| Message ID | 20230905213416.1290219-1-mcanal@igalia.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | drm/v3d: Expose GPU usage stats | expand |
I've just tested this series, and it is working perfectly. For the two patches of the series: Acked-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com> Thanks Maíra for taking care of upstreaming this feature. Chema El 5/9/23 a las 23:06, Maíra Canal escribió: > This patchset exposes GPU usages stats both globally and per-file > descriptor. > > The first patch exposes the accumulated amount of active time per client > through the fdinfo infrastructure. The amount of active time is exposed > for each V3D queue. Moreover, it exposes the number of jobs submitted to > each queue. > > The second patch exposes the accumulated amount of active time for each > V3D queue, independent of the client. This data is exposed through the > sysfs interface. > > With these patches, it is possible to calculate the GPU usage percentage > per queue globally and per-file descriptor. > > * Example fdinfo output: > > $ cat /proc/1140/fdinfo/4 > pos: 0 > flags: 02400002 > mnt_id: 24 > ino: 209 > drm-driver: v3d > drm-client-id: 44 > drm-engine-bin: 1661076898 ns > v3d-jobs-bin: 19576 jobs > drm-engine-render: 31469427170 ns > v3d-jobs-render: 19575 jobs > drm-engine-tfu: 5002964 ns > v3d-jobs-tfu: 13 jobs > drm-engine-csd: 188038329691 ns > v3d-jobs-csd: 250393 jobs > drm-engine-cache_clean: 27736024038 ns > v3d-jobs-cache_clean: 250392 job > > * Example gputop output: > > DRM minor 128 > PID bin render tfu csd cache_clean NAME > 1140 |▎ ||██▋ || ||█████████████▍ ||█▋ | computecloth > 1158 |▍ ||████████▉ || || || | gears > 1002 |▏ ||█▎ || || || | chromium-browse > > Best Regards, > - Maíra > > --- > > v1 -> v2:https://lore.kernel.org/dri-devel/20230727142929.1275149-1-mcanal@igalia.com/T/ > > * Use sysfs to expose global GPU stats (Tvrtko Ursulin) > > v2 -> v3:https://lore.kernel.org/dri-devel/20230807211849.49867-1-mcanal@igalia.com/T/ > > * Document the expected behavior in case of a GPU reset (Melissa Wen) > * Add a brief description about the sysfs outputs (Melissa Wen) > * Instead of having multiple sysfs files, use only one sysfs file, > called gpu_stats, with all the information (Chema Casanova) > * Add the number of jobs submitted in the global GPU stats (Chema Casanova) > * Now, the number of jobs submitted is only incremented if the job was > completed > > Maíra Canal (2): > drm/v3d: Implement show_fdinfo() callback for GPU usage stats > drm/v3d: Expose the total GPU usage stats on sysfs > > drivers/gpu/drm/v3d/Makefile | 3 +- > drivers/gpu/drm/v3d/v3d_drv.c | 45 ++++++++++++++++++++- > drivers/gpu/drm/v3d/v3d_drv.h | 31 +++++++++++++++ > drivers/gpu/drm/v3d/v3d_gem.c | 7 +++- > drivers/gpu/drm/v3d/v3d_irq.c | 49 +++++++++++++++++++++++ > drivers/gpu/drm/v3d/v3d_sched.c | 33 ++++++++++++++++ > drivers/gpu/drm/v3d/v3d_sysfs.c | 69 +++++++++++++++++++++++++++++++++ > 7 files changed, 234 insertions(+), 3 deletions(-) > create mode 100644 drivers/gpu/drm/v3d/v3d_sysfs.c > > -- > 2.41.0 > >
Hi, I've just applied this patchset to drm-misc/drm-misc-next. Thanks Melissa and Chema for reviewing it! Best Regards, - Maíra On 9/5/23 18:06, Maíra Canal wrote: > This patchset exposes GPU usages stats both globally and per-file > descriptor. > > The first patch exposes the accumulated amount of active time per client > through the fdinfo infrastructure. The amount of active time is exposed > for each V3D queue. Moreover, it exposes the number of jobs submitted to > each queue. > > The second patch exposes the accumulated amount of active time for each > V3D queue, independent of the client. This data is exposed through the > sysfs interface. > > With these patches, it is possible to calculate the GPU usage percentage > per queue globally and per-file descriptor. > > * Example fdinfo output: > > $ cat /proc/1140/fdinfo/4 > pos: 0 > flags: 02400002 > mnt_id: 24 > ino: 209 > drm-driver: v3d > drm-client-id: 44 > drm-engine-bin: 1661076898 ns > v3d-jobs-bin: 19576 jobs > drm-engine-render: 31469427170 ns > v3d-jobs-render: 19575 jobs > drm-engine-tfu: 5002964 ns > v3d-jobs-tfu: 13 jobs > drm-engine-csd: 188038329691 ns > v3d-jobs-csd: 250393 jobs > drm-engine-cache_clean: 27736024038 ns > v3d-jobs-cache_clean: 250392 job > > * Example gputop output: > > DRM minor 128 > PID bin render tfu csd cache_clean NAME > 1140 |▎ ||██▋ || ||█████████████▍ ||█▋ | computecloth > 1158 |▍ ||████████▉ || || || | gears > 1002 |▏ ||█▎ || || || | chromium-browse > > Best Regards, > - Maíra > > --- > > v1 -> v2: https://lore.kernel.org/dri-devel/20230727142929.1275149-1-mcanal@igalia.com/T/ > > * Use sysfs to expose global GPU stats (Tvrtko Ursulin) > > v2 -> v3: https://lore.kernel.org/dri-devel/20230807211849.49867-1-mcanal@igalia.com/T/ > > * Document the expected behavior in case of a GPU reset (Melissa Wen) > * Add a brief description about the sysfs outputs (Melissa Wen) > * Instead of having multiple sysfs files, use only one sysfs file, > called gpu_stats, with all the information (Chema Casanova) > * Add the number of jobs submitted in the global GPU stats (Chema Casanova) > * Now, the number of jobs submitted is only incremented if the job was > completed > > Maíra Canal (2): > drm/v3d: Implement show_fdinfo() callback for GPU usage stats > drm/v3d: Expose the total GPU usage stats on sysfs > > drivers/gpu/drm/v3d/Makefile | 3 +- > drivers/gpu/drm/v3d/v3d_drv.c | 45 ++++++++++++++++++++- > drivers/gpu/drm/v3d/v3d_drv.h | 31 +++++++++++++++ > drivers/gpu/drm/v3d/v3d_gem.c | 7 +++- > drivers/gpu/drm/v3d/v3d_irq.c | 49 +++++++++++++++++++++++ > drivers/gpu/drm/v3d/v3d_sched.c | 33 ++++++++++++++++ > drivers/gpu/drm/v3d/v3d_sysfs.c | 69 +++++++++++++++++++++++++++++++++ > 7 files changed, 234 insertions(+), 3 deletions(-) > create mode 100644 drivers/gpu/drm/v3d/v3d_sysfs.c > > -- > 2.41.0 >