Commit 4ff7d8ba authored by Nirmoy Das's avatar Nirmoy Das Committed by Alex Deucher
Browse files

drm/amdgpu: disable gpu_sched load balancer for vcn jobs 



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: default avatarNirmoy Das <nirmoy.das@amd.com>
Reviewed-by: default avatarChristian König <christian.koenig@amd.com>
Signed-off-by: default avatarAlex Deucher <alexander.deucher@amd.com>
parent ec2edcc2

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

+8 −4
Original line number Diff line number Diff line
@@ -121,12 +121,16 @@ 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,

					    adev->vcn.num_vcn_dec_sched);

		scheds = &sched;

		num_scheds = 1;

		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,

					    adev->vcn.num_vcn_enc_sched);

		scheds = &sched;

		num_scheds = 1;

		break;

	case AMDGPU_HW_IP_VCN_JPEG:

		scheds = adev->jpeg.jpeg_sched;

CRA Git | Maintained and supported by SUSTech CRA and CCSE