possible deadlock in proc_pid_stack

[syzbot]

Hello,

syzbot found the following crash on:

HEAD commit: b3a99fd3 Linux 4.19.129
git tree: linux-4.19.y
console output: https://syzkaller.appspot.com/x/log.txt?x=164bbcf9100000
kernel config: https://syzkaller.appspot.com/x/.config?x=afe60417c22a7f7a
dashboard link: https://syzkaller.appspot.com/bug?extid=920b788cbadd967f2ca4
compiler: gcc (GCC) 10.1.0-syz 20200507

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+920b78...@syzkaller.appspotmail.com

libceph: connect [d::]:6789 error -101
libceph: mon0 [d::]:6789 connect error
======================================================
WARNING: possible circular locking dependency detected
4.19.129-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor.1/7856 is trying to acquire lock:
000000005faf416f (&sig->cred_guard_mutex){+.+.}, at: lock_trace fs/proc/base.c:402 [inline]
000000005faf416f (&sig->cred_guard_mutex){+.+.}, at: proc_pid_stack+0x160/0x350 fs/proc/base.c:452

but task is already holding lock:
00000000d1f52354 (&p->lock){+.+.}, at: seq_read+0x6b/0x1160 fs/seq_file.c:161

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #3 (&p->lock){+.+.}:
seq_read+0x6b/0x1160 fs/seq_file.c:161
proc_reg_read+0x1bd/0x2d0 fs/proc/inode.c:231
do_loop_readv_writev fs/read_write.c:701 [inline]
do_loop_readv_writev fs/read_write.c:688 [inline]
do_iter_read+0x471/0x630 fs/read_write.c:925
vfs_readv+0xe5/0x150 fs/read_write.c:987
kernel_readv fs/splice.c:362 [inline]
default_file_splice_read+0x457/0xa00 fs/splice.c:417
do_splice_to+0x10e/0x160 fs/splice.c:881
splice_direct_to_actor+0x2b9/0x8d0 fs/splice.c:959
do_splice_direct+0x1a7/0x270 fs/splice.c:1068
do_sendfile+0x550/0xc30 fs/read_write.c:1447
__do_sys_sendfile64 fs/read_write.c:1508 [inline]
__se_sys_sendfile64+0x147/0x160 fs/read_write.c:1494
do_syscall_64+0xf9/0x620 arch/x86/entry/common.c:293
entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #2 (sb_writers#3){.+.+}:
sb_start_write include/linux/fs.h:1579 [inline]
mnt_want_write+0x3a/0xb0 fs/namespace.c:360
ovl_do_remove+0xf0/0xdb0 fs/overlayfs/dir.c:843
vfs_rmdir.part.0+0x10f/0x3d0 fs/namei.c:3882
vfs_rmdir fs/namei.c:3868 [inline]
do_rmdir+0x3fd/0x490 fs/namei.c:3943
do_syscall_64+0xf9/0x620 arch/x86/entry/common.c:293
entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #1 (&ovl_i_mutex_dir_key[depth]){++++}:
inode_lock_shared include/linux/fs.h:758 [inline]
do_last fs/namei.c:3326 [inline]
path_openat+0x17ec/0x2df0 fs/namei.c:3537
do_filp_open+0x18c/0x3f0 fs/namei.c:3567
do_open_execat+0x11d/0x5b0 fs/exec.c:853
__do_execve_file+0x1a8b/0x2360 fs/exec.c:1757
do_execveat_common fs/exec.c:1866 [inline]
do_execve+0x35/0x50 fs/exec.c:1883
__do_sys_execve fs/exec.c:1964 [inline]
__se_sys_execve fs/exec.c:1959 [inline]
__x64_sys_execve+0x7c/0xa0 fs/exec.c:1959
do_syscall_64+0xf9/0x620 arch/x86/entry/common.c:293
entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&sig->cred_guard_mutex){+.+.}:
__mutex_lock_common kernel/locking/mutex.c:925 [inline]
__mutex_lock+0xd7/0x1260 kernel/locking/mutex.c:1072
lock_trace fs/proc/base.c:402 [inline]
proc_pid_stack+0x160/0x350 fs/proc/base.c:452
proc_single_show+0xeb/0x170 fs/proc/base.c:755
seq_read+0x4be/0x1160 fs/seq_file.c:229
do_loop_readv_writev fs/read_write.c:701 [inline]
do_loop_readv_writev fs/read_write.c:688 [inline]
do_iter_read+0x471/0x630 fs/read_write.c:925
vfs_readv+0xe5/0x150 fs/read_write.c:987
do_preadv fs/read_write.c:1071 [inline]
__do_sys_preadv fs/read_write.c:1121 [inline]
__se_sys_preadv fs/read_write.c:1116 [inline]
__x64_sys_preadv+0x22b/0x310 fs/read_write.c:1116
do_syscall_64+0xf9/0x620 arch/x86/entry/common.c:293
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 --> &p->lock

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&p->lock);
lock(sb_writers#3);
lock(&p->lock);
lock(&sig->cred_guard_mutex);

*** DEADLOCK ***

1 lock held by syz-executor.1/7856:
#0: 00000000d1f52354 (&p->lock){+.+.}, at: seq_read+0x6b/0x1160 fs/seq_file.c:161

stack backtrace:
CPU: 0 PID: 7856 Comm: syz-executor.1 Not tainted 4.19.129-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+0x1fc/0x2fe lib/dump_stack.c:118
print_circular_bug.constprop.0.cold+0x2d7/0x41e kernel/locking/lockdep.c:1221
check_prev_add kernel/locking/lockdep.c:1865 [inline]
check_prevs_add kernel/locking/lockdep.c:1978 [inline]
validate_chain kernel/locking/lockdep.c:2419 [inline]
__lock_acquire+0x30c9/0x3ff0 kernel/locking/lockdep.c:3415
lock_acquire+0x170/0x3c0 kernel/locking/lockdep.c:3907
__mutex_lock_common kernel/locking/mutex.c:925 [inline]
__mutex_lock+0xd7/0x1260 kernel/locking/mutex.c:1072
lock_trace fs/proc/base.c:402 [inline]
proc_pid_stack+0x160/0x350 fs/proc/base.c:452
proc_single_show+0xeb/0x170 fs/proc/base.c:755
seq_read+0x4be/0x1160 fs/seq_file.c:229
do_loop_readv_writev fs/read_write.c:701 [inline]
do_loop_readv_writev fs/read_write.c:688 [inline]
do_iter_read+0x471/0x630 fs/read_write.c:925
vfs_readv+0xe5/0x150 fs/read_write.c:987
do_preadv fs/read_write.c:1071 [inline]
__do_sys_preadv fs/read_write.c:1121 [inline]
__se_sys_preadv fs/read_write.c:1116 [inline]
__x64_sys_preadv+0x22b/0x310 fs/read_write.c:1116
do_syscall_64+0xf9/0x620 arch/x86/entry/common.c:293
entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x45cb09
Code: 0d b7 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 db b6 fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007f3085805c78 EFLAGS: 00000246 ORIG_RAX: 0000000000000127
RAX: ffffffffffffffda RBX: 00000000004fb6e0 RCX: 000000000045cb09
RDX: 00000000000002fb RSI: 00000000200017c0 RDI: 0000000000000004
RBP: 000000000078bf00 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 00000000ffffffff
R13: 0000000000000884 R14: 00000000004cb7c8 R15: 00007f30858066d4
libceph: connect [d::]:6789 error -101
libceph: mon0 [d::]:6789 connect error
libceph: connect [d::]:6789 error -101
libceph: mon0 [d::]:6789 connect error
libceph: mon1 [::6]:6789 socket closed (con state CONNECTING)
libceph: mon1 [::6]:6789 socket closed (con state CONNECTING)
libceph: connect [d::]:6789 error -101
libceph: mon0 [d::]:6789 connect error
libceph: connect [d::]:6789 error -101
libceph: mon0 [d::]:6789 connect error
libceph: connect [d::]:6789 error -101
libceph: mon0 [d::]:6789 connect error
libceph: mon1 [::6]:6789 socket closed (con state CONNECTING)


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

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

[syzbot]

Auto-closing this bug as obsolete.
Crashes did not happen for a while, no reproducer and no activity.