Re: [PATCH net-next 1/3] udp_tunnel: allow to turn off path mtu discovery on encap sockets

From: Florian Westphal <fw@strlen.de>
To: David Ahern <dsahern@gmail.com>
Cc: Stefano Brivio <sbrivio@redhat.com>,
	Florian Westphal <fw@strlen.de>,
	netdev@vger.kernel.org, aconole@redhat.com
Subject: Re: [PATCH net-next 1/3] udp_tunnel: allow to turn off path mtu discovery on encap sockets
Date: Fri, 17 Jul 2020 20:43:05 +0200	[thread overview]
Message-ID: <20200717184305.GV32005@breakpoint.cc> (raw)
In-Reply-To: <89e5ec7b-845f-ab23-5043-73e797a29a14@gmail.com>

David Ahern <dsahern@gmail.com> wrote:
> On 7/17/20 6:27 AM, Stefano Brivio wrote:
> >> Every type of bridge port that needs to add additional header on egress
> >> has this problem in the bridge scenario once the peer of the IP tunnel
> >> signals a PMTU event.
> > 
> > Yes :(
> > 
> 
> The vxlan/tunnel device knows it is a bridge port, and it knows it is
> going to push a udp and ip{v6} header. So why not use that information
> in setting / updating the MTU? That's what I was getting at on Monday
> with my comment about lwtunnel_headroom equivalent.

What action should be taken in the vxlan driver?  Say, here:

static inline void skb_dst_update_pmtu_no_confirm(struct sk_buff *skb,
	u32 mtu)
{
 struct dst_entry *dst = skb_dst(skb);

 if (dst && dst->ops->update_pmtu)
    dst->ops->update_pmtu(dst, NULL, skb, mtu, false);
 else
    /* ??? HERE */
 }

We hit the (non-existent) else branch as skb has no dst entry.