Received-SPF: pass (google.com: domain of linux-kernel+bounces-85466-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Precedence: bulk
MIME-Version: 1.0
References: <df300a49-7811-4126-a56a-a77100c8841b@gmail.com>
 <79a8ba83-86bf-4c22-845c-8f285c2d1396@gmail.com> <CAHsH6GvX7zYSoA7JVemRtunWWSaew1S11Y996WAGt6B9d8=cOA@mail.gmail.com>
 <aa5f1c11-4528-4d53-91f3-5ce8c02363ac@gmail.com>
In-Reply-To: <aa5f1c11-4528-4d53-91f3-5ce8c02363ac@gmail.com>
From: Eyal Birger <eyal.birger@gmail.com>
Date: Wed, 28 Feb 2024 09:51:14 -0800
Message-ID: <CAHsH6GspzSsSJX2Xuerp3NApJnobv7_eoNdwMSX=Mqw3EVhAzg@mail.gmail.com>
Subject: Re: [PATCH net-next 2/2] net: geneve: enable local address bind for
 geneve sockets
To: Richard Gobert <richardbgobert@gmail.com>
Cc: davem@davemloft.net, edumazet@google.com, kuba@kernel.org, 
	pabeni@redhat.com, idosch@nvidia.com, razor@blackwall.org, amcohen@nvidia.com, 
	petrm@nvidia.com, jbenc@redhat.com, b.galvani@gmail.com, bpoirier@nvidia.com, 
	gavinl@nvidia.com, martin.lau@kernel.org, daniel@iogearbox.net, 
	herbert@gondor.apana.org.au, netdev@vger.kernel.org, 
	linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Tue, Feb 27, 2024 at 1:03=E2=80=AFAM Richard Gobert <richardbgobert@gmai=
l.com> wrote:
>
> Eyal Birger wrote:
> > Hi,
> >
> > On Thu, Feb 22, 2024 at 12:54=E2=80=AFPM Richard Gobert
> > <richardbgobert@gmail.com> wrote:
> >>
> >> This patch adds support for binding to a local address in geneve socke=
ts.
> >
> > Thanks for adding this.
> >
> >> It achieves this by adding a geneve_addr union to represent local addr=
ess
> >> to bind to, and copying it to udp_port_cfg in geneve_create_sock.
> >
> > AFICT in geneve_sock_add(), geneve_socket_create() is only called if th=
ere's
> > no existing open socket with the GENEVE destination port. As such, woul=
dn't
> > this bind work only for the first socket in the namespace?
> >
> > If that is the case, then perhaps binding the socket isn't the right
> > approach, and instead geneve_lookup() should search for the tunnel base=
d on
> > both the source and destination IPs.
> >
> > Am I missing something?
> >
> > Eyal
>
> You are right, I missed it.
> Binding the socket is the main reason for the patch, to prevent exposing
> the geneve port on all interfaces.

I see. The use case I had in mind is allowing to differentiate between
tunnels based on local IP, but not exposing the port sounds like
a good use case too.

> I think it should be searched in geneve{6}_lookup and in geneve_find_sock=
:

If the socket is bound to a specific IP, i'm not sure you'd need to
change geneve{6}_lookup() - only packets matching that IP would arrive
there no? In that case I think the change you suggested to geneve_find_sock=
()
should be enough.

>
> static struct geneve_sock *geneve_find_sock(struct geneve_net *gn,
>                                             sa_family_t family,
>                                             union geneve_addr *saddr)
>  {
>         struct geneve_sock *gs;
>
>         list_for_each_entry(gs, &gn->sock_list, list) {
>                 struct inet_sock *inet =3D inet_sk(gs->sock->sk);
>
>                 if (inet->inet_sport =3D=3D dst_port && geneve_get_sk_fam=
ily(gs) =3D=3D family) {
>                         if (family =3D=3D AF_INET && inet->inet_rcv_saddr=
 =3D=3D saddr->sin.sin_addr.s_addr)
>                                 return gs;
>         ...
>
> This is also true for VXLAN

I haven't looked at the VXLAN code, but if the lookup is similar there
I would guess the same.

Eyal.