linux-kernel - Re: [syzbot] [ext4?] general protection fault in hrtimer

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 01 Nov 2023 13:58:51 +0100
From:   Thomas Gleixner <tglx@...utronix.de>
To:     syzbot <syzbot+b408cd9b40ec25380ee1@...kaller.appspotmail.com>,
        adilger.kernel@...ger.ca, linux-ext4@...r.kernel.org,
        linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
        syzkaller-bugs@...glegroups.com, tytso@....edu
Subject: Re: [syzbot] [ext4?] general protection fault in hrtimer_nanosleep

On Tue, Oct 31 2023 at 22:36, syzbot wrote:
> general protection fault, probably for non-canonical address 0xdffffc003ffff113: 0000 [#1] PREEMPT SMP KASAN
> KASAN: probably user-memory-access in range [0x00000001ffff8898-0x00000001ffff889f]
> CPU: 1 PID: 5308 Comm: syz-executor.4 Not tainted 6.6.0-rc7-syzkaller-00142-g888cf78c29e2 #0
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/09/2023
> RIP: 0010:lookup_object lib/debugobjects.c:195 [inline]
> RIP: 0010:lookup_object_or_alloc lib/debugobjects.c:564 [inline]
> RIP: 0010:__debug_object_init+0xf3/0x2b0 lib/debugobjects.c:634
> Code: d8 48 c1 e8 03 42 80 3c 20 00 0f 85 85 01 00 00 48 8b 1b 48 85 db 0f 84 9f 00 00 00 48 8d 7b 18 83 c5 01 48 89 f8 48 c1 e8 03 <42> 80 3c 20 00 0f 85 4c 01 00 00 4c 3b 73 18 75 c3 48 8d 7b 10 48
> RSP: 0018:ffffc900050e7d08 EFLAGS: 00010012
> RAX: 000000003ffff113 RBX: 00000001ffff8880 RCX: ffffffff8169123e
> RDX: 1ffffffff249b149 RSI: 0000000000000004 RDI: 00000001ffff8898
> RBP: 0000000000000003 R08: 0000000000000001 R09: 0000000000000216
> R10: 0000000000000003 R11: 0000000000000000 R12: dffffc0000000000
> R13: ffffffff924d8a48 R14: ffffc900050e7d90 R15: ffffffff924d8a50
> FS:  0000555556eec480(0000) GS:ffff8880b9900000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 00007fa23ab065ee CR3: 000000007e5c1000 CR4: 0000000000350ee0

So this dies in debugobjects::lookup_object()

hlist_for_each_entry()

>   10:	48 8b 1b             	mov    (%rbx),%rbx

Gets the next entry

>   13:	48 85 db             	test   %rbx,%rbx
>   16:	0f 84 9f 00 00 00    	je     0xbb

Checks for the termination condition (NULL pointer)

>   1c:	48 8d 7b 18          	lea    0x18(%rbx),%rdi

Calculates the address of obj->object

>   20:	83 c5 01             	add    $0x1,%ebp

cnt++;

>   23:	48 89 f8             	mov    %rdi,%rax
>   26:	48 c1 e8 03          	shr    $0x3,%rax

KASAN shadow address calculation

> * 2a:	42 80 3c 20 00       	cmpb   $0x0,(%rax,%r12,1) <-- trapping instruction

Kasan accesses 0xdffffc003ffff113 and dies.

RBX contains the pointer to the next object: 0x00000001ffff8880 which is
clearly a user space address, but I have no idea where that might come
from. It's obviously data corruption of unknown provenience.

Unfortunately repro.syz does not hold up to its name and refuses to
reproduce.

Thanks,

        tglx