On 2026-03-30, Polarian <[email protected]> wrote: > Hey, > >> OpenBSD does not support to use a IPv6 nexthop for IPv4 routes at the >> moment. It is something I want to fix at some point but it is not >> high on my todo list. You can inject routes like this into the kernel >> but they do not work because the way ifp_output is called makes too >> many assumptions. > > There seems to be varying responses to this when someone on IRC asked > on mas [1].
being able to run the 'route add' command does not contradict what Claudio said :) >> Assume you have a /29 from the ISP that is routed to you via a >> separate /32 link? There are some games you can play to maximise >> use of this before we get v6 nexthop. >> >> If you're using nat(/rdr) for some hosts, you can use the network >> and broadcast addresses for these. >> >> Slightly dirtier but if you don't need to contact addresses each >> side of your allocation (i.e. other customers of your ISP) you >> can configure a wider subnet than you really have and make use >> of net/bcast addresses directly. >> >> Or you can configure a separate rfc1918 network on a normal >> ethernet interface, add a static route for the /32 pointing over >> that, and configure the routable address on a dummy interface >> (lo1 or vether or something would do). > > I wanted to avoid nat/rdr, however in the end binat-to is just so > simple to use I question why I even went through the debugging process. > > Another win for pf being so simple, yet so powerful. It is pretty easy, as long as the protocols you're using play ball. (If not then keep the "add a static route and configure on a dummy interface" option in mind, it's very handy to have in the toolbox). -- Please keep replies on the mailing list.
