[syzbot] [bluetooth?] possible deadlock in hci_dev_do_close (2)

[syzbot <syzbot+c55f7bc8d4809b2bad59@syzkaller.appspotmail.com>]

Hello,

syzbot found the following issue on:

HEAD commit:    bb7a2467e6be Add linux-next specific files for 20240426
git tree:       linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=1337aaa0980000
kernel config:  https://syzkaller.appspot.com/x/.config?x=5c6a0288262dd108
dashboard link: https://syzkaller.appspot.com/bug?extid=c55f7bc8d4809b2bad59
compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40

Unfortunately, I don't have any reproducer for this issue yet.

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/5175af7dda64/disk-bb7a2467.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/70db0462e868/vmlinux-bb7a2467.xz
kernel image: https://storage.googleapis.com/syzbot-assets/3217fb825698/bzImage-bb7a2467.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+c55f7bc8d4809b2bad59@syzkaller.appspotmail.com

Bluetooth: hci6: Opcode 0x0c1a failed: -110
Bluetooth: hci6: Error when powering off device on rfkill (-110)
======================================================
WARNING: possible circular locking dependency detected
6.9.0-rc5-next-20240426-syzkaller #0 Not tainted
------------------------------------------------------
kworker/0:3/5122 is trying to acquire lock:
ffff88805c0888d0 ((work_completion)(&hdev->cmd_sync_work)){+.+.}-{0:0}, at: rcu_lock_acquire include/linux/rcupdate.h:329 [inline]
ffff88805c0888d0 ((work_completion)(&hdev->cmd_sync_work)){+.+.}-{0:0}, at: rcu_read_lock include/linux/rcupdate.h:841 [inline]
ffff88805c0888d0 ((work_completion)(&hdev->cmd_sync_work)){+.+.}-{0:0}, at: start_flush_work kernel/workqueue.c:4113 [inline]
ffff88805c0888d0 ((work_completion)(&hdev->cmd_sync_work)){+.+.}-{0:0}, at: __flush_work+0xe6/0xd00 kernel/workqueue.c:4172

but task is already holding lock:
ffff88805c089060 (&hdev->req_lock){+.+.}-{3:3}, at: hci_dev_do_close+0x28/0x90 net/bluetooth/hci_core.c:559

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (&hdev->req_lock){+.+.}-{3:3}:
       lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5754
       __mutex_lock_common kernel/locking/mutex.c:608 [inline]
       __mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
       hci_cmd_sync_work+0x1ec/0x400 net/bluetooth/hci_sync.c:309
       process_one_work kernel/workqueue.c:3222 [inline]
       process_scheduled_works+0xa2c/0x1830 kernel/workqueue.c:3303
       worker_thread+0x86d/0xd70 kernel/workqueue.c:3384
       kthread+0x2f0/0x390 kernel/kthread.c:389
       ret_from_fork+0x4b/0x80 arch/x86/kernel/process.c:147
       ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244

-> #0 ((work_completion)(&hdev->cmd_sync_work)){+.+.}-{0:0}:
       check_prev_add kernel/locking/lockdep.c:3134 [inline]
       check_prevs_add kernel/locking/lockdep.c:3253 [inline]
       validate_chain+0x18cb/0x58e0 kernel/locking/lockdep.c:3869
       __lock_acquire+0x1346/0x1fd0 kernel/locking/lockdep.c:5137
       lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5754
       touch_work_lockdep_map kernel/workqueue.c:3885 [inline]
       start_flush_work kernel/workqueue.c:4139 [inline]
       __flush_work+0x73c/0xd00 kernel/workqueue.c:4172
       __cancel_work_sync+0xbc/0x110 kernel/workqueue.c:4322
       hci_cmd_sync_clear+0x30/0x220 net/bluetooth/hci_sync.c:588
       hci_dev_close_sync+0xbae/0x1060 net/bluetooth/hci_sync.c:5188
       hci_dev_do_close+0x30/0x90 net/bluetooth/hci_core.c:561
       hci_rfkill_set_block+0x232/0x300 net/bluetooth/hci_core.c:992
       rfkill_set_block+0x1f1/0x440 net/rfkill/core.c:346
       rfkill_epo+0x84/0x180 net/rfkill/core.c:466
       __rfkill_handle_global_op net/rfkill/input.c:60 [inline]
       rfkill_op_handler+0x121/0x280 net/rfkill/input.c:108
       process_one_work kernel/workqueue.c:3222 [inline]
       process_scheduled_works+0xa2c/0x1830 kernel/workqueue.c:3303
       worker_thread+0x86d/0xd70 kernel/workqueue.c:3384
       kthread+0x2f0/0x390 kernel/kthread.c:389
       ret_from_fork+0x4b/0x80 arch/x86/kernel/process.c:147
       ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244

other info that might help us debug this:

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(&hdev->req_lock);
                               lock((work_completion)(&hdev->cmd_sync_work));
                               lock(&hdev->req_lock);
  lock((work_completion)(&hdev->cmd_sync_work));

 *** DEADLOCK ***

5 locks held by kworker/0:3/5122:
 #0: ffff888015080948 ((wq_completion)events){+.+.}-{0:0}, at: process_one_work kernel/workqueue.c:3197 [inline]
 #0: ffff888015080948 ((wq_completion)events){+.+.}-{0:0}, at: process_scheduled_works+0x90a/0x1830 kernel/workqueue.c:3303
 #1: ffffc9000387fd00 ((rfkill_op_work).work){+.+.}-{0:0}, at: process_one_work kernel/workqueue.c:3198 [inline]
 #1: ffffc9000387fd00 ((rfkill_op_work).work){+.+.}-{0:0}, at: process_scheduled_works+0x945/0x1830 kernel/workqueue.c:3303
 #2: ffffffff8f8b13e8 (rfkill_global_mutex){+.+.}-{3:3}, at: rfkill_epo+0x4f/0x180 net/rfkill/core.c:462
 #3: ffff88805c089060 (&hdev->req_lock){+.+.}-{3:3}, at: hci_dev_do_close+0x28/0x90 net/bluetooth/hci_core.c:559
 #4: ffffffff8e333ba0 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire include/linux/rcupdate.h:329 [inline]
 #4: ffffffff8e333ba0 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock include/linux/rcupdate.h:841 [inline]
 #4: ffffffff8e333ba0 (rcu_read_lock){....}-{1:2}, at: start_flush_work kernel/workqueue.c:4113 [inline]
 #4: ffffffff8e333ba0 (rcu_read_lock){....}-{1:2}, at: __flush_work+0xe6/0xd00 kernel/workqueue.c:4172

stack backtrace:
CPU: 0 PID: 5122 Comm: kworker/0:3 Not tainted 6.9.0-rc5-next-20240426-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/27/2024
Workqueue: events rfkill_op_handler
Call Trace:
 <TASK>
 __dump_stack lib/dump_stack.c:88 [inline]
 dump_stack_lvl+0x241/0x360 lib/dump_stack.c:114
 check_noncircular+0x36a/0x4a0 kernel/locking/lockdep.c:2187
 check_prev_add kernel/locking/lockdep.c:3134 [inline]
 check_prevs_add kernel/locking/lockdep.c:3253 [inline]
 validate_chain+0x18cb/0x58e0 kernel/locking/lockdep.c:3869
 __lock_acquire+0x1346/0x1fd0 kernel/locking/lockdep.c:5137
 lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5754
 touch_work_lockdep_map kernel/workqueue.c:3885 [inline]
 start_flush_work kernel/workqueue.c:4139 [inline]
 __flush_work+0x73c/0xd00 kernel/workqueue.c:4172
 __cancel_work_sync+0xbc/0x110 kernel/workqueue.c:4322
 hci_cmd_sync_clear+0x30/0x220 net/bluetooth/hci_sync.c:588
 hci_dev_close_sync+0xbae/0x1060 net/bluetooth/hci_sync.c:5188
 hci_dev_do_close+0x30/0x90 net/bluetooth/hci_core.c:561
 hci_rfkill_set_block+0x232/0x300 net/bluetooth/hci_core.c:992
 rfkill_set_block+0x1f1/0x440 net/rfkill/core.c:346
 rfkill_epo+0x84/0x180 net/rfkill/core.c:466
 __rfkill_handle_global_op net/rfkill/input.c:60 [inline]
 rfkill_op_handler+0x121/0x280 net/rfkill/input.c:108
 process_one_work kernel/workqueue.c:3222 [inline]
 process_scheduled_works+0xa2c/0x1830 kernel/workqueue.c:3303
 worker_thread+0x86d/0xd70 kernel/workqueue.c:3384
 kthread+0x2f0/0x390 kernel/kthread.c:389
 ret_from_fork+0x4b/0x80 arch/x86/kernel/process.c:147
 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244
 </TASK>


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzkaller@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup