On Thu, Jan 7, 2016 at 7:36 AM, Lou Berger <[email protected]> wrote:
>
>
> On 1/7/2016 10:21 AM, Vivek Venkatraman wrote:
> >
> > On Thu, Jan 7, 2016 at 5:48 AM, Lou Berger <[email protected]
> > <mailto:[email protected]>> wrote:
> >
> > Vivek,
> >
> > to close this one off.
> >
> > On 1/5/2016 8:18 AM, Vivek Venkatraman wrote:
> > >
> > >
> > > +#if 0 /* duplicates mpattr
> > included in
> > > bgp_packet.c */
> > > if (p && !(afi == AFI_IP && safi == SAFI_UNICAST))
> > >
> > changed to:
> > if (p && afi == AFI_IP6 && safi == SAFI_UNICAST)
> >
> >
> > Right now, the default-originate command is installed for multiple
> > address families (e.g., multicast). This is also an area where we have
> > done some (minor) changes due to RFC 5549. So, my preference would be
> > to keep the check as it was before.
>
> do you think default originate apply to other safi's such as ENCAP, VPN,
> EVPN and whatever comes next? I'm leaning towards no...
>
I feel the same way. The default_originate command need not be installed
for those AFI/SAFI. But I believe they make sense for the multicast AFs and
others like labeled-unicast (1/4 or 2/4).
> Lou
>
> > The 'mpattrlen_pos' variable can of course be moved within this block.
>
>
> >
> >
> > > {
> > > + size_t mpattrlen_pos = 0;
> > > mpattrlen_pos = bgp_packet_mpattr_start(s, afi, safi,
> > attr);
> > > bgp_packet_mpattr_prefix(s, afi, safi, p, prd, tag);
> > > bgp_packet_mpattr_end(s, mpattrlen_pos);
> > > }
> > > -
> > > +#endif
> > >
> > >
> > > This part of the code (which has been placed under "#if 0") is
> > used by
> > > the default originate code. I'm not sure how that will work
> > after this
> > > change. Can you please check?
> >
> > Good catch!
> >
> > Lou
> >
> >
>
>
>
_______________________________________________
Quagga-dev mailing list
[email protected]
https://lists.quagga.net/mailman/listinfo/quagga-dev
