linux/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c, branch v4.7

Revert "drm/amdgpu: add pipeline sync while vmid switch in same ctx"

2016-06-13T22:59:17Z

This reverts commit 2ba272d7bde27e1db2cf1c6cee49b01b7ea08989. The issue fixed by this patch is specific to compute rings and the previous patch was enough. Additionally, this patch as been traced to strange behavior on some CZ systems so we might as well drop it.

drm/amdgpu: add pipeline sync while vmid switch in same ctx

2016-06-09T14:49:01Z

Since vmid-mgr supports vmid sharing in one vm, the same ctx could get different vmids for two emits without vm flush, vm_flush could be done in another ring. Signed-off-by: Chunming Zhou Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher

drm/amdgpu: Encapsulate some VM table update parameters (v2)

2016-05-11T17:53:54Z

Bundle some VM table parameters into amdgpu_vm_update_params structure, so that number of function parameters can be reduced. Only structural change, no logic change. v2: agd: squash in fix from Harish Signed-off-by: Harish Kasiviswanathan Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: remove define for reserved client ID

2016-05-11T16:31:20Z

Just set it to zero instead. Signed-off-by: Christian König Reviewed-by: Chunming Zhou Signed-off-by: Alex Deucher

drm/amdgpu: remove owner cleanup v2

2016-05-11T16:31:19Z

The client ID is now unique, so no need to resert the owner fields any more. v2: remove unused variables as well Signed-off-by: Christian König Reviewed-by: Chunming Zhou (v1) Signed-off-by: Alex Deucher

drm/amdgpu: make the VMID owner always 64bit

2016-05-11T16:31:18Z

Otherwise we could (in theory) run into problems on 32bit systems. Signed-off-by: Christian König Reviewed-by: Chunming Zhou Signed-off-by: Alex Deucher

drm/amdgpu: add pipeline sync for compute job

2016-05-11T16:31:16Z

hardware ring is async processed, the job is executed in parallel. In some case, this will result vm fault, like jobs with different vmids. This works around a CPC hw issue which will eventually be fixed in fw. Signed-off-by: Chunming Zhou Reviewed-by: Alex Deucher Reviewed-by: Christian König Reviewed-by: Monk Liu Signed-off-by: Alex Deucher

drm/amdgpu: keep vm in job instead of ib (v2)

2016-05-11T16:31:16Z

ib.vm is a legacy way to get vm, after scheduler implemented vm should be get from job, and all ibs from one job share the same vm, no need to keep ib.vm just move vm field to job. this patch as well add job as paramter to ib_schedule so it can get vm from job->vm. v2: agd: sqaush in: drm/amdgpu: check if ring emit_vm_flush exists in vm flush No vm flush on engines that don't support VM. bug: https://bugs.freedesktop.org/show_bug.cgi?id=95195 Signed-off-by: Monk Liu Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher

drm/amdgpu: make vmid owner be client_id

2016-05-11T16:31:15Z

Using the pointer is not adequate. Signed-off-by: Chunming Zhou Reviewed-by: Alex Deucher Reviewed-by: Monk Liu Signed-off-by: Alex Deucher

drm/amdgpu: add client id for every vm

2016-05-11T16:31:14Z

This adds a unique id for each vm client so we can properly track them. Signed-off-by: Chunming Zhou Reviewed-by: Alex Deucher Reviewed-by: Monk Liu Signed-off-by: Alex Deucher