possible deadlock in refcount_dec_and_mutex_lock

From: syzbot
Date: Wed Sep 25 2019 - 02:19:13 EST


Hello,

syzbot found the following crash on:

HEAD commit: b5b3bd89 Add linux-next specific files for 20190920
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=12bd8429600000
kernel config: https://syzkaller.appspot.com/x/.config?x=eadcd01c1bd90b7f
dashboard link: https://syzkaller.appspot.com/bug?extid=22987eb604bffbcb717a
compiler: gcc (GCC) 9.0.0 20181231 (experimental)

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

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+22987eb604bffbcb717a@xxxxxxxxxxxxxxxxxxxxxxxxx

block nbd3: Receive control failed (result -22)
======================================================
WARNING: possible circular locking dependency detected
5.3.0-next-20190920 #0 Not tainted
------------------------------------------------------
kworker/u5:0/1524 is trying to acquire lock:
ffff8880a3030978 (&nbd->config_lock){+.+.}, at: refcount_dec_and_mutex_lock lib/refcount.c:319 [inline]
ffff8880a3030978 (&nbd->config_lock){+.+.}, at: refcount_dec_and_mutex_lock+0x56/0x90 lib/refcount.c:314

but task is already holding lock:
ffff8880a63ffdc0 ((work_completion)(&args->work)){+.+.}, at: process_one_work+0x8c1/0x1740 kernel/workqueue.c:2244

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #2 ((work_completion)(&args->work)){+.+.}:
process_one_work+0x91c/0x1740 kernel/workqueue.c:2245
worker_thread+0x98/0xe40 kernel/workqueue.c:2415
kthread+0x361/0x430 kernel/kthread.c:255
ret_from_fork+0x24/0x30 arch/x86/entry/entry_64.S:352

-> #1 ((wq_completion)knbd0-recv){+.+.}:
flush_workqueue+0x126/0x14c0 kernel/workqueue.c:2774
drain_workqueue+0x1b4/0x470 kernel/workqueue.c:2939
destroy_workqueue+0x74/0x750 kernel/workqueue.c:4335
nbd_config_put+0x3dd/0x870 drivers/block/nbd.c:1181
nbd_release+0x103/0x150 drivers/block/nbd.c:1460
__blkdev_put+0x4d1/0x810 fs/block_dev.c:1867
blkdev_put+0x98/0x560 fs/block_dev.c:1929
blkdev_close+0x8b/0xb0 fs/block_dev.c:1936
__fput+0x2ff/0x890 fs/file_table.c:280
____fput+0x16/0x20 fs/file_table.c:313
task_work_run+0x145/0x1c0 kernel/task_work.c:113
exit_task_work include/linux/task_work.h:22 [inline]
do_exit+0x904/0x2e60 kernel/exit.c:879
do_group_exit+0x135/0x360 kernel/exit.c:983
__do_sys_exit_group kernel/exit.c:994 [inline]
__se_sys_exit_group kernel/exit.c:992 [inline]
__x64_sys_exit_group+0x44/0x50 kernel/exit.c:992
do_syscall_64+0xfa/0x760 arch/x86/entry/common.c:290
entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&nbd->config_lock){+.+.}:
check_prev_add kernel/locking/lockdep.c:2476 [inline]
check_prevs_add kernel/locking/lockdep.c:2581 [inline]
validate_chain kernel/locking/lockdep.c:2971 [inline]
__lock_acquire+0x2596/0x4a00 kernel/locking/lockdep.c:3955
lock_acquire+0x190/0x410 kernel/locking/lockdep.c:4487
__mutex_lock_common kernel/locking/mutex.c:956 [inline]
__mutex_lock+0x156/0x13c0 kernel/locking/mutex.c:1103
mutex_lock_nested+0x16/0x20 kernel/locking/mutex.c:1118
refcount_dec_and_mutex_lock lib/refcount.c:319 [inline]
refcount_dec_and_mutex_lock+0x56/0x90 lib/refcount.c:314
nbd_config_put+0x31/0x870 drivers/block/nbd.c:1159
recv_work+0x19b/0x200 drivers/block/nbd.c:787
process_one_work+0x9af/0x1740 kernel/workqueue.c:2269
worker_thread+0x98/0xe40 kernel/workqueue.c:2415
kthread+0x361/0x430 kernel/kthread.c:255
ret_from_fork+0x24/0x30 arch/x86/entry/entry_64.S:352

other info that might help us debug this:

Chain exists of:
&nbd->config_lock --> (wq_completion)knbd0-recv --> (work_completion)(&args->work)

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock((work_completion)(&args->work));
lock((wq_completion)knbd0-recv);
lock((work_completion)(&args->work));
lock(&nbd->config_lock);

*** DEADLOCK ***

2 locks held by kworker/u5:0/1524:
#0: ffff8880948d3d28 ((wq_completion)knbd3-recv){+.+.}, at: __write_once_size include/linux/compiler.h:226 [inline]
#0: ffff8880948d3d28 ((wq_completion)knbd3-recv){+.+.}, at: arch_atomic64_set arch/x86/include/asm/atomic64_64.h:34 [inline]
#0: ffff8880948d3d28 ((wq_completion)knbd3-recv){+.+.}, at: atomic64_set include/asm-generic/atomic-instrumented.h:855 [inline]
#0: ffff8880948d3d28 ((wq_completion)knbd3-recv){+.+.}, at: atomic_long_set include/asm-generic/atomic-long.h:40 [inline]
#0: ffff8880948d3d28 ((wq_completion)knbd3-recv){+.+.}, at: set_work_data kernel/workqueue.c:620 [inline]
#0: ffff8880948d3d28 ((wq_completion)knbd3-recv){+.+.}, at: set_work_pool_and_clear_pending kernel/workqueue.c:647 [inline]
#0: ffff8880948d3d28 ((wq_completion)knbd3-recv){+.+.}, at: process_one_work+0x88b/0x1740 kernel/workqueue.c:2240
#1: ffff8880a63ffdc0 ((work_completion)(&args->work)){+.+.}, at: process_one_work+0x8c1/0x1740 kernel/workqueue.c:2244

stack backtrace:
CPU: 1 PID: 1524 Comm: kworker/u5:0 Not tainted 5.3.0-next-20190920 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Workqueue: knbd3-recv recv_work
Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x172/0x1f0 lib/dump_stack.c:113
print_circular_bug.isra.0.cold+0x163/0x172 kernel/locking/lockdep.c:1685
check_noncircular+0x32e/0x3e0 kernel/locking/lockdep.c:1809
check_prev_add kernel/locking/lockdep.c:2476 [inline]
check_prevs_add kernel/locking/lockdep.c:2581 [inline]
validate_chain kernel/locking/lockdep.c:2971 [inline]
__lock_acquire+0x2596/0x4a00 kernel/locking/lockdep.c:3955
lock_acquire+0x190/0x410 kernel/locking/lockdep.c:4487
__mutex_lock_common kernel/locking/mutex.c:956 [inline]
__mutex_lock+0x156/0x13c0 kernel/locking/mutex.c:1103
mutex_lock_nested+0x16/0x20 kernel/locking/mutex.c:1118
refcount_dec_and_mutex_lock lib/refcount.c:319 [inline]
refcount_dec_and_mutex_lock+0x56/0x90 lib/refcount.c:314
nbd_config_put+0x31/0x870 drivers/block/nbd.c:1159
recv_work+0x19b/0x200 drivers/block/nbd.c:787
process_one_work+0x9af/0x1740 kernel/workqueue.c:2269
worker_thread+0x98/0xe40 kernel/workqueue.c:2415
kthread+0x361/0x430 kernel/kthread.c:255
ret_from_fork+0x24/0x30 arch/x86/entry/entry_64.S:352
block nbd3: shutting down sockets


---
This bug 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@xxxxxxxxxxxxxxxxx

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