Re: [ECOS] TCP/IP checksum routine performance

[Grant Edwards <grante at visi dot com>]

On Fri, Apr 28, 2000 at 10:18:44AM -0600, Gary Thomas wrote:

> > Yea, I've got a test in there now.  I'm pretty sure that none
> > of the standard protocol code is going to try to split a word.
> > IP and TCP headers are always in their own mbufs, and I don't
> > think that TCP data segments are going to get split at odd byte
> > boundaries.  I also can't see any way that it could be made to
> > happen by user application code.  But, somebody went to a fair
> > amount of trouble to make the current routine handle it, so I'm
> > a bit uneasy.
> 
> I also am not convinced this will ever happen, but it's best to
> err on the safe side.

I keep thinking it must have happened to somebody or they
wouldn't have put code in there to handle it.  An assert would
be almost good enough, but on an embedded system with no
console output located at a customer site in Lower Nowhere it's
darned hard to figure out exatly why the box keeps resetting
itself (and of course it won't ever do it back at the factory).

> I have already made sure that the ARM GCC maintainers are aware
> of your findings.

OK, I'll push that to the bottom of the todo list.

-- 
Grant Edwards
grante@visi.com