[syzbot] [ext4?] possible deadlock in ext4_xattr_inode_iget (3)

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: fe46a7dd189e Merge tag 'sound-6.9-rc1' of git://git.kernel..
git tree: upstream
console+strace: https://syzkaller.appspot.com/x/log.txt?x=11a1e52d180000
kernel config: https://syzkaller.appspot.com/x/.config?x=1a07d5da4eb21586
dashboard link: https://syzkaller.appspot.com/bug?extid=ee72b9a7aad1e5a77c5c
compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=12407f45180000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=140d9db1180000

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/b42ab0fd4947/disk-fe46a7dd.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/b8a6e7231930/vmlinux-fe46a7dd.xz
kernel image: https://storage.googleapis.com/syzbot-assets/4fbf3e4ce6f8/bzImage-fe46a7dd.xz
mounted in repro: https://storage.googleapis.com/syzbot-assets/5d293cee060a/mount_0.gz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+ee72b9...@syzkaller.appspotmail.com

======================================================
WARNING: possible circular locking dependency detected
6.8.0-syzkaller-08951-gfe46a7dd189e #0 Not tainted
------------------------------------------------------
syz-executor545/5275 is trying to acquire lock:
ffff888077730400 (&ea_inode->i_rwsem#8/1){+.+.}-{3:3}, at: inode_lock include/linux/fs.h:793 [inline]
ffff888077730400 (&ea_inode->i_rwsem#8/1){+.+.}-{3:3}, at: ext4_xattr_inode_iget+0x173/0x440 fs/ext4/xattr.c:461

but task is already holding lock:
ffff888077730c88 (&ei->i_data_sem/3){++++}-{3:3}, at: ext4_setattr+0x1ba0/0x29d0 fs/ext4/inode.c:5417

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (&ei->i_data_sem/3){++++}-{3:3}:
down_write+0x3a/0x50 kernel/locking/rwsem.c:1579
ext4_update_i_disksize fs/ext4/ext4.h:3383 [inline]
ext4_xattr_inode_write fs/ext4/xattr.c:1446 [inline]
ext4_xattr_inode_lookup_create fs/ext4/xattr.c:1594 [inline]
ext4_xattr_set_entry+0x3a14/0x3cf0 fs/ext4/xattr.c:1719
ext4_xattr_ibody_set+0x126/0x380 fs/ext4/xattr.c:2287
ext4_xattr_set_handle+0x98d/0x1480 fs/ext4/xattr.c:2444
ext4_xattr_set+0x149/0x380 fs/ext4/xattr.c:2558
__vfs_setxattr+0x176/0x1e0 fs/xattr.c:200
__vfs_setxattr_noperm+0x127/0x5e0 fs/xattr.c:234
__vfs_setxattr_locked+0x182/0x260 fs/xattr.c:295
vfs_setxattr+0x146/0x350 fs/xattr.c:321
do_setxattr+0x146/0x170 fs/xattr.c:629
setxattr+0x15d/0x180 fs/xattr.c:652
path_setxattr+0x179/0x1e0 fs/xattr.c:671
__do_sys_lsetxattr fs/xattr.c:694 [inline]
__se_sys_lsetxattr fs/xattr.c:690 [inline]
__x64_sys_lsetxattr+0xc1/0x160 fs/xattr.c:690
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xd5/0x260 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x6d/0x75

-> #0 (&ea_inode->i_rwsem#8/1){+.+.}-{3:3}:
check_prev_add kernel/locking/lockdep.c:3134 [inline]
check_prevs_add kernel/locking/lockdep.c:3253 [inline]
validate_chain kernel/locking/lockdep.c:3869 [inline]
__lock_acquire+0x2478/0x3b30 kernel/locking/lockdep.c:5137
lock_acquire kernel/locking/lockdep.c:5754 [inline]
lock_acquire+0x1b1/0x540 kernel/locking/lockdep.c:5719
down_write+0x3a/0x50 kernel/locking/rwsem.c:1579
inode_lock include/linux/fs.h:793 [inline]
ext4_xattr_inode_iget+0x173/0x440 fs/ext4/xattr.c:461
ext4_xattr_inode_get+0x16c/0x870 fs/ext4/xattr.c:535
ext4_xattr_move_to_block fs/ext4/xattr.c:2640 [inline]
ext4_xattr_make_inode_space fs/ext4/xattr.c:2742 [inline]
ext4_expand_extra_isize_ea+0x1367/0x1ae0 fs/ext4/xattr.c:2834
__ext4_expand_extra_isize+0x346/0x480 fs/ext4/inode.c:5789
ext4_try_to_expand_extra_isize fs/ext4/inode.c:5832 [inline]
__ext4_mark_inode_dirty+0x55a/0x860 fs/ext4/inode.c:5910
ext4_setattr+0x1c14/0x29d0 fs/ext4/inode.c:5420
notify_change+0x745/0x11c0 fs/attr.c:497
do_truncate+0x15c/0x220 fs/open.c:65
handle_truncate fs/namei.c:3300 [inline]
do_open fs/namei.c:3646 [inline]
path_openat+0x24b9/0x2990 fs/namei.c:3799
do_filp_open+0x1dc/0x430 fs/namei.c:3826
do_sys_openat2+0x17a/0x1e0 fs/open.c:1406
do_sys_open fs/open.c:1421 [inline]
__do_sys_openat fs/open.c:1437 [inline]
__se_sys_openat fs/open.c:1432 [inline]
__x64_sys_openat+0x175/0x210 fs/open.c:1432
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xd5/0x260 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x6d/0x75

other info that might help us debug this:

Possible unsafe locking scenario:

CPU0 CPU1
---- ----
lock(&ei->i_data_sem/3);
lock(&ea_inode->i_rwsem#8/1);
lock(&ei->i_data_sem/3);
lock(&ea_inode->i_rwsem#8/1);

*** DEADLOCK ***

5 locks held by syz-executor545/5275:
#0: ffff888022da6420 (sb_writers#4){.+.+}-{0:0}, at: do_open fs/namei.c:3635 [inline]
#0: ffff888022da6420 (sb_writers#4){.+.+}-{0:0}, at: path_openat+0x1fba/0x2990 fs/namei.c:3799
#1: ffff888077730e00 (&sb->s_type->i_mutex_key#8){++++}-{3:3}, at: inode_lock include/linux/fs.h:793 [inline]
#1: ffff888077730e00 (&sb->s_type->i_mutex_key#8){++++}-{3:3}, at: do_truncate+0x14b/0x220 fs/open.c:63
#2: ffff888077730fa0 (mapping.invalidate_lock){++++}-{3:3}, at: filemap_invalidate_lock include/linux/fs.h:838 [inline]
#2: ffff888077730fa0 (mapping.invalidate_lock){++++}-{3:3}, at: ext4_setattr+0xdfd/0x29d0 fs/ext4/inode.c:5378
#3: ffff888077730c88 (&ei->i_data_sem/3){++++}-{3:3}, at: ext4_setattr+0x1ba0/0x29d0 fs/ext4/inode.c:5417
#4: ffff888077730ac8 (&ei->xattr_sem){++++}-{3:3}, at: ext4_write_trylock_xattr fs/ext4/xattr.h:162 [inline]
#4: ffff888077730ac8 (&ei->xattr_sem){++++}-{3:3}, at: ext4_try_to_expand_extra_isize fs/ext4/inode.c:5829 [inline]
#4: ffff888077730ac8 (&ei->xattr_sem){++++}-{3:3}, at: __ext4_mark_inode_dirty+0x4cf/0x860 fs/ext4/inode.c:5910

stack backtrace:
CPU: 1 PID: 5275 Comm: syz-executor545 Not tainted 6.8.0-syzkaller-08951-gfe46a7dd189e #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 03/27/2024
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:114
check_noncircular+0x31a/0x400 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 kernel/locking/lockdep.c:3869 [inline]
__lock_acquire+0x2478/0x3b30 kernel/locking/lockdep.c:5137
lock_acquire kernel/locking/lockdep.c:5754 [inline]
lock_acquire+0x1b1/0x540 kernel/locking/lockdep.c:5719
down_write+0x3a/0x50 kernel/locking/rwsem.c:1579
inode_lock include/linux/fs.h:793 [inline]
ext4_xattr_inode_iget+0x173/0x440 fs/ext4/xattr.c:461
ext4_xattr_inode_get+0x16c/0x870 fs/ext4/xattr.c:535
ext4_xattr_move_to_block fs/ext4/xattr.c:2640 [inline]
ext4_xattr_make_inode_space fs/ext4/xattr.c:2742 [inline]
ext4_expand_extra_isize_ea+0x1367/0x1ae0 fs/ext4/xattr.c:2834
__ext4_expand_extra_isize+0x346/0x480 fs/ext4/inode.c:5789
ext4_try_to_expand_extra_isize fs/ext4/inode.c:5832 [inline]
__ext4_mark_inode_dirty+0x55a/0x860 fs/ext4/inode.c:5910
ext4_setattr+0x1c14/0x29d0 fs/ext4/inode.c:5420
notify_change+0x745/0x11c0 fs/attr.c:497
do_truncate+0x15c/0x220 fs/open.c:65
handle_truncate fs/namei.c:3300 [inline]
do_open fs/namei.c:3646 [inline]
path_openat+0x24b9/0x2990 fs/namei.c:3799
do_filp_open+0x1dc/0x430 fs/namei.c:3826
do_sys_openat2+0x17a/0x1e0 fs/open.c:1406
do_sys_open fs/open.c:1421 [inline]
__do_sys_openat fs/open.c:1437 [inline]
__se_sys_openat fs/open.c:1432 [inline]
__x64_sys_openat+0x175/0x210 fs/open.c:1432
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xd5/0x260 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x6d/0x75
RIP: 0033:0x7fc7c030b2e9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 21 18 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 b8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007ffc3c4a0608 EFLAGS: 00000246 ORIG_RAX: 0000000000000101
RAX: ffffffffffffffda RBX: 0031656c69662f2e RCX: 00007fc7c030b2e9
RDX: 0000000000143362 RSI: 00000000200000c0 RDI: 00000000ffffff9c
RBP: 6c6c616c65646f6e R08: 00007ffc3c4a0640 R09: 00007ffc3c4a0640
R10: 000000000a000000 R11: 0000000000000246 R12: 00007ffc3c4a062c
R13: 0000000000000040 R14: 431bde82d7b634db R15: 00007ffc3c4a0660
</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 syzbot to run the reproducer, reply with:
#syz test: git://repo/address.git branch-or-commit-hash
If you attach or paste a git patch, syzbot will apply it before testing.

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

[Hillf Danton]

On Wed, 03 Apr 2024 00:45:29 -0700

> syzbot found the following issue on:
>
> HEAD commit: fe46a7dd189e Merge tag 'sound-6.9-rc1' of git://git.kernel..
> git tree: upstream

> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=140d9db1180000

#syz test https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git fe46a7dd189e

--- x/fs/ext4/inode.c
+++ y/fs/ext4/inode.c
@@ -5414,6 +5414,7 @@ int ext4_setattr(struct mnt_idmap *idmap
(attr->ia_size > 0 ? attr->ia_size - 1 : 0) >>
inode->i_sb->s_blocksize_bits);

+ down_read(&EXT4_I(inode)->xattr_sem);
down_write(&EXT4_I(inode)->i_data_sem);
old_disksize = EXT4_I(inode)->i_disksize;
EXT4_I(inode)->i_disksize = attr->ia_size;
@@ -5430,6 +5431,7 @@ int ext4_setattr(struct mnt_idmap *idmap
else
EXT4_I(inode)->i_disksize = old_disksize;
up_write(&EXT4_I(inode)->i_data_sem);
+ up_read(&EXT4_I(inode)->xattr_sem);
ext4_journal_stop(handle);
if (error)
goto out_mmap_sem;
--

[syzbot]

Hello,

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

Reported-and-tested-by: syzbot+ee72b9...@syzkaller.appspotmail.com

Tested on:

commit: fe46a7dd Merge tag 'sound-6.9-rc1' of git://git.kernel..
git tree: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
console output: https://syzkaller.appspot.com/x/log.txt?x=10fceed3180000

kernel config: https://syzkaller.appspot.com/x/.config?x=1a07d5da4eb21586
dashboard link: https://syzkaller.appspot.com/bug?extid=ee72b9a7aad1e5a77c5c
compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40

patch: https://syzkaller.appspot.com/x/patch.diff?x=132ad4f6180000

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

[Edward Adam Davis]

[Syzbot reported]

[Fix]
According to mark inode dirty context, it does not need to be protected by lock
i_data_sem, and if it is protected by i_data_sem, a deadlock will occur.

Reported-by: syzbot+ee72b9...@syzkaller.appspotmail.com
Signed-off-by: Edward Adam Davis <ead...@qq.com>
---
fs/ext4/inode.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 537803250ca9..d2cbe3dddfab 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -5417,6 +5417,7 @@ int ext4_setattr(struct mnt_idmap *idmap, struct dentry *dentry,

down_write(&EXT4_I(inode)->i_data_sem);
old_disksize = EXT4_I(inode)->i_disksize;
EXT4_I(inode)->i_disksize = attr->ia_size;

+ up_write(&EXT4_I(inode)->i_data_sem);
rc = ext4_mark_inode_dirty(handle, inode);
if (!error)
error = rc;
@@ -5425,6 +5426,7 @@ int ext4_setattr(struct mnt_idmap *idmap, struct dentry *dentry,
* with i_disksize to avoid races with writeback code
* running ext4_wb_update_i_disksize().
*/
+ down_write(&EXT4_I(inode)->i_data_sem);
if (!error)
i_size_write(inode, attr->ia_size);
else
--
2.43.0

[Andreas Dilger]

This seems like a false positive? There are two different inodes involved
in this case - the "real" inode, and the inode holding the xattr. I guess
this needs to be annotated so that lockdep doesn't complain about the order.

Cheers, Andreas

Cheers, Andreas