block nbd4: shutting down sockets
============================================
WARNING: possible recursive locking detected
5.4.0-rc7+ #0 Not tainted
--------------------------------------------
kworker/u5:1/24212 is trying to acquire lock:
ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: flush_workqueue+0xf7/0x14c0 kernel/workqueue.c:2771

but task is already holding lock:
ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: __write_once_size include/linux/compiler.h:226 [inline]
ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: arch_atomic64_set arch/x86/include/asm/atomic64_64.h:34 [inline]
ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: atomic64_set include/asm-generic/atomic-instrumented.h:855 [inline]
ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: atomic_long_set include/asm-generic/atomic-long.h:40 [inline]
ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: set_work_data kernel/workqueue.c:620 [inline]
ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: set_work_pool_and_clear_pending kernel/workqueue.c:647 [inline]
ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: process_one_work+0x88b/0x1740 kernel/workqueue.c:2240

other info that might help us debug this:
 Possible unsafe locking scenario:

       CPU0
       ----
  lock((wq_completion)knbd4-recv);
  lock((wq_completion)knbd4-recv);

 *** DEADLOCK ***

 May be due to missing lock nesting notation

3 locks held by kworker/u5:1/24212:
 #0: ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: __write_once_size include/linux/compiler.h:226 [inline]
 #0: ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: arch_atomic64_set arch/x86/include/asm/atomic64_64.h:34 [inline]
 #0: ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: atomic64_set include/asm-generic/atomic-instrumented.h:855 [inline]
 #0: ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: atomic_long_set include/asm-generic/atomic-long.h:40 [inline]
 #0: ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: set_work_data kernel/workqueue.c:620 [inline]
 #0: ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: set_work_pool_and_clear_pending kernel/workqueue.c:647 [inline]
 #0: ffff888098906528 ((wq_completion)knbd4-recv){+.+.}, at: process_one_work+0x88b/0x1740 kernel/workqueue.c:2240
 #1: ffff88809a647dc0 ((work_completion)(&args->work)){+.+.}, at: process_one_work+0x8c1/0x1740 kernel/workqueue.c:2244
 #2: ffff888218d9a578 (&nbd->config_lock){+.+.}, at: refcount_dec_and_mutex_lock lib/refcount.c:319 [inline]
 #2: ffff888218d9a578 (&nbd->config_lock){+.+.}, at: refcount_dec_and_mutex_lock+0x56/0x90 lib/refcount.c:314

stack backtrace:
CPU: 1 PID: 24212 Comm: kworker/u5:1 Not tainted 5.4.0-rc7+ #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Workqueue: knbd4-recv recv_work
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x197/0x210 lib/dump_stack.c:118
 print_deadlock_bug kernel/locking/lockdep.c:2371 [inline]
 check_deadlock kernel/locking/lockdep.c:2412 [inline]
 validate_chain kernel/locking/lockdep.c:2955 [inline]
 __lock_acquire.cold+0x15d/0x385 kernel/locking/lockdep.c:3955
 lock_acquire+0x190/0x410 kernel/locking/lockdep.c:4487
 flush_workqueue+0x126/0x14c0 kernel/workqueue.c:2774
 drain_workqueue+0x1b4/0x470 kernel/workqueue.c:2939
 destroy_workqueue+0x21/0x700 kernel/workqueue.c:4329
 nbd_config_put+0x3dd/0x870 drivers/block/nbd.c:1208
 recv_work+0x19b/0x200 drivers/block/nbd.c:792
 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