[syzbot ci] Re: vfs: add O_CREAT|O_DIRECTORY to open*(2)

[syzbot ci]

syzbot ci has tested the following series

[v3] vfs: add O_CREAT|O_DIRECTORY to open*(2)
https://lore.kernel.org/all/20260517170244.18...@xs4all.nl
* [RFC PATCH v3 1/2] vfs: add O_CREAT|O_DIRECTORY to open*(2)
* [RFC PATCH v3 2/2] selftest: add tests for open*(O_CREAT|O_DIRECTORY)

and found the following issues:
* WARNING: lock held when returning to user space in filename_create
* WARNING: lock held when returning to user space in start_creating
* possible deadlock in mnt_want_write

Full report is available here:
https://ci.syzbot.org/series/6c2681e8-f8f3-4287-8f97-bd6ea26a767f

***

WARNING: lock held when returning to user space in filename_create

tree: torvalds
URL: https://kernel.googlesource.com/pub/scm/linux/kernel/git/torvalds/linux
base: 6916d5703ddf9a38f1f6c2cc793381a24ee914c6
arch: amd64
compiler: Debian clang version 21.1.8 (++20251221033036+2078da43e25a-1~exp1~20251221153213.50), Debian LLD 21.1.8
config: https://ci.syzbot.org/builds/7a0474b8-fd0f-4804-8833-2232742e06e3/config
syz repro: https://ci.syzbot.org/findings/512f2832-177b-4b65-b835-bdff3e4402bc/syz_repro

hpfs: hpfs_map_4sectors(): unaligned read
hpfs: hpfs_map_4sectors(): unaligned read
hpfs: filesystem error: unable to find root dir
================================================
WARNING: lock held when returning to user space!
syzkaller #0 Not tainted
------------------------------------------------
syz.0.17/5814 is leaving the kernel with locks still held!
1 lock held by syz.0.17/5814:
#0: ffff8881bab5b878 (&type->i_mutex_dir_key#8/1){+.+.}-{4:4}, at: inode_lock_nested include/linux/fs.h:1074 [inline]
#0: ffff8881bab5b878 (&type->i_mutex_dir_key#8/1){+.+.}-{4:4}, at: __start_dirop fs/namei.c:2919 [inline]
#0: ffff8881bab5b878 (&type->i_mutex_dir_key#8/1){+.+.}-{4:4}, at: start_dirop fs/namei.c:2943 [inline]
#0: ffff8881bab5b878 (&type->i_mutex_dir_key#8/1){+.+.}-{4:4}, at: filename_create+0x200/0x370 fs/namei.c:4984


***

WARNING: lock held when returning to user space in start_creating

tree: torvalds
URL: https://kernel.googlesource.com/pub/scm/linux/kernel/git/torvalds/linux
base: 6916d5703ddf9a38f1f6c2cc793381a24ee914c6
arch: amd64
compiler: Debian clang version 21.1.8 (++20251221033036+2078da43e25a-1~exp1~20251221153213.50), Debian LLD 21.1.8
config: https://ci.syzbot.org/builds/7a0474b8-fd0f-4804-8833-2232742e06e3/config
syz repro: https://ci.syzbot.org/findings/7bdf5f73-8cce-4785-aa9e-dd89622cf613/syz_repro

overlayfs: failed to create directory ./bus/work (errno: 126); mounting read-only
================================================
WARNING: lock held when returning to user space!
syzkaller #0 Not tainted
------------------------------------------------
syz.1.18/5833 is leaving the kernel with locks still held!
1 lock held by syz.1.18/5833:
#0: ffff8881ba0c4518 (&type->i_mutex_dir_key#3/1){+.+.}-{4:4}, at: inode_lock_nested include/linux/fs.h:1074 [inline]
#0: ffff8881ba0c4518 (&type->i_mutex_dir_key#3/1){+.+.}-{4:4}, at: __start_dirop fs/namei.c:2919 [inline]
#0: ffff8881ba0c4518 (&type->i_mutex_dir_key#3/1){+.+.}-{4:4}, at: start_dirop fs/namei.c:2943 [inline]
#0: ffff8881ba0c4518 (&type->i_mutex_dir_key#3/1){+.+.}-{4:4}, at: start_creating+0xbe/0x100 fs/namei.c:3412


***

possible deadlock in mnt_want_write

tree: torvalds
URL: https://kernel.googlesource.com/pub/scm/linux/kernel/git/torvalds/linux
base: 6916d5703ddf9a38f1f6c2cc793381a24ee914c6
arch: amd64
compiler: Debian clang version 21.1.8 (++20251221033036+2078da43e25a-1~exp1~20251221153213.50), Debian LLD 21.1.8
config: https://ci.syzbot.org/builds/7a0474b8-fd0f-4804-8833-2232742e06e3/config
syz repro: https://ci.syzbot.org/findings/6d4f386f-fb66-4862-b644-4ac19c79f6a3/syz_repro

======================================================
WARNING: possible circular locking dependency detected
syzkaller #0 Not tainted
------------------------------------------------------
syz.0.17/5836 is trying to acquire lock:
ffff888118cba410
(sb_writers#12){.+.+}-{0:0}, at: mnt_want_write+0x41/0x90 fs/namespace.c:493

but task is already holding lock:
ffff88811f1627e0 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}, at: inode_lock_nested include/linux/fs.h:1074 [inline]
ffff88811f1627e0 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}, at: __start_dirop fs/namei.c:2919 [inline]
ffff88811f1627e0 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}, at: start_dirop fs/namei.c:2943 [inline]
ffff88811f1627e0 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}, at: filename_create+0x200/0x370 fs/namei.c:4984

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}:
down_write_nested+0x9d/0x210 kernel/locking/rwsem.c:1751
inode_lock_nested include/linux/fs.h:1074 [inline]
__start_dirop fs/namei.c:2919 [inline]
start_dirop fs/namei.c:2943 [inline]
filename_unlinkat+0x2a7/0x610 fs/namei.c:5599
__do_sys_unlink fs/namei.c:5653 [inline]
__se_sys_unlink+0x2e/0x140 fs/namei.c:5650
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0x15f/0xf80 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

-> #0 (sb_writers#12){.+.+}-{0:0}:
check_prev_add kernel/locking/lockdep.c:3165 [inline]
check_prevs_add kernel/locking/lockdep.c:3284 [inline]
validate_chain kernel/locking/lockdep.c:3908 [inline]
__lock_acquire+0x15a5/0x2cf0 kernel/locking/lockdep.c:5237
lock_acquire+0x106/0x350 kernel/locking/lockdep.c:5868
percpu_down_read_internal include/linux/percpu-rwsem.h:53 [inline]
percpu_down_read_freezable include/linux/percpu-rwsem.h:83 [inline]
__sb_start_write include/linux/fs/super.h:19 [inline]
sb_start_write+0x4d/0x1c0 include/linux/fs/super.h:125
mnt_want_write+0x41/0x90 fs/namespace.c:493
filename_create+0x154/0x370 fs/namei.c:4977
filename_mkdirat+0xd2/0x510 fs/namei.c:5337
__do_sys_mkdirat fs/namei.c:5365 [inline]
__se_sys_mkdirat+0x35/0x150 fs/namei.c:5362
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0x15f/0xf80 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

other info that might help us debug this:

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&ovl_i_mutex_dir_key[depth]/1);
lock(sb_writers#12);
lock(&ovl_i_mutex_dir_key[depth]/1);
rlock(sb_writers#12);

*** DEADLOCK ***

1 lock held by syz.0.17/5836:
#0: ffff88811f1627e0 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}, at: inode_lock_nested include/linux/fs.h:1074 [inline]
#0: ffff88811f1627e0 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}, at: __start_dirop fs/namei.c:2919 [inline]
#0: ffff88811f1627e0 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}, at: start_dirop fs/namei.c:2943 [inline]
#0: ffff88811f1627e0 (&ovl_i_mutex_dir_key[depth]/1){+.+.}-{4:4}, at: filename_create+0x200/0x370 fs/namei.c:4984

stack backtrace:
CPU: 0 UID: 0 PID: 5836 Comm: syz.0.17 Not tainted syzkaller #0 PREEMPT(full)
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.2-debian-1.16.2-1 04/01/2014
Call Trace:
<TASK>
dump_stack_lvl+0xe8/0x150 lib/dump_stack.c:120
print_circular_bug+0x2e1/0x300 kernel/locking/lockdep.c:2043
check_noncircular+0x12e/0x150 kernel/locking/lockdep.c:2175
check_prev_add kernel/locking/lockdep.c:3165 [inline]
check_prevs_add kernel/locking/lockdep.c:3284 [inline]
validate_chain kernel/locking/lockdep.c:3908 [inline]
__lock_acquire+0x15a5/0x2cf0 kernel/locking/lockdep.c:5237
lock_acquire+0x106/0x350 kernel/locking/lockdep.c:5868
percpu_down_read_internal include/linux/percpu-rwsem.h:53 [inline]
percpu_down_read_freezable include/linux/percpu-rwsem.h:83 [inline]
__sb_start_write include/linux/fs/super.h:19 [inline]
sb_start_write+0x4d/0x1c0 include/linux/fs/super.h:125
mnt_want_write+0x41/0x90 fs/namespace.c:493
filename_create+0x154/0x370 fs/namei.c:4977
filename_mkdirat+0xd2/0x510 fs/namei.c:5337
__do_sys_mkdirat fs/namei.c:5365 [inline]
__se_sys_mkdirat+0x35/0x150 fs/namei.c:5362
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0x15f/0xf80 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7f2f5e99ce59
Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 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 e8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f2f5f771028 EFLAGS: 00000246 ORIG_RAX: 0000000000000102
RAX: ffffffffffffffda RBX: 00007f2f5ec15fa0 RCX: 00007f2f5e99ce59
RDX: 0000000000000010 RSI: 0000200000002040 RDI: ffffffffffffff9c
RBP: 00007f2f5ea32d6f R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007f2f5ec16038 R14: 00007f2f5ec15fa0 R15: 00007ffe96c961d8
</TASK>


***

If these findings have caused you to resend the series or submit a
separate fix, please add the following tag to your commit message:
Tested-by: syz...@syzkaller.appspotmail.com

---
This report is generated by a bot. It may contain errors.
syzbot ci engineers can be reached at syzk...@googlegroups.com.

To test a patch for this bug, please reply with `#syz test`
(should be on a separate line).

The patch should be attached to the email.
Note: arguments like custom git repos and branches are not supported.