On 11/04/2017 22:47, Robert Elz wrote:
>     Date:        Tue, 11 Apr 2017 19:25:33 +0100
>     From:        Roy Marples <[email protected]>
>     Message-ID:  <[email protected]>
> 
>   | I didn't see any other RTM_* consumers in our tree.
> 
> sbin/routed and usr.sbin/route6d are likely to benefit.

These need to be re-worked by someone who uses them.
There is a bit of code which does <= RTM_CHANGE which will break at some
point if we re-number it.

I've left them alone.

> There are also a whole bunch of processes that write to the routing socket,
> then read it, to verify that the message they sent was received - most of
> those are essentially one-off uses (do something and exit, like sbin/route)
> but a couple (rarpd ldpd - more? didn't see any, but...) are daemons and
> might possibly benefit.

Fixed rarpd and ldpd.

Roy

Reply via email to