[PATCH 5/7] drm/amdgpu: no need pipeline sync after gpu reset

Christian König deathsimple at vodafone.de
Wed Jun 29 08:59:22 UTC 2016


Am 29.06.2016 um 10:09 schrieb Chunming Zhou:
> Change-Id: I69ed200110a0b000280cd84d07b03fb647d8e3a7
> Signed-off-by: Chunming Zhou <David1.Zhou at amd.com>

The reason is that the original hardware fence isn't signaled, isn't it?

It would be better if we don't need this and use new hardware fence 
numbers for the resubmission.

Christian.

> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c | 3 ++-
>   1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> index cced2f6..6ba2b0a 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
> @@ -378,7 +378,8 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring,
>   		id->oa_size != oa_size);
>   	int r;
>   
> -	if (ring->funcs->emit_pipeline_sync && (
> +	if (!amdgpu_vm_is_gpu_reset(adev, id) &&
> +	    ring->funcs->emit_pipeline_sync && (
>   	    pd_addr != AMDGPU_VM_NO_FLUSH || gds_switch_needed ||
>   	    amdgpu_vm_ring_has_compute_vm_bug(ring)))
>   		amdgpu_ring_emit_pipeline_sync(ring);



More information about the amd-gfx mailing list