[PATCH V3 3/4] blk-mq: add tagset quiesce interface

[Ming Lei]

drivers that have shared tagsets may need to quiesce potentially a lot
of request queues that all share a single tagset (e.g. nvme). Add an interface
to quiesce all the queues on a given tagset. This interface is useful because
it can speedup the quiesce by doing it in parallel.

For tagsets that have BLK_MQ_F_BLOCKING set, we kill request queue's dispatch
percpu-refcount such that all of them wait for the counter becoming zero. For
tagsets that don't have BLK_MQ_F_BLOCKING set, we simply call a single
synchronize_rcu as this is sufficient.

This patch is against Sagi's original post.
Signed-off-by: Ming Lei <ming.lei at redhat.com>
Signed-off-by: Sagi <ming.lei at redhat.com>
Cc: Sagi Grimberg <sagi at grimberg.me>
Cc: Bart Van Assche <bvanassche at acm.org>
Cc: Johannes Thumshirn <Johannes.Thumshirn at wdc.com>
Cc: Chao Leng <lengchao at huawei.com>
---
 block/blk-mq.c         | 68 ++++++++++++++++++++++++++++++++----------
 include/linux/blk-mq.h |  2 ++
 2 files changed, 55 insertions(+), 15 deletions(-)

diff --git a/block/blk-mq.c b/block/blk-mq.c
index 60630a720449..49933d7dceba 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -209,16 +209,7 @@ void blk_mq_quiesce_queue_nowait(struct request_queue *q)
 }
 EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue_nowait);
 
-/**
- * blk_mq_quiesce_queue() - wait until all ongoing dispatches have finished
- * @q: request queue.
- *
- * Note: this function does not prevent that the struct request end_io()
- * callback function is invoked. Once this function is returned, we make
- * sure no dispatch can happen until the queue is unquiesced via
- * blk_mq_unquiesce_queue().
- */
-void blk_mq_quiesce_queue(struct request_queue *q)
+static void __blk_mq_quiesce_queue(struct request_queue *q, bool wait)
 {
 	bool blocking = !!(q->tag_set->flags & BLK_MQ_F_BLOCKING);
 
@@ -230,14 +221,30 @@ void blk_mq_quiesce_queue(struct request_queue *q)
 			percpu_ref_kill(&q->dispatch_counter);
 	}
 
-	if (blocking)
-		wait_event(q->mq_quiesce_wq,
-			   percpu_ref_is_zero(&q->dispatch_counter));
-	else
-		synchronize_rcu();
+	if (wait) {
+		if (blocking)
+			wait_event(q->mq_quiesce_wq,
+				   percpu_ref_is_zero(&q->dispatch_counter));
+		else
+			synchronize_rcu();
+	}
 
 	mutex_unlock(&q->mq_quiesce_lock);
 }
+
+/**
+ * blk_mq_quiesce_queue() - wait until all ongoing dispatches have finished
+ * @q: request queue.
+ *
+ * Note: this function does not prevent that the struct request end_io()
+ * callback function is invoked. Once this function is returned, we make
+ * sure no dispatch can happen until the queue is unquiesced via
+ * blk_mq_unquiesce_queue().
+ */
+void blk_mq_quiesce_queue(struct request_queue *q)
+{
+	__blk_mq_quiesce_queue(q, true);
+}
 EXPORT_SYMBOL_GPL(blk_mq_quiesce_queue);
 
 /*
@@ -265,6 +272,37 @@ void blk_mq_unquiesce_queue(struct request_queue *q)
 }
 EXPORT_SYMBOL_GPL(blk_mq_unquiesce_queue);
 
+void blk_mq_quiesce_tagset(struct blk_mq_tag_set *set)
+{
+	struct request_queue *q;
+
+	mutex_lock(&set->tag_list_lock);
+	list_for_each_entry(q, &set->tag_list, tag_set_list)
+		__blk_mq_quiesce_queue(q, false);
+
+	/* wait until all queues' quiesce is done */
+	if (set->flags & BLK_MQ_F_BLOCKING) {
+		list_for_each_entry(q, &set->tag_list, tag_set_list)
+			wait_event(q->mq_quiesce_wq,
+				   percpu_ref_is_zero(&q->dispatch_counter));
+	} else {
+		synchronize_rcu();
+	}
+	mutex_unlock(&set->tag_list_lock);
+}
+EXPORT_SYMBOL_GPL(blk_mq_quiesce_tagset);
+
+void blk_mq_unquiesce_tagset(struct blk_mq_tag_set *set)
+{
+	struct request_queue *q;
+
+	mutex_lock(&set->tag_list_lock);
+	list_for_each_entry(q, &set->tag_list, tag_set_list)
+		blk_mq_unquiesce_queue(q);
+	mutex_unlock(&set->tag_list_lock);
+}
+EXPORT_SYMBOL_GPL(blk_mq_unquiesce_tagset);
+
 void blk_mq_wake_waiters(struct request_queue *q)
 {
 	struct blk_mq_hw_ctx *hctx;
diff --git a/include/linux/blk-mq.h b/include/linux/blk-mq.h
index df642055f02c..90da3582b91d 100644
--- a/include/linux/blk-mq.h
+++ b/include/linux/blk-mq.h
@@ -519,6 +519,8 @@ int blk_mq_map_queues(struct blk_mq_queue_map *qmap);
 void blk_mq_update_nr_hw_queues(struct blk_mq_tag_set *set, int nr_hw_queues);
 
 void blk_mq_quiesce_queue_nowait(struct request_queue *q);
+void blk_mq_quiesce_tagset(struct blk_mq_tag_set *set);
+void blk_mq_unquiesce_tagset(struct blk_mq_tag_set *set);
 
 unsigned int blk_mq_rq_cpu(struct request *rq);
 
-- 
2.25.2