possible deadlock in proc_pid_attr_write (2)

From: syzbot
Date: Wed Feb 19 2020 - 12:37:14 EST

Next message: Andy Shevchenko: "Re: [PATCH v2] vsprintf: don't obfuscate NULL and error pointers"
Previous message: Paul E. McKenney: "Re: [PATCH v3 16/22] locking/atomics, kcsan: Add KCSAN instrumentation"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hello,

syzbot found the following crash on:

HEAD commit: 9f01828e Add linux-next specific files for 20200214
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=1328a229e00000
kernel config: https://syzkaller.appspot.com/x/.config?x=ffb2752ee30931a0
dashboard link: https://syzkaller.appspot.com/bug?extid=ac4fc482d8eee6b61151
compiler: gcc (GCC) 9.0.0 20181231 (experimental)

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

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+ac4fc482d8eee6b61151@xxxxxxxxxxxxxxxxxxxxxxxxx

======================================================
WARNING: possible circular locking dependency detected
5.6.0-rc1-next-20200214-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor.2/3299 is trying to acquire lock:
ffff88804c4e5150 (&sig->cred_guard_mutex){+.+.}, at: proc_pid_attr_write+0x268/0x6b0 fs/proc/base.c:2683

but task is already holding lock:
ffff888099519c60 (&pipe->mutex/1){+.+.}, at: pipe_lock_nested fs/pipe.c:66 [inline]
ffff888099519c60 (&pipe->mutex/1){+.+.}, at: pipe_lock+0x65/0x80 fs/pipe.c:74

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #3 (&pipe->mutex/1){+.+.}:
__mutex_lock_common kernel/locking/mutex.c:956 [inline]
__mutex_lock+0x156/0x13c0 kernel/locking/mutex.c:1103
mutex_lock_nested+0x16/0x20 kernel/locking/mutex.c:1118
pipe_lock_nested fs/pipe.c:66 [inline]
pipe_lock+0x65/0x80 fs/pipe.c:74
iter_file_splice_write+0x18b/0xc10 fs/splice.c:709
do_splice_from fs/splice.c:863 [inline]
do_splice+0xbae/0x1690 fs/splice.c:1170
__do_sys_splice fs/splice.c:1447 [inline]
__se_sys_splice fs/splice.c:1427 [inline]
__x64_sys_splice+0x2c6/0x330 fs/splice.c:1427
do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294
entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #2 (sb_writers#3){.+.+}:
percpu_down_read include/linux/percpu-rwsem.h:51 [inline]
__sb_start_write+0x255/0x4a0 fs/super.c:1674
sb_start_write include/linux/fs.h:1649 [inline]
mnt_want_write+0x3f/0xc0 fs/namespace.c:354
ovl_want_write+0x76/0xa0 fs/overlayfs/util.c:21
ovl_setattr+0xdd/0x930 fs/overlayfs/inode.c:27
notify_change+0xb6d/0x1060 fs/attr.c:336
chmod_common+0x217/0x460 fs/open.c:561
do_fchmodat+0xbe/0x150 fs/open.c:599
__do_sys_chmod fs/open.c:617 [inline]
__se_sys_chmod fs/open.c:615 [inline]
__x64_sys_chmod+0x5c/0x80 fs/open.c:615
do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294
entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #1 (&ovl_i_mutex_dir_key[depth]){++++}:
down_read+0x95/0x440 kernel/locking/rwsem.c:1492
inode_lock_shared include/linux/fs.h:801 [inline]
do_last fs/namei.c:3400 [inline]
path_openat+0x1c4f/0x33f0 fs/namei.c:3607
do_filp_open+0x192/0x260 fs/namei.c:3637
do_open_execat+0x13b/0x6d0 fs/exec.c:860
__do_execve_file.isra.0+0x16d5/0x2270 fs/exec.c:1791
do_execveat_common fs/exec.c:1897 [inline]
do_execve fs/exec.c:1914 [inline]
__do_sys_execve fs/exec.c:1990 [inline]
__se_sys_execve fs/exec.c:1985 [inline]
__x64_sys_execve+0x8f/0xc0 fs/exec.c:1985
do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294
entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&sig->cred_guard_mutex){+.+.}:
check_prev_add kernel/locking/lockdep.c:2481 [inline]
check_prevs_add kernel/locking/lockdep.c:2586 [inline]
validate_chain kernel/locking/lockdep.c:3203 [inline]
__lock_acquire+0x29cd/0x6320 kernel/locking/lockdep.c:4190
lock_acquire+0x190/0x410 kernel/locking/lockdep.c:4720
__mutex_lock_common kernel/locking/mutex.c:956 [inline]
__mutex_lock+0x156/0x13c0 kernel/locking/mutex.c:1103
mutex_lock_interruptible_nested+0x16/0x20 kernel/locking/mutex.c:1140
proc_pid_attr_write+0x268/0x6b0 fs/proc/base.c:2683
__vfs_write+0x8a/0x110 fs/read_write.c:494
__kernel_write+0x11b/0x3b0 fs/read_write.c:515
write_pipe_buf+0x15d/0x1f0 fs/splice.c:809
splice_from_pipe_feed fs/splice.c:512 [inline]
__splice_from_pipe+0x3ee/0x7c0 fs/splice.c:636
splice_from_pipe+0x108/0x170 fs/splice.c:671
default_file_splice_write+0x3c/0x90 fs/splice.c:821
do_splice_from fs/splice.c:863 [inline]
do_splice+0xbae/0x1690 fs/splice.c:1170
__do_sys_splice fs/splice.c:1447 [inline]
__se_sys_splice fs/splice.c:1427 [inline]
__x64_sys_splice+0x2c6/0x330 fs/splice.c:1427
do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294
entry_SYSCALL_64_after_hwframe+0x49/0xbe

other info that might help us debug this:

Chain exists of:
&sig->cred_guard_mutex --> sb_writers#3 --> &pipe->mutex/1

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&pipe->mutex/1);
lock(sb_writers#3);
lock(&pipe->mutex/1);
lock(&sig->cred_guard_mutex);

*** DEADLOCK ***

2 locks held by syz-executor.2/3299:
#0: ffff8880a9996418 (sb_writers#6){.+.+}, at: file_start_write include/linux/fs.h:2903 [inline]
#0: ffff8880a9996418 (sb_writers#6){.+.+}, at: do_splice+0xf52/0x1690 fs/splice.c:1169
#1: ffff888099519c60 (&pipe->mutex/1){+.+.}, at: pipe_lock_nested fs/pipe.c:66 [inline]
#1: ffff888099519c60 (&pipe->mutex/1){+.+.}, at: pipe_lock+0x65/0x80 fs/pipe.c:74

stack backtrace:
CPU: 1 PID: 3299 Comm: syz-executor.2 Not tainted 5.6.0-rc1-next-20200214-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+0x197/0x210 lib/dump_stack.c:118
print_circular_bug.isra.0.cold+0x163/0x172 kernel/locking/lockdep.c:1688
check_noncircular+0x32e/0x3e0 kernel/locking/lockdep.c:1812
check_prev_add kernel/locking/lockdep.c:2481 [inline]
check_prevs_add kernel/locking/lockdep.c:2586 [inline]
validate_chain kernel/locking/lockdep.c:3203 [inline]
__lock_acquire+0x29cd/0x6320 kernel/locking/lockdep.c:4190
lock_acquire+0x190/0x410 kernel/locking/lockdep.c:4720
__mutex_lock_common kernel/locking/mutex.c:956 [inline]
__mutex_lock+0x156/0x13c0 kernel/locking/mutex.c:1103
mutex_lock_interruptible_nested+0x16/0x20 kernel/locking/mutex.c:1140
proc_pid_attr_write+0x268/0x6b0 fs/proc/base.c:2683
__vfs_write+0x8a/0x110 fs/read_write.c:494
__kernel_write+0x11b/0x3b0 fs/read_write.c:515
write_pipe_buf+0x15d/0x1f0 fs/splice.c:809
splice_from_pipe_feed fs/splice.c:512 [inline]
__splice_from_pipe+0x3ee/0x7c0 fs/splice.c:636
splice_from_pipe+0x108/0x170 fs/splice.c:671
default_file_splice_write+0x3c/0x90 fs/splice.c:821
do_splice_from fs/splice.c:863 [inline]
do_splice+0xbae/0x1690 fs/splice.c:1170
__do_sys_splice fs/splice.c:1447 [inline]
__se_sys_splice fs/splice.c:1427 [inline]
__x64_sys_splice+0x2c6/0x330 fs/splice.c:1427
do_syscall_64+0xfa/0x790 arch/x86/entry/common.c:294
entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x45c6c9
Code: ad b6 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 7b b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007ffa3925ac78 EFLAGS: 00000246 ORIG_RAX: 0000000000000113
RAX: ffffffffffffffda RBX: 00007ffa3925b6d4 RCX: 000000000045c6c9
RDX: 0000000000000005 RSI: 0000000000000000 RDI: 0000000000000004
RBP: 000000000076bf20 R08: 0000000100000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00000000ffffffff
R13: 0000000000000b6e R14: 00000000004ce22b R15: 000000000076bf2c

---
This bug 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@xxxxxxxxxxxxxxxxx

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

Next message: Andy Shevchenko: "Re: [PATCH v2] vsprintf: don't obfuscate NULL and error pointers"
Previous message: Paul E. McKenney: "Re: [PATCH v3 16/22] locking/atomics, kcsan: Add KCSAN instrumentation"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]