Path: utzoo!attcan!uunet!spool2.mu.edu!samsung!rex!uflorida!gatech!ncsuvx!news
From: kdarling@hobbes.ncsu.edu (Kevin Darling)
Newsgroups: comp.os.os9
Subject: Re: GIME & IRQ
Message-ID: <1991Jan28.023048.24497@ncsuvx.ncsu.edu>
Date: 28 Jan 91 02:30:48 GMT
References: <1991Jan27.011002.21385@lsuc.on.ca>
Sender: news@ncsuvx.ncsu.edu (USENET News System)
Distribution: na
Organization: NCSU Computing Center
Lines: 16

jimomura@lsuc.on.ca (Jim Omura) asks:
> Since the problem was in the GIME chip, I was wondering if
> anybody knew whether it was necessary to have the diode fix
> with the new GIME chip?

Hi Jim! Unless you're using a hacked clock module (perhaps Bruce's),
then yes, it is necessary. The GIME interrupt internals haven't changed.

Clarification: the problem is really that the software expected
different hardware interaction.  Apparently the interrupt hardware
was changed shortly before CC3 production started and yep, as
Murphy's Law dictates <g>, the info didn't get to the programmers.

Basically, the GIME acts as an edge-triggered device.  The stock L-II
interrupt routines expected the more common level-sensitive type.
 best - kevin <kdarling@catt.ncsu.edu>