drm/i915/gvt: Add VM healthy check for workload_thread

When a scan error occurs in dispatch_workload, this patch is to
check the healthy state and free all the queued workloads before
the failsafe mode is entered.
Signed-off-by: fred gao <fred.gao@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>

drivers/gpu/drm/i915/gvt/gvt.h

@@ -193,6 +193,10 @@ struct intel_vgpu {
 #endif
 };
 
+/* validating GM healthy status*/
+#define vgpu_is_vm_unhealthy(ret_val) \
+	(((ret_val) == -EBADRQC) || ((ret_val) == -EFAULT))
+
 struct intel_gvt_gm {
 	unsigned long vgpu_allocated_low_gm_size;
 	unsigned long vgpu_allocated_high_gm_size;
@@ -497,6 +501,7 @@ int intel_vgpu_emulate_opregion_request(struct intel_vgpu *vgpu, u32 swsci);
 void populate_pvinfo_page(struct intel_vgpu *vgpu);
 
 int intel_gvt_scan_and_shadow_workload(struct intel_vgpu_workload *workload);
+void enter_failsafe_mode(struct intel_vgpu *vgpu, int reason);
 
 struct intel_gvt_ops {
 	int (*emulate_cfg_read)(struct intel_vgpu *, unsigned int, void *,
@@ -519,6 +524,7 @@ struct intel_gvt_ops {
 enum {
 	GVT_FAILSAFE_UNSUPPORTED_GUEST,
 	GVT_FAILSAFE_INSUFFICIENT_RESOURCE,
+	GVT_FAILSAFE_GUEST_ERR,
 };
 
 static inline void mmio_hw_access_pre(struct drm_i915_private *dev_priv)

drivers/gpu/drm/i915/gvt/handlers.c

@@ -157,7 +157,7 @@ static int render_mmio_to_ring_id(struct intel_gvt *gvt, unsigned int reg)
 	(num * 8 + i915_mmio_reg_offset(FENCE_REG_GEN6_LO(0)))
 
 
-static void enter_failsafe_mode(struct intel_vgpu *vgpu, int reason)
+void enter_failsafe_mode(struct intel_vgpu *vgpu, int reason)
 {
 	switch (reason) {
 	case GVT_FAILSAFE_UNSUPPORTED_GUEST:
@@ -165,6 +165,8 @@ static void enter_failsafe_mode(struct intel_vgpu *vgpu, int reason)
 		break;
 	case GVT_FAILSAFE_INSUFFICIENT_RESOURCE:
 		pr_err("Graphics resource is not enough for the guest\n");
+	case GVT_FAILSAFE_GUEST_ERR:
+		pr_err("GVT Internal error  for the guest\n");
 	default:
 		break;
 	}

drivers/gpu/drm/i915/gvt/scheduler.c

@@ -634,6 +634,13 @@ static int workload_thread(void *priv)
 					FORCEWAKE_ALL);
 
 		intel_runtime_pm_put(gvt->dev_priv);
+		if (ret && (vgpu_is_vm_unhealthy(ret))) {
+			mutex_lock(&gvt->lock);
+			intel_vgpu_clean_execlist(vgpu);
+			mutex_unlock(&gvt->lock);
+			enter_failsafe_mode(vgpu, GVT_FAILSAFE_GUEST_ERR);
+		}
+
 	}
 	return 0;
 }