(usagi-users 03861) Re: Bug related to advanced IPv6 Socket API

"Boutin Maël" <[email protected]> Fri, 22 Jun 2007 17:01:19 +0200
Newsgroups gmane.linux.ipv6.usagi.users
Message-ID <[email protected]>
Ok, i did not subscribe to this mailing list ;)

Regards,

Maël

On 6/22/07, Romain KUNTZ <[email protected]> wrote:
> Hi Mael,
>
> This bug has been reported by Masahide Nakamura from USAGI on the
> MIPL ML. He has provided a patch here:
> http://www.mobile-ipv6.org/pipermail/mipl/2007-May/006853.html
>
> You'll find a similar patch on this list in a mail I sent today with
> several UMIP patches.
>
> Cheers,
> romain
>
> On 2007/06/22, at 15:08, Boutin Maël wrote:
>
> > Hi,
> >
> > I have found a bug related to the Advanced Ipv6 Socket API support :
> > If the system does not provide the needed functions then the files in
> > libmissing are compiled and everything is fine. However, on FC6 the
> > functions are defined (more recent versions of libc ?) and therefore
> > mipl uses these ones.
> >
> > With FC6 functions Compilation is OK, but during the execution the
> > problem appears :
> > The MN sends a BU to the HA, The HA receives the BU and generates a BA
> > with a good status. Debug messages returns that the HA has sent a BA
> > however the BA never leaves the HA.
> > After having browsed the code it appears that the problem is related
> > to inet6_rth_init() which is called by mh_send(). This function
> > returns NULL while it does not return NULL with libmissing functions.
> >
> > I didnot have time to try to fix this, i only disabled the check for
> > advanced ipv6 socket api support in configure.ac and makefiles.
> >
> > Regards,
> >
> > Maël Boutin
> >
> >
>
> --
> Romain KUNTZ
> [email protected]
> Louis Pasteur University - Networks and Protocols Team
>
>
>
>
>