drm/scheduler: Fix UAF race in drm_sched_entity_push_job()

After a job is pushed into the queue, it is owned by the scheduler core
and may be freed at any time, so we can't write nor read the submit
timestamp after that point.

Fixes oopses observed with the drm/asahi driver, found with kASAN.

Signed-off-by: Asahi Lina <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Luben Tuikov <[email protected]>
Signed-off-by: Luben Tuikov <[email protected]>

1 files changed, 9 insertions, 2 deletions

diff --git a/drivers/gpu/drm/scheduler/sched_entity.c b/drivers/gpu/drm/scheduler/sched_entity.c
index 15d04a0ec623..e0a8890a62e2 100644
--- a/drivers/gpu/drm/scheduler/sched_entity.c
+++ b/drivers/gpu/drm/scheduler/sched_entity.c
@@ -507,12 +507,19 @@ void drm_sched_entity_push_job(struct drm_sched_job *sched_job)
 {
 	struct drm_sched_entity *entity = sched_job->entity;
 	bool first;
+	ktime_t submit_ts;
 
 	trace_drm_sched_job(sched_job, entity);
 	atomic_inc(entity->rq->sched->score);
 	WRITE_ONCE(entity->last_user, current->group_leader);
+
+	/*
+	 * After the sched_job is pushed into the entity queue, it may be
+	 * completed and freed up at any time. We can no longer access it.
+	 * Make sure to set the submit_ts first, to avoid a race.
+	 */
+	sched_job->submit_ts = submit_ts = ktime_get();
 	first = spsc_queue_push(&entity->job_queue, &sched_job->queue_node);
-	sched_job->submit_ts = ktime_get();
 
 	/* first job wakes up scheduler */
 	if (first) {
@@ -529,7 +536,7 @@ void drm_sched_entity_push_job(struct drm_sched_job *sched_job)
 		spin_unlock(&entity->rq_lock);
 
 		if (drm_sched_policy == DRM_SCHED_POLICY_FIFO)
-			drm_sched_rq_update_fifo(entity, sched_job->submit_ts);
+			drm_sched_rq_update_fifo(entity, submit_ts);
 
 		drm_sched_wakeup(entity->rq->sched);
 	}