Date:   Tue, 30 Nov 2021 07:23:08 -0800
From:   Jakub Kicinski <kuba@kernel.org>
To:     Menglong Dong <menglong8.dong@gmail.com>
Cc:     David Miller <davem@davemloft.net>,
        Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>,
        dsahern@kernel.org, Eric Dumazet <edumazet@google.com>,
        Menglong Dong <imagedong@tencent.com>,
        Yuchung Cheng <ycheng@google.com>, kuniyu@amazon.co.jp,
        LKML <linux-kernel@vger.kernel.org>,
        netdev <netdev@vger.kernel.org>
Subject: Re: [PATCH v2 net-next] net: snmp: add statistics for tcp small
 queue check
Message-ID: <20211130072308.76cc711c@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
In-Reply-To: <CADxym3YJwgs1-hYZURUf+K56zTtQmWHbwAvEG27s_w8FwQrkQQ@mail.gmail.com>
References: <20211128060102.6504-1-imagedong@tencent.com>
        <20211129075707.47ab0ffe@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
        <CADxym3YJwgs1-hYZURUf+K56zTtQmWHbwAvEG27s_w8FwQrkQQ@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, 30 Nov 2021 22:36:59 +0800 Menglong Dong wrote:
> On Mon, Nov 29, 2021 at 11:57 PM Jakub Kicinski <kuba@kernel.org> wrote:
> >
> > On Sun, 28 Nov 2021 14:01:02 +0800 menglong8.dong@gmail.com wrote: =20
> > > Once tcp small queue check failed in tcp_small_queue_check(), the
> > > throughput of tcp will be limited, and it's hard to distinguish
> > > whether it is out of tcp congestion control.
> > >
> > > Add statistics of LINUX_MIB_TCPSMALLQUEUEFAILURE for this scene. =20
> >
> > Isn't this going to trigger all the time and alarm users because of the
> > "Failure" in the TCPSmallQueueFailure name?  Isn't it perfectly fine
> > for TCP to bake full TSQ amount of data and have it paced out onto the
> > wire? What's your link speed? =20
>=20
> Well, it's a little complex. In my case, there is a guest in kvm, and vir=
tio_net
> is used with napi_tx enabled.
>=20
> With napi_tx enabled, skb won't be orphaned after it is passed to virtio_=
net,
> until it is released. The point is that the sending interrupt of
> virtio_net will be
> turned off and the skb can't be released until the next net_rx interrupt =
comes.
> So, wmem_alloc can't decrease on time, and the bandwidth is limited. When
> this happens, the bandwidth can decrease from 500M to 10M.
>=20
> In fact, this issue of uapi_tx is fixed in this commit:
> https://lore.kernel.org/lkml/20210719144949.935298466@linuxfoundation.org/
>=20
> I added this statistics to monitor the sending failure (may be called
> sending delay)
> caused by qdisc and net_device. When something happen, maybe users can
> raise =E2=80=98/proc/sys/net/ipv4/tcp_pacing_ss_ratio=E2=80=99 to get bet=
ter bandwidth.

Sounds very second-order and particular to a buggy driver :/
Let's see what Eric says but I vote revert.