[PATCH 6/6] drm/xe: Clean up GuC software state after a wedge

[Stuart Summers <stuart.summers@intel.com>]

Comments also added to the code, but in the event of
a wedge or a hardware failure while communication with
GuC is outstanding (e.g. during a schedule disable or
context deregistration), the driver doesn't automatically
reset the software state as it would in a typical GT reset
since we are trying to save the state for debug. However
once the user unbinds the driver we still need to go through
and clean everything up for these exec queues so we don't
leak memory on the DRM side (e.g. LRC or LRC BO).

Add a kick start to the DRM scheduler to handle any outstanding
messages on hold during the wedge and go through the GuC stop
flow to simulate that reset on teardown.

Signed-off-by: Stuart Summers <stuart.summers@intel.com>
---
 drivers/gpu/drm/xe/xe_guc_submit.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/drivers/gpu/drm/xe/xe_guc_submit.c b/drivers/gpu/drm/xe/xe_guc_submit.c
index 5ec1e4a83d68..0bbae336c722 100644
--- a/drivers/gpu/drm/xe/xe_guc_submit.c
+++ b/drivers/gpu/drm/xe/xe_guc_submit.c
@@ -276,6 +276,14 @@ static void guc_submit_fini(struct drm_device *drm, void *arg)
 	struct xe_gt *gt = guc_to_gt(guc);
 	int ret;
 
+	/*
+	 * If GuC stopped responding during deregistration
+	 * some queues can be left in a bad state. Ensure
+	 * these are all cleaned up by going through the
+	 * GuC software reset flow.
+	 */
+	xe_guc_stop(guc);
+
 	ret = wait_event_timeout(guc->submission_state.fini_wq,
 				 xa_empty(&guc->submission_state.exec_queue_lookup),
 				 HZ * 5);
@@ -295,6 +303,13 @@ static void guc_submit_wedged_fini(void *arg)
 
 	mutex_lock(&guc->submission_state.lock);
 	xa_for_each(&guc->submission_state.exec_queue_lookup, index, q) {
+		/*
+		 * Kick start the scheduler since some messages
+		 * might have been added while the scheduler was
+		 * stopped during a wedge event.
+		 */
+		xe_sched_submission_start(&q->guc->sched);
+
 		if (exec_queue_wedged(q)) {
 			mutex_unlock(&guc->submission_state.lock);
 			xe_exec_queue_put(q);
-- 
2.34.1