Re: [syzbot] inconsistent lock state in io_link_timeout_fn

From: Pavel Begunkov
Date: Fri Nov 26 2021 - 09:52:36 EST


On 11/26/21 09:37, syzbot wrote:
Hello,

syzbot found the following issue on:

#syz dup: inconsistent lock state in io_poll_remove_all


HEAD commit: a4849f6000e2 Merge tag 'drm-fixes-2021-11-26' of git://ano..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1542553eb00000
kernel config: https://syzkaller.appspot.com/x/.config?x=75f05fb8d1a152d3
dashboard link: https://syzkaller.appspot.com/bug?extid=3368aadcd30425ceb53b
compiler: gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2

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

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+3368aadcd30425ceb53b@xxxxxxxxxxxxxxxxxxxxxxxxx

================================
WARNING: inconsistent lock state
5.16.0-rc2-syzkaller #0 Not tainted
--------------------------------
inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage.
syz-executor.1/14925 [HC1[1]:SC0[0]:HE0:SE1] takes:
ffff888028b7e418 (&ctx->timeout_lock){?.+.}-{2:2}, at: io_link_timeout_fn+0x6b/0x470 fs/io_uring.c:6904
{HARDIRQ-ON-W} state was registered at:
__trace_hardirqs_on_caller kernel/locking/lockdep.c:4224 [inline]
lockdep_hardirqs_on_prepare kernel/locking/lockdep.c:4292 [inline]
lockdep_hardirqs_on_prepare+0x135/0x400 kernel/locking/lockdep.c:4244
trace_hardirqs_on+0x5b/0x1c0 kernel/trace/trace_preemptirq.c:49
__raw_spin_unlock_irq include/linux/spinlock_api_smp.h:159 [inline]
_raw_spin_unlock_irq+0x1f/0x40 kernel/locking/spinlock.c:202
spin_unlock_irq include/linux/spinlock.h:399 [inline]
__io_poll_remove_one fs/io_uring.c:5669 [inline]
__io_poll_remove_one fs/io_uring.c:5654 [inline]
io_poll_remove_one+0x236/0x870 fs/io_uring.c:5680
io_poll_remove_all+0x1af/0x235 fs/io_uring.c:5709
io_ring_ctx_wait_and_kill+0x1cc/0x322 fs/io_uring.c:9534
io_uring_release+0x42/0x46 fs/io_uring.c:9554
__fput+0x286/0x9f0 fs/file_table.c:280
task_work_run+0xdd/0x1a0 kernel/task_work.c:164
tracehook_notify_resume include/linux/tracehook.h:189 [inline]
exit_to_user_mode_loop kernel/entry/common.c:175 [inline]
exit_to_user_mode_prepare+0x27e/0x290 kernel/entry/common.c:207
__syscall_exit_to_user_mode_work kernel/entry/common.c:289 [inline]
syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:300
do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86
entry_SYSCALL_64_after_hwframe+0x44/0xae
irq event stamp: 1634
hardirqs last enabled at (1633): [<ffffffff8946bc7f>] __raw_spin_unlock_irq include/linux/spinlock_api_smp.h:159 [inline]
hardirqs last enabled at (1633): [<ffffffff8946bc7f>] _raw_spin_unlock_irq+0x1f/0x40 kernel/locking/spinlock.c:202
hardirqs last disabled at (1634): [<ffffffff8943d3fb>] sysvec_apic_timer_interrupt+0xb/0xc0 arch/x86/kernel/apic/apic.c:1097
softirqs last enabled at (238): [<ffffffff8146a9b3>] invoke_softirq kernel/softirq.c:432 [inline]
softirqs last enabled at (238): [<ffffffff8146a9b3>] __irq_exit_rcu+0x123/0x180 kernel/softirq.c:636
softirqs last disabled at (233): [<ffffffff8146a9b3>] invoke_softirq kernel/softirq.c:432 [inline]
softirqs last disabled at (233): [<ffffffff8146a9b3>] __irq_exit_rcu+0x123/0x180 kernel/softirq.c:636

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

CPU0
----
lock(&ctx->timeout_lock);
<Interrupt>
lock(&ctx->timeout_lock);

*** DEADLOCK ***

1 lock held by syz-executor.1/14925:
#0: ffff888028b7e0a8 (&ctx->uring_lock){+.+.}-{3:3}, at: __do_sys_io_uring_enter+0xf60/0x1f50 fs/io_uring.c:10042

stack backtrace:
CPU: 0 PID: 14925 Comm: syz-executor.1 Not tainted 5.16.0-rc2-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
<IRQ>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
print_usage_bug kernel/locking/lockdep.c:203 [inline]
valid_state kernel/locking/lockdep.c:3945 [inline]
mark_lock_irq kernel/locking/lockdep.c:4148 [inline]
mark_lock.cold+0x61/0x8e kernel/locking/lockdep.c:4605
mark_usage kernel/locking/lockdep.c:4497 [inline]
__lock_acquire+0x149d/0x54a0 kernel/locking/lockdep.c:4981
lock_acquire kernel/locking/lockdep.c:5637 [inline]
lock_acquire+0x1ab/0x510 kernel/locking/lockdep.c:5602
__raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
_raw_spin_lock_irqsave+0x39/0x50 kernel/locking/spinlock.c:162
io_link_timeout_fn+0x6b/0x470 fs/io_uring.c:6904
__run_hrtimer kernel/time/hrtimer.c:1685 [inline]
__hrtimer_run_queues+0x609/0xe50 kernel/time/hrtimer.c:1749
hrtimer_interrupt+0x31c/0x790 kernel/time/hrtimer.c:1811
local_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1086 [inline]
__sysvec_apic_timer_interrupt+0x146/0x530 arch/x86/kernel/apic/apic.c:1103
sysvec_apic_timer_interrupt+0x8e/0xc0 arch/x86/kernel/apic/apic.c:1097
</IRQ>
<TASK>
asm_sysvec_apic_timer_interrupt+0x12/0x20 arch/x86/include/asm/idtentry.h:638
RIP: 0010:__raw_spin_unlock_irq include/linux/spinlock_api_smp.h:160 [inline]
RIP: 0010:_raw_spin_unlock_irq+0x25/0x40 kernel/locking/spinlock.c:202
Code: 0f 1f 44 00 00 55 48 8b 74 24 08 48 89 fd 48 83 c7 18 e8 6e f0 15 f8 48 89 ef e8 b6 66 16 f8 e8 81 20 37 f8 fb bf 01 00 00 00 <e8> 86 30 09 f8 65 8b 05 6f b3 bb 76 85 c0 74 02 5d c3 e8 8b 88 b9
RSP: 0018:ffffc90004827ae8 EFLAGS: 00000202
RAX: 0000000000000661 RBX: ffff888028b7e000 RCX: 1ffffffff1ffa99e
RDX: 0000000000000000 RSI: 0000000000000001 RDI: 0000000000000001
RBP: ffff888028b7e400 R08: 0000000000000001 R09: ffffffff8ff71b07
R10: 0000000000000001 R11: 0000000000000000 R12: 0000000000000001
R13: ffff888028b7e400 R14: ffff888028b7e188 R15: ffff88802650a290
spin_unlock_irq include/linux/spinlock.h:399 [inline]
io_queue_linked_timeout+0x292/0x430 fs/io_uring.c:6943
io_queue_sqe_arm_apoll+0x15d/0x1a0 fs/io_uring.c:6967
__io_queue_sqe fs/io_uring.c:6991 [inline]
io_queue_sqe fs/io_uring.c:7018 [inline]
io_submit_sqe fs/io_uring.c:7221 [inline]
io_submit_sqes+0x796a/0x8a20 fs/io_uring.c:7327
__do_sys_io_uring_enter+0xf6e/0x1f50 fs/io_uring.c:10043
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7ff8e9ef4ae9
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 bc ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007ff8e746a188 EFLAGS: 00000246 ORIG_RAX: 00000000000001aa
RAX: ffffffffffffffda RBX: 00007ff8ea007f60 RCX: 00007ff8e9ef4ae9
RDX: 0000000000000000 RSI: 0000000000006700 RDI: 0000000000000005
RBP: 00007ff8e9f4ef6d R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007ff8ea53bb2f R14: 00007ff8e746a300 R15: 0000000000022000
</TASK>
----------------
Code disassembly (best guess):
0: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
5: 55 push %rbp
6: 48 8b 74 24 08 mov 0x8(%rsp),%rsi
b: 48 89 fd mov %rdi,%rbp
e: 48 83 c7 18 add $0x18,%rdi
12: e8 6e f0 15 f8 callq 0xf815f085
17: 48 89 ef mov %rbp,%rdi
1a: e8 b6 66 16 f8 callq 0xf81666d5
1f: e8 81 20 37 f8 callq 0xf83720a5
24: fb sti
25: bf 01 00 00 00 mov $0x1,%edi
* 2a: e8 86 30 09 f8 callq 0xf80930b5 <-- trapping instruction
2f: 65 8b 05 6f b3 bb 76 mov %gs:0x76bbb36f(%rip),%eax # 0x76bbb3a5
36: 85 c0 test %eax,%eax
38: 74 02 je 0x3c
3a: 5d pop %rbp
3b: c3 retq
3c: e8 .byte 0xe8
3d: 8b .byte 0x8b
3e: 88 .byte 0x88
3f: b9 .byte 0xb9


---
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@xxxxxxxxxxxxxxxx.

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


--
Pavel Begunkov