linux-stable.git/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c, branch v4.8 Linux kernel stable tree drm/amdgpu: cleanup hw reference handling in the IB tests 2016-07-14T20:46:05+00:00 Christian König christian.koenig@amd.com 2016-07-05T12:48:17+00:00 22a77cf6d87e0cfadf91bf7d09bae71359bc85f6 Reference should be taken when we make the assignment, not anywhere else. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
Reference should be taken when we make the assignment, not anywhere else.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: add a bool to specify if needing vm flush V2 2016-07-07T19:06:16+00:00 Chunming Zhou David1.Zhou@amd.com 2016-07-01T09:59:01+00:00 fd53be302f0efabead8e37553eaeed1572d0f727 which avoids job->vm_pd_addr be changed. V2: pass job structure to amdgpu_vm_grab_id and amdgpu_vm_flush directly. Signed-off-by: Chunming Zhou <David1.Zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
which avoids job->vm_pd_addr be changed.

V2: pass job structure to amdgpu_vm_grab_id and amdgpu_vm_flush directly.

Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: put old hw fence of job if gpu reset 2016-07-07T19:06:12+00:00 Chunming Zhou David1.Zhou@amd.com 2016-06-30T09:30:42+00:00 c7c5fbcdc3b064943491d8dd1229cb25479e9093 Signed-off-by: Chunming Zhou <David1.Zhou@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: remove fence parameter from amd_sched_job_init 2016-07-07T19:06:11+00:00 Christian König christian.koenig@amd.com 2016-06-30T08:52:03+00:00 595a9cd68c132e474ee5daf97067d4d15c618739 We return the fence as part of the job structur anyway, no need to do this twice. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
We return the fence as part of the job structur anyway,
no need to do this twice.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: earlier free SA resources 2016-07-07T19:06:10+00:00 Christian König christian.koenig@amd.com 2016-06-29T13:10:31+00:00 a5fb4ec29c74a16ce1c269e52bc85ca86ee41e81 Keep the time we don't have a fence associated with the resource smaller. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
Keep the time we don't have a fence associated with the resource smaller.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: shorten amdgpu_job_free_resources 2016-07-07T19:06:10+00:00 Christian König christian.koenig@amd.com 2016-06-29T11:29:57+00:00 a79a5bdcefc4d283b58eef46804c43c88789dd1f The fence and the sync object are not hardware resources. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
The fence and the sync object are not hardware resources.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: fix user fence handling once more 2016-07-07T19:06:09+00:00 Christian König christian.koenig@amd.com 2016-06-29T11:26:41+00:00 b5f5acbc87052e1bd8ada6915e1dedd856da767d Same problem as with the VM page tables. The user fence address must be determined before the job is scheduled, not when the IB is executed. This fixes a security problem where user fences could be used to overwrite any part of VRAM. Signed-off-by: Christian König <christian.koenig@amd.com> Reviewed-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
Same problem as with the VM page tables. The user fence address must be
determined before the job is scheduled, not when the IB is executed.

This fixes a security problem where user fences could be used to overwrite
any part of VRAM.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: use a fence array for VMID management 2016-07-07T18:51:23+00:00 Christian König christian.koenig@amd.com 2016-06-01T08:47:36+00:00 1fbb2e929902ab6e161ebcfb2f4d6de1c4613473 Just wait for any fence to become available, instead of waiting for the last entry of the LRU. Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
Just wait for any fence to become available, instead
of waiting for the last entry of the LRU.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: add optional ring to amdgpu_sync_is_idle 2016-07-07T18:51:21+00:00 Christian König christian.koenig@amd.com 2016-05-23T12:26:39+00:00 354202389614bc7c29b6fbb98677326fcc248055 Check if the sync object is idle depending on the ring a submission works with. Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
Check if the sync object is idle depending on the ring a submission works with.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: remove amdgpu_sync_wait 2016-07-07T18:51:21+00:00 Christian König christian.koenig@amd.com 2016-05-20T11:06:09+00:00 a7e7a93e5766dcd2929f8e2ca280967ef1572d27 Stop hiding bugs, instead print a proper error when the scheduler doesn't handle all dependencies. Acked-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Christian König <christian.koenig@amd.com> Acked-by: Chunming Zhou <david1.zhou@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 
Stop hiding bugs, instead print a proper error when the scheduler
doesn't handle all dependencies.

Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>