From mboxrd@z Thu Jan  1 00:00:00 1970
From: Liran Alon <liran.alon@oracle.com>
Subject: Re: [PATCH] net: dev_forward_skb(): Scrub packet's per-netns info
 only when crossing netns
Date: Thu, 15 Mar 2018 08:01:03 -0700 (PDT)
Message-ID: <58e67195-56f6-4d01-8747-f8322a382358@default>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Cc: <netdev@vger.kernel.org>, <shmulik.ladkani@gmail.com>,
        <daniel@iogearbox.net>, <davem@davemloft.net>,
        <linux-kernel@vger.kernel.org>, <yuval.shaia@oracle.com>,
        <idan.brown@oracle.com>
To: <mrv@mojatatu.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
Sender: linux-kernel-owner@vger.kernel.org
List-Id: netdev.vger.kernel.org


----- mrv@mojatatu.com wrote:

> Liran Alon <liran.alon@oracle.com> writes:
>=20
>=20
> [...]
>=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 compared to
> > others is basically a bug which could easily break with a little bit
> of more 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 instead.
>=20
> One valid use case could be preserving a source namespace nsid in
> skb->mark when a packet crosses netns.

Before and after this commit, veth peers that crosses netns zero skb->mark.
Therefore, what you suggest is a new feature unrelated to the
issue fixed by this commit.

I still think that default behavior should be to zero skb->mark only when s=
kb
cross netdevs in different netns. And for backwards comparability, we can
consider adding a /proc/sys/net/core file to let dev_forward_skb() to alway=
s
scrub packet regardless if it crosses netdevs or not.