linux-stable.git/drivers/md/md.c, branch v6.4 Linux kernel stable tree Merge tag 'sysctl-6.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/mcgrof/linux 2023-04-27T23:52:33+00:00 Linus Torvalds torvalds@linux-foundation.org 2023-04-27T23:52:33+00:00 888d3c9f7f3ae44101a3fd76528d3dd6f96e9fd0 Pull sysctl updates from Luis Chamberlain: "This only does a few sysctl moves from the kernel/sysctl.c file, the rest of the work has been put towards deprecating two API calls which incur recursion and prevent us from simplifying the registration process / saving memory per move. Most of the changes have been soaking on linux-next since v6.3-rc3. I've slowed down the kernel/sysctl.c moves due to Matthew Wilcox's feedback that we should see if we could *save* memory with these moves instead of incurring more memory. We currently incur more memory since when we move a syctl from kernel/sysclt.c out to its own file we end up having to add a new empty sysctl used to register it. To achieve saving memory we want to allow syctls to be passed without requiring the end element being empty, and just have our registration process rely on ARRAY_SIZE(). Without this, supporting both styles of sysctls would make the sysctl registration pretty brittle, hard to read and maintain as can be seen from Meng Tang's efforts to do just this [0]. Fortunately, in order to use ARRAY_SIZE() for all sysctl registrations also implies doing the work to deprecate two API calls which use recursion in order to support sysctl declarations with subdirectories. And so during this development cycle quite a bit of effort went into this deprecation effort. I've annotated the following two APIs are deprecated and in few kernel releases we should be good to remove them: - register_sysctl_table() - register_sysctl_paths() During this merge window we should be able to deprecate and unexport register_sysctl_paths(), we can probably do that towards the end of this merge window. Deprecating register_sysctl_table() will take a bit more time but this pull request goes with a few example of how to do this. As it turns out each of the conversions to move away from either of these two API calls *also* saves memory. And so long term, all these changes *will* prove to have saved a bit of memory on boot. The way I see it then is if remove a user of one deprecated call, it gives us enough savings to move one kernel/sysctl.c out from the generic arrays as we end up with about the same amount of bytes. Since deprecating register_sysctl_table() and register_sysctl_paths() does not require maintainer coordination except the final unexport you'll see quite a bit of these changes from other pull requests, I've just kept the stragglers after rc3" Link: https://lkml.kernel.org/r/ZAD+cpbrqlc5vmry@bombadil.infradead.org [0] * tag 'sysctl-6.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/mcgrof/linux: (29 commits) fs: fix sysctls.c built mm: compaction: remove incorrect #ifdef checks mm: compaction: move compaction sysctl to its own file mm: memory-failure: Move memory failure sysctls to its own file arm: simplify two-level sysctl registration for ctl_isa_vars ia64: simplify one-level sysctl registration for kdump_ctl_table utsname: simplify one-level sysctl registration for uts_kern_table ntfs: simplfy one-level sysctl registration for ntfs_sysctls coda: simplify one-level sysctl registration for coda_table fs/cachefiles: simplify one-level sysctl registration for cachefiles_sysctls xfs: simplify two-level sysctl registration for xfs_table nfs: simplify two-level sysctl registration for nfs_cb_sysctls nfs: simplify two-level sysctl registration for nfs4_cb_sysctls lockd: simplify two-level sysctl registration for nlm_sysctls proc_sysctl: enhance documentation xen: simplify sysctl registration for balloon md: simplify sysctl registration hv: simplify sysctl registration scsi: simplify sysctl registration with register_sysctl() csky: simplify alignment sysctl registration ... 
Pull sysctl updates from Luis Chamberlain:
 "This only does a few sysctl moves from the kernel/sysctl.c file, the
  rest of the work has been put towards deprecating two API calls which
  incur recursion and prevent us from simplifying the registration
  process / saving memory per move. Most of the changes have been
  soaking on linux-next since v6.3-rc3.

  I've slowed down the kernel/sysctl.c moves due to Matthew Wilcox's
  feedback that we should see if we could *save* memory with these moves
  instead of incurring more memory. We currently incur more memory since
  when we move a syctl from kernel/sysclt.c out to its own file we end
  up having to add a new empty sysctl used to register it. To achieve
  saving memory we want to allow syctls to be passed without requiring
  the end element being empty, and just have our registration process
  rely on ARRAY_SIZE(). Without this, supporting both styles of sysctls
  would make the sysctl registration pretty brittle, hard to read and
  maintain as can be seen from Meng Tang's efforts to do just this [0].
  Fortunately, in order to use ARRAY_SIZE() for all sysctl registrations
  also implies doing the work to deprecate two API calls which use
  recursion in order to support sysctl declarations with subdirectories.

  And so during this development cycle quite a bit of effort went into
  this deprecation effort. I've annotated the following two APIs are
  deprecated and in few kernel releases we should be good to remove
  them:

   - register_sysctl_table()
   - register_sysctl_paths()

  During this merge window we should be able to deprecate and unexport
  register_sysctl_paths(), we can probably do that towards the end of
  this merge window.

  Deprecating register_sysctl_table() will take a bit more time but this
  pull request goes with a few example of how to do this.

  As it turns out each of the conversions to move away from either of
  these two API calls *also* saves memory. And so long term, all these
  changes *will* prove to have saved a bit of memory on boot.

  The way I see it then is if remove a user of one deprecated call, it
  gives us enough savings to move one kernel/sysctl.c out from the
  generic arrays as we end up with about the same amount of bytes.

  Since deprecating register_sysctl_table() and register_sysctl_paths()
  does not require maintainer coordination except the final unexport
  you'll see quite a bit of these changes from other pull requests, I've
  just kept the stragglers after rc3"

Link: https://lkml.kernel.org/r/ZAD+cpbrqlc5vmry@bombadil.infradead.org [0]

* tag 'sysctl-6.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/mcgrof/linux: (29 commits)
  fs: fix sysctls.c built
  mm: compaction: remove incorrect #ifdef checks
  mm: compaction: move compaction sysctl to its own file
  mm: memory-failure: Move memory failure sysctls to its own file
  arm: simplify two-level sysctl registration for ctl_isa_vars
  ia64: simplify one-level sysctl registration for kdump_ctl_table
  utsname: simplify one-level sysctl registration for uts_kern_table
  ntfs: simplfy one-level sysctl registration for ntfs_sysctls
  coda: simplify one-level sysctl registration for coda_table
  fs/cachefiles: simplify one-level sysctl registration for cachefiles_sysctls
  xfs: simplify two-level sysctl registration for xfs_table
  nfs: simplify two-level sysctl registration for nfs_cb_sysctls
  nfs: simplify two-level sysctl registration for nfs4_cb_sysctls
  lockd: simplify two-level sysctl registration for nlm_sysctls
  proc_sysctl: enhance documentation
  xen: simplify sysctl registration for balloon
  md: simplify sysctl registration
  hv: simplify sysctl registration
  scsi: simplify sysctl registration with register_sysctl()
  csky: simplify alignment sysctl registration
  ...
Merge tag 'for-6.4/block-2023-04-21' of git://git.kernel.dk/linux 2023-04-26T19:52:58+00:00 Linus Torvalds torvalds@linux-foundation.org 2023-04-26T19:52:58+00:00 9dd6956b38923dc1b7b349ca1eee3c0bb1f0163a Pull block updates from Jens Axboe: - drbd patches, bringing us closer to unifying the out-of-tree version and the in tree one (Andreas, Christoph) - support for auto-quiesce for the s390 dasd driver (Stefan) - MD pull request via Song: - md/bitmap: Optimal last page size (Jon Derrick) - Various raid10 fixes (Yu Kuai, Li Nan) - md: add error_handlers for raid0 and linear (Mariusz Tkaczyk) - NVMe pull request via Christoph: - Drop redundant pci_enable_pcie_error_reporting (Bjorn Helgaas) - Validate nvmet module parameters (Chaitanya Kulkarni) - Fence TCP socket on receive error (Chris Leech) - Fix async event trace event (Keith Busch) - Minor cleanups (Chaitanya Kulkarni, zhenwei pi) - Fix and cleanup nvmet Identify handling (Damien Le Moal, Christoph Hellwig) - Fix double blk_mq_complete_request race in the timeout handler (Lei Yin) - Fix irq locking in nvme-fcloop (Ming Lei) - Remove queue mapping helper for rdma devices (Sagi Grimberg) - use structured request attribute checks for nbd (Jakub) - fix blk-crypto race conditions between keyslot management (Eric) - add sed-opal support for reading read locking range attributes (Ondrej) - make fault injection configurable for null_blk (Akinobu) - clean up the request insertion API (Christoph) - clean up the queue running API (Christoph) - blkg config helper cleanups (Tejun) - lazy init support for blk-iolatency (Tejun) - various fixes and tweaks to ublk (Ming) - remove hybrid polling. It hasn't really been useful since we got async polled IO support, and these days we don't support sync polled IO at all (Keith) - misc fixes, cleanups, improvements (Zhong, Ondrej, Colin, Chengming, Chaitanya, me) * tag 'for-6.4/block-2023-04-21' of git://git.kernel.dk/linux: (118 commits) nbd: fix incomplete validation of ioctl arg ublk: don't return 0 in case of any failure sed-opal: geometry feature reporting command null_blk: Always check queue mode setting from configfs block: ublk: switch to ioctl command encoding blk-mq: fix the blk_mq_add_to_requeue_list call in blk_kick_flush block, bfq: Fix division by zero error on zero wsum fault-inject: fix build error when FAULT_INJECTION_CONFIGFS=y and CONFIGFS_FS=m block: store bdev->bd_disk->fops->submit_bio state in bdev block: re-arrange the struct block_device fields for better layout md/raid5: remove unused working_disks variable md/raid10: don't call bio_start_io_acct twice for bio which experienced read error md/raid10: fix memleak of md thread md/raid10: fix memleak for 'conf->bio_split' md/raid10: fix leak of 'r10bio->remaining' for recovery md/raid10: don't BUG_ON() in raise_barrier() md: fix soft lockup in status_resync md: add error_handlers for raid0 and linear md: Use optimal I/O size for last bitmap page md: Fix types in sb writer ... 
Pull block updates from Jens Axboe:

 - drbd patches, bringing us closer to unifying the out-of-tree version
   and the in tree one (Andreas, Christoph)

 - support for auto-quiesce for the s390 dasd driver (Stefan)

 - MD pull request via Song:
      - md/bitmap: Optimal last page size (Jon Derrick)
      - Various raid10 fixes (Yu Kuai, Li Nan)
      - md: add error_handlers for raid0 and linear (Mariusz Tkaczyk)

 - NVMe pull request via Christoph:
      - Drop redundant pci_enable_pcie_error_reporting (Bjorn Helgaas)
      - Validate nvmet module parameters (Chaitanya Kulkarni)
      - Fence TCP socket on receive error (Chris Leech)
      - Fix async event trace event (Keith Busch)
      - Minor cleanups (Chaitanya Kulkarni, zhenwei pi)
      - Fix and cleanup nvmet Identify handling (Damien Le Moal,
        Christoph Hellwig)
      - Fix double blk_mq_complete_request race in the timeout handler
        (Lei Yin)
      - Fix irq locking in nvme-fcloop (Ming Lei)
      - Remove queue mapping helper for rdma devices (Sagi Grimberg)

 - use structured request attribute checks for nbd (Jakub)

 - fix blk-crypto race conditions between keyslot management (Eric)

 - add sed-opal support for reading read locking range attributes
   (Ondrej)

 - make fault injection configurable for null_blk (Akinobu)

 - clean up the request insertion API (Christoph)

 - clean up the queue running API (Christoph)

 - blkg config helper cleanups (Tejun)

 - lazy init support for blk-iolatency (Tejun)

 - various fixes and tweaks to ublk (Ming)

 - remove hybrid polling. It hasn't really been useful since we got
   async polled IO support, and these days we don't support sync polled
   IO at all (Keith)

 - misc fixes, cleanups, improvements (Zhong, Ondrej, Colin, Chengming,
   Chaitanya, me)

* tag 'for-6.4/block-2023-04-21' of git://git.kernel.dk/linux: (118 commits)
  nbd: fix incomplete validation of ioctl arg
  ublk: don't return 0 in case of any failure
  sed-opal: geometry feature reporting command
  null_blk: Always check queue mode setting from configfs
  block: ublk: switch to ioctl command encoding
  blk-mq: fix the blk_mq_add_to_requeue_list call in blk_kick_flush
  block, bfq: Fix division by zero error on zero wsum
  fault-inject: fix build error when FAULT_INJECTION_CONFIGFS=y and CONFIGFS_FS=m
  block: store bdev->bd_disk->fops->submit_bio state in bdev
  block: re-arrange the struct block_device fields for better layout
  md/raid5: remove unused working_disks variable
  md/raid10: don't call bio_start_io_acct twice for bio which experienced read error
  md/raid10: fix memleak of md thread
  md/raid10: fix memleak for 'conf->bio_split'
  md/raid10: fix leak of 'r10bio->remaining' for recovery
  md/raid10: don't BUG_ON() in raise_barrier()
  md: fix soft lockup in status_resync
  md: add error_handlers for raid0 and linear
  md: Use optimal I/O size for last bitmap page
  md: Fix types in sb writer
  ...
md: fix soft lockup in status_resync 2023-04-14T05:20:24+00:00 Yu Kuai yukuai3@huawei.com 2023-03-10T07:38:51+00:00 6efddf1e32e2a264694766ca485a4f5e04ee82a7 status_resync() will calculate 'curr_resync - recovery_active' to show user a progress bar like following: [============>........] resync = 61.4% 'curr_resync' and 'recovery_active' is updated in md_do_sync(), and status_resync() can read them concurrently, hence it's possible that 'curr_resync - recovery_active' can overflow to a huge number. In this case status_resync() will be stuck in the loop to print a large amount of '=', which will end up soft lockup. Fix the problem by setting 'resync' to MD_RESYNC_ACTIVE in this case, this way resync in progress will be reported to user. Signed-off-by: Yu Kuai <yukuai3@huawei.com> Signed-off-by: Song Liu <song@kernel.org> Link: https://lore.kernel.org/r/20230310073855.1337560-3-yukuai1@huaweicloud.com 
status_resync() will calculate 'curr_resync - recovery_active' to show
user a progress bar like following:

[============>........]  resync = 61.4%

'curr_resync' and 'recovery_active' is updated in md_do_sync(), and
status_resync() can read them concurrently, hence it's possible that
'curr_resync - recovery_active' can overflow to a huge number. In this
case status_resync() will be stuck in the loop to print a large amount
of '=', which will end up soft lockup.

Fix the problem by setting 'resync' to MD_RESYNC_ACTIVE in this case,
this way resync in progress will be reported to user.

Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20230310073855.1337560-3-yukuai1@huaweicloud.com
md: add error_handlers for raid0 and linear 2023-04-14T05:20:24+00:00 Mariusz Tkaczyk mariusz.tkaczyk@linux.intel.com 2023-03-06T13:03:17+00:00 c31fea2f8e2a72c817f318016bbc327095175a9f After the commit 9631abdbf406c("md: Set MD_BROKEN for RAID1 and RAID10") MD_BROKEN must be set if array is failed because state_store() checks it. If it is set then -EBUSY is returned to userspace. For raid0 and linear MD_BROKEN is not set by error_handler(). As a result mdadm is unable to trigger clean-up actions. It is a regression. This patch adds appropriate error_handler for raid0 and linear. The error handler sets MD_BROKEN for this device. Reviewed-by: Xiao Ni <xni@redhat.com> Signed-off-by: Mariusz Tkaczyk <mariusz.tkaczyk@linux.intel.com> Signed-off-by: Song Liu <song@kernel.org> Link: https://lore.kernel.org/r/20230306130317.3418-1-mariusz.tkaczyk@linux.intel.com 
After the commit 9631abdbf406c("md: Set MD_BROKEN for RAID1 and RAID10")
MD_BROKEN must be set if array is failed because state_store() checks it.
If it is set then -EBUSY is returned to userspace.

For raid0 and linear MD_BROKEN is not set by error_handler(). As a result
mdadm is unable to trigger clean-up actions. It is a regression.

This patch adds appropriate error_handler for raid0 and linear. The
error handler sets MD_BROKEN for this device.

Reviewed-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Mariusz Tkaczyk <mariusz.tkaczyk@linux.intel.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20230306130317.3418-1-mariusz.tkaczyk@linux.intel.com
md: make kobj_type structures constant 2023-04-14T05:20:23+00:00 Thomas Weißschuh linux@weissschuh.net 2023-02-14T03:19:22+00:00 4d72a9de2f002c77267c720ebc5595a17f0d66e9 Since commit ee6d3dd4ed48 ("driver core: make kobj_type constant.") the driver core allows the usage of const struct kobj_type. Take advantage of this to constify the structure definitions to prevent modification at runtime. Signed-off-by: Thomas Weißschuh <linux@weissschuh.net> Signed-off-by: Song Liu <song@kernel.org> Link: https://lore.kernel.org/r/20230214-kobj_type-md-v1-1-d6853f707f11@weissschuh.net 
Since commit ee6d3dd4ed48 ("driver core: make kobj_type constant.")
the driver core allows the usage of const struct kobj_type.

Take advantage of this to constify the structure definitions to prevent
modification at runtime.

Signed-off-by: Thomas Weißschuh <linux@weissschuh.net>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20230214-kobj_type-md-v1-1-d6853f707f11@weissschuh.net
md: simplify sysctl registration 2023-04-13T18:49:20+00:00 Luis Chamberlain mcgrof@kernel.org 2023-03-02T20:46:09+00:00 9adcf9d3d5c9ed6bb9d493b59594501f79ae3bed register_sysctl_table() is a deprecated compatibility wrapper. register_sysctl() can do the directory creation for you so just use that. Signed-off-by: Luis Chamberlain <mcgrof@kernel.org> Acked-by: Song Liu <song@kernel.org> 
register_sysctl_table() is a deprecated compatibility wrapper.
register_sysctl() can do the directory creation for you so just use
that.

Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Acked-by: Song Liu <song@kernel.org>
md: fix regression for null-ptr-deference in __md_stop() 2023-03-29T18:30:20+00:00 Yu Kuai yukuai3@huawei.com 2023-03-28T09:44:00+00:00 433279beba1d4872da10b7b60a539e0cb828b32b Commit 3e453522593d ("md: Free resources in __md_stop") tried to fix null-ptr-deference for 'active_io' by moving percpu_ref_exit() to __md_stop(), however, the commit also moving 'writes_pending' to __md_stop(), and this will cause mdadm tests broken: BUG: kernel NULL pointer dereference, address: 0000000000000038 Oops: 0000 [#1] PREEMPT SMP CPU: 15 PID: 17830 Comm: mdadm Not tainted 6.3.0-rc3-next-20230324-00009-g520d37 RIP: 0010:free_percpu+0x465/0x670 Call Trace: <TASK> __percpu_ref_exit+0x48/0x70 percpu_ref_exit+0x1a/0x90 __md_stop+0xe9/0x170 do_md_stop+0x1e1/0x7b0 md_ioctl+0x90c/0x1aa0 blkdev_ioctl+0x19b/0x400 vfs_ioctl+0x20/0x50 __x64_sys_ioctl+0xba/0xe0 do_syscall_64+0x6c/0xe0 entry_SYSCALL_64_after_hwframe+0x63/0xcd And the problem can be reporduced 100% by following test: mdadm -CR /dev/md0 -l1 -n1 /dev/sda --force echo inactive > /sys/block/md0/md/array_state echo read-auto > /sys/block/md0/md/array_state echo inactive > /sys/block/md0/md/array_state Root cause: // start raid raid1_run mddev_init_writes_pending percpu_ref_init // inactive raid array_state_store do_md_stop __md_stop percpu_ref_exit // start raid again array_state_store do_md_run raid1_run mddev_init_writes_pending if (mddev->writes_pending.percpu_count_ptr) // won't reinit // inactive raid again ... percpu_ref_exit -> null-ptr-deference Before the commit, 'writes_pending' is exited when mddev is freed, and it's safe to restart raid because mddev_init_writes_pending() already make sure that 'writes_pending' will only be initialized once. Fix the prblem by moving 'writes_pending' back, it's a litter hard to find the relationship between alloc memory and free memory, however, code changes is much less and we lived with this for a long time already. Fixes: 3e453522593d ("md: Free resources in __md_stop") Signed-off-by: Yu Kuai <yukuai3@huawei.com> Reviewed-by: Xiao Ni <xni@redhat.com> Signed-off-by: Song Liu <song@kernel.org> Link: https://lore.kernel.org/r/20230328094400.1448955-1-yukuai1@huaweicloud.com 
Commit 3e453522593d ("md: Free resources in __md_stop") tried to fix
null-ptr-deference for 'active_io' by moving percpu_ref_exit() to
__md_stop(), however, the commit also moving 'writes_pending' to
__md_stop(), and this will cause mdadm tests broken:

BUG: kernel NULL pointer dereference, address: 0000000000000038
Oops: 0000 [#1] PREEMPT SMP
CPU: 15 PID: 17830 Comm: mdadm Not tainted 6.3.0-rc3-next-20230324-00009-g520d37
RIP: 0010:free_percpu+0x465/0x670
Call Trace:
 <TASK>
 __percpu_ref_exit+0x48/0x70
 percpu_ref_exit+0x1a/0x90
 __md_stop+0xe9/0x170
 do_md_stop+0x1e1/0x7b0
 md_ioctl+0x90c/0x1aa0
 blkdev_ioctl+0x19b/0x400
 vfs_ioctl+0x20/0x50
 __x64_sys_ioctl+0xba/0xe0
 do_syscall_64+0x6c/0xe0
 entry_SYSCALL_64_after_hwframe+0x63/0xcd

And the problem can be reporduced 100% by following test:

mdadm -CR /dev/md0 -l1 -n1 /dev/sda --force
echo inactive > /sys/block/md0/md/array_state
echo read-auto  > /sys/block/md0/md/array_state
echo inactive > /sys/block/md0/md/array_state

Root cause:

// start raid
raid1_run
 mddev_init_writes_pending
  percpu_ref_init

// inactive raid
array_state_store
 do_md_stop
  __md_stop
   percpu_ref_exit

// start raid again
array_state_store
 do_md_run
  raid1_run
   mddev_init_writes_pending
    if (mddev->writes_pending.percpu_count_ptr)
    // won't reinit

// inactive raid again
...
percpu_ref_exit
-> null-ptr-deference

Before the commit, 'writes_pending' is exited when mddev is freed, and
it's safe to restart raid because mddev_init_writes_pending() already make
sure that 'writes_pending' will only be initialized once.

Fix the prblem by moving 'writes_pending' back, it's a litter hard to find
the relationship between alloc memory and free memory, however, code
changes is much less and we lived with this for a long time already.

Fixes: 3e453522593d ("md: Free resources in __md_stop")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Reviewed-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20230328094400.1448955-1-yukuai1@huaweicloud.com
md: avoid signed overflow in slot_store() 2023-03-13T19:50:54+00:00 NeilBrown neilb@suse.de 2023-03-05T22:36:25+00:00 3bc57292278a0b6ac4656cad94c14f2453344b57 slot_store() uses kstrtouint() to get a slot number, but stores the result in an "int" variable (by casting a pointer). This can result in a negative slot number if the unsigned int value is very large. A negative number means that the slot is empty, but setting a negative slot number this way will not remove the device from the array. I don't think this is a serious problem, but it could cause confusion and it is best to fix it. Reported-by: Dan Carpenter <error27@gmail.com> Signed-off-by: NeilBrown <neilb@suse.de> Signed-off-by: Song Liu <song@kernel.org> 
slot_store() uses kstrtouint() to get a slot number, but stores the
result in an "int" variable (by casting a pointer).
This can result in a negative slot number if the unsigned int value is
very large.

A negative number means that the slot is empty, but setting a negative
slot number this way will not remove the device from the array.  I don't
think this is a serious problem, but it could cause confusion and it is
best to fix it.

Reported-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Song Liu <song@kernel.org>
md: Free resources in __md_stop 2023-03-13T17:56:54+00:00 Xiao Ni xni@redhat.com 2023-02-22T03:59:16+00:00 3e453522593d74a87cf68a38e14aa36ebca1dbcd If md_run() fails after ->active_io is initialized, then percpu_ref_exit is called in error path. However, later md_free_disk will call percpu_ref_exit again which leads to a panic because of null pointer dereference. It can also trigger this bug when resources are initialized but are freed in error path, then will be freed again in md_free_disk. BUG: kernel NULL pointer dereference, address: 0000000000000038 Oops: 0000 [#1] PREEMPT SMP Workqueue: md_misc mddev_delayed_delete RIP: 0010:free_percpu+0x110/0x630 Call Trace: <TASK> __percpu_ref_exit+0x44/0x70 percpu_ref_exit+0x16/0x90 md_free_disk+0x2f/0x80 disk_release+0x101/0x180 device_release+0x84/0x110 kobject_put+0x12a/0x380 kobject_put+0x160/0x380 mddev_delayed_delete+0x19/0x30 process_one_work+0x269/0x680 worker_thread+0x266/0x640 kthread+0x151/0x1b0 ret_from_fork+0x1f/0x30 For creating raid device, md raid calls do_md_run->md_run, dm raid calls md_run. We alloc those memory in md_run. For stopping raid device, md raid calls do_md_stop->__md_stop, dm raid calls md_stop->__md_stop. So we can free those memory resources in __md_stop. Fixes: 72adae23a72c ("md: Change active_io to percpu") Reported-and-tested-by: Yu Kuai <yukuai3@huawei.com> Signed-off-by: Xiao Ni <xni@redhat.com> Signed-off-by: Song Liu <song@kernel.org> 
If md_run() fails after ->active_io is initialized, then percpu_ref_exit
is called in error path. However, later md_free_disk will call
percpu_ref_exit again which leads to a panic because of null pointer
dereference. It can also trigger this bug when resources are initialized
but are freed in error path, then will be freed again in md_free_disk.

BUG: kernel NULL pointer dereference, address: 0000000000000038
Oops: 0000 [#1] PREEMPT SMP
Workqueue: md_misc mddev_delayed_delete
RIP: 0010:free_percpu+0x110/0x630
Call Trace:
 <TASK>
 __percpu_ref_exit+0x44/0x70
 percpu_ref_exit+0x16/0x90
 md_free_disk+0x2f/0x80
 disk_release+0x101/0x180
 device_release+0x84/0x110
 kobject_put+0x12a/0x380
 kobject_put+0x160/0x380
 mddev_delayed_delete+0x19/0x30
 process_one_work+0x269/0x680
 worker_thread+0x266/0x640
 kthread+0x151/0x1b0
 ret_from_fork+0x1f/0x30

For creating raid device, md raid calls do_md_run->md_run, dm raid calls
md_run. We alloc those memory in md_run. For stopping raid device, md raid
calls do_md_stop->__md_stop, dm raid calls md_stop->__md_stop. So we can
free those memory resources in __md_stop.

Fixes: 72adae23a72c ("md: Change active_io to percpu")
Reported-and-tested-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Song Liu <song@kernel.org>
md: account io_acct_set usage with active_io 2023-02-08T23:46:57+00:00 Xiao Ni xni@redhat.com 2023-02-03T05:13:44+00:00 76fed01420bb8b0e282745a4945925b25751d42b io_acct_set was enabled for raid0/raid5 io accounting. bios that contain md_io_acct are allocated in the i/o path. There isn't a good method to monitor if these bios are all finished and freed. In the takeover process, io_acct_set (which is used for bios with md_io_acct) need to be freed. However, if some bios finish after io_acct_set is freed, it may trigger the following panic: [ 6973.767999] RIP: 0010:mempool_free+0x52/0x80 [ 6973.786098] Call Trace: [ 6973.786549] md_end_io_acct+0x31/0x40 [ 6973.787227] blk_update_request+0x224/0x380 [ 6973.787994] blk_mq_end_request+0x1a/0x130 [ 6973.788739] blk_complete_reqs+0x35/0x50 [ 6973.789456] __do_softirq+0xd7/0x2c8 [ 6973.790114] ? sort_range+0x20/0x20 [ 6973.790763] run_ksoftirqd+0x2a/0x40 [ 6973.791400] smpboot_thread_fn+0xb5/0x150 [ 6973.792114] kthread+0x10b/0x130 [ 6973.792724] ? set_kthread_struct+0x50/0x50 [ 6973.793491] ret_from_fork+0x1f/0x40 Fix this by increasing and decreasing active_io for each bio with md_io_acct so that mddev_suspend() will wait until all bios from io_acct_set finish before freeing io_acct_set. Reported-by: Fine Fan <ffan@redhat.com> Signed-off-by: Xiao Ni <xni@redhat.com> Signed-off-by: Song Liu <song@kernel.org> 
io_acct_set was enabled for raid0/raid5 io accounting. bios that contain
md_io_acct are allocated in the i/o path. There isn't a good method to
monitor if these bios are all finished and freed. In the takeover process,
io_acct_set (which is used for bios with md_io_acct) need to be freed.
However, if some bios finish after io_acct_set is freed, it may trigger
the following panic:

[ 6973.767999] RIP: 0010:mempool_free+0x52/0x80
[ 6973.786098] Call Trace:
[ 6973.786549]  md_end_io_acct+0x31/0x40
[ 6973.787227]  blk_update_request+0x224/0x380
[ 6973.787994]  blk_mq_end_request+0x1a/0x130
[ 6973.788739]  blk_complete_reqs+0x35/0x50
[ 6973.789456]  __do_softirq+0xd7/0x2c8
[ 6973.790114]  ? sort_range+0x20/0x20
[ 6973.790763]  run_ksoftirqd+0x2a/0x40
[ 6973.791400]  smpboot_thread_fn+0xb5/0x150
[ 6973.792114]  kthread+0x10b/0x130
[ 6973.792724]  ? set_kthread_struct+0x50/0x50
[ 6973.793491]  ret_from_fork+0x1f/0x40

Fix this by increasing and decreasing active_io for each bio with
md_io_acct so that mddev_suspend() will wait until all bios from
io_acct_set finish before freeing io_acct_set.

Reported-by: Fine Fan <ffan@redhat.com>
Signed-off-by: Xiao Ni <xni@redhat.com>
Signed-off-by: Song Liu <song@kernel.org>