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;
Content-Type: text/plain;
        charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 14.0 \(3654.120.0.1.13\))
Subject: Re: [PATCH 1/2] io_uring: clear TIF_NOTIFY_SIGNAL when running task
 work
From:   Nadav Amit <nadav.amit@gmail.com>
In-Reply-To: <bbd25a42-eac0-a8f9-0e54-3c8c8e9894fd@gmail.com>
Date:   Tue, 10 Aug 2021 19:33:28 -0700
Cc:     Olivier Langlois <olivier@trillion01.com>,
        Jens Axboe <axboe@kernel.dk>, io-uring@vger.kernel.org,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Content-Transfer-Encoding: quoted-printable
Message-Id: <FD8FD9BD-1E94-4A84-88EB-3A1531BCF556@gmail.com>
References: <20210808001342.964634-1-namit@vmware.com>
 <20210808001342.964634-2-namit@vmware.com>
 <fdd54421f4d4e825152192e327c838d035352945.camel@trillion01.com>
 <A4DC14BA-74CA-41DB-BE08-D7B693C11AE0@gmail.com>
 <bbd25a42-eac0-a8f9-0e54-3c8c8e9894fd@gmail.com>
To:     Pavel Begunkov <asml.silence@gmail.com>
Precedence: bulk


> On Aug 10, 2021, at 2:32 PM, Pavel Begunkov <asml.silence@gmail.com> =
wrote:
>=20
> On 8/10/21 9:28 AM, Nadav Amit wrote:
>>=20
>> Unfortunately, there seems to be yet another issue (unless my code
>> somehow caused it). It seems that when SQPOLL is used, there are =
cases
>> in which we get stuck in io_uring_cancel_sqpoll() when =
tctx_inflight()
>> never goes down to zero.
>>=20
>> Debugging... (while also trying to make some progress with my code)
>=20
> It's most likely because a request has been lost (mis-refcounted).
> Let us know if you need any help. Would be great to solve it for 5.14.
> quick tips:=20
>=20
> 1) if not already, try out Jens' 5.14 branch
> git://git.kernel.dk/linux-block io_uring-5.14
>=20
> 2) try to characterise the io_uring use pattern. Poll requests?
> Read/write requests? Send/recv? Filesystem vs bdev vs sockets?
>=20
> If easily reproducible, you can match io_alloc_req() with it
> getting into io_dismantle_req();

So actually the problem is more of a missing IO-uring functionality that =
I need. When an I/O is queued for async completion (i.e., after =
returning -EIOCBQUEUED), there should be a way for io-uring to cancel =
these I/Os if needed. Otherwise they might potentially never complete, =
as happens in my use-case.

AIO has ki_cancel() for this matter. So I presume the proper solution =
would be to move ki_cancel() from aio_kiocb to kiocb so it can be used =
by both io-uring and aio. And then - to use this infrastructure.

But it is messy. There is already a bug in the (few) uses of =
kiocb_set_cancel_fn() that blindly assume AIO is used and not IO-uring. =
Then, I am not sure about some things in the AIO code. Oh boy. I=E2=80=99l=
l work on an RFC.