KASAN: use-after-free Read in ext4_xattr_set_entry (4)

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: f8ad8187 fs/pipe: allow sendfile() to pipe again
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=11b8ad10d00000
kernel config: https://syzkaller.appspot.com/x/.config?x=96b123631a6700e9
dashboard link: https://syzkaller.appspot.com/bug?extid=4cb1e27475bf90a9b926
compiler: gcc (GCC) 10.1.0-syz 20200507
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=11131f94d00000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=15c3761b500000

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+4cb1e2...@syzkaller.appspotmail.com

==================================================================
BUG: KASAN: use-after-free in ext4_xattr_set_entry+0x3151/0x3780 fs/ext4/xattr.c:1586
Read of size 4 at addr ffff888030c00004 by task syz-executor392/11280

CPU: 0 PID: 11280 Comm: syz-executor392 Not tainted 5.11.0-rc5-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
__dump_stack lib/dump_stack.c:79 [inline]
dump_stack+0x107/0x163 lib/dump_stack.c:120
print_address_description.constprop.0.cold+0x5b/0x2f8 mm/kasan/report.c:230
__kasan_report mm/kasan/report.c:396 [inline]
kasan_report.cold+0x79/0xd5 mm/kasan/report.c:413
ext4_xattr_set_entry+0x3151/0x3780 fs/ext4/xattr.c:1586
ext4_xattr_ibody_set+0x78/0x2b0 fs/ext4/xattr.c:2224
ext4_xattr_set_handle+0x8f4/0x13e0 fs/ext4/xattr.c:2380
ext4_xattr_set+0x13a/0x340 fs/ext4/xattr.c:2493
__vfs_setxattr+0x10e/0x170 fs/xattr.c:177
__vfs_setxattr_noperm+0x11a/0x4c0 fs/xattr.c:208
__vfs_setxattr_locked+0x1bf/0x250 fs/xattr.c:266
vfs_setxattr+0x135/0x320 fs/xattr.c:291
setxattr+0x1ff/0x290 fs/xattr.c:553
path_setxattr+0x170/0x190 fs/xattr.c:572
__do_sys_setxattr fs/xattr.c:587 [inline]
__se_sys_setxattr fs/xattr.c:583 [inline]
__x64_sys_setxattr+0xc0/0x160 fs/xattr.c:583
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x445589
Code: e8 4c b3 02 00 48 83 c4 18 c3 0f 1f 80 00 00 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 0f 83 0b cc fb ff c3 66 2e 0f 1f 84 00 00 00 00
RSP: 002b:00007fffdf5f9b28 EFLAGS: 00000246 ORIG_RAX: 00000000000000bc
RAX: ffffffffffffffda RBX: 0000000000000016 RCX: 0000000000445589
RDX: 0000000000000000 RSI: 0000000020000140 RDI: 0000000020000100
RBP: 0000000001a18c30 R08: 0000000000000001 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 000000000000073d
R13: 00007fffdf5f9ba0 R14: 0000000000000000 R15: 0000000000000000

The buggy address belongs to the page:
page:00000000f7ed6945 refcount:0 mapcount:-128 mapping:0000000000000000 index:0x1 pfn:0x30c00
flags: 0xfff00000000000()
raw: 00fff00000000000 ffffea0000c56508 ffffea0000c17d88 0000000000000000
raw: 0000000000000001 0000000000000001 00000000ffffff7f 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
ffff888030bfff00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
ffff888030bfff80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>ffff888030c00000: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
^
ffff888030c00080: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
ffff888030c00100: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
==================================================================


---
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.
syzbot can test patches for this issue, for details see:
https://goo.gl/tpsmEJ#testing-patches

[syzbot]

syzbot suspects this issue was fixed by commit:

commit 6e47a3cc68fc525428297a00524833361ebbb0e9
Author: Lukas Czerner <lcze...@redhat.com>
Date: Wed Oct 27 14:18:52 2021 +0000

ext4: get rid of super block and sbi from handle_mount_ops()

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=100bc10b700000
start commit: f8ad8187c3b5 fs/pipe: allow sendfile() to pipe again
git tree: upstream

kernel config: https://syzkaller.appspot.com/x/.config?x=96b123631a6700e9
dashboard link: https://syzkaller.appspot.com/bug?extid=4cb1e27475bf90a9b926

syz repro: https://syzkaller.appspot.com/x/repro.syz?x=11131f94d00000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=15c3761b500000

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

#syz fix: ext4: get rid of super block and sbi from handle_mount_ops()

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

[Dmitry Vyukov]

Looks reasonable:

[Ritesh Harjani]

Sorry, I might have missed some discussion maybe.
But why do we think that this patch could fix the reported bug?
Because I see this patch from Lukas is a part of "ext4: new mount API
conversion" patch series. This I guess has nothing to do with the reported call
stack, no?

Or am I missing anything?

BUG: KASAN: use-after-free in ext4_xattr_set_entry+0x3151/0x3780 fs/ext4/xattr.c:1586
Read of size 4 at addr ffff888030c00004 by task syz-executor392/11280

CPU: 0 PID: 11280 Comm: syz-executor392 Not tainted 5.11.0-rc5-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
__dump_stack lib/dump_stack.c:79 [inline]
dump_stack+0x107/0x163 lib/dump_stack.c:120
print_address_description.constprop.0.cold+0x5b/0x2f8 mm/kasan/report.c:230
__kasan_report mm/kasan/report.c:396 [inline]
kasan_report.cold+0x79/0xd5 mm/kasan/report.c:413
ext4_xattr_set_entry+0x3151/0x3780 fs/ext4/xattr.c:1586
ext4_xattr_ibody_set+0x78/0x2b0 fs/ext4/xattr.c:2224
ext4_xattr_set_handle+0x8f4/0x13e0 fs/ext4/xattr.c:2380
ext4_xattr_set+0x13a/0x340 fs/ext4/xattr.c:2493
__vfs_setxattr+0x10e/0x170 fs/xattr.c:177
__vfs_setxattr_noperm+0x11a/0x4c0 fs/xattr.c:208
__vfs_setxattr_locked+0x1bf/0x250 fs/xattr.c:266
vfs_setxattr+0x135/0x320 fs/xattr.c:291
setxattr+0x1ff/0x290 fs/xattr.c:553
path_setxattr+0x170/0x190 fs/xattr.c:572
__do_sys_setxattr fs/xattr.c:587 [inline]
__se_sys_setxattr fs/xattr.c:583 [inline]
__x64_sys_setxattr+0xc0/0x160 fs/xattr.c:583
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46

-ritesh

[Dmitry Vyukov]

Hi Ritesh,

It looked reasonable because the identified patch is in ext4 and the
bug is also in ext4 + the bisection log looked reasonable + there were
no other suggestions/corrections. In such a case it's better to close
the bug with at least some reasonable fix, then let it pile to
hundreds of other unclosed bugs and prevent reporting of new bugs.