deadline-iosched: Introduce zone locking support

Introduce zone write locking to avoid write request reordering with
zoned block devices. This is achieved using a finer selection of the
next request to dispatch:
1) Any non-write request is always allowed to proceed.
2) Any write to a conventional zone is always allowed to proceed.
3) For a write to a sequential zone, the zone lock is first checked.
   a) If the zone is not locked, the write is allowed to proceed after
      its target zone is locked.
   b) If the zone is locked, the write request is skipped and the next
      request in the dispatch queue tested (back to step 1).

For a write request that has locked its target zone, the zone is
unlocked either when the request completes and the method
deadline_request_completed() is called, or when the request is requeued
using the method deadline_add_request().

Requests targeting a locked zone are always left in the scheduler queue
to preserve the initial write order. If no write request can be
dispatched, allow reads to be dispatched even if the write batch is not
done.

If the device used is not a zoned block device, or if zoned block device
support is disabled, this patch does not modify deadline behavior.

Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

1 files changed, 68 insertions, 3 deletions

diff --git a/block/deadline-iosched.c b/block/deadline-iosched.c
index 81e3f0897457..9de9f156e203 100644
--- a/block/deadline-iosched.c
+++ b/block/deadline-iosched.c
@@ -98,6 +98,12 @@ deadline_add_request(struct request_queue *q, struct request *rq)
 	struct deadline_data *dd = q->elevator->elevator_data;
 	const int data_dir = rq_data_dir(rq);
 
+	/*
+	 * This may be a requeue of a write request that has locked its
+	 * target zone. If it is the case, this releases the zone lock.
+	 */
+	blk_req_zone_write_unlock(rq);
+
 	deadline_add_rq_rb(dd, rq);
 
 	/*
@@ -188,6 +194,12 @@ deadline_move_to_dispatch(struct deadline_data *dd, struct request *rq)
 {
 	struct request_queue *q = rq->q;
 
+	/*
+	 * For a zoned block device, write requests must write lock their
+	 * target zone.
+	 */
+	blk_req_zone_write_lock(rq);
+
 	deadline_remove_request(q, rq);
 	elv_dispatch_add_tail(q, rq);
 }
@@ -235,13 +247,28 @@ static inline int deadline_check_fifo(struct deadline_data *dd, int ddir)
 static struct request *
 deadline_fifo_request(struct deadline_data *dd, int data_dir)
 {
+	struct request *rq;
+
 	if (WARN_ON_ONCE(data_dir != READ && data_dir != WRITE))
 		return NULL;
 
 	if (list_empty(&dd->fifo_list[data_dir]))
 		return NULL;
 
-	return rq_entry_fifo(dd->fifo_list[data_dir].next);
+	rq = rq_entry_fifo(dd->fifo_list[data_dir].next);
+	if (data_dir == READ || !blk_queue_is_zoned(rq->q))
+		return rq;
+
+	/*
+	 * Look for a write request that can be dispatched, that is one with
+	 * an unlocked target zone.
+	 */
+	list_for_each_entry(rq, &dd->fifo_list[WRITE], queuelist) {
+		if (blk_req_can_dispatch_to_zone(rq))
+			return rq;
+	}
+
+	return NULL;
 }
 
 /*
@@ -251,10 +278,29 @@ deadline_fifo_request(struct deadline_data *dd, int data_dir)
 static struct request *
 deadline_next_request(struct deadline_data *dd, int data_dir)
 {
+	struct request *rq;
+
 	if (WARN_ON_ONCE(data_dir != READ && data_dir != WRITE))
 		return NULL;
 
-	return dd->next_rq[data_dir];
+	rq = dd->next_rq[data_dir];
+	if (!rq)
+		return NULL;
+
+	if (data_dir == READ || !blk_queue_is_zoned(rq->q))
+		return rq;
+
+	/*
+	 * Look for a write request that can be dispatched, that is one with
+	 * an unlocked target zone.
+	 */
+	while (rq) {
+		if (blk_req_can_dispatch_to_zone(rq))
+			return rq;
+		rq = deadline_latter_request(rq);
+	}
+
+	return NULL;
 }
 
 /*
@@ -288,7 +334,8 @@ static int deadline_dispatch_requests(struct request_queue *q, int force)
 	if (reads) {
 		BUG_ON(RB_EMPTY_ROOT(&dd->sort_list[READ]));
 
-		if (writes && (dd->starved++ >= dd->writes_starved))
+		if (deadline_fifo_request(dd, WRITE) &&
+		    (dd->starved++ >= dd->writes_starved))
 			goto dispatch_writes;
 
 		data_dir = READ;
@@ -333,6 +380,13 @@ dispatch_find_request:
 		rq = next_rq;
 	}
 
+	/*
+	 * For a zoned block device, if we only have writes queued and none of
+	 * them can be dispatched, rq will be NULL.
+	 */
+	if (!rq)
+		return 0;
+
 	dd->batching = 0;
 
 dispatch_request:
@@ -345,6 +399,16 @@ dispatch_request:
 	return 1;
 }
 
+/*
+ * For zoned block devices, write unlock the target zone of completed
+ * write requests.
+ */
+static void
+deadline_completed_request(struct request_queue *q, struct request *rq)
+{
+	blk_req_zone_write_unlock(rq);
+}
+
 static void deadline_exit_queue(struct elevator_queue *e)
 {
 	struct deadline_data *dd = e->elevator_data;
@@ -466,6 +530,7 @@ static struct elevator_type iosched_deadline = {
 		.elevator_merged_fn =		deadline_merged_request,
 		.elevator_merge_req_fn =	deadline_merged_requests,
 		.elevator_dispatch_fn =		deadline_dispatch_requests,
+		.elevator_completed_req_fn =	deadline_completed_request,
 		.elevator_add_req_fn =		deadline_add_request,
 		.elevator_former_req_fn =	elv_rb_former_request,
 		.elevator_latter_req_fn =	elv_rb_latter_request,