Adding NOP packets one by one in the ring
does not use the CP efficiently.
Solution:
Use CP optimization while adding NOP packet's so PFP
can discard NOP packets based on information of count
from the Header instead of fetching all NOP packets
one by one.
Reviewed-by: Christian König <christian.koenig@amd.com> Cc: Christian König <christian.koenig@amd.com> Cc: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com> Cc: Tvrtko Ursulin <tursulin@igalia.com> Cc: Marek Olšák <marek.olsak@amd.com> Signed-off-by: Sunil Khatri <sunil.khatri@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Add NULL check for function pointer in dcn32_set_output_transfer_func
This commit adds a null check for the set_output_gamma function pointer
in the dcn32_set_output_transfer_func function. Previously,
set_output_gamma was being checked for null, but then it was being
dereferenced without any null check. This could lead to a null pointer
dereference if set_output_gamma is null.
To fix this, we now ensure that set_output_gamma is not null before
dereferencing it. We do this by adding a null check for set_output_gamma
before the call to set_output_gamma.
Cc: Tom Chung <chiahsuan.chung@amd.com> Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Cc: Roman Li <roman.li@amd.com> Cc: Alex Hung <alex.hung@amd.com> Cc: Aurabindo Pillai <aurabindo.pillai@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Add NULL check for function pointer in dcn401_set_output_transfer_func
This commit adds a null check for the set_output_gamma function pointer
in the dcn401_set_output_transfer_func function. Previously,
set_output_gamma was being checked for null, but then it was being
dereferenced without any null check. This could lead to a null pointer
dereference if set_output_gamma is null.
To fix this, we now ensure that set_output_gamma is not null before
dereferencing it. We do this by adding a null check for set_output_gamma
before the call to set_output_gamma.
Cc: Tom Chung <chiahsuan.chung@amd.com> Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Cc: Roman Li <roman.li@amd.com> Cc: Alex Hung <alex.hung@amd.com> Cc: Aurabindo Pillai <aurabindo.pillai@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Add NULL check for function pointer in dcn20_set_output_transfer_func
This commit adds a null check for the set_output_gamma function pointer
in the dcn20_set_output_transfer_func function. Previously,
set_output_gamma was being checked for null at line 1030, but then it
was being dereferenced without any null check at line 1048. This could
potentially lead to a null pointer dereference error if set_output_gamma
is null.
To fix this, we now ensure that set_output_gamma is not null before
dereferencing it. We do this by adding a null check for set_output_gamma
before the call to set_output_gamma at line 1048.
Cc: Tom Chung <chiahsuan.chung@amd.com> Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Cc: Roman Li <roman.li@amd.com> Cc: Alex Hung <alex.hung@amd.com> Cc: Aurabindo Pillai <aurabindo.pillai@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Adding NOP packets one by one in the ring
does not use the CP efficiently.
Solution:
Use CP optimization while adding NOP packet's so PFP
can discard NOP packets based on information of count
from the Header instead of fetching all NOP packets
one by one.
Reviewed-by: Christian König <christian.koenig@amd.com> Cc: Christian König <christian.koenig@amd.com> Cc: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com> Cc: Tvrtko Ursulin <tursulin@igalia.com> Cc: Marek Olšák <marek.olsak@amd.com> Signed-off-by: Sunil Khatri <sunil.khatri@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jonathan Kim [Mon, 22 Jul 2024 17:26:08 +0000 (13:26 -0400)]
drm/amdkfd: fix debug watchpoints for logical devices
The number of watchpoints should be set and constrained per logical
partition device, not by the socket device.
Signed-off-by: Jonathan Kim <jonathan.kim@amd.com> Reviewed-by: Harish Kasiviswanathan <harish.kasiviswanathan@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: do not call insert_nop fn for zero count
Do not make a function call for zero size NOP as it
does not add anything in the ring and is unnecessary
function call.
Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Sunil Khatri <sunil.khatri@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jonathan Kim [Tue, 25 Jun 2024 15:22:50 +0000 (11:22 -0400)]
drm/amdkfd: support per-queue reset on gfx9
Support per-queue reset for GFX9. The recommendation is for the driver
to target reset the HW queue via a SPI MMIO register write.
Since this requires pipe and HW queue info and MEC FW is limited to
doorbell reports of hung queues after an unmap failure, scan the HW
queue slots defined by SET_RESOURCES first to identify the user queue
candidates to reset.
Only signal reset events to processes that have had a queue reset.
If queue reset fails, fall back to GPU reset.
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>
Adding NOP packets one by one in the ring
does not use the CP efficiently.
Solution:
Use CP optimization while adding NOP packet's so PFP
can discard NOP packets based on information of count
from the Header instead of fetching all NOP packets
one by one.
Cc: Christian König <christian.koenig@amd.com> Cc: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com> Cc: Tvrtko Ursulin <tursulin@igalia.com> Cc: Marek Olšák <marek.olsak@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Sunil Khatri <sunil.khatri@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Align 'dpp401_dscl_program_isharp' with actual function parameters
This commit corrects the function comment for
'dpp401_dscl_program_isharp' in 'dcn401_dpp_dscl.c'. The comment
previously included a description for a non-existent parameter
'bs_coeffs_updated'. This parameter description has been removed to
reflect the function's actual parameters.
Fixes the below with gcc W=1:
drivers/gpu/drm/amd/amdgpu/../display/dc/dpp/dcn401/dcn401_dpp_dscl.c:981: warning: Excess function parameter 'bs_coeffs_updated' description in 'dpp401_dscl_program_isharp'
Cc: Tom Chung <chiahsuan.chung@amd.com> Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Cc: Roman Li <roman.li@amd.com> Cc: Alex Hung <alex.hung@amd.com> Cc: Aurabindo Pillai <aurabindo.pillai@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: Clean up the register dump via debugfs list
debugfs register list for dump is cleaned as it have
some issues related to proper power state of the IP
before register read.
Since the above mentioned is removed we no longer want
this to be dumped part of the devcoredump and hence
removed.
Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Sunil Khatri <sunil.khatri@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
There are some problem with existing amdgpu_reset_dump_register_list
debugfs node. It is supposed to read a list of registers but there
could be cases when the IP is not in correct power state. Register
read in such cases could lead to more problems.
We are taking care of all such power states in devcoredump and
dumping the registers of need for debugging. So cleaning this code
and we dont need this functionality via debugfs anymore.
Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Sunil Khatri <sunil.khatri@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/pm: add overdrive support on smu v14.0.2/3
add overdrive support on smu v14.0.2/3
Signed-off-by: Kenneth Feng <kenneth.feng@amd.com> Reviewed-by: Yang Wang <kevinyang.wang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Handle null 'stream_status' in 'planes_changed_for_existing_stream'
This commit adds a null check for 'stream_status' in the function
'planes_changed_for_existing_stream'. Previously, the code assumed
'stream_status' could be null, but did not handle the case where it was
actually null. This could lead to a null pointer dereference.
Reported by smatch:
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc_resource.c:3784 planes_changed_for_existing_stream() error: we previously assumed 'stream_status' could be null (see line 3774)
Cc: Tom Chung <chiahsuan.chung@amd.com> Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Cc: Roman Li <roman.li@amd.com> Cc: Alex Hung <alex.hung@amd.com> Cc: Aurabindo Pillai <aurabindo.pillai@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/pm: update powerplay structure on smu v14.0.2/3
update powerplay structure on smu v14.0.2/3
Signed-off-by: Kenneth Feng <kenneth.feng@amd.com> Reviewed-by: Yang Wang <kevinyang.wang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Thu, 25 Jul 2024 23:10:43 +0000 (19:10 -0400)]
drm/amdkfd: Fix compile error if HMM support not enabled
Fixes the below if kernel config not enable HMM support
>> drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_queue.c:107:26: error:
implicit declaration of function 'svm_range_from_addr'
>> drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_queue.c:107:24: error:
assignment to 'struct svm_range *' from 'int' makes pointer from integer
without a cast [-Wint-conversion]
>> drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_queue.c:111:28: error:
invalid use of undefined type 'struct svm_range'
Fixes: b049504e211e ("drm/amdkfd: Validate user queue svm memory residency") Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202407252127.zvnxaKRA-lkp@intel.com/ 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/amd/display: Add missing DET segments programming
The commit 5034b935f62a ("drm/amd/display: Modify DHCUB waterwark
structures and functions") introduced a code refactor for DCHUB, but
during the merge process into amd-staging-drm-next, the program det
segments were removed. This commit adds the DET segment programming for
DCN35.
Fixes: 5034b935f62a ("drm/amd/display: Modify DHCUB waterwark structures and functions") Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alvin Lee [Fri, 26 Apr 2024 19:24:06 +0000 (15:24 -0400)]
drm/amd/display: For FAMS2 don't program P-State force from driver
P-State force programming is handled entirely by FW in FAMS2. Remove
any programming from driver side to prevent incorrect programming from
driver side (which may override FW programming)
Signed-off-by: Alvin Lee <Alvin.Lee2@amd.com> Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Rodrigo Siqueira [Mon, 4 Dec 2023 21:35:04 +0000 (16:35 -0500)]
drm/amd/display: Replace dm_execute_dmub_cmd with dc_wake_and_execute_dmub_cmd
In the commit c2cec7a872b6 ("drm/amd/display: Wake DMCUB before sending
a command for replay feature"), replaced dm_execute_dmub_cmd with
dc_wake_and_execute_dmub_cmd in multiple areas, but due to merge issues
the replacement of this function in the dmub_replay_copy_settings was
missed. This commit replaces the old dm_execute_dmub_cmd with
dc_wake_and_execute_dmub_cmd.
Fixes: 3601a35a2e9d ("drm/amd/display: Wake DMCUB before sending a command for replay feature") Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
SPL has a control field for controlling the two pixels per container
that is not in use yet. This commit adds a proper initialization for
this feature.
Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Hung [Tue, 16 Jul 2024 20:05:12 +0000 (14:05 -0600)]
drm/amd/display: Check UnboundedRequestEnabled's value
CalculateSwathAndDETConfiguration_params_st's UnboundedRequestEnabled is
a pointer (i.e. dml_bool_t *UnboundedRequestEnabled), and thus
p->UnboundedRequestEnabled checks its address, not bool value.
To check value, *p->UnboundedRequestEnabled is used instead.
This fixes 1 REVERSE_INULL issue reported by Coverity.
Signed-off-by: Alex Hung <alex.hung@amd.com> Reviewed-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Daniel Sa [Fri, 19 Jul 2024 17:39:09 +0000 (13:39 -0400)]
drm/amd/display: Underflow Seen on DCN401 eGPU
[WHY]
In dcn401 we read clock values before FW is loaded. These incorrect
values cause the driver to believe that we are running higher clocks
than what we actually have. This then causes corruption/underflow for
the eGPU.
[HOW]
When new values are read from HW, update internal structures to
propagate the new/correct value. Fixes issue
Signed-off-by: Daniel Sa <Daniel.Sa@amd.com> Reviewed-by: Alvin Lee <alvin.lee2@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Enable aux transfer path via dmub for dp tunneling
[Why]
Aux transfer retries path does not support dp tunneling.
[How]
Based on ddc pin check, aux will be issued in legacy path or DMUB.
Signed-off-by: Meenakshikumar Somasundaram <meenakshikumar.somasundaram@amd.com> Reviewed-by: Eric Yang <eric.yang@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Add new enable and disable functions for DCN35
Add new enable and disable functions based on DCCG spec.
Signed-off-by: Hansen Dsouza <Hansen.Dsouza@amd.com> Reviewed-by: Muhammad Ahmed <ahmed.ahmed@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Nicholas Kazlauskas [Thu, 18 Jul 2024 15:53:31 +0000 (11:53 -0400)]
drm/amd/display: Use gpuvm_min_page_size_kbytes for DML2 surfaces
[Why]
It's currently hard coded to 256 when it should be using the SOC
provided values. This can result in corruption with linear surfaces
where we prefetch more PTE than the buffer can hold.
[How]
Update the min page size correctly for the plane.
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: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Hung [Mon, 15 Jul 2024 15:57:01 +0000 (09:57 -0600)]
drm/amd/display: Initialize get_bytes_per_element's default to 1
Variables, used as denominators and maybe not assigned to other values,
should not be 0. bytes_per_element_y & bytes_per_element_c are
initialized by get_bytes_per_element() which should never return 0.
This fixes 10 DIVIDE_BY_ZERO issues reported by Coverity.
Signed-off-by: Alex Hung <alex.hung@amd.com> Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Hung [Tue, 16 Jul 2024 20:58:06 +0000 (14:58 -0600)]
drm/amd/display: Check null values from functions
Functions get_per_method_common_meta and get_expanded_strategy_list can
return null and thus it is necessary to check their returned values
before dereferencing.
This fixes 3 NULL_RETURNS issues reported by Coverity.
Signed-off-by: Alex Hung <alex.hung@amd.com> Reviewed-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The DP tunnel AUX reply is received through Outbox1. Print the Outbox1
status if an issue occurs.
Signed-off-by: Cruise <Cruise.Hung@amd.com> Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Reviewed-by: Wayne Lin <wayne.lin@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Add new enable and disable functions
Add new enable and disable functions based on DCCG spec.
Signed-off-by: Hansen Dsouza <Hansen.Dsouza@amd.com> Reviewed-by: Muhammad Ahmed <ahmed.ahmed@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Thu, 25 Jul 2024 21:30:37 +0000 (17:30 -0400)]
drm/amdgpu: Fix APU handling in amdgpu_pm_load_smu_firmware()
We only need to skip this on modern APUs. It's required
on older APUs as it's where start_smu gets called from.
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3502 Fixes: 064d92436b69 ("drm/amd/pm: avoid to load smu firmware for APUs") Reviewed-by: Tim Huang <tim.huang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Cc: Tim Huang <Tim.Huang@amd.com>
drm/amdgpu: trigger ip dump before suspend of IP's
Problem:
IP dump right now is done post suspend of all
IP's which for some IP's could change power
state and software state too which we do not want
to reflect in the dump as it might not be same at
the time of hang.
Solution:
IP should be dumped as close to the HW state when
the GPU was in hung state without trying to reinitialize
any resource.
Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sunil Khatri <sunil.khatri@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Hung [Wed, 17 Jul 2024 15:17:56 +0000 (09:17 -0600)]
drm/amd/display: Avoid overflow assignment in link_dp_cts
sampling_rate is an uint8_t but is assigned an unsigned int, and thus it
can overflow. As a result, sampling_rate is changed to uint32_t.
Similarly, LINK_QUAL_PATTERN_SET has a size of 2 bits, and it should
only be assigned to a value less or equal than 4.
This fixes 2 INTEGER_OVERFLOW issues reported by Coverity.
Signed-off-by: Alex Hung <alex.hung@amd.com> Reviewed-by: Wenjing Liu <wenjing.liu@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Need to use cm3_helper function with DCN401 to avoid cases where high
RGB component values can get set to zero if using the TF curve generated
by cm_helper.
Signed-off-by: Ilya Bakoulin <Ilya.Bakoulin@amd.com> Reviewed-by: Alvin Lee <alvin.lee2@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Nicholas Kazlauskas [Tue, 16 Jul 2024 21:41:54 +0000 (17:41 -0400)]
drm/amd/display: Add seamless boot support for more DIG operation modes
[Why]
When pre-OS firmware enables display support for displays that operate
the DIG in 2 pixels per cycle processing modes the inferred pixel rate
from get_pixel_clk_frequency_100hz does not account for the true pixel
rate since we're outputting 2 per cycle past the stream encoder.
This causes seamless boot validation to abort early.
[How]
Add a new stream encoder function for getting pixels per cycle from the
stream encoder. If the pixels per cycle is greater than 1 and the driver
policy is to enable 2 pixels per cycle for post-OS then allow seamless
boot to continue.
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Reviewed-by: Duncan Ma <duncan.ma@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom Chung [Fri, 12 Jul 2024 09:29:07 +0000 (17:29 +0800)]
drm/amd/display: Reset VRR config during resume
[Why]
After resume the system, the new_crtc_state->vrr_infopacket does not
synchronize with the current state. It will affect the
update_freesync_state_on_stream() does not update the state correctly.
The previous patch causes a PSR SU regression that cannot let panel go
into self-refresh mode.
[How]
Reset the VRR config during resume to force update the VRR config later.
Fixes: eb6dfbb7a9c6 ("drm/amd/display: Reset freesync config before update new state") Signed-off-by: Tom Chung <chiahsuan.chung@amd.com> Reviewed-by: Sun peng Li <sunpeng.li@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tom Chung [Fri, 12 Jul 2024 10:02:30 +0000 (18:02 +0800)]
drm/amd/display: Add a missing PSR state
[Why & How]
Add a missing PSR state to make the dmub_psr_get_state() return a
correct PSR state.
Signed-off-by: Tom Chung <chiahsuan.chung@amd.com> Reviewed-by: Sun peng Li <sunpeng.li@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Charlene Liu [Tue, 16 Jul 2024 19:58:35 +0000 (15:58 -0400)]
drm/amd/display: sync dmub output event type.
[why]
dmubfw added a new event type, update amdgpu to avoid "notify type 6
invalid"
Signed-off-by: Charlene Liu <Charlene.Liu@amd.com> Reviewed-by: Chris Park <chris.park@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Charlene Liu [Tue, 16 Jul 2024 17:47:43 +0000 (13:47 -0400)]
drm/amd/display: restore immediate_disable_crtc for w/a
[why]
immediate_disable_crtc does not reset ODM. if switching to disable_crtc
which will disable ODM as well. i.e. need to restore ODM mem cfg at
reenable it at end of w/a.
Signed-off-by: Charlene Liu <Charlene.Liu@amd.com> Reviewed-by: Xi (Alex) Liu <xi.liu@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Michael Chen [Tue, 23 Jul 2024 21:45:23 +0000 (17:45 -0400)]
drm/amdgpu: increase mes log buffer size for gfx12
MES firmware requires larger log buffer for gfx12. Allocate
proper buffer respectively for gfx11 and gfx12.
Signed-off-by: Michael Chen <michael.chen@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Hung [Sat, 8 Jun 2024 04:09:53 +0000 (22:09 -0600)]
drm/amd/display: Fix possible overflow in integer multiplication
[WHAT & HOW]
Integer multiplies integer may overflow in context that expects an
expression of unsigned long long (64 bits). This can be fixed by casting
integer to unsigned long long to force 64 bits results.
This fixes 2 OVERFLOW_BEFORE_WIDEN issues reported by Coverity.
Signed-off-by: Alex Hung <alex.hung@amd.com> Reviewed-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Nicholas Kazlauskas [Mon, 15 Jul 2024 19:52:46 +0000 (15:52 -0400)]
drm/amd/display: Request 0MHz dispclk for zero display case
[Why]
If we aren't entering RCG/IPS2 or CLKSTOP is not supported by PMFW then
we should be requesting a dispclk value of 0MHz to PMFW.
Currenly we run at max clock since there's an assumption in APU clock
table formulation where we can run at any DISPCLK at any state so the
real clock value ends up as 1200Mhz - the maximum.
[How]
Set to 0 instead of the minimum value in the state array.
Signed-off-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Reviewed-by: Duncan Ma <duncan.ma@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Chris Park [Fri, 12 Jul 2024 16:50:48 +0000 (12:50 -0400)]
drm/amd/display: Add two dmmuy I2C entry for GPIO port mapping issue
[Why]
When only 4 I2C is declared, two dummies are required to correctly map
GPIO port.
[How]
Add one more I2C dummy entry to match GPIO port.
Signed-off-by: Chris Park <chris.park@amd.com> Reviewed-by: Alvin Lee <alvin.lee2@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Leo Li [Thu, 11 Jul 2024 18:38:11 +0000 (14:38 -0400)]
drm/amd/display: Run idle optimizations at end of vblank handler
[Why & How]
1. After allowing idle optimizations, hw programming is disallowed.
2. Before hw programming, we need to disallow idle optimizations.
Otherwise, in scenario 1, we will immediately kick hw out of idle
optimizations with register access.
Scenario 2 is less of a concern, since any register access will kick
hw out of idle optimizations. But we'll do it early for correctness.
Signed-off-by: Leo Li <sunpeng.li@amd.com> Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Leo Li [Thu, 11 Jul 2024 18:31:27 +0000 (14:31 -0400)]
drm/amd/display: Let drm_crtc_vblank_on/off manage interrupts
[Why]
We manage interrupts for CRTCs in two places:
1. In manage_dm_interrupts(), when CRTC get enabled or disabled
2. When drm_vblank_get/put() starts or kills the vblank counter, calling
into amdgpu_dm_crtc_set_vblank()
The interrupts managed by these twp places should be identical.
[How]
Since manage_dm_interrupts() already use drm_crtc_vblank_on/off(), just
move all CRTC interrupt management into amdgpu_dm_crtc_set_vblank().
This has the added benefit of disabling all CRTC and HUBP interrupts
when there are no vblank requestors.
Note that there is a TODO item - unchanged from when it was first
introduced - to properly identify the HUBP instance from the OTG
instance, rather than just assume direct mapping.
Signed-off-by: Leo Li <sunpeng.li@amd.com> Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Samson Tam [Sun, 14 Jul 2024 20:31:05 +0000 (16:31 -0400)]
drm/amd/display: roll back quality EASF and ISHARP and dc dependency changes
[Why]
Seeing several regressions related to quality EASF and ISHARP changes
and removing dc dependency changes.
[How]
Roll back SPL changes
Signed-off-by: Samson Tam <Samson.Tam@amd.com> Reviewed-by: Martin Leung <martin.leung@amd.com> Tested-by: Daniel Wheeler <daniel.wheeler@amd.com> Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdkfd: Fix missing error code in kfd_queue_acquire_buffers
The fix involves setting 'err' to '-EINVAL' before each 'goto
out_err_unreserve'.
Fixes the below:
drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_queue.c:265 kfd_queue_acquire_buffers()
warn: missing error code 'err'
drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_queue.c
226 int kfd_queue_acquire_buffers(struct kfd_process_device *pdd, struct queue_properties *properties)
227 {
228 struct kfd_topology_device *topo_dev;
229 struct amdgpu_vm *vm;
230 u32 total_cwsr_size;
231 int err;
232
233 topo_dev = kfd_topology_device_by_id(pdd->dev->id);
234 if (!topo_dev)
235 return -EINVAL;
236
237 vm = drm_priv_to_vm(pdd->drm_priv);
238 err = amdgpu_bo_reserve(vm->root.bo, false);
239 if (err)
240 return err;
241
242 err = kfd_queue_buffer_get(vm, properties->write_ptr, &properties->wptr_bo, PAGE_SIZE);
243 if (err)
244 goto out_err_unreserve;
245
246 err = kfd_queue_buffer_get(vm, properties->read_ptr, &properties->rptr_bo, PAGE_SIZE);
247 if (err)
248 goto out_err_unreserve;
249
250 err = kfd_queue_buffer_get(vm, (void *)properties->queue_address,
251 &properties->ring_bo, properties->queue_size);
252 if (err)
253 goto out_err_unreserve;
254
255 /* only compute queue requires EOP buffer and CWSR area */
256 if (properties->type != KFD_QUEUE_TYPE_COMPUTE)
257 goto out_unreserve;
This is clearly a success path.
258
259 /* EOP buffer is not required for all ASICs */
260 if (properties->eop_ring_buffer_address) {
261 if (properties->eop_ring_buffer_size != topo_dev->node_props.eop_buffer_size) {
262 pr_debug("queue eop bo size 0x%lx not equal to node eop buf size 0x%x\n",
263 properties->eop_buf_bo->tbo.base.size,
264 topo_dev->node_props.eop_buffer_size);
--> 265 goto out_err_unreserve;
Fixes: 629568d25fea ("drm/amdkfd: Validate queue cwsr area and eop buffer size") Reported-by: Dan Carpenter <dan.carpenter@linaro.org> Cc: Philip Yang <Philip.Yang@amd.com> Cc: Felix Kuehling <Felix.Kuehling@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: Philip Yang <Philip.Yang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Add null check for top_pipe_to_program in commit_planes_for_stream
This commit addresses a null pointer dereference issue in the
`commit_planes_for_stream` function at line 4140. The issue could occur
when `top_pipe_to_program` is null.
The fix adds a check to ensure `top_pipe_to_program` is not null before
accessing its stream_res. This prevents a null pointer dereference.
Reported by smatch:
drivers/gpu/drm/amd/amdgpu/../display/dc/core/dc.c:4140 commit_planes_for_stream() error: we previously assumed 'top_pipe_to_program' could be null (see line 3906)
Cc: Tom Chung <chiahsuan.chung@amd.com> Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Cc: Roman Li <roman.li@amd.com> Cc: Alex Hung <alex.hung@amd.com> Cc: Aurabindo Pillai <aurabindo.pillai@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Add null check for pipe_ctx->plane_state in dcn20_program_pipe
This commit addresses a null pointer dereference issue in the
`dcn20_program_pipe` function. The issue could occur when
`pipe_ctx->plane_state` is null.
The fix adds a check to ensure `pipe_ctx->plane_state` is not null
before accessing. This prevents a null pointer dereference.
Reported by smatch:
drivers/gpu/drm/amd/amdgpu/../display/dc/hwss/dcn20/dcn20_hwseq.c:1925 dcn20_program_pipe() error: we previously assumed 'pipe_ctx->plane_state' could be null (see line 1877)
Cc: Tom Chung <chiahsuan.chung@amd.com> Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Cc: Roman Li <roman.li@amd.com> Cc: Alex Hung <alex.hung@amd.com> Cc: Aurabindo Pillai <aurabindo.pillai@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
ISP I2C bus device can't be enumerated via ACPI mechanism
since it shares the memory map with the AMDGPU.
So use the MFD mechanism for registering the ISP I2C device
and add the required resources.
Signed-off-by: Venkata Narendra Kumar Gutta <vengutta@amd.com> Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Christian König [Wed, 24 Jul 2024 07:24:02 +0000 (09:24 +0200)]
drm/amdgpu: fix contiguous handling for IB parsing v2
Otherwise we won't get correct access to the IB.
v2: keep setting AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS to avoid problems in
the VRAM backend.
Signed-off-by: Christian König <christian.koenig@amd.com> Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3501 Fixes: e362b7c8f8c7 ("drm/amdgpu: Modify the contiguous flags behaviour") Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Cc: stable@vger.kernel.org Tested-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: add macro to calculate offset with instance
Add macro definition which calculate offset of the
register with index override.
This is useful in case when there is an array of
registers which is common for all instances.
To read registers in that case it is easy to define
registers once and the index value is manually passed
to calculate proper offset of register for each instance.
Signed-off-by: Sunil Khatri <sunil.khatri@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Jonathan Kim [Tue, 21 May 2024 17:22:15 +0000 (13:22 -0400)]
drm/amdkfd: allow users to target recommended SDMA engines
Certain GPUs have better copy performance over xGMI on specific
SDMA engines depending on the source and destination GPU.
Allow users to create SDMA queues on these recommended engines.
Close to 2x overall performance has been observed with this
optimization.
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/amdgpu/pm: support gpu_metrics sysfs interface for smu v14.0.2/3
support gpu_metrics sysfs interface for smu v14.0.2/3
Signed-off-by: Kenneth Feng <kenneth.feng@amd.com> Reviewed-by: Yang Wang <kevinyang.wang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Commit 2563391e57b5 ("drm/amd/display: DML2.1 resynchronization") blew
away the compiler warning fix from commit 2fde4fdddc1f
("drm/amd/display: Avoid -Wenum-float-conversion in
add_margin_and_round_to_dfs_grainularity()"), causing the warning to
reappear.
drivers/gpu/drm/amd/amdgpu/../display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.c:183:58: error: arithmetic between enumeration type 'enum dentist_divider_range' and floating-point type 'double' [-Werror,-Wenum-float-conversion]
183 | divider = (unsigned int)(DFS_DIVIDER_RANGE_SCALE_FACTOR * (vco_freq_khz / clock_khz));
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~
1 error generated.
The comment in the vbios structure says:
// = 128 means EDID length is 128 bytes, otherwise the EDID length = ucFakeEDIDLength*128
This fake edid struct has not been used in a long time, so I'm
not sure if there were actually any boards out there with a non-128 byte
EDID, but align the code with the comment.
Reviewed-by: Thomas Weißschuh <linux@weissschuh.net> Reported-by: Thomas Weißschuh <linux@weissschuh.net> Link: https://lists.freedesktop.org/archives/amd-gfx/2024-June/109964.html Fixes: c324acd5032f ("drm/radeon/kms: parse the extended LCD info block") Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The comment in the vbios structure says:
// = 128 means EDID length is 128 bytes, otherwise the EDID length = ucFakeEDIDLength*128
This fake edid struct has not been used in a long time, so I'm
not sure if there were actually any boards out there with a non-128 byte
EDID, but align the code with the comment.
Reviewed-by: Thomas Weißschuh <linux@weissschuh.net> Reported-by: Thomas Weißschuh <linux@weissschuh.net> Link: https://lists.freedesktop.org/archives/amd-gfx/2024-June/109964.html Fixes: d38ceaf99ed0 ("drm/amdgpu: add core driver (v4)") Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Wed, 26 Jun 2024 19:03:05 +0000 (15:03 -0400)]
drm/amdkfd: Validate queue cwsr area and eop buffer size
When creating KFD user compute queue, check if queue eop buffer size,
cwsr area size, ctl stack size equal to the size of KFD node
properities.
Check the entire cwsr area which may split into multiple svm ranges
aligned to granularity boundary.
Signed-off-by: Philip Yang <Philip.Yang@amd.com> Reviewed-by: Felix Kuehling <felix.kuehling@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Wed, 26 Jun 2024 18:52:28 +0000 (14:52 -0400)]
drm/amdkfd: Store queue cwsr area size to node properties
Use the queue eop buffer size, cwsr area size, ctl stack size
calculation from Thunk, store the value to KFD node properties.
Those will be used to validate queue eop buffer size, cwsr area size,
ctl stack size when creating KFD user compute queue.
Those will be exposed to user space via sysfs KFD node properties, to
remove the duplicate calculation code from Thunk.
Signed-off-by: Philip Yang <Philip.Yang@amd.com> Reviewed-by: Felix Kuehling <felix.kuehling@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: reset vm state machine after gpu reset(vram lost)
[Why]
Page table of compute VM in the VRAM will lost after gpu reset.
VRAM won't be restored since compute VM has no shadows.
[How]
Use higher 32-bit of vm->generation to record a vram_lost_counter.
Reset the VM state machine when vm->genertaion is not equal to
the new generation token.
v2: Check vm->generation instead of calling drm_sched_entity_error
in amdgpu_vm_validate.
v3: Use new generation token instead of vram_lost_counter for check.
Signed-off-by: ZhenGuo Yin <zhenguo.yin@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amd/display: Add null check for set_output_gamma in dcn30_set_output_transfer_func
This commit adds a null check for the set_output_gamma function pointer
in the dcn30_set_output_transfer_func function. Previously,
set_output_gamma was being checked for nullity at line 386, but then it
was being dereferenced without any nullity check at line 401. This
could potentially lead to a null pointer dereference error if
set_output_gamma is indeed null.
To fix this, we now ensure that set_output_gamma is not null before
dereferencing it. We do this by adding a nullity check for
set_output_gamma before the call to set_output_gamma at line 401. If
set_output_gamma is null, we log an error message and do not call the
function.
This fix prevents a potential null pointer dereference error.
drivers/gpu/drm/amd/amdgpu/../display/dc/hwss/dcn30/dcn30_hwseq.c:401 dcn30_set_output_transfer_func()
error: we previously assumed 'mpc->funcs->set_output_gamma' could be null (see line 386)
drivers/gpu/drm/amd/amdgpu/../display/dc/hwss/dcn30/dcn30_hwseq.c
373 bool dcn30_set_output_transfer_func(struct dc *dc,
374 struct pipe_ctx *pipe_ctx,
375 const struct dc_stream_state *stream)
376 {
377 int mpcc_id = pipe_ctx->plane_res.hubp->inst;
378 struct mpc *mpc = pipe_ctx->stream_res.opp->ctx->dc->res_pool->mpc;
379 const struct pwl_params *params = NULL;
380 bool ret = false;
381
382 /* program OGAM or 3DLUT only for the top pipe*/
383 if (pipe_ctx->top_pipe == NULL) {
384 /*program rmu shaper and 3dlut in MPC*/
385 ret = dcn30_set_mpc_shaper_3dlut(pipe_ctx, stream);
386 if (ret == false && mpc->funcs->set_output_gamma) {
^^^^^^^^^^^^^^^^^^^^^^^^^^^^ If this is NULL
387 if (stream->out_transfer_func.type == TF_TYPE_HWPWL)
388 params = &stream->out_transfer_func.pwl;
389 else if (pipe_ctx->stream->out_transfer_func.type ==
390 TF_TYPE_DISTRIBUTED_POINTS &&
391 cm3_helper_translate_curve_to_hw_format(
392 &stream->out_transfer_func,
393 &mpc->blender_params, false))
394 params = &mpc->blender_params;
395 /* there are no ROM LUTs in OUTGAM */
396 if (stream->out_transfer_func.type == TF_TYPE_PREDEFINED)
397 BREAK_TO_DEBUGGER();
398 }
399 }
400
--> 401 mpc->funcs->set_output_gamma(mpc, mpcc_id, params);
^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Then it will crash
402 return ret;
403 }
Fixes: d99f13878d6f ("drm/amd/display: Add DCN3 HWSEQ") Reported-by: Dan Carpenter <dan.carpenter@linaro.org> Cc: Tom Chung <chiahsuan.chung@amd.com> Cc: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Cc: Roman Li <roman.li@amd.com> Cc: Hersen Wu <hersenxs.wu@amd.com> Cc: Alex Hung <alex.hung@amd.com> Cc: Aurabindo Pillai <aurabindo.pillai@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Hamza Mahfooz <hamza.mahfooz@amd.com> Signed-off-by: Srinivasan Shanmugam <srinivasan.shanmugam@amd.com> Reviewed-by: Tom Chung <chiahsuan.chung@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Thu, 20 Jun 2024 17:00:48 +0000 (13:00 -0400)]
drm/amdkfd: Validate user queue update
Ensure update queue new ring buffer is mapped on GPU with correct size.
Decrease queue old ring_bo queue_refcount and increase new ring_bo
queue_refcount.
Signed-off-by: Philip Yang <Philip.Yang@amd.com> Reviewed-by: Felix Kuehling <felix.kuehling@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Philip Yang [Thu, 20 Jun 2024 16:44:57 +0000 (12:44 -0400)]
drm/amdkfd: Validate user queue svm memory residency
Queue CWSR area maybe registered to GPU as svm memory, create queue to
ensure svm mapped to GPU with KFD_IOCTL_SVM_FLAG_GPU_ALWAYS_MAPPED flag.
Add queue_refcount to struct svm_range, to track queue CWSR area usage.
Because unmap mmu notifier callback return value is ignored, if
application unmap the CWSR area while queue is active, pr_warn message
in dmesg log. To be safe, evict user queue.
Signed-off-by: Philip Yang <Philip.Yang@amd.com> Reviewed-by: Felix Kuehling <felix.kuehling@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Fri, 12 Jul 2024 22:57:14 +0000 (18:57 -0400)]
drm/amdgpu/gfx9.4.3: Enable bad opcode interrupt
For the bad opcode case, it will cause CP/ME hang.
The firmware will prevent the ME side from hanging by raising a bad opcode interrupt.
And the driver needs to perform a vmid reset when receiving the interrupt.
Acked-by: Felix Kuehling <felix.kuehling@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Alex Deucher [Fri, 12 Jul 2024 22:50:26 +0000 (18:50 -0400)]
drm/amdgpu/gfx9: Enable bad opcode interrupt
For the bad opcode case, it will cause CP/ME hang.
The firmware will prevent the ME side from hanging by raising a bad opcode interrupt.
And the driver needs to perform a vmid reset when receiving the interrupt.
Acked-by: Felix Kuehling <felix.kuehling@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For the bad opcode case, it will cause CP/ME hang.
The firmware will prevent the ME side from hanging by raising a bad opcode interrupt.
And the driver needs to perform a vmid reset when receiving the interrupt.
v2: update irq naming (drop priv) (Alex)
Acked-by: Felix Kuehling <felix.kuehling@amd.com> Signed-off-by: Jesse Zhang <jesse.zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For the bad opcode case, it will cause CP/ME hang.
The firmware will prevent the ME side from hanging by raising a bad opcode interrupt.
And the driver needs to perform a vmid reset when receiving the interrupt.
v2: update irq naming (drop priv) (Alex)
Acked-by: Felix Kuehling <felix.kuehling@amd.com> Signed-off-by: Jesse Zhang <jesse.zhang@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For the bad opcode case, it will cause CP/ME hang.
The firmware will prevent the ME side from hanging by raising a bad opcode interrupt.
And the driver needs to perform a vmid reset when receiving the interrupt.
v2: update irq naming (drop priv) (Alex)
Acked-by: Felix Kuehling <felix.kuehling@amd.com> Signed-off-by: Jesse Zhang <Jesse.Zhang@amd.com> Reviewed-by: Prike Liang <Prike.Liang@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
projects / users / hch / configfs.git / log