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: <20230620145338.1300897-1-dhowells@redhat.com> <20230620145338.1300897-2-dhowells@redhat.com>
 <20230622111234.23aadd87@kernel.org>
In-Reply-To: <20230622111234.23aadd87@kernel.org>
From:   Alexander Duyck <alexander.duyck@gmail.com>
Date:   Thu, 22 Jun 2023 11:28:56 -0700
Message-ID: <CAKgT0Ucp-62cOA_r1p=QwmSZvtVijW=9EEEgWYSp88=79wVfgA@mail.gmail.com>
Subject: Re: [PATCH net-next v3 01/18] net: Copy slab data for sendmsg(MSG_SPLICE_PAGES)
To:     Jakub Kicinski <kuba@kernel.org>
Cc:     David Howells <dhowells@redhat.com>, netdev@vger.kernel.org,
        "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        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, Menglong Dong <imagedong@tencent.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Thu, Jun 22, 2023 at 11:12=E2=80=AFAM Jakub Kicinski <kuba@kernel.org> w=
rote:
>
> On Tue, 20 Jun 2023 15:53:20 +0100 David Howells wrote:
> > If sendmsg() is passed MSG_SPLICE_PAGES and is given a buffer that cont=
ains
> > some data that's resident in the slab, copy it rather than returning EI=
O.
>
> How did that happen? I thought MSG_SPLICE_PAGES comes from former
> sendpage users and sendpage can't operate on slab pages.
>
> > This can be made use of by a number of drivers in the kernel, including=
:
> > iwarp, ceph/rds, dlm, nvme, ocfs2, drdb.  It could also be used by iscs=
i,
> > rxrpc, sunrpc, cifs and probably others.
> >
> > skb_splice_from_iter() is given it's own fragment allocator as
> > page_frag_alloc_align() can't be used because it does no locking to pre=
vent
> > parallel callers from racing.
>
> The locking is to local_bh_disable(). Does the milliont^w new frag
> allocator have any additional benefits?

Actually I would be concerned about it causing confusion since it is
called out as being how we do it for sockets but we already have
several different socket based setups using skb_page_frag_refill() and
the like.