Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20230623225513.2732256-1-dhowells@redhat.com> <20230623225513.2732256-4-dhowells@redhat.com>
 <CAOi1vP9vjLfk3W+AJFeexC93jqPaPUn2dD_4NrzxwoZTbYfOnw@mail.gmail.com> <3068221.1687788027@warthog.procyon.org.uk>
In-Reply-To: <3068221.1687788027@warthog.procyon.org.uk>
From:   Ilya Dryomov <idryomov@gmail.com>
Date:   Mon, 26 Jun 2023 17:41:03 +0200
Message-ID: <CAOi1vP8AyL=nsqDw-QKjPsC8wMsEnq+hSh29PobADYLm-L9ZNg@mail.gmail.com>
Subject: Re: [PATCH net-next v5 03/16] ceph: Use sendmsg(MSG_SPLICE_PAGES)
 rather than sendpage
To:     David Howells <dhowells@redhat.com>
Cc:     netdev@vger.kernel.org,
        Alexander Duyck <alexander.duyck@gmail.com>,
        "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>,
        Willem de Bruijn <willemdebruijn.kernel@gmail.com>,
        David Ahern <dsahern@kernel.org>,
        Matthew Wilcox <willy@infradead.org>,
        Jens Axboe <axboe@kernel.dk>, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org, Xiubo Li <xiubli@redhat.com>,
        Jeff Layton <jlayton@kernel.org>, ceph-devel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Mon, Jun 26, 2023 at 4:00=E2=80=AFPM David Howells <dhowells@redhat.com>=
 wrote:
>
> Ilya Dryomov <idryomov@gmail.com> wrote:
>
> > Same here...  I would suggest that you keep ceph_tcp_sendpage() functio=
n
> > and make only minimal modifications to avoid regressions.
>
> This is now committed to net-next.

This needs to be dropped from linux-next because both this and
especially the other (net/ceph/messenger_v2.c) patch introduce
regressions.

> I can bring ceph_tcp_sendpage() back into
> existence or fix it in place for now if you have a preference.

I already mentioned that I would prefer if ceph_tcp_sendpage() was
brought back into existence.

>
> Note that I'm working on patches to rework the libceph transmission path =
so
> that it isn't dealing with transmitting a single page at a time, but it's=
 not
> ready yet.

That is a worthwhile improvement now that sock_sendmsg() can take
advantage of multiple pages!  It would be pretty invasive though so
I think ceph_tcp_sendpage() is better to remain in place until then.

Thanks,

                Ilya