BUG: Dentry still in use [unmount of hugetlbfs hugetlbfs] (2)

[syzbot]

Hello,

syzbot found the following crash on:

HEAD commit: c68c5373 Add linux-next specific files for 20191107
git tree: linux-next
console output: https://syzkaller.appspot.com/x/log.txt?x=13f62294e00000
kernel config: https://syzkaller.appspot.com/x/.config?x=742545dcdea21726
dashboard link: https://syzkaller.appspot.com/bug?extid=136d2439a4e6561ea00c
compiler: gcc (GCC) 9.0.0 20181231 (experimental)
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=12ed5264e00000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=15fe2a64e00000

The bug was bisected to:

commit efe323d5cfecd2ad9dd4aa485e05312f0fa7617a
Author: Piotr Sarna <p.s...@tlen.pl>
Date: Wed Nov 6 05:06:34 2019 +0000

hugetlbfs: add O_TMPFILE support

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=14ac944ae00000
final crash: https://syzkaller.appspot.com/x/report.txt?x=16ac944ae00000
console output: https://syzkaller.appspot.com/x/log.txt?x=12ac944ae00000

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: syzbot+136d24...@syzkaller.appspotmail.com
Fixes: efe323d5cfec ("hugetlbfs: add O_TMPFILE support")

BUG: Dentry 00000000c225b886{i=685d,n=#26717} still in use (1) [unmount of
hugetlbfs hugetlbfs]
------------[ cut here ]------------
WARNING: CPU: 1 PID: 8874 at fs/dcache.c:1595 umount_check fs/dcache.c:1586
[inline]
WARNING: CPU: 1 PID: 8874 at fs/dcache.c:1595 umount_check.cold+0xe9/0x10a
fs/dcache.c:1576
Kernel panic - not syncing: panic_on_warn set ...
CPU: 1 PID: 8874 Comm: syz-executor027 Not tainted 5.4.0-rc6-next-20191107
#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
panic+0x2e3/0x75c kernel/panic.c:221
__warn.cold+0x2f/0x35 kernel/panic.c:582
report_bug+0x289/0x300 lib/bug.c:195
fixup_bug arch/x86/kernel/traps.c:174 [inline]
fixup_bug arch/x86/kernel/traps.c:169 [inline]
do_error_trap+0x11b/0x200 arch/x86/kernel/traps.c:267
do_invalid_op+0x37/0x50 arch/x86/kernel/traps.c:286
invalid_op+0x23/0x30 arch/x86/entry/entry_64.S:1027
RIP: 0010:umount_check fs/dcache.c:1595 [inline]
RIP: 0010:umount_check.cold+0xe9/0x10a fs/dcache.c:1576
Code: 89 ff e8 70 8f ef ff 48 81 c3 88 06 00 00 45 89 e8 4c 89 e1 53 4d 8b
0f 4c 89 f2 4c 89 e6 48 c7 c7 20 d9 b6 87 e8 b1 f4 9c ff <0f> 0b 58 e9 7d
1b ff ff e8 40 8f ef ff e9 29 ff ff ff 45 31 f6 e9
RSP: 0018:ffff888095c0fbc8 EFLAGS: 00010282
RAX: 0000000000000060 RBX: ffff888098736688 RCX: 0000000000000000
RDX: 0000000000000000 RSI: ffffffff815d0b86 RDI: ffffed1012b81f6b
RBP: ffff888095c0fbf8 R08: 0000000000000060 R09: ffffed1015d26161
R10: ffffed1015d26160 R11: ffff8880ae930b07 R12: ffff888090c7b300
R13: 0000000000000001 R14: 000000000000685d R15: ffffffff8910a5e0
d_walk+0x27e/0x930 fs/dcache.c:1305
do_one_tree+0x28/0x40 fs/dcache.c:1602
shrink_dcache_for_umount+0x72/0x170 fs/dcache.c:1618
generic_shutdown_super+0x6d/0x370 fs/super.c:447
kill_anon_super+0x3e/0x60 fs/super.c:1106
kill_litter_super+0x50/0x60 fs/super.c:1115
deactivate_locked_super+0x95/0x100 fs/super.c:335
deactivate_super fs/super.c:366 [inline]
deactivate_super+0x1b2/0x1d0 fs/super.c:362
cleanup_mnt+0x351/0x4c0 fs/namespace.c:1102
__cleanup_mnt+0x16/0x20 fs/namespace.c:1109
task_work_run+0x145/0x1c0 kernel/task_work.c:113
tracehook_notify_resume include/linux/tracehook.h:188 [inline]
exit_to_usermode_loop+0x316/0x380 arch/x86/entry/common.c:163
prepare_exit_to_usermode arch/x86/entry/common.c:194 [inline]
syscall_return_slowpath arch/x86/entry/common.c:274 [inline]
do_syscall_64+0x65f/0x760 arch/x86/entry/common.c:300
entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x449707
Code: 44 00 00 b8 08 00 00 00 0f 05 48 3d 01 f0 ff ff 0f 83 ed dc fb ff c3
66 2e 0f 1f 84 00 00 00 00 00 66 90 b8 a6 00 00 00 0f 05 <48> 3d 01 f0 ff
ff 0f 83 cd dc fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007ffe00511778 EFLAGS: 00000202 ORIG_RAX: 00000000000000a6
RAX: 0000000000000000 RBX: 00000000000108a6 RCX: 0000000000449707
RDX: 0000000000400df0 RSI: 0000000000000002 RDI: 00007ffe00511820
RBP: 00000000000022cb R08: 0000000000000000 R09: 0000000000000009
R10: 0000000000000005 R11: 0000000000000202 R12: 00007ffe005128d0
R13: 0000000002133880 R14: 0000000000000000 R15: 0000000000000000
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

[Piotr Sarna]

The obvious culprit seems to be the unnecessary call to d_get() when
creating a tmp file with tmpfile() call, which is then never countered
by dput()... sloppy of me.  I don't know what the procedure is now, but
I hope that my patch is simply dequeued from /next. I'll provide a fixed
and retested version once I'm back from a conference.

[Eric Biggers]

On Fri, Nov 08, 2019 at 07:05:41AM +0100, Piotr Sarna wrote:
> The obvious culprit seems to be the unnecessary call to d_get() when
> creating a tmp file with tmpfile() call, which is then never countered by
> dput()... sloppy of me.  I don't know what the procedure is now, but I hope
> that my patch is simply dequeued from /next. I'll provide a fixed and
> retested version once I'm back from a conference.
>

Fixed in next-20191115 and later, thanks Piotr. Apparently, Andrew dropped the
Reported-by syzbot line that you tried to add in v2, so let's manually
invalidate this bug report instead:

#syz invalid