linux-stable.git/drivers/md, branch v5.4.32

Revert "dm: always call blk_queue_split() in dm_process_bio()"

2020-04-08T07:08:43+00:00

commit 120c9257f5f19e5d1e87efcbb5531b7cd81b7d74 upstream.

This reverts commit effd58c95f277744f75d6e08819ac859dbcbd351.

blk_queue_split() is causing excessive IO splitting -- because
blk_max_size_offset() depends on 'chunk_sectors' limit being set and
if it isn't (as is the case for DM targets!) it falls back to
splitting on a 'max_sectors' boundary regardless of offset.

"Fix" this by reverting back to _not_ using blk_queue_split() in
dm_process_bio() for normal IO (reads and writes).  Long-term fix is
still TBD but it should focus on training blk_max_size_offset() to
call into a DM provided hook (to call DM's max_io_len()).

Test results from simple misaligned IO test on 4-way dm-striped device
with chunksize of 128K and stripesize of 512K:

xfs_io -d -c 'pread -b 2m 224s 4072s' /dev/mapper/stripe_dev

before this revert:

253,0   21        1     0.000000000  2206  Q   R 224 + 4072 [xfs_io]
253,0   21        2     0.000008267  2206  X   R 224 / 480 [xfs_io]
253,0   21        3     0.000010530  2206  X   R 224 / 256 [xfs_io]
253,0   21        4     0.000027022  2206  X   R 480 / 736 [xfs_io]
253,0   21        5     0.000028751  2206  X   R 480 / 512 [xfs_io]
253,0   21        6     0.000033323  2206  X   R 736 / 992 [xfs_io]
253,0   21        7     0.000035130  2206  X   R 736 / 768 [xfs_io]
253,0   21        8     0.000039146  2206  X   R 992 / 1248 [xfs_io]
253,0   21        9     0.000040734  2206  X   R 992 / 1024 [xfs_io]
253,0   21       10     0.000044694  2206  X   R 1248 / 1504 [xfs_io]
253,0   21       11     0.000046422  2206  X   R 1248 / 1280 [xfs_io]
253,0   21       12     0.000050376  2206  X   R 1504 / 1760 [xfs_io]
253,0   21       13     0.000051974  2206  X   R 1504 / 1536 [xfs_io]
253,0   21       14     0.000055881  2206  X   R 1760 / 2016 [xfs_io]
253,0   21       15     0.000057462  2206  X   R 1760 / 1792 [xfs_io]
253,0   21       16     0.000060999  2206  X   R 2016 / 2272 [xfs_io]
253,0   21       17     0.000062489  2206  X   R 2016 / 2048 [xfs_io]
253,0   21       18     0.000066133  2206  X   R 2272 / 2528 [xfs_io]
253,0   21       19     0.000067507  2206  X   R 2272 / 2304 [xfs_io]
253,0   21       20     0.000071136  2206  X   R 2528 / 2784 [xfs_io]
253,0   21       21     0.000072764  2206  X   R 2528 / 2560 [xfs_io]
253,0   21       22     0.000076185  2206  X   R 2784 / 3040 [xfs_io]
253,0   21       23     0.000077486  2206  X   R 2784 / 2816 [xfs_io]
253,0   21       24     0.000080885  2206  X   R 3040 / 3296 [xfs_io]
253,0   21       25     0.000082316  2206  X   R 3040 / 3072 [xfs_io]
253,0   21       26     0.000085788  2206  X   R 3296 / 3552 [xfs_io]
253,0   21       27     0.000087096  2206  X   R 3296 / 3328 [xfs_io]
253,0   21       28     0.000093469  2206  X   R 3552 / 3808 [xfs_io]
253,0   21       29     0.000095186  2206  X   R 3552 / 3584 [xfs_io]
253,0   21       30     0.000099228  2206  X   R 3808 / 4064 [xfs_io]
253,0   21       31     0.000101062  2206  X   R 3808 / 3840 [xfs_io]
253,0   21       32     0.000104956  2206  X   R 4064 / 4096 [xfs_io]
253,0   21       33     0.001138823     0  C   R 4096 + 200 [0]

after this revert:

253,0   18        1     0.000000000  4430  Q   R 224 + 3896 [xfs_io]
253,0   18        2     0.000018359  4430  X   R 224 / 256 [xfs_io]
253,0   18        3     0.000028898  4430  X   R 256 / 512 [xfs_io]
253,0   18        4     0.000033535  4430  X   R 512 / 768 [xfs_io]
253,0   18        5     0.000065684  4430  X   R 768 / 1024 [xfs_io]
253,0   18        6     0.000091695  4430  X   R 1024 / 1280 [xfs_io]
253,0   18        7     0.000098494  4430  X   R 1280 / 1536 [xfs_io]
253,0   18        8     0.000114069  4430  X   R 1536 / 1792 [xfs_io]
253,0   18        9     0.000129483  4430  X   R 1792 / 2048 [xfs_io]
253,0   18       10     0.000136759  4430  X   R 2048 / 2304 [xfs_io]
253,0   18       11     0.000152412  4430  X   R 2304 / 2560 [xfs_io]
253,0   18       12     0.000160758  4430  X   R 2560 / 2816 [xfs_io]
253,0   18       13     0.000183385  4430  X   R 2816 / 3072 [xfs_io]
253,0   18       14     0.000190797  4430  X   R 3072 / 3328 [xfs_io]
253,0   18       15     0.000197667  4430  X   R 3328 / 3584 [xfs_io]
253,0   18       16     0.000218751  4430  X   R 3584 / 3840 [xfs_io]
253,0   18       17     0.000226005  4430  X   R 3840 / 4096 [xfs_io]
253,0   18       18     0.000250404  4430  Q   R 4120 + 176 [xfs_io]
253,0   18       19     0.000847708     0  C   R 4096 + 24 [0]
253,0   18       20     0.000855783     0  C   R 4120 + 176 [0]

Fixes: effd58c95f27774 ("dm: always call blk_queue_split() in dm_process_bio()")
Cc: stable@vger.kernel.org
Reported-by: Andreas Gruenbacher 
Tested-by: Barry Marson 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm integrity: use dm_bio_record and dm_bio_restore

2020-03-25T07:25:48+00:00

[ Upstream commit 248aa2645aa7fc9175d1107c2593cc90d4af5a4e ]

In cases where dec_in_flight() has to requeue the integrity_bio_wait
work to transfer the rest of the data, the bio's __bi_remaining might
already have been decremented to 0, e.g.: if bio passed to underlying
data device was split via blk_queue_split().

Use dm_bio_{record,restore} rather than effectively open-coding them in
dm-integrity -- these methods now manage __bi_remaining too.

Depends-on: f7f0b057a9c1 ("dm bio record: save/restore bi_end_io and bi_integrity")
Reported-by: Daniel Glöckner 
Suggested-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Sasha Levin

dm bio record: save/restore bi_end_io and bi_integrity

2020-03-25T07:25:48+00:00

[ Upstream commit 1b17159e52bb31f982f82a6278acd7fab1d3f67b ]

Also, save/restore __bi_remaining in case the bio was used in a
BIO_CHAIN (e.g. due to blk_queue_split).

Suggested-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Sasha Levin

dm: fix congested_fn for request-based device

2020-03-12T12:00:24+00:00

commit 974f51e8633f0f3f33e8f86bbb5ae66758aa63c7 upstream.

We neither assign congested_fn for requested-based blk-mq device nor
implement it correctly. So fix both.

Also, remove incorrect comment from dm_init_normal_md_queue and rename
it to dm_init_congested_fn.

Fixes: 4aa9c692e052 ("bdi: separate out congested state into a separate struct")
Cc: stable@vger.kernel.org
Signed-off-by: Hou Tao 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm zoned: Fix reference counter initial value of chunk works

2020-03-12T12:00:24+00:00

commit ee63634bae02e13c8c0df1209a6a0ca5326f3189 upstream.

Dm-zoned initializes reference counters of new chunk works with zero
value and refcount_inc() is called to increment the counter. However, the
refcount_inc() function handles the addition to zero value as an error
and triggers the warning as follows:

refcount_t: addition on 0; use-after-free.
WARNING: CPU: 7 PID: 1506 at lib/refcount.c:25 refcount_warn_saturate+0x68/0xf0
...
CPU: 7 PID: 1506 Comm: systemd-udevd Not tainted 5.4.0+ #134
...
Call Trace:
 dmz_map+0x2d2/0x350 [dm_zoned]
 __map_bio+0x42/0x1a0
 __split_and_process_non_flush+0x14a/0x1b0
 __split_and_process_bio+0x83/0x240
 ? kmem_cache_alloc+0x165/0x220
 dm_process_bio+0x90/0x230
 ? generic_make_request_checks+0x2e7/0x680
 dm_make_request+0x3e/0xb0
 generic_make_request+0xcf/0x320
 ? memcg_drain_all_list_lrus+0x1c0/0x1c0
 submit_bio+0x3c/0x160
 ? guard_bio_eod+0x2c/0x130
 mpage_readpages+0x182/0x1d0
 ? bdev_evict_inode+0xf0/0xf0
 read_pages+0x6b/0x1b0
 __do_page_cache_readahead+0x1ba/0x1d0
 force_page_cache_readahead+0x93/0x100
 generic_file_read_iter+0x83a/0xe40
 ? __seccomp_filter+0x7b/0x670
 new_sync_read+0x12a/0x1c0
 vfs_read+0x9d/0x150
 ksys_read+0x5f/0xe0
 do_syscall_64+0x5b/0x180
 entry_SYSCALL_64_after_hwframe+0x44/0xa9
...

After this warning, following refcount API calls for the counter all fail
to change the counter value.

Fix this by setting the initial reference counter value not zero but one
for the new chunk works. Instead, do not call refcount_inc() via
dmz_get_chunk_work() for the new chunks works.

The failure was observed with linux version 5.4 with CONFIG_REFCOUNT_FULL
enabled. Refcount rework was merged to linux version 5.5 by the
commit 168829ad09ca ("Merge branch 'locking-core-for-linus' of
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip"). After this
commit, CONFIG_REFCOUNT_FULL was removed and the failure was observed
regardless of kernel configuration.

Linux version 4.20 merged the commit 092b5648760a ("dm zoned: target: use
refcount_t for dm zoned reference counters"). Before this commit, dm
zoned used atomic_t APIs which does not check addition to zero, then this
fix is not necessary.

Fixes: 092b5648760a ("dm zoned: target: use refcount_t for dm zoned reference counters")
Cc: stable@vger.kernel.org # 5.4+
Signed-off-by: Shin'ichiro Kawasaki 
Reviewed-by: Damien Le Moal 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm writecache: verify watermark during resume

2020-03-12T12:00:24+00:00

commit 41c526c5af46d4c4dab7f72c99000b7fac0b9702 upstream.

Verify the watermark upon resume - so that if the target is reloaded
with lower watermark, it will start the cleanup process immediately.

Fixes: 48debafe4f2f ("dm: add writecache target")
Cc: stable@vger.kernel.org # 4.18+
Signed-off-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm: report suspended device during destroy

2020-03-12T12:00:23+00:00

commit adc0daad366b62ca1bce3e2958a40b0b71a8b8b3 upstream.

The function dm_suspended returns true if the target is suspended.
However, when the target is being suspended during unload, it returns
false.

An example where this is a problem: the test "!dm_suspended(wc->ti)" in
writecache_writeback is not sufficient, because dm_suspended returns
zero while writecache_suspend is in progress.  As is, without an
enhanced dm_suspended, simply switching from flush_workqueue to
drain_workqueue still emits warnings:
workqueue writecache-writeback: drain_workqueue() isn't complete after 10 tries
workqueue writecache-writeback: drain_workqueue() isn't complete after 100 tries
workqueue writecache-writeback: drain_workqueue() isn't complete after 200 tries
workqueue writecache-writeback: drain_workqueue() isn't complete after 300 tries
workqueue writecache-writeback: drain_workqueue() isn't complete after 400 tries

writecache_suspend calls flush_workqueue(wc->writeback_wq) - this function
flushes the current work. However, the workqueue may re-queue itself and
flush_workqueue doesn't wait for re-queued works to finish. Because of
this - the function writecache_writeback continues execution after the
device was suspended and then concurrently with writecache_dtr, causing
a crash in writecache_writeback.

We must use drain_workqueue - that waits until the work and all re-queued
works finish.

As a prereq for switching to drain_workqueue, this commit fixes
dm_suspended to return true after the presuspend hook and before the
postsuspend hook - just like during a normal suspend. It allows
simplifying the dm-integrity and dm-writecache targets so that they
don't have to maintain suspended flags on their own.

With this change use of drain_workqueue() can be used effectively.  This
change was tested with the lvm2 testsuite and cryptsetup testsuite and
the are no regressions.

Fixes: 48debafe4f2f ("dm: add writecache target")
Cc: stable@vger.kernel.org # 4.18+
Reported-by: Corey Marthaler 
Signed-off-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm cache: fix a crash due to incorrect work item cancelling

2020-03-12T12:00:23+00:00

commit 7cdf6a0aae1cccf5167f3f04ecddcf648b78e289 upstream.

The crash can be reproduced by running the lvm2 testsuite test
lvconvert-thin-external-cache.sh for several minutes, e.g.:
  while :; do make check T=shell/lvconvert-thin-external-cache.sh; done

The crash happens in this call chain:
do_waker -> policy_tick -> smq_tick -> end_hotspot_period -> clear_bitset
-> memset -> __memset -- which accesses an invalid pointer in the vmalloc
area.

The work entry on the workqueue is executed even after the bitmap was
freed. The problem is that cancel_delayed_work doesn't wait for the
running work item to finish, so the work item can continue running and
re-submitting itself even after cache_postsuspend. In order to make sure
that the work item won't be running, we must use cancel_delayed_work_sync.

Also, change flush_workqueue to drain_workqueue, so that if some work item
submits itself or another work item, we are properly waiting for both of
them.

Fixes: c6b4fcbad044 ("dm: add cache target")
Cc: stable@vger.kernel.org # v3.9
Signed-off-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm integrity: fix invalid table returned due to argument count mismatch

2020-03-12T12:00:23+00:00

commit 7fc2e47f40dd77ab1fcbda6db89614a0173d89c7 upstream.

If the flag SB_FLAG_RECALCULATE is present in the superblock, but it was
not specified on the command line (i.e. ic->recalculate_flag is false),
dm-integrity would return invalid table line - the reported number of
arguments would not match the real number.

Fixes: 468dfca38b1a ("dm integrity: add a bitmap mode")
Cc: stable@vger.kernel.org # v5.2+
Reported-by: Ondrej Kozina 
Signed-off-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman

dm integrity: fix a deadlock due to offloading to an incorrect workqueue

2020-03-12T12:00:23+00:00

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@vger.kernel.org # v4.12+
Reported-by: Heinz Mauelshagen 
Tested-by: Heinz Mauelshagen 
Signed-off-by: Heinz Mauelshagen 
Signed-off-by: Mikulas Patocka 
Signed-off-by: Mike Snitzer 
Signed-off-by: Greg Kroah-Hartman