block: split the blk-mq case from elevator_init

There is almost no shared logic, which leads to a very confusing code
flow.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Damien Le Moal <damien.lemoal@wdc.com>
Tested-by: Damien Le Moal <damien.lemoal@wdc.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

3 files changed, 48 insertions, 32 deletions

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 858d6edff4d3..6332940ca118 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -2573,7 +2573,7 @@ struct request_queue *blk_mq_init_allocated_queue(struct blk_mq_tag_set *set,
 	if (!(set->flags & BLK_MQ_F_NO_SCHED)) {
 		int ret;
 
-		ret = elevator_init(q);
+		ret = elevator_init_mq(q);
 		if (ret)
 			return ERR_PTR(ret);
 	}
diff --git a/block/blk.h b/block/blk.h
index ab5ec2dadc7b..8d23aea96ce9 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -232,6 +232,7 @@ static inline void elv_deactivate_rq(struct request_queue *q, struct request *rq
 }
 
 int elevator_init(struct request_queue *);
+int elevator_init_mq(struct request_queue *q);
 void elevator_exit(struct request_queue *, struct elevator_queue *);
 int elv_register_queue(struct request_queue *q);
 void elv_unregister_queue(struct request_queue *q);
diff --git a/block/elevator.c b/block/elevator.c
index d6480e70816e..fa828b5bfd4b 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -199,6 +199,11 @@ static void elevator_release(struct kobject *kobj)
 	kfree(e);
 }
 
+/*
+ * Use the default elevator specified by config boot param for non-mq devices,
+ * or by config option.  Don't try to load modules as we could be running off
+ * async and request_module() isn't allowed from async.
+ */
 int elevator_init(struct request_queue *q)
 {
 	struct elevator_type *e = NULL;
@@ -212,46 +217,22 @@ int elevator_init(struct request_queue *q)
 	if (unlikely(q->elevator))
 		goto out_unlock;
 
-	/*
-	 * Use the default elevator specified by config boot param for
-	 * non-mq devices, or by config option. Don't try to load modules
-	 * as we could be running off async and request_module() isn't
-	 * allowed from async.
-	 */
-	if (!q->mq_ops && *chosen_elevator) {
+	if (*chosen_elevator) {
 		e = elevator_get(q, chosen_elevator, false);
 		if (!e)
 			printk(KERN_ERR "I/O scheduler %s not found\n",
 							chosen_elevator);
 	}
 
+	if (!e)
+		e = elevator_get(q, CONFIG_DEFAULT_IOSCHED, false);
 	if (!e) {
-		/*
-		 * For blk-mq devices, we default to using mq-deadline,
-		 * if available, for single queue devices. If deadline
-		 * isn't available OR we have multiple queues, default
-		 * to "none".
-		 */
-		if (q->mq_ops) {
-			if (q->nr_hw_queues == 1)
-				e = elevator_get(q, "mq-deadline", false);
-			if (!e)
-				goto out_unlock;
-		} else
-			e = elevator_get(q, CONFIG_DEFAULT_IOSCHED, false);
-
-		if (!e) {
-			printk(KERN_ERR
-				"Default I/O scheduler not found. " \
-				"Using noop.\n");
-			e = elevator_get(q, "noop", false);
-		}
+		printk(KERN_ERR
+			"Default I/O scheduler not found. Using noop.\n");
+		e = elevator_get(q, "noop", false);
 	}
 
-	if (e->uses_mq)
-		err = blk_mq_init_sched(q, e);
-	else
-		err = e->ops.sq.elevator_init_fn(q, e);
+	err = e->ops.sq.elevator_init_fn(q, e);
 	if (err)
 		elevator_put(e);
 out_unlock:
@@ -993,6 +974,40 @@ out:
 }
 
 /*
+ * For blk-mq devices, we default to using mq-deadline, if available, for single
+ * queue devices.  If deadline isn't available OR we have multiple queues,
+ * default to "none".
+ */
+int elevator_init_mq(struct request_queue *q)
+{
+	struct elevator_type *e;
+	int err = 0;
+
+	if (q->nr_hw_queues != 1)
+		return 0;
+
+	/*
+	 * q->sysfs_lock must be held to provide mutual exclusion between
+	 * elevator_switch() and here.
+	 */
+	mutex_lock(&q->sysfs_lock);
+	if (unlikely(q->elevator))
+		goto out_unlock;
+
+	e = elevator_get(q, "mq-deadline", false);
+	if (!e)
+		goto out_unlock;
+
+	err = blk_mq_init_sched(q, e);
+	if (err)
+		elevator_put(e);
+out_unlock:
+	mutex_unlock(&q->sysfs_lock);
+	return err;
+}
+
+
+/*
  * switch to new_e io scheduler. be careful not to introduce deadlocks -
  * we don't free the old io scheduler, before we have allocated what we
  * need for the new one. this way we have a chance of going back to the old