1 files changed, 32 insertions, 16 deletions

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 820d131a6893..a9c181603cbd 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -362,7 +362,7 @@ static struct request *blk_mq_get_request(struct request_queue *q,
 	}
 	if (likely(!data->hctx))
 		data->hctx = blk_mq_map_queue(q, data->cmd_flags,
-						data->ctx->cpu);
+						data->ctx);
 	if (data->cmd_flags & REQ_NOWAIT)
 		data->flags |= BLK_MQ_REQ_NOWAIT;
 
@@ -733,12 +733,20 @@ static void blk_mq_requeue_work(struct work_struct *work)
 	spin_unlock_irq(&q->requeue_lock);
 
 	list_for_each_entry_safe(rq, next, &rq_list, queuelist) {
-		if (!(rq->rq_flags & RQF_SOFTBARRIER))
+		if (!(rq->rq_flags & (RQF_SOFTBARRIER | RQF_DONTPREP)))
 			continue;
 
 		rq->rq_flags &= ~RQF_SOFTBARRIER;
 		list_del_init(&rq->queuelist);
-		blk_mq_sched_insert_request(rq, true, false, false);
+		/*
+		 * If RQF_DONTPREP, rq has contained some driver specific
+		 * data, so insert it to hctx dispatch list to avoid any
+		 * merge.
+		 */
+		if (rq->rq_flags & RQF_DONTPREP)
+			blk_mq_request_bypass_insert(rq, false);
+		else
+			blk_mq_sched_insert_request(rq, true, false, false);
 	}
 
 	while (!list_empty(&rq_list)) {
@@ -1902,7 +1910,7 @@ static blk_qc_t blk_mq_make_request(struct request_queue *q, struct bio *bio)
 {
 	const int is_sync = op_is_sync(bio->bi_opf);
 	const int is_flush_fua = op_is_flush(bio->bi_opf);
-	struct blk_mq_alloc_data data = { .flags = 0, .cmd_flags = bio->bi_opf };
+	struct blk_mq_alloc_data data = { .flags = 0};
 	struct request *rq;
 	struct blk_plug *plug;
 	struct request *same_queue_rq = NULL;
@@ -1924,6 +1932,7 @@ static blk_qc_t blk_mq_make_request(struct request_queue *q, struct bio *bio)
 
 	rq_qos_throttle(q, bio);
 
+	data.cmd_flags = bio->bi_opf;
 	rq = blk_mq_get_request(q, bio, &data);
 	if (unlikely(!rq)) {
 		rq_qos_cleanup(q, bio);
@@ -2056,7 +2065,7 @@ struct blk_mq_tags *blk_mq_alloc_rq_map(struct blk_mq_tag_set *set,
 	struct blk_mq_tags *tags;
 	int node;
 
-	node = blk_mq_hw_queue_to_node(&set->map[0], hctx_idx);
+	node = blk_mq_hw_queue_to_node(&set->map[HCTX_TYPE_DEFAULT], hctx_idx);
 	if (node == NUMA_NO_NODE)
 		node = set->numa_node;
 
@@ -2112,7 +2121,7 @@ int blk_mq_alloc_rqs(struct blk_mq_tag_set *set, struct blk_mq_tags *tags,
 	size_t rq_size, left;
 	int node;
 
-	node = blk_mq_hw_queue_to_node(&set->map[0], hctx_idx);
+	node = blk_mq_hw_queue_to_node(&set->map[HCTX_TYPE_DEFAULT], hctx_idx);
 	if (node == NUMA_NO_NODE)
 		node = set->numa_node;
 
@@ -2411,7 +2420,7 @@ static void blk_mq_map_swqueue(struct request_queue *q)
 	 * If the cpu isn't present, the cpu is mapped to first hctx.
 	 */
 	for_each_possible_cpu(i) {
-		hctx_idx = set->map[0].mq_map[i];
+		hctx_idx = set->map[HCTX_TYPE_DEFAULT].mq_map[i];
 		/* unmapped hw queue can be remapped after CPU topo changed */
 		if (!set->tags[hctx_idx] &&
 		    !__blk_mq_alloc_rq_map(set, hctx_idx)) {
@@ -2421,16 +2430,19 @@ static void blk_mq_map_swqueue(struct request_queue *q)
 			 * case, remap the current ctx to hctx[0] which
 			 * is guaranteed to always have tags allocated
 			 */
-			set->map[0].mq_map[i] = 0;
+			set->map[HCTX_TYPE_DEFAULT].mq_map[i] = 0;
 		}
 
 		ctx = per_cpu_ptr(q->queue_ctx, i);
 		for (j = 0; j < set->nr_maps; j++) {
-			if (!set->map[j].nr_queues)
+			if (!set->map[j].nr_queues) {
+				ctx->hctxs[j] = blk_mq_map_queue_type(q,
+						HCTX_TYPE_DEFAULT, i);
 				continue;
+			}
 
 			hctx = blk_mq_map_queue_type(q, j, i);
-
+			ctx->hctxs[j] = hctx;
 			/*
 			 * If the CPU is already set in the mask, then we've
 			 * mapped this one already. This can happen if
@@ -2450,6 +2462,10 @@ static void blk_mq_map_swqueue(struct request_queue *q)
 			 */
 			BUG_ON(!hctx->nr_ctx);
 		}
+
+		for (; j < HCTX_MAX_TYPES; j++)
+			ctx->hctxs[j] = blk_mq_map_queue_type(q,
+					HCTX_TYPE_DEFAULT, i);
 	}
 
 	mutex_unlock(&q->sysfs_lock);
@@ -2721,7 +2737,7 @@ static void blk_mq_realloc_hw_ctxs(struct blk_mq_tag_set *set,
 		int node;
 		struct blk_mq_hw_ctx *hctx;
 
-		node = blk_mq_hw_queue_to_node(&set->map[0], i);
+		node = blk_mq_hw_queue_to_node(&set->map[HCTX_TYPE_DEFAULT], i);
 		/*
 		 * If the hw queue has been mapped to another numa node,
 		 * we need to realloc the hctx. If allocation fails, fallback
@@ -2825,9 +2841,6 @@ struct request_queue *blk_mq_init_allocated_queue(struct blk_mq_tag_set *set,
 	    set->map[HCTX_TYPE_POLL].nr_queues)
 		blk_queue_flag_set(QUEUE_FLAG_POLL, q);
 
-	if (!(set->flags & BLK_MQ_F_SG_MERGE))
-		blk_queue_flag_set(QUEUE_FLAG_NO_SG_MERGE, q);
-
 	q->sg_reserved_size = INT_MAX;
 
 	INIT_DELAYED_WORK(&q->requeue_work, blk_mq_requeue_work);
@@ -2955,7 +2968,7 @@ static int blk_mq_update_queue_map(struct blk_mq_tag_set *set)
 		return set->ops->map_queues(set);
 	} else {
 		BUG_ON(set->nr_maps > 1);
-		return blk_mq_map_queues(&set->map[0]);
+		return blk_mq_map_queues(&set->map[HCTX_TYPE_DEFAULT]);
 	}
 }
 
@@ -3077,6 +3090,9 @@ int blk_mq_update_nr_requests(struct request_queue *q, unsigned int nr)
 	if (!set)
 		return -EINVAL;
 
+	if (q->nr_requests == nr)
+		return 0;
+
 	blk_mq_freeze_queue(q);
 	blk_mq_quiesce_queue(q);
 
@@ -3222,7 +3238,7 @@ fallback:
 			pr_warn("Increasing nr_hw_queues to %d fails, fallback to %d\n",
 					nr_hw_queues, prev_nr_hw_queues);
 			set->nr_hw_queues = prev_nr_hw_queues;
-			blk_mq_map_queues(&set->map[0]);
+			blk_mq_map_queues(&set->map[HCTX_TYPE_DEFAULT]);
 			goto fallback;
 		}
 		blk_mq_map_swqueue(q);