Gu@amd.com>
Reviewed-by: Emily.Deng <Emily.Deng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.122

This version brings along following fixes:
- Fix error handling in dc/dcn10/dcn10_hw_sequencer_debug.c
- DIO Supported for virtual displays
- memset dmub_rb_cmd to 0 which fixes issues talking with firmware
- fix 64bit division issue on 32bit OS
- move edp sink present detection to hw init to fix hang
- reverts a patch that caused 3x4K60 displays over MST
  with DSC enabled to not light up

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: [FW Promotion] Release 0.0.51

Add new definition and union for dmub_rb_cmd_dig1_transmitter_control

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix type mismatch error for return variable

It is possible for ret_vsnprintf to be assigned negative value in
error cases. As an unsigned variable, negative values which
are stored in their 2's complement form gets treated as a positive
number. This will led to treating bad cases as good ones.
eg: -1 gets stored as 0xFFFFFFFF on a 32 bit system

Signed-off-by: Dale Zhao <dale.zhao@amd.com>
Reviewed-by: Yongqiang Sun <yongqiang.sun@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: DIO Supported for virtual displays

[WHY]
Virtual displays do not use the backend of the pipe, and so have
infinite backend bandwidth.

[HOW]
Add a skip_dio_check bool to the VBA struct, which is used to override
the DIOSupport calculations.

Signed-off-by: Wesley Chalmers <Wesley.Chalmers@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Acked-by: Chris Park <Chris.Park@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Initialize dmub_rb_cmd unions to 0

[Why]
Since dmub_rb_cmds are not initialized to 0,
the header is filled with invalid data.
This is causing issues on the fw side.

[How]
Initialize dmub_rb_cmd unions to 0.

Signed-off-by: Wyatt Wood <wyatt.wood@amd.com>
Reviewed-by: Josip Pavic <Josip.Pavic@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix 64bit division issue on 32bit OS

Replace "/" with div_u64 for 64bit division on 32bit OS.

Signed-off-by: Lang Yu <Lang.Yu@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Acked-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: move edp sink present detection to hw init

[Why]
At SW init, we may not be ready to do detect eDP sink.

Signed-off-by: Eric Yang <Eric.Yang2@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: revert support for DID2.0 dsc passthrough

3x4K60 displays over MST with DSC enabled was not able to light up
due to this patch.

Signed-off-by: Jun Lei <jun.lei@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: DP HDCP Compliance 1A-08/09 tests fail

[Why]
Current implementation of mod_hdcp_hdcp2_validate_ake_cert()
does not process HDCP status message
TA_HDCP2_MSG_AUTHENTICATION_STATUS__SIGNATURE_CERTIFICAT_ERROR.
As a result, when there is a signature certificate error,
mod_hdcp_hdcp2_validate_ake_cert would return the default status,
which is success.

[How]
For all messages other than TA_HDCP2_MSG_AUTHENTICATION_STATUS__SUCCESS
and TA_HDCP2_MSG_AUTHENTICATION_STATUS__RECEIVERID_REVOKED, return status
as failure.

Signed-off-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Reviewed-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add DIG_CLOCK_PATTERN in the transmitter control

[Why and How]
VBIOS program DIG_CLK_PATTERN using engine ID instead of PHY ID.
Workaround by writing value for 0x1f (for HDMI) after calling vbios.

Signed-off-by: Derek Lai <Derek.Lai@amd.com>
Reviewed-by: Tony Cheng <Tony.Cheng@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add return code instead of boolean for future use

Signed-off-by: Fangzhi Zuo <Jerry.Zuo@amd.com>
Reviewed-by: Mikita Lipski <Mikita.Lipski@amd.com>
Acked-by: Anson Jacob <Anson.Jacob@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable gpu recovery for dimgrey_cavefish

As dimgrey_cavefish driver is stable enough, set gpu recovery as default
in HW hang for dimgrey_cavefish.

Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Jiansong Chen <Jiansong.Chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: enable LCLK DS

Enable LCLK deep sleep and it works if we enable ASPM:
modprobe amdgpu aspm=1

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: reset runpm flag if device suspend fails

If device suspend fails when we attempt to runtime suspend,
reset the runpm flag.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: use runpm flag rather than fbcon for kfd runtime suspend (v2)

the flag used by kfd is not actually related to fbcon, it just happens
to align.  Use the runpm flag instead so that we can decouple it from
the fbcon flag.

v2: fix resume as well

Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Rajneesh Bhardwaj <rajneesh.bhardwaj@amd.com>
Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: drop extra drm_kms_helper_poll_enable/disable calls

These are already called in amdgpu_device_suspend/resume which
are already called in the same functions.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/nv: add PCI reset support

Use generic PCI reset for GPU reset if the user specifies
PCI reset as the reset mechanism.  This should in general
only be used for validation.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/soc15: add PCI reset support

Use generic PCI reset for GPU reset if the user specifies
PCI reset as the reset mechanism.  This should in general
only be used for validation.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/si: add PCI reset support

Use generic PCI reset for GPU reset if the user specifies
PCI reset as the reset mechanism.  This should in general
only be used for validation.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add generic pci reset as an option

This allows us to use generic PCI reset mechanisms (FLR, SBR) as
a reset mechanism to verify that the generic PCI reset mechanisms
are working properly.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/vi: minor clean up of reset code

Drop duplicate reset method logging, whitespace changes.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/cik: minor clean up of reset code

Drop duplicate reset method logging, whitespace changes.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/si: minor clean up of reset code

Drop duplicate reset method logging, whitespace changes.

Acked-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: add support for hwmon control of slow and fast PPT limit on vangogh

Implement hwmon API for reading/setting slow and fast PPT limit.

APU power is managed to system-level requirements through the PPT
(package power tracking) feature. PPT is intended to limit power to the
requirements of the power source and could be dynamically updated to
maximize APU performance within the system power budget.

Here FAST_PPT_LIMIT manages the ~10 ms moving average of APU power,
while SLOW_PPT_LIMIT manages the configurable, thermally significant
moving average of APU power (default ~5000 ms).

User could read slow/fast ppt limit using command "cat power*_cap" or
"sensors" in the hwmon device directory. User could adjust values of
slow/fast ppt limit as needed depending on workloads through command
"echo ## > power*_cap".

Example:
$ echo 15000000 > power1_cap
$ echo 18000000 > power2_cap
$ sensors
amdgpu-pci-0300
Adapter: PCI adapter
slowPPT:     9.04W (cap = 15.00 W)
fastPPT:     9.04W (cap = 18.00 W)

v2: align with existing interfaces for the getting/setting of PPT
    limits. Encode the upper 8 bits of limit value to distinguish
    slow and fast power limit type.

Signed-off-by: Xiaomeng Hou <Xiaomeng.Hou@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: modify the power limit level parameter from bool to enum type

The original smu_get_power_limit callback accepts the power limit level
parameter as bool which limits to max and current. For possible needs to
retrieve other level like min, extend the parameter type using enum.

Signed-off-by: Xiaomeng Hou <Xiaomeng.Hou@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: update the smu v11.5 smc header for vangogh

Add PP messages for reading/setting Fast PPT and Slow PPT limit.

Signed-off-by: Xiaomeng Hou <Xiaomeng.Hou@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable wave limit on non high prio cs pipes

To achieve the best QoS for high priority compute jobs it is
required to limit waves on other compute pipes as well.
This patch will set min value in non high priority
mmSPI_WCL_PIPE_PERCENT_CS[0-3] registers to minimize the
impact of normal/low priority compute jobs over high priority
compute jobs.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add otg vertical interrupt0 support in DCN1.0

[Why & How]
On DCN1.0, need otg vertical line interrupt to get appropriate timing
to achieve specific feature request.

Add otg vertical interrupt0 support for registers which operation is
vertical sensitive.

Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add otg vertical IRQ Source

[Why & How]
In order to get appropriate timing for registers which
read/write is vertical line sensitive, add new IRQ source variable.
This interrupt is triggered by specific vertical line,

Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: optimize list operation in amdgpu_xgmi

simplify the list operation.

Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm/swsmu: unify the init soft gpu metrics function

the soft gpu metrics is not asic related data structure.
unify them to reduce duplicate code.

Signed-off-by: Kevin Wang <kevin1.wang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: support rom clockgating related function for NV family

Add functions to support enable/disable rom clock gating and get rom
clock gating status.

Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: switch to use smuio callbacks for NV family

Switch to smuio callbacks: use smuio v11_0_6 callbacks for
Sienna_cichlid and forward ASIC, use smuio v11_0 callbacks for the
other NV family ASIC.

Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: implement smuio v11_0_6 callbacks

Implement smuio v11_0_6 callbacks which will used by Sienna_Cichlid and
forward ASIC.

Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add SMUIO 11.0.6 register headers

Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: support ASPM for some specific ASIC

Support to program ASPM and LTR for Sienna Cichlid and forward ASIC.
Disable ASPM for Sienna Cichlid and forward ASIC by default.

Signed-off-by: Likun Gao <Likun.Gao@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: enable DCS

Enable DCS

V1: Enable Async DCS.
V2: Add the ppfeaturemask bit to enable from the modprobe parameter.
V3:
1. add the flag to skip APU support.
2. remove the hunk for workload selection since
it doesn't impact the function.

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/gmc9: fix mmhub client mapping for arcturus

The hw interface changed on arcturus so the old numbering
scheme doesn't work.

Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/smu12: fix power reporting on renoir

Align with Vangogh.

Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/1467
Acked-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Get unique_id dynamically v2

Instead of caching the value during amdgpu_device_init, just call the
function directly. This avoids issues where the unique_id hasn't been
saved by the time that KFD's topology snapshot is done (e.g. Arcturus).

KFD's topology information from the amdgpu_device was initially cached
at KFD initialization due to amdkfd and amdgpu being separate modules.
Now that they are combined together, we can directly call the functions
that we need and avoid this unnecessary duplication and complexity.

As a side-effect of this change, we also remove unique_id=0 for CPUs,
which is obviously not unique.

v2: Drop previous patch printing unique_id in hex

Signed-off-by: Kent Russell <kent.russell@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: enable ACDC feature

The power limit and clock ragne are different in AC mode and DC mode.
Firmware does the setting after this feature is enabled.

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable gfx wave limiting for high priority compute jobs

Enable gfx wave limiting for gfx jobs before pushing high priority
compute jobs so that high priority compute jobs gets more resources
to finish early.

v2: use ring priority instead of job priority.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add wave limit functionality for gfx8,9

Wave limiting can be use to load balance high priority
compute jobs along with gfx jobs. When enabled, this will reserve
~75% of waves for compute jobs.

We do not need this from gfx10 onwards because >=gfx10 has
asynchronous compute tunneling to replace wave limit requirement.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: enable only one high prio compute queue

For high priority compute to work properly we need to enable
wave limiting on gfx pipe. Wave limiting is done through writing
into mmSPI_WCL_PIPE_PERCENT_GFX register. Enable only one high
priority compute queue to avoid race condition between multiple
high priority compute queues writing that register simultaneously.

Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: cleanup struct amdgpu_ring

This patch consist of below related changes:

1 Rename ring->priority to ring->hw_prio.
2 Assign correct hardware ring priority.
3 Remove ring->priority_mutex as ring priority remains unchanged
  after initialization.
4 Remove unused ring->num_jobs.

v3: remove ring->num_jobs.
v2: remove ring->priority_mutex.

Fixes: 33abcb1f5a17 ("drm/amdgpu: set compute queue priority at mqd_init")
Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix unused variable warning

[Why]
The dm struct is only being used if DCN config is defined and this
causes a unused variable warning if DCN option is not set.

[How]
Remove the compile flag so the variable is used (there also seems to be
a duplicate guard due to a bad rebase) so remove the outer guard to fix
the warning.

Fixes this warning
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c: In function 'dm_set_vblank':
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c:5380:33: warning: unused variable 'dm' [-Wunused-variable]

Fixes: 98ab5f3513f9 ("drm/amd/display: Fix deadlock during gpu reset v3")
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Bhawanpreet Lakha <Bhawanpreet.Lakha@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: fix unused variable warning"

This reverts commit 4c3a3292730c56591472717d8c5c0faf74f6c6bb.

Bhawan's fix is better (less ifdefs), so revert this and apply Bhawan's.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix unused variable warning

After all users of the 'dm' warnings got hidden in an #ifdef,
the compiler started warning about it being unused:

drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:5380:33: error: unused variable 'dm' [-Werror,-Wunused-variable]

Add another such #ifdef.

Fixes: 98ab5f3513f9 ("drm/amd/display: Fix deadlock during gpu reset v3")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210125124849.102037-1-arnd@kernel.org

Merge tag 'amd-drm-next-5.12-2021-02-03' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-5.12-2021-02-03:

amdgpu:
- Display fixes and cleanups
- Vangogh fixes
- Fix possible race when there are timeouts on two rings
- SR-IOV fixes
- Add missing license
- DCE 10/12 bpc fixes
- Display MALL fixes
- Fix SMU user preference settings persistence
- Fix retry in gem allocate
- Add new PCI DID
- Fix for manual fan speed control on cards where it was problematic
- Fix regression in pinning GTT
- Misc display fixes
- Misc code cleanups

amdkfd:
- Fix config handling
- Fix regression in buffer free

From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20210204045717.3823-1-alexander.deucher@amd.com
Signed-off-by: Dave Airlie <airlied@redhat.com>

Merge tag 'mediatek-drm-next-5.12' of https://git./linux/kernel/git/chunkuang.hu/linux into drm-next

Mediatek DRM Next for Linux 5.12

1. Decouple Mediatek DRM sub driver
2. Share mtk mutex driver for both DRM and MDP
3. Add support for SoC MT8183

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Chun-Kuang Hu <chunkuang.hu@kernel.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20210204151750.7550-1-chunkuang.hu@kernel.org

drm/mediatek: Add matrix_bits private data for ccorr

Add matrix_bits and coeffs_precision to ccorr private data:
- matrix bits of mt8183 is 10
- matrix bits of mt8192 is 11

Signed-off-by: Yongqiang Niu <yongqiang.niu@mediatek.com>
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: CK Hu <ck.hu@mediatek.com>
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>

drm/mediatek: Fix ccorr size config

Fix setting to follow hardware datasheet. The original error setting
affects mt8192 display.

Signed-off-by: Yongqiang Niu <yongqiang.niu@mediatek.com>
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: CK Hu <ck.hu@mediatek.com>
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>

drm/mediatek: Separate ccorr module

ccorr ctm matrix bits will be different in mt8192.

Signed-off-by: Yongqiang Niu <yongqiang.niu@mediatek.com>
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: CK Hu <ck.hu@mediatek.com>
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>