possible deadlock in send_sigio (2)

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: dff9f829 Add linux-next specific files for 20200908
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=17521b35900000
kernel config: https://syzkaller.appspot.com/x/.config?x=37b3426c77bda44c
dashboard link: https://syzkaller.appspot.com/bug?extid=907b8537e3b0e55151fc
compiler: gcc (GCC) 10.1.0-syz 20200507

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+907b85...@syzkaller.appspotmail.com

=====================================================
WARNING: SOFTIRQ-safe -> SOFTIRQ-unsafe lock order detected
5.9.0-rc4-next-20200908-syzkaller #0 Not tainted
-----------------------------------------------------
syz-executor.5/15112 [HC0[0]:SC0[0]:HE0:SE1] is trying to acquire:
ffff8880a828e478 (&f->f_owner.lock){.+.?}-{2:2}, at: send_sigio+0x24/0x320 fs/fcntl.c:786

and this task is already holding:
ffff88809d98b9f0 (&new->fa_lock){....}-{2:2}, at: kill_fasync_rcu fs/fcntl.c:1002 [inline]
ffff88809d98b9f0 (&new->fa_lock){....}-{2:2}, at: kill_fasync fs/fcntl.c:1023 [inline]
ffff88809d98b9f0 (&new->fa_lock){....}-{2:2}, at: kill_fasync+0x14b/0x460 fs/fcntl.c:1016
which would create a new lock dependency:
(&new->fa_lock){....}-{2:2} -> (&f->f_owner.lock){.+.?}-{2:2}

but this new dependency connects a SOFTIRQ-irq-safe lock:
(fasync_lock){+.+.}-{2:2}

... which became SOFTIRQ-irq-safe at:
lock_acquire+0x1f3/0xaf0 kernel/locking/lockdep.c:5398
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
spin_lock include/linux/spinlock.h:354 [inline]
fasync_remove_entry+0x2e/0x1f0 fs/fcntl.c:875
fasync_helper+0x9e/0xb0 fs/fcntl.c:983
__tty_fasync drivers/tty/tty_io.c:2126 [inline]
tty_release+0x16d/0xf60 drivers/tty/tty_io.c:1664
__fput+0x285/0x920 fs/file_table.c:281
task_work_run+0xdd/0x190 kernel/task_work.c:141
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_user_mode_loop kernel/entry/common.c:163 [inline]
exit_to_user_mode_prepare+0x1e2/0x1f0 kernel/entry/common.c:190
syscall_exit_to_user_mode+0x7e/0x2e0 kernel/entry/common.c:265
entry_SYSCALL_64_after_hwframe+0x44/0xa9

to a SOFTIRQ-irq-unsafe lock:
(&f->f_owner.lock){.+.?}-{2:2}

... which became SOFTIRQ-irq-unsafe at:
...
lock_acquire+0x1f3/0xaf0 kernel/locking/lockdep.c:5398
__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x36/0x70 kernel/locking/spinlock.c:223
send_sigurg+0x1e/0xa60 fs/fcntl.c:824
sk_send_sigurg+0x76/0x300 net/core/sock.c:2930
tcp_check_urg.isra.0+0x1f4/0x710 net/ipv4/tcp_input.c:5485
tcp_urg net/ipv4/tcp_input.c:5526 [inline]
tcp_rcv_established+0x10b9/0x1eb0 net/ipv4/tcp_input.c:5858
tcp_v4_do_rcv+0x5d1/0x870 net/ipv4/tcp_ipv4.c:1643
tcp_v4_rcv+0x2d10/0x3750 net/ipv4/tcp_ipv4.c:2025
ip_protocol_deliver_rcu+0x5c/0x880 net/ipv4/ip_input.c:204
ip_local_deliver_finish+0x20a/0x370 net/ipv4/ip_input.c:231
NF_HOOK include/linux/netfilter.h:301 [inline]
NF_HOOK include/linux/netfilter.h:295 [inline]
ip_local_deliver+0x1b3/0x200 net/ipv4/ip_input.c:252
dst_input include/net/dst.h:449 [inline]
ip_rcv_finish+0x1da/0x2f0 net/ipv4/ip_input.c:428
NF_HOOK include/linux/netfilter.h:301 [inline]
NF_HOOK include/linux/netfilter.h:295 [inline]
ip_rcv+0xaa/0xd0 net/ipv4/ip_input.c:539
__netif_receive_skb_one_core+0x114/0x180 net/core/dev.c:5286
__netif_receive_skb+0x27/0x1c0 net/core/dev.c:5400
process_backlog+0x2e1/0x8e0 net/core/dev.c:6242
napi_poll net/core/dev.c:6688 [inline]
net_rx_action+0x4f8/0xf90 net/core/dev.c:6758
__do_softirq+0x1f7/0xa91 kernel/softirq.c:298
run_ksoftirqd kernel/softirq.c:652 [inline]
run_ksoftirqd+0xcf/0x170 kernel/softirq.c:644
smpboot_thread_fn+0x655/0x9e0 kernel/smpboot.c:165
kthread+0x3af/0x4a0 kernel/kthread.c:292
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294

other info that might help us debug this:

Chain exists of:
fasync_lock --> &new->fa_lock --> &f->f_owner.lock

Possible interrupt unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&f->f_owner.lock);
local_irq_disable();
lock(fasync_lock);
lock(&new->fa_lock);
<Interrupt>
lock(fasync_lock);

*** DEADLOCK ***

8 locks held by syz-executor.5/15112:
#0: ffff88809cee8160 (&evdev->mutex){+.+.}-{3:3}, at: evdev_write+0x1cd/0x750 drivers/input/evdev.c:513
#1: ffff88809da12230 (&dev->event_lock){-...}-{2:2}, at: input_inject_event+0xa6/0x310 drivers/input/input.c:466
#2: ffffffff89c68540 (rcu_read_lock){....}-{1:2}, at: is_event_supported drivers/input/input.c:53 [inline]
#2: ffffffff89c68540 (rcu_read_lock){....}-{1:2}, at: is_event_supported drivers/input/input.c:50 [inline]
#2: ffffffff89c68540 (rcu_read_lock){....}-{1:2}, at: input_inject_event+0x92/0x310 drivers/input/input.c:465
#3: ffffffff89c68540 (rcu_read_lock){....}-{1:2}, at: input_pass_values.part.0+0x0/0x700 drivers/input/input.c:833
#4: ffffffff89c68540 (rcu_read_lock){....}-{1:2}, at: evdev_events+0x55/0x330 drivers/input/evdev.c:297
#5: ffff888097fef028 (&client->buffer_lock){....}-{2:2}, at: spin_lock include/linux/spinlock.h:354 [inline]
#5: ffff888097fef028 (&client->buffer_lock){....}-{2:2}, at: evdev_pass_values+0x195/0xa30 drivers/input/evdev.c:262
#6: ffffffff89c68540 (rcu_read_lock){....}-{1:2}, at: kill_fasync+0x3d/0x460 fs/fcntl.c:1021
#7: ffff88809d98b9f0 (&new->fa_lock){....}-{2:2}, at: kill_fasync_rcu fs/fcntl.c:1002 [inline]
#7: ffff88809d98b9f0 (&new->fa_lock){....}-{2:2}, at: kill_fasync fs/fcntl.c:1023 [inline]
#7: ffff88809d98b9f0 (&new->fa_lock){....}-{2:2}, at: kill_fasync+0x14b/0x460 fs/fcntl.c:1016

the dependencies between SOFTIRQ-irq-safe lock and the holding lock:
-> (fasync_lock){+.+.}-{2:2} {
HARDIRQ-ON-W at:
lock_acquire+0x1f3/0xaf0 kernel/locking/lockdep.c:5398
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
spin_lock include/linux/spinlock.h:354 [inline]
fasync_remove_entry+0x2e/0x1f0 fs/fcntl.c:875
fasync_helper+0x9e/0xb0 fs/fcntl.c:983
__tty_fasync drivers/tty/tty_io.c:2126 [inline]
tty_release+0x16d/0xf60 drivers/tty/tty_io.c:1664
__fput+0x285/0x920 fs/file_table.c:281
task_work_run+0xdd/0x190 kernel/task_work.c:141
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_user_mode_loop kernel/entry/common.c:163 [inline]
exit_to_user_mode_prepare+0x1e2/0x1f0 kernel/entry/common.c:190
syscall_exit_to_user_mode+0x7e/0x2e0 kernel/entry/common.c:265
entry_SYSCALL_64_after_hwframe+0x44/0xa9
SOFTIRQ-ON-W at:
lock_acquire+0x1f3/0xaf0 kernel/locking/lockdep.c:5398
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
spin_lock include/linux/spinlock.h:354 [inline]
fasync_remove_entry+0x2e/0x1f0 fs/fcntl.c:875
fasync_helper+0x9e/0xb0 fs/fcntl.c:983
__tty_fasync drivers/tty/tty_io.c:2126 [inline]
tty_release+0x16d/0xf60 drivers/tty/tty_io.c:1664
__fput+0x285/0x920 fs/file_table.c:281
task_work_run+0xdd/0x190 kernel/task_work.c:141
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_user_mode_loop kernel/entry/common.c:163 [inline]
exit_to_user_mode_prepare+0x1e2/0x1f0 kernel/entry/common.c:190
syscall_exit_to_user_mode+0x7e/0x2e0 kernel/entry/common.c:265
entry_SYSCALL_64_after_hwframe+0x44/0xa9
INITIAL USE at:
lock_acquire+0x1f3/0xaf0 kernel/locking/lockdep.c:5398
__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
spin_lock include/linux/spinlock.h:354 [inline]
fasync_remove_entry+0x2e/0x1f0 fs/fcntl.c:875
fasync_helper+0x9e/0xb0 fs/fcntl.c:983
__tty_fasync drivers/tty/tty_io.c:2126 [inline]
tty_release+0x16d/0xf60 drivers/tty/tty_io.c:1664
__fput+0x285/0x920 fs/file_table.c:281
task_work_run+0xdd/0x190 kernel/task_work.c:141
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_user_mode_loop kernel/entry/common.c:163 [inline]
exit_to_user_mode_prepare+0x1e2/0x1f0 kernel/entry/common.c:190
syscall_exit_to_user_mode+0x7e/0x2e0 kernel/entry/common.c:265
entry_SYSCALL_64_after_hwframe+0x44/0xa9
}
... key at: [<ffffffff89d445f8>] fasync_lock+0x18/0x660
... acquired at:
__raw_write_lock_irq include/linux/rwlock_api_smp.h:196 [inline]
_raw_write_lock_irq+0x94/0xd0 kernel/locking/spinlock.c:311
fasync_remove_entry+0xb6/0x1f0 fs/fcntl.c:880
fasync_helper+0x9e/0xb0 fs/fcntl.c:983
lease_modify fs/locks.c:1522 [inline]
lease_modify+0x28a/0x370 fs/locks.c:1509
locks_remove_lease fs/locks.c:2770 [inline]
locks_remove_file+0x2be/0x580 fs/locks.c:2795
__fput+0x1b8/0x920 fs/file_table.c:273
task_work_run+0xdd/0x190 kernel/task_work.c:141
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_user_mode_loop kernel/entry/common.c:163 [inline]
exit_to_user_mode_prepare+0x1e2/0x1f0 kernel/entry/common.c:190
syscall_exit_to_user_mode+0x7e/0x2e0 kernel/entry/common.c:265
entry_SYSCALL_64_after_hwframe+0x44/0xa9

-> (&new->fa_lock){....}-{2:2} {
INITIAL USE at:
lock_acquire+0x1f3/0xaf0 kernel/locking/lockdep.c:5398
__raw_write_lock_irq include/linux/rwlock_api_smp.h:196 [inline]
_raw_write_lock_irq+0x94/0xd0 kernel/locking/spinlock.c:311
fasync_remove_entry+0xb6/0x1f0 fs/fcntl.c:880
fasync_helper+0x9e/0xb0 fs/fcntl.c:983
lease_modify fs/locks.c:1522 [inline]
lease_modify+0x28a/0x370 fs/locks.c:1509
locks_remove_lease fs/locks.c:2770 [inline]
locks_remove_file+0x2be/0x580 fs/locks.c:2795
__fput+0x1b8/0x920 fs/file_table.c:273
task_work_run+0xdd/0x190 kernel/task_work.c:141
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_user_mode_loop kernel/entry/common.c:163 [inline]
exit_to_user_mode_prepare+0x1e2/0x1f0 kernel/entry/common.c:190
syscall_exit_to_user_mode+0x7e/0x2e0 kernel/entry/common.c:265
entry_SYSCALL_64_after_hwframe+0x44/0xa9
(null) at:
general protection fault, probably for non-canonical address 0xdffffc0000000002: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017]
CPU: 0 PID: 15112 Comm: syz-executor.5 Not tainted 5.9.0-rc4-next-20200908-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
RIP: 0010:print_lock_trace kernel/locking/lockdep.c:1751 [inline]
RIP: 0010:print_lock_class_header kernel/locking/lockdep.c:2240 [inline]
RIP: 0010:print_shortest_lock_dependencies.cold+0x110/0x2af kernel/locking/lockdep.c:2263
Code: 48 8b 04 24 48 c1 e8 03 42 80 3c 20 00 74 09 48 8b 3c 24 e8 dd fb de f9 48 8b 04 24 48 8b 00 48 8d 78 14 48 89 fa 48 c1 ea 03 <42> 0f b6 0c 22 48 89 fa 83 e2 07 83 c2 03 38 ca 7c 08 84 c9 0f 85
RSP: 0018:ffffc900057875c0 EFLAGS: 00010003
RAX: 0000000000000001 RBX: ffffc90005787718 RCX: 0000000000000000
RDX: 0000000000000002 RSI: ffffffff815bc817 RDI: 0000000000000015
RBP: ffffc90005787718 R08: 0000000000000004 R09: ffff8880ae620f8b
R10: 0000000000000000 R11: 6c6c756e28202020 R12: dffffc0000000000
R13: ffffffff8c709bb8 R14: 0000000000000009 R15: 0000000000000000
FS: 00007f7930053700(0000) GS:ffff8880ae600000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000004e9cf0 CR3: 000000008f8ef000 CR4: 00000000001506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
print_bad_irq_dependency kernel/locking/lockdep.c:2395 [inline]
check_irq_usage.cold+0x42d/0x5b0 kernel/locking/lockdep.c:2634
check_prev_add kernel/locking/lockdep.c:2823 [inline]
check_prevs_add kernel/locking/lockdep.c:2944 [inline]
validate_chain kernel/locking/lockdep.c:3562 [inline]
__lock_acquire+0x2800/0x55d0 kernel/locking/lockdep.c:4796
lock_acquire+0x1f3/0xaf0 kernel/locking/lockdep.c:5398
__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x5b/0x70 kernel/locking/spinlock.c:223
send_sigio+0x24/0x320 fs/fcntl.c:786
kill_fasync_rcu fs/fcntl.c:1009 [inline]
kill_fasync fs/fcntl.c:1023 [inline]
kill_fasync+0x205/0x460 fs/fcntl.c:1016
__pass_event drivers/input/evdev.c:240 [inline]
evdev_pass_values+0x715/0xa30 drivers/input/evdev.c:279
evdev_events+0x20c/0x330 drivers/input/evdev.c:307
input_to_handler+0x2a0/0x4c0 drivers/input/input.c:115
input_pass_values.part.0+0x284/0x700 drivers/input/input.c:145
input_pass_values drivers/input/input.c:134 [inline]
input_handle_event+0x324/0x1390 drivers/input/input.c:399
input_inject_event+0x2f5/0x310 drivers/input/input.c:471
evdev_write+0x424/0x750 drivers/input/evdev.c:530
vfs_write+0x28e/0x700 fs/read_write.c:593
ksys_write+0x1ee/0x250 fs/read_write.c:648
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x45d5b9
Code: 5d b4 fb ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 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 0f 83 2b b4 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007f7930052c78 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 0000000000038f80 RCX: 000000000045d5b9
RDX: 0000000000000373 RSI: 0000000020000040 RDI: 0000000000000003
RBP: 000000000118cf80 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 000000000118cf4c
R13: 00007ffe25f626af R14: 00007f79300539c0 R15: 000000000118cf4c
Modules linked in:
---[ end trace 7926691370b595f2 ]---
RIP: 0010:print_lock_trace kernel/locking/lockdep.c:1751 [inline]
RIP: 0010:print_lock_class_header kernel/locking/lockdep.c:2240 [inline]
RIP: 0010:print_shortest_lock_dependencies.cold+0x110/0x2af kernel/locking/lockdep.c:2263
Code: 48 8b 04 24 48 c1 e8 03 42 80 3c 20 00 74 09 48 8b 3c 24 e8 dd fb de f9 48 8b 04 24 48 8b 00 48 8d 78 14 48 89 fa 48 c1 ea 03 <42> 0f b6 0c 22 48 89 fa 83 e2 07 83 c2 03 38 ca 7c 08 84 c9 0f 85
RSP: 0018:ffffc900057875c0 EFLAGS: 00010003
RAX: 0000000000000001 RBX: ffffc90005787718 RCX: 0000000000000000
RDX: 0000000000000002 RSI: ffffffff815bc817 RDI: 0000000000000015
RBP: ffffc90005787718 R08: 0000000000000004 R09: ffff8880ae620f8b
R10: 0000000000000000 R11: 6c6c756e28202020 R12: dffffc0000000000
R13: ffffffff8c709bb8 R14: 0000000000000009 R15: 0000000000000000
FS: 00007f7930053700(0000) GS:ffff8880ae600000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00000000004e9cf0 CR3: 000000008f8ef000 CR4: 00000000001506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400


---
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 syzk...@googlegroups.com.

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

[syzbot]

syzbot has found a reproducer for the following issue on:

HEAD commit: 4ef8451b Merge tag 'perf-tools-for-v5.10-2020-11-03' of gi..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1321012a500000
kernel config: https://syzkaller.appspot.com/x/.config?x=4e93228e2f17fb12

dashboard link: https://syzkaller.appspot.com/bug?extid=907b8537e3b0e55151fc
compiler: gcc (GCC) 10.1.0-syz 20200507

syz repro: https://syzkaller.appspot.com/x/repro.syz?x=160ab4b2500000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=12cb8e2a500000

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+907b85...@syzkaller.appspotmail.com

========================================================
WARNING: possible irq lock inversion dependency detected
5.10.0-rc2-syzkaller #0 Not tainted
--------------------------------------------------------
syz-executor814/8489 just changed the state of lock:
ffff888018f97438 (&f->f_owner.lock){.+..}-{2:2}, at: send_sigio+0x24/0x360 fs/fcntl.c:786
but this lock was taken by another, HARDIRQ-safe lock in the past:

(&dev->event_lock){-...}-{2:2}

and interrupts could create inverse lock ordering between them.

other info that might help us debug this:
Chain exists of:

&dev->event_lock --> &new->fa_lock --> &f->f_owner.lock

Possible interrupt unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&f->f_owner.lock);
local_irq_disable();

lock(&dev->event_lock);
lock(&new->fa_lock);
<Interrupt>
lock(&dev->event_lock);

*** DEADLOCK ***

2 locks held by syz-executor814/8489:
#0: ffffffff8ef6adf8 (&fsnotify_mark_srcu){....}-{0:0}, at: fsnotify+0x2e6/0x10a0 fs/notify/fsnotify.c:478
#1: ffff88802de9ebf0 (&mark->lock){+.+.}-{2:2}, at: spin_lock include/linux/spinlock.h:354 [inline]
#1: ffff88802de9ebf0 (&mark->lock){+.+.}-{2:2}, at: dnotify_handle_event+0x47/0x280 fs/notify/dnotify/dnotify.c:89

the shortest dependencies between 2nd lock and 1st lock:
-> (&dev->event_lock){-...}-{2:2} {
IN-HARDIRQ-W at:
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401
__raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
_raw_spin_lock_irqsave+0x39/0x50 kernel/locking/spinlock.c:159
input_event drivers/input/input.c:440 [inline]
input_event+0x7b/0xb0 drivers/input/input.c:433
input_report_key include/linux/input.h:417 [inline]
psmouse_report_standard_buttons+0x2c/0x80 drivers/input/mouse/psmouse-base.c:123
psmouse_report_standard_packet drivers/input/mouse/psmouse-base.c:141 [inline]
psmouse_process_byte+0x1e1/0x890 drivers/input/mouse/psmouse-base.c:232
psmouse_handle_byte+0x41/0x1b0 drivers/input/mouse/psmouse-base.c:274
psmouse_interrupt+0x304/0xf00 drivers/input/mouse/psmouse-base.c:426
serio_interrupt+0x88/0x150 drivers/input/serio/serio.c:1002
i8042_interrupt+0x27a/0x520 drivers/input/serio/i8042.c:598
__handle_irq_event_percpu+0x303/0x8f0 kernel/irq/handle.c:156
handle_irq_event_percpu kernel/irq/handle.c:196 [inline]
handle_irq_event+0x102/0x290 kernel/irq/handle.c:213
handle_edge_irq+0x25f/0xd00 kernel/irq/chip.c:819
asm_call_irq_on_stack+0xf/0x20
__run_irq_on_irqstack arch/x86/include/asm/irq_stack.h:48 [inline]
run_irq_on_irqstack_cond arch/x86/include/asm/irq_stack.h:101 [inline]
handle_irq arch/x86/kernel/irq.c:230 [inline]
__common_interrupt arch/x86/kernel/irq.c:249 [inline]
common_interrupt+0x120/0x200 arch/x86/kernel/irq.c:239
asm_common_interrupt+0x1e/0x40 arch/x86/include/asm/idtentry.h:622
native_restore_fl arch/x86/include/asm/irqflags.h:41 [inline]
arch_local_irq_restore arch/x86/include/asm/irqflags.h:84 [inline]
__raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:160 [inline]
_raw_spin_unlock_irqrestore+0x25/0x50 kernel/locking/spinlock.c:191
spin_unlock_irqrestore include/linux/spinlock.h:409 [inline]
i8042_command+0x111/0x130 drivers/input/serio/i8042.c:348
i8042_aux_write+0xd7/0x120 drivers/input/serio/i8042.c:383
serio_write include/linux/serio.h:125 [inline]
ps2_do_sendbyte+0x2ca/0x710 drivers/input/serio/libps2.c:40
ps2_sendbyte+0x58/0x150 drivers/input/serio/libps2.c:92
cypress_ps2_sendbyte+0x2e/0x160 drivers/input/mouse/cypress_ps2.c:42
cypress_ps2_read_cmd_status drivers/input/mouse/cypress_ps2.c:116 [inline]
cypress_send_ext_cmd+0x1d0/0x8d0 drivers/input/mouse/cypress_ps2.c:189
cypress_detect+0x75/0x190 drivers/input/mouse/cypress_ps2.c:205
psmouse_do_detect drivers/input/mouse/psmouse-base.c:1009 [inline]
psmouse_try_protocol+0x211/0x370 drivers/input/mouse/psmouse-base.c:1023
psmouse_extensions+0x557/0x930 drivers/input/mouse/psmouse-base.c:1146
psmouse_switch_protocol+0x52a/0x740 drivers/input/mouse/psmouse-base.c:1542
psmouse_connect+0x5e6/0xfc0 drivers/input/mouse/psmouse-base.c:1632
serio_connect_driver drivers/input/serio/serio.c:47 [inline]
serio_driver_probe+0x72/0xa0 drivers/input/serio/serio.c:778
really_probe+0x291/0xde0 drivers/base/dd.c:554
driver_probe_device+0x26b/0x3d0 drivers/base/dd.c:738
device_driver_attach+0x228/0x290 drivers/base/dd.c:1013
__driver_attach+0x15b/0x2f0 drivers/base/dd.c:1090
bus_for_each_dev+0x147/0x1d0 drivers/base/bus.c:305
serio_attach_driver drivers/input/serio/serio.c:808 [inline]
serio_handle_event+0x5f6/0xa30 drivers/input/serio/serio.c:227
process_one_work+0x933/0x15a0 kernel/workqueue.c:2272
worker_thread+0x64c/0x1120 kernel/workqueue.c:2418
kthread+0x3af/0x4a0 kernel/kthread.c:292
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:296
INITIAL USE at:
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401
__raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
_raw_spin_lock_irqsave+0x39/0x50 kernel/locking/spinlock.c:159
input_inject_event+0xa6/0x310 drivers/input/input.c:466
__led_set_brightness drivers/leds/led-core.c:48 [inline]
led_set_brightness_nopm drivers/leds/led-core.c:275 [inline]
led_set_brightness_nosleep+0xe6/0x1a0 drivers/leds/led-core.c:292
led_set_brightness+0x134/0x170 drivers/leds/led-core.c:267
led_trigger_event drivers/leds/led-triggers.c:387 [inline]
led_trigger_event+0x70/0xd0 drivers/leds/led-triggers.c:377
kbd_led_trigger_activate+0xfa/0x130 drivers/tty/vt/keyboard.c:1010
led_trigger_set+0x61e/0xbd0 drivers/leds/led-triggers.c:195
led_trigger_set_default drivers/leds/led-triggers.c:259 [inline]
led_trigger_set_default+0x1a6/0x230 drivers/leds/led-triggers.c:246
led_classdev_register_ext+0x5b1/0x7c0 drivers/leds/led-class.c:417
led_classdev_register include/linux/leds.h:190 [inline]
input_leds_connect+0x3fb/0x740 drivers/input/input-leds.c:139
input_attach_handler+0x180/0x1f0 drivers/input/input.c:1031
input_register_device.cold+0xf0/0x307 drivers/input/input.c:2229
atkbd_connect+0x736/0xa00 drivers/input/keyboard/atkbd.c:1293
serio_connect_driver drivers/input/serio/serio.c:47 [inline]
serio_driver_probe+0x72/0xa0 drivers/input/serio/serio.c:778
really_probe+0x291/0xde0 drivers/base/dd.c:554
driver_probe_device+0x26b/0x3d0 drivers/base/dd.c:738
device_driver_attach+0x228/0x290 drivers/base/dd.c:1013
__driver_attach+0x15b/0x2f0 drivers/base/dd.c:1090
bus_for_each_dev+0x147/0x1d0 drivers/base/bus.c:305
serio_attach_driver drivers/input/serio/serio.c:808 [inline]
serio_handle_event+0x5f6/0xa30 drivers/input/serio/serio.c:227
process_one_work+0x933/0x15a0 kernel/workqueue.c:2272
worker_thread+0x64c/0x1120 kernel/workqueue.c:2418
kthread+0x3af/0x4a0 kernel/kthread.c:292
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:296
}
... key at: [<ffffffff8fa3b080>] __key.8+0x0/0x40
... acquired at:

__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
spin_lock include/linux/spinlock.h:354 [inline]

evdev_pass_values.part.0+0xf6/0x970 drivers/input/evdev.c:261
evdev_pass_values drivers/input/evdev.c:253 [inline]
evdev_events+0x28b/0x3f0 drivers/input/evdev.c:306

input_to_handler+0x2a0/0x4c0 drivers/input/input.c:115
input_pass_values.part.0+0x284/0x700 drivers/input/input.c:145
input_pass_values drivers/input/input.c:134 [inline]

input_handle_event+0x324/0x1400 drivers/input/input.c:399
input_inject_event+0x2f5/0x310 drivers/input/input.c:471
evdev_write+0x430/0x760 drivers/input/evdev.c:530
vfs_write+0x28e/0x700 fs/read_write.c:603
ksys_write+0x1ee/0x250 fs/read_write.c:658
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9

-> (&client->buffer_lock){....}-{2:2} {
INITIAL USE at:
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401

__raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
_raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
spin_lock include/linux/spinlock.h:354 [inline]

evdev_pass_values.part.0+0xf6/0x970 drivers/input/evdev.c:261
evdev_pass_values drivers/input/evdev.c:253 [inline]
evdev_events+0x28b/0x3f0 drivers/input/evdev.c:306

input_to_handler+0x2a0/0x4c0 drivers/input/input.c:115
input_pass_values.part.0+0x284/0x700 drivers/input/input.c:145
input_pass_values drivers/input/input.c:134 [inline]

input_handle_event+0x324/0x1400 drivers/input/input.c:399
input_inject_event+0x2f5/0x310 drivers/input/input.c:471
evdev_write+0x430/0x760 drivers/input/evdev.c:530
vfs_write+0x28e/0x700 fs/read_write.c:603
ksys_write+0x1ee/0x250 fs/read_write.c:658
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
}
... key at: [<ffffffff8fa3b580>] __key.4+0x0/0x40
... acquired at:

__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x5b/0x70 kernel/locking/spinlock.c:223

kill_fasync_rcu fs/fcntl.c:1002 [inline]
kill_fasync fs/fcntl.c:1023 [inline]
kill_fasync+0x14b/0x460 fs/fcntl.c:1016
__pass_event drivers/input/evdev.c:240 [inline]
evdev_pass_values.part.0+0x64e/0x970 drivers/input/evdev.c:278
evdev_pass_values drivers/input/evdev.c:253 [inline]
evdev_events+0x28b/0x3f0 drivers/input/evdev.c:306

input_to_handler+0x2a0/0x4c0 drivers/input/input.c:115
input_pass_values.part.0+0x284/0x700 drivers/input/input.c:145
input_pass_values drivers/input/input.c:134 [inline]

input_handle_event+0x324/0x1400 drivers/input/input.c:399
input_inject_event+0x2f5/0x310 drivers/input/input.c:471
evdev_write+0x430/0x760 drivers/input/evdev.c:530
vfs_write+0x28e/0x700 fs/read_write.c:603
ksys_write+0x1ee/0x250 fs/read_write.c:658
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46

entry_SYSCALL_64_after_hwframe+0x44/0xa9

-> (&new->fa_lock){....}-{2:2} {

INITIAL READ USE at:
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401

__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x5b/0x70 kernel/locking/spinlock.c:223

kill_fasync_rcu fs/fcntl.c:1002 [inline]
kill_fasync fs/fcntl.c:1023 [inline]
kill_fasync+0x14b/0x460 fs/fcntl.c:1016
__pass_event drivers/input/evdev.c:240 [inline]
evdev_pass_values.part.0+0x64e/0x970 drivers/input/evdev.c:278
evdev_pass_values drivers/input/evdev.c:253 [inline]
evdev_events+0x28b/0x3f0 drivers/input/evdev.c:306

input_to_handler+0x2a0/0x4c0 drivers/input/input.c:115
input_pass_values.part.0+0x284/0x700 drivers/input/input.c:145
input_pass_values drivers/input/input.c:134 [inline]

input_handle_event+0x324/0x1400 drivers/input/input.c:399
input_inject_event+0x2f5/0x310 drivers/input/input.c:471
evdev_write+0x430/0x760 drivers/input/evdev.c:530
vfs_write+0x28e/0x700 fs/read_write.c:603
ksys_write+0x1ee/0x250 fs/read_write.c:658
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
}
... key at: [<ffffffff8ef677e0>] __key.0+0x0/0x40
... acquired at:

__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x5b/0x70 kernel/locking/spinlock.c:223

send_sigio+0x24/0x360 fs/fcntl.c:786

kill_fasync_rcu fs/fcntl.c:1009 [inline]
kill_fasync fs/fcntl.c:1023 [inline]
kill_fasync+0x205/0x460 fs/fcntl.c:1016
__pass_event drivers/input/evdev.c:240 [inline]

evdev_pass_values.part.0+0x64e/0x970 drivers/input/evdev.c:278
evdev_pass_values drivers/input/evdev.c:253 [inline]
evdev_events+0x28b/0x3f0 drivers/input/evdev.c:306

input_to_handler+0x2a0/0x4c0 drivers/input/input.c:115
input_pass_values.part.0+0x284/0x700 drivers/input/input.c:145
input_pass_values drivers/input/input.c:134 [inline]

input_handle_event+0x324/0x1400 drivers/input/input.c:399
input_inject_event+0x2f5/0x310 drivers/input/input.c:471
evdev_write+0x430/0x760 drivers/input/evdev.c:530
vfs_write+0x28e/0x700 fs/read_write.c:603
ksys_write+0x1ee/0x250 fs/read_write.c:658
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9

-> (&f->f_owner.lock){.+..}-{2:2} {
HARDIRQ-ON-R at:
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401

__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x5b/0x70 kernel/locking/spinlock.c:223

send_sigio+0x24/0x360 fs/fcntl.c:786
dnotify_handle_event+0x148/0x280 fs/notify/dnotify/dnotify.c:97
fsnotify_handle_event fs/notify/fsnotify.c:263 [inline]
send_to_group fs/notify/fsnotify.c:326 [inline]
fsnotify+0xbc1/0x10a0 fs/notify/fsnotify.c:504
fsnotify_parent include/linux/fsnotify.h:71 [inline]
fsnotify_file include/linux/fsnotify.h:90 [inline]
fsnotify_access include/linux/fsnotify.h:247 [inline]
do_iter_read+0x531/0x6e0 fs/read_write.c:806
vfs_readv+0xe5/0x150 fs/read_write.c:921
do_preadv fs/read_write.c:1013 [inline]
__do_sys_preadv fs/read_write.c:1063 [inline]
__se_sys_preadv fs/read_write.c:1058 [inline]
__x64_sys_preadv+0x231/0x310 fs/read_write.c:1058
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
INITIAL USE at:
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401
__raw_write_lock_irq include/linux/rwlock_api_smp.h:196 [inline]
_raw_write_lock_irq+0x32/0x50 kernel/locking/spinlock.c:311
f_modown+0x2a/0x390 fs/fcntl.c:90
fcntl_dirnotify+0x736/0xbd0 fs/notify/dnotify/dnotify.c:351
do_fcntl+0x269/0x1070 fs/fcntl.c:413
__do_sys_fcntl fs/fcntl.c:463 [inline]
__se_sys_fcntl fs/fcntl.c:448 [inline]
__x64_sys_fcntl+0x165/0x1e0 fs/fcntl.c:448
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
INITIAL READ USE at:
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401

__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x5b/0x70 kernel/locking/spinlock.c:223

send_sigio+0x24/0x360 fs/fcntl.c:786

kill_fasync_rcu fs/fcntl.c:1009 [inline]
kill_fasync fs/fcntl.c:1023 [inline]
kill_fasync+0x205/0x460 fs/fcntl.c:1016
__pass_event drivers/input/evdev.c:240 [inline]

evdev_pass_values.part.0+0x64e/0x970 drivers/input/evdev.c:278
evdev_pass_values drivers/input/evdev.c:253 [inline]
evdev_events+0x28b/0x3f0 drivers/input/evdev.c:306

input_to_handler+0x2a0/0x4c0 drivers/input/input.c:115
input_pass_values.part.0+0x284/0x700 drivers/input/input.c:145
input_pass_values drivers/input/input.c:134 [inline]

input_handle_event+0x324/0x1400 drivers/input/input.c:399
input_inject_event+0x2f5/0x310 drivers/input/input.c:471
evdev_write+0x430/0x760 drivers/input/evdev.c:530
vfs_write+0x28e/0x700 fs/read_write.c:603
ksys_write+0x1ee/0x250 fs/read_write.c:658
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
}
... key at: [<ffffffff8ef66a00>] __key.5+0x0/0x40
... acquired at:
mark_usage kernel/locking/lockdep.c:4311 [inline]
__lock_acquire+0x11f5/0x5590 kernel/locking/lockdep.c:4785
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401

__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x5b/0x70 kernel/locking/spinlock.c:223

send_sigio+0x24/0x360 fs/fcntl.c:786
dnotify_handle_event+0x148/0x280 fs/notify/dnotify/dnotify.c:97
fsnotify_handle_event fs/notify/fsnotify.c:263 [inline]
send_to_group fs/notify/fsnotify.c:326 [inline]
fsnotify+0xbc1/0x10a0 fs/notify/fsnotify.c:504
fsnotify_parent include/linux/fsnotify.h:71 [inline]
fsnotify_file include/linux/fsnotify.h:90 [inline]
fsnotify_access include/linux/fsnotify.h:247 [inline]
do_iter_read+0x531/0x6e0 fs/read_write.c:806
vfs_readv+0xe5/0x150 fs/read_write.c:921
do_preadv fs/read_write.c:1013 [inline]
__do_sys_preadv fs/read_write.c:1063 [inline]
__se_sys_preadv fs/read_write.c:1058 [inline]
__x64_sys_preadv+0x231/0x310 fs/read_write.c:1058
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9


stack backtrace:
CPU: 0 PID: 8489 Comm: syz-executor814 Not tainted 5.10.0-rc2-syzkaller #0

Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011

Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x107/0x163 lib/dump_stack.c:118
print_irq_inversion_bug kernel/locking/lockdep.c:3739 [inline]
check_usage_backwards kernel/locking/lockdep.c:3883 [inline]
mark_lock_irq kernel/locking/lockdep.c:3973 [inline]
mark_lock.cold+0x1a/0x74 kernel/locking/lockdep.c:4410
mark_usage kernel/locking/lockdep.c:4311 [inline]
__lock_acquire+0x11f5/0x5590 kernel/locking/lockdep.c:4785
lock_acquire kernel/locking/lockdep.c:5436 [inline]
lock_acquire+0x2a3/0x8c0 kernel/locking/lockdep.c:5401

__raw_read_lock include/linux/rwlock_api_smp.h:149 [inline]
_raw_read_lock+0x5b/0x70 kernel/locking/spinlock.c:223

send_sigio+0x24/0x360 fs/fcntl.c:786
dnotify_handle_event+0x148/0x280 fs/notify/dnotify/dnotify.c:97
fsnotify_handle_event fs/notify/fsnotify.c:263 [inline]
send_to_group fs/notify/fsnotify.c:326 [inline]
fsnotify+0xbc1/0x10a0 fs/notify/fsnotify.c:504
fsnotify_parent include/linux/fsnotify.h:71 [inline]
fsnotify_file include/linux/fsnotify.h:90 [inline]
fsnotify_access include/linux/fsnotify.h:247 [inline]
do_iter_read+0x531/0x6e0 fs/read_write.c:806
vfs_readv+0xe5/0x150 fs/read_write.c:921
do_preadv fs/read_write.c:1013 [inline]
__do_sys_preadv fs/read_write.c:1063 [inline]
__se_sys_preadv fs/read_write.c:1058 [inline]
__x64_sys_preadv+0x231/0x310 fs/read_write.c:1058
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x447ca9
Code: e8 fc ab 02 00 48 83 c4 18 c3 0f 1f 80 00 00 00 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 0f 83 3b cd fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007ffd6e267ef8 EFLAGS: 00000246 ORIG_RAX: 0000000000000127
RAX: ffffffffffffffda RBX: 00007ffd6e267f10 RCX: 0000000000447ca9
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000005
RBP: 00000000006d2018 R08: 0000000000000000 R09: 00000000000000c2
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000404e70
R13: 0000000000404f00 R14: 0000000000000000 R15: 0000000000000000

[syzbot]

syzbot has bisected this issue to:

commit e918188611f073063415f40fae568fa4d86d9044
Author: Boqun Feng <boqun...@gmail.com>
Date: Fri Aug 7 07:42:20 2020 +0000

locking: More accurate annotations for read_lock()

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=15fbef32500000
start commit: 4ef8451b Merge tag 'perf-tools-for-v5.10-2020-11-03' of gi..
git tree: upstream
final oops: https://syzkaller.appspot.com/x/report.txt?x=17fbef32500000
console output: https://syzkaller.appspot.com/x/log.txt?x=13fbef32500000

kernel config: https://syzkaller.appspot.com/x/.config?x=4e93228e2f17fb12
dashboard link: https://syzkaller.appspot.com/bug?extid=907b8537e3b0e55151fc

syz repro: https://syzkaller.appspot.com/x/repro.syz?x=160ab4b2500000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=12cb8e2a500000

Reported-by: syzbot+907b85...@syzkaller.appspotmail.com
Fixes: e918188611f0 ("locking: More accurate annotations for read_lock()")

For information about bisection process see: https://goo.gl/tpsmEJ#bisection

[syzbot]

syzbot suspects this issue was fixed by commit:

commit 8d1ddb5e79374fb277985a6b3faa2ed8631c5b4c
Author: Boqun Feng <boqun...@gmail.com>
Date: Thu Nov 5 06:23:51 2020 +0000

fcntl: Fix potential deadlock in send_sig{io, urg}()

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=17455db4d00000
start commit: 7b1b868e Merge tag 'for-linus' of git://git.kernel.org/pub..
git tree: upstream
kernel config: https://syzkaller.appspot.com/x/.config?x=3416bb960d5c705d
dashboard link: https://syzkaller.appspot.com/bug?extid=907b8537e3b0e55151fc
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=163e046b500000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=12f8b623500000

If the result looks correct, please mark the issue as fixed by replying with:

#syz fix: fcntl: Fix potential deadlock in send_sig{io, urg}()

[syzbot]

> #syz fix: fcntl: Fix potential deadlock in send_sig{io, urg}()

Your 'fix:' command is accepted, but please keep syzkall...@googlegroups.com mailing list in CC next time. It serves as a history of what happened with each bug report. Thank you.

>
> --
> Jeff Layton <jla...@kernel.org>
>

[Dmitry Vyukov]