[why]
It's important to make sure SMU messages
are logged by default to improve debugging for
power optimization use cases.
[how]
Change logs to warnings when SMU message
returns non-success id.
Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Acked-by: Tom Chung <chiahsuan.chung@amd.com>
Signed-off-by: Sung Joon Kim <sungkim@amd.com>
Tested-by: Daniel Wheeler <daniel.wheeler@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The amdgpu_ras_get_context may return NULL if device
not support ras feature, so add check before using.
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>
With S/G display becomes stable, enable S/G display for recent APUs
by default rather than white list.
v2: explicitly disable sg on pre-CZ chips (Alex)
v3: add parens for every clause (Alex)
Co-authored-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Yifan Zhang <yifan1.zhang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add support to read Manufacturer Name and FRU File Id fields. Also add
sysfs device attributes for external usage.
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>
[what]
does_configuration_meet_sw_policies check was not done in the
validate_only portion of dml2, so some unsupported modes were passing bw
validation, only to fail the same check later in validate_and_build. now
we add the check to validate_only.
Also add line in dcn35_resource to ensure that value set for
enable_windowed_mpo_odm gets passed to dml.
[why]
Immediate black screen during video playback at 4k144hz. The debugger
showed that we were failing validation in dml on every updateplanes().
Reviewed-by: Wenjing Liu <wenjing.liu@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Gabe Teeger <gabe.teeger@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
There are certain cases during a transition to ODM that might cause
corruption on the display. This occurs when we choose certain pipes in a
particular state.
[How]
We now will store the pipe indexes of the any pipes that might be
problematic to switch to during an ODM transition, and only use them as
a last resort.
Reviewed-by: Dmytro Laktyushkin <dmytro.laktyushkin@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Saaem Rizvi <syedsaaem.rizvi@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
dc->dml2_options.use_native_pstate_optimization flag will make driver
use dcn32 legacy_svp_drr related tuning. Set this to false fixed the
stutter underflow issue also based on HW suggest disable ODM by default
and let DML choose it.
Reviewed-by: Zhan Liu <zhan.liu@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Charlene Liu <charlene.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[why]
Surface height/width for Chroma has another variable that it should be
intialized to, chroma_size. Fixing this will help pass DML2.0 validation
for YCbCr420 tests, DCHB006.109,129, DCHB014.011,012.
[how]
Assign SurfaceHeight/WidthC to chroma_size.height/width
Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Sung Joon Kim <sungkim@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[why]
Regression from DML1.0 where we use differen DET buffer sizes for each
pipe. From the spec, we need to use DET buffer size of 384 kb for each
pipe
[how]
Ensure to use 384 kb DET buffer sizes for each available pipe.
Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Sung Joon Kim <sungkim@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[why]
There are cases where more than 1 stream can be mapped to the same
surface. DML2.0 does not seem to handle these cases.
[how]
Make sure to account for the stream id when deriving the plane id. By
doing this, each plane id will be unique based on the stream id.
Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Acked-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
Signed-off-by: Sung Joon Kim <sungkim@amd.com>
Signed-off-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
DC is transitioning from DML to DML2, and this commit introduces all the
required changes for some of the already available ASICs and adds the
required code infra to support new ASICs under DML2. DML2 is also a
generated code that provides better mode verification and programming
models for software/hardware, and it enables a better way to create
validation tools. This version is more like a middle step to the
complete transition to the DML2 version.
Changes since V1:
- Alex: Fix typos
Changes since V2:
- Update DC includes
Changes since V3:
- Fix 32 bit compilation issues on x86
Changes since V4:
- Avoid compilation of DML2 on some not supported 32-bit architecture
- Update commit message
Co-developed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Co-developed-by: Roman Li <roman.li@amd.com>
Signed-off-by: Roman Li <roman.li@amd.com>
Signed-off-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
For some reason, the dml code is not guarded under CONFIG_DRM_AMD_DC_FP
in the Makefile. This commit moves the dml code under the DC_FP guard.
Reviewed-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
bw_fixed does not need any FPU operation, and it is used on DCE and DCN.
For this reason, this commit moves bw_fixed to the basic folder outside
DML.
Reviewed-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The custom_float file does not have any FPU operation, so it should be
inside DML. This commit moves the file to the basic folder.
Reviewed-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
dce_calcs does not have FPU operations, and it is required for DCE and
DCN. Remove this file from the DML folder and add it to the basic folder
visible for DCE and DCN.
Reviewed-by: Qingqing Zhuo <Qingqing.Zhuo@amd.com>
Signed-off-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When ROCm is active enable additional SMU 13.0.0 optimizations.
This reuses the unused powersave profile on PMFW.
v2: move to the swsmu code since we need both bits active in
the workload mask.
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
dcn20_validate_bandwidth_fp() is invoked while FPU access has been
enabled. FPU access requires disabling preemption even on PREEMPT_RT.
It is not possible to allocate memory with disabled preemption even with
GFP_ATOMIC on PREEMPT_RT.
Move the memory allocation before FPU access is enabled.
To preserve previous "clean" state of "pipes" add a memset() before the
second invocation of dcn20_validate_bandwidth_internal() where the
variable is used.
Acked-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add a warning if the FPU is used from any context other than task
context. This is only precaution since the code is not able to be used
from softirq while the API allows it on x86 for instance.
Acked-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
fix wrong ip count INFO on spatial partitions. update the query
to return the instance count corresponding to the partition id.
v2:
initialize variables only when required to be (Christian)
move variable declarations to the beginning of function (Christian)
Signed-off-by: Sathishkumar S <sathishkumar.sundararaju@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch implements partial migration in gpu page fault according to migration
granularity(default 2MB) and not split svm range in cpu page fault handling.
A svm range may include pages from both system ram and vram of one gpu now.
These chagnes are expected to improve migration performance and reduce mmu
callback and TLB flush workloads.
Signed-off-by: Xiaogang Chen <xiaogang.chen@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The fpu_recursion_depth counter is used to ensure that dc_fpu_begin()
can be invoked multiple times while the FPU-disable function itself is
only invoked once. Also the counter part (dc_fpu_end()) is ballanced
properly.
Instead of using the get_cpu_ptr() dance around the inc it is simpler to
increment the per-CPU variable directly. Also the per-CPU variable has
to be incremented and decremented on the same CPU. This is ensured by
the inner-part which disables preemption. This is kind of not obvious,
works and the preempt-counter is touched a few times for no reason.
Disable preemption before incrementing fpu_recursion_depth for the first
time. Keep preemption disabled until dc_fpu_end() where the counter is
decremented making it obvious that the preemption has to stay disabled
while the counter is non-zero.
Use simple inc/dec functions.
Remove the nested preempt_disable/enable functions which are now not
needed.
Acked-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This is a revert of the commit mentioned below while it is not wrong, as
in the kernel will explode, having migrate_disable() here it is
complete waste of resources.
Additionally commit message is plain wrong the review tag does not make
it any better. The migrate_disable() interface has a fat comment
describing it and it includes the word "undesired" in the headline which
should tickle people to read it before using it.
Initially I assumed it is worded too harsh but now I beg to differ.
The reviewer of the original commit, even not understanding what
migrate_disable() does should ask the following:
- migrate_disable() is added only to the CONFIG_X86 block and it claims
to protect fpu_recursion_depth. Why are the other the architectures
excluded?
- migrate_disable() is added after fpu_recursion_depth was modified.
Shouldn't it be added before the modification or referencing takes
place?
Moving on.
Disabling preemption DOES prevent CPU migration. A task, that can not be
pushed away from the CPU by the scheduler (due to disabled preemption)
can not be pushed or migrated to another CPU.
Disabling migration DOES NOT ensure consistency of per-CPU variables. It
only ensures that the task acts always on the same per-CPU variable. The
task remains preemptible meaning multiple tasks can access the same
per-CPU variable. This in turn leads to inconsistency for the statement
*pcpu -= 1;
with two tasks on one CPU and a preemption point during the RMW
operation:
Task A Task B
read pcpu to reg # 0
inc reg # 0 -> 1
read pcpu to reg # 0
inc reg # 0 -> 1
write reg to pcpu # 1
write reg to pcpu # 1
At the end pcpu reads 1 but should read 2 instead. Boom.
get_cpu_ptr() already contains a preempt_disable() statement. That means
that the per-CPU variable can only be referenced by a single task which
is currently running. The only inconsistency that can occur if the
variable is additionally accessed from an interrupt.
Remove migrate_disable/enable() from dc_fpu_begin/end().
Cc: Tianci Yin <tianci.yin@amd.com>
Cc: Aurabindo Pillai <aurabindo.pillai@amd.com>
Fixes: 0c316556d1 ("drm/amd/display: Disable migration to ensure consistency of per-CPU variable")
Acked-by: Harry Wentland <harry.wentland@amd.com>
Reviewed-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Hamza Mahfooz <hamza.mahfooz@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Don't update the fault cache if status is 0. In the multiple
fault case, subsequent faults will return a 0 status which is
useless for userspace and replaces the useful fault status, so
only update if status is non-0.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
[Why]
Unplug mst hub will cause warning. That's because
dm_helpers_construct_old_payload() is changed to be called after
payload removement from dc link.
In dm_helpers_construct_old_payload(), We refer to the vcpi in
payload allocation table of dc link to construct the old payload
and payload is no longer in the table when we call the function
now.
[How]
Refer to the mst_state to construct the number of time slot for old
payload now. Note that dm_helpers_construct_old_payload() is just
a quick workaround before and we are going to abandon it soon.
Fixes: 5aa1dfcdf0 ("drm/mst: Refactor the flow for payload allocation/removement")
Reviewed-by: Jerry Zuo <jerry.zuo@amd.com>
Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231005080405.169841-1-Wayne.Lin@amd.com
On some systems with Navi3x dGPU will attempt to use BACO for runtime
PM but fails to resume properly. This is because on these systems
the root port goes into D3cold which is incompatible with BACO.
This happens because in this case dGPU is connected to a bridge between
root port which causes BOCO detection logic to fail. Fix the intent of
the logic by looking at root port, not the immediate upstream bridge for
_PR3.
Cc: stable@vger.kernel.org
Suggested-by: Jun Ma <Jun.Ma2@amd.com>
Tested-by: David Perry <David.Perry@amd.com>
Fixes: b10c1c5b3a ("drm/amdgpu: add check for ACPI power resources")
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
