New ICMP limits
James
james at towardex.com
Wed Dec 8 20:02:54 PST 2004
On Wed, Dec 08, 2004 at 06:21:07PM -0800, Bruce M Simpson wrote:
> On Wed, Dec 08, 2004 at 03:53:07PM +0100, Andre Oppermann wrote:
> > I'll take care of this but I'm busy right now. Look into it later this week.
>
> Thanks for looking into this, this is one of the items which came up on
> the TODO lists of three separate projects (TowardEX's, XORP's, and the
> Network Junta's). If you aren't able to look at it let us know so someone
> else can step up to the mic.
>
> Of course, the sooner we can remove ARP's special meaning from RTF_REJECT,
> the better - that would let us implement RTF_REJECT in the fastforwarding
> path without further worry.
When we have routing table cleaned up (e.g. remove arp off of it), I'll look
into getting out some patch for installing /32 host routes for all
receive-adjacent addresses. This way we don't have to run a hash lookup at
ip_fastforward() to find out whether address belongs to us. We can simply
either route it to lo0 as a receive-path (like in Cisco GSR/7500 rcvpath and
Juniper loopback path) or send it to a separate input handling routine with
packet filtering before returning to ip_input.
Unless ofcourse, someone already has it made up and ready to go -- in that
case there is no need ;)
Thanks,
-J
--
James Jun TowardEX Technologies, Inc.
Technical Lead Boston IPv4/IPv6 Web Hosting, Colocation and
james at towardex.com Network design/consulting & configuration services
cell: 1(978)-394-2867 web: http://www.towardex.com , noc: www.twdx.net
More information about the freebsd-net
mailing list