Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: INFO: task hung in io_uring_cancel_task_requests
To:     Pavel Begunkov <asml.silence@gmail.com>,
        syzbot <syzbot+695b03d82fa8e4901b06@syzkaller.appspotmail.com>,
        io-uring@vger.kernel.org, linux-fsdevel@vger.kernel.org,
        linux-kernel@vger.kernel.org, syzkaller-bugs@googlegroups.com,
        viro@zeniv.linux.org.uk
References: <00000000000075f90305baf0d732@google.com>
 <4ccfa4de-b63e-10c6-786a-d745cdf22535@gmail.com>
From:   Jens Axboe <axboe@kernel.dk>
Message-ID: <563f69d6-6e93-f375-c3b9-dcb3afdae025@kernel.dk>
Date:   Tue, 9 Feb 2021 20:02:11 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <4ccfa4de-b63e-10c6-786a-d745cdf22535@gmail.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 2/9/21 7:57 PM, Pavel Begunkov wrote:
> On 10/02/2021 00:54, syzbot wrote:
>> Hello,
>>
>> syzbot found the following issue on:
>>
>> HEAD commit:    dd86e7fa Merge tag 'pci-v5.11-fixes-2' of git://git.kernel..
>> git tree:       upstream
>> console output: https://syzkaller.appspot.com/x/log.txt?x=13e43f90d00000
>> kernel config:  https://syzkaller.appspot.com/x/.config?x=e83e68d0a6aba5f6
>> dashboard link: https://syzkaller.appspot.com/bug?extid=695b03d82fa8e4901b06
>> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=1490f0d4d00000
>> C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=17aedf1cd00000
>>
>> IMPORTANT: if you fix the issue, please add the following tag to the commit:
>> Reported-by: syzbot+695b03d82fa8e4901b06@syzkaller.appspotmail.com
> 
> It looks like SQPOLL. I wonder if that's due to parked SQPOLL task that
> won't be able to do task_work run, and so reap poll-cancelled requests
> killed by io_put_deferred().
> 
> I'll test it out tomorrow.

It is indeed SQPOLL. From a quick look, it's doing a POLL_ADD on the ring
fd itself.

-- 
Jens Axboe