KASAN: slab-out-of-bounds Read in squashfs_get_id

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: 171d4ff7 Merge tag 'mmc-v5.9-rc4-2' of git://git.kernel.or..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=1597ead3900000
kernel config: https://syzkaller.appspot.com/x/.config?x=af502ec9a451c9fc
dashboard link: https://syzkaller.appspot.com/bug?extid=8e28bba73ed1772a6802
compiler: clang version 10.0.0 (https://github.com/llvm/llvm-project/ c2443155a0fb245c8f17f2c1c72b6ea391e86e81)
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=172ff481900000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=17c3e6c5900000

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+8e28bb...@syzkaller.appspotmail.com

==================================================================
BUG: KASAN: slab-out-of-bounds in squashfs_get_id+0xb9/0x1c0 fs/squashfs/id.c:38
Read of size 8 at addr ffff8880a9684b98 by task syz-executor329/6836

CPU: 1 PID: 6836 Comm: syz-executor329 Not tainted 5.9.0-rc6-syzkaller #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+0x1d6/0x29e lib/dump_stack.c:118
print_address_description+0x66/0x620 mm/kasan/report.c:383
__kasan_report mm/kasan/report.c:513 [inline]
kasan_report+0x132/0x1d0 mm/kasan/report.c:530
squashfs_get_id+0xb9/0x1c0 fs/squashfs/id.c:38
squashfs_new_inode fs/squashfs/inode.c:51 [inline]
squashfs_read_inode+0x155/0x2170 fs/squashfs/inode.c:120
squashfs_fill_super+0x1478/0x1790 fs/squashfs/super.c:310
get_tree_bdev+0x3e9/0x5f0 fs/super.c:1342
vfs_get_tree+0x88/0x270 fs/super.c:1547
do_new_mount fs/namespace.c:2875 [inline]
path_mount+0x179d/0x29e0 fs/namespace.c:3192
do_mount fs/namespace.c:3205 [inline]
__do_sys_mount fs/namespace.c:3413 [inline]
__se_sys_mount+0x126/0x180 fs/namespace.c:3390
do_syscall_64+0x31/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x446d1a
Code: b8 08 00 00 00 0f 05 48 3d 01 f0 ff ff 0f 83 fd ad fb ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 49 89 ca b8 a5 00 00 00 0f 05 <48> 3d 01 f0 ff ff 0f 83 da ad fb ff c3 66 0f 1f 84 00 00 00 00 00
RSP: 002b:00007ffd7dd4f8b8 EFLAGS: 00000293 ORIG_RAX: 00000000000000a5
RAX: ffffffffffffffda RBX: 00007ffd7dd4f910 RCX: 0000000000446d1a
RDX: 0000000020000000 RSI: 0000000020000100 RDI: 00007ffd7dd4f8d0
RBP: 00007ffd7dd4f8d0 R08: 00007ffd7dd4f910 R09: 00007ffd00000015
R10: 0000000000000000 R11: 0000000000000293 R12: 0000000000000001
R13: 0000000000000004 R14: 0000000000000003 R15: 0000000000000003

Allocated by task 3913:
kasan_save_stack mm/kasan/common.c:48 [inline]
kasan_set_track mm/kasan/common.c:56 [inline]
__kasan_kmalloc+0x100/0x130 mm/kasan/common.c:461
kmalloc_node include/linux/slab.h:577 [inline]
__vmalloc_area_node mm/vmalloc.c:2429 [inline]
__vmalloc_node_range+0x2c7/0x870 mm/vmalloc.c:2511
module_alloc+0x7e/0x90 arch/x86/kernel/module.c:75
bpf_jit_binary_alloc+0x123/0x230 kernel/bpf/core.c:871
bpf_int_jit_compile+0x7995/0x8920 arch/x86/net/bpf_jit_comp.c:1911
bpf_prog_select_runtime+0x76d/0xa60 kernel/bpf/core.c:1807
bpf_migrate_filter net/core/filter.c:1290 [inline]
bpf_prepare_filter+0xec2/0x1140 net/core/filter.c:1338
bpf_prog_create_from_user+0x2ad/0x3e0 net/core/filter.c:1432
seccomp_prepare_filter kernel/seccomp.c:567 [inline]
seccomp_prepare_user_filter kernel/seccomp.c:604 [inline]
seccomp_set_mode_filter kernel/seccomp.c:1546 [inline]
do_seccomp+0x852/0x20b0 kernel/seccomp.c:1661
do_syscall_64+0x31/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xa9

The buggy address belongs to the object at ffff8880a9684b80
which belongs to the cache kmalloc-32 of size 32
The buggy address is located 24 bytes inside of
32-byte region [ffff8880a9684b80, ffff8880a9684ba0)
The buggy address belongs to the page:
page:00000000f697ca3d refcount:1 mapcount:0 mapping:0000000000000000 index:0xffff8880a9684fc1 pfn:0xa9684
flags: 0xfffe0000000200(slab)
raw: 00fffe0000000200 ffffea0002a5d5c8 ffffea0002a98588 ffff8880aa440100
raw: ffff8880a9684fc1 ffff8880a9684000 000000010000003f 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
ffff8880a9684a80: fa fb fb fb fc fc fc fc fa fb fb fb fc fc fc fc
ffff8880a9684b00: 00 fc fc fc fc fc fc fc fa fb fb fb fc fc fc fc
>ffff8880a9684b80: 00 fc fc fc fc fc fc fc fa fb fb fb fc fc fc fc
^
ffff8880a9684c00: fa fb fb fb fc fc fc fc fa fb fb fb fc fc fc fc
ffff8880a9684c80: 00 00 01 fc fc fc fc fc fa fb fb fb fc fc fc fc
==================================================================


---
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

[Fox Chen]

Hi, 

I found this bug was caused by either uid/gid info in superblocks or id_index_table is corrupted. The uid/gid index is larger than the size of msblk->id_table.

Should I add a sanity check to squashfs_get_id?? 

The complete solution is to record the size of msblk->id_table in msblk and check uid/gid index each time in squashfs_get_id. However, this requires a change to msblk struct.

A simple solution is to calculate the max available room for uid/gid table by doing msblk->xattr_table - msblk->id_table[0] and check if index is larger than this. While this provides some sort of check, it is imperfect because id_table can be smaller than that. 

Both of them work out for this bug.

thanks,

fox

[Fox Chen]

Hi,


I found this bug was caused by either uid/gid info in superblocks or
id_index_table is corrupted. The uid/gid index is larger than the size
of msblk->id_table.

Should I add a sanity check to squashfs_get_id??

The complete solution is to record the size of msblk->id_table in msblk
and check uid/gid index each time in squashfs_get_id. However, this
requires a change to msblk struct.

A simple solution is to calculate the max available room for uid/gid
table by doing msblk->xattr_table - msblk->id_table[0] and check if
index is larger than this. While this provides some sort of check, it is
imperfect because id_table can be smaller than that.
Both of them work out for this bug.


thanks,
fox

On Friday, September 25, 2020 at 10:48:18 PM UTC+8 syzbot wrote:

[Dmitry Vyukov]

FYI Gmail has a setting of making Replay All the default button.

--
You received this message because you are subscribed to the Google Groups "syzkaller-bugs" group.
To unsubscribe from this group and stop receiving emails from it, send an email to syzkaller-bug...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/syzkaller-bugs/300c32a1-a0de-4f05-b467-0822b3d21733n%40googlegroups.com.

[Fox Chen]

Oh, Thank you, Dmitry!

I firstly replied to this from
https://groups.google.com/g/syzkaller-bugs/c/SPtlDDI7jmw/m/BtgjaCcDAwAJ.
I don't know why "reply all" only replies to syzkaller-bugs and I
didn't notice that.

Sorry about that.

[Dmitry Vyukov]

On Wed, Oct 14, 2020 at 10:02 AM Fox Chen <foxh...@gmail.com> wrote:
>
> Oh, Thank you, Dmitry!
>
> I firstly replied to this from
> https://groups.google.com/g/syzkaller-bugs/c/SPtlDDI7jmw/m/BtgjaCcDAwAJ.
> I don't know why "reply all" only replies to syzkaller-bugs and I
> didn't notice that.
>
> Sorry about that.

Oh, from groups. Yes, that won't work for the kernel. That also
generated an HTML email, which is generally not welcome on kernel
mailing lists.

[Fox Chen]

On Wed, Oct 14, 2020 at 4:27 PM Dmitry Vyukov <dvy...@google.com> wrote:
>
> Oh, from groups. Yes, that won't work for the kernel. That also
> generated an HTML email, which is generally not welcome on kernel
> mailing lists.
>

Got it! thank you!

If I miss the original email, without using "groups", how can I reply
to the thread??
Apart from manually copying/pasting the content, recipients, and
setting in-reply-to, Is there any convenient way??

[Dmitry Vyukov]

On Wed, Oct 14, 2020 at 11:13 AM Fox Chen <foxh...@gmail.com> wrote:
> > Oh, from groups. Yes, that won't work for the kernel. That also
> > generated an HTML email, which is generally not welcome on kernel
> > mailing lists.
> >
>
> Got it! thank you!
>
> If I miss the original email, without using "groups", how can I reply
> to the thread??
> Apart from manually copying/pasting the content, recipients, and
> setting in-reply-to, Is there any convenient way??

No, unfortunately, no convenient way. Nature of email-based
development workflow.

I think I've seen some web archives in the past that allowed you to
forward an email to yourself, and then you can reply to it normally.
But I don't remember where it was. Maybe you can find it.

[Fox Chen]

umm, bittersweet :)

Thank you, I'll check it out.

[Lukas Bulwahn]

On Wed, Oct 14, 2020 at 11:21 AM Dmitry Vyukov via
Linux-kernel-mentees <linux-kern...@lists.linuxfoundation.org>
wrote:

If the mailing list is archived on lore.kernel.org
(https://lore.kernel.org/lists.html), you can search the email there
and you will find a suitable mailto link (which might work for your
client)

Alternatively, you can download an mbox, load that with an email
client, and then respond from mutt, pine etc.

syzkaller-bugs is not archived at lore.kernel.org, but I assumed all
those emails there also go out to lkml.

If useful, we could ask to archive syzkaller-bugs on lore.kernel.org as well.

Lukas

[Fox Chen]

Hi Lukas,

> If the mailing list is archived on lore.kernel.org
> (https://lore.kernel.org/lists.html), you can search the email there
> and you will find a suitable mailto link (which might work for your
> client)
>
> Alternatively, you can download an mbox, load that with an email
> client, and then respond from mutt, pine etc.

Oh, Great!

> syzkaller-bugs is not archived at lore.kernel.org, but I assumed all
> those emails there also go out to lkml.
>
> If useful, we could ask to archive syzkaller-bugs on lore.kernel.org as well.
>

Yes, that would be helpful.


thanks,
fox

[Dmitry Vyukov]

All syzbot emails are also CCed to LKML, so searching linux-kernel should do.

[Greg KH]

On Wed, Oct 14, 2020 at 05:13:29PM +0800, Fox Chen wrote:
> On Wed, Oct 14, 2020 at 4:27 PM Dmitry Vyukov <dvy...@google.com> wrote:
> >
> > Oh, from groups. Yes, that won't work for the kernel. That also
> > generated an HTML email, which is generally not welcome on kernel
> > mailing lists.
> >
>
> Got it! thank you!
>
> If I miss the original email, without using "groups", how can I reply
> to the thread??

If the tool lets you, just download the "raw" message, and load it in
your email client and respond to it from there.

I know lore.kernel.org has the option to save it this way, then just use
a mail client like mutt:
mutt -f message_downloaded.txt

and away you go.

If your email client does not let you do this, perhaps you need a better
email client :)

thanks,

greg k-h

[Fox Chen]

Hi Anant, Dmitry, Greg,

Wow, mbox & raw work. lore has mailto link, just on-click, even more convenient.

And, yes, I've found the mail on linux-kernel list.


Thank you all for the help! :)


Cheers!
fox

[syzbot]

syzbot suspects this issue was fixed by commit:

commit e812cbbbbbb15adbbbee176baa1e8bda53059bf0
Author: Phillip Lougher <phi...@squashfs.org.uk>
Date: Tue Feb 9 21:41:50 2021 +0000

squashfs: avoid out of bounds writes in decompressors

bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=11bfa48ad00000
start commit: cd796ed3 Merge tag 'trace-v5.10-rc7' of git://git.kernel.o..
git tree: upstream
kernel config: https://syzkaller.appspot.com/x/.config?x=59df2a4dced5f928
dashboard link: https://syzkaller.appspot.com/bug?extid=8e28bba73ed1772a6802
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=1138f80f500000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=125e080f500000

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

#syz fix: squashfs: avoid out of bounds writes in decompressors

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

[Dmitry Vyukov]

Looks reasonable: