Commit 8b2ac103 authored Jun 12, 2016 by Chunming Zhou Committed by Alex Deucher Jul 07, 2016

drm/amdgpu: evict vram when gpu reset



On workstation cards with ECC vram, the entirety of vram is cleared to 0
on asic init to set the ECC status correctly.  On non ECC boards, I don't
think they do any explicit clearing, but the vram controller is reset
which may cause issues with the data there.

Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

parent 0875dc9e

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

+4 −0

Original line number	Diff line number	Diff line
		@@ -1896,6 +1896,9 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev)

		atomic_inc(&adev->gpu_reset_counter);

		/* evict vram memory */
		amdgpu_bo_evict_vram(adev);

		/* block scheduler */
		for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
		struct amdgpu_ring *ring = adev->rings[i];
		@@ -1904,6 +1907,7 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev)
		continue;
		kthread_park(ring->sched.thread);
		}

		/* block TTM */
		resched = ttm_bo_lock_delayed_workqueue(&adev->mman.bdev);