summaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c
Commit message (Collapse)AuthorAgeFilesLines
...
* drm/amdgpu: fix scheduler thread creation error checkingChristian König2015-08-251-7/+6
| | | | | Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: fix entity wakeup race conditionChristian König2015-08-251-22/+21
| | | | | | | That actually didn't worked at all. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: cleanup entity pickingChristian König2015-08-251-6/+10
| | | | | | | | Cleanup function name, stop checking scheduler ready twice, but check if kernel thread should stop instead. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: remove some more unused entity members v2Christian König2015-08-251-8/+3
| | | | | | | | | None of them are used any more. v2: fix type in error message Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: rework scheduler submission handling.Christian König2015-08-251-33/+10
| | | | | | | | | | | | | Remove active_hw_rq and it's protecting queue_lock, they are unused. User 32bit atomic for hw_rq_count, 64bits for counting to three is a bit overkill. Cleanup the function name and remove incorrect comments. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Jammy Zhou <Jammy.Zhou@amd.com>
* drm/amdgpu: remove v_seq handling from the scheduler v2Christian König2015-08-251-18/+1
| | | | | | | | | | | Simply not used any more. Only keep 32bit atomic for fence sequence numbering. v2: trivial rebase Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1) Reviewed-by: Jammy Zhou <Jammy.Zhou@amd.com> (v1) Reviewed-by: Chunming Zhou <david1.zhou@amd.com> (v1)
* drm/amdgpu: use a spinlock instead of a mutex for the rqChristian König2015-08-201-32/+17
| | | | | | | More appropriate and fixes some nasty lockdep warnings. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: abstract amdgpu_job for schedulerChunming Zhou2015-08-201-27/+15
| | | | | Signed-off-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian K?nig <christian.koenig@amd.com>
* drm/amdgpu: cleanup sheduler rq handling v2Christian König2015-08-171-59/+44
| | | | | | | | | Rework run queue implementation, especially remove the odd list handling. v2: cleanup the code only, no algorithem change. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: fix unnecessary wake upChunming Zhou2015-08-171-4/+6
| | | | | | | | decrease CPU extra overhead. Signed-off-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Christian K?nig <christian.koenig@amd.com>
* drm/amdgpu: add reference for **fenceChunming Zhou2015-08-171-0/+1
| | | | | | | | fix fence is released when pass to **fence sometimes. add reference for it. Signed-off-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian K?nig <christian.koenig@amd.com>
* drm/amdgpu: remove scheduler fence list v2Christian König2015-08-171-1/+0
| | | | | | | | | Unused and missing proper locking. v2: add locking comment to commit message. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> (v1)
* drm/amdgpu: remove amd_sched_wait_emit v2Christian König2015-08-171-61/+0
| | | | | | | | | Not used any more. v2: remove amd_sched_emit as well. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amd: add scheduler fence implementation (v2)Chunming Zhou2015-08-171-3/+23
| | | | | | | | | scheduler fence is based on kernel fence framework. v2: squash in Christian's build fix Signed-off-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian K?nig <christian.koenig@amd.com>
* drm/amdgpu: use amd_sched_job in its backend opsChunming Zhou2015-08-171-16/+14
| | | | | Signed-off-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian K?nig <christian.koenig@amd.com>
* drm/amdgpu: cleanup and fix scheduler fence handling v2Christian König2015-08-171-28/+33
| | | | | | | | v2: rebased Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (v1) Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: merge amd_sched_entity and amd_context_entity v2Christian König2015-08-171-39/+28
| | | | | | | | | Avoiding a couple of casts. v2: rename c_entity to entity as well Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: fix coding style in a couple of placesChristian König2015-08-171-7/+10
| | | | | | Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: remove unused parent entityChristian König2015-08-171-3/+0
| | | | | | Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: process sched job exactly triggered by fence signalChunming Zhou2015-08-171-36/+41
| | | | | Signed-off-by: Chunming Zhou <david1.zhou@amd.com> Reviewed-by: Christian K?nig <christian.koenig@amd.com>
* Revert "drm/amdgpu: return new seq_no for amd_sched_push_job"Chunming Zhou2015-08-171-4/+6
| | | | | | | | | | This reverts commit d1d33da8eb86b8ca41dd9ed95738030df5267b95. Reviewed-by: Christian K?nig <christian.koenig@amd.com> Conflicts: drivers/gpu/drm/amd/amdgpu/amdgpu_sched.c drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
* drm/amdgpu: stop leaking the ctx id into the scheduler v2Christian König2015-08-171-7/+3
| | | | | | | | | Id's are for the IOCTL ABI only. v2: remove tgid as well Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: add amd_sched_next_queued_seq functionJammy Zhou2015-08-171-0/+12
| | | | | | | | This function is used to get the next queued sequence number Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: make last_handled_seq atomicJammy Zhou2015-08-171-3/+3
| | | | | | Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: add amd_sched_commitJammy Zhou2015-08-171-0/+12
| | | | | | | | | | This function is to update last_emitted_v_seq and wake up the waiters. It should be called by driver in the run_job backend function Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: return new seq_no for amd_sched_push_jobJammy Zhou2015-08-171-6/+4
| | | | | | | | It is clean to update last_queued_v_seq in the scheduler module Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
* drm/amdgpu: clean up amd sched wait_ts and wait_signalChristian König2015-08-171-40/+7
| | | | | | | Remove code not used at the moment. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: add amdgpu.sched_hw_submission optionJammy Zhou2015-08-171-2/+3
| | | | | | | | This option can be used to specify the max number of submissions in the active HW queue. The default value is 2 now. Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: add amdgpu.sched_jobs optionJammy Zhou2015-08-171-2/+4
| | | | | | | | This option can be used to specify the max job number in the job queue, and it is 16 by default. Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amdgpu: silent the message for GPU scheduler creationJammy Zhou2015-08-171-2/+0
| | | | | Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
* drm/amd: add basic scheduling frameworkJammy Zhou2015-08-171-0/+531
run queue: A set of entities scheduling commands for the same ring. It implements the scheduling policy that selects the next entity to emit commands from. entity: A scheduler entity is a wrapper around a job queue or a group of other entities. This can be used to build hierarchies of entities. For example all job queue entities belonging to the same process may be placed in a higher level entity and scheduled against other process entities. Entities take turns emitting jobs from their job queue to the corresponding hardware ring, in accordance with the scheduler policy. Signed-off-by: Shaoyun Liu <Shaoyun.Liu@amd.com> Signed-off-by: Chunming Zhou <David1.Zhou@amd.com> Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> Acked-by: Christian K?nig <christian.koenig@amd.com> Reviewed-by: Jammy Zhou <Jammy.Zhou@amd.com>
OpenPOWER on IntegriCloud