[3292] in SIPB_Linux_Development
Re: Netword card problems
daemon@ATHENA.MIT.EDU (Derek Atkins)
Sun Jan 7 19:31:07 2001
To: Alex Coventry <alex_c@MIT.EDU>
Cc: linux-dev@MIT.EDU
From: Derek Atkins <warlord@MIT.EDU>
Date: 07 Jan 2001 19:31:00 -0500
In-Reply-To: Alex Coventry's message of "Sun, 7 Jan 2001 15:26:42 -0500"
Message-Id: <sjmbstiq3zv.fsf@rcn.ihtfp.org>
Alex Coventry <alex_c@MIT.EDU> writes:
> > I suspect that the only changes we'll need are the same changes that
> > Seph made, namely changing the 3c905C to use the 3c90x driver. I
> > suspect that you just need to change the pcitable that gets installed
> > on the install server in .../RedHat/instimage/modules to match the
> > changes made in the install disk.
>
> Yeah, I think you're right, but it'd still be prudent to test it in a
> separate NFS-served directory before actually putting it on the
> server-proper, imo.
Except I'm not convinced I'm right -- I think that not only do we need
to change this, but we also need to add the 3c90x module to the
modules.cgz that sits on the server. Unfortuately this is a DIFFERENT
modules.cgz, so we can't just copy it from trees/network. Uggh.
> I had to build kudzu in order to get a seperate version of the
> second-stage installer that I could test from the command line. It was
> a relatively simple matter. Is it possible that we could just drop that
> in, or are these horrid module dependancies likely to bite us, in that
> case, too?
I was actually referring to a change in the Kudzu RPM
(kudzu-0.36-2.i386.rpm, or whatever the most recent version is). The
problem is that the installed machine needs to know about the new
driver.
I think this warrents:
a) a check on RH7 to see if this behavior has changed, and if not
b) a bug report to Red Hat.
> Alex.
-derek
--
Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
Member, MIT Student Information Processing Board (SIPB)
URL: http://web.mit.edu/warlord/ PP-ASEL-IA N1NWH
warlord@MIT.EDU PGP key available
