Subject: Re: CVS commit: syssrc
To: None <thorpej@nas.nasa.gov>
From: Christoph Badura <bad@ora.de>
List: tech-net
Date: 05/30/1999 20:34:00
Jason Thorpe wrote:

> On Sat, 29 May 1999 17:39:08 -0700 (PDT) 
>  Christoph Badura <bad@netbsd.org> wrote:

>  > Modified Files:
>  > 	syssrc/sys/net: if_token.h if_tokensubr.c
>  > 	syssrc/sys/netinet: if_arp.c
>  > 
>  > Log Message:
>  > Fix thinko of mine in previous.  The source route info is not at m->m_data
>  > after various m_adj()s have been done.  Kludge around this with a cheesy
>  > macro that knows where the drivers put the mac header in the first mbuf.
>  > 
>  > XXX There should be a better way to do this.

> There is... make ETHERTYPE_ARP packets NOT m_adj() the header/routing
> info.  Then let the ARP code adjust it as necessary.

Ah, but delaying the m_adj() won't make a difference.  The handling of
the Token-Ring source routing info is extremely kludgy.  You don't
really want to know.  I think we need a better interface for this in the
long term.

-- 
Christoph Badura					www.netbsd.org

	Anything that can be done in O(N) can be done in O(N^2).
	-- Ralf Schuettau (after looking at a particular piece of code)