Re: KASAN: null-ptr-deref Write in io_uring_cancel_sqpoll

[Dmitry Vyukov]

On Mon, Apr 26, 2021 at 5:58 PM Palash Oswal <oswalpalash@xxxxxxxxx> wrote:
> On Friday, February 26, 2021 at 3:03:16 PM UTC+5:30 syzbot wrote:
>>
>> Hello,
>>
>> syzbot found the following issue on:
>>
>> HEAD commit: d01f2f7e Add linux-next specific files for 20210226
>> git tree: linux-next
>> console output: https://syzkaller.appspot.com/x/log.txt?x=108dc5a8d00000
>> kernel config: https://syzkaller.appspot.com/x/.config?x=a1746d2802a82a05
>> dashboard link: https://syzkaller.appspot.com/bug?extid=be51ca5a4d97f017cd50
>>
>> Unfortunately, I don't have any reproducer for this issue yet.
>>
>> IMPORTANT: if you fix the issue, please add the following tag to the commit:
>> Reported-by: syzbot+be51ca...@xxxxxxxxxxxxxxxxxxxxxxxxx
>>
>> ==================================================================
>> BUG: KASAN: null-ptr-deref in instrument_atomic_read_write include/linux/instrumented.h:101 [inline]
>> BUG: KASAN: null-ptr-deref in atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline]
>> BUG: KASAN: null-ptr-deref in io_uring_cancel_sqpoll+0x2c7/0x450 fs/io_uring.c:8871
>> Write of size 4 at addr 0000000000000110 by task iou-sqp-19439/19447
>>
>> CPU: 0 PID: 19447 Comm: iou-sqp-19439 Not tainted 5.11.0-next-20210226-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+0xfa/0x151 lib/dump_stack.c:120
>> __kasan_report mm/kasan/report.c:403 [inline]
>> kasan_report.cold+0x5f/0xd8 mm/kasan/report.c:416
>> check_region_inline mm/kasan/generic.c:180 [inline]
>> kasan_check_range+0x13d/0x180 mm/kasan/generic.c:186
>> instrument_atomic_read_write include/linux/instrumented.h:101 [inline]
>> atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline]
>> io_uring_cancel_sqpoll+0x2c7/0x450 fs/io_uring.c:8871
>> io_sq_thread+0x1109/0x1ae0 fs/io_uring.c:6782
>> ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294
>> ==================================================================
>> Kernel panic - not syncing: panic_on_warn set ...
>> CPU: 0 PID: 19447 Comm: iou-sqp-19439 Tainted: G B 5.11.0-next-20210226-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+0xfa/0x151 lib/dump_stack.c:120
>> panic+0x306/0x73d kernel/panic.c:231
>> end_report mm/kasan/report.c:102 [inline]
>> end_report.cold+0x5a/0x5a mm/kasan/report.c:88
>> __kasan_report mm/kasan/report.c:406 [inline]
>> kasan_report.cold+0x6a/0xd8 mm/kasan/report.c:416
>> check_region_inline mm/kasan/generic.c:180 [inline]
>> kasan_check_range+0x13d/0x180 mm/kasan/generic.c:186
>> instrument_atomic_read_write include/linux/instrumented.h:101 [inline]
>> atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline]
>> io_uring_cancel_sqpoll+0x2c7/0x450 fs/io_uring.c:8871
>> io_sq_thread+0x1109/0x1ae0 fs/io_uring.c:6782
>> ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294
>> Kernel Offset: disabled
>> Rebooting in 86400 seconds..
>>
>>
>> ---
>> 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...@xxxxxxxxxxxxxxxx.
>>
>> syzbot will keep track of this issue. See:
>> https://goo.gl/tpsmEJ#status for how to communicate with syzbot.
>
>
> My syzkaller instance reported a syz-repro for this bug:
> Syzkaller reproducer: # {Threaded:true Collide:true Repeat:true RepeatTimes:0 Procs:2 Slowdown:1 Sandbox:none Fault:false FaultCall:-1 FaultNth:0 Leak:false NetInjection:true NetDevices:true NetReset:true Cgroups:true BinfmtMisc:true CloseFDs:true KCSAN:false DevlinkPCI:false USB:false VhciInjection:false Wifi:false IEEE802154:false Sysctl:true UseTmpDir:true HandleSegv:true Repro:false Trace:false}
> r0 = fsmount(0xffffffffffffffff, 0x1, 0xc)
> syz_io_uring_setup(0x329b, &(0x7f0000000080)={0x0, 0x850e, 0x2, 0x2, 0x1b4}, &(0x7f0000ffc000/0x4000)=nil, &(0x7f0000ffa000/0x4000)=nil, 0x0, 0x0)
> syz_io_uring_setup(0x3de2, &(0x7f0000001480)={0x0, 0x4f62, 0x4, 0x2, 0x75}, &(0x7f0000ffb000/0x3000)=nil, &(0x7f0000ffd000/0x3000)=nil, 0x0, 0x0)
> fsetxattr$trusted_overlay_nlink(r0, &(0x7f0000000140), 0x0, 0x0, 0x0)
>
> I'm working to get a c reproducer for it that is consistent. This syz-repro does not produce a working reproducer for me just yet.
> Initial suspicion is that io_sq_thread_stop sets set_bit(IO_SQ_THREAD_SHOULD_STOP, &sqd->state);
> And subsequently after a return from fork, where the process receives a SIGKILL and io_uring_cancel_sqpoll(ctx) is called with a NULL ctx in io_sq_thread(). I haven't connected all of the dots yet, working on it.

+kernel lists and syzbot email
(almost nobody is reading syzkaller-bugs@ itself)