Commit 96cf8271 authored by John Brooks's avatar John Brooks Committed by Alex Deucher
Browse files

drm/amdgpu: Set/clear CPU_ACCESS flag on page fault and move to VRAM 



When a BO is moved to VRAM, clear AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED.
This allows it to potentially later move to invisible VRAM if the CPU
does not access it again.

Setting the CPU_ACCESS flag in amdgpu_bo_fault_reserve_notify() also means
that we can remove the loop to restrict lpfn to the end of visible VRAM,
because amdgpu_ttm_placement_init() will do it for us.

v3 [Michel Dänzer]
* Use AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED instead of a new flag
  (Christian König)
* Clear flag in amdgpu_bo_move instead of amdgpu_move_ram_vram
  (Christian)
* Explicitly mention amdgpu_bo_fault_reserve_notify in amdgpu_bo_move
* Also clear flag in amdgpu_bo_create_restricted

Suggested-by: default avatarMichel Dänzer <michel.daenzer@amd.com>
Signed-off-by: default avatarJohn Brooks <john@fastquake.com>
Reviewed-by: default avatarChristian König <christian.koenig@amd.com>
Signed-off-by: default avatarMichel Dänzer <michel.daenzer@amd.com>
Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
parent 00f06b24

drivers/gpu/drm/amd/amdgpu/amdgpu_object.c

+10 −10
Original line number Original line Diff line number Diff line
@@ -432,6 +432,10 @@ int amdgpu_bo_create_restricted(struct amdgpu_device *adev, 




	trace_amdgpu_bo_create(bo);

	trace_amdgpu_bo_create(bo);





	/* Treat CPU_ACCESS_REQUIRED only as a hint if given by UMD */

	if (type == ttm_bo_type_device)

		bo->flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;



	return 0;

	return 0;





fail_unreserve:

fail_unreserve:
@@ -945,13 +949,17 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) 
{

{

	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev);

	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev);

	struct amdgpu_bo *abo;

	struct amdgpu_bo *abo;

	unsigned long offset, size, lpfn;

	unsigned long offset, size;

	int i, r;

	int r;





	if (!amdgpu_ttm_bo_is_amdgpu_bo(bo))

	if (!amdgpu_ttm_bo_is_amdgpu_bo(bo))

		return 0;

		return 0;





	abo = container_of(bo, struct amdgpu_bo, tbo);

	abo = container_of(bo, struct amdgpu_bo, tbo);



	/* Remember that this BO was accessed by the CPU */

	abo->flags |= AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;



	if (bo->mem.mem_type != TTM_PL_VRAM)

	if (bo->mem.mem_type != TTM_PL_VRAM)

		return 0;

		return 0;
@@ -967,14 +975,6 @@ int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) 
	/* hurrah the memory is not visible ! */

	/* hurrah the memory is not visible ! */

	atomic64_inc(&adev->num_vram_cpu_page_faults);

	atomic64_inc(&adev->num_vram_cpu_page_faults);

	amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM);

	amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM);

	lpfn =	adev->mc.visible_vram_size >> PAGE_SHIFT;

	for (i = 0; i < abo->placement.num_placement; i++) {

		/* Force into visible VRAM */

		if ((abo->placements[i].flags & TTM_PL_FLAG_VRAM) &&

		    (!abo->placements[i].lpfn ||

		     abo->placements[i].lpfn > lpfn))

			abo->placements[i].lpfn = lpfn;

	}

	r = ttm_bo_validate(bo, &abo->placement, false, false);

	r = ttm_bo_validate(bo, &abo->placement, false, false);

	if (unlikely(r == -ENOMEM)) {

	if (unlikely(r == -ENOMEM)) {

		amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT);

		amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT);

drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c

+9 −0
Original line number Original line Diff line number Diff line
@@ -499,6 +499,15 @@ memcpy: 
		}

		}

	}

	}





	if (bo->type == ttm_bo_type_device &&

	    new_mem->mem_type == TTM_PL_VRAM &&

	    old_mem->mem_type != TTM_PL_VRAM) {

		/* amdgpu_bo_fault_reserve_notify will re-set this if the CPU

		 * accesses the BO after it's moved.

		 */

		abo->flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;

	}



	/* update statistics */

	/* update statistics */

	atomic64_add((u64)bo->num_pages << PAGE_SHIFT, &adev->num_bytes_moved);

	atomic64_add((u64)bo->num_pages << PAGE_SHIFT, &adev->num_bytes_moved);

	return 0;

	return 0;

CRA Git | Maintained and supported by SUSTech CRA and CCSE