kernel/linux.git/drivers/gpu/drm/amd/amdgpu/amdgpu_pmu.c, branch v6.12.80

drm/amdgpu: Use local64_try_cmpxchg in amdgpu_perf_read

2023-08-09T22:08:51+00:00

Use local64_try_cmpxchg instead of local64_cmpxchg (*ptr, old, new) == old in amdgpu_perf_read. x86 CMPXCHG instruction returns success in ZF flag, so this change saves a compare after cmpxchg (and related move instruction in front of cmpxchg). Also, try_cmpxchg implicitly assigns old *ptr value to "old" when cmpxchg fails. There is no need to re-read the value in the loop. No functional change intended. Cc: Alex Deucher Cc: "Christian König" Cc: "Pan, Xinhui" Cc: David Airlie Cc: Daniel Vetter Signed-off-by: Uros Bizjak Signed-off-by: Alex Deucher

drm/amdgpu: check df_funcs and its callback pointers

2021-12-13T21:33:16+00:00

in case they are not avaiable in early phase Signed-off-by: Hawking Zhang Reviewed-by: Le Ma Signed-off-by: Alex Deucher

drm/amdgpu: fix an error code in init_pmu_entry_by_type_and_add()

2021-04-15T20:32:45+00:00

If the kmemdup() fails then this should return a negative error code but it currently returns success Fixes: b4a7db71ea06 ("drm/amdgpu: add per device user friendly xgmi events for vega20") Signed-off-by: Dan Carpenter Signed-off-by: Alex Deucher

drm/amdgpu: fix semicolon.cocci warnings

2020-10-14T19:13:48+00:00

drivers/gpu/drm/amd/amdgpu/amdgpu_pmu.c:608:2-3: Unneeded semicolon Remove unneeded semicolon. Generated by: scripts/coccinelle/misc/semicolon.cocci Fixes: b4a7db71ea06 ("drm/amdgpu: add per device user friendly xgmi events for vega20") CC: Jonathan Kim Signed-off-by: kernel test robot Signed-off-by: Alex Deucher

drm/amdgpu: add xgmi perfmons for arcturus

2020-10-07T18:44:40+00:00

Add xgmi perfmons for Arcturus. v4: Align with patch 2 with struct improvements v3: Align with patch 2 streamlining perf types versus event config types. v2: Resend for re-review with alignment for v3 in patch 2. Signed-off-by: Jonathan Kim Tested-by: Chris Freehill Reviewed-by: Harish Kasiviswanathan Signed-off-by: Alex Deucher

drm/amdgpu: add per device user friendly xgmi events for vega20

2020-10-07T18:44:33+00:00

Non-outbound data metrics are non useful so mark them as legacy. Bucket new perf counters into device and not device ip. Bind events to chip instead of IP. Report available event counters and not number of hw counter banks. Move DF public macros to private since not needed outside of IP version. v5: cleanup by moving per chip configs into structs v4: After more discussion, replace *_LEGACY references with IP references to indicate concept of pmu-typed versus event-config-typed event registration. v3: attr groups const array is global but attr groups are allocated per device which doesn't work and causes problems on memory allocation and de-allocation for pmu unregister. Switch to building const attr groups per pmu instead to simplify solution. v2: add comments on sysfs structure and formatting. Signed-off-by: Jonathan Kim Reviewed-by: Harish Kasiviswanathan Signed-off-by: Alex Deucher

drm/amdgpu: fix xgmi perfmon a-b-a problem

2020-10-07T18:44:16+00:00

Mapping hw counters per event config will cause ABA problems so map per event instead. v2: Discontinue starting perf counters if add fails. Make it clear what's happening with pmc_start. Signed-off-by: Jonathan Kim Reviewed-by: Harish Kasiviswanathan Signed-off-by: Alex Deucher

drm/amdgpu: Get DRM dev from adev by inline-f

2020-08-24T17:06:06+00:00

Add a static inline adev_to_drm() to obtain the DRM device pointer from an amdgpu_device pointer. Signed-off-by: Luben Tuikov Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher

drm/amdgpu: fix amdgpu pmu to use hwc->config instead of hwc->conf

2020-02-11T20:35:54+00:00

hwc->conf was designated specifically for AMD APU IOMMU purposes. This could cause problems in performance and/or function since APU IOMMU implementation is elsewhere. Also hwc->conf and hwc->config are different members of an anonymous union so hwc->conf aliases as hw->last_tag. Signed-off-by: Jonathan Kim Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher

drm/amdgpu: Create generic DF struct in adev

2020-01-14T15:18:41+00:00

The only data fabric information the adev struct currently contains is a function pointer table. In the near future, we will be adding some cached DF information into adev. As such, this patch creates a new amdgpu_df struct for adev. Right now, it only containst the old function pointer table, but new stuff will be added soon. Signed-off-by: Joseph Greathouse Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher