Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
Message-ID: <be0d17cf-12d5-440c-adee-e943ccb199c9@default>
Date:   Thu, 15 Mar 2018 05:23:41 -0700 (PDT)
From:   Liran Alon <liran.alon@oracle.com>
To:     <daniel@iogearbox.net>
Cc:     <netdev@vger.kernel.org>, <shmulik.ladkani@gmail.com>,
        <davem@davemloft.net>, <linux-kernel@vger.kernel.org>,
        <yuval.shaia@oracle.com>, <idan.brown@oracle.com>
Subject: Re: [PATCH] net: dev_forward_skb(): Scrub packet's per-netns info
 only when crossing netns
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


----- daniel@iogearbox.net wrote:

> On 03/15/2018 10:21 AM, Shmulik Ladkani wrote:
> > Regarding the premise of this commit, this "reduces" the
> > ipvs/orphan/mark scrubbing in the following *non* xnet situations:
> >=20
> >  1. mac2vlan port xmit to other macvlan ports in Bridge Mode
> >  2. similarly for ipvlan
> >  3. veth xmit
> >  4. l2tp_eth_dev_recv
> >  5. bpf redirect/clone_redirect ingress actions
> >=20
> > Regarding l2tp recv, this commit seems to align the srubbing
> behavior
> > with ip tunnels (full scrub only if crossing netns, see
> ip_tunnel_rcv).
> >=20
> > Regarding veth xmit, it does makes sense to preserve the fields if
> not
> > crossing netns. This is also the case when one uses tc mirred.
> >=20
> > Regarding bpf redirect, well, it depends on the expectations of each
> bpf
> > program.
> > I'd argue that preserving the fields (at least the mark field) in
> the
> > *non* xnet makes sense and provides more information and therefore
> more
> > capabilities; Alas this might change behavior already being relied
> on.
> >=20
> > Maybe Daniel can comment on the matter.
>=20
> Overall I think it might be nice to not need scrubbing skb in such
> cases,
> although my concern would be that this has potential to break
> existing
> setups when they would expect mark being zero on other veth peer in
> any
> case since it's the behavior for a long time already. The safer
> option
> would be to have some sort of explicit opt-in e.g. on link creation to
> let
> the skb->mark pass through unscrubbed. This would definitely be a
> useful
> option e.g. when mark is set in the netns facing veth via
> clsact/egress
> on xmit and when the container is unprivileged anyway.
>=20
> Thanks,
> Daniel

I see your point in regards to backwards comparability.
However, not scrubbing skb when it cross netns via some kernel functions co=
mpared to
others is basically a bug which could easily break with a little bit of mor=
e refactoring.
Therefore, it seems a bit weird to me to from now on, we will force
every user on link creation to consider that once there was a bug leading
to this weird behavior on specific netdevs.
Thus, I suggest to maybe control this via a global /proc/sys/net file inste=
ad.

-Liran