Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
In-Reply-To: <20180911063335.GA3315@infradead.org>
References: <000000000000cbb35d05757f7a3a@google.com> <20180910165317.GA3237@infradead.org>
 <CAJfpeguWqJgU8EfjvOtsBYRbbY2J7Ck2oNUcLNS8my9g2eNZfA@mail.gmail.com> <20180911063335.GA3315@infradead.org>
From:   Miklos Szeredi <miklos@szeredi.hu>
Date:   Tue, 11 Sep 2018 09:20:40 +0200
Message-ID: <CAJfpegtS8QMfhNMFV3i6GycqmXsLNYXvKjRm3wG79PFuF1VvuA@mail.gmail.com>
Subject: Re: possible deadlock in aio_poll
To:     Christoph Hellwig <hch@infradead.org>
Cc:     syzbot <syzbot+5b1df0420c523b45a953@syzkaller.appspotmail.com>,
        bcrl <bcrl@kvack.org>, linux-aio <linux-aio@kvack.org>,
        linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
        syzkaller-bugs <syzkaller-bugs@googlegroups.com>,
        Al Viro <viro@zeniv.linux.org.uk>,
        Andrea Arcangeli <aarcange@redhat.com>,
        Andrew Morton <akpm@linux-foundation.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Sep 11, 2018 at 8:33 AM, Christoph Hellwig <hch@infradead.org> wrote:
> On Mon, Sep 10, 2018 at 08:14:20PM +0200, Miklos Szeredi wrote:
>> Why do pollable waitqueues need to disable interrupts generally?
>
> Any waitqueue needs to disable interrupts for consistency.  We
> always use spin_lock_irqsave in __wake_up_common_lock() for example.

There are the _locked (non _irq) variants that do not.

And poll/select/etc don't impose non-interuptibility on wakeups
either.  So it looks like it's just aio that has weird spin lock
dependencies that forces this requirement on a waitq used in ->poll().

Thanks,
Miklos