cvs commit: src/sys/netinet6 in6.h in6_proto.c route6.c

Kevin Oberman oberman at
Wed Apr 25 13:33:40 UTC 2007

> Date: Wed, 25 Apr 2007 10:51:39 +0900
> From: gnn at
> At Tue, 24 Apr 2007 19:43:25 +0200,
> Simon L. Nielsen wrote:
> > > > Can you "MTC" this so we can turn off and on RH0 processing on current
> > > > systems? It would greatly simplify things for me if I didn't have to
> > > > build two kernels and move back and forth to test things.
> > > 
> > > +1 ... also is this going to be in the supported RELENG_X_Y branches as 
> > > well?
> > 
> > Yes, once the foodwork required is done.
> You'll still have to build two kernels, it is just off by default
> completely in HEAD.  The IETF is currently discussing retiring the
> option in which case the code will disappear completely from HEAD.
> Do you really think the option is required in HEAD?

I see RH0 as a very useful feature for troubleshooting in a network, but
I really can't see why a non-routing system should every need or want
it. It looks to me like it should be non-existent when routing is not
enabled and off by default when it is enabled.

That said, in these days of MPLS and wide-area VLANs, the usefulness of
this option is far less than it was in the earlier days of the net and I
can't remember the last time I used the loose-source route capability of
IPv4. I have never used it in IPv6.

If it does live on (in any form), it is important that IPFW2 be able to
filter it properly (ala Bjoren's patches) so it can be limited to
desired connections, typically internal connections.
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at			Phone: +1 510 486-8634
Key fingerprint:059B 2DDF 031C 9BA3 14A4  EADA 927D EBB3 987B 3751
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 224 bytes
Desc: not available
Url :

More information about the cvs-src mailing list