[PATCH 2/2] drm/amdgpu: disable gpu_sched load balancer for vcn jobs

[Christian König]

Am 13.03.20 um 14:59 schrieb Nirmoy Das:
> VCN HW doesn't support dynamic load balance on multiple instances
> for a context. This patch initializes VNC entities with only one
> drm_gpu_scheduler picked by drm_sched_pick_best(). Picking a
> drm_gpu_scheduler using drm_sched_pick_best() ensures that we
> do load balance among multiple contexts but not among multiple
> jobs in a context.
>
> Signed-off-by: Nirmoy Das <nirmoy.das at amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 7 ++++++-
>   1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
> index fa575bdc03c8..8f1e3c05812e 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
> @@ -121,12 +121,17 @@ static int amdgpu_ctx_init_entity(struct amdgpu_ctx *ctx, const u32 hw_ip, const
>   			num_scheds = 1;
>   			break;
>   		case AMDGPU_HW_IP_VCN_DEC:
> -			scheds = adev->vcn.vcn_dec_sched;
>   			num_scheds =  adev->vcn.num_vcn_dec_sched;
> +			sched = drm_sched_pick_best(adev->vcn.vcn_dec_sched,
> +						    num_scheds);
> +			scheds = &sched;
>   			break;
>   		case AMDGPU_HW_IP_VCN_ENC:
>   			scheds = adev->vcn.vcn_enc_sched;
>   			num_scheds =  adev->vcn.num_vcn_enc_sched;
> +			sched = drm_sched_pick_best(adev->vcn.vcn_enc_sched,
> +						    num_scheds);
> +			scheds = &sched;

num_scheds need to be set to 1 here and above.

And maybe add a separate patch to fix the indentation of the case labels.

Christian.

>   			break;
>   		case AMDGPU_HW_IP_VCN_JPEG:
>   			scheds = adev->jpeg.jpeg_sched;