drm/i915/gvt: remove the vgpu->released and its sanity check

The life cycle of a vGPU, which is represented by a vfio_device, has been
managed by the VFIO core logic. Remove the vgpu->released, which was used
for a sanity check on the removal path of the vGPU instance. The sanity
check has already been covered in the VFIO core logic.

Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: Kevin Tian <kevin.tian@intel.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: intel-gvt-dev@lists.freedesktop.org
Suggested-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Zhi Wang <zhi.a.wang@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20221104145652.1570-1-zhi.a.wang@intel.comReviewed-by: Zhenyu Wang <zhenyuw@linux.intel.com>

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

@@ -226,8 +226,6 @@ struct intel_vgpu {
 	unsigned long nr_cache_entries;
 	struct mutex cache_lock;
 
-	atomic_t released;
-
 	struct kvm_page_track_notifier_node track_node;
 #define NR_BKT (1 << 18)
 	struct hlist_head ptable[NR_BKT];

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

@@ -785,7 +785,6 @@ static int intel_vgpu_open_device(struct vfio_device *vfio_dev)
 
 	intel_gvt_activate_vgpu(vgpu);
 
-	atomic_set(&vgpu->released, 0);
 	return 0;
 }
 
@@ -807,9 +806,6 @@ static void intel_vgpu_close_device(struct vfio_device *vfio_dev)
 	if (!vgpu->attached)
 		return;
 
-	if (atomic_cmpxchg(&vgpu->released, 0, 1))
-		return;
-
 	intel_gvt_release_vgpu(vgpu);
 
 	debugfs_remove(debugfs_lookup(KVMGT_DEBUGFS_FILENAME, vgpu->debugfs));