Commit e2195f7d authored Nov 26, 2019 by Monk Liu Committed by Alex Deucher Dec 03, 2019

drm/amdgpu: use CPU to flush vmhub if sched stopped



otherwse the flush_gpu_tlb will hang if we unload the
KMD becuase the schedulers already stopped

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

parent c38402fe

drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c

+2 −1

Original line number	Diff line number	Diff line
		@@ -326,7 +326,8 @@ static void gmc_v10_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,

		if (!adev->mman.buffer_funcs_enabled \|\|
		!adev->ib_pool_ready \|\|
		adev->in_gpu_reset) {
		adev->in_gpu_reset \|\|
		ring->sched.ready == false) {
		gmc_v10_0_flush_vm_hub(adev, vmid, AMDGPU_GFXHUB_0, 0);
		mutex_unlock(&adev->mman.gtt_window_lock);
		return;

Admin message