drm/amdgpu: Increase hmm range get pages timeout

If hmm_range_fault returns -EBUSY, we should call hmm_range_fault again
to validate the remaining pages. On one system with NUMA auto balancing
enabled, hmm_range_fault takes 6 seconds for 1GB range because CPU
migrate the range one page at a time. To be safe, increase timeout value
to 1 second for 128MB range.

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

drm/amdgpu: Enable translate further for GC v9.4.3

To extend UTCL2 reach.

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

drm/amdgpu: Remove unused NBIO interface

Set compute partition mode interface in NBIO is no longer used. Remove
the only implementation from NBIO v7.9

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Le Ma <le.ma@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: bump kfd ioctl minor version for event age availability

Bump the minor version to declare event age tracking feature is now
available.

In kernel amdgpu driver, kfd_wait_on_events is used to support user
space signal event wait function. For multiple threads waiting on same
event scenery, race condition could occur since some threads after
checking signal condition, before calling kfd_wait_on_events, the
event interrupt could be fired and wake up other thread which are
sleeping on this event. Then those threads could fall into sleep
without waking up again. Adding event age tracking in both kernel and
user mode, will help avoiding this race condition.

Proposed ROCT-Thunk-Interface:
https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/commit/efdbf6cfbc026bd68ac3c35d00dacf84370eb81e
https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/commit/1820ae0a2db85b6f584611dc0cde1a00e7c22915

Proposed ROCR-Runtime:
https://github.com/RadeonOpenCompute/ROCR-Runtime/compare/master...zhums:ROCR-Runtime:new_event_wait_review
https://github.com/RadeonOpenCompute/ROCR-Runtime/commit/e1f5bdb88eb882ac798aeca2c00ea3fbb2dba459
https://github.com/RadeonOpenCompute/ROCR-Runtime/commit/7d26afd14107b5c2a754c1a3f415d89f3aabb503

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: update user space last_event_age

Update user space last_event_age when event age is enabled.
It is only for KFD_EVENT_TYPE_SIGNAL which is checked by user space.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: set activated flag true when event age unmatchs

Set waiter's activated flag true when event age unmatchs with last_event_age.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: add event_age tracking when receiving interrupt

Add event_age tracking when receiving interrupt.

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: add event age tracking

Add event age tracking

Signed-off-by: James Zhu <James.Zhu@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/scheduler: avoid infinite loop if entity's dependency is a scheduled error fence

[Why]
drm_sched_entity_add_dependency_cb ignores the scheduled fence and return false.
If entity's dependency is a scheduler error fence and drm_sched_stop is called
due to TDR, drm_sched_entity_pop_job will wait for the dependency infinitely.

[How]
Do not wait or ignore the scheduled error fence, add drm_sched_entity_wakeup
callback for the dependency with scheduled error fence.

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

drm/amdgpu: add entity error check in amdgpu_ctx_get_entity

[Why]
UMD is not aware of entity error, and will keep submitting jobs
into the error entity.

[How]
Add entity error check when getting entity from ctx.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: ZhenGuo Yin <zhenguo.yin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add VM generation token

Instead of using the VRAM lost counter add a 64bit token which indicates
if a context or job is still valid to use.

Should the VRAM be lost or the page tables need re-creation the token will
change indicating that userspace needs to act and re-create the contexts
and re-submit the work.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: reset VM when an error is detected

When some problem with the updates of page tables is detected reset the
state machine of the VM and re-create all page tables from scratch.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: abort submissions during prepare on error

Forward errors from previous submissions to this one.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: mark soft recovered fences with -ENODATA

Set the fence error code before trying to soft-recover it.

It gets overwritten when a hard recovery is required.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: mark force completed fences with -ECANCELED

When we force complete fences we should mark them as canceled.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add amdgpu_error_* debugfs file

This allows us to insert some error codes into the bottom of the pipeline
on an engine.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: mark GC 9.4.3 experimental for now

Mark as experimental for now until we get closer to production
to avoid possible undesireable behavior when mixing newer
boards with older kernels.

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

drm/amdgpu: Use PSP FW API for partition switch

Use PSP firmware interface for switching compute partitions.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Le Ma <le.ma@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Change nbio v7.9 xcp status definition

PARTITION_MODE field in PARTITION_COMPUTE_STATUS register is defined as
below by firmware.

SPX = 0, DPX = 1, TPX = 2, QPX = 3, CPX = 4

Change driver definition accordingly.

Signed-off-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Reviewed-by: Le Ma <le.ma@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: make sure that BOs have a backing store

It's perfectly possible that the BO is about to be destroyed and doesn't
have a backing store associated with it.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Guchun Chen <guchun.chen@amd.com>
Tested-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: make sure BOs are locked in amdgpu_vm_get_memory

We need to grab the lock of the BO or otherwise can run into a crash
when we try to inspect the current location.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Guchun Chen <guchun.chen@amd.com>
Tested-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add checking mc_vram_size

Do not compare injection address with mc_vram_size
if mc_vram_size is zero.

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

drm/amdgpu: Optimize checking ras supported

Using "is_app_apu" to identify device in the native
APU mode or carveout mode.

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

drm/amdgpu: Add channel_dis_num to ras init flags

Add disabled channel number to ras init flags.

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Update total channel number for umc v8_10

Update total channel number for umc v8_10.

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: Align eccinfo table structure with smu v13_0_0 interface

Update eccinfo table structure according to smu v13_0_0 interface.

v2: Calculate array size instead of using macro definition.

Signed-off-by: Candice Li <candice.li@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Convert to kdoc formats in dc/core/dc.c

Fixes the following gcc with W=1:

drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c:3483: warning: Cannot understand  * *******************************************************************************
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c:4204: warning: Cannot understand  * *******************************************************************************

Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: decrement queue count on mes queue destroy

Queue count should decrement on queue destruction regardless of HWS
support type.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: enable more Pstates profile levels for SMU v13.0.5

This patch enables following UMD stable Pstates profile
levels for power_dpm_force_performance_level interface.

- profile_peak
- profile_min_sclk
- profile_standard

Signed-off-by: Tim Huang <Tim.Huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon: Fix missing prototypes in radeon_atpx_handler.c

Fixes the following gcc with W=1:

drivers/gpu/drm/radeon/radeon_atpx_handler.c:64:6: warning: no previous prototype for ‘radeon_has_atpx’ [-Wmissing-prototypes]
   64 | bool 4(void) {
      |      ^~~~~~~~~~~~~~~
drivers/gpu/drm/radeon/radeon_atpx_handler.c:68:6: warning: no previous prototype for ‘radeon_has_atpx_dgpu_power_cntl’ [-Wmissing-prototypes]
   68 | bool radeon_has_atpx_dgpu_power_cntl(void) {
      |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/radeon/radeon_atpx_handler.c:72:6: warning: no previous prototype for ‘radeon_is_atpx_hybrid’ [-Wmissing-prototypes]
   72 | bool radeon_is_atpx_hybrid(void) {
      |      ^~~~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/radeon/radeon_atpx_handler.c:77:6: warning: no previous prototype for ‘radeon_atpx_dgpu_req_power_for_displays’ [-Wmissing-prototypes]
   77 | bool radeon_atpx_dgpu_req_power_for_displays(void) {
      |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/radeon/radeon_atpx_handler.c:596:6: warning: no previous prototype for ‘radeon_register_atpx_handler’ [-Wmissing-prototypes]
  596 | void radeon_register_atpx_handler(void)
      |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/radeon/radeon_atpx_handler.c:614:6: warning: no previous prototype for ‘radeon_unregister_atpx_handler’ [-Wmissing-prototypes]
  614 | void radeon_unregister_atpx_handler(void)
      |      ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/gpu/drm/radeon/radeon_atpx_handler.c:159: warning: expecting prototype for radeon_atpx_validate_functions(). Prototype was for radeon_atpx_validate() instead

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

drm/amdgpu: Fix usage of UMC fill record in RAS

The fixed commit listed in the Fixes tag below, introduced a bug in
amdgpu_ras.c::amdgpu_reserve_page_direct(), in that when introducing the new
amdgpu_umc_fill_error_record() and internally in that new function the physical
address (argument "uint64_t retired_page"--wrong name) is right-shifted by
AMDGPU_GPU_PAGE_SHIFT. Thus, in amdgpu_reserve_page_direct() when we pass
"address" to that new function, we should NOT right-shift it, since this
results, erroneously, in the page address to be 0 for first
2^(2*AMDGPU_GPU_PAGE_SHIFT) memory addresses.

This commit fixes this bug.

Cc: Tao Zhou <tao.zhou1@amd.com>
Cc: Hawking Zhang <Hawking.Zhang@amd.com>
Cc: Alex Deucher <Alexander.Deucher@amd.com>
Fixes: 400013b268cb ("drm/amdgpu: add umc_fill_error_record to make code more simple")
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Link: https://lore.kernel.org/r/20230610113536.10621-1-luben.tuikov@amd.com
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu/sdma4: set align mask to 255

The wptr needs to be incremented at at least 64 dword intervals,
use 256 to align with windows.  This should fix potential hangs
with unaligned updates.

Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Aaron Liu <aaron.liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Report ras_num_recs in debugfs

Report the number of records stored in the RAS EEPROM table in debugfs.

This can be used by user-space to calculate the capacity of the RAS EEPROM
table since "bad_page_cnt_threshold" is also reported in the same place in
debugfs.

See commit 7fb640714547 ("drm/amdgpu: Add bad_page_cnt_threshold to debugfs").

ras_num_recs can already be inferred by dumping the RAS EEPROM table, also in
the same debugfs location, see commit reference c65b0805e77919 (drm/amdgpu:
RAS EEPROM table is now in debugfs, 2021-04-08). This commit makes it an
integer value easily shown in a single file.

Cc: Alex Deucher <Alexander.Deucher@amd.com>
Cc: Hawking Zhang <Hawking.Zhang@amd.com>
Cc: Tao Zhou <tao.zhou1@amd.com>
Cc: Stanley Yang <Stanley.Yang@amd.com>
Cc: John Clements <john.clements@amd.com>
Signed-off-by: Luben Tuikov <luben.tuikov@amd.com>
Link: https://lore.kernel.org/r/20230603051043.211548-1-luben.tuikov@amd.com
Acked-by: Alex Deucher <Alexander.Deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: Remove DUMMY_VRAM_SIZE

Remove DUMMY_VRAM_SIZE as it is not needed and can result
in reporting incorrect memory size.

Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Release SDMAv4.4.2 ecc irq properly

Release ECC irq only if irq is enabled - only when RAS feature is enabled
ECC irq gets enabled.

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

drm/amdgpu: add wait_for helper for spirom update

Spirom update typically requires extremely long
duration for command execution, and special helper
function to wait for it completion.

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/amd/display: Clean up dcn10_optc.c kdoc

Fixes the following W=1 kernel build warning:

display/dc/dcn10/dcn10_optc.c:45: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* apply_front_porch_workaround  TODO FPGA still need?
display/dc/dcn10/dcn10_optc.c:136: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
 * program_timing_generator   used by mode timing set
display/dc/dcn10/dcn10_optc.c:391: warning: Function parameter or member 'optc' not described in 'optc1_set_timing_double_buffer'
display/dc/dcn10/dcn10_optc.c:391: warning: Function parameter or member 'enable' not described in 'optc1_set_timing_double_buffer'
display/dc/dcn10/dcn10_optc.c:404: warning: Function parameter or member 'optc' not described in 'optc1_unblank_crtc'
display/dc/dcn10/dcn10_optc.c:404: warning: expecting prototype for unblank_crtc(). Prototype was for optc1_unblank_crtc() instead
display/dc/dcn10/dcn10_optc.c:427: warning: Function parameter or member 'optc' not described in 'optc1_blank_crtc'
display/dc/dcn10/dcn10_optc.c:427: warning: expecting prototype for blank_crtc(). Prototype was for optc1_blank_crtc() instead
display/dc/dcn10/dcn10_optc.c:496: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
 * Enable CRTC
display/dc/dcn10/dcn10_optc.c:895: warning: Cannot understand  *****************************************************************************
 on line 895 - I thought it was a doc line

Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Cc: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Correct kdoc formats in dcn32_resource_helpers.c

Fixes the following gcc with W=1:

drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource_helpers.c:285: warning: Function parameter or member 'dc' not described in 'dcn32_determine_det_override'
drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource_helpers.c:285: warning: Function parameter or member 'context' not described in 'dcn32_determine_det_override'
drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource_helpers.c:285: warning: Function parameter or member 'pipes' not described in 'dcn32_determine_det_override'
drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource_helpers.c:624: warning: Cannot understand  * *****************************************************************
drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_resource_helpers.c:676: warning: Cannot understand  * *****************************************************************

Cc: Alvin Lee <alvin.lee2@amd.com>
Cc: Stylon Wang <stylon.wang@amd.com>
Cc: Hamza Mahfooz <hamza.mahfooz@amd.com>
Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Reviewed-by: Alvin Lee <alvin.lee2@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Provide function name for 'optc32_enable_crtc()'

Fixes the following W=1 kernel build warning(s):

drivers/gpu/drm/amd/amdgpu/../display/dc/dcn32/dcn32_optc.c:109: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
 * Enable CRTC

Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Cc: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Correct and remove excess function parameter names in kdoc

Fixes the following gcc with W=1:

drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/dcn32_fpu.c:872: warning: Excess function parameter 'drr_pipe' description in 'subvp_drr_schedulable'
drivers/gpu/drm/amd/amdgpu/../display/dc/dml/dcn32/dcn32_fpu.c:1030: warning: Cannot understand  * ****************************************************

Cc: Stylon Wang <stylon.wang@amd.com>
Cc: Hamza Mahfooz <hamza.mahfooz@amd.com>
Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Correct kdoc formats in dcn10_opp.c

Fixes the following W=1 kernel build warning:

display/dc/dcn10/dcn10_opp.c:52: warning: Function parameter or member 'oppn10' not described in 'opp1_set_truncation'
display/dc/dcn10/dcn10_opp.c:52: warning: Function parameter or member 'params' not described in 'opp1_set_truncation'
display/dc/dcn10/dcn10_opp.c:52: warning: expecting prototype for set_truncation(). Prototype was for opp1_set_truncation() instead
display/dc/dcn10/dcn10_opp.c:161: warning: Function parameter or member 'oppn10' not described in 'opp1_set_pixel_encoding'
display/dc/dcn10/dcn10_opp.c:161: warning: Function parameter or member 'params' not described in 'opp1_set_pixel_encoding'
display/dc/dcn10/dcn10_opp.c:161: warning: expecting prototype for set_pixel_encoding(). Prototype was for opp1_set_pixel_encoding() instead
display/dc/dcn10/dcn10_opp.c:183: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
 *      Set Clamping

Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Cc: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Add missing function parameter 'optc' & 'enable' to kdoc in optc3_set_timing_double_buffer()

Fixes the following gcc with W=1:

drivers/gpu/drm/amd/amdgpu/../display/dc/dcn30/dcn30_optc.c:285: warning: Function parameter or member 'optc' not described in 'optc3_set_timing_double_buffer'
drivers/gpu/drm/amd/amdgpu/../display/dc/dcn30/dcn30_optc.c:285: warning: Function parameter or member 'enable' not described in 'optc3_set_timing_double_buffer'

Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Print client id for the unregistered interrupt resource

Modify the debug information and print the clien id for these
interrupts as well.

Signed-off-by: Ma Jun <Jun.Ma2@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: To enable traps for GC_11_0_4 and up

Flag trap_en should be enabled for trap handler.

Signed-off-by: Ruili Ji <ruiliji2@amd.com>
Signed-off-by: Aaron Liu <aaron.liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: don't free stolen console memory during suspend

Don't free the memory if we are hitting this as part of suspend.
This way we don't free any memory during suspend; see
amdgpu_bo_free_kernel().  The memory will be freed in the first
non-suspend modeset or when the driver is torn down.

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2568
Acked-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Guchun Chen <guchun.chen@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: fix dpms_off issue when disabling bios mode"

This reverts commit 58e67bb3c131da5ee14e4842b08e53f4888dce0a.

This patch was reverted, but came back again as
commit 58e67bb3c131 ("drm/amd/display: fix dpms_off issue when disabling bios mode")
Revert it again as it breaks Asus G513QY / 6800M laptops.

Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/2259
Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Zhongwei <Zhongwei.Zhang@amd.com>
Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Mike Lothian <mike@fireburn.co.uk>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: fix null queue check on debug setting exceptions

Null check should be done on queue struct itself and not on the
process queue list node.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: enable vclk and dclk Pstates for SMU v13.0.5

Add the ability to control the vclk and dclk frequency by
power_dpm_force_performance_level interface.

Signed-off-by: Tim Huang <Tim.Huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix memcpy() in sienna_cichlid_append_powerplay_table function.

Fixes the following gcc with W=1:

In file included from ./include/linux/string.h:253,
                 from ./include/linux/bitmap.h:11,
                 from ./include/linux/cpumask.h:12,
                 from ./arch/x86/include/asm/cpumask.h:5,
                 from ./arch/x86/include/asm/msr.h:11,
                 from ./arch/x86/include/asm/processor.h:22,
                 from ./arch/x86/include/asm/cpufeature.h:5,
                 from ./arch/x86/include/asm/thread_info.h:53,
                 from ./include/linux/thread_info.h:60,
                 from ./arch/x86/include/asm/preempt.h:7,
                 from ./include/linux/preempt.h:78,
                 from ./include/linux/spinlock.h:56,
                 from ./include/linux/mmzone.h:8,
                 from ./include/linux/gfp.h:7,
                 from ./include/linux/firmware.h:7,
                 from drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu11/sienna_cichlid_ppt.c:26:
In function ‘fortify_memcpy_chk’,
    inlined from ‘sienna_cichlid_append_powerplay_table’ at drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu11/sienna_cichlid_ppt.c:444:2,
    inlined from ‘sienna_cichlid_setup_pptable’ at drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu11/sienna_cichlid_ppt.c:506:8,
    inlined from ‘sienna_cichlid_setup_pptable’ at drivers/gpu/drm/amd/amdgpu/../pm/swsmu/smu11/sienna_cichlid_ppt.c:494:12:
./include/linux/fortify-string.h:413:4: warning: call to ‘__read_overflow2_field’ declared with attribute warning: detected read beyond size of field (2nd parameter); maybe use struct_group()? [-Wattribute-warning]
  413 |    __read_overflow2_field(q_size_field, size);
      |    ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~  ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

the compiler complains about the size calculation in the memcpy() -
"sizeof(*smc_dpm_table) - sizeof(smc_dpm_table->table_header)" is much
larger than what fits into table_member.

Hence, reuse 'smu_memcpy_trailing' for nv1x

Fixes: 7077b19a38240 ("drm/amd/pm: use macro to get pptable members")
Suggested-by: Evan Quan <Evan.Quan@amd.com>
Cc: Evan Quan <Evan.Quan@amd.com>
Cc: Chengming Gui <Jack.Gui@amd.com>
Cc: Christian König <christian.koenig@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: fix vclk setting failed for SMU v13.0.5

PMFW use the left-shifted 16 bits argument to set the VCLK
DPM frequency for SMU v13.0.5.

Signed-off-by: Tim Huang <Tim.Huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: 3.2.239

This version brings along following fixes:
- Improvement on eDP
- PSR bug fixes
- SubVP bug fixes
- Improvements on pipe handling to address potential issues
- Freesync bug fix
- DPIA bug fix
- Fix multi-display issues

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

drm/amd/display: fix pixel rate update sequence

The k1/k2 pixel rate dividers in dccg should only be updated on stream enable
and do not actually depend on whether odm combine is active.

This removes an on flip update of these and fixes the calculate function
to ignore odm status for dp steams.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Dmytro Laktyushkin <dmytro.laktyushkin@amd.com>
Reviewed-by: Ariel Bernstein <Eric.Bernstein@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: reallocate DET for dual displays with high pixel rate ratio"

Revert commit 5f3401eeb064 ("drm/amd/display: reallocate DET for dual displays with high pixel rate ratio")

[Why]
Previously had modified DET allocation algorithm to allocate less DET
segments for lower pixel rate display and more DET segments for higher
pixel rate display.  But noticed it is causing underflow when higher
pixel rate display is not displaying at higher mode

[How]
Roll back change

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Samson Tam <samson.tam@amd.com>
Reviewed-by: Alvin Lee <Alvin.Lee2@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: limit DPIA link rate to HBR3

[Why]
DPIA doesn't support UHBR, driver should not enable UHBR
for dp tunneling

[How]
limit DPIA link rate to HBR3

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Peichen Huang <peichen.huang@amd.com>
Reviewed-by: Mustapha Ghaddar <Mustapha.Ghaddar@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Include CSC updates in new fast update path

[Description]
- Missed color / CSC updates in fast update path
  which caused Custom Color to break.
- Add color / CSC updates to new fast update path
  to fix custom color

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Reviewed-by: Jun Lei <jun.lei@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Limit Minimum FreeSync Refresh Rate

Why:
Some EDIDs report a minimum refresh rate lower than what HW can support

How:
Add a check to calculate minimum supported refresh rate with current timing
and use that as the minimum if a lower one is passed in

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Austin Zheng <austin.zheng@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Bug fix in dcn315_populate_dml_pipes_from_context

[Why]
When iterating over all pipes in the loop, the CRB allocation algorithm
may potentially skip some of the pipes. Previously, the current pipe
index didn't get updated in this case, causing incorrect outcome.

[How]
Increment the pipe index when skipping over a pipe in the loop.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Artem Grishin <artem.grishin@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Block SubVP + DRR if the DRR is PSR capable

[Description]
PSR implementation in FW has inline polling which can poll for up
to 1ms. This will interfere with SubVP because SubVP is timing
sensitive and can't tolerate up to 1ms worth of delay before
handling vertical or VLINE interrupts. Therefore block SubVP + DRR
cases if DRR is PSR capable

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Reviewed-by: Saaem Rizvi <SyedSaaem.Rizvi@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Do not disable phantom pipes in driver

[Why and How]
We should not disable phantom pipes in this sequence, as this should be
controlled by FW. Furthermore, the previous programming sequence would
have enabled the phantom pipe in driver as well, causing corruption.
This change should avoid this from occuring.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Saaem Rizvi <syedsaaem.rizvi@amd.com>
Reviewed-by: Alvin Lee <Alvin.Lee2@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Re-enable SubVP high refresh

Re-enable SubVP high refresh now that it is fixed
for displays with high refresh rates.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Reviewed-by: Saaem Rizvi <SyedSaaem.Rizvi@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Re-enable DPP/HUBP Power Gating

[Why & How]
Bugs preventing DPP/HUBP power gating have been addressed
so this should be reenabled on dcn314 for sufficient
hardware rev versions

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

drm/amd/display: SubVP high refresh only if all displays >= 120hz

[Description]
- SubVP high refresh should only be enabled if all displays
  are >= 120hz. We do not want to accidentally enables configs
  such as 60hz[SubVP] + 120hz[SubVP]
- Ensure that the SubVP config generation code does not produce
  configs such as 60hz[SubVP] + 120hz[SubVP]
- Also add admissibility checks to ensure these configs do not
  pass as valid configs

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Reviewed-by: Dillon Varone <Dillon.Varone@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Fix disbling PSR slow response issue

[Why]
dmub_psr_get_state() return an invalid PSR state while disable
the PSR because convert_psr_state() doesn't recognize the state
that return from DMCUB.

[How]
Add a PSR state to make the dmub_psr_get_state() return a
correct PSR state.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Tom Chung <chiahsuan.chung@amd.com>
Reviewed-by: Wayne Lin <Wayne.Lin@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: fix the system hang while disable PSR

[Why]
When the PSR enabled. If you try to adjust the timing parameters,
it may cause system hang. Because the timing mismatch with the
DMCUB settings.

[How]
Disable the PSR before adjusting timing parameters.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Tom Chung <chiahsuan.chung@amd.com>
Reviewed-by: Wayne Lin <Wayne.Lin@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Promote DAL to 3.2.238

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

drm/amd/display: Add Error Code for Dml Validation Failure

Any invalid mode from hw perspective should be given corresponding
error code, otherwise it leads to confusing warning message
"[drm] Mode Validation Warning: Validation OK failed validation."

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Fangzhi Zuo <jerry.zuo@amd.com>
Reviewed-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Add DP2 Metrics

Why:
Log DP2 data to be used for debugging purposes

How:
Check the reported link rate of the DP connection and
translate it to the DP version.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Austin Zheng <austin.zheng@amd.com>
Reviewed-by: Martin Leung <Martin.Leung@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add debugfs for allow_edp_hotplug_detection

[Why] within dc_link_detect, edp edid is read only for the first time
and saved. edid will not be read after the first time read. to run edp
edid read test, need read edp edid for each dc_link_detect. dc->config
flag allow_edp_hotplug_detection could be used for edp edid test.

[How] add debugfs for dc->config.allow_edp_hotplug_detection

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: edp do not add non-edid timings

[Why] most edp support only timings from edid. applying
non-edid timings, especially those timings out of edp
bandwidth, may damage edp.

[How] do not add non-edid timings for edp.

Cc: Mario Limonciello <mario.limonciello@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Roman Li <roman.li@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amdgpu: remove TOPDOWN flags when allocating VRAM in large bar system"

This reverts commit c105518679b6e87232874ffc989ec403bee59664.

This patch disables the TOPDOWN flag for APU and few dGPU cards
which has the VRAM size equal to the BAR size.

When we enable the TOPDOWN flag, we get the free blocks at
the highest available memory region and we don't split the
lower order blocks. This change is required to keep off
the fragmentation related issues particularly in ASIC
which has VRAM space <= 500MiB

Hence, we are reverting this patch.

Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2270
Signed-off-by: Arunpravin Paneer Selvam <Arunpravin.PaneerSelvam@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: expose num_hops and num_links xgmi info through dev attr

Add these two dev attrs for xgmi info details which is helpful for
developers checking the xgmi topology by catting the sys file directly.

Take 4 cards with xgmi connection as an example, get the num_hops for each
device or node through xmig_hive_info dir like,
cat /sys/bus/pci/devices/0000:41:00.0/xgmi_hive_info/node1/num_hops
will return "00 41 41 41" where "00" stands for the hops to node1 itself
and "41" is the hops in hex format to every other node in the same hive.
There are node1/node2/node3/node4 representing 4 cards in the hive.

The same for num_links dev attr.

Signed-off-by: Shiwu Zhang <shiwu.zhang@amd.com>
Acked-by: Le Ma <le.ma@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: vcn_4_0 set instance 0 init sched score to 1

Only vcn0 can process AV1 codecx. In order to use both vcn0 and
vcn1 in h264/265 transcode to AV1 cases, set vcn0 sched score to 1
at initialization time.

Signed-off-by: Sonny Jiang <sonjiang@amd.com>
Reviewed-by: Leo Liu <leo.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/radeon: Disable outputs when releasing fbdev client

Disable the modesetting pipeline before release the radeon's fbdev
client. Fixes the following error:

[   17.217408] WARNING: CPU: 5 PID: 1464 at drivers/gpu/drm/ttm/ttm_bo.c:326 ttm_bo_release+0x27e/0x2d0 [ttm]
[   17.217418] Modules linked in: edac_mce_amd radeon(+) drm_ttm_helper ttm video drm_suballoc_helper drm_display_helper kvm irqbypass drm_kms_helper syscopyarea crc32_pclmul sysfillrect sha512_ssse3 sysimgblt sha512_generic cfbfillrect cfbimgblt wmi_bmof aesni_intel cfbcopyarea crypto_simd cryptd k10temp acpi_cpufreq wmi dm_mod
[   17.217432] CPU: 5 PID: 1464 Comm: systemd-udevd Not tainted 6.4.0-rc4+ #1
[   17.217436] Hardware name: Micro-Star International Co., Ltd. MS-7A38/B450M PRO-VDH MAX (MS-7A38), BIOS B.G0 07/26/2022
[   17.217438] RIP: 0010:ttm_bo_release+0x27e/0x2d0 [ttm]
[   17.217444] Code: 48 89 43 38 48 89 43 40 48 8b 5c 24 30 48 8b b5 40 08 00 00 48 8b 6c 24 38 48 83 c4 58 e9 7a 49 f7 e0 48 89 ef e9 6c fe ff ff <0f> 0b 48 83 7b 20 00 0f 84 b7 fd ff ff 0f 0b 0f 1f 00 e9 ad fd ff
[   17.217448] RSP: 0018:ffffc9000095fbb0 EFLAGS: 00010202
[   17.217451] RAX: 0000000000000001 RBX: ffff8881052c8de0 RCX: 0000000000000000
[   17.217453] RDX: 0000000000000001 RSI: 0000000000000000 RDI: ffff8881052c8de0
[   17.217455] RBP: ffff888104a66e00 R08: ffff8881052c8de0 R09: ffff888104a7cf08
[   17.217457] R10: ffffc9000095fbe0 R11: ffffc9000095fbe8 R12: ffff8881052c8c78
[   17.217458] R13: ffff8881052c8c78 R14: dead000000000100 R15: ffff88810528b108
[   17.217460] FS:  00007f319fcbb8c0(0000) GS:ffff88881a540000(0000) knlGS:0000000000000000
[   17.217463] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   17.217464] CR2: 000055dc8b0224a0 CR3: 000000010373d000 CR4: 0000000000750ee0
[   17.217466] PKRU: 55555554
[   17.217468] Call Trace:
[   17.217470]  <TASK>
[   17.217472]  ? __warn+0x97/0x160
[   17.217476]  ? ttm_bo_release+0x27e/0x2d0 [ttm]
[   17.217481]  ? report_bug+0x1ec/0x200
[   17.217487]  ? handle_bug+0x3c/0x70
[   17.217490]  ? exc_invalid_op+0x1f/0x90
[   17.217493]  ? preempt_count_sub+0xb5/0x100
[   17.217496]  ? asm_exc_invalid_op+0x16/0x20
[   17.217500]  ? ttm_bo_release+0x27e/0x2d0 [ttm]
[   17.217505]  ? ttm_resource_move_to_lru_tail+0x1ab/0x1d0 [ttm]
[   17.217511]  radeon_bo_unref+0x1a/0x30 [radeon]
[   17.217547]  radeon_gem_object_free+0x20/0x30 [radeon]
[   17.217579]  radeon_fbdev_fb_destroy+0x57/0x90 [radeon]
[   17.217616]  unregister_framebuffer+0x72/0x110
[   17.217620]  drm_client_dev_unregister+0x6d/0xe0
[   17.217623]  drm_dev_unregister+0x2e/0x90
[   17.217626]  drm_put_dev+0x26/0x90
[   17.217628]  pci_device_remove+0x44/0xc0
[   17.217631]  really_probe+0x257/0x340
[   17.217635]  __driver_probe_device+0x73/0x120
[   17.217638]  driver_probe_device+0x2c/0xb0
[   17.217641]  __driver_attach+0xa0/0x150
[   17.217643]  ? __pfx___driver_attach+0x10/0x10
[   17.217646]  bus_for_each_dev+0x67/0xa0
[   17.217649]  bus_add_driver+0x10e/0x210
[   17.217651]  driver_register+0x5c/0x120
[   17.217653]  ? __pfx_radeon_module_init+0x10/0x10 [radeon]
[   17.217681]  do_one_initcall+0x44/0x220
[   17.217684]  ? kmalloc_trace+0x37/0xc0
[   17.217688]  do_init_module+0x64/0x240
[   17.217691]  __do_sys_finit_module+0xb2/0x100
[   17.217694]  do_syscall_64+0x3b/0x90
[   17.217697]  entry_SYSCALL_64_after_hwframe+0x72/0xdc
[   17.217700] RIP: 0033:0x7f319feaa5a9
[   17.217702] Code: 08 89 e8 5b 5d c3 66 2e 0f 1f 84 00 00 00 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 27 08 0d 00 f7 d8 64 89 01 48
[   17.217706] RSP: 002b:00007ffc6bf3e7f8 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
[   17.217709] RAX: ffffffffffffffda RBX: 00005607204f3170 RCX: 00007f319feaa5a9
[   17.217710] RDX: 0000000000000000 RSI: 00007f31a002eefd RDI: 0000000000000018
[   17.217712] RBP: 00007f31a002eefd R08: 0000000000000000 R09: 00005607204f1860
[   17.217714] R10: 0000000000000018 R11: 0000000000000246 R12: 0000000000020000
[   17.217716] R13: 0000000000000000 R14: 0000560720522450 R15: 0000560720255899
[   17.217718]  </TASK>
[   17.217719] ---[ end trace 0000000000000000 ]---

The buffer object backing the fbdev emulation got pinned twice: by the
fb_probe helper radeon_fbdev_create_pinned_object() and the modesetting
code when the framebuffer got displayed. It only got unpinned once by
the fbdev helper radeon_fbdev_destroy_pinned_object(). Hence TTM's BO-
release function complains about the pin counter. Forcing the outputs
off also undoes the modesettings pin increment.

Tested-by: Borislav Petkov (AMD) <bp@alien8.de>
Reported-by: Borislav Petkov <bp@alien8.de>
Closes: https://lore.kernel.org/dri-devel/20230603174814.GCZHt83pN+wNjf63sC@fat_crate.local/
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Fixes: e317a69fe891 ("drm/radeon: Implement client-based fbdev emulation")
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Thomas Zimmermann <tzimmermann@suse.de>
Cc: "Christian König" <christian.koenig@amd.com>
Cc: "Pan, Xinhui" <Xinhui.Pan@amd.com>
Cc: amd-gfx@lists.freedesktop.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/amdgpu: enable W=1 for amdgpu

We have a clean build with W=1 as of
commit c168feed5d03 ("drm/amd/display/amdgpu_dm/amdgpu_dm_helpers: Move
SYNAPTICS_DEVICE_ID into CONFIG_DRM_AMD_DC_DCN ifdef"). So, let's enable
these checks unconditionally for the entire module to catch these errors
during development.

Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: Nathan Chancellor <nathan@kernel.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Fix kdoc warning

Fixes the following gcc with W=1:

drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c:76: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
 * EEPROM Table structure v1
drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c:98: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
 * EEPROM Table structrue v2.1

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

drm/amd/display: Enable dcn314 DPP RCO

[Why and How]
Add back debug bits enabling RCO for dcn314 as underflow
associated with this change has been resolved

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Daniel Miess <daniel.miess@amd.com>
Reviewed-by: Jun Lei <jun.lei@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Rename DRM schedulers in amdgpu TTM

Rename mman.entity to mman.high_pr to make the distinction
clearer that this is a high priority scheduler. Similarly,
rename the recently added mman.delayed to mman.low_pr to
make it clear it is a low priority scheduler.
No functional change in this patch.

Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display/amdgpu_dm/amdgpu_dm_helpers: Move SYNAPTICS_DEVICE_ID into CONFIG_DRM_AMD_DC_DCN ifdef

Fixes the following W=1 kernel build warning(s):

 drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm_helpers.c:48:22: warning: ‘SYNAPTICS_DEVICE_ID’ defined but not used [-Wunused-const-variable=]

Cc: Harry Wentland <harry.wentland@amd.com>
Cc: Leo Li <sunpeng.li@amd.com>
Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: "Christian König" <christian.koenig@amd.com>
Cc: "Pan, Xinhui" <Xinhui.Pan@amd.com>
Cc: David Airlie <airlied@gmail.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: amd-gfx@lists.freedesktop.org
Cc: dri-devel@lists.freedesktop.org
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Lee Jones <lee@kernel.org>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: Skip DPP DTO update if root clock is gated

[Why]
Hardware implements root clock gating by utilizing the DPP DTO registers
with a special case of DTO enabled, phase = 0, modulo = 1. This
conflicts with our policy to always update the DPPDTO for cases where
it's expected to be disabled.

The pipes unexpectedly enter a higher power state than expected because
of this programming flow.

[How]
Guard the upper layers of HWSS against this hardware quirk with
programming the register with an internal state flag in DCCG.

While technically acting as global state for the DCCG, HWSS shouldn't be
expected to understand the hardware quirk for having DTO disabled
causing more power than DTO enabled with this specific setting.

This also prevents sequencing errors from occuring in the future if
we have to program DPP DTO in multiple locations.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com>
Reviewed-by: Jun Lei <jun.lei@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Merge tag 'amd-drm-next-6.5-2023-06-09' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.5-2023-06-02:

amdgpu:
- SR-IOV fixes
- Warning fixes
- Misc code cleanups and spelling fixes
- DCN 3.2 updates
- Improved DC FAMS support for better power management
- Improved DC SubVP support for better power management
- DCN 3.1.x fixes
- Max IB size query
- DC GPU reset fixes
- RAS updates
- DCN 3.0.x fixes
- S/G display fixes
- CP shadow buffer support
- Implement connector force callback
- Z8 power improvements
- PSP 13.0.10 vbflash support
- Mode2 reset fixes
- Store MQDs in VRAM to improve queue switch latency
- VCN 3.x fixes
- JPEG 3.x fixes
- Enable DC_FP on LoongArch
- GFXOFF fixes
- GC 9.4.3 partition support
- SDMA 4.4.2 partition support
- VCN/JPEG 4.0.3 partition support
- VCN 4.0.3 updates
- NBIO 7.9 updates
- GC 9.4.3 updates
- Take NUMA into account when allocating memory
- Handle NUMA for partitions
- SMU 13.0.6 updates
- GC 9.4.3 RAS updates
- Stop including unused swiotlb.h
- SMU 13.0.7 fixes
- Fix clock output ordering on some APUs
- Clean up DC FPGA code
- GFX9 preemption fixes
- Misc irq fixes
- S0ix fixes
- Add new DRM_AMDGPU_WERROR config parameter to help with CI
- PCIe fix for RDNA2
- kdoc fixes
- Documentation updates

amdkfd:
- Query TTM mem limit rather than hardcoding it
- GC 9.4.3 partition support
- Handle NUMA for partitions

radeon:
- Fix possible double free
- Stop including unused swiotlb.h
- Fix possible division by zero

ttm:
- Add query for TTM mem limit
- Add NUMA awareness to pools
- Export ttm_pool_fini()

UAPI:
- Add new ctx query flag to better handle GPU resets
  Mesa MR: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22290
- Add new interface to query and set shadow buffer for RDNA3
  Mesa MR: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21986
- Add new INFO query for max IB size
  Proposed userspace: https://gitlab.freedesktop.org/bnieuwenhuizen/mesa/-/commits/ib-rejection-v3

amd-drm-next-6.5-2023-06-09:

amdgpu:
- S0ix fixes
- Initial SMU13 Overdrive support
- kdoc fixes
- Misc clode cleanups
- Flexible array fixes
- Display OTG fixes
- SMU 13.0.6 updates
- Revert some broken clock counter updates
- Misc display fixes
- GFX9 preemption fixes
- Add support for newer EEPROM bad page table format
- Add missing radeon secondary id
- Add support for new colorspace KMS API
- CSA fix
- Stable pstate fixes for APUs
- make vbl interface admin only
- Handle PCI accelerator class

amdkfd:
- Add debugger support for gdb

radeon:
- Fix possible UAF

drm:
- Add Colorspace functionality

UAPI:
- Add debugger interface for enabling gdb
  Proposed userspace: https://github.com/ROCm-Developer-Tools/ROCdbgapi/tree/wip-dbgapi
- Add KMS colorspace API
  Discussion: https://lists.freedesktop.org/archives/dri-devel/2023-June/408128.html

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230609174817.7764-1-alexander.deucher@amd.com

drm/amd/display: Filter out AC mode frequencies on DC mode systems

Why:
Limit maximum clock speeds to DC mode limits for DC mode systems
How:
Store DC mode limits when individual clocks are initialized and
cap the values when building the clock table

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Austin Zheng <austin.zheng@amd.com>
Reviewed-by: Alvin Lee <Alvin.Lee2@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: DSC Programming Deltas

[Why]
Programming register delta for DSC sub-block

[How]
Change DSC, resource files for programming register delta.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Sridevi <sarvinde@amd.com>
Reviewed-by: Chris Park <Chris.Park@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "drm/amd/display: cache trace buffer size"

Revert commit 257e9891db0b ("drm/amd/display: cache trace buffer size")
to fix regression found in tests.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Leo Ma <hanghong.ma@amd.com>
Reviewed-by: Josip Pavic <Josip.Pavic@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: workaround for compute workload type on some skus

On smu 13.0.0, the compute workload type cannot be set on all the skus
due to some other problems. This workaround is to make sure compute workload type
can also run on some specific skus.

v2: keep the variable consistent

Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Acked-by: Lijo Lazar <lijo.lazar@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/display: add NULL pointer check

[why]
check dmub_Srv exist or not before accessing dmub.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Charlene Liu <charlene.liu@amd.com>
Reviewed-by: Zhan Liu <Zhan.Liu@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: enable more Pstates profile levels for yellow_carp

This patch enables following UMD stable Pstates profile levels for
power_dpm_force_performance_level interface.

- profile_peak
- profile_min_mclk
- profile_min_sclk
- profile_standard

Signed-off-by: shikaguo <shikai.guo@amd.com>
Reviewed-by: Tim Huang <Tim.Huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add option params to enforce process isolation between graphics and compute

enforce process isolation between graphics and compute via using the same reserved vmid.

v2: remove params "struct amdgpu_vm *vm" from
    amdgpu_vmid_alloc_reserved and amdgpu_vmid_free_reserved.

Signed-off-by: Chong Li <chongli2@amd.com>
Reviewed-by: Christian Koenig <Christian.Koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: Wrap -Wunused-but-set-variable in cc-option

-Wunused-but-set-variable was only supported in clang starting with
13.0.0, so earlier versions will emit a warning, which is turned into a
hard error for the kernel to mirror GCC:

  error: unknown warning option '-Wunused-but-set-variable'; did you mean '-Wunused-const-variable'? [-Werror,-Wunknown-warning-option]

The minimum supported version of clang for building the kernel is
11.0.0, so match the rest of the kernel and wrap
-Wunused-but-set-variable in a cc-option call, so that it is only used
when supported by the compiler.

Closes: https://github.com/ClangBuiltLinux/linux/issues/1869
Fixes: 1b320ad3f5a8 ("drm/amd/amdgpu: introduce DRM_AMDGPU_WERROR")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: add the accelerator PCIe class

Add the accelerator PCIe class and match the
class in amdgpu for 0x1002 devices of that class.

From PCI spec:
"PCI Code and ID Assignment, r1.9, sec 1, 1.19"

Signed-off-by: Shiwu Zhang <shiwu.zhang@amd.com>
Acked-by: Lijo Lazar <lijo.lazar@amd.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com> # pci_ids.h
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Revert "Revert drm/amd/display: Enable Freesync Video Mode by default"

This reverts commit 4243c84aa082d8fba70c45f48eb2bb5c19799060.

Enables freesync video by default, since the hang and corruption issue
on eDP panels are now fixed.

Acked-by: Stylon Wang <stylon.wang@amd.com>
Signed-off-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Reviewed-by: Rodrigo Siqueira <rjordrigo@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdkfd: fix and enable debugging for gfx11

There are a couple of fixes required to enable gfx11 debugging.

First, ADD_QUEUE.trap_en is an inappropriate place to toggle
a per-process register so move it to SET_SHADER_DEBUGGER.trap_en.
When ADD_QUEUE.skip_process_ctx_clear is set, MES will prioritize
the SET_SHADER_DEBUGGER.trap_en setting.

Second, to preserve correct save/restore priviledged wave states
in coordination with the trap enablement setting, resume suspended
waves early in the disable call.

Signed-off-by: Jonathan Kim <jonathan.kim@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: Tighten permissions on VBIOS flashing attributes

Non-root users shouldn't be able to try to trigger a VBIOS flash
or query the flashing status.  This should be reserved for users with the
appropriate permissions.

Cc: stable@vger.kernel.org
Fixes: 8424f2ccb3c0 ("drm/amdgpu/psp: Add vbflash sysfs interface support")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd: Make sure image is written to trigger VBIOS image update flow

The VBIOS image update flow requires userspace to:
1) Write the image to `psp_vbflash`
2) Read `psp_vbflash`
3) Poll `psp_vbflash_status` to check for completion

If userspace reads `psp_vbflash` before writing an image, it's
possible that it causes problems that can put the dGPU into an invalid
state.

Explicitly check that an image has been written before letting a read
succeed.

Cc: stable@vger.kernel.org
Fixes: 8424f2ccb3c0 ("drm/amdgpu/psp: Add vbflash sysfs interface support")
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: enable more Pstates profile levels for SMU v13.0.4

This patch enables following UMD stable Pstates profile
levels for power_dpm_force_performance_level interface.

- profile_peak
- profile_min_mclk
- profile_min_sclk
- profile_standard

Signed-off-by: Tim Huang <Tim.Huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: enable vclk and dclk Pstates for SMU v13.0.4

Add the ability to control the vclk and dclk frequency by
power_dpm_force_performance_level interface.

Signed-off-by: Tim Huang <Tim.Huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amd/pm: fix vclk setting failed for SMU v13.0.4

PMFW use the left-shifted 16 bits argument to set the VCLK
DPM frequency for SMU v13.0.4.

Signed-off-by: Tim Huang <Tim.Huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Yifan Zhang <yifan1.zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: skip to resume rlcg for gc 9.4.3 in vf side

skip to resume rlcg, because rlcg is already enabled in pf side.

Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Lijo Lazar <lijo.lazar@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: disable virtual display support on APP device

virtual display is not support on APP device.

Signed-off-by: Yang Wang <kevinyang.wang@amd.com>
Signed-off-by: Gavin Wan <gavin.wan@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

drm/amdgpu: unmap and remove csa_va properly

Root PD BO should be reserved before unmap and remove
a bo_va from VM otherwise lockdep will complain.

v2: check fpriv->csa_va is not NULL instead of amdgpu_mcbp (christian)

[14616.936827] WARNING: CPU: 6 PID: 1711 at drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c:1762 amdgpu_vm_bo_del+0x399/0x3f0 [amdgpu]
[14616.937096] Call Trace:
[14616.937097]  <TASK>
[14616.937102]  amdgpu_driver_postclose_kms+0x249/0x2f0 [amdgpu]
[14616.937187]  drm_file_free+0x1d6/0x300 [drm]
[14616.937207]  drm_close_helper.isra.0+0x62/0x70 [drm]
[14616.937220]  drm_release+0x5e/0x100 [drm]
[14616.937234]  __fput+0x9f/0x280
[14616.937239]  ____fput+0xe/0x20
[14616.937241]  task_work_run+0x61/0x90
[14616.937246]  exit_to_user_mode_prepare+0x215/0x220
[14616.937251]  syscall_exit_to_user_mode+0x2a/0x60
[14616.937254]  do_syscall_64+0x48/0x90
[14616.937257]  entry_SYSCALL_64_after_hwframe+0x63/0xcd

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

drm/amd/display: Refactor avi_info_frame colorimetry determination

Replace the messy two if-else chains here that were
on the same value with a switch on the enum.

Signed-off-by: Joshua Ashton <joshua@froggi.es>
Signed-off-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Cc: Pekka Paalanen <ppaalanen@gmail.com>
Cc: Sebastian Wick <sebastian.wick@redhat.com>
Cc: Vitaly.Prosyak@amd.com
Cc: Joshua Ashton <joshua@froggi.es>
Cc: Simon Ser <contact@emersion.fr>
Cc: Melissa Wen <mwen@igalia.com>
Cc: dri-devel@lists.freedesktop.org
Cc: amd-gfx@lists.freedesktop.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>