Skip to contentSkip to conversation

drm/i915/gvt: do not deliver a workload if its creation fails

PATCH New Yan Zhao 1 participant 2 messages Mar 27, 2019

Subsystems (4)
lore
Share & tools
lore

Yan Zhao

author
#0 lore raw

PATCH v2 1/2 drm/i915/gvt: do not deliver a workload if its creation fails

in workload creation routine, if any failure occurs, do not queue this workload for delivery. if this failure is fatal, enter into failsafe mode.

v2: add fixes tag and cc stable kernel (zhenyu wang)

Fixes 6d76303553ba ("drm/i915/gvt: Move common vGPU workload creation into

scheduler.c")

---
 drivers/gpu/drm/i915/gvt/scheduler.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/gpu/drm/i915/gvt/scheduler.c +3 -2
index c2e3d7bcac82..8aff97de2c01 100644
@@ -1503,8 +1503,9 @@ intel_vgpu_create_workload(struct intel_vgpu *vgpu, int ring_id,
 		intel_runtime_pm_put_unchecked(dev_priv);
 	}
 
-	if (ret && (vgpu_is_vm_unhealthy(ret))) {
-		enter_failsafe_mode(vgpu, GVT_FAILSAFE_GUEST_ERR);
+	if (ret) {
+		if (vgpu_is_vm_unhealthy(ret))
+			enter_failsafe_mode(vgpu, GVT_FAILSAFE_GUEST_ERR);
 		intel_vgpu_destroy_workload(workload);
 		return ERR_PTR(ret);
 	}
Signature · 1 line
2.17.1
author
#1 lore raw

PATCH v2 2/2 drm/i915/gvt: do not let pin count of shadow mm go negative

shadow mm's pin count got increased in workload preparation phase, which is after workload scanning.
it will get decreased in complete_current_workload() anyway after workload completion.
Sometimes, if a workload meets a scanning error, its shadow mm pin count will not get increased but will get decreased in the end. This patch lets shadow mm's pin count not go below 0.

v2: add fixes tag and cc stable kernel (zhenyu wang)

Fixes 2707e4446688 ("drm/i915/gvt: vGPU graphics memory virtualization")
---
 drivers/gpu/drm/i915/gvt/gtt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/i915/gvt/gtt.c +1 -1
index f4c992d96087..88ed2e9df253 100644
@@ -1946,7 +1946,7 @@ void _intel_vgpu_mm_release(struct kref *mm_ref)
  */
 void intel_vgpu_unpin_mm(struct intel_vgpu_mm *mm)
 {
-	atomic_dec(&mm->pincount);
+	atomic_dec_if_positive(&mm->pincount);
 }
 
 /**
Signature · 1 line
2.17.1