All of lore.kernel.org
 help / color / mirror / Atom feed
From: David Sterba <dsterba@suse.cz>
To: Josef Bacik <josef@toxicpanda.com>
Cc: linux-btrfs@vger.kernel.org, kernel-team@fb.com
Subject: Re: [PATCH v2 5/7] btrfs: delay blkdev_put until after the device remove
Date: Thu, 2 Sep 2021 14:16:30 +0200	[thread overview]
Message-ID: <20210902121630.GQ3379@twin.jikos.cz> (raw)
In-Reply-To: <e6af22a1b116e908d26359b55c0d6e2d50fe3105.1627419595.git.josef@toxicpanda.com>

On Tue, Jul 27, 2021 at 05:01:17PM -0400, Josef Bacik wrote:
> When removing the device we call blkdev_put() on the device once we've
> removed it, and because we have an EXCL open we need to take the
> ->open_mutex on the block device to clean it up.  Unfortunately during
> device remove we are holding the sb writers lock, which results in the
> following lockdep splat
> 
> ======================================================
> WARNING: possible circular locking dependency detected
> 5.14.0-rc2+ #407 Not tainted
> ------------------------------------------------------
> losetup/11595 is trying to acquire lock:
> ffff973ac35dd138 ((wq_completion)loop0){+.+.}-{0:0}, at: flush_workqueue+0x67/0x5e0
> 
> but task is already holding lock:
> ffff973ac9812c68 (&lo->lo_mutex){+.+.}-{3:3}, at: __loop_clr_fd+0x41/0x660 [loop]
> 
> which lock already depends on the new lock.
> 
> the existing dependency chain (in reverse order) is:
> 
> -> #4 (&lo->lo_mutex){+.+.}-{3:3}:
>        __mutex_lock+0x7d/0x750
>        lo_open+0x28/0x60 [loop]
>        blkdev_get_whole+0x25/0xf0
>        blkdev_get_by_dev.part.0+0x168/0x3c0
>        blkdev_open+0xd2/0xe0
>        do_dentry_open+0x161/0x390
>        path_openat+0x3cc/0xa20
>        do_filp_open+0x96/0x120
>        do_sys_openat2+0x7b/0x130
>        __x64_sys_openat+0x46/0x70
>        do_syscall_64+0x38/0x90
>        entry_SYSCALL_64_after_hwframe+0x44/0xae
> 
> -> #3 (&disk->open_mutex){+.+.}-{3:3}:
>        __mutex_lock+0x7d/0x750
>        blkdev_put+0x3a/0x220
>        btrfs_rm_device.cold+0x62/0xe5
>        btrfs_ioctl+0x2a31/0x2e70
>        __x64_sys_ioctl+0x80/0xb0
>        do_syscall_64+0x38/0x90
>        entry_SYSCALL_64_after_hwframe+0x44/0xae
> 
> -> #2 (sb_writers#12){.+.+}-{0:0}:
>        lo_write_bvec+0xc2/0x240 [loop]
>        loop_process_work+0x238/0xd00 [loop]
>        process_one_work+0x26b/0x560
>        worker_thread+0x55/0x3c0
>        kthread+0x140/0x160
>        ret_from_fork+0x1f/0x30
> 
> -> #1 ((work_completion)(&lo->rootcg_work)){+.+.}-{0:0}:
>        process_one_work+0x245/0x560
>        worker_thread+0x55/0x3c0
>        kthread+0x140/0x160
>        ret_from_fork+0x1f/0x30
> 
> -> #0 ((wq_completion)loop0){+.+.}-{0:0}:
>        __lock_acquire+0x10ea/0x1d90
>        lock_acquire+0xb5/0x2b0
>        flush_workqueue+0x91/0x5e0
>        drain_workqueue+0xa0/0x110
>        destroy_workqueue+0x36/0x250
>        __loop_clr_fd+0x9a/0x660 [loop]
>        block_ioctl+0x3f/0x50
>        __x64_sys_ioctl+0x80/0xb0
>        do_syscall_64+0x38/0x90
>        entry_SYSCALL_64_after_hwframe+0x44/0xae
> 
> other info that might help us debug this:
> 
> Chain exists of:
>   (wq_completion)loop0 --> &disk->open_mutex --> &lo->lo_mutex
> 
>  Possible unsafe locking scenario:
> 
>        CPU0                    CPU1
>        ----                    ----
>   lock(&lo->lo_mutex);
>                                lock(&disk->open_mutex);
>                                lock(&lo->lo_mutex);
>   lock((wq_completion)loop0);
> 
>  *** DEADLOCK ***
> 
> 1 lock held by losetup/11595:
>  #0: ffff973ac9812c68 (&lo->lo_mutex){+.+.}-{3:3}, at: __loop_clr_fd+0x41/0x660 [loop]
> 
> stack backtrace:
> CPU: 0 PID: 11595 Comm: losetup Not tainted 5.14.0-rc2+ #407
> Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014
> Call Trace:
>  dump_stack_lvl+0x57/0x72
>  check_noncircular+0xcf/0xf0
>  ? stack_trace_save+0x3b/0x50
>  __lock_acquire+0x10ea/0x1d90
>  lock_acquire+0xb5/0x2b0
>  ? flush_workqueue+0x67/0x5e0
>  ? lockdep_init_map_type+0x47/0x220
>  flush_workqueue+0x91/0x5e0
>  ? flush_workqueue+0x67/0x5e0
>  ? verify_cpu+0xf0/0x100
>  drain_workqueue+0xa0/0x110
>  destroy_workqueue+0x36/0x250
>  __loop_clr_fd+0x9a/0x660 [loop]
>  ? blkdev_ioctl+0x8d/0x2a0
>  block_ioctl+0x3f/0x50
>  __x64_sys_ioctl+0x80/0xb0
>  do_syscall_64+0x38/0x90
>  entry_SYSCALL_64_after_hwframe+0x44/0xae
> RIP: 0033:0x7fc21255d4cb
> 
> So instead save the bdev and do the put once we've dropped the sb
> writers lock in order to avoid the lockdep recursion.
> 
> Signed-off-by: Josef Bacik <josef@toxicpanda.com>

Added to misc-next, thanks.

  parent reply	other threads:[~2021-09-02 12:16 UTC|newest]

Thread overview: 39+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-27 21:01 [PATCH v2 0/7] Josef Bacik
2021-07-27 21:01 ` [PATCH v2 1/7] btrfs: do not call close_fs_devices in btrfs_rm_device Josef Bacik
2021-09-01  8:13   ` Anand Jain
2021-07-27 21:01 ` [PATCH v2 2/7] btrfs: do not take the uuid_mutex " Josef Bacik
2021-09-01 12:01   ` Anand Jain
2021-09-01 17:08     ` David Sterba
2021-09-01 17:10     ` Josef Bacik
2021-09-01 19:49       ` Anand Jain
2021-09-02 12:58   ` David Sterba
2021-09-02 14:10     ` Josef Bacik
2021-09-17 14:33       ` David Sterba
2021-09-20  7:45   ` Anand Jain
2021-09-20  8:26     ` David Sterba
2021-09-20  9:41       ` Anand Jain
2021-09-23  4:33         ` Anand Jain
2021-09-21 11:59   ` Filipe Manana
2021-09-21 12:17     ` Filipe Manana
2021-09-22 15:33       ` Filipe Manana
2021-09-23  4:15         ` Anand Jain
2021-09-23  3:58   ` [PATCH] btrfs: drop lockdep assert in close_fs_devices() Anand Jain
2021-09-23  4:04     ` Anand Jain
2021-07-27 21:01 ` [PATCH v2 3/7] btrfs: do not read super look for a device path Josef Bacik
2021-08-25  2:00   ` Anand Jain
2021-09-27 15:32     ` Josef Bacik
2021-09-28 11:50       ` Anand Jain
2021-07-27 21:01 ` [PATCH v2 4/7] btrfs: update the bdev time directly when closing Josef Bacik
2021-08-25  0:35   ` Anand Jain
2021-09-02 12:16   ` David Sterba
2021-07-27 21:01 ` [PATCH v2 5/7] btrfs: delay blkdev_put until after the device remove Josef Bacik
2021-08-25  1:00   ` Anand Jain
2021-09-02 12:16   ` David Sterba [this message]
2021-07-27 21:01 ` [PATCH v2 6/7] btrfs: unify common code for the v1 and v2 versions of " Josef Bacik
2021-08-25  1:19   ` Anand Jain
2021-09-01 14:05   ` Nikolay Borisov
2021-07-27 21:01 ` [PATCH v2 7/7] btrfs: do not take the device_list_mutex in clone_fs_devices Josef Bacik
2021-08-24 22:08   ` Anand Jain
2021-09-01 13:35   ` Nikolay Borisov
2021-09-02 12:59   ` David Sterba
2021-09-17 15:06 ` [PATCH v2 0/7] David Sterba

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210902121630.GQ3379@twin.jikos.cz \
    --to=dsterba@suse.cz \
    --cc=josef@toxicpanda.com \
    --cc=kernel-team@fb.com \
    --cc=linux-btrfs@vger.kernel.org \
    --subject='Re: [PATCH v2 5/7] btrfs: delay blkdev_put until after the device remove' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.