diff --git a/block/blk-core.c b/block/blk-core.c
index cfe5c2b46e4692286428c9663c838df44b748dba..01f930e62d4def3d01ccadad83d47458834fa175 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -958,7 +958,7 @@ int blk_queue_enter(struct request_queue *q, blk_mq_req_flags_t flags)
 		smp_rmb();
 
 		wait_event(q->mq_freeze_wq,
-			   (atomic_read(&q->mq_freeze_depth) == 0 &&
+			   (!q->mq_freeze_depth &&
 			    (pm || !blk_queue_pm_only(q))) ||
 			   blk_queue_dying(q));
 		if (blk_queue_dying(q))
@@ -1073,6 +1073,7 @@ struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id,
 	queue_flag_set_unlocked(QUEUE_FLAG_BYPASS, q);
 
 	init_waitqueue_head(&q->mq_freeze_wq);
+	mutex_init(&q->mq_freeze_lock);
 
 	/*
 	 * Init percpu_ref in atomic mode so that it's faster to shutdown.
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 15a2d6a4370e2236d908ee97feec4c1c09154e9c..73efd46c2f209b21301697d4e120b047d02ad43a 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -168,13 +168,14 @@ void blk_mq_in_hang_rw(struct request_queue *q, struct hd_struct *part,
 
 void blk_freeze_queue_start(struct request_queue *q)
 {
-	int freeze_depth;
-
-	freeze_depth = atomic_inc_return(&q->mq_freeze_depth);
-	if (freeze_depth == 1) {
+	mutex_lock(&q->mq_freeze_lock);
+	if (++q->mq_freeze_depth == 1) {
 		percpu_ref_kill(&q->q_usage_counter);
+		mutex_unlock(&q->mq_freeze_lock);
 		if (q->mq_ops)
 			blk_mq_run_hw_queues(q, false);
+	} else {
+		mutex_unlock(&q->mq_freeze_lock);
 	}
 }
 EXPORT_SYMBOL_GPL(blk_freeze_queue_start);
@@ -225,14 +226,14 @@ EXPORT_SYMBOL_GPL(blk_mq_freeze_queue);
 
 void blk_mq_unfreeze_queue(struct request_queue *q)
 {
-	int freeze_depth;
-
-	freeze_depth = atomic_dec_return(&q->mq_freeze_depth);
-	WARN_ON_ONCE(freeze_depth < 0);
-	if (!freeze_depth) {
+	mutex_lock(&q->mq_freeze_lock);
+	q->mq_freeze_depth--;
+	WARN_ON_ONCE(q->mq_freeze_depth < 0);
+	if (!q->mq_freeze_depth) {
 		percpu_ref_resurrect(&q->q_usage_counter);
 		wake_up_all(&q->mq_freeze_wq);
 	}
+	mutex_unlock(&q->mq_freeze_lock);
 }
 EXPORT_SYMBOL_GPL(blk_mq_unfreeze_queue);
 
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index f8f4b2598141affed2f31443e49ef75ac72327b1..8f2bd2ed6c24f2d0a5432146a9205191315cfe70 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -653,7 +653,7 @@ struct request_queue {
 	struct mutex		sysfs_lock;
 
 	int			bypass_depth;
-	atomic_t		mq_freeze_depth;
+	int			mq_freeze_depth;
 
 #if defined(CONFIG_BLK_DEV_BSG)
 	bsg_job_fn		*bsg_job_fn;
@@ -666,6 +666,11 @@ struct request_queue {
 #endif
 	struct rcu_head		rcu_head;
 	wait_queue_head_t	mq_freeze_wq;
+	/*
+	 * Protect concurrent access to q_usage_counter by
+	 * percpu_ref_kill() and percpu_ref_reinit().
+	 */
+	struct mutex		mq_freeze_lock;
 	struct percpu_ref	q_usage_counter;
 	struct list_head	all_q_node;