Hi all,
I try to implement an IPv4 in IPv6 encapsulator/decapsulator on my own.
I know that ip4ip6 tunnel exists in Linux but I need to do it for some
special features!
So in order to do that, I use libipq (which is quite obsolete I know).
The principle is the following:
IPv4/IPv6 IPv4/IPv6
Host A ---------------> Host B
@IPv4A @IPv4B
@IPv6A @IPv6B
Host A pings host B in IPv4 (ping @IPv4B).
With libipq, I encapsulate the V4 ping in V6 which gives the following
packet: [@IPv6A @IPv6B][@IPv4A @IPv4B][ICMPv4].
Host B receives the IPv4 in IPv6 ping. I decapsulate the IPv6 header
with libipq. It gives me the following IPv4 echo request: [@IPv4A
@IPv4B][ICMPv4].
Then with a raw socket where sin->addr is set to @IPv4B, I "send" the
IPv4 echo request in order to have the right echo reply.
The kernel produces me an echo reply but it's where the problem begins.
The echo reply has as source addr @IPv4B which is logical but also
@IPv4B as destination address wereas I'm waiting @IPv4A to reply to the
ping. I dont understand why. I think it has something to deal with the
fact that the IPv4 echo request is "generated localy" when I decaspulate
the IPv6 header but I dont get it... Note also that I just
decapsulate/encapsulate the header, I never manipulate src or dst address.
If you have any tips for solving that, that would be great. If i post on
netfilter mailing list, It's because I think someone with a great
knowledge on the netfilter chain can easily indentify the problem. If
I'm in the wrong place, sorry...
Thank you,
cheers.
Tanguy
--
To unsubscribe from this list: send the line "unsubscribe netfilter" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
