This is a note to let you know that I've just added the patch titled dm mpath: requeue after a small delay if blk_get_request() fails to the 4.11-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary The filename of the patch is: dm-mpath-requeue-after-a-small-delay-if-blk_get_request-fails.patch and it can be found in the queue-4.11 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable@xxxxxxxxxxxxxxx> know about it. >From 06eb061f48594aa369f6e852b352410298b317a8 Mon Sep 17 00:00:00 2001 From: Bart Van Assche <bart.vanassche@xxxxxxxxxxx> Date: Fri, 7 Apr 2017 16:50:44 -0700 Subject: dm mpath: requeue after a small delay if blk_get_request() fails From: Bart Van Assche <bart.vanassche@xxxxxxxxxxx> commit 06eb061f48594aa369f6e852b352410298b317a8 upstream. If blk_get_request() returns ENODEV then multipath_clone_and_map() causes a request to be requeued immediately. This can cause a kworker thread to spend 100% of the CPU time of a single core in __blk_mq_run_hw_queue() and also can cause device removal to never finish. Avoid this by only requeuing after a delay if blk_get_request() fails. Additionally, reduce the requeue delay. Signed-off-by: Bart Van Assche <bart.vanassche@xxxxxxxxxxx> Signed-off-by: Mike Snitzer <snitzer@xxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- drivers/md/dm-mpath.c | 5 ++--- drivers/md/dm-rq.c | 2 +- 2 files changed, 3 insertions(+), 4 deletions(-) --- a/drivers/md/dm-mpath.c +++ b/drivers/md/dm-mpath.c @@ -484,7 +484,6 @@ static int multipath_clone_and_map(struc struct request **__clone) { struct multipath *m = ti->private; - int r = DM_MAPIO_REQUEUE; size_t nr_bytes = blk_rq_bytes(rq); struct pgpath *pgpath; struct block_device *bdev; @@ -503,7 +502,7 @@ static int multipath_clone_and_map(struc } else if (test_bit(MPATHF_QUEUE_IO, &m->flags) || test_bit(MPATHF_PG_INIT_REQUIRED, &m->flags)) { pg_init_all_paths(m); - return r; + return DM_MAPIO_REQUEUE; } memset(mpio, 0, sizeof(*mpio)); @@ -517,7 +516,7 @@ static int multipath_clone_and_map(struc GFP_ATOMIC); if (IS_ERR(clone)) { /* EBUSY, ENODEV or EWOULDBLOCK: requeue */ - return r; + return DM_MAPIO_DELAY_REQUEUE; } clone->bio = clone->biotail = NULL; clone->rq_disk = bdev->bd_disk; --- a/drivers/md/dm-rq.c +++ b/drivers/md/dm-rq.c @@ -280,7 +280,7 @@ static void dm_requeue_original_request( if (!rq->q->mq_ops) dm_old_requeue_request(rq); else - dm_mq_delay_requeue_request(rq, delay_requeue ? 5000 : 0); + dm_mq_delay_requeue_request(rq, delay_requeue ? 100/*ms*/ : 0); rq_completed(md, rw, false); } Patches currently in stable-queue which might be from bart.vanassche@xxxxxxxxxxx are queue-4.11/dm-mpath-requeue-after-a-small-delay-if-blk_get_request-fails.patch queue-4.11/dm-mpath-avoid-that-path-removal-can-trigger-an-infinite-loop.patch queue-4.11/dm-mpath-split-and-rename-activate_path-to-prepare-for-its-expanded-use.patch queue-4.11/dm-mpath-delay-requeuing-while-path-initialization-is-in-progress.patch