drm/i915/gvt: Don't mark vgpu context as inactive when preempted

We shouldn't mark inactive for vGPU context if preempted,
which would still be re-scheduled later. So keep active state.

Fixes: d6c0511300 ("drm/i915/execlists: Distinguish the incomplete context notifies")
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
This commit is contained in:
Zhenyu Wang 2017-12-01 14:59:53 +08:00
parent c130456cef
commit da5f99eacc

View File

@ -188,10 +188,12 @@ static int shadow_context_status_change(struct notifier_block *nb,
atomic_set(&workload->shadow_ctx_active, 1);
break;
case INTEL_CONTEXT_SCHEDULE_OUT:
case INTEL_CONTEXT_SCHEDULE_PREEMPTED:
save_ring_hw_state(workload->vgpu, ring_id);
atomic_set(&workload->shadow_ctx_active, 0);
break;
case INTEL_CONTEXT_SCHEDULE_PREEMPTED:
save_ring_hw_state(workload->vgpu, ring_id);
break;
default:
WARN_ON(1);
return NOTIFY_OK;