[PATCH] drm/amdgpu: skip ring soft reset on SR-IOV VF

Christian K├Ânig ckoenig.leichtzumerken at gmail.com
Wed Nov 14 10:24:05 UTC 2018


Am 14.11.18 um 09:23 schrieb Trigger Huang:
> When TDR happened on SR-IOV VF, the host Hypervisor may has already
> done the VF FLR, then, the re-initialization flow in KMD is a must.
> The same logic is in amdgpu_device_ip_check_soft_reset, too

NAK, this is only called in the timeout code path and not when the 
Hypervisor signals a FLR.

Christian.

>
> Signed-off-by: Trigger Huang <Trigger.Huang at amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c | 3 +++
>   1 file changed, 3 insertions(+)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
> index 5b75bdc..14bd67a 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
> @@ -397,6 +397,9 @@ bool amdgpu_ring_soft_recovery(struct amdgpu_ring *ring, unsigned int vmid,
>   {
>   	ktime_t deadline = ktime_add_us(ktime_get(), 10000);
>   
> +	if (amdgpu_sriov_vf(ring->adev))
> +		return false;
> +
>   	if (!ring->funcs->soft_recovery)
>   		return false;
>   



More information about the amd-gfx mailing list