[PATCH 5.4 105/168] dm integrity: fix a deadlock due to offloading to an incorrect workqueue

From: Greg Kroah-Hartman
Date: Tue Mar 10 2020 - 08:52:41 EST


From: Mikulas Patocka <mpatocka@xxxxxxxxxx>

commit 53770f0ec5fd417429775ba006bc4abe14002335 upstream.

If we need to perform synchronous I/O in dm_integrity_map_continue(),
we must make sure that we are not in the map function - in order to
avoid the deadlock due to bio queuing in generic_make_request. To
avoid the deadlock, we offload the request to metadata_wq.

However, metadata_wq also processes metadata updates for write requests.
If there are too many requests that get offloaded to metadata_wq at the
beginning of dm_integrity_map_continue, the workqueue metadata_wq
becomes clogged and the system is incapable of processing any metadata
updates.

This causes a deadlock because all the requests that need to do metadata
updates wait for metadata_wq to proceed and metadata_wq waits inside
wait_and_add_new_range until some existing request releases its range
lock (which doesn't happen because the range lock is released after
metadata update).

In order to fix the deadlock, we create a new workqueue offload_wq and
offload requests to it - so that processing of offload_wq is independent
from processing of metadata_wq.

Fixes: 7eada909bfd7 ("dm: add integrity target")
Cc: stable@xxxxxxxxxxxxxxx # v4.12+
Reported-by: Heinz Mauelshagen <heinzm@xxxxxxxxxx>
Tested-by: Heinz Mauelshagen <heinzm@xxxxxxxxxx>
Signed-off-by: Heinz Mauelshagen <heinzm@xxxxxxxxxx>
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-integrity.c | 19 +++++++++++++++----
1 file changed, 15 insertions(+), 4 deletions(-)

--- a/drivers/md/dm-integrity.c
+++ b/drivers/md/dm-integrity.c
@@ -210,6 +210,7 @@ struct dm_integrity_c {
struct list_head wait_list;
wait_queue_head_t endio_wait;
struct workqueue_struct *wait_wq;
+ struct workqueue_struct *offload_wq;

unsigned char commit_seq;
commit_id_t commit_ids[N_COMMIT_IDS];
@@ -1434,7 +1435,7 @@ static void dec_in_flight(struct dm_inte
dio->range.logical_sector += dio->range.n_sectors;
bio_advance(bio, dio->range.n_sectors << SECTOR_SHIFT);
INIT_WORK(&dio->work, integrity_bio_wait);
- queue_work(ic->wait_wq, &dio->work);
+ queue_work(ic->offload_wq, &dio->work);
return;
}
do_endio_flush(ic, dio);
@@ -1860,7 +1861,7 @@ static void dm_integrity_map_continue(st

if (need_sync_io && from_map) {
INIT_WORK(&dio->work, integrity_bio_wait);
- queue_work(ic->metadata_wq, &dio->work);
+ queue_work(ic->offload_wq, &dio->work);
return;
}

@@ -2496,7 +2497,7 @@ static void bitmap_block_work(struct wor
dio->range.n_sectors, BITMAP_OP_TEST_ALL_SET)) {
remove_range(ic, &dio->range);
INIT_WORK(&dio->work, integrity_bio_wait);
- queue_work(ic->wait_wq, &dio->work);
+ queue_work(ic->offload_wq, &dio->work);
} else {
block_bitmap_op(ic, ic->journal, dio->range.logical_sector,
dio->range.n_sectors, BITMAP_OP_SET);
@@ -2519,7 +2520,7 @@ static void bitmap_block_work(struct wor

remove_range(ic, &dio->range);
INIT_WORK(&dio->work, integrity_bio_wait);
- queue_work(ic->wait_wq, &dio->work);
+ queue_work(ic->offload_wq, &dio->work);
}

queue_delayed_work(ic->commit_wq, &ic->bitmap_flush_work, ic->bitmap_flush_interval);
@@ -3825,6 +3826,14 @@ static int dm_integrity_ctr(struct dm_ta
goto bad;
}

+ ic->offload_wq = alloc_workqueue("dm-integrity-offload", WQ_MEM_RECLAIM,
+ METADATA_WORKQUEUE_MAX_ACTIVE);
+ if (!ic->offload_wq) {
+ ti->error = "Cannot allocate workqueue";
+ r = -ENOMEM;
+ goto bad;
+ }
+
ic->commit_wq = alloc_workqueue("dm-integrity-commit", WQ_MEM_RECLAIM, 1);
if (!ic->commit_wq) {
ti->error = "Cannot allocate workqueue";
@@ -4129,6 +4138,8 @@ static void dm_integrity_dtr(struct dm_t
destroy_workqueue(ic->metadata_wq);
if (ic->wait_wq)
destroy_workqueue(ic->wait_wq);
+ if (ic->offload_wq)
+ destroy_workqueue(ic->offload_wq);
if (ic->commit_wq)
destroy_workqueue(ic->commit_wq);
if (ic->writer_wq)