After commit
396eaf21ee17 ("blk-mq: improve DM's blk-mq IO merging via
blk_insert_cloned_request feedback"), map_request() will requeue the tio
when issued clone request return BLK_STS_RESOURCE or BLK_STS_DEV_RESOURCE.
Thus, if device driver status is error, a tio may be requeued multiple
times until the return value is not DM_MAPIO_REQUEUE. That means
type->start_io may be called multiple times, while type->end_io is only
called when IO complete.
In fact, even without commit
396eaf21ee17, setup_clone() failure can
also cause tio requeue and associated missed call to type->end_io.
The service-time path selector selects path based on in_flight_size,
which is increased by st_start_io() and decreased by st_end_io().
Missed calls to st_end_io() can lead to in_flight_size count error and
will cause the selector to make the wrong choice. In addition,
queue-length path selector will also be affected.
To fix the problem, call type->end_io in ->release_clone_rq before tio
requeue. map_info is passed to ->release_clone_rq() for map_request()
error path that result in requeue.
Fixes:
396eaf21ee17 ("blk-mq: improve DM's blk-mq IO merging via blk_insert_cloned_request feedback")
Cc: stable@vger.kernl.org
Signed-off-by: Yufen Yu <yuyufen@huawei.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
return DM_MAPIO_REMAPPED;
}
-static void multipath_release_clone(struct request *clone)
+static void multipath_release_clone(struct request *clone,
+ union map_info *map_context)
{
+ if (unlikely(map_context)) {
+ /*
+ * non-NULL map_context means caller is still map
+ * method; must undo multipath_clone_and_map()
+ */
+ struct dm_mpath_io *mpio = get_mpio(map_context);
+ struct pgpath *pgpath = mpio->pgpath;
+
+ if (pgpath && pgpath->pg->ps.type->end_io)
+ pgpath->pg->ps.type->end_io(&pgpath->pg->ps,
+ &pgpath->path,
+ mpio->nr_bytes);
+ }
+
blk_put_request(clone);
}
struct request *rq = tio->orig;
blk_rq_unprep_clone(clone);
- tio->ti->type->release_clone_rq(clone);
+ tio->ti->type->release_clone_rq(clone, NULL);
rq_end_stats(md, rq);
blk_mq_end_request(rq, error);
rq_end_stats(md, rq);
if (tio->clone) {
blk_rq_unprep_clone(tio->clone);
- tio->ti->type->release_clone_rq(tio->clone);
+ tio->ti->type->release_clone_rq(tio->clone, NULL);
}
dm_mq_delay_requeue_request(rq, delay_ms);
case DM_MAPIO_REMAPPED:
if (setup_clone(clone, rq, tio, GFP_ATOMIC)) {
/* -ENOMEM */
- ti->type->release_clone_rq(clone);
+ ti->type->release_clone_rq(clone, &tio->info);
return DM_MAPIO_REQUEUE;
}
ret = dm_dispatch_clone_request(clone, rq);
if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE) {
blk_rq_unprep_clone(clone);
- tio->ti->type->release_clone_rq(clone);
+ tio->ti->type->release_clone_rq(clone, &tio->info);
tio->clone = NULL;
return DM_MAPIO_REQUEUE;
}
return DM_MAPIO_KILL;
}
-static void io_err_release_clone_rq(struct request *clone)
+static void io_err_release_clone_rq(struct request *clone,
+ union map_info *map_context)
{
}
struct request *rq,
union map_info *map_context,
struct request **clone);
-typedef void (*dm_release_clone_request_fn) (struct request *clone);
+typedef void (*dm_release_clone_request_fn) (struct request *clone,
+ union map_info *map_context);
/*
* Returns: