Subject: Re: Q950 (Re: 1.2-current machine status update)
To: Colin Wood <ender@is.rice.edu>
From: Jonathan Lennox <lennox@cs.columbia.edu>
List: port-mac68k
Date: 10/23/1996 02:42:23
On Tue, October 22 1996, "Colin Wood" wrote to "Yasuhiro Mitsui, ender@is.rice.edu, port-mac68k@NetBSD.ORG" saying:

> Ugh.  What is it about the Quadra 950 that makes it so different from 
> earlier Quadras?  You would think that they would simply reuse the Q900 
> design rather than introducing more confusion.

They did.  According to the Developer Note on the Quadra 950, the 950 is
essentially just a speeded-up 900.  The 900, however, is apparently one
bizarre beast.

According to the 900's Developer Note, the 900 (and thus the 950) use the
same IOPs (I/O Processors) as the IIfx: one that controls the SWIM and ADB,
and another that controls the serial ports.  (I think I saw something that
implies that the latter can be put into a compatibility mode, though, so
hopefully it's only for the ADB that the IOPs will be a problem.)

There are also some other weird chips mentioned, but I don't know how
different they are from a programmer's point of view -- I don't know enough
about how these things work.  There's a map of the I/O space provided, if
that'll help anyone else out there.  (The Developer's Notes are in
<ftp://ftpdev.info.apple.com/Developer_Services/Technical_Documentation/
Developer_Notes/Macintosh_CPUs_-_68K_Desktop/> for those interested.)

So...it looks like ADB on the 900/950 will be a bit of a bear, although
presumably if we can get the IOPs working it's a win in the long run over
the direct CPU control of these things.  Is there anything I can do to help
anyone figure out anything about these chips, or to test anything else out
on my 950?

-- 
Jonathan Lennox
lennox@cs.columbia.edu