[syzbot] [overlayfs?] possible deadlock in seq_read_iter (3)

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: 841c35169323 Linux 6.8-rc4
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1215e120180000
kernel config: https://syzkaller.appspot.com/x/.config?x=fbd950b5071b7ea3
dashboard link: https://syzkaller.appspot.com/bug?extid=af9aa785e14a67796a87
compiler: Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40

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

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/ef821afd15d3/disk-841c3516.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/e41102f18e6e/vmlinux-841c3516.xz
kernel image: https://storage.googleapis.com/syzbot-assets/a17352b259d8/bzImage-841c3516.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+af9aa7...@syzkaller.appspotmail.com

F2FS-fs (loop1): Found nat_bits in checkpoint
F2FS-fs (loop1): Mounted with checkpoint version = 48b305e5
======================================================
WARNING: possible circular locking dependency detected
6.8.0-rc4-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor.1/640 is trying to acquire lock:
ffff888072750310 (&p->lock){+.+.}-{3:3}, at: seq_read_iter+0xb7/0xd60 fs/seq_file.c:182

but task is already holding lock:
ffff888023f61c68 (&pipe->mutex/1){+.+.}-{3:3}, at: splice_file_to_pipe+0x2e/0x500 fs/splice.c:1292

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #4 (&pipe->mutex/1){+.+.}-{3:3}:
lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
__mutex_lock_common kernel/locking/mutex.c:608 [inline]
__mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
iter_file_splice_write+0x335/0x14e0 fs/splice.c:687
do_splice_from fs/splice.c:941 [inline]
do_splice+0xd77/0x1880 fs/splice.c:1354
__do_splice fs/splice.c:1436 [inline]
__do_sys_splice fs/splice.c:1652 [inline]
__se_sys_splice+0x331/0x4a0 fs/splice.c:1634
do_syscall_64+0xf9/0x240
entry_SYSCALL_64_after_hwframe+0x6f/0x77

-> #3 (sb_writers#4){.+.+}-{0:0}:
lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
percpu_down_read include/linux/percpu-rwsem.h:51 [inline]
__sb_start_write include/linux/fs.h:1639 [inline]
sb_start_write+0x4d/0x1c0 include/linux/fs.h:1775
mnt_want_write+0x3f/0x90 fs/namespace.c:409
ovl_fix_origin fs/overlayfs/namei.c:908 [inline]
ovl_lookup+0x1394/0x2a60 fs/overlayfs/namei.c:1143
__lookup_slow+0x28c/0x3f0 fs/namei.c:1693
lookup_slow+0x53/0x70 fs/namei.c:1710
walk_component fs/namei.c:2001 [inline]
link_path_walk+0x9cd/0xe80 fs/namei.c:2328
path_lookupat+0xa9/0x450 fs/namei.c:2481
filename_lookup+0x255/0x610 fs/namei.c:2511
user_path_at_empty+0x42/0x60 fs/namei.c:2920
user_path_at include/linux/namei.h:57 [inline]
__do_sys_chdir fs/open.c:556 [inline]
__se_sys_chdir+0xbf/0x220 fs/open.c:550
do_syscall_64+0xf9/0x240
entry_SYSCALL_64_after_hwframe+0x6f/0x77

-> #2 (&ovl_i_mutex_dir_key[depth]){++++}-{3:3}:
lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
down_read+0xb1/0xa40 kernel/locking/rwsem.c:1526
inode_lock_shared include/linux/fs.h:812 [inline]
lookup_slow+0x45/0x70 fs/namei.c:1709
walk_component+0x2e1/0x410 fs/namei.c:2001
lookup_last fs/namei.c:2458 [inline]
path_lookupat+0x16f/0x450 fs/namei.c:2482
filename_lookup+0x255/0x610 fs/namei.c:2511
kern_path+0x35/0x50 fs/namei.c:2619
lookup_bdev+0xc5/0x290 block/bdev.c:1014
resume_store+0x1a0/0x710 kernel/power/hibernate.c:1183
kernfs_fop_write_iter+0x3a4/0x500 fs/kernfs/file.c:334
call_write_iter include/linux/fs.h:2085 [inline]
new_sync_write fs/read_write.c:497 [inline]
vfs_write+0xa81/0xcb0 fs/read_write.c:590
ksys_write+0x1a0/0x2c0 fs/read_write.c:643
do_syscall_64+0xf9/0x240
entry_SYSCALL_64_after_hwframe+0x6f/0x77

-> #1 (&of->mutex){+.+.}-{3:3}:
lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
__mutex_lock_common kernel/locking/mutex.c:608 [inline]
__mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
kernfs_seq_start+0x53/0x3b0 fs/kernfs/file.c:154
seq_read_iter+0x3d0/0xd60 fs/seq_file.c:225
call_read_iter include/linux/fs.h:2079 [inline]
new_sync_read fs/read_write.c:395 [inline]
vfs_read+0x978/0xb70 fs/read_write.c:476
ksys_read+0x1a0/0x2c0 fs/read_write.c:619
do_syscall_64+0xf9/0x240
entry_SYSCALL_64_after_hwframe+0x6f/0x77

-> #0 (&p->lock){+.+.}-{3:3}:
check_prev_add kernel/locking/lockdep.c:3134 [inline]
check_prevs_add kernel/locking/lockdep.c:3253 [inline]
validate_chain+0x18ca/0x58e0 kernel/locking/lockdep.c:3869
__lock_acquire+0x1345/0x1fd0 kernel/locking/lockdep.c:5137
lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
__mutex_lock_common kernel/locking/mutex.c:608 [inline]
__mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
seq_read_iter+0xb7/0xd60 fs/seq_file.c:182
proc_reg_read_iter+0x1c3/0x290 fs/proc/inode.c:302
call_read_iter include/linux/fs.h:2079 [inline]
copy_splice_read+0x661/0xb60 fs/splice.c:365
do_splice_read fs/splice.c:985 [inline]
splice_file_to_pipe+0x299/0x500 fs/splice.c:1295
do_sendfile+0x515/0xdc0 fs/read_write.c:1301
__do_sys_sendfile64 fs/read_write.c:1356 [inline]
__se_sys_sendfile64+0x100/0x1e0 fs/read_write.c:1348
do_syscall_64+0xf9/0x240
entry_SYSCALL_64_after_hwframe+0x6f/0x77

other info that might help us debug this:

Chain exists of:
&p->lock --> sb_writers#4 --> &pipe->mutex/1

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&pipe->mutex/1);
lock(sb_writers#4);
lock(&pipe->mutex/1);
lock(&p->lock);

*** DEADLOCK ***

1 lock held by syz-executor.1/640:
#0: ffff888023f61c68 (&pipe->mutex/1){+.+.}-{3:3}, at: splice_file_to_pipe+0x2e/0x500 fs/splice.c:1292

stack backtrace:
CPU: 1 PID: 640 Comm: syz-executor.1 Not tainted 6.8.0-rc4-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/25/2024
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0x1e7/0x2e0 lib/dump_stack.c:106
check_noncircular+0x36a/0x4a0 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+0x18ca/0x58e0 kernel/locking/lockdep.c:3869
__lock_acquire+0x1345/0x1fd0 kernel/locking/lockdep.c:5137
lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
__mutex_lock_common kernel/locking/mutex.c:608 [inline]
__mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
seq_read_iter+0xb7/0xd60 fs/seq_file.c:182
proc_reg_read_iter+0x1c3/0x290 fs/proc/inode.c:302
call_read_iter include/linux/fs.h:2079 [inline]
copy_splice_read+0x661/0xb60 fs/splice.c:365
do_splice_read fs/splice.c:985 [inline]
splice_file_to_pipe+0x299/0x500 fs/splice.c:1295
do_sendfile+0x515/0xdc0 fs/read_write.c:1301
__do_sys_sendfile64 fs/read_write.c:1356 [inline]
__se_sys_sendfile64+0x100/0x1e0 fs/read_write.c:1348
do_syscall_64+0xf9/0x240
entry_SYSCALL_64_after_hwframe+0x6f/0x77
RIP: 0033:0x7f3acde7dda9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 e1 20 00 00 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 73 01 c3 48 c7 c1 b0 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f3acec5e0c8 EFLAGS: 00000246 ORIG_RAX: 0000000000000028
RAX: ffffffffffffffda RBX: 00007f3acdfac050 RCX: 00007f3acde7dda9
RDX: 0000000020000000 RSI: 0000000000000004 RDI: 0000000000000000
RBP: 00007f3acdeca47a R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000007 R11: 0000000000000246 R12: 0000000000000000
R13: 000000000000000b R14: 00007f3acdfac050 R15: 00007fff9463d788
</TASK>


---
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.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

[Miklos Szeredi]

[Cc: Rafael]

plain fs sb_writers --> pipe->mutex

This is just a plain splice from a pipe to a regular file (which can
be on the upper layer of overlayfs).

> -> #3 (sb_writers#4){.+.+}-{0:0}:
> lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
> percpu_down_read include/linux/percpu-rwsem.h:51 [inline]
> __sb_start_write include/linux/fs.h:1639 [inline]
> sb_start_write+0x4d/0x1c0 include/linux/fs.h:1775
> mnt_want_write+0x3f/0x90 fs/namespace.c:409
> ovl_fix_origin fs/overlayfs/namei.c:908 [inline]
> ovl_lookup+0x1394/0x2a60 fs/overlayfs/namei.c:1143
> __lookup_slow+0x28c/0x3f0 fs/namei.c:1693
> lookup_slow+0x53/0x70 fs/namei.c:1710
> walk_component fs/namei.c:2001 [inline]
> link_path_walk+0x9cd/0xe80 fs/namei.c:2328
> path_lookupat+0xa9/0x450 fs/namei.c:2481
> filename_lookup+0x255/0x610 fs/namei.c:2511
> user_path_at_empty+0x42/0x60 fs/namei.c:2920
> user_path_at include/linux/namei.h:57 [inline]
> __do_sys_chdir fs/open.c:556 [inline]
> __se_sys_chdir+0xbf/0x220 fs/open.c:550
> do_syscall_64+0xf9/0x240
> entry_SYSCALL_64_after_hwframe+0x6f/0x77

overlayfs directory i_mutex --> plain fs sb_writers

This is perfectly normal lock ordering for overlayfs: lock the
overlayfs inode, then call mnt_want_write() on the upper filesystem.

> -> #2 (&ovl_i_mutex_dir_key[depth]){++++}-{3:3}:
> lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
> down_read+0xb1/0xa40 kernel/locking/rwsem.c:1526
> inode_lock_shared include/linux/fs.h:812 [inline]

> lookup_slow+0x45/0x70 fs/namei.c:17091

> walk_component+0x2e1/0x410 fs/namei.c:2001
> lookup_last fs/namei.c:2458 [inline]
> path_lookupat+0x16f/0x450 fs/namei.c:2482
> filename_lookup+0x255/0x610 fs/namei.c:2511
> kern_path+0x35/0x50 fs/namei.c:2619
> lookup_bdev+0xc5/0x290 block/bdev.c:1014
> resume_store+0x1a0/0x710 kernel/power/hibernate.c:1183
> kernfs_fop_write_iter+0x3a4/0x500 fs/kernfs/file.c:334
> call_write_iter include/linux/fs.h:2085 [inline]
> new_sync_write fs/read_write.c:497 [inline]
> vfs_write+0xa81/0xcb0 fs/read_write.c:590
> ksys_write+0x1a0/0x2c0 fs/read_write.c:643
> do_syscall_64+0xf9/0x240
> entry_SYSCALL_64_after_hwframe+0x6f/0x77

kernfs of->mutex --> overlayfs directory i_mutex

The device name is written to /sys/power/resume. When performing the
lookup of the name, the inode lock is taken on the directory, which
happens to be on overlayfs.

> -> #1 (&of->mutex){+.+.}-{3:3}:
> lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
> __mutex_lock_common kernel/locking/mutex.c:608 [inline]
> __mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
> kernfs_seq_start+0x53/0x3b0 fs/kernfs/file.c:154
> seq_read_iter+0x3d0/0xd60 fs/seq_file.c:225
> call_read_iter include/linux/fs.h:2079 [inline]
> new_sync_read fs/read_write.c:395 [inline]
> vfs_read+0x978/0xb70 fs/read_write.c:476
> ksys_read+0x1a0/0x2c0 fs/read_write.c:619
> do_syscall_64+0xf9/0x240
> entry_SYSCALL_64_after_hwframe+0x6f/0x77

seqfile p->lock --> kernfs of->mutex

Reading an attribute. It could be "/sys/power/resume", but here we
don't know, it's just the same lock class.

> -> #0 (&p->lock){+.+.}-{3:3}:
> check_prev_add kernel/locking/lockdep.c:3134 [inline]
> check_prevs_add kernel/locking/lockdep.c:3253 [inline]
> validate_chain+0x18ca/0x58e0 kernel/locking/lockdep.c:3869
> __lock_acquire+0x1345/0x1fd0 kernel/locking/lockdep.c:5137
> lock_acquire+0x1e3/0x530 kernel/locking/lockdep.c:5754
> __mutex_lock_common kernel/locking/mutex.c:608 [inline]
> __mutex_lock+0x136/0xd70 kernel/locking/mutex.c:752
> seq_read_iter+0xb7/0xd60 fs/seq_file.c:182
> proc_reg_read_iter+0x1c3/0x290 fs/proc/inode.c:302
> call_read_iter include/linux/fs.h:2079 [inline]
> copy_splice_read+0x661/0xb60 fs/splice.c:365
> do_splice_read fs/splice.c:985 [inline]
> splice_file_to_pipe+0x299/0x500 fs/splice.c:1295
> do_sendfile+0x515/0xdc0 fs/read_write.c:1301
> __do_sys_sendfile64 fs/read_write.c:1356 [inline]
> __se_sys_sendfile64+0x100/0x1e0 fs/read_write.c:1348
> do_syscall_64+0xf9/0x240
> entry_SYSCALL_64_after_hwframe+0x6f/0x77

pipe->mutex --> seqfile p->lock

This is a sendfile from a seqfile. Could have been
"/sys/power/resume" as well, AFAICS.

I don't really know if/how this needs fixing, but that path lookup in
resume_store() looks a bit nasty.

Thanks,
Miklos