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;
MIME-Version: 1.0
References: <1615603667-22568-1-git-send-email-linyunsheng@huawei.com>
 <1615777818-13969-1-git-send-email-linyunsheng@huawei.com>
 <20210315115332.1647e92b@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
 <CAM_iQpXvVZxBRHF6PBDOYSOSCj08nPyfcY0adKuuTg=cqffV+w@mail.gmail.com>
 <87eegddhsj.fsf@toke.dk> <CAHmME9qDU7VRmBV+v0tzLiUpMJykjswSDwqc9P43ZwG1UD7mzw@mail.gmail.com>
 <3bae7b26-9d7f-15b8-d466-ff5c26d08b35@huawei.com> <CAM_iQpVvR1eUQxgihWrZ==X=xQjaaeH_qkehvU0Y2R6i9eM-Qw@mail.gmail.com>
 <9d045462-051e-0cde-24d0-349dd397e2b7@huawei.com>
In-Reply-To: <9d045462-051e-0cde-24d0-349dd397e2b7@huawei.com>
From:   Cong Wang <xiyou.wangcong@gmail.com>
Date:   Tue, 23 Mar 2021 18:49:06 -0700
Message-ID: <CAM_iQpVgARDaUd3jdvSA11j=Q_K6KvcKfn7DQavGYXUWmvLZtw@mail.gmail.com>
Subject: Re: [Linuxarm] Re: [RFC v2] net: sched: implement TCQ_F_CAN_BYPASS
 for lockless qdisc
To:     Yunsheng Lin <linyunsheng@huawei.com>
Cc:     "Jason A. Donenfeld" <Jason@zx2c4.com>,
        =?UTF-8?B?VG9rZSBIw7hpbGFuZC1Kw7hyZ2Vuc2Vu?= <toke@redhat.com>,
        Jakub Kicinski <kuba@kernel.org>,
        David Miller <davem@davemloft.net>,
        Vladimir Oltean <olteanv@gmail.com>,
        Alexei Starovoitov <ast@kernel.org>,
        Daniel Borkmann <daniel@iogearbox.net>,
        Andrii Nakryiko <andriin@fb.com>,
        Eric Dumazet <edumazet@google.com>,
        Wei Wang <weiwan@google.com>,
        "Cong Wang ." <cong.wang@bytedance.com>,
        Taehee Yoo <ap420073@gmail.com>,
        Linux Kernel Network Developers <netdev@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>, linuxarm@openeuler.org,
        Marc Kleine-Budde <mkl@pengutronix.de>,
        linux-can@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Sun, Mar 21, 2021 at 5:55 PM Yunsheng Lin <linyunsheng@huawei.com> wrote=
:
>
> On 2021/3/20 2:15, Cong Wang wrote:
> > On Thu, Mar 18, 2021 at 12:33 AM Yunsheng Lin <linyunsheng@huawei.com> =
wrote:
> >>
> >> On 2021/3/17 21:45, Jason A. Donenfeld wrote:
> >>> On 3/17/21, Toke H=C3=B8iland-J=C3=B8rgensen <toke@redhat.com> wrote:
> >>>> Cong Wang <xiyou.wangcong@gmail.com> writes:
> >>>>
> >>>>> On Mon, Mar 15, 2021 at 2:07 PM Jakub Kicinski <kuba@kernel.org> wr=
ote:
> >>>>>>
> >>>>>> I thought pfifo was supposed to be "lockless" and this change
> >>>>>> re-introduces a lock between producer and consumer, no?
> >>>>>
> >>>>> It has never been truly lockless, it uses two spinlocks in the ring
> >>>>> buffer
> >>>>> implementation, and it introduced a q->seqlock recently, with this =
patch
> >>>>> now we have priv->lock, 4 locks in total. So our "lockless" qdisc e=
nds
> >>>>> up having more locks than others. ;) I don't think we are going to =
a
> >>>>> right direction...
> >>>>
> >>>> Just a thought, have you guys considered adopting the lockless MSPC =
ring
> >>>> buffer recently introduced into Wireguard in commit:
> >>>>
> >>>> 8b5553ace83c ("wireguard: queueing: get rid of per-peer ring buffers=
")
> >>>>
> >>>> Jason indicated he was willing to work on generalising it into a
> >>>> reusable library if there was a use case for it. I haven't quite tho=
ugh
> >>>> through the details of whether this would be such a use case, but
> >>>> figured I'd at least mention it :)
> >>>
> >>> That offer definitely still stands. Generalization sounds like a lot =
of fun.
> >>>
> >>> Keep in mind though that it's an eventually consistent queue, not an
> >>> immediately consistent one, so that might not match all use cases. It
> >>> works with wg because we always trigger the reader thread anew when i=
t
> >>> finishes, but that doesn't apply to everyone's queueing setup.
> >>
> >> Thanks for mentioning this.
> >>
> >> "multi-producer, single-consumer" seems to match the lockless qdisc's
> >> paradigm too, for now concurrent enqueuing/dequeuing to the pfifo_fast=
's
> >> queues() is not allowed, it is protected by producer_lock or consumer_=
lock.
> >>
> >> So it would be good to has lockless concurrent enqueuing, while dequeu=
ing
> >> can be protected by qdisc_lock() or q->seqlock, which meets the "multi=
-producer,
> >> single-consumer" paradigm.
> >
> > I don't think so. Usually we have one queue for each CPU so we can expe=
ct
> > each CPU has a lockless qdisc assigned, but we can not assume this in
> > the code, so we still have to deal with multiple CPU's sharing a lockle=
ss qdisc,
> > and we usually enqueue and dequeue in process context, so it means we c=
ould
> > have multiple producers and multiple consumers.
>
> For lockless qdisc, dequeuing is always within the qdisc_run_begin() and
> qdisc_run_end(), so multiple consumers is protected with each other by
> q->seqlock .

So are you saying you will never go lockless for lockless qdisc? I thought
you really want to go lockless with Jason's proposal of MPMC ring buffer
code.

>
> For enqueuing, multiple consumers is protected by producer_lock, see
> pfifo_fast_enqueue() -> skb_array_produce() -> ptr_ring_produce().

I think you seriously misunderstand how we classify MPMC or MPSC,
it is not about how we lock them, it is about whether we truly have
a single or multiple consumers regardless of locks used, because the
goal is to go lockless.

> I am not sure if lockless MSPC can work with the process context, but
> even if not, the enqueuing is also protected by rcu_read_lock_bh(),
> which provides some kind of atomicity, so that producer_lock can be
> reomved when lockless MSPC is used.


Not sure if I can even understand what you are saying here, Jason's
code only disables preemption with busy wait, I can't see why it can
not be used in the process context.

Thanks.