From: Theo Buehler <tb@theobuehler.org>
Subject: Re: bgpd: fix nexthop qualify via bgp
To: tech@openbsd.org
Date: Wed, 4 Sep 2024 16:20:50 +0200

On Wed, Sep 04, 2024 at 04:01:14PM +0200, Claudio Jeker wrote:
> Qualifying nexthops via BGP is currently busted because the nexthops are
> not rechecked once a new BGP route is added. We need to keep track of
> nexthops both on inserts (kroute_insert) and on change (krX_change but
> only for AID_INET and AID_INET6 -- no nexthops in the other tables)
> 
> We don't want to always validate all nexthops so only do this if 'nexthop
> qualify via bgp' is enabled. For route changes we can depend on the
> F_NEXTHOP flag.

Thanks for the clear summary. With these explanations, it all makes sense.

ok tb