Re: [syzbot] possible deadlock in vmci_qp_broker_detach
From: syzbot
Date: Mon Apr 12 2021 - 13:29:18 EST
syzbot has found a reproducer for the following issue on:
HEAD commit: d434405a Linux 5.12-rc7
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1661482ed00000
kernel config: https://syzkaller.appspot.com/x/.config?x=9c3d8981d2bdb103
dashboard link: https://syzkaller.appspot.com/bug?extid=44e40ac2cfe68e8ce207
compiler: Debian clang version 11.0.1-2
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=102336a6d00000
IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+44e40ac2cfe68e8ce207@xxxxxxxxxxxxxxxxxxxxxxxxx
============================================
WARNING: possible recursive locking detected
5.12.0-rc7-syzkaller #0 Not tainted
--------------------------------------------
syz-executor.0/10571 is trying to acquire lock:
ffffffff8ce6c1f8 (qp_broker_list.mutex){+.+.}-{3:3}, at: vmci_qp_broker_detach+0xd3/0x10c0 drivers/misc/vmw_vmci/vmci_queue_pair.c:2093
but task is already holding lock:
ffffffff8ce6c1f8 (qp_broker_list.mutex){+.+.}-{3:3}, at: vmci_qp_broker_detach+0xd3/0x10c0 drivers/misc/vmw_vmci/vmci_queue_pair.c:2093
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(qp_broker_list.mutex);
lock(qp_broker_list.mutex);
*** DEADLOCK ***
May be due to missing lock nesting notation
1 lock held by syz-executor.0/10571:
#0: ffffffff8ce6c1f8 (qp_broker_list.mutex){+.+.}-{3:3}, at: vmci_qp_broker_detach+0xd3/0x10c0 drivers/misc/vmw_vmci/vmci_queue_pair.c:2093
stack backtrace:
CPU: 1 PID: 10571 Comm: syz-executor.0 Not tainted 5.12.0-rc7-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
__dump_stack lib/dump_stack.c:79 [inline]
dump_stack+0x176/0x24e lib/dump_stack.c:120
__lock_acquire+0x2303/0x5e60 kernel/locking/lockdep.c:4739
lock_acquire+0x126/0x650 kernel/locking/lockdep.c:5511
__mutex_lock_common+0x167/0x2eb0 kernel/locking/mutex.c:949
__mutex_lock kernel/locking/mutex.c:1096 [inline]
mutex_lock_nested+0x1a/0x20 kernel/locking/mutex.c:1111
vmci_qp_broker_detach+0xd3/0x10c0 drivers/misc/vmw_vmci/vmci_queue_pair.c:2093
ctx_free_ctx drivers/misc/vmw_vmci/vmci_context.c:444 [inline]
kref_put include/linux/kref.h:65 [inline]
vmci_ctx_put+0x722/0xe00 drivers/misc/vmw_vmci/vmci_context.c:497
vmci_ctx_enqueue_datagram+0x3a7/0x440 drivers/misc/vmw_vmci/vmci_context.c:360
dg_dispatch_as_host drivers/misc/vmw_vmci/vmci_datagram.c:275 [inline]
vmci_datagram_dispatch+0x3ec/0xb40 drivers/misc/vmw_vmci/vmci_datagram.c:339
qp_notify_peer drivers/misc/vmw_vmci/vmci_queue_pair.c:1479 [inline]
vmci_qp_broker_detach+0x9fa/0x10c0 drivers/misc/vmw_vmci/vmci_queue_pair.c:2186
ctx_free_ctx drivers/misc/vmw_vmci/vmci_context.c:444 [inline]
kref_put include/linux/kref.h:65 [inline]
vmci_ctx_put+0x722/0xe00 drivers/misc/vmw_vmci/vmci_context.c:497
vmci_host_close+0x96/0x160 drivers/misc/vmw_vmci/vmci_host.c:143
__fput+0x352/0x7b0 fs/file_table.c:280
task_work_run+0x146/0x1c0 kernel/task_work.c:140
tracehook_notify_resume include/linux/tracehook.h:189 [inline]
exit_to_user_mode_loop kernel/entry/common.c:174 [inline]
exit_to_user_mode_prepare+0x10b/0x1e0 kernel/entry/common.c:208
__syscall_exit_to_user_mode_work kernel/entry/common.c:290 [inline]
syscall_exit_to_user_mode+0x26/0x70 kernel/entry/common.c:301
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x41926b
Code: 0f 05 48 3d 00 f0 ff ff 77 45 c3 0f 1f 40 00 48 83 ec 18 89 7c 24 0c e8 63 fc ff ff 8b 7c 24 0c 41 89 c0 b8 03 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 35 44 89 c7 89 44 24 0c e8 a1 fc ff ff 8b 44
RSP: 002b:00007ffee76536f0 EFLAGS: 00000293 ORIG_RAX: 0000000000000003
RAX: 0000000000000000 RBX: 0000000000000004 RCX: 000000000041926b
RDX: 0000000000570698 RSI: 0000000000000001 RDI: 0000000000000003
RBP: 0000000000000001 R08: 0000000000000000 R09: 0000001b30e200a8
R10: 00007ffee76537e0 R11: 0000000000000293 R12: 00000000000688ea
R13: 00000000000003e8 R14: 000000000056bf60 R15: 00000000000688cf