Re: [PATCH nf-next] netfilter: add support for matching IPv4 options

[Pablo Neira Ayuso <pablo@netfilter.org>]

On Sat, Jun 01, 2019 at 11:04:29AM -0400, Stephen Suryaputra wrote:
> On Sat, Jun 01, 2019 at 02:22:30AM +0200, Pablo Neira Ayuso wrote:
> > > It is the same as the IPv6 one. The offset returned is the offset to the
> > > specific option (target) or the byte beyond the options if the target
> > > isn't specified (< 0).
> > 
> > Thanks for explaining. So you are using ipv6_find_hdr() as reference,
> > but not sure this offset parameter is useful for this patchset since
> > this is always set to zero, do you have plans to use this in a follow
> > up patchset?
> 
> I developed this patchset to suit my employer needs and there is no plan
> for a follow up patchset, however I think non-zero offset might be useful
> in the future for tunneled packets.

For tunneled traffic, we can store the network offset in the
nft_pktinfo object. Then, add a new extension to update this network
offset to point to the network offset inside the tunnel header, and
use this pkt->network_offset everywhere.

I think this new IPv4 options extension should use priv->offset to
match fields inside the IPv4 option specifically, just like in the
IPv6 extensions and TCP options do. If you look on how the
priv->offset is used in the existing code, this offset points to
values that the specific option field conveys.

Thanks.