[PATCH] ext4: WARNING: locking bug in ext4_ioctl

[PATCH] ext4: WARNING: locking bug in ext4_ioctl

[Remington Brasga]

Fixed the locking bug with the bisection found by syzbot.

Link to the syzbot bug report:
https://syzkaller.appspot.com/bug?extid=a3c8e9ac9f9d77240afd

Reported-by: syzbot+a3c8e9ac9f9d77240afd@syzkaller.appspotmail.com
Signed-off-by: Remington Brasga <rbrasga@uci.edu>
---
First attempt at patching this issue. Verified the bug exists, and that
this patch fixes it, and EXT4 KUnit tests pass. Though I am not sure if
the patch should be higher up the stack.

Open to questions or any suggestions for improvements.

#syz test

 fs/ext4/xattr.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c
index e0e1956dcdd3..2b30b9571fd8 100644
--- a/fs/ext4/xattr.c
+++ b/fs/ext4/xattr.c
@@ -124,8 +124,12 @@ void ext4_xattr_inode_set_class(struct inode *ea_inode)
 	struct ext4_inode_info *ei = EXT4_I(ea_inode);
 
 	lockdep_set_subclass(&ea_inode->i_rwsem, 1);
-	(void) ei;	/* shut up clang warning if !CONFIG_LOCKDEP */
-	lockdep_set_subclass(&ei->i_data_sem, I_DATA_SEM_EA);
+
+	if (ei->i_flags & EXT4_EA_INODE_FL) {
+		lockdep_set_subclass(&ei->i_data_sem, I_DATA_SEM_EA);
+	} else {
+		lockdep_set_subclass(&ei->i_data_sem, I_DATA_SEM_NORMAL);
+	}
 }
 #endif
 
-- 
2.34.1

Re: [syzbot] [ext4?] WARNING: locking bug in ext4_ioctl

[syzbot]

Hello,

syzbot has tested the proposed patch but the reproducer is still triggering an issue:
WARNING: locking bug in ext4_ioctl

loop0: detected capacity change from 0 to 512
ext4 filesystem being mounted at /root/syzkaller-testdir500386566/syzkaller.6PgxEV/36/bus supports timestamps until 2038-01-19 (0x7fffffff)
------------[ cut here ]------------
Looking for class "&ei->i_data_sem" with key init_once.__key.799, but found a different class "&ei->i_data_sem" with the same key
WARNING: CPU: 1 PID: 5815 at kernel/locking/lockdep.c:939 look_up_lock_class+0xdc/0x170 kernel/locking/lockdep.c:936
Modules linked in:
CPU: 1 UID: 0 PID: 5815 Comm: syz-executor.0 Not tainted 6.11.0-syzkaller-11673-geee280841e1c-dirty #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/13/2024
RIP: 0010:look_up_lock_class+0xdc/0x170 kernel/locking/lockdep.c:936
Code: 01 0f 85 8a 00 00 00 c6 05 b3 f8 46 04 01 90 49 8b 16 49 8b 76 18 48 8b 8b b8 00 00 00 48 c7 c7 60 ea ea 8b e8 75 2c bf f5 90 <0f> 0b 90 90 eb 61 90 e8 08 44 18 f9 e8 53 5e de f5 48 c7 c7 a0 e9
RSP: 0018:ffffc900049ef4f0 EFLAGS: 00010046
RAX: 3592bb3d4927ad00 RBX: ffffffff931c0730 RCX: ffff888027c18000
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000
RBP: ffffc900049ef600 R08: ffffffff8155a9c2 R09: 1ffff1101722519a
R10: dffffc0000000000 R11: ffffed101722519b R12: ffff888077aadbb0
R13: ffff888077aadbb0 R14: ffff888077aadbb0 R15: ffffffff94d9f1c1
FS:  00007f214c1256c0(0000) GS:ffff8880b9100000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000000002000e000 CR3: 0000000027aec000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 register_lock_class+0x102/0x980 kernel/locking/lockdep.c:1290
 __lock_acquire+0xf0/0x2050 kernel/locking/lockdep.c:5074
 lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5822
 down_write_nested+0xa2/0x220 kernel/locking/rwsem.c:1695
 swap_inode_boot_loader fs/ext4/ioctl.c:432 [inline]
 __ext4_ioctl fs/ext4/ioctl.c:1436 [inline]
 ext4_ioctl+0x4124/0x5590 fs/ext4/ioctl.c:1626
 vfs_ioctl fs/ioctl.c:51 [inline]
 __do_sys_ioctl fs/ioctl.c:907 [inline]
 __se_sys_ioctl+0xf9/0x170 fs/ioctl.c:893
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7f214b47dea9
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:00007f214c1250c8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
RAX: ffffffffffffffda RBX: 00007f214b5abf80 RCX: 00007f214b47dea9
RDX: 0000000000000000 RSI: 0000000000006611 RDI: 0000000000000004
RBP: 00007f214b4ca4a4 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 000000000000000b R14: 00007f214b5abf80 R15: 00007fff61b66448
 </TASK>


Tested on:

commit:         eee28084 Merge tag 'mm-hotfixes-stable-2024-09-27-09-4..
git tree:       upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=161cde27980000
kernel config:  https://syzkaller.appspot.com/x/.config?x=6e73c530afb43e82
dashboard link: https://syzkaller.appspot.com/bug?extid=a3c8e9ac9f9d77240afd
compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
patch:          https://syzkaller.appspot.com/x/patch.diff?x=17082507980000

Re: [syzbot] [ext4?] WARNING: locking bug in ext4_ioctl

[syzbot]

syzbot suspects this issue was fixed by commit:

commit d7fe143cb115076fed0126ad8cf5ba6c3e575e43
Author: Ahmed Ehab <bottaawesome633@gmail.com>
Date:   Sat Aug 24 22:10:30 2024 +0000

    locking/lockdep: Avoid creating new name string literals in lockdep_set_subclass()

bisection log:  https://syzkaller.appspot.com/x/bisect.txt?x=16c9ffd8580000
start commit:   e42b1a9a2557 Merge tag 'spi-fix-v6.12-rc5' of git://git.ke..
git tree:       upstream
kernel config:  https://syzkaller.appspot.com/x/.config?x=4aec7739e14231a7
dashboard link: https://syzkaller.appspot.com/bug?extid=a3c8e9ac9f9d77240afd
syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=123fef57980000
C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=1598eebb980000

If the result looks correct, please mark the issue as fixed by replying with:

#syz fix: locking/lockdep: Avoid creating new name string literals in lockdep_set_subclass()

For information about bisection process see: https://goo.gl/tpsmEJ#bisection