IPv4-mapped IPv6 addresses

Grigory Kljuchnikov grn@ispras.ru
Thu, 21 Dec 2000 14:25:11 +0300 (MSK)


Hello,

We have a project with SIIT, and have had the problem with IPv4-mapped
addresses. We use FreeBSD with KAME, and there is an implementation
of using IPv4-mapped addresses in KAME as documented in RFC2553
section 3.7. Therefore we use addresses like ::ffff:ffff:a.b.c.d
instead ::ffff:a.b.c.d for our SIIT implementation.

Grigory.

P.S. The origin message from KAME developer is:

------------------------------------
Date: Wed, 29 Nov 2000 22:05:49 +0900
From: itojun@iijlab.net
Reply-To: snap-users@kame.net
To: User Dimka <dimka@ispras.ru>
Cc: snap-users@kame.net
Subject: (KAME-snap 3684) Re: Problem with my pseudo-iface and IPv6(FreeBSD)


>>         ::ffff:xxxx:xxxx is used for very specific purposes in IPv6
>>         specification.  you should not use it for other purposes.
>But this is precise my case: I know for what this addresses is used and
>I'm attempting playing with siit - so...
>Tell me, please, what do KAME stack  with  ::ffff:xxxx:xxxx ?
>Thanks anyway.

        as documented in INSTALL, KAME does not support SIIT environment.
        the address (::ffff:xxxx:xxxx) is used as documented in RFC2553
        section 3.7.

itojun
-----------------------------------------------

On Wed, 20 Dec 2000, Julio Baixauli wrote:

> 	Are you sure the communication is made in IPv6? The kernel handles
> these addresses like it would be IPv4 addresses (...I think). How you
> can skip this default conduct?
> 
> 	In my test program I make:
> 
> socket() with PF_INET6 & SOCK_STREAM
> connect() to ::ffff:a.b.c.d
> 
> ...and I can see, at least, ARP packets for a.b.c.d (if any route is
> found in the route table). This implies that the connection will be made
> in IPv4, isn't it?
> 
> 	Could you send me any code that can clarify me?	
> 
> 	Thank you very much!!
> 
> -- 
> ********************************************
> 
> 	Julio Baixauli Garreta
> 	baixauli@mat.upc.es
> 
> ********************************************
> 


Grigory Klyuchnikov, System Engineer,
Institute for System Programming 
Russian Academy of Sciences