Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!mips!spool.mu.edu!think.com!snorkelwacker.mit.edu!stanford.edu!neon.Stanford.EDU!torrie
From: torrie@cs.stanford.edu (Evan Torrie)
Newsgroups: comp.sys.mac.hardware
Subject: Re: 8-24GC and System 7.0
Message-ID: <1991Jun6.170950.18129@neon.Stanford.EDU>
Date: 6 Jun 91 17:09:50 GMT
References: <1991May22.153034.1227@bwdls61.bnr.ca> <1991May23.004756.15977@terminator.cc.umich.edu> <17@fleet.UUCP> <1991Jun5.174223.4786@palace.uucp>
Sender: torrie@neon.Stanford.EDU (Evan James Torrie)
Organization: Computer Science Department, Stanford University, Ca , USA
Lines: 36

dmt@palace.uucp (Darryl Trujillo) writes:

>mel@fleet.UUCP (mel) writes:

>>According to a message I read on Applelink the 8-24GC Software drivers
>>won't be available to work under 7.0 in accelerated mode until the
>>FALL quarter. 

>I think the main problem here is that the INIT is not really a driver
>but a rewrite of quickdraw.  From how I understood it, aside from caching
>GWorlds and other information, the processor on the GC executes Quickdraw
>primitives in parallel witht he CPU.  Do make this work they basically
>re-wrote quickdraw in C, then compiled it for the on-board processor.
>The INIT downloads this code onto the board.  This is a non-trivial
>task;  much more difficult then messing with driver code.

  Actually, a big problem is the use of virtual memory in System 7.
Often, you'll have pixmaps and things stored in main memory, which you
want to draw into with QuickDraw.  Now, it's quite likely that these
pixmaps will actually be paged out to disk as part of the normal
virtual memory mechanism.  
  However, the 29K CPU on the GC card has no way of knowing whether
this is the case or not.  It, somehow, has to interrogate the 68030 to
find out whether the pages are actually in memory or not.  If not,
they have to be brought in, and guaranteed not to move, before the 29K
can start executing those QuickDraw commands.  I think it's possible
to do this (just by a quick glance at the Memory Manager chapter of IM
VI), but I'm not surprised it's taking a while..

(I think Paul Campbell from Taniwha could give a better explanation
of this, since he's well-versed in designing graphics cards).
  
-- 
------------------------------------------------------------------------------
Evan Torrie.  Stanford University, Class of 199?       torrie@cs.stanford.edu   
Fame, fame, fame...  What's it good for?  Ab-so-lute-ly nothing