1 files changed, 84 insertions, 46 deletions

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 100fb0c3114f..7ed6b9469f97 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -747,11 +747,16 @@ static void req_bio_endio(struct request *rq, struct bio *bio,
 		/*
 		 * Partial zone append completions cannot be supported as the
 		 * BIO fragments may end up not being written sequentially.
+		 * For such case, force the completed nbytes to be equal to
+		 * the BIO size so that bio_advance() sets the BIO remaining
+		 * size to 0 and we end up calling bio_endio() before returning.
 		 */
-		if (bio->bi_iter.bi_size != nbytes)
+		if (bio->bi_iter.bi_size != nbytes) {
 			bio->bi_status = BLK_STS_IOERR;
-		else
+			nbytes = bio->bi_iter.bi_size;
+		} else {
 			bio->bi_iter.bi_sector = rq->__sector;
+		}
 	}
 
 	bio_advance(bio, nbytes);
@@ -1500,14 +1505,26 @@ void blk_mq_delay_kick_requeue_list(struct request_queue *q,
 }
 EXPORT_SYMBOL(blk_mq_delay_kick_requeue_list);
 
+static bool blk_is_flush_data_rq(struct request *rq)
+{
+	return (rq->rq_flags & RQF_FLUSH_SEQ) && !is_flush_rq(rq);
+}
+
 static bool blk_mq_rq_inflight(struct request *rq, void *priv)
 {
 	/*
 	 * If we find a request that isn't idle we know the queue is busy
 	 * as it's checked in the iter.
 	 * Return false to stop the iteration.
+	 *
+	 * In case of queue quiesce, if one flush data request is completed,
+	 * don't count it as inflight given the flush sequence is suspended,
+	 * and the original flush data request is invisible to driver, just
+	 * like other pending requests because of quiesce
 	 */
-	if (blk_mq_request_started(rq)) {
+	if (blk_mq_request_started(rq) && !(blk_queue_quiesced(rq->q) &&
+				blk_is_flush_data_rq(rq) &&
+				blk_mq_request_completed(rq))) {
 		bool *busy = priv;
 
 		*busy = true;
@@ -1848,6 +1865,22 @@ static bool blk_mq_mark_tag_wait(struct blk_mq_hw_ctx *hctx,
 	__add_wait_queue(wq, wait);
 
 	/*
+	 * Add one explicit barrier since blk_mq_get_driver_tag() may
+	 * not imply barrier in case of failure.
+	 *
+	 * Order adding us to wait queue and allocating driver tag.
+	 *
+	 * The pair is the one implied in sbitmap_queue_wake_up() which
+	 * orders clearing sbitmap tag bits and waitqueue_active() in
+	 * __sbitmap_queue_wake_up(), since waitqueue_active() is lockless
+	 *
+	 * Otherwise, re-order of adding wait queue and getting driver tag
+	 * may cause __sbitmap_queue_wake_up() to wake up nothing because
+	 * the waitqueue_active() may not observe us in wait queue.
+	 */
+	smp_mb();
+
+	/*
 	 * It's possible that a tag was freed in the window between the
 	 * allocation failure and adding the hardware queue to the wait
 	 * queue.
@@ -2855,11 +2888,8 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
 	};
 	struct request *rq;
 
-	if (unlikely(bio_queue_enter(bio)))
-		return NULL;
-
 	if (blk_mq_attempt_bio_merge(q, bio, nsegs))
-		goto queue_exit;
+		return NULL;
 
 	rq_qos_throttle(q, bio);
 
@@ -2875,35 +2905,23 @@ static struct request *blk_mq_get_new_requests(struct request_queue *q,
 	rq_qos_cleanup(q, bio);
 	if (bio->bi_opf & REQ_NOWAIT)
 		bio_wouldblock_error(bio);
-queue_exit:
-	blk_queue_exit(q);
 	return NULL;
 }
 
-static inline struct request *blk_mq_get_cached_request(struct request_queue *q,
-		struct blk_plug *plug, struct bio **bio, unsigned int nsegs)
+/* return true if this @rq can be used for @bio */
+static bool blk_mq_can_use_cached_rq(struct request *rq, struct blk_plug *plug,
+		struct bio *bio)
 {
-	struct request *rq;
-	enum hctx_type type, hctx_type;
-
-	if (!plug)
-		return NULL;
-	rq = rq_list_peek(&plug->cached_rq);
-	if (!rq || rq->q != q)
-		return NULL;
+	enum hctx_type type = blk_mq_get_hctx_type(bio->bi_opf);
+	enum hctx_type hctx_type = rq->mq_hctx->type;
 
-	if (blk_mq_attempt_bio_merge(q, *bio, nsegs)) {
-		*bio = NULL;
-		return NULL;
-	}
+	WARN_ON_ONCE(rq_list_peek(&plug->cached_rq) != rq);
 
-	type = blk_mq_get_hctx_type((*bio)->bi_opf);
-	hctx_type = rq->mq_hctx->type;
 	if (type != hctx_type &&
 	    !(type == HCTX_TYPE_READ && hctx_type == HCTX_TYPE_DEFAULT))
-		return NULL;
-	if (op_is_flush(rq->cmd_flags) != op_is_flush((*bio)->bi_opf))
-		return NULL;
+		return false;
+	if (op_is_flush(rq->cmd_flags) != op_is_flush(bio->bi_opf))
+		return false;
 
 	/*
 	 * If any qos ->throttle() end up blocking, we will have flushed the
@@ -2911,11 +2929,11 @@ static inline struct request *blk_mq_get_cached_request(struct request_queue *q,
 	 * before we throttle.
 	 */
 	plug->cached_rq = rq_list_next(rq);
-	rq_qos_throttle(q, *bio);
+	rq_qos_throttle(rq->q, bio);
 
-	rq->cmd_flags = (*bio)->bi_opf;
+	rq->cmd_flags = bio->bi_opf;
 	INIT_LIST_HEAD(&rq->queuelist);
-	return rq;
+	return true;
 }
 
 static void bio_set_ioprio(struct bio *bio)
@@ -2944,31 +2962,51 @@ void blk_mq_submit_bio(struct bio *bio)
 	struct request_queue *q = bdev_get_queue(bio->bi_bdev);
 	struct blk_plug *plug = blk_mq_plug(bio);
 	const int is_sync = op_is_sync(bio->bi_opf);
-	struct request *rq;
+	struct request *rq = NULL;
 	unsigned int nr_segs = 1;
 	blk_status_t ret;
 
 	bio = blk_queue_bounce(bio, q);
-	if (bio_may_exceed_limits(bio, &q->limits)) {
-		bio = __bio_split_to_limits(bio, &q->limits, &nr_segs);
-		if (!bio)
-			return;
-	}
-
-	if (!bio_integrity_prep(bio))
-		return;
-
 	bio_set_ioprio(bio);
 
-	rq = blk_mq_get_cached_request(q, plug, &bio, nr_segs);
-	if (!rq) {
-		if (!bio)
+	if (plug) {
+		rq = rq_list_peek(&plug->cached_rq);
+		if (rq && rq->q != q)
+			rq = NULL;
+	}
+	if (rq) {
+		if (unlikely(bio_may_exceed_limits(bio, &q->limits))) {
+			bio = __bio_split_to_limits(bio, &q->limits, &nr_segs);
+			if (!bio)
+				return;
+		}
+		if (!bio_integrity_prep(bio))
 			return;
-		rq = blk_mq_get_new_requests(q, plug, bio, nr_segs);
-		if (unlikely(!rq))
+		if (blk_mq_attempt_bio_merge(q, bio, nr_segs))
 			return;
+		if (blk_mq_can_use_cached_rq(rq, plug, bio))
+			goto done;
+		percpu_ref_get(&q->q_usage_counter);
+	} else {
+		if (unlikely(bio_queue_enter(bio)))
+			return;
+		if (unlikely(bio_may_exceed_limits(bio, &q->limits))) {
+			bio = __bio_split_to_limits(bio, &q->limits, &nr_segs);
+			if (!bio)
+				goto fail;
+		}
+		if (!bio_integrity_prep(bio))
+			goto fail;
+	}
+
+	rq = blk_mq_get_new_requests(q, plug, bio, nr_segs);
+	if (unlikely(!rq)) {
+fail:
+		blk_queue_exit(q);
+		return;
 	}
 
+done:
 	trace_block_getrq(bio);
 
 	rq_qos_track(q, rq, bio);