linux/drivers/md
Yu Kuai a865b96c51 Revert "md: unlock mddev before reap sync_thread in action_store"
This reverts commit 9dfbdafda3.

Because it will introduce a defect that sync_thread can be running while
MD_RECOVERY_RUNNING is cleared, which will cause some unexpected problems,
for example:

list_add corruption. prev->next should be next (ffff0001ac1daba0), but was ffff0000ce1a02a0. (prev=ffff0000ce1a02a0).
Call trace:
 __list_add_valid+0xfc/0x140
 insert_work+0x78/0x1a0
 __queue_work+0x500/0xcf4
 queue_work_on+0xe8/0x12c
 md_check_recovery+0xa34/0xf30
 raid10d+0xb8/0x900 [raid10]
 md_thread+0x16c/0x2cc
 kthread+0x1a4/0x1ec
 ret_from_fork+0x10/0x18

This is because work is requeued while it's still inside workqueue:

t1:			t2:
action_store
 mddev_lock
  if (mddev->sync_thread)
   mddev_unlock
   md_unregister_thread
   // first sync_thread is done
			md_check_recovery
			 mddev_try_lock
			 /*
			  * once MD_RECOVERY_DONE is set, new sync_thread
			  * can start.
			  */
			 set_bit(MD_RECOVERY_RUNNING, &mddev->recovery)
			 INIT_WORK(&mddev->del_work, md_start_sync)
			 queue_work(md_misc_wq, &mddev->del_work)
			  test_and_set_bit(WORK_STRUCT_PENDING_BIT, ...)
			  // set pending bit
			  insert_work
			   list_add_tail
			 mddev_unlock
   mddev_lock_nointr
   md_reap_sync_thread
   // MD_RECOVERY_RUNNING is cleared
 mddev_unlock

t3:

// before queued work started from t2
md_check_recovery
 // MD_RECOVERY_RUNNING is not set, a new sync_thread can be started
 INIT_WORK(&mddev->del_work, md_start_sync)
  work->data = 0
  // work pending bit is cleared
 queue_work(md_misc_wq, &mddev->del_work)
  insert_work
   list_add_tail
   // list is corrupted

The above commit is reverted to fix the problem, the deadlock this
commit tries to fix will be fixed in following patches.

Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20230529132037.2124527-2-yukuai1@huaweicloud.com
2023-07-27 00:13:28 -07:00
..
bcache Merge branch 'for-6.5/block-late' into block-6.5 2023-06-28 16:08:19 -06:00
persistent-data dm thin metadata: Fix ABBA deadlock by resetting dm_bufio_client 2023-06-16 18:24:13 -04:00
dm-audit.c dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-audit.h dm: introduce audit event module for device mapper 2021-10-27 16:53:47 -04:00
dm-bio-prison-v1.c dm: improve hash_locks sizing and hash function 2023-03-30 15:57:51 -04:00
dm-bio-prison-v1.h dm bio prison v1: add dm_cell_key_has_valid_range 2023-03-30 15:57:51 -04:00
dm-bio-prison-v2.c dm: address space issues relative to switch/while/for/... 2023-02-14 14:23:06 -05:00
dm-bio-prison-v2.h dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-bio-record.h dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-bufio.c dm: get rid of GFP_NOIO workarounds for __vmalloc and kvmalloc 2023-06-27 16:06:54 -04:00
dm-builtin.c dm: adjust EXPORT_SYMBOL() to follow functions immediately 2023-02-14 14:23:07 -05:00
dm-cache-background-tracker.c dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-cache-background-tracker.h dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-cache-block-types.h dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-cache-metadata.c Revert "mm: shrinkers: convert shrinker_rwsem to mutex" 2023-06-19 13:19:33 -07:00
dm-cache-metadata.h dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-cache-policy-internal.h dm: add missing empty lines 2023-02-14 14:23:06 -05:00
dm-cache-policy-smq.c dm: add missing empty lines 2023-02-14 14:23:06 -05:00
dm-cache-policy.c dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-cache-policy.h dm: address indent/space issues 2023-02-14 14:23:06 -05:00
dm-cache-target.c block: replace fmode_t with a block-specific type for block open flags 2023-06-12 08:04:05 -06:00
dm-clone-metadata.c dm clone metadata: remove unused function 2021-04-19 13:20:31 -04:00
dm-clone-metadata.h dm clone metadata: Fix return type of dm_clone_nr_of_hydrated_regions() 2020-03-27 14:42:51 -04:00
dm-clone-target.c block: replace fmode_t with a block-specific type for block open flags 2023-06-12 08:04:05 -06:00
dm-core.h dm: support turning off block-core's io stats accounting 2023-06-16 18:24:13 -04:00
dm-crypt.c This update includes the following changes: 2023-06-30 21:27:13 -07:00
dm-delay.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-dust.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-ebs-target.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-era-target.c block: replace fmode_t with a block-specific type for block open flags 2023-06-12 08:04:05 -06:00
dm-exception-store.c dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-exception-store.h dm: avoid spaces before function arguments or in favour of tabs 2023-02-14 14:23:06 -05:00
dm-flakey.c dm flakey: introduce random_read_corrupt and random_write_corrupt options 2023-06-16 18:24:13 -04:00
dm-ima.c dm: avoid inline filenames 2023-02-14 14:23:07 -05:00
dm-ima.h dm: avoid inline filenames 2023-02-14 14:23:07 -05:00
dm-init.c dm: open code dm_get_dev_t in dm_init_init 2023-06-05 10:57:40 -06:00
dm-integrity.c - Update DM crypt to allocate compound pages if possible. 2023-06-30 12:16:00 -07:00
dm-io-rewind.c dm: avoid void function return statements 2023-02-14 14:23:07 -05:00
dm-io-tracker.h dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-io.c dm: remove unnecessary (void*) conversions 2023-04-11 12:01:01 -04:00
dm-ioctl.c - Update DM crypt to allocate compound pages if possible. 2023-06-30 12:16:00 -07:00
dm-kcopyd.c dm: remove unnecessary (void*) conversions 2023-04-11 12:01:01 -04:00
dm-linear.c dm: remove unnecessary (void*) conversions 2023-04-11 12:01:01 -04:00
dm-log-userspace-base.c dm: avoid void function return statements 2023-02-14 14:23:07 -05:00
dm-log-userspace-transfer.c dm: avoid split of quoted strings where possible 2023-02-14 14:23:07 -05:00
dm-log-userspace-transfer.h dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-log-writes.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-log.c dm: remove unnecessary (void*) conversions 2023-04-11 12:01:01 -04:00
dm-mpath.c dm: push error reporting down to dm_register_target() 2023-04-11 12:01:01 -04:00
dm-mpath.h dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-path-selector.c dm: adjust EXPORT_SYMBOL() to follow functions immediately 2023-02-14 14:23:07 -05:00
dm-path-selector.h dm: avoid spaces before function arguments or in favour of tabs 2023-02-14 14:23:06 -05:00
dm-ps-historical-service-time.c dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-ps-io-affinity.c dm: address space issues relative to switch/while/for/... 2023-02-14 14:23:06 -05:00
dm-ps-queue-length.c dm: avoid spaces before function arguments or in favour of tabs 2023-02-14 14:23:06 -05:00
dm-ps-round-robin.c dm: correct block comments format. 2023-02-14 14:23:06 -05:00
dm-ps-service-time.c dm: avoid spaces before function arguments or in favour of tabs 2023-02-14 14:23:06 -05:00
dm-raid1.c dm: remove unnecessary (void*) conversions 2023-04-11 12:01:01 -04:00
dm-raid.c Revert "md: unlock mddev before reap sync_thread in action_store" 2023-07-27 00:13:28 -07:00
dm-region-hash.c dm: correct block comments format. 2023-02-14 14:23:06 -05:00
dm-rq.c dm: avoid using symbolic permissions 2023-02-14 14:23:07 -05:00
dm-rq.h dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-snap-persistent.c dm: remove unnecessary (void*) conversions 2023-04-11 12:01:01 -04:00
dm-snap-transient.c dm: avoid split of quoted strings where possible 2023-02-14 14:23:07 -05:00
dm-snap.c block: replace fmode_t with a block-specific type for block open flags 2023-06-12 08:04:05 -06:00
dm-stats.c dm stats: check for and propagate alloc_percpu failure 2023-03-16 13:37:06 -04:00
dm-stats.h dm stats: check for and propagate alloc_percpu failure 2023-03-16 13:37:06 -04:00
dm-stripe.c dm: remove unnecessary (void*) conversions 2023-04-11 12:01:01 -04:00
dm-switch.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-sysfs.c dm sysfs: make kobj_type structure constant 2023-02-14 14:23:08 -05:00
dm-table.c block: replace fmode_t with a block-specific type for block open flags 2023-06-12 08:04:05 -06:00
dm-target.c dm: push error reporting down to dm_register_target() 2023-04-11 12:01:01 -04:00
dm-thin-metadata.c - Update DM crypt to allocate compound pages if possible. 2023-06-30 12:16:00 -07:00
dm-thin-metadata.h dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-thin.c - Update DM crypt to allocate compound pages if possible. 2023-06-30 12:16:00 -07:00
dm-uevent.c dm: avoid spaces before function arguments or in favour of tabs 2023-02-14 14:23:06 -05:00
dm-uevent.h dm: fix undue/missing spaces 2023-02-14 14:23:06 -05:00
dm-unstripe.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-verity-fec.c block: replace fmode_t with a block-specific type for block open flags 2023-06-12 08:04:05 -06:00
dm-verity-fec.h dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-verity-loadpin.c dm: verity-loadpin: Add NULL pointer check for 'bdev' parameter 2023-06-28 10:43:04 -07:00
dm-verity-target.c block: replace fmode_t with a block-specific type for block open flags 2023-06-12 08:04:05 -06:00
dm-verity-verify-sig.c dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-verity-verify-sig.h dm: add missing SPDX-License-Indentifiers 2023-02-14 14:23:06 -05:00
dm-verity.h dm: change "unsigned" to "unsigned int" 2023-02-14 14:23:06 -05:00
dm-writecache.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-zero.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-zone.c dm zone: Use the bitmap API to allocate bitmaps 2023-06-16 18:24:13 -04:00
dm-zoned-metadata.c dm: dm-zoned: use __bio_add_page for adding single metadata page 2023-05-31 09:50:02 -06:00
dm-zoned-reclaim.c dm kcopyd: avoid useless atomic operations 2021-06-04 12:07:24 -04:00
dm-zoned-target.c dm: add helper macro for simple DM target module init and exit 2023-04-11 12:09:08 -04:00
dm-zoned.h dm/dm-zoned: Use the enum req_op type 2022-07-14 12:14:31 -06:00
dm.c - Update DM crypt to allocate compound pages if possible. 2023-06-30 12:16:00 -07:00
dm.h dm: remove stale/redundant dm_internal_{suspend,resume} prototypes in dm.h 2023-06-16 18:24:14 -04:00
Kconfig Merge branch 'md-fixes' of https://git.kernel.org/pub/scm/linux/kernel/git/song/md into block-6.3 2023-03-15 12:18:07 -06:00
Makefile hardening updates for v5.20-rc1 2022-08-02 14:38:59 -07:00
md-autodetect.c init: improve the name_to_dev_t interface 2023-06-05 10:56:46 -06:00
md-bitmap.c md/md-bitmap: add a new helper to unplug bitmap asynchrously 2023-06-13 15:25:44 -07:00
md-bitmap.h md/md-bitmap: add a new helper to unplug bitmap asynchrously 2023-06-13 15:25:44 -07:00
md-cluster.c md: protect md_thread with rcu 2023-06-13 15:25:39 -07:00
md-cluster.h
md-faulty.c block: pass a block_device to bio_clone_fast 2022-02-04 07:43:18 -07:00
md-linear.c md: add error_handlers for raid0 and linear 2023-04-13 22:20:24 -07:00
md-linear.h md/raid1: Replace zero-length array with flexible-array 2020-05-13 12:02:23 -07:00
md-multipath.c md: protect md_thread with rcu 2023-06-13 15:25:39 -07:00
md-multipath.h
md.c Revert "md: unlock mddev before reap sync_thread in action_store" 2023-07-27 00:13:28 -07:00
md.h md: fix 'delete_mutex' deadlock 2023-06-23 09:41:47 -07:00
raid0.c md/raid0: add discard support for the 'original' layout 2023-06-30 15:43:50 -07:00
raid0.h md/raid0: add discard support for the 'original' layout 2023-06-30 15:43:50 -07:00
raid1-10.c md/raid1-10: fix casting from randomized structure in raid1_submit_write() 2023-06-23 09:33:16 -07:00
raid1.c md/raid1-10: limit the number of plugged bio 2023-06-13 15:25:44 -07:00
raid1.h md: protect md_thread with rcu 2023-06-13 15:25:39 -07:00
raid5-cache.c md: protect md_thread with rcu 2023-06-13 15:25:39 -07:00
raid5-log.h md/raid5-ppl: Drop unused argument from ppl_handle_flush_request() 2022-08-02 17:14:31 -06:00
raid5-ppl.c md: raid5: use __bio_add_page to add single page to new bio 2023-05-31 09:50:02 -06:00
raid5.c hardening updates for v6.5-rc1 2023-06-27 21:24:18 -07:00
raid5.h hardening updates for v6.5-rc1 2023-06-27 21:24:18 -07:00
raid10.c raid10: avoid spin_lock from fastpath from raid10_unplug() 2023-06-23 09:41:50 -07:00
raid10.h md: protect md_thread with rcu 2023-06-13 15:25:39 -07:00