Subject: Re: Quadra 610 (and others) internal video
To: None <port-mac68k@NetBSD.ORG>
From: Michael R. Zucca <mrz5149@cs.rit.edu>
List: port-mac68k
Date: 01/21/1997 16:07:01
>I want to address this before I forget (and as you might imagine from the
>lag between the original message an this response, I'm swamped).

Don't worry. I haven't had time to read my mail lately :)

>There is currently a table in machdep.c for the express purpose of making
>some machines work that currently do not.  It is only partly a hack,
>though; please imagine with me a world where the Mac OS is never loaded,
>and the MMU hasn't been enabled.  This world assumes the existence of
>native boot code, and while I won't pretend that it's going to happen any
>time soon, some sort of table to initialize the console may well be
>necessary for machines where we can't find the video start address and
>length through configuration ROMs.  The table that currently exists is
>intended in the long run to be used only in such cases, not in the general
>case.

I can envision a table that perhaps contains the start of video RAM. That makes
some sense. However, you *must* probe for the video RAM size since you can't
know how large video RAM is on a particular machine. On many machines there
are several VRAM configurations and you can't just assume lowest common
denominator every time. So we must probe.

Even if we made a boot loader that was 100% native with boot-sector code,
the ROMs still initialize with the default memory map before we get a chance
to reek any havoc to the MMU. Before we toy with the MMU, then, it is supremely
important to size up the VRAM regardless of how native NetBSD becomes in
the future.

How else can we possibly know the size of VRAM? I'd say the only possible way
is to look at how the ROMs figure this sort of thing out. Maybe there's a
hardware register someplace? Outside of that, we're forced to probe. Tables
are too static.

_______________________________________________________________________
 Michael Zucca - mrz5149@rit.cs.rit.edu - http://www.rit.edu/~mrz5149/
 "I will choose a path that's clear. I will choose Freewill. "
  --Rush, Freewill
_______________________________________________________________________