dm rq: do not update rq partially in each ending bio

We don't need to update the original dm request partially when ending
each cloned bio: just update original dm request once when the whole
cloned request is finished.  This still allows full support for partial
completion because a new 'completed' counter accounts for incremental
progress as the clone bios complete.

Partial request update can be a bit expensive, so we should try to avoid
it, especially because it is run in softirq context.

Avoiding all the partial request updates fixes both hard lockup and
soft lockups that were easily reproduced while running Laurence's
test[1] on IB/SRP.

BTW, after d4acf3650c7c ("block: Make blk_mq_delay_kick_requeue_list()
rerun the queue at a quiet time"), we need to make the test more
aggressive for reproducing the lockup:

	1) run hammer_write.sh 32 or 64 concurrently.
	2) write 8M each time

[1] https://marc.info/?l=linux-block&m=150220185510245&w=2

Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>

2 files changed, 8 insertions, 11 deletions

diff --git a/drivers/md/dm-rq.c b/drivers/md/dm-rq.c
index cbee09054d1e..eadfcfd106ff 100644
--- a/drivers/md/dm-rq.c
+++ b/drivers/md/dm-rq.c
@@ -117,9 +117,9 @@ static void end_clone_bio(struct bio *clone)
 	struct dm_rq_clone_bio_info *info =
 		container_of(clone, struct dm_rq_clone_bio_info, clone);
 	struct dm_rq_target_io *tio = info->tio;
-	struct bio *bio = info->orig;
 	unsigned int nr_bytes = info->orig->bi_iter.bi_size;
 	blk_status_t error = clone->bi_status;
+	bool is_last = !clone->bi_next;
 
 	bio_put(clone);
 
@@ -137,28 +137,23 @@ static void end_clone_bio(struct bio *clone)
 		 * when the request is completed.
 		 */
 		tio->error = error;
-		return;
+		goto exit;
 	}
 
 	/*
 	 * I/O for the bio successfully completed.
 	 * Notice the data completion to the upper layer.
 	 */
-
-	/*
-	 * bios are processed from the head of the list.
-	 * So the completing bio should always be rq->bio.
-	 * If it's not, something wrong is happening.
-	 */
-	if (tio->orig->bio != bio)
-		DMERR("bio completion is going in the middle of the request");
+	tio->completed += nr_bytes;
 
 	/*
 	 * Update the original request.
 	 * Do not use blk_end_request() here, because it may complete
 	 * the original request before the clone, and break the ordering.
 	 */
-	blk_update_request(tio->orig, BLK_STS_OK, nr_bytes);
+	if (is_last)
+ exit:
+		blk_update_request(tio->orig, BLK_STS_OK, tio->completed);
 }
 
 static struct dm_rq_target_io *tio_from_request(struct request *rq)
@@ -456,6 +451,7 @@ static void init_tio(struct dm_rq_target_io *tio, struct request *rq,
 	tio->clone = NULL;
 	tio->orig = rq;
 	tio->error = 0;
+	tio->completed = 0;
 	/*
 	 * Avoid initializing info for blk-mq; it passes
 	 * target-specific data through info.ptr
diff --git a/drivers/md/dm-rq.h b/drivers/md/dm-rq.h
index 9813922e4fe5..f43c45460aac 100644
--- a/drivers/md/dm-rq.h
+++ b/drivers/md/dm-rq.h
@@ -29,6 +29,7 @@ struct dm_rq_target_io {
 	struct dm_stats_aux stats_aux;
 	unsigned long duration_jiffies;
 	unsigned n_sectors;
+	unsigned completed;
 };
 
 /*