| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
reshape
During a reshape or a RAID6 array such as expanding by adding an additional
disk, I/Os to the region of the array which have not yet been reshaped can
stall indefinitely. This is from errors in the stripe_ahead_of_reshape
function causing md to think the I/O is to a region in the actively
undergoing the reshape.
stripe_ahead_of_reshape fails to account for the q disk having a sector
value of 0. By not excluding the q disk from the for loop, raid6 will always
generate a min_sector value of 0, causing a return value which stalls.
The function's max_sector calculation also uses min() when it should use
max(), causing the max_sector value to always be 0. During a backwards
rebuild this can cause the opposite problem where it allows I/O to advance
when it should wait.
Fixing these errors will allow safe I/O to advance in a timely manner and
delay only I/O which is unsafe due to stripes in the middle of undergoing
the reshape.
Fixes: 486f60558607 ("md/raid5: Check all disks in a stripe_head for reshape progress")
Cc: stable@vger.kernel.org # v6.0+
Signed-off-by: David Jeffery <djeffery@redhat.com>
Tested-by: Laurence Oberman <loberman@redhat.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20231128181233.6187-1-djeffery@redhat.com
|
|
Pull block fixes from Jens Axboe:
- NVMe pull request via Keith:
- Invalid namespace identification error handling (Marizio Ewan,
Keith)
- Fabrics keep-alive tuning (Mark)
- Fix for a bad error check regression in bcache (Markus)
- Fix for a performance regression with O_DIRECT (Ming)
- Fix for a flush related deadlock (Ming)
- Make the read-only warn on per-partition (Yu)
* tag 'block-6.7-2023-12-01' of git://git.kernel.dk/linux:
nvme-core: check for too small lba shift
blk-mq: don't count completed flush data request as inflight in case of quiesce
block: Document the role of the two attribute groups
block: warn once for each partition in bio_check_ro()
block: move .bd_inode into 1st cacheline of block_device
nvme: check for valid nvme_identify_ns() before using it
nvme-core: fix a memory leak in nvme_ns_info_from_identify()
nvme: fine-tune sending of first keep-alive
bcache: revert replacing IS_ERR_OR_NULL with IS_ERR
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm
Pull device mapper fixes from Mike Snitzer:
- Fix DM verity target's FEC support to always initialize IO before it
frees it. Also fix alignment of struct dm_verity_fec_io within the
per-bio-data
- Fix DM verity target to not FEC failed readahead IO
- Update DM flakey target to use MAX_ORDER rather than MAX_ORDER - 1
* tag 'dm-6.7/dm-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
dm-flakey: start allocating with MAX_ORDER
dm-verity: align struct dm_verity_fec_io properly
dm verity: don't perform FEC for failed readahead IO
dm verity: initialize fec io before freeing it
|
|
Pull more bcachefs bugfixes from Kent Overstreet:
- bcache & bcachefs were broken with CFI enabled; patch for closures to
fix type punning
- mark erasure coding as extra-experimental; there are incompatible
disk space accounting changes coming for erasure coding, and I'm
still seeing checksum errors in some tests
- several fixes for durability-related issues (durability is a device
specific setting where we can tell bcachefs that data on a given
device should be counted as replicated x times)
- a fix for a rare livelock when a btree node merge then updates a
parent node that is almost full
- fix a race in the device removal path, where dropping a pointer in a
btree node to a device would be clobbered by an in flight btree write
updating the btree node key on completion
- fix one SRCU lock hold time warning in the btree gc code - ther's
still a bunch more of these to fix
- fix a rare race where we'd start copygc before initializing the "are
we rw" percpu refcount; copygc would think we were already ro and die
immediately
* tag 'bcachefs-2023-11-29' of https://evilpiepirate.org/git/bcachefs: (23 commits)
bcachefs: Extra kthread_should_stop() calls for copygc
bcachefs: Convert gc_alloc_start() to for_each_btree_key2()
bcachefs: Fix race between btree writes and metadata drop
bcachefs: move journal seq assertion
bcachefs: -EROFS doesn't count as move_extent_start_fail
bcachefs: trace_move_extent_start_fail() now includes errcode
bcachefs: Fix split_race livelock
bcachefs: Fix bucket data type for stripe buckets
bcachefs: Add missing validation for jset_entry_data_usage
bcachefs: Fix zstd compress workspace size
bcachefs: bpos is misaligned on big endian
bcachefs: Fix ec + durability calculation
bcachefs: Data update path won't accidentaly grow replicas
bcachefs: deallocate_extra_replicas()
bcachefs: Proper refcounting for journal_keys
bcachefs: preserve device path as device name
bcachefs: Fix an endianness conversion
bcachefs: Start gc, copygc, rebalance threads after initing writes ref
bcachefs: Don't stop copygc thread on device resize
bcachefs: Make sure bch2_move_ratelimit() also waits for move_ops
...
|
|
Commit 23baf831a32c ("mm, treewide: redefine MAX_ORDER sanely")
changed the meaning of MAX_ORDER from exclusive to inclusive. So, we
can allocate compound pages with up to 1 << MAX_ORDER pages.
Reflect this change in dm-flakey and start trying to allocate compound
pages with MAX_ORDER.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
dm_verity_fec_io is placed after the end of two hash digests. If the hash
digest has unaligned length, struct dm_verity_fec_io could be unaligned.
This commit fixes the placement of struct dm_verity_fec_io, so that it's
aligned.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Fixes: a739ff3f543a ("dm verity: add support for forward error correction")
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
We found an issue under Android OTA scenario that many BIOs have to do
FEC where the data under dm-verity is 100% complete and no corruption.
Android OTA has many dm-block layers, from upper to lower:
dm-verity
dm-snapshot
dm-origin & dm-cow
dm-linear
ufs
DM tables have to change 2 times during Android OTA merging process.
When doing table change, the dm-snapshot will be suspended for a while.
During this interval, many readahead IOs are submitted to dm_verity
from filesystem. Then the kverity works are busy doing FEC process
which cost too much time to finish dm-verity IO. This causes needless
delay which feels like system is hung.
After adding debugging it was found that each readahead IO needed
around 10s to finish when this situation occurred. This is due to IO
amplification:
dm-snapshot suspend
erofs_readahead // 300+ io is submitted
dm_submit_bio (dm_verity)
dm_submit_bio (dm_snapshot)
bio return EIO
bio got nothing, it's empty
verity_end_io
verity_verify_io
forloop range(0, io->n_blocks) // each io->nblocks ~= 20
verity_fec_decode
fec_decode_rsb
fec_read_bufs
forloop range(0, v->fec->rsn) // v->fec->rsn = 253
new_read
submit_bio (dm_snapshot)
end loop
end loop
dm-snapshot resume
Readahead BIOs get nothing while dm-snapshot is suspended, so all of
them will cause verity's FEC.
Each readahead BIO needs to verify ~20 (io->nblocks) blocks.
Each block needs to do FEC, and every block needs to do 253
(v->fec->rsn) reads.
So during the suspend interval(~200ms), 300 readahead BIOs trigger
~1518000 (300*20*253) IOs to dm-snapshot.
As readahead IO is not required by userspace, and to fix this issue,
it is best to pass readahead errors to upper layer to handle it.
Cc: stable@vger.kernel.org
Fixes: a739ff3f543a ("dm verity: add support for forward error correction")
Signed-off-by: Wu Bo <bo.wu@vivo.com>
Reviewed-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
If BIO error, verity_end_io() can call verity_finish_io() before
verity_fec_init_io(). Therefore, fec_io->rs is not initialized and
may crash when doing memory freeing in verity_fec_finish_io().
Crash call stack:
die+0x90/0x2b8
__do_kernel_fault+0x260/0x298
do_bad_area+0x2c/0xdc
do_translation_fault+0x3c/0x54
do_mem_abort+0x54/0x118
el1_abort+0x38/0x5c
el1h_64_sync_handler+0x50/0x90
el1h_64_sync+0x64/0x6c
free_rs+0x18/0xac
fec_rs_free+0x10/0x24
mempool_free+0x58/0x148
verity_fec_finish_io+0x4c/0xb0
verity_end_io+0xb8/0x150
Cc: stable@vger.kernel.org # v6.0+
Fixes: 5721d4e5a9cd ("dm verity: Add optional "try_verify_in_tasklet" feature")
Signed-off-by: Wu Bo <bo.wu@vivo.com>
Reviewed-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
Because it's safe to accees rdev from conf:
- If any spinlock is held, because synchronize_rcu() from
md_kick_rdev_from_array() will prevent 'rdev' to be freed until
spinlock is released;
- If there is normal IO inflight, because mddev_suspend() will prevent
rdev to be added or removed from array;
And these will cover all the scenarios in md-multipath.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20231125081604.3939938-6-yukuai1@huaweicloud.com
|
|
Because it's safe to accees rdev from conf:
- If any spinlock is held, because synchronize_rcu() from
md_kick_rdev_from_array() will prevent 'rdev' to be freed until
spinlock is released;
- If 'reconfig_lock' is held, because rdev can't be added or removed from
array;
- If there is normal IO inflight, because mddev_suspend() will prevent
rdev to be added or removed from array;
- If there is sync IO inflight, because 'MD_RECOVERY_RUNNING' is
checked in remove_and_add_spares().
And these will cover all the scenarios in raid456.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20231125081604.3939938-5-yukuai1@huaweicloud.com
|
|
Because it's safe to accees rdev from conf:
- If any spinlock is held, because synchronize_rcu() from
md_kick_rdev_from_array() will prevent 'rdev' to be freed until
spinlock is released;
- If 'reconfig_lock' is held, because rdev can't be added or removed from
array;
- If there is normal IO inflight, because mddev_suspend() will prevent
rdev to be added or removed from array;
- If there is sync IO inflight, because 'MD_RECOVERY_RUNNING' is
checked in remove_and_add_spares().
And these will cover all the scenarios in raid1.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20231125081604.3939938-4-yukuai1@huaweicloud.com
|
|
Because it's safe to accees rdev from conf:
- If any spinlock is held, because synchronize_rcu() from
md_kick_rdev_from_array() will prevent 'rdev' to be freed until
spinlock is released;
- If 'reconfig_lock' is held, because rdev can't be added or removed from
array;
- If there is normal IO inflight, because mddev_suspend() will prevent
rdev to be added or removed from array;
- If there is sync IO inflight, because 'MD_RECOVERY_RUNNING' is
checked in remove_and_add_spares().
And these will cover all the scenarios in raid10.
This patch also cleanup the code to handle the case that replacement
replace rdev while IO is still inflight.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20231125081604.3939938-3-yukuai1@huaweicloud.com
|
|
rcu is not used correctly here, because synchronize_rcu() is called
before replacing old value, for example:
remove_and_add_spares // other path
synchronize_rcu
// called before replacing old value
set_bit(RemoveSynchronized)
rcu_read_lock()
rdev = conf->mirros[].rdev
pers->hot_remove_disk
conf->mirros[].rdev = NULL;
if (!test_bit(RemoveSynchronized))
synchronize_rcu
/*
* won't be called, and won't wait
* for concurrent readers to be done.
*/
// access rdev after remove_and_add_spares()
rcu_read_unlock()
Fortunately, there is a separate rcu protection to prevent such rdev
to be freed:
md_kick_rdev_from_array //other path
rcu_read_lock()
rdev = conf->mirros[].rdev
list_del_rcu(&rdev->same_set)
rcu_read_unlock()
/*
* rdev can be removed from conf, but
* rdev won't be freed.
*/
synchronize_rcu()
free rdev
Hence remove this useless flag and prepare to remove rcu protection to
access rdev from 'conf'.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20231125081604.3939938-2-yukuai1@huaweicloud.com
|
|
This reverts commit 5e2cf333b7bd5d3e62595a44d598a254c697cd74.
That commit introduced the following race and can cause system hung.
md_write_start: raid5d:
// mddev->in_sync == 1
set "MD_SB_CHANGE_PENDING"
// running before md_write_start wakeup it
waiting "MD_SB_CHANGE_PENDING" cleared
>>>>>>>>> hung
wakeup mddev->thread
...
waiting "MD_SB_CHANGE_PENDING" cleared
>>>> hung, raid5d should clear this flag
but get hung by same flag.
The issue reverted commit fixing is fixed by last patch in a new way.
Fixes: 5e2cf333b7bd ("md/raid5: Wait for MD_SB_CHANGE_PENDING in raid5d")
Cc: stable@vger.kernel.org # v5.19+
Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>
Reviewed-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20231108182216.73611-2-junxiao.bi@oracle.com
|
|
commit 5e2cf333b7bd ("md/raid5: Wait for MD_SB_CHANGE_PENDING in raid5d")
introduced a hung bug and will be reverted in next patch, since the issue
that commit is fixing is due to md superblock write is throttled by wbt,
to fix it, we can have superblock write bypass block layer throttle.
Fixes: 5e2cf333b7bd ("md/raid5: Wait for MD_SB_CHANGE_PENDING in raid5d")
Cc: stable@vger.kernel.org # v5.19+
Suggested-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20231108182216.73611-1-junxiao.bi@oracle.com
|
|
Commit 028ddcac477b ("bcache: Remove unnecessary NULL point check in
node allocations") replaced IS_ERR_OR_NULL by IS_ERR. This leads to a
NULL pointer dereference.
BUG: kernel NULL pointer dereference, address: 0000000000000080
Call Trace:
? __die_body.cold+0x1a/0x1f
? page_fault_oops+0xd2/0x2b0
? exc_page_fault+0x70/0x170
? asm_exc_page_fault+0x22/0x30
? btree_node_free+0xf/0x160 [bcache]
? up_write+0x32/0x60
btree_gc_coalesce+0x2aa/0x890 [bcache]
? bch_extent_bad+0x70/0x170 [bcache]
btree_gc_recurse+0x130/0x390 [bcache]
? btree_gc_mark_node+0x72/0x230 [bcache]
bch_btree_gc+0x5da/0x600 [bcache]
? cpuusage_read+0x10/0x10
? bch_btree_gc+0x600/0x600 [bcache]
bch_gc_thread+0x135/0x180 [bcache]
The relevant code starts with:
new_nodes[0] = NULL;
for (i = 0; i < nodes; i++) {
if (__bch_keylist_realloc(&keylist, bkey_u64s(&r[i].b->key)))
goto out_nocoalesce;
// ...
out_nocoalesce:
// ...
for (i = 0; i < nodes; i++)
if (!IS_ERR(new_nodes[i])) { // IS_ERR_OR_NULL before
028ddcac477b
btree_node_free(new_nodes[i]); // new_nodes[0] is NULL
rw_unlock(true, new_nodes[i]);
}
This patch replaces IS_ERR() by IS_ERR_OR_NULL() to fix this.
Fixes: 028ddcac477b ("bcache: Remove unnecessary NULL point check in node allocations")
Link: https://lore.kernel.org/all/3DF4A87A-2AC1-4893-AE5F-E921478419A9@suse.de/
Cc: stable@vger.kernel.org
Cc: Zheng Wang <zyytlz.wz@163.com>
Cc: Coly Li <colyli@suse.de>
Signed-off-by: Markus Weippert <markus@gekmihesg.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Control flow integrity is now checking that type signatures match on
indirect function calls. That breaks closures, which embed a work_struct
in a closure in such a way that a closure_fn may also be used as a
workqueue fn by the underlying closure code.
So we have to change closure fns to take a work_struct as their
argument - but that results in a loss of clarity, as closure fns have
different semantics from normal workqueue functions (they run owning a
ref on the closure, which must be released with continue_at() or
closure_return()).
Thus, this patc introduces CLOSURE_CALLBACK() and closure_type() macros
as suggested by Kees, to smooth things over a bit.
Suggested-by: Kees Cook <keescook@chromium.org>
Cc: Coly Li <colyli@suse.de>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
|
|
Pull block fixes from Jens Axboe:
"A bit bigger than usual at this time, but nothing really earth
shattering:
- NVMe pull request via Keith:
- TCP TLS fixes (Hannes)
- Authentifaction fixes (Mark, Hannes)
- Properly terminate target names (Christoph)
- MD pull request via Song, fixing a raid5 corruption issue
- Disentanglement of the dependency mess in nvme introduced with the
tls additions. Now it should actually build on all configs (Arnd)
- Series of bcache fixes (Coly)
- Removal of a dead helper (Damien)
- s390 dasd fix (Muhammad, Jan)
- lockdep blk-cgroup fixes (Ming)"
* tag 'block-6.7-2023-11-23' of git://git.kernel.dk/linux: (33 commits)
nvme: tcp: fix compile-time checks for TLS mode
nvme: target: fix Kconfig select statements
nvme: target: fix nvme_keyring_id() references
nvme: move nvme_stop_keep_alive() back to original position
nbd: pass nbd_sock to nbd_read_reply() instead of index
s390/dasd: protect device queue against concurrent access
s390/dasd: resolve spelling mistake
block/null_blk: Fix double blk_mq_start_request() warning
nvmet-tcp: always initialize tls_handshake_tmo_work
nvmet: nul-terminate the NQNs passed in the connect command
nvme: blank out authentication fabrics options if not configured
nvme: catch errors from nvme_configure_metadata()
nvme-tcp: only evaluate 'tls' option if TLS is selected
nvme-auth: set explanation code for failure2 msgs
nvme-auth: unlock mutex in one place only
block: Remove blk_set_runtime_active()
nbd: fix null-ptr-dereference while accessing 'nbd->config'
nbd: factor out a helper to get nbd_config without holding 'config_lock'
nbd: fold nbd config initialization into nbd_alloc_config()
bcache: avoid NULL checking to c->root in run_cache_set()
...
|
|
https://git.kernel.org/pub/scm/linux/kernel/git/song/md into block-6.7
Pull MD fix from Song.
* tag 'md-fixes-20231120' of https://git.kernel.org/pub/scm/linux/kernel/git/song/md:
md: fix bi_status reporting in md_end_clone_io
|
|
In run_cache_set() after c->root returned from bch_btree_node_get(), it
is checked by IS_ERR_OR_NULL(). Indeed it is unncessary to check NULL
because bch_btree_node_get() will not return NULL pointer to caller.
This patch replaces IS_ERR_OR_NULL() by IS_ERR() for the above reason.
Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20231120052503.6122-11-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
This patch adds code comments to bch_btree_node_get() and
__bch_btree_node_alloc() that NULL pointer will not be returned and it
is unnecessary to check NULL pointer by the callers of these routines.
Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20231120052503.6122-10-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Commit 028ddcac477b ("bcache: Remove unnecessary NULL point check in
node allocations") do the following change inside btree_gc_coalesce(),
31 @@ -1340,7 +1340,7 @@ static int btree_gc_coalesce(
32 memset(new_nodes, 0, sizeof(new_nodes));
33 closure_init_stack(&cl);
34
35 - while (nodes < GC_MERGE_NODES && !IS_ERR_OR_NULL(r[nodes].b))
36 + while (nodes < GC_MERGE_NODES && !IS_ERR(r[nodes].b))
37 keys += r[nodes++].keys;
38
39 blocks = btree_default_blocks(b->c) * 2 / 3;
At line 35 the original r[nodes].b is not always allocatored from
__bch_btree_node_alloc(), and possibly initialized as NULL pointer by
caller of btree_gc_coalesce(). Therefore the change at line 36 is not
correct.
This patch replaces the mistaken IS_ERR() by IS_ERR_OR_NULL() to avoid
potential issue.
Fixes: 028ddcac477b ("bcache: Remove unnecessary NULL point check in node allocations")
Cc: <stable@vger.kernel.org> # 6.5+
Cc: Zheng Wang <zyytlz.wz@163.com>
Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20231120052503.6122-9-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We get a kernel crash about "unable to handle kernel paging request":
```dmesg
[368033.032005] BUG: unable to handle kernel paging request at ffffffffad9ae4b5
[368033.032007] PGD fc3a0d067 P4D fc3a0d067 PUD fc3a0e063 PMD 8000000fc38000e1
[368033.032012] Oops: 0003 [#1] SMP PTI
[368033.032015] CPU: 23 PID: 55090 Comm: bch_dirtcnt[0] Kdump: loaded Tainted: G OE --------- - - 4.18.0-147.5.1.es8_24.x86_64 #1
[368033.032017] Hardware name: Tsinghua Tongfang THTF Chaoqiang Server/072T6D, BIOS 2.4.3 01/17/2017
[368033.032027] RIP: 0010:native_queued_spin_lock_slowpath+0x183/0x1d0
[368033.032029] Code: 8b 02 48 85 c0 74 f6 48 89 c1 eb d0 c1 e9 12 83 e0
03 83 e9 01 48 c1 e0 05 48 63 c9 48 05 c0 3d 02 00 48 03 04 cd 60 68 93
ad <48> 89 10 8b 42 08 85 c0 75 09 f3 90 8b 42 08 85 c0 74 f7 48 8b 02
[368033.032031] RSP: 0018:ffffbb48852abe00 EFLAGS: 00010082
[368033.032032] RAX: ffffffffad9ae4b5 RBX: 0000000000000246 RCX: 0000000000003bf3
[368033.032033] RDX: ffff97b0ff8e3dc0 RSI: 0000000000600000 RDI: ffffbb4884743c68
[368033.032034] RBP: 0000000000000001 R08: 0000000000000000 R09: 000007ffffffffff
[368033.032035] R10: ffffbb486bb01000 R11: 0000000000000001 R12: ffffffffc068da70
[368033.032036] R13: 0000000000000003 R14: 0000000000000000 R15: 0000000000000000
[368033.032038] FS: 0000000000000000(0000) GS:ffff97b0ff8c0000(0000) knlGS:0000000000000000
[368033.032039] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[368033.032040] CR2: ffffffffad9ae4b5 CR3: 0000000fc3a0a002 CR4: 00000000003626e0
[368033.032042] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[368033.032043] bcache: bch_cached_dev_attach() Caching rbd479 as bcache462 on set 8cff3c36-4a76-4242-afaa-7630206bc70b
[368033.032045] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[368033.032046] Call Trace:
[368033.032054] _raw_spin_lock_irqsave+0x32/0x40
[368033.032061] __wake_up_common_lock+0x63/0xc0
[368033.032073] ? bch_ptr_invalid+0x10/0x10 [bcache]
[368033.033502] bch_dirty_init_thread+0x14c/0x160 [bcache]
[368033.033511] ? read_dirty_submit+0x60/0x60 [bcache]
[368033.033516] kthread+0x112/0x130
[368033.033520] ? kthread_flush_work_fn+0x10/0x10
[368033.034505] ret_from_fork+0x35/0x40
```
The crash occurred when call wake_up(&state->wait), and then we want
to look at the value in the state. However, bch_sectors_dirty_init()
is not found in the stack of any task. Since state is allocated on
the stack, we guess that bch_sectors_dirty_init() has exited, causing
bch_dirty_init_thread() to be unable to handle kernel paging request.
In order to verify this idea, we added some printing information during
wake_up(&state->wait). We find that "wake up" is printed twice, however
we only expect the last thread to wake up once.
```dmesg
[ 994.641004] alcache: bch_dirty_init_thread() wake up
[ 994.641018] alcache: bch_dirty_init_thread() wake up
[ 994.641523] alcache: bch_sectors_dirty_init() init exit
```
There is a race. If bch_sectors_dirty_init() exits after the first wake
up, the second wake up will trigger this bug("unable to handle kernel
paging request").
Proceed as follows:
bch_sectors_dirty_init
kthread_run ==============> bch_dirty_init_thread(bch_dirtcnt[0])
... ...
atomic_inc(&state.started) ...
... ...
atomic_read(&state.enough) ...
... atomic_set(&state->enough, 1)
kthread_run ======================================================> bch_dirty_init_thread(bch_dirtcnt[1])
... atomic_dec_and_test(&state->started) ...
atomic_inc(&state.started) ... ...
... wake_up(&state->wait) ...
atomic_read(&state.enough) atomic_dec_and_test(&state->started)
... ...
wait_event(state.wait, atomic_read(&state.started) == 0) ...
return ...
wake_up(&state->wait)
We believe it is very common to wake up twice if there is no dirty, but
crash is an extremely low probability event. It's hard for us to reproduce
this issue. We attached and detached continuously for a week, with a total
of more than one million attaches and only one crash.
Putting atomic_inc(&state.started) before kthread_run() can avoid waking
up twice.
Fixes: b144e45fc576 ("bcache: make bch_sectors_dirty_init() to be multithreaded")
Signed-off-by: Mingzhe Zou <mingzhe.zou@easystack.cn>
Cc: <stable@vger.kernel.org>
Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20231120052503.6122-8-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We had a problem with io hung because it was waiting for c->root to
release the lock.
crash> cache_set.root -l cache_set.list ffffa03fde4c0050
root = 0xffff802ef454c800
crash> btree -o 0xffff802ef454c800 | grep rw_semaphore
[ffff802ef454c858] struct rw_semaphore lock;
crash> struct rw_semaphore ffff802ef454c858
struct rw_semaphore {
count = {
counter = -4294967297
},
wait_list = {
next = 0xffff00006786fc28,
prev = 0xffff00005d0efac8
},
wait_lock = {
raw_lock = {
{
val = {
counter = 0
},
{
locked = 0 '\000',
pending = 0 '\000'
},
{
locked_pending = 0,
tail = 0
}
}
}
},
osq = {
tail = {
counter = 0
}
},
owner = 0xffffa03fdc586603
}
The "counter = -4294967297" means that lock count is -1 and a write lock
is being attempted. Then, we found that there is a btree with a counter
of 1 in btree_cache_freeable.
crash> cache_set -l cache_set.list ffffa03fde4c0050 -o|grep btree_cache
[ffffa03fde4c1140] struct list_head btree_cache;
[ffffa03fde4c1150] struct list_head btree_cache_freeable;
[ffffa03fde4c1160] struct list_head btree_cache_freed;
[ffffa03fde4c1170] unsigned int btree_cache_used;
[ffffa03fde4c1178] wait_queue_head_t btree_cache_wait;
[ffffa03fde4c1190] struct task_struct *btree_cache_alloc_lock;
crash> list -H ffffa03fde4c1140|wc -l
973
crash> list -H ffffa03fde4c1150|wc -l
1123
crash> cache_set.btree_cache_used -l cache_set.list ffffa03fde4c0050
btree_cache_used = 2097
crash> list -s btree -l btree.list -H ffffa03fde4c1140|grep -E -A2 "^ lock = {" > btree_cache.txt
crash> list -s btree -l btree.list -H ffffa03fde4c1150|grep -E -A2 "^ lock = {" > btree_cache_freeable.txt
[root@node-3 127.0.0.1-2023-08-04-16:40:28]# pwd
/var/crash/127.0.0.1-2023-08-04-16:40:28
[root@node-3 127.0.0.1-2023-08-04-16:40:28]# cat btree_cache.txt|grep counter|grep -v "counter = 0"
[root@node-3 127.0.0.1-2023-08-04-16:40:28]# cat btree_cache_freeable.txt|grep counter|grep -v "counter = 0"
counter = 1
We found that this is a bug in bch_sectors_dirty_init() when locking c->root:
(1). Thread X has locked c->root(A) write.
(2). Thread Y failed to lock c->root(A), waiting for the lock(c->root A).
(3). Thread X bch_btree_set_root() changes c->root from A to B.
(4). Thread X releases the lock(c->root A).
(5). Thread Y successfully locks c->root(A).
(6). Thread Y releases the lock(c->root B).
down_write locked ---(1)----------------------┐
| |
| down_read waiting ---(2)----┐ |
| | ┌-------------┐ ┌-------------┐
bch_btree_set_root ===(3)========>> | c->root A | | c->root B |
| | └-------------┘ └-------------┘
up_write ---(4)---------------------┘ | |
| | |
down_read locked ---(5)-----------┘ |
| |
up_read ---(6)-----------------------------┘
Since c->root may change, the correct steps to lock c->root should be
the same as bch_root_usage(), compare after locking.
static unsigned int bch_root_usage(struct cache_set *c)
{
unsigned int bytes = 0;
struct bkey *k;
struct btree *b;
struct btree_iter iter;
goto lock_root;
do {
rw_unlock(false, b);
lock_root:
b = c->root;
rw_lock(false, b, b->level);
} while (b != c->root);
for_each_key_filter(&b->keys, k, &iter, bch_ptr_bad)
bytes += bkey_bytes(k);
rw_unlock(false, b);
return (bytes * 100) / btree_bytes(c);
}
Fixes: b144e45fc576 ("bcache: make bch_sectors_dirty_init() to be multithreaded")
Signed-off-by: Mingzhe Zou <mingzhe.zou@easystack.cn>
Cc: <stable@vger.kernel.org>
Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20231120052503.6122-7-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
We found that after long run, the dirty_data of the bcache device
will have errors. This error cannot be eliminated unless re-register.
We also found that reattach after detach, this error can accumulate.
In bch_sectors_dirty_init(), all inode <= d->id keys will be recounted
again. This is wrong, we only need to count the keys of the current
device.
Fixes: b144e45fc576 ("bcache: make bch_sectors_dirty_init() to be multithreaded")
Signed-off-by: Mingzhe Zou <mingzhe.zou@easystack.cn>
Cc: <stable@vger.kernel.org>
Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20231120052503.6122-6-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
In SHOW(), the variable 'n' is of type 'size_t.' While there is a
conditional check to verify that 'n' is not equal to zero before
executing the 'do_div' macro, concerns arise regarding potential
division by zero error in 64-bit environments.
The concern arises when 'n' is 64 bits in size, greater than zero, and
the lower 32 bits of it are zeros. In such cases, the conditional check
passes because 'n' is non-zero, but the 'do_div' macro casts 'n' to
'uint32_t,' effectively truncating it to its lower 32 bits.
Consequently, the 'n' value becomes zero.
To fix this potential division by zero error and ensure precise
division handling, this commit replaces the 'do_div' macro with
div64_u64(). div64_u64() is designed to work with 64-bit operands,
guaranteeing that division is performed correctly.
This change enhances the robustness of the code, ensuring that division
operations yield accurate results in all scenarios, eliminating the
possibility of division by zero, and improving compatibility across
different 64-bit environments.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Signed-off-by: Rand Deeb <rand.sec96@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20231120052503.6122-5-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Variable cur_idx is being initialized with a value that is never read,
it is being re-assigned later in a while-loop. Remove the redundant
assignment. Cleans up clang scan build warning:
drivers/md/bcache/writeback.c:916:2: warning: Value stored to 'cur_idx'
is never read [deadcode.DeadStores]
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Reviewed-by: Coly Li <colyli@suse.de>
Signed-off-by: Coly Li <colyli@suse.de>
Link: https://lore.kernel.org/r/20231120052503.6122-4-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
In btree_gc_rewrite_node(), pointer 'n' is not checked after it returns
from btree_gc_rewrite_node(). There is potential possibility that 'n' is
a non NULL ERR_PTR(), referencing such error code is not permitted in
following code. Therefore a return value checking is necessary after 'n'
is back from btree_node_alloc_replacement().
Signed-off-by: Coly Li <colyli@suse.de>
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20231120052503.6122-3-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Arraies bcache->stripe_sectors_dirty and bcache->full_dirty_stripes are
used for dirty data writeback, their sizes are decided by backing device
capacity and stripe size. Larger backing device capacity or smaller
stripe size make these two arraies occupies more dynamic memory space.
Currently bcache->stripe_size is directly inherited from
queue->limits.io_opt of underlying storage device. For normal hard
drives, its limits.io_opt is 0, and bcache sets the corresponding
stripe_size to 1TB (1<<31 sectors), it works fine 10+ years. But for
devices do declare value for queue->limits.io_opt, small stripe_size
(comparing to 1TB) becomes an issue for oversize memory allocations of
bcache->stripe_sectors_dirty and bcache->full_dirty_stripes, while the
capacity of hard drives gets much larger in recent decade.
For example a raid5 array assembled by three 20TB hardrives, the raid
device capacity is 40TB with typical 512KB limits.io_opt. After the math
calculation in bcache code, these two arraies will occupy 400MB dynamic
memory. Even worse Andrea Tomassetti reports that a 4KB limits.io_opt is
declared on a new 2TB hard drive, then these two arraies request 2GB and
512MB dynamic memory from kzalloc(). The result is that bcache device
always fails to initialize on his system.
To avoid the oversize memory allocation, bcache->stripe_size should not
directly inherited by queue->limits.io_opt from the underlying device.
This patch defines BCH_MIN_STRIPE_SZ (4MB) as minimal bcache stripe size
and set bcache device's stripe size against the declared limits.io_opt
value from the underlying storage device,
- If the declared limits.io_opt > BCH_MIN_STRIPE_SZ, bcache device will
set its stripe size directly by this limits.io_opt value.
- If the declared limits.io_opt < BCH_MIN_STRIPE_SZ, bcache device will
set its stripe size by a value multiplying limits.io_opt and euqal or
large than BCH_MIN_STRIPE_SZ.
Then the minimal stripe size of a bcache device will always be >= 4MB.
For a 40TB raid5 device with 512KB limits.io_opt, memory occupied by
bcache->stripe_sectors_dirty and bcache->full_dirty_stripes will be 50MB
in total. For a 2TB hard drive with 4KB limits.io_opt, memory occupied
by these two arraies will be 2.5MB in total.
Such mount of memory allocated for bcache->stripe_sectors_dirty and
bcache->full_dirty_stripes is reasonable for most of storage devices.
Reported-by: Andrea Tomassetti <andrea.tomassetti-opensource@devo.com>
Signed-off-by: Coly Li <colyli@suse.de>
Reviewed-by: Eric Wheeler <bcache@lists.ewheeler.net>
Link: https://lore.kernel.org/r/20231120052503.6122-2-colyli@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
md_end_clone_io() may overwrite error status in orig_bio->bi_status with
BLK_STS_OK. This could happen when orig_bio has BIO_CHAIN (split by
md_submit_bio => bio_split_to_limits, for example). As a result, upper
layer may miss error reported from md (or the device) and consider the
failed IO was successful.
Fix this by only update orig_bio->bi_status when current bio reports
error and orig_bio is BLK_STS_OK. This is the same behavior as
__bio_chain_endio().
Fixes: 10764815ff47 ("md: add io accounting for raid0 and raid5")
Cc: stable@vger.kernel.org # v5.14+
Reported-by: Bhanu Victor DiCara <00bvd0+linux@gmail.com>
Closes: https://lore.kernel.org/regressions/5727380.DvuYhMxLoT@bvd0/
Signed-off-by: Song Liu <song@kernel.org>
Tested-by: Xiao Ni <xni@redhat.com>
Reviewed-by: Yu Kuai <yukuai3@huawei.com>
Acked-by: Guoqing Jiang <guoqing.jiang@linux.dev>
|
|
We have bdev_mark_dead() etc and we're going to move block device
freezing to holder ops in the next patch. Make the naming consistent:
* freeze_bdev() -> bdev_freeze()
* thaw_bdev() -> bdev_thaw()
Also document the return code.
Link: https://lore.kernel.org/r/20231024-vfs-super-freeze-v2-2-599c19f4faac@kernel.org
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>
|
|
Commit 23baf831a32c ("mm, treewide: redefine MAX_ORDER sanely")
changed the meaning of MAX_ORDER from exclusive to inclusive. So, we
can allocate compound pages with up to 1 << MAX_ORDER pages.
Reflect this change in dm-crypt and start trying to allocate compound
pages with MAX_ORDER.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
The commit 5721d4e5a9cd enhanced dm-verity, so that it can verify blocks
from tasklets rather than from workqueues. This reportedly improves
performance significantly.
However, dm-verity was using the flag CRYPTO_TFM_REQ_MAY_SLEEP from
tasklets which resulted in warnings about sleeping function being called
from non-sleeping context.
BUG: sleeping function called from invalid context at crypto/internal.h:206
in_atomic(): 1, irqs_disabled(): 0, non_block: 0, pid: 14, name: ksoftirqd/0
preempt_count: 100, expected: 0
RCU nest depth: 0, expected: 0
CPU: 0 PID: 14 Comm: ksoftirqd/0 Tainted: G W 6.7.0-rc1 #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014
Call Trace:
<TASK>
dump_stack_lvl+0x32/0x50
__might_resched+0x110/0x160
crypto_hash_walk_done+0x54/0xb0
shash_ahash_update+0x51/0x60
verity_hash_update.isra.0+0x4a/0x130 [dm_verity]
verity_verify_io+0x165/0x550 [dm_verity]
? free_unref_page+0xdf/0x170
? psi_group_change+0x113/0x390
verity_tasklet+0xd/0x70 [dm_verity]
tasklet_action_common.isra.0+0xb3/0xc0
__do_softirq+0xaf/0x1ec
? smpboot_thread_fn+0x1d/0x200
? sort_range+0x20/0x20
run_ksoftirqd+0x15/0x30
smpboot_thread_fn+0xed/0x200
kthread+0xdc/0x110
? kthread_complete_and_exit+0x20/0x20
ret_from_fork+0x28/0x40
? kthread_complete_and_exit+0x20/0x20
ret_from_fork_asm+0x11/0x20
</TASK>
This commit fixes dm-verity so that it doesn't use the flags
CRYPTO_TFM_REQ_MAY_SLEEP and CRYPTO_TFM_REQ_MAY_BACKLOG from tasklets. The
crypto API would do GFP_ATOMIC allocation instead, it could return -ENOMEM
and we catch -ENOMEM in verity_tasklet and requeue the request to the
workqueue.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org # v6.0+
Fixes: 5721d4e5a9cd ("dm verity: Add optional "try_verify_in_tasklet" feature")
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
dm-bufio has a no-sleep mode. When activated (with the
DM_BUFIO_CLIENT_NO_SLEEP flag), the bufio client is read-only and we
could call dm_bufio_get from tasklets. This is used by dm-verity.
Unfortunately, commit 450e8dee51aa ("dm bufio: improve concurrent IO
performance") broke this and the kernel would warn that cache_get()
was calling down_read() from no-sleeping context. The bug can be
reproduced by using "veritysetup open" with the "--use-tasklets"
flag.
This commit fixes dm-bufio, so that the tasklet mode works again, by
expanding use of the 'no_sleep_enabled' static_key to conditionally
use either a rw_semaphore or rwlock_t (which are colocated in the
buffer_tree structure using a union).
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org # v6.4
Fixes: 450e8dee51aa ("dm bufio: improve concurrent IO performance")
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
This is small refactoring of dm-delay - we avoid duplicate logic in
flush_delayed_bios and flush_delayed_bios_fast and join these two
functions into one.
We also add cond_resched() to flush_delayed_bios because the list may have
unbounded number of entries.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
This commit fixes the following bugs introduced by commit 70bbeb29fab0
("dm delay: for short delays, use kthread instead of timers and wq"):
* the function flush_worker_fn has no exit path - on unload, this
function will just loop and consume 100% CPU without any progress
* the wake-up mechanism in flush_worker_fn is racy - a wake up will be
missed if the process adds entries to the delayed_bios list just
before set_current_state(TASK_INTERRUPTIBLE)
* flush_delayed_bios_fast submits a bio while holding a global mutex;
this may deadlock if we have multiple stacked dm-delay devices and
the underlying device attempts to acquire the mutex too
* if the target constructor fails, it will call delay_dtr. delay_dtr
would attempt to free dc->timer_lock without it being initialized by
the constructor.
* if the target constructor's kthread allocation fails, delay_dtr
would crash trying to dereference dc->worker because it is non-NULL
due to ERR_PTR.
Fixes: 70bbeb29fab0 ("dm delay: for short delays, use kthread instead of timers and wq")
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Mike Snitzer <snitzer@kernel.org>
|
|
In delay_presuspend, we set the atomic variable may_delay and then stop
the timer and flush pending bios. The intention here is to prevent the
delay target from re-arming the timer again.
However, this test is racy. Suppose that one thread goes to delay_bio,
sees that dc->may_delay is one and proceeds; now, another thread executes
delay_presuspend, it sets dc->may_delay to zero, deletes the timer and
flushes pending bios. Then, the first thread continues and adds the bio to
delayed->list despite the fact that dc->may_delay is false.
Fix this bug by changing may_delay's type from atomic_t to bool and
only access it while holding the delayed_bios_lock mutex. Note that we
d |
