Re: [patch net-next 2/3] netfilter: ip6_tables: use reasm skb for matching

[Patrick McHardy <kaber@trash.net>]

On Tue, Nov 05, 2013 at 09:55:20PM +0100, Jiri Pirko wrote:
> Tue, Nov 05, 2013 at 07:16:33PM CET, kaber@trash.net wrote:
>
> >> I'm a bit lost. What "nfct_frag" are you reffering to here?
> >
> >I meant nfct_reasm of course.
> 
> The patch is not moving this to struct sk_buff. It is already there.

Right again, sorry, I was replying to Florian's mail without the
quoted patch, so I seem to have mixed up things in between :)

> >> >So if someone wants to change this, simply *only* pass the reassembled
> >> >packet through the netfilter hooks and drop the fragments, as in IPv4.
> >> 
> >> This is unfortunatelly not possible because in forwarding use case, the
> >> fragments have to be send out as they come in.
> >
> >No, the IPv6 NAT patches fixed that, we still do proper refragmentation
> >and we still respect the original fragment sizes, thus are not responsible
> >for potentially exceeding the PMTU on the following path.
> 
> Ok. So the plan is to remove net/ipv6/netfilter/nf_conntrack_reasm.c
> code entirely and use net/ipv6/reassembly.c code directly from
> nf_defrag_ipv6. This would result in very similar code currently ipv4
> has. 

If its possible to use net/ipv6/reassembly.c directly, even better,
so far I was thinking of just getting rid of the fragment replay,
but you're most likely right that this is the proper way to do it.