bin/111493: routed doesn't use multicasts for RIPv2 via P2P interfaces

[James Carlson]

The following reply was made to PR bin/111493; it has been noted by GNATS.

From: James Carlson <carlsonj at workingcode.com>
To: Bruce M Simpson <bms at incunabulum.net>
Cc: Vernon Schryver <vjs at calcite.rhyolite.com>, dan at obluda.cz,
        freebsd-gnats-submit at FreeBSD.org
Subject: Re: bin/111493: routed doesn't use multicasts for RIPv2 via P2P interfaces
Date: Fri, 10 Aug 2007 09:15:51 -0400

 Bruce M Simpson writes:
 > I would just suggest that it should default to off to avoid confusing 
 > folk.
 
 Defaulting what to "off?"  The point of the patch (as I understood it)
 was to use multicast on interfaces where the interface flags say
 multicast is supported (IFF_MULTICAST) and where RIP-2 is configured
 to run and multicast hasn't been explicitly disabled.
 
 Why would this work differently than it does on Ethernet?
 
 > If Zebra/Quagga is disallowing multicasted RIP across 
 > non-multicast links, that's another story entirely.
 
 Yes, it is doing that, and so is the proposed routed patch.  For
 Zebra, see:
 
 http://cvs.quagga.net/cgi-bin/viewcvs.cgi/zebra/zebra-cvs/ripd/rip_interface.c?rev=1.1.1.1&content-type=text/vnd.viewcvs-markup
 
 and for Quagga, see:
 
 http://cvs.quagga.net/cgi-bin/viewcvs.cgi/quagga/ripd/rip_interface.c?annotate=1.36#163
 
 The function of interest is rip_request_interface_send().  It checks
 for version 2 and whether the interface supports multicast.  If it
 does, then it calls through to rip_request_send and rip_send_packet --
 which uses the well-known multicast destination.
 
 It doesn't check for point-to-point first.
 
 (Sorry about the lame link to Zebra.  They don't seem to have a web
 interface.)
 
 -- 
 James Carlson         42.703N 71.076W         <carlsonj at workingcode.com>