[PATCH 2/2] blk: Fix lock inversion between ioc lock and bfqd lock

@@ -2317,9 +2317,9 @@ static bool bfq_bio_merge(struct request_queue *q, struct bio *bio,
 
 	ret = blk_mq_sched_try_merge(q, bio, nr_segs, &free);
 
+	spin_unlock_irq(&bfqd->lock);
 	if (free)
 		blk_mq_free_request(free);
-	spin_unlock_irq(&bfqd->lock);
 
 	return ret;
 }

@@ -5933,6 +5933,7 @@ static void bfq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
 	spin_lock_irq(&bfqd->lock);
 	if (blk_mq_sched_try_insert_merge(q, rq)) {
 		spin_unlock_irq(&bfqd->lock);
+		blk_put_request(rq);
 		return;
 	}
 

@@ -6376,6 +6377,7 @@ static void bfq_finish_requeue_request(struct request *rq)
 {
 	struct bfq_queue *bfqq = RQ_BFQQ(rq);
 	struct bfq_data *bfqd;
+	unsigned long flags;
 
 	/*
 	 * rq either is not associated with any icq, or is an already

@@ -6393,18 +6395,12 @@ static void bfq_finish_requeue_request(struct request *rq)
 					     rq->io_start_time_ns,
 					     rq->cmd_flags);
 
+	spin_lock_irqsave(&bfqd->lock, flags);
 	if (likely(rq->rq_flags & RQF_STARTED)) {
-		unsigned long flags;
-
-		spin_lock_irqsave(&bfqd->lock, flags);
-
 		if (rq == bfqd->waited_rq)
 			bfq_update_inject_limit(bfqd, bfqq);
 
 		bfq_completed_request(bfqq, bfqd);
-		bfq_finish_requeue_request_body(bfqq);
-
-		spin_unlock_irqrestore(&bfqd->lock, flags);
 	} else {
 		/*
 		 * Request rq may be still/already in the scheduler,

@@ -6414,18 +6410,16 @@ static void bfq_finish_requeue_request(struct request *rq)
 		 * inconsistencies in the time interval from the end
 		 * of this function to the start of the deferred work.
 		 * This situation seems to occur only in process
-		 * context, as a consequence of a merge. In the
-		 * current version of the code, this implies that the
-		 * lock is held.
+		 * context, as a consequence of a merge.
 		 */
-
 		if (!RB_EMPTY_NODE(&rq->rb_node)) {
 			bfq_remove_request(rq->q, rq);
 			bfqg_stats_update_io_remove(bfqq_group(bfqq),
 						    rq->cmd_flags);
 		}
-		bfq_finish_requeue_request_body(bfqq);
 	}
+	bfq_finish_requeue_request_body(bfqq);
+	spin_unlock_irqrestore(&bfqd->lock, flags);
 
 	/*
 	 * Reset private fields. In case of a requeue, this allows

@@ -846,18 +846,15 @@ static struct request *attempt_front_merge(struct request_queue *q,
 	return NULL;
 }
 
-int blk_attempt_req_merge(struct request_queue *q, struct request *rq,
-			  struct request *next)
+/*
+ * Try to merge 'next' into 'rq'. Return true if the merge happened, false
+ * otherwise. The caller is responsible for freeing 'next' if the merge
+ * happened.
+ */
+bool blk_attempt_req_merge(struct request_queue *q, struct request *rq,
+			   struct request *next)
 {
-	struct request *free;
-
-	free = attempt_merge(q, rq, next);
-	if (free) {
-		blk_put_request(free);
-		return 1;
-	}
-
-	return 0;
+	return attempt_merge(q, rq, next);
 }
 
 bool blk_rq_merge_ok(struct request *rq, struct bio *bio)

@@ -225,7 +225,7 @@ ssize_t part_timeout_store(struct device *, struct device_attribute *,
 void __blk_queue_split(struct bio **bio, unsigned int *nr_segs);
 int ll_back_merge_fn(struct request *req, struct bio *bio,
 		unsigned int nr_segs);
-int blk_attempt_req_merge(struct request_queue *q, struct request *rq,
+bool blk_attempt_req_merge(struct request_queue *q, struct request *rq,
 				struct request *next);
 unsigned int blk_recalc_rq_segments(struct request *rq);
 void blk_rq_set_mixed_merge(struct request *rq);

@@ -494,8 +494,10 @@ static void dd_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq,
 	 */
 	blk_req_zone_write_unlock(rq);
 
-	if (blk_mq_sched_try_insert_merge(q, rq))
+	if (blk_mq_sched_try_insert_merge(q, rq)) {
+		blk_put_request(rq);
 		return;
+	}
 
 	trace_block_rq_insert(rq);