[Intel-xe] [PATCH 1/1] fixup! drm/xe: standardize vm-less kernel submissions

[Daniele Ceraolo Spurio]

VM-less permanent queues exists for the whole lifetime of the driver, so
they can't hold a wakeref. Given that these queues are never
disabled/deregistered and should never hang, we don't expect any G2H
messages outside of the submission flow and therefore we don't need the
extra ref at all as long as the driver makes sure to hold a wakeref when
submitting.

Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio at intel.com>
Cc: Matthew Brost <matthew.brost at intel.com>
---
 drivers/gpu/drm/xe/xe_exec_queue.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_exec_queue.c b/drivers/gpu/drm/xe/xe_exec_queue.c
index cb284c4ad049..85574740bc1e 100644
--- a/drivers/gpu/drm/xe/xe_exec_queue.c
+++ b/drivers/gpu/drm/xe/xe_exec_queue.c
@@ -95,7 +95,7 @@ static struct xe_exec_queue *__xe_exec_queue_create(struct xe_device *xe,
 	 * can perform GuC CT actions when needed. Caller is expected to have
 	 * already grabbed the rpm ref outside any sensitive locks.
 	 */
-	if (q->flags & EXEC_QUEUE_FLAG_VM || !vm)
+	if (!(q->flags & EXEC_QUEUE_FLAG_PERMANENT) && (q->flags & EXEC_QUEUE_FLAG_VM || !vm))
 		drm_WARN_ON(&xe->drm, !xe_device_mem_access_get_if_ongoing(xe));
 
 	return q;
@@ -172,7 +172,7 @@ void xe_exec_queue_fini(struct xe_exec_queue *q)
 
 	for (i = 0; i < q->width; ++i)
 		xe_lrc_finish(q->lrc + i);
-	if (q->flags & EXEC_QUEUE_FLAG_VM || !q->vm)
+	if (!(q->flags & EXEC_QUEUE_FLAG_PERMANENT) && (q->flags & EXEC_QUEUE_FLAG_VM || !q->vm))
 		xe_device_mem_access_put(gt_to_xe(q->gt));
 	if (q->vm)
 		xe_vm_put(q->vm);
-- 
2.43.0