syzbot


possible deadlock in smc_release

Status: upstream: reported on 2025/11/19 23:52
Reported-by: syzbot+65e45a84f76e6f44eaff@syzkaller.appspotmail.com
First crash: 53d, last: 53d
Similar bugs (3)
Kernel Title Rank 🛈 Repro Cause bisect Fix bisect Count Last Reported Patched Status
linux-5.15 possible deadlock in smc_release missing-backport origin:upstream 4 C error 762 3d01h 474d 0/3 upstream: reported C repro on 2024/09/24 05:56
upstream possible deadlock in smc_release smc 4 C error 184 1d12h 709d 0/29 upstream: reported C repro on 2024/02/02 13:26
linux-6.1 possible deadlock in smc_release origin:upstream missing-backport 4 C 1012 23h23m 713d 0/3 upstream: reported C repro on 2024/01/29 16:17

Sample crash report:
======================================================
WARNING: possible circular locking dependency detected
syzkaller #0 Not tainted
------------------------------------------------------
syz.0.1542/9363 is trying to acquire lock:
ffff8880213d1458 ((work_completion)(&new_smc->smc_listen_work)){+.+.}-{0:0}, at: rcu_lock_acquire include/linux/rcupdate.h:334 [inline]
ffff8880213d1458 ((work_completion)(&new_smc->smc_listen_work)){+.+.}-{0:0}, at: rcu_read_lock include/linux/rcupdate.h:786 [inline]
ffff8880213d1458 ((work_completion)(&new_smc->smc_listen_work)){+.+.}-{0:0}, at: start_flush_work kernel/workqueue.c:3371 [inline]
ffff8880213d1458 ((work_completion)(&new_smc->smc_listen_work)){+.+.}-{0:0}, at: __flush_work+0xe6/0x9f0 kernel/workqueue.c:3429

but task is already holding lock:
ffff8880213d0130 (sk_lock-AF_SMC/1){+.+.}-{0:0}, at: smc_release+0x255/0x530 net/smc/af_smc.c:-1

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (sk_lock-AF_SMC/1){+.+.}-{0:0}:
       lock_sock_nested+0x48/0x100 net/core/sock.c:3532
       smc_listen_out+0x109/0x3d0 net/smc/af_smc.c:1912
       process_one_work kernel/workqueue.c:2634 [inline]
       process_scheduled_works+0xa45/0x15b0 kernel/workqueue.c:2711
       worker_thread+0xa55/0xfc0 kernel/workqueue.c:2792
       kthread+0x2fa/0x390 kernel/kthread.c:388
       ret_from_fork+0x48/0x80 arch/x86/kernel/process.c:152
       ret_from_fork_asm+0x11/0x20 arch/x86/entry/entry_64.S:293

-> #0 ((work_completion)(&new_smc->smc_listen_work)){+.+.}-{0:0}:
       check_prev_add kernel/locking/lockdep.c:3134 [inline]
       check_prevs_add kernel/locking/lockdep.c:3253 [inline]
       validate_chain kernel/locking/lockdep.c:3869 [inline]
       __lock_acquire+0x2ddb/0x7c80 kernel/locking/lockdep.c:5137
       lock_acquire+0x197/0x410 kernel/locking/lockdep.c:5754
       touch_work_lockdep_map kernel/workqueue.c:3145 [inline]
       start_flush_work kernel/workqueue.c:3397 [inline]
       __flush_work+0x5ed/0x9f0 kernel/workqueue.c:3429
       __cancel_work_timer+0x3b0/0x520 kernel/workqueue.c:3517
       smc_clcsock_release+0x60/0xf0 net/smc/smc_close.c:29
       __smc_release+0x661/0x7d0 net/smc/af_smc.c:301
       smc_close_non_accepted+0xd5/0x1f0 net/smc/af_smc.c:1843
       smc_close_cleanup_listen net/smc/smc_close.c:45 [inline]
       smc_close_active+0xb61/0xf00 net/smc/smc_close.c:225
       __smc_release+0x8d/0x7d0 net/smc/af_smc.c:277
       smc_release+0x2ce/0x530 net/smc/af_smc.c:344
       __sock_release net/socket.c:659 [inline]
       sock_close+0xbd/0x230 net/socket.c:1421
       __fput+0x234/0x970 fs/file_table.c:384
       task_work_run+0x1ce/0x250 kernel/task_work.c:239
       resume_user_mode_work include/linux/resume_user_mode.h:49 [inline]
       exit_to_user_mode_loop+0xe6/0x110 kernel/entry/common.c:177
       exit_to_user_mode_prepare+0xf6/0x180 kernel/entry/common.c:210
       __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
       syscall_exit_to_user_mode+0x1a/0x50 kernel/entry/common.c:302
       do_syscall_64+0x61/0xb0 arch/x86/entry/common.c:87
       entry_SYSCALL_64_after_hwframe+0x68/0xd2

other info that might help us debug this:

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(sk_lock-AF_SMC/1);
                               lock((work_completion)(&new_smc->smc_listen_work));
                               lock(sk_lock-AF_SMC/1);
  lock((work_completion)(&new_smc->smc_listen_work));

 *** DEADLOCK ***

3 locks held by syz.0.1542/9363:
 #0: ffff88805b95be20 (&sb->s_type->i_mutex_key#11){+.+.}-{3:3}, at: inode_lock include/linux/fs.h:804 [inline]
 #0: ffff88805b95be20 (&sb->s_type->i_mutex_key#11){+.+.}-{3:3}, at: __sock_release net/socket.c:658 [inline]
 #0: ffff88805b95be20 (&sb->s_type->i_mutex_key#11){+.+.}-{3:3}, at: sock_close+0x9b/0x230 net/socket.c:1421
 #1: ffff8880213d0130 (sk_lock-AF_SMC/1){+.+.}-{0:0}, at: smc_release+0x255/0x530 net/smc/af_smc.c:-1
 #2: ffffffff8cd2fee0 (rcu_read_lock){....}-{1:2}, at: rcu_lock_acquire include/linux/rcupdate.h:334 [inline]
 #2: ffffffff8cd2fee0 (rcu_read_lock){....}-{1:2}, at: rcu_read_lock include/linux/rcupdate.h:786 [inline]
 #2: ffffffff8cd2fee0 (rcu_read_lock){....}-{1:2}, at: start_flush_work kernel/workqueue.c:3371 [inline]
 #2: ffffffff8cd2fee0 (rcu_read_lock){....}-{1:2}, at: __flush_work+0xe6/0x9f0 kernel/workqueue.c:3429

stack backtrace:
CPU: 1 PID: 9363 Comm: syz.0.1542 Not tainted syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/02/2025
Call Trace:
 <TASK>
 dump_stack_lvl+0x16c/0x230 lib/dump_stack.c:106
 check_noncircular+0x2bd/0x3c0 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 kernel/locking/lockdep.c:3869 [inline]
 __lock_acquire+0x2ddb/0x7c80 kernel/locking/lockdep.c:5137
 lock_acquire+0x197/0x410 kernel/locking/lockdep.c:5754
 touch_work_lockdep_map kernel/workqueue.c:3145 [inline]
 start_flush_work kernel/workqueue.c:3397 [inline]
 __flush_work+0x5ed/0x9f0 kernel/workqueue.c:3429
 __cancel_work_timer+0x3b0/0x520 kernel/workqueue.c:3517
 smc_clcsock_release+0x60/0xf0 net/smc/smc_close.c:29
 __smc_release+0x661/0x7d0 net/smc/af_smc.c:301
 smc_close_non_accepted+0xd5/0x1f0 net/smc/af_smc.c:1843
 smc_close_cleanup_listen net/smc/smc_close.c:45 [inline]
 smc_close_active+0xb61/0xf00 net/smc/smc_close.c:225
 __smc_release+0x8d/0x7d0 net/smc/af_smc.c:277
 smc_release+0x2ce/0x530 net/smc/af_smc.c:344
 __sock_release net/socket.c:659 [inline]
 sock_close+0xbd/0x230 net/socket.c:1421
 __fput+0x234/0x970 fs/file_table.c:384
 task_work_run+0x1ce/0x250 kernel/task_work.c:239
 resume_user_mode_work include/linux/resume_user_mode.h:49 [inline]
 exit_to_user_mode_loop+0xe6/0x110 kernel/entry/common.c:177
 exit_to_user_mode_prepare+0xf6/0x180 kernel/entry/common.c:210
 __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline]
 syscall_exit_to_user_mode+0x1a/0x50 kernel/entry/common.c:302
 do_syscall_64+0x61/0xb0 arch/x86/entry/common.c:87
 entry_SYSCALL_64_after_hwframe+0x68/0xd2
RIP: 0033:0x7f00a4b8f749
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 a8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f00a5943038 EFLAGS: 00000246 ORIG_RAX: 00000000000001b4
RAX: 0000000000000000 RBX: 00007f00a4de5fa0 RCX: 00007f00a4b8f749
RDX: 0000000000000000 RSI: ffffffffffffffff RDI: 0000000000000000
RBP: 00007f00a4c13f91 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007f00a4de6038 R14: 00007f00a4de5fa0 R15: 00007ffe0417a9d8
 </TASK>

Crashes (1):
Time Kernel Commit Syzkaller Config Log Report Syz repro C repro VM info Assets (help?) Manager Title
2025/11/19 23:51 linux-6.6.y 0a805b6ea8cd 26ee5237 .config console log report info [disk image] [vmlinux] [kernel image] ci2-linux-6-6-kasan possible deadlock in smc_release
* Struck through repros no longer work on HEAD.