WARNING in fanotify_handle_event

[syzbot]

Hello,

syzbot found the following crash on:

HEAD commit: 963172d9 Merge branch 'x86-urgent-for-linus' of git://git...
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=17c090eaa00000
kernel config: https://syzkaller.appspot.com/x/.config?x=fa9f7e1b6a8bb586
dashboard link: https://syzkaller.appspot.com/bug?extid=c277e8e2f46414645508
compiler: gcc (GCC) 9.0.0 20181231 (experimental)
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=15a32f46a00000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=13a7dc9ea00000

The bug was bisected to:

commit 77115225acc67d9ac4b15f04dd138006b9cd1ef2
Author: Amir Goldstein <amir...@gmail.com>
Date: Thu Jan 10 17:04:37 2019 +0000

fanotify: cache fsid in fsnotify_mark_connector

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=12bfcb66a00000
final crash: https://syzkaller.appspot.com/x/report.txt?x=11bfcb66a00000
console output: https://syzkaller.appspot.com/x/log.txt?x=16bfcb66a00000

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+c277e8...@syzkaller.appspotmail.com
Fixes: 77115225acc6 ("fanotify: cache fsid in fsnotify_mark_connector")

WARNING: CPU: 0 PID: 8994 at fs/notify/fanotify/fanotify.c:359
fanotify_get_fsid fs/notify/fanotify/fanotify.c:359 [inline]
WARNING: CPU: 0 PID: 8994 at fs/notify/fanotify/fanotify.c:359
fanotify_handle_event+0x5ff/0xc6d fs/notify/fanotify/fanotify.c:418
Kernel panic - not syncing: panic_on_warn set ...
CPU: 0 PID: 8994 Comm: rs:main Q:Reg Not tainted 5.2.0-rc4+ #27
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+0x172/0x1f0 lib/dump_stack.c:113
panic+0x2cb/0x744 kernel/panic.c:219
__warn.cold+0x20/0x4d kernel/panic.c:576
report_bug+0x263/0x2b0 lib/bug.c:186
fixup_bug arch/x86/kernel/traps.c:179 [inline]
fixup_bug arch/x86/kernel/traps.c:174 [inline]
do_error_trap+0x11b/0x200 arch/x86/kernel/traps.c:272
do_invalid_op+0x37/0x50 arch/x86/kernel/traps.c:291
invalid_op+0x14/0x20 arch/x86/entry/entry_64.S:986
RIP: 0010:fanotify_get_fsid fs/notify/fanotify/fanotify.c:359 [inline]
RIP: 0010:fanotify_handle_event+0x5ff/0xc6d
fs/notify/fanotify/fanotify.c:418
Code: 06 00 00 8b 5b 40 31 ff 8b b5 fc fe ff ff 09 de 89 b5 f0 fe ff ff e8
f0 64 ab ff 8b b5 f0 fe ff ff 85 f6 75 55 e8 61 63 ab ff <0f> 0b e8 5a 63
ab ff 41 83 c6 01 bf 03 00 00 00 44 89 f6 e8 c9 64
RSP: 0018:ffff8880a66dfb80 EFLAGS: 00010293
RAX: ffff888087bf0640 RBX: 0000000000000000 RCX: ffffffff81c55df0
RDX: 0000000000000000 RSI: ffffffff81c55dff RDI: 0000000000000005
RBP: ffff8880a66dfcc8 R08: ffff888087bf0640 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000002
R13: 0000000000000000 R14: 0000000000000002 R15: dffffc0000000000
send_to_group fs/notify/fsnotify.c:271 [inline]
fsnotify+0x71f/0xbc0 fs/notify/fsnotify.c:409
fsnotify_path include/linux/fsnotify.h:54 [inline]
fsnotify_path include/linux/fsnotify.h:47 [inline]
fsnotify_modify include/linux/fsnotify.h:230 [inline]
vfs_write+0x4dc/0x580 fs/read_write.c:560
ksys_write+0x14f/0x290 fs/read_write.c:611
__do_sys_write fs/read_write.c:623 [inline]
__se_sys_write fs/read_write.c:620 [inline]
__x64_sys_write+0x73/0xb0 fs/read_write.c:620
do_syscall_64+0xfd/0x680 arch/x86/entry/common.c:301
entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x7f4cdf18519d
Code: d1 20 00 00 75 10 b8 01 00 00 00 0f 05 48 3d 01 f0 ff ff 73 31 c3 48
83 ec 08 e8 be fa ff ff 48 89 04 24 b8 01 00 00 00 0f 05 <48> 8b 3c 24 48
89 c2 e8 07 fb ff ff 48 89 d0 48 83 c4 08 48 3d 01
RSP: 002b:00007f4cdd726000 EFLAGS: 00000293 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 0000000000000155 RCX: 00007f4cdf18519d
RDX: 0000000000000155 RSI: 00000000008bda90 RDI: 0000000000000001
RBP: 00000000008bda90 R08: 6573753a725f7463 R09: 745f656d6f685f72
R10: 0000000000000000 R11: 0000000000000293 R12: 0000000000000000
R13: 00007f4cdd726480 R14: 0000000000000001 R15: 00000000008bd890
Kernel Offset: disabled
Rebooting in 86400 seconds..


---
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.
For information about bisection process see: https://goo.gl/tpsmEJ#bisection
syzbot can test patches for this bug, for details see:
https://goo.gl/tpsmEJ#testing-patches

[Amir Goldstein]

Oops, we forgot to update conn->fsid when the first mark added
for inode has no fsid (e.g. inotify) and the second mark has fid,
which is more or less the only thing the repro does.
And if we are going to update conn->fsid, we do no have the
cmpxchg to guaranty setting fsid atomically.

I am thinking a set-once flag on connector FSNOTIFY_CONN_HAS_FSID
checked before smp_rmb() in fanotify_get_fsid().
If the flag is not set then call vfs_get_fsid() instead of using fsid cache.
conn->fsid can be updated in fsnotify_add_mark_list() under conn->lock,
and flag set after smp_wmb().

Does that sound correct?

Thanks,
Amir.

[Amir Goldstein]

Actually, we don't need to call vfs_get_fsid() in race we just drop the event.

> conn->fsid can be updated in fsnotify_add_mark_list() under conn->lock,
> and flag set after smp_wmb().
>
> Does that sound correct?
>

Something like this:

#syz test: https://github.com/amir73il/linux.git fsnotify-fix-fsid-cache

It passed my modified ltp test:
https://github.com/amir73il/ltp/commits/fanotify_dirent

Thanks,
Amir.

[syzbot]

Hello,

syzbot has tested the proposed patch and the reproducer did not trigger
crash:

Reported-and-tested-by:
syzbot+c277e8...@syzkaller.appspotmail.com

Tested on:

commit: a6a3fd5c fanotify: update connector fsid cache on add mark
git tree: https://github.com/amir73il/linux.git
fsnotify-fix-fsid-cache
kernel config: https://syzkaller.appspot.com/x/.config?x=fa9f7e1b6a8bb586

compiler: gcc (GCC) 9.0.0 20181231 (experimental)

Note: testing is done by a robot and is best-effort only.