| Age | Commit message (Collapse) | Author |
|---|
|
Stale && dirty keys can be produced in the follow way:
After writeback in write_dirty_finish(), dirty keys k1 will
replace by clean keys k2
==>ret = bch_btree_insert(dc->disk.c, &keys, NULL, &w->key);
==>btree_insert_fn(struct btree_op *b_op, struct btree *b)
==>static int bch_btree_insert_node(struct btree *b,
struct btree_op *op,
struct keylist *insert_keys,
atomic_t *journal_ref,
Then two steps:
A) update k1 to k2 in btree node memory;
bch_btree_insert_keys(b, op, insert_keys, replace_key)
B) Write the bset(contains k2) to cache disk by a 30s delay work
bch_btree_leaf_dirty(b, journal_ref).
But before the 30s delay work write the bset to cache device,
these things happened:
A) GC works, and reclaim the bucket k2 point to;
B) Allocator works, and invalidate the bucket k2 point to,
and increase the gen of the bucket, and place it into free_inc
fifo;
C) Until now, the 30s delay work still does not finish work,
so in the disk, the key still is k1, it is dirty and stale
(its gen is smaller than the gen of the bucket). and then the
machine power off suddenly happens;
D) When the machine power on again, after the btree reconstruction,
the stale dirty key appear.
In bch_extent_bad(), when expensive_debug_checks is off, it would
treat the dirty key as good even it is stale keys, and it would
cause bellow probelms:
A) In read_dirty() it would cause machine crash:
BUG_ON(ptr_stale(dc->disk.c, &w->key, 0));
B) It could be worse when reads hits stale dirty keys, it would
read old incorrect data.
This patch tolerate the existence of these stale && dirty keys,
and treat them as bad key in bch_extent_bad().
(Coly Li: fix indent which was modified by sender's email client)
Signed-off-by: Tang Junhui <tang.junhui.linux@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
There is a hunk of code that is indented one level too deep, fix this
by removing the extra tabs.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
When there are multiple bcache devices, after a reboot the name of
bcache devices may change (e.g. current /dev/bcache1 was /dev/bcache0
before reboot). Therefore we need the backing device UUID (sb.uuid) to
identify each bcache device.
Backing device uuid can be found by program bcache-super-show, but
directly exporting backing_dev_uuid by sysfs file
/sys/block/bcache<?>/bcache/backing_dev_uuid is a much simpler method.
With backing_dev_uuid, and partition uuids from /dev/disk/by-partuuid/,
now we can identify each bcache device and its partitions conveniently.
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
This patch export dc->backing_dev_name to sysfs file
/sys/block/bcache<?>/bcache/backing_dev_name, then people or user space
tools may know the backing device name of this bcache device.
Of cause it can be done by parsing sysfs links, but this method can be
much simpler to find the link between bcache device and backing device.
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
In stats.c:bch_cache_accounting_clear(), a hard coded number '7' is
used in memset(). It is because in struct cache_stats, there are 7
atomic_t type members. This is not good when new members added into
struct stats, the hard coded number will only clear part of memory.
This patch replaces 'sizeof(unsigned long) * 7' by more generic
'sizeof(struct cache_stats))', to avoid potential error if new
member added into struct cache_stats.
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Some users see panics like the following when performing fstrim on a
bcached volume:
[ 529.803060] BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
[ 530.183928] #PF error: [normal kernel read fault]
[ 530.412392] PGD 8000001f42163067 P4D 8000001f42163067 PUD 1f42168067 PMD 0
[ 530.750887] Oops: 0000 [#1] SMP PTI
[ 530.920869] CPU: 10 PID: 4167 Comm: fstrim Kdump: loaded Not tainted 5.0.0-rc1+ #3
[ 531.290204] Hardware name: HP ProLiant DL360 Gen9/ProLiant DL360 Gen9, BIOS P89 12/27/2015
[ 531.693137] RIP: 0010:blk_queue_split+0x148/0x620
[ 531.922205] Code: 60 38 89 55 a0 45 31 db 45 31 f6 45 31 c9 31 ff 89 4d 98 85 db 0f 84 7f 04 00 00 44 8b 6d 98 4c 89 ee 48 c1 e6 04 49 03 70 78 <8b> 46 08 44 8b 56 0c 48
8b 16 44 29 e0 39 d8 48 89 55 a8 0f 47 c3
[ 532.838634] RSP: 0018:ffffb9b708df39b0 EFLAGS: 00010246
[ 533.093571] RAX: 00000000ffffffff RBX: 0000000000046000 RCX: 0000000000000000
[ 533.441865] RDX: 0000000000000200 RSI: 0000000000000000 RDI: 0000000000000000
[ 533.789922] RBP: ffffb9b708df3a48 R08: ffff940d3b3fdd20 R09: 0000000000000000
[ 534.137512] R10: ffffb9b708df3958 R11: 0000000000000000 R12: 0000000000000000
[ 534.485329] R13: 0000000000000000 R14: 0000000000000000 R15: ffff940d39212020
[ 534.833319] FS: 00007efec26e3840(0000) GS:ffff940d1f480000(0000) knlGS:0000000000000000
[ 535.224098] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 535.504318] CR2: 0000000000000008 CR3: 0000001f4e256004 CR4: 00000000001606e0
[ 535.851759] Call Trace:
[ 535.970308] ? mempool_alloc_slab+0x15/0x20
[ 536.174152] ? bch_data_insert+0x42/0xd0 [bcache]
[ 536.403399] blk_mq_make_request+0x97/0x4f0
[ 536.607036] generic_make_request+0x1e2/0x410
[ 536.819164] submit_bio+0x73/0x150
[ 536.980168] ? submit_bio+0x73/0x150
[ 537.149731] ? bio_associate_blkg_from_css+0x3b/0x60
[ 537.391595] ? _cond_resched+0x1a/0x50
[ 537.573774] submit_bio_wait+0x59/0x90
[ 537.756105] blkdev_issue_discard+0x80/0xd0
[ 537.959590] ext4_trim_fs+0x4a9/0x9e0
[ 538.137636] ? ext4_trim_fs+0x4a9/0x9e0
[ 538.324087] ext4_ioctl+0xea4/0x1530
[ 538.497712] ? _copy_to_user+0x2a/0x40
[ 538.679632] do_vfs_ioctl+0xa6/0x600
[ 538.853127] ? __do_sys_newfstat+0x44/0x70
[ 539.051951] ksys_ioctl+0x6d/0x80
[ 539.212785] __x64_sys_ioctl+0x1a/0x20
[ 539.394918] do_syscall_64+0x5a/0x110
[ 539.568674] entry_SYSCALL_64_after_hwframe+0x44/0xa9
We have observed it where both:
1) LVM/devmapper is involved (bcache backing device is LVM volume) and
2) writeback cache is involved (bcache cache_mode is writeback)
On one machine, we can reliably reproduce it with:
# echo writeback > /sys/block/bcache0/bcache/cache_mode
(not sure whether above line is required)
# mount /dev/bcache0 /test
# for i in {0..10}; do
file="$(mktemp /test/zero.XXX)"
dd if=/dev/zero of="$file" bs=1M count=256
sync
rm $file
done
# fstrim -v /test
Observing this with tracepoints on, we see the following writes:
fstrim-18019 [022] .... 91107.302026: bcache_write: 73f95583-561c-408f-a93a-4cbd2498f5c8 inode 0 DS 4260112 + 196352 hit 0 bypass 1
fstrim-18019 [022] .... 91107.302050: bcache_write: 73f95583-561c-408f-a93a-4cbd2498f5c8 inode 0 DS 4456464 + 262144 hit 0 bypass 1
fstrim-18019 [022] .... 91107.302075: bcache_write: 73f95583-561c-408f-a93a-4cbd2498f5c8 inode 0 DS 4718608 + 81920 hit 0 bypass 1
fstrim-18019 [022] .... 91107.302094: bcache_write: 73f95583-561c-408f-a93a-4cbd2498f5c8 inode 0 DS 5324816 + 180224 hit 0 bypass 1
fstrim-18019 [022] .... 91107.302121: bcache_write: 73f95583-561c-408f-a93a-4cbd2498f5c8 inode 0 DS 5505040 + 262144 hit 0 bypass 1
fstrim-18019 [022] .... 91107.302145: bcache_write: 73f95583-561c-408f-a93a-4cbd2498f5c8 inode 0 DS 5767184 + 81920 hit 0 bypass 1
fstrim-18019 [022] .... 91107.308777: bcache_write: 73f95583-561c-408f-a93a-4cbd2498f5c8 inode 0 DS 6373392 + 180224 hit 1 bypass 0
<crash>
Note the final one has different hit/bypass flags.
This is because in should_writeback(), we were hitting a case where
the partial stripe condition was returning true and so
should_writeback() was returning true early.
If that hadn't been the case, it would have hit the would_skip test, and
as would_skip == s->iop.bypass == true, should_writeback() would have
returned false.
Looking at the git history from 'commit 72c270612bd3 ("bcache: Write out
full stripes")', it looks like the idea was to optimise for raid5/6:
* If a stripe is already dirty, force writes to that stripe to
writeback mode - to help build up full stripes of dirty data
To fix this issue, make sure that should_writeback() on a discard op
never returns true.
More details of debugging:
https://www.spinics.net/lists/linux-bcache/msg06996.html
Previous reports:
- https://bugzilla.kernel.org/show_bug.cgi?id=201051
- https://bugzilla.kernel.org/show_bug.cgi?id=196103
- https://www.spinics.net/lists/linux-bcache/msg06885.html
(Coly Li: minor modification to follow maximum 75 chars per line rule)
Cc: Kent Overstreet <koverstreet@google.com>
Cc: stable@vger.kernel.org
Fixes: 72c270612bd3 ("bcache: Write out full stripes")
Signed-off-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
bio_trim() has an early return, which makes it _not_ idempotent, if the
offset is 0 and the bio's bi_size already matches the requested size.
Prior to DM, all users of bio_trim() were fine with this. But DM has
exposed the fact that bio_trim()'s early return is incompatible with a
cloned bio whose integrity payload must be trimmed via
bio_integrity_trim().
Fix this by reverting DM back to doing the equivalent of bio_trim() but
in an idempotent manner (so bio_integrity_trim is always performed).
Follow-on work is needed to assess what benefit bio_trim()'s early
return is providing to its existing callers.
Reported-by: Milan Broz <gmazyland@gmail.com>
Fixes: 57c36519e4b94 ("dm: fix clone_bio() to trigger blk_recount_segments()")
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Block core changes to switch bio-based IO accounting to be percpu had a
side-effect of altering DM core to now rely on calling waitqueue_active
(in both bio-based and request-based) to check if another task is in
dm_wait_for_completion().
A memory barrier is needed before calling waitqueue_active(). DM core
doesn't piggyback on a preceding memory barrier so it must explicitly
use its own.
For more details on why using waitqueue_active() without a preceding
barrier is unsafe, please see the comment before the waitqueue_active()
definition in include/linux/wait.h.
Add the missing memory barrier by switching to using wq_has_sleeper().
Fixes: 6f75723190d8 ("dm: remove the pending IO accounting")
Fixes: c4576aed8d85 ("dm: fix request-based dm's use of dm_wait_for_completion")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Remove redundance set_bit and let code simplify.
Signed-off-by: Yufen Yu <yuyufen@huawei.com>
Signed-off-by: Song Liu <songliubraving@fb.com>
|
|
One of the more common cases of allocation size calculations is finding the
size of a structure that has a zero-sized array at the end, along with memory
for some number of elements for that array. For example:
struct foo {
int stuff;
void *entry[];
};
instance = kzalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL);
Instead of leaving these open-coded and prone to type mistakes, we can now
use the new struct_size() helper:
instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);
This code was detected with the help of Coccinelle.
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Song Liu <songliubraving@fb.com>
|
|
This fixes the case when md array assembly fails because of raid cache recovery
unable to allocate a stripe, despite attempts to replay stripes and increase
cache size. This happens because stripes released by r5c_recovery_replay_stripes
and raid5_set_cache_size don't become available for allocation immediately.
Released stripes first are placed on conf->released_stripes list and require
md thread to merge them on conf->inactive_list before they can be allocated.
Patch allows final allocation attempt during cache recovery to wait for
new stripes to become availabe for allocation.
Cc: linux-raid@vger.kernel.org
Cc: Shaohua Li <shli@kernel.org>
Cc: linux-stable <stable@vger.kernel.org> # 4.10+
Fixes: b4c625c67362 ("md/r5cache: r5cache recovery: part 1")
Signed-off-by: Alexei Naberezhnov <anaberezhnov@fb.com>
Signed-off-by: Song Liu <songliubraving@fb.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm
Pull device mapper fixes from Mike Snitzer:
- Fix DM crypt's parsing of extended IV arguments.
- Fix DM thinp's discard passdown to properly account for extra
reference that is taken to guard against reallocating a block before
a discard has been issued.
- Fix bio-based DM's redundant IO accounting that was occurring for
bios that must be split due to the nature of the DM target (e.g.
dm-stripe, dm-thinp, etc).
* tag 'for-5.0/dm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
dm: add missing trace_block_split() to __split_and_process_bio()
dm: fix dm_wq_work() to only use __split_and_process_bio() if appropriate
dm: fix redundant IO accounting for bios that need splitting
dm: fix clone_bio() to trigger blk_recount_segments()
dm thin: fix passdown_double_checking_shared_status()
dm crypt: fix parsing of extended IV arguments
|
|
Provides useful context about bio splits in blktrace.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Otherwise targets that don't support/expect IO splitting could resubmit
bios using code paths with unnecessary IO splitting complexity.
Depends-on: 24113d487843 ("dm: avoid indirect call in __dm_make_request")
Fixes: 978e51ba38e00 ("dm: optimize bio-based NVMe IO submission")
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
The risk of redundant IO accounting was not taken into consideration
when commit 18a25da84354 ("dm: ensure bio submission follows a
depth-first tree walk") introduced IO splitting in terms of recursion
via generic_make_request().
Fix this by subtracting the split bio's payload from the IO stats that
were already accounted for by start_io_acct() upon dm_make_request()
entry. This repeat oscillation of the IO accounting, up then down,
isn't ideal but refactoring DM core's IO splitting to pre-split bios
_before_ they are accounted turned out to be an excessive amount of
change that will need a full development cycle to refine and verify.
Before this fix:
/dev/mapper/stripe_dev is a 4-way stripe using a 32k chunksize, so
bios are split on 32k boundaries.
# fio --name=16M --filename=/dev/mapper/stripe_dev --rw=write --bs=64k --size=16M \
--iodepth=1 --ioengine=libaio --direct=1 --refill_buffers
with debugging added:
[103898.310264] device-mapper: core: start_io_acct: dm-2 WRITE bio->bi_iter.bi_sector=0 len=128
[103898.318704] device-mapper: core: __split_and_process_bio: recursing for following split bio:
[103898.329136] device-mapper: core: start_io_acct: dm-2 WRITE bio->bi_iter.bi_sector=64 len=64
...
16M written yet 136M (278528 * 512b) accounted:
# cat /sys/block/dm-2/stat | awk '{ print $7 }'
278528
After this fix:
16M written and 16M (32768 * 512b) accounted:
# cat /sys/block/dm-2/stat | awk '{ print $7 }'
32768
Fixes: 18a25da84354 ("dm: ensure bio submission follows a depth-first tree walk")
Cc: stable@vger.kernel.org # 4.16+
Reported-by: Bryan Gurney <bgurney@redhat.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
DM's clone_bio() now benefits from using bio_trim() by fixing the fact
that clone_bio() wasn't clearing BIO_SEG_VALID like bio_trim() does;
which triggers blk_recount_segments() via bio_phys_segments().
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Commit 00a0ea33b495 ("dm thin: do not queue freed thin mapping for next
stage processing") changed process_prepared_discard_passdown_pt1() to
increment all the blocks being discarded until after the passdown had
completed to avoid them being prematurely reused.
IO issued to a thin device that breaks sharing with a snapshot, followed
by a discard issued to snapshot(s) that previously shared the block(s),
results in passdown_double_checking_shared_status() being called to
iterate through the blocks double checking their reference count is zero
and issuing the passdown if so. So a side effect of commit 00a0ea33b495
is passdown_double_checking_shared_status() was broken.
Fix this by checking if the block reference count is greater than 1.
Also, rename dm_pool_block_is_used() to dm_pool_block_is_shared().
Fixes: 00a0ea33b495 ("dm thin: do not queue freed thin mapping for next stage processing")
Cc: stable@vger.kernel.org # 4.9+
Reported-by: ryan.p.norwood@gmail.com
Signed-off-by: Joe Thornber <ejt@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
bio_alloc_bioset returns a bio pointer or NULL, so we can avoid storing
the returned data into a new variable.
Acked-by: Guoqing Jiang <gqjiang@suse.com>
Acked-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
Signed-off-by: Marcos Paulo de Souza <marcos.souza.org@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
The dm-crypt cipher specification in a mapping table is defined as:
cipher[:keycount]-chainmode-ivmode[:ivopts]
or (new crypt API format):
capi:cipher_api_spec-ivmode[:ivopts]
For ESSIV, the parameter includes hash specification, for example:
aes-cbc-essiv:sha256
The implementation expected that additional IV option to never include
another dash '-' character.
But, with SHA3, there are names like sha3-256; so the mapping table
parser fails:
dmsetup create test --table "0 8 crypt aes-cbc-essiv:sha3-256 9c1185a5c5e9fc54612808977ee8f5b9e 0 /dev/sdb 0"
or (new crypt API format)
dmsetup create test --table "0 8 crypt capi:cbc(aes)-essiv:sha3-256 9c1185a5c5e9fc54612808977ee8f5b9e 0 /dev/sdb 0"
device-mapper: crypt: Ignoring unexpected additional cipher options
device-mapper: table: 253:0: crypt: Error creating IV
device-mapper: ioctl: error adding target to table
Fix the dm-crypt constructor to ignore additional dash in IV options and
also remove a bogus warning (that is ignored anyway).
Cc: stable@vger.kernel.org # 4.12+
Signed-off-by: Milan Broz <gmazyland@gmail.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/shli/md into for-linus
Pull the pending 4.21 changes for md from Shaohua.
* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md:
md: fix raid10 hang issue caused by barrier
raid10: refactor common wait code from regular read/write request
md: remvoe redundant condition check
lib/raid6: add option to skip algo benchmarking
lib/raid6: sort algos in rough performance order
lib/raid6: check for assembler SSSE3 support
lib/raid6: avoid __attribute_const__ redefinition
lib/raid6: add missing include for raid6test
md: remove set but not used variable 'bi_rdev'
|
|
Merge misc updates from Andrew Morton:
- large KASAN update to use arm's "software tag-based mode"
- a few misc things
- sh updates
- ocfs2 updates
- just about all of MM
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (167 commits)
kernel/fork.c: mark 'stack_vm_area' with __maybe_unused
memcg, oom: notify on oom killer invocation from the charge path
mm, swap: fix swapoff with KSM pages
include/linux/gfp.h: fix typo
mm/hmm: fix memremap.h, move dev_page_fault_t callback to hmm
hugetlbfs: Use i_mmap_rwsem to fix page fault/truncate race
hugetlbfs: use i_mmap_rwsem for more pmd sharing synchronization
memory_hotplug: add missing newlines to debugging output
mm: remove __hugepage_set_anon_rmap()
include/linux/vmstat.h: remove unused page state adjustment macro
mm/page_alloc.c: allow error injection
mm: migrate: drop unused argument of migrate_page_move_mapping()
blkdev: avoid migration stalls for blkdev pages
mm: migrate: provide buffer_migrate_page_norefs()
mm: migrate: move migrate_page_lock_buffers()
mm: migrate: lock buffers before migrate_page_move_mapping()
mm: migration: factor out code to compute expected number of page references
mm, page_alloc: enable pcpu_drain with zone capability
kmemleak: add config to select auto scan
mm/page_alloc.c: don't call kasan_free_pages() at deferred mem init
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm
Pull device mapper updates from Mike Snitzer:
- Eliminate a couple indirect calls from bio-based DM core.
- Fix DM to allow reads that exceed readahead limits by setting
io_pages in the backing_dev_info.
- A couple code cleanups in request-based DM.
- Fix various DM targets to check for device sector overflow if
CONFIG_LBDAF is not set.
- Use u64 instead of sector_t to store iv_offset in DM crypt; sector_t
isn't large enough on 32bit when CONFIG_LBDAF is not set.
- Performance fixes to DM's kcopyd and the snapshot target focused on
limiting memory use and workqueue stalls.
- Fix typos in the integrity and writecache targets.
- Log which algorithm is used for dm-crypt's encryption and
dm-integrity's hashing.
- Fix false -EBUSY errors in DM raid target's handling of check/repair
messages.
- Fix DM flakey target's corrupt_bio_byte feature to reliably corrupt
the Nth byte in a bio's payload.
* tag 'for-4.21/dm-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
dm: do not allow readahead to limit IO size
dm raid: fix false -EBUSY when handling check/repair message
dm rq: cleanup leftover code from recently removed q->mq_ops branching
dm verity: log the hash algorithm implementation
dm crypt: log the encryption algorithm implementation
dm integrity: fix spelling mistake in workqueue name
dm flakey: Properly corrupt multi-page bios.
dm: Check for device sector overflow if CONFIG_LBDAF is not set
dm crypt: use u64 instead of sector_t to store iv_offset
dm kcopyd: Fix bug causing workqueue stalls
dm snapshot: Fix excessive memory usage and workqueue stalls
dm bufio: update comment in dm-bufio.c
dm writecache: fix typo in error msg for creating writecache_flush_thread
dm: remove indirect calls from __send_changing_extent_only()
dm mpath: only flush workqueue when needed
dm rq: remove unused arguments from rq_completed()
dm: avoid indirect call in __dm_make_request
|
|
Pull block updates from Jens Axboe:
"This is the main pull request for block/storage for 4.21.
Larger than usual, it was a busy round with lots of goodies queued up.
Most notable is the removal of the old IO stack, which has been a long
time coming. No new features for a while, everything coming in this
week has all been fixes for things that were previously merged.
This contains:
- Use atomic counters instead of semaphores for mtip32xx (Arnd)
- Cleanup of the mtip32xx request setup (Christoph)
- Fix for circular locking dependency in loop (Jan, Tetsuo)
- bcache (Coly, Guoju, Shenghui)
* Optimizations for writeback caching
* Various fixes and improvements
- nvme (Chaitanya, Christoph, Sagi, Jay, me, Keith)
* host and target support for NVMe over TCP
* Error log page support
* Support for separate read/write/poll queues
* Much improved polling
* discard OOM fallback
* Tracepoint improvements
- lightnvm (Hans, Hua, Igor, Matias, Javier)
* Igor added packed metadata to pblk. Now drives without metadata
per LBA can be used as well.
* Fix from Geert on uninitialized value on chunk metadata reads.
* Fixes from Hans and Javier to pblk recovery and write path.
* Fix from Hua Su to fix a race condition in the pblk recovery
code.
* Scan optimization added to pblk recovery from Zhoujie.
* Small geometry cleanup from me.
- Conversion of the last few drivers that used the legacy path to
blk-mq (me)
- Removal of legacy IO path in SCSI (me, Christoph)
- Removal of legacy IO stack and schedulers (me)
- Support for much better polling, now without interrupts at all.
blk-mq adds support for multiple queue maps, which enables us to
have a map per type. This in turn enables nvme to have separate
completion queues for polling, which can then be interrupt-less.
Also means we're ready for async polled IO, which is hopefully
coming in the next release.
- Killing of (now) unused block exports (Christoph)
- Unification of the blk-rq-qos and blk-wbt wait handling (Josef)
- Support for zoned testing with null_blk (Masato)
- sx8 conversion to per-host tag sets (Christoph)
- IO priority improvements (Damien)
- mq-deadline zoned fix (Damien)
- Ref count blkcg series (Dennis)
- Lots of blk-mq improvements and speedups (me)
- sbitmap scalability improvements (me)
- Make core inflight IO accounting per-cpu (Mikulas)
- Export timeout setting in sysfs (Weiping)
- Cleanup the direct issue path (Jianchao)
- Export blk-wbt internals in block debugfs for easier debugging
(Ming)
- Lots of other fixes and improvements"
* tag 'for-4.21/block-20181221' of git://git.kernel.dk/linux-block: (364 commits)
kyber: use sbitmap add_wait_queue/list_del wait helpers
sbitmap: add helpers for add/del wait queue handling
block: save irq state in blkg_lookup_create()
dm: don't reuse bio for flushes
nvme-pci: trace SQ status on completions
nvme-rdma: implement polling queue map
nvme-fabrics: allow user to pass in nr_poll_queues
nvme-fabrics: allow nvmf_connect_io_queue to poll
nvme-core: optionally poll sync commands
block: make request_to_qc_t public
nvme-tcp: fix spelling mistake "attepmpt" -> "attempt"
nvme-tcp: fix endianess annotations
nvmet-tcp: fix endianess annotations
nvme-pci: refactor nvme_poll_irqdisable to make sparse happy
nvme-pci: only set nr_maps to 2 if poll queues are supported
nvmet: use a macro for default error location
nvmet: fix comparison of a u16 with -1
blk-mq: enable IO poll if .nr_queues of type poll > 0
blk-mq: change blk_mq_queue_busy() to blk_mq_queue_inflight()
blk-mq: skip zero-queue maps in blk_mq_map_swqueue
...
|
|
totalram_pages and totalhigh_pages are made static inline function.
Main motivation was that managed_page_count_lock handling was complicating
things. It was discussed in length here,
https://lore.kernel.org/patchwork/patch/995739/#1181785 So it seemes
better to remove the lock and convert variables to atomic, with preventing
poteintial store-to-read tearing as a bonus.
[akpm@linux-foundation.org: coding style fixes]
Link: http://lkml.kernel.org/r/1542090790-21750-4-git-send-email-arunks@codeaurora.org
Signed-off-by: Arun KS <arunks@codeaurora.org>
Suggested-by: Michal Hocko <mhocko@suse.com>
Suggested-by: Vlastimil Babka <vbabka@suse.cz>
Reviewed-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Reviewed-by: Pavel Tatashin <pasha.tatashin@soleen.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: David Hildenbrand <david@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Pull crypto updates from Herbert Xu:
"API:
- Add 1472-byte test to tcrypt for IPsec
- Reintroduced crypto stats interface with numerous changes
- Support incremental algorithm dumps
Algorithms:
- Add xchacha12/20
- Add nhpoly1305
- Add adiantum
- Add streebog hash
- Mark cts(cbc(aes)) as FIPS allowed
Drivers:
- Improve performance of arm64/chacha20
- Improve performance of x86/chacha20
- Add NEON-accelerated nhpoly1305
- Add SSE2 accelerated nhpoly1305
- Add AVX2 accelerated nhpoly1305
- Add support for 192/256-bit keys in gcmaes AVX
- Add SG support in gcmaes AVX
- ESN for inline IPsec tx in chcr
- Add support for CryptoCell 703 in ccree
- Add support for CryptoCell 713 in ccree
- Add SM4 support in ccree
- Add SM3 support in ccree
- Add support for chacha20 in caam/qi2
- Add support for chacha20 + poly1305 in caam/jr
- Add support for chacha20 + poly1305 in caam/qi2
- Add AEAD cipher support in cavium/nitrox"
* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (130 commits)
crypto: skcipher - remove remnants of internal IV generators
crypto: cavium/nitrox - Fix build with !CONFIG_DEBUG_FS
crypto: salsa20-generic - don't unnecessarily use atomic walk
crypto: skcipher - add might_sleep() to skcipher_walk_virt()
crypto: x86/chacha - avoid sleeping under kernel_fpu_begin()
crypto: cavium/nitrox - Added AEAD cipher support
crypto: mxc-scc - fix build warnings on ARM64
crypto: api - document missing stats member
crypto: user - remove unused dump functions
crypto: chelsio - Fix wrong error counter increments
crypto: chelsio - Reset counters on cxgb4 Detach
crypto: chelsio - Handle PCI shutdown event
crypto: chelsio - cleanup:send addr as value in function argument
crypto: chelsio - Use same value for both channel in single WR
crypto: chelsio - Swap location of AAD and IV sent in WR
crypto: chelsio - remove set but not used variable 'kctx_len'
crypto: ux500 - Use proper enum in hash_set_dma_transfer
crypto: ux500 - Use proper enum in cryp_set_dma_transfer
crypto: aesni - Add scatter/gather avx stubs, and use them in C
crypto: aesni - Introduce partial block macro
..
|
|
When both regular IO and resync IO happen at the same time,
and if we also need to split regular. Then we can see tasks
hang due to barrier.
1. resync thread
[ 1463.757205] INFO: task md1_resync:5215 blocked for more than 480 seconds.
[ 1463.757207] Not tainted 4.19.5-1-default #1
[ 1463.757209] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1463.757212] md1_resync D 0 5215 2 0x80000000
[ 1463.757216] Call Trace:
[ 1463.757223] ? __schedule+0x29a/0x880
[ 1463.757231] ? raise_barrier+0x8d/0x140 [raid10]
[ 1463.757236] schedule+0x78/0x110
[ 1463.757243] raise_barrier+0x8d/0x140 [raid10]
[ 1463.757248] ? wait_woken+0x80/0x80
[ 1463.757257] raid10_sync_request+0x1f6/0x1e30 [raid10]
[ 1463.757265] ? _raw_spin_unlock_irq+0x22/0x40
[ 1463.757284] ? is_mddev_idle+0x125/0x137 [md_mod]
[ 1463.757302] md_do_sync.cold.78+0x404/0x969 [md_mod]
[ 1463.757311] ? wait_woken+0x80/0x80
[ 1463.757336] ? md_rdev_init+0xb0/0xb0 [md_mod]
[ 1463.757351] md_thread+0xe9/0x140 [md_mod]
[ 1463.757358] ? _raw_spin_unlock_irqrestore+0x2e/0x60
[ 1463.757364] ? __kthread_parkme+0x4c/0x70
[ 1463.757369] kthread+0x112/0x130
[ 1463.757374] ? kthread_create_worker_on_cpu+0x40/0x40
[ 1463.757380] ret_from_fork+0x3a/0x50
2. regular IO
[ 1463.760679] INFO: task kworker/0:8:5367 blocked for more than 480 seconds.
[ 1463.760683] Not tainted 4.19.5-1-default #1
[ 1463.760684] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1463.760687] kworker/0:8 D 0 5367 2 0x80000000
[ 1463.760718] Workqueue: md submit_flushes [md_mod]
[ 1463.760721] Call Trace:
[ 1463.760731] ? __schedule+0x29a/0x880
[ 1463.760741] ? wait_barrier+0xdd/0x170 [raid10]
[ 1463.760746] schedule+0x78/0x110
[ 1463.760753] wait_barrier+0xdd/0x170 [raid10]
[ 1463.760761] ? wait_woken+0x80/0x80
[ 1463.760768] raid10_write_request+0xf2/0x900 [raid10]
[ 1463.760774] ? wait_woken+0x80/0x80
[ 1463.760778] ? mempool_alloc+0x55/0x160
[ 1463.760795] ? md_write_start+0xa9/0x270 [md_mod]
[ 1463.760801] ? try_to_wake_up+0x44/0x470
[ 1463.760810] raid10_make_request+0xc1/0x120 [raid10]
[ 1463.760816] ? wait_woken+0x80/0x80
[ 1463.760831] md_handle_request+0x121/0x190 [md_mod]
[ 1463.760851] md_make_request+0x78/0x190 [md_mod]
[ 1463.760860] generic_make_request+0x1c6/0x470
[ 1463.760870] raid10_write_request+0x77a/0x900 [raid10]
[ 1463.760875] ? wait_woken+0x80/0x80
[ 1463.760879] ? mempool_alloc+0x55/0x160
[ 1463.760895] ? md_write_start+0xa9/0x270 [md_mod]
[ 1463.760904] raid10_make_request+0xc1/0x120 [raid10]
[ 1463.760910] ? wait_woken+0x80/0x80
[ 1463.760926] md_handle_request+0x121/0x190 [md_mod]
[ 1463.760931] ? _raw_spin_unlock_irq+0x22/0x40
[ 1463.760936] ? finish_task_switch+0x74/0x260
[ 1463.760954] submit_flushes+0x21/0x40 [md_mod]
So resync io is waiting for regular write io to complete to
decrease nr_pending (conf->barrier++ is called before waiting).
The regular write io splits another bio after call wait_barrier
which call nr_pending++, then the splitted bio would continue
with raid10_write_request -> wait_barrier, so the splitted bio
has to wait for barrier to be zero, then deadlock happens as
follows.
resync io regular io
raise_barrier
wait_barrier
generic_make_request
wait_barrier
To resolve the issue, we need to call allow_barrier to decrease
nr_pending before generic_make_request since regular IO is not
issued to underlying devices, and wait_barrier is called again
to ensure no internal IO happening.
Fixes: fc9977dd069e ("md/raid10: simplify the splitting of requests.")
Reported-and-tested-by: Siniša Bandin <sinisa@4net.rs>
Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
|
|
Both raid10_read_request and raid10_write_request share
the same code at the beginning of them, so introduce
regular_request_wait to clean up code, and call it in
both request functions.
Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
|
|
mempool_destroy() can handle NULL pointer correctly,
so there is no need to check NULL pointer before calling
mempool_destroy().
Signed-off-by: Chengguang Xu <cgxu519@gmx.com>
Signed-off-by: Shaohua Li <shli@fb.com>
|
|
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/md/md.c: In function 'md_integrity_add_rdev':
drivers/md/md.c:2149:24: warning:
variable 'bi_rdev' set but not used [-Wunused-but-set-variable]
It not used any more after commit
1501efadc524 ("md/raid: only permit hot-add of compatible integrity profiles")
Signed-off-by: Yue Haibing <yuehaibing@huawei.com>
Signed-off-by: Shaohua Li <shli@fb.com>
|
|
DM currently has a statically allocated bio that it uses to issue empty
flushes. It doesn't submit this bio, it just uses it for maintaining
state while setting up clones. Multiple users can access this bio at the
same time. This wasn't previously an issue, even if it was a bit iffy,
but with the blkg associations it can become one.
We setup the blkg association, then clone bio's and submit, then remove
the blkg assocation again. But since we can have multiple tasks doing
this at the same time, against multiple blkg's, then we can either lose
references to a blkg, or put it twice. The latter causes complaints on
the percpu ref being <= 0 when released, and can cause use-after-free as
well. Ming reports that xfstest generic/475 triggers this:
------------[ cut here ]------------
percpu ref (blkg_release) <= 0 (0) after switching to atomic
WARNING: CPU: 13 PID: 0 at lib/percpu-refcount.c:155 percpu_ref_switch_to_atomic_rcu+0x2c9/0x4a0
Switch to just using an on-stack bio for this, and get rid of the
embedded bio.
Fixes: 5cdf2e3fea5e ("blkcg: associate blkg when associating a device")
Reported-by: Ming Lei <ming.lei@redhat.com>
Tested-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Update DM to set the bdi's io_pages. This fixes reads to be capped at
the device's max request size (even if user's read IO exceeds the
established readahead setting).
Fixes: 9491ae4a ("mm: don't cap request size based on read-ahead setting")
Cc: stable@vger.kernel.org
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Sending a check/repair message infrequently leads to -EBUSY instead of
properly identifying an active resync. This occurs because
raid_message() is testing recovery bits in a racy way.
Fix by calling decipher_sync_action() from raid_message() to properly
identify the idle state of the RAID device.
Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
When commit 6a23e05c2fe3c6 ("dm: remove legacy request-based IO path")
removed some q->mq_ops branching from map_request() it left in place a
goto that was only needed if that branching (and conditional 'r'
assignment) existed. Now that the branching is gone map_request()'s
goto can be removed too.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Log the hash algorithm's driver name when a dm-verity target is created.
This will help people determine whether the expected implementation is
being used. It can make an enormous difference; e.g., SHA-256 on ARM
can be 8x faster with the crypto extensions than without. It can also
be useful to know if an implementation using an external crypto
accelerator is being used instead of a software implementation.
Example message:
[ 35.281945] device-mapper: verity: sha256 using implementation "sha256-ce"
We've already found the similar message in fs/crypto/keyinfo.c to be
very useful.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Log the encryption algorithm's driver name when a dm-crypt target is
created. This will help people determine whether the expected
implementation is being used. In some cases we've seen people do
benchmarks and reject using encryption for performance reasons, when in
fact they used a much slower implementation than was possible on the
hardware. It can make an enormous difference; e.g., AES-XTS on ARM can
be over 10x faster with the crypto extensions than without. It can also
be useful to know if an implementation using an external crypto
accelerator is being used instead of a software implementation.
Example message:
[ 29.307629] device-mapper: crypt: xts(aes) using implementation "xts-aes-ce"
We've already found the similar message in fs/crypto/keyinfo.c to be
very useful.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Rename the workqueue from dm-intergrity-recalc to dm-integrity-recalc.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
The flakey target is documented to be able to corrupt the Nth byte in
a bio, but does not corrupt byte indices after the first biovec in the
bio. Change the corrupting function to actually corrupt the Nth byte
no matter in which biovec that index falls.
A test device generating two-page bios, atop a flakey device configured
to corrupt a byte index on the second page, verified both the failure
to corrupt before this patch and the expected corruption after this
change.
Signed-off-by: John Dorminy <jdorminy@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Reference to a device in device-mapper table contains offset in sectors.
If the sector_t is 32bit integer (CONFIG_LBDAF is not set), then
several device-mapper targets can overflow this offset and validity
check is then performed on a wrong offset and a wrong table is activated.
See for example (on 32bit without CONFIG_LBDAF) this overflow:
# dmsetup create test --table "0 2048 linear /dev/sdg 4294967297"
# dmsetup table test
0 2048 linear 8:96 1
This patch adds explicit check for overflow if the offset is sector_t type.
Signed-off-by: Milan Broz <gmazyland@gmail.com>
Reviewed-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
The iv_offset in the mapping table of crypt target is a 64bit number when
IV algorithm is plain64, plain64be, essiv or benbi. It will be assigned to
iv_offset of struct crypt_config, cc_sector of struct convert_context and
iv_sector of struct dm_crypt_request. These structures members are defined
as a sector_t. But sector_t is 32bit when CONFIG_LBDAF is not set in 32bit
kernel. In this situation sector_t is not big enough to store the 64bit
iv_offset.
Here is a reproducer.
Prepare test image and device (loop is automatically allocated by cryptsetup):
# dd if=/dev/zero of=tst.img bs=1M count=1
# echo "tst"|cryptsetup open --type plain -c aes-xts-plain64 \
--skip 500000000000000000 tst.img test
On 32bit system (use IV offset value that overflows to 64bit; CONFIG_LBDAF if off)
and device checksum is wrong:
# dmsetup table test --showkeys
0 2048 crypt aes-xts-plain64 dfa7cfe3c481f2239155739c42e539ae8f2d38f304dcc89d20b26f69daaf0933 3551657984 7:0 0
# sha256sum /dev/mapper/test
533e25c09176632b3794f35303488c4a8f3f965dffffa6ec2df347c168cb6c19 /dev/mapper/test
On 64bit system (and on 32bit system with the patch), table and checksum is now correct:
# dmsetup table test --showkeys
0 2048 crypt aes-xts-plain64 dfa7cfe3c481f2239155739c42e539ae8f2d38f304dcc89d20b26f69daaf0933 500000000000000000 7:0 0
# sha256sum /dev/mapper/test
5d16160f9d5f8c33d8051e65fdb4f003cc31cd652b5abb08f03aa6fce0df75fc /dev/mapper/test
Signed-off-by: AliOS system security <alios_sys_security@linux.alibaba.com>
Tested-and-Reviewed-by: Milan Broz <gmazyland@gmail.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
When using kcopyd to run callbacks through dm_kcopyd_do_callback() or
submitting copy jobs with a source size of 0, the jobs are pushed
directly to the complete_jobs list, which could be under processing by
the kcopyd thread. As a result, the kcopyd thread can continue running
completed jobs indefinitely, without releasing the CPU, as long as
someone keeps submitting new completed jobs through the aforementioned
paths. Processing of work items, queued for execution on the same CPU as
the currently running kcopyd thread, is thus stalled for excessive
amounts of time, hurting performance.
Running the following test, from the device mapper test suite [1],
dmtest run --suite snapshot -n parallel_io_to_many_snaps_N
, with 8 active snapshots, we get, in dmesg, messages like the
following:
[68899.948523] BUG: workqueue lockup - pool cpus=0 node=0 flags=0x0 nice=0 stuck for 95s!
[68899.949282] Showing busy workqueues and worker pools:
[68899.949288] workqueue events: flags=0x0
[68899.949295] pwq 0: cpus=0 node=0 flags=0x0 nice=0 active=2/256
[68899.949306] pending: vmstat_shepherd, cache_reap
[68899.949331] workqueue mm_percpu_wq: flags=0x8
[68899.949337] pwq 0: cpus=0 node=0 flags=0x0 nice=0 active=1/256
[68899.949345] pending: vmstat_update
[68899.949387] workqueue dm_bufio_cache: flags=0x8
[68899.949392] pwq 4: cpus=2 node=0 flags=0x0 nice=0 active=1/256
[68899.949400] pending: work_fn [dm_bufio]
[68899.949423] workqueue kcopyd: flags=0x8
[68899.949429] pwq 0: cpus=0 node=0 flags=0x0 nice=0 active=1/256
[68899.949437] pending: do_work [dm_mod]
[68899.949452] workqueue kcopyd: flags=0x8
[68899.949458] pwq 0: cpus=0 node=0 flags=0x0 nice=0 active=2/256
[68899.949466] in-flight: 13:do_work [dm_mod]
[68899.949474] pending: do_work [dm_mod]
[68899.949487] workqueue kcopyd: flags=0x8
[68899.949493] pwq 0: cpus=0 node=0 flags=0x0 nice=0 active=1/256
[68899.949501] pending: do_work [dm_mod]
[68899.949515] workqueue kcopyd: flags=0x8
[68899.949521] pwq 0: cpus=0 node=0 flags=0x0 nice=0 active=1/256
[68899.949529] pending: do_work [dm_mod]
[68899.949541] workqueue kcopyd: flags=0x8
[68899.949547] pwq 0: cpus=0 node=0 flags=0x0 nice=0 active=1/256
[68899.949555] pending: do_work [dm_mod]
[68899.949568] pool 0: cpus=0 node=0 flags=0x0 nice=0 hung=95s workers=4 idle: 27130 27223 1084
Fix this by splitting the complete_jobs list into two parts: A user
facing part, named callback_jobs, and one used internally by kcopyd,
retaining the name complete_jobs. dm_kcopyd_do_callback() and
dispatch_job() now push their jobs to the callback_jobs list, which is
spliced to the complete_jobs list once, every time the kcopyd thread
wakes up. This prevents kcopyd from hogging the CPU indefinitely and
causing workqueue stalls.
Re-running the aforementioned test:
* Workqueue stalls are eliminated
* The maximum writing time among all targets is reduced from 09m37.10s
to 06m04.85s and the total run time of the test is reduced from
10m43.591s to 7m19.199s
[1] https://github.com/jthornber/device-mapper-test-suite
Signed-off-by: Nikos Tsironis <ntsironis@arrikto.com>
Signed-off-by: Ilias Tsitsimpis <iliastsi@arrikto.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
kcopyd has no upper limit to the number of jobs one can allocate and
issue. Under certain workloads this can lead to excessive memory usage
and workqueue stalls. For example, when creating multiple dm-snapshot
targets with a 4K chunk size and then writing to the origin through the
page cache. Syncing the page cache causes a large number of BIOs to be
issued to the dm-snapshot origin target, which itself issues an even
larger (because of the BIO splitting taking place) number of kcopyd
jobs.
Running the following test, from the device mapper test suite [1],
dmtest run --suite snapshot -n many_snapshots_of_same_volume_N
, with 8 active snapshots, results in the kcopyd job slab cache growing
to 10G. Depending on the available system RAM this can lead to the OOM
killer killing user processes:
[463.492878] kthreadd invoked oom-killer: gfp_mask=0x6040c0(GFP_KERNEL|__GFP_COMP),
nodemask=(null), order=1, oom_score_adj=0
[463.492894] kthreadd cpuset=/ mems_allowed=0
[463.492948] CPU: 7 PID: 2 Comm: kthreadd Not tainted 4.19.0-rc7 #3
[463.492950] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014
[463.492952] Call Trace:
[463.492964] dump_stack+0x7d/0xbb
[463.492973] dump_header+0x6b/0x2fc
[463.492987] ? lockdep_hardirqs_on+0xee/0x190
[463.493012] oom_kill_process+0x302/0x370
[463.493021] out_of_memory+0x113/0x560
[463.493030] __alloc_pages_slowpath+0xf40/0x1020
[463.493055] __alloc_pages_nodemask+0x348/0x3c0
[463.493067] cache_grow_begin+0x81/0x8b0
[463.493072] ? cache_grow_begin+0x874/0x8b0
[463.493078] fallback_alloc+0x1e4/0x280
[463.493092] kmem_cache_alloc_node+0xd6/0x370
[463.493098] ? copy_process.part.31+0x1c5/0x20d0
[463.493105] copy_process.part.31+0x1c5/0x20d0
[463.493115] ? __lock_acquire+0x3cc/0x1550
[463.493121] ? __switch_to_asm+0x34/0x70
[463.493129] ? kthread_create_worker_on_cpu+0x70/0x70
[463.493135] ? finish_task_switch+0x90/0x280
[463.493165] _do_fork+0xe0/0x6d0
[463.493191] ? kthreadd+0x19f/0x220
[463.493233] kernel_thread+0x25/0x30
[463.493235] kthreadd+0x1bf/0x220
[463.493242] ? kthread_create_on_cpu+0x90/0x90
[463.493248] ret_from_fork+0x3a/0x50
[463.493279] Mem-Info:
[463.493285] active_anon:20631 inactive_anon:4831 isolated_anon:0
[463.493285] active_file:80216 inactive_file:80107 isolated_file:435
[463.493285] unevictable:0 dirty:51266 writeback:109372 unstable:0
[463.493285] slab_reclaimable:31191 slab_unreclaimable:3483521
[463.493285] mapped:526 shmem:4903 pagetables:1759 bounce:0
[463.493285] free:33623 free_pcp:2392 free_cma:0
...
[463.493489] Unreclaimable slab info:
[463.493513] Name Used Total
[463.493522] bio-6 1028KB 1028KB
[463.493525] bio-5 1028KB 1028KB
[463.493528] dm_snap_pending_exception 236783KB 243789KB
[463.493531] dm_exception 41KB 42KB
[463.493534] bio-4 1216KB 1216KB
[463.493537] bio-3 439396KB 439396KB
[463.493539] kcopyd_job 6973427KB 6973427KB
...
[463.494340] Out of memory: Kill process 1298 (ruby2.3) score 1 or sacrifice child
[463.494673] Killed process 1298 (ruby2.3) total-vm:435740kB, anon-rss:20180kB, file-rss:4kB, shmem-rss:0kB
[463.506437] oom_reaper: reaped process 1298 (ruby2.3), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB
Moreover, issuing a large number of kcopyd jobs results in kcopyd
hogging the CPU, while processing them. As a result, processing of work
items, queued for execution on the same CPU as the currently running
kcopyd thread, is stalled for long periods of time, hurting performance.
Running the aforementioned test we get, in dmesg, messages like the
following:
[67501.194592] BUG: workqueue lockup - pool cpus=4 node=0 flags=0x0 nice=0 stuck for 27s!
[67501.195586] Showing busy workqueues and worker pools:
[67501.195591] workqueue events: flags=0x0
[67501.195597] pwq 8: cpus=4 node=0 flags=0x0 nice=0 active=1/256
[67501.195611] pending: cache_reap
[67501.195641] workqueue mm_percpu_wq: flags=0x8
[67501.195645] pwq 8: cpus=4 node=0 flags=0x0 nice=0 active=1/256
[67501.195656] pending: vmstat_update
[67501.195682] workqueue kblockd: flags=0x18
[67501.195687] pwq 5: cpus=2 node=0 flags=0x0 nice=-20 active=1/256
[67501.195698] pending: blk_timeout_work
[67501.195753] workqueue kcopyd: flags=0x8
[67501.195757] pwq 8: cpus=4 node=0 flags=0x0 nice=0 active=1/256
[67501.195768] pending: do_work [dm_mod]
[67501.195802] workqueue kcopyd: flags=0x8
[67501.195806] pwq 8: cpus=4 node=0 flags=0x0 nice=0 active=1/256
[67501.195817] pending: do_work [dm_mod]
[67501.195834] workqueue kcopyd: flags=0x8
[67501.195838] pwq 8: cpus=4 node=0 flags=0x0 nice=0 active=1/256
[67501.195848] pending: do_work [dm_mod]
[67501.195881] workqueue kcopyd: flags=0x8
[67501.195885] pwq 8: cpus=4 node=0 flags=0x0 nice=0 active=1/256
[67501.195896] pending: do_work [dm_mod]
[67501.195920] workqueue kcopyd: flags=0x8
[67501.195924] pwq 8: cpus=4 node=0 flags=0x0 nice=0 active=2/256
[67501.195935] in-flight: 67:do_work [dm_mod]
[67501.195945] pending: do_work [dm_mod]
[67501.195961] pool 8: cpus=4 node=0 flags=0x0 nice=0 hung=27s workers=3 idle: 129 23765
The root cause for these issues is the way dm-snapshot uses kcopyd. In
particular, the lack of an explicit or implicit limit to the maximum
number of in-flight COW jobs. The merging path is not affected because
it implicitly limits the in-flight kcopyd jobs to one.
Fix these issues by using a semaphore to limit the maximum number of
in-flight kcopyd jobs. We grab the semaphore before allocating a new
kcopyd job in start_copy() and start_full_bio() and release it after the
job finishes in copy_callback().
The initial semaphore value is configurable through a module parameter,
to allow fine tuning the maximum number of in-flight COW jobs. Setting
this parameter to zero initializes the semaphore to INT_MAX.
A default value of 2048 maximum in-flight kcopyd jobs was chosen. This
value was decided experimentally as a trade-off between memory
consumption, stalling the kernel's workqueues and maintaining a high
enough throughput.
Re-running the aforementioned test:
* Workqueue stalls are eliminated
* kcopyd's job slab cache uses a maximum of 130MB
* The time taken by the test to write to the snapshot-origin target is
reduced from 05m20.48s to 03m26.38s
[1] https://github.com/jthornber/device-mapper-test-suite
Signed-off-by: Nikos Tsironis <ntsironis@arrikto.com>
Signed-off-by: Ilias Tsitsimpis <iliastsi@arrikto.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
* Hashtable has been replaced by rbtree to manage buffers.
Update the comment.
* Fix typo in the comment for dm_bufio_issue_flush
Signed-off-by: Shenghui Wang <shhuiw@foxmail.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
The error msg should be "flush thread" instead of "endio thread"
for writecache_flush_thread.
Signed-off-by: Shenghui Wang <shhuiw@foxmail.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
No need to be so fancy.
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
The workqueues are shared by many multipath devices, only flush whole
workqueue when necessary. Otherwise, we just flush works as needed.
Signed-off-by: wuzhouhui <wuzhouhui14@mails.ucas.ac.cn>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
Indirect calls are inefficient because of retpolines that are used for
spectre workaround. This patch replaces an indirect call with a condition
(that can be predicted by the branch predictor).
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
|
|
There's a single user of this function, dm, and dm just wants
to check if IO is inflight, not that it's just allocated.
This fixes a hang with srp/002 in blktests with dm, where it tries
to suspend but waits for inflight IO to finish first. As it checks
for just allocated requests, this fails.
Tested-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Sometimes flush journal may be very frequent, so it's useful to dump
number of keys every time write journal.
Signed-off-by: Guoju Fang <fangguoju@gmail.com>
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Because CUTOFF_WRITEBACK is defined as 40, so before the changes of
dynamic cutoff writeback values, writeback_percent is limited to [0,
CUTOFF_WRITEBACK]. Any value larger than CUTOFF_WRITEBACK will be fixed
up to 40.
Now cutof writeback limit is a dynamic value bch_cutoff_writeback, so
the range of writeback_percent can be a more flexible range as [0,
bch_cutoff_writeback]. The flexibility is, it can be expended to a
larger or smaller range than [0, 40], depends on how value
bch_cutoff_writeback is specified.
The default value is still strongly recommended to most of users for
most of workloads. But for people who want to do research on bcache
writeback perforamnce tuning, they may have chance to specify more
flexible writeback_percent in range [0, 70].
Signed-off-by: Coly Li <colyli@suse.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
