[PATCH 4.9 41/44] dm: flush queued bios when process blocks to avoid deadlock

[Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>]

4.9-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Mikulas Patocka <mpatocka@xxxxxxxxxx>

commit d67a5f4b5947aba4bfe9a80a2b86079c215ca755 upstream.

Commit df2cb6daa4 ("block: Avoid deadlocks with bio allocation by
stacking drivers") created a workqueue for every bio set and code
in bio_alloc_bioset() that tries to resolve some low-memory deadlocks
by redirecting bios queued on current->bio_list to the workqueue if the
system is low on memory.  However other deadlocks (see below **) may
happen, without any low memory condition, because generic_make_request
is queuing bios to current->bio_list (rather than submitting them).

** the related dm-snapshot deadlock is detailed here:
https://www.redhat.com/archives/dm-devel/2016-July/msg00065.html

Fix this deadlock by redirecting any bios on current->bio_list to the
bio_set's rescue workqueue on every schedule() call.  Consequently,
when the process blocks on a mutex, the bios queued on
current->bio_list are dispatched to independent workqueus and they can
complete without waiting for the mutex to be available.

The structure blk_plug contains an entry cb_list and this list can contain
arbitrary callback functions that are called when the process blocks.
To implement this fix DM (ab)uses the onstack plug's cb_list interface
to get its flush_current_bio_list() called at schedule() time.

This fixes the snapshot deadlock - if the map method blocks,
flush_current_bio_list() will be called and it redirects bios waiting
on current->bio_list to appropriate workqueues.

Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=1267650
Depends-on: df2cb6daa4 ("block: Avoid deadlocks with bio allocation by stacking drivers")
Signed-off-by: Mikulas Patocka <mpatocka@xxxxxxxxxx>
Signed-off-by: Mike Snitzer <snitzer@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>


---
 drivers/md/dm.c |   55 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 55 insertions(+)

--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -972,10 +972,61 @@ void dm_accept_partial_bio(struct bio *b
 }
 EXPORT_SYMBOL_GPL(dm_accept_partial_bio);
 
+/*
+ * Flush current->bio_list when the target map method blocks.
+ * This fixes deadlocks in snapshot and possibly in other targets.
+ */
+struct dm_offload {
+	struct blk_plug plug;
+	struct blk_plug_cb cb;
+};
+
+static void flush_current_bio_list(struct blk_plug_cb *cb, bool from_schedule)
+{
+	struct dm_offload *o = container_of(cb, struct dm_offload, cb);
+	struct bio_list list;
+	struct bio *bio;
+
+	INIT_LIST_HEAD(&o->cb.list);
+
+	if (unlikely(!current->bio_list))
+		return;
+
+	list = *current->bio_list;
+	bio_list_init(current->bio_list);
+
+	while ((bio = bio_list_pop(&list))) {
+		struct bio_set *bs = bio->bi_pool;
+		if (unlikely(!bs) || bs == fs_bio_set) {
+			bio_list_add(current->bio_list, bio);
+			continue;
+		}
+
+		spin_lock(&bs->rescue_lock);
+		bio_list_add(&bs->rescue_list, bio);
+		queue_work(bs->rescue_workqueue, &bs->rescue_work);
+		spin_unlock(&bs->rescue_lock);
+	}
+}
+
+static void dm_offload_start(struct dm_offload *o)
+{
+	blk_start_plug(&o->plug);
+	o->cb.callback = flush_current_bio_list;
+	list_add(&o->cb.list, &current->plug->cb_list);
+}
+
+static void dm_offload_end(struct dm_offload *o)
+{
+	list_del(&o->cb.list);
+	blk_finish_plug(&o->plug);
+}
+
 static void __map_bio(struct dm_target_io *tio)
 {
 	int r;
 	sector_t sector;
+	struct dm_offload o;
 	struct bio *clone = &tio->clone;
 	struct dm_target *ti = tio->ti;
 
@@ -988,7 +1039,11 @@ static void __map_bio(struct dm_target_i
 	 */
 	atomic_inc(&tio->io->io_count);
 	sector = clone->bi_iter.bi_sector;
+
+	dm_offload_start(&o);
 	r = ti->type->map(ti, clone);
+	dm_offload_end(&o);
+
 	if (r == DM_MAPIO_REMAPPED) {
 		/* the bio has been remapped so dispatch it */