[quagga-dev 9073] Re: Help with specifying both gateway and ifindex in IPv4

Joakim Tjernlund joakim.tjernlund at transmode.se
Thu Feb 16 07:29:13 GMT 2012


David Lamparter <equinox at diac24.net> wrote on 2012/02/16 06:51:25:

> From: David Lamparter <equinox at diac24.net>
> To: Joakim Tjernlund <joakim.tjernlund at transmode.se>, Quagga development list <quagga-dev at lists.quagga.net>
> Date: 2012/02/16 06:51
> Subject: Re: [quagga-dev 9058] Re: Help with specifying both gateway and ifindex in IPv4
>
> On Mon, Feb 13, 2012 at 01:17:34PM +0100, Joakim Tjernlund wrote:
> > > > >> Is the limitation to either a nexthop or an ifindex in IPv4 by design?
> > > > >> Is it a bug?  If the latter, could somebody point me at the point in the
> > > > >> code where to fix it?
> > > >
> > > > > It is a case of no one have impl. yet. There is
> > > > > ZEBRA_NEXTHOP_IPV4_IFINDEX but the code that impl. is yet to be
> > > > > written. It is easy to impl. though, I did it for OSPF
> > > >
> > > > Is the code in RE-testing-0.99?
> > >
> > > No, it is not public as I haven't bothered to clean it up and submit it.
> >
> > BTW, here is the zebra part, no idea if it still applies:
>
> Cool - out of genuine coincidence I just added a counterpart to
> zsend_ipv4_nexthop_lookup (it's in the pimd merge). (Will push to a
> separate branch of master in a bit)

Why separate? If it is a new feature, nothing old should break.

>
> Either way I'm interested in this as well since I have the same problem
> with OLSR. Kernel code needs to be added to push gw+ifindex, I presume?
> Or does that magically work as-is?

Should work as is, was some time ago I added this and someone else tested it.
Nothing broke though :)

 Jocke




More information about the Quagga-dev mailing list