Subject: Re: UDP checksum trouble in -current
To: None <>
From: Miles Nordin <carton@Ivy.NET>
List: tech-net
Date: 01/21/2005 20:19:24
Content-Type: text/plain; charset=US-ASCII

>>>>> "was" == William Allen Simpson <> writes:

   was> Both IP and TCP checksums MUST be performed.

uh, yeah, pull the handle and the RFC talks, but this is on the
loopback interface.  What are you checking for, memory corruption?

If you think no checksums on loopback is a bad idea, then do checksums
on loopback.  But if you're going to omit checksums for loopback, omit
them.  ``because the RFC says'' or ``because the person reading
tcpdump might forget he set a sysctl for no checksums'' are arguments
for backing out the no-checksums-on-loopback change.  You may or may
not find them convincing arguments.  But they're not arguments for
being schitzophrenic about whether you're doing checksums or not.  You
have to make up your mind.

As long as checksums are verified for packets that arrive on a real
interface but are destined to a loopback alias, I don't see the big
deal either way.

Content-Type: application/pgp-signature
Content-Transfer-Encoding: 7bit

Version: GnuPG v1.2.6 (NetBSD)