Subject: Re: SLOTMAN??? on Q650
To: CTAPYXA" , "MacBSD Newsletter <port-mac68k@NetBSD.ORG>
From: None <wb2oyc@bellatlantic.net>
List: port-mac68k
Date: 01/10/1998 07:03:19
> Anybody with similar setup runs MacBSD successfully?
>Please respond.
Yuri,
     Well, I suspect that Paul will comment on what you've provided,
and it'll be very helpful info.  Yes, I run SLOTMAN using the same
setup, except not on a Q650.  In my case the Mac is a IIci and with
the same video card as you are using.  Mine may be a different Rev
but I do have several of them.

    The fault you are seeing seems eerily familiar to me, as I was
recently dealing with very similar kernel boot faults using it on
a card (Raster Ops) that has some significant differences from many
of the others.  Paul (SLOTMAN's author) resolved it and added a new
supported card very quickly.  I sent him one of mine to work with 
and now I can use it and run X on it too!

    But, we learned an important lesson from that, and that is, I
have two of those Raster Ops cards.  One is Rev 1.1 and the other
is 1.3 on the ROM code.  I had sent Paul the 1.3.  The 1.1 still
has a problem while he was able to resolve the issues completely
on the other!  The differences between the two revisions are in
this case, that significant.

    I'm wondering if what you are experiencing may be similar to
that situation, as I also have several Mac II Hi Res cards which
work properly with SLOTMAN, yet your's does not.  Also, one of the
differences with SLOTMAN is that it attempts to build an interface
between NetBSD's video calls and the on card ROM based code.  It
uses that ROM code to carry out the work associated with the call.
SLOTMAN interfaces that ROM code with the kernel.

    The amazing thing is, the differences between them, AND, that
although they may DO or NOT DO some things in response to requests
from MacOS, or do things very differently even between revisions 
of the on card ROM code of the same model card (like in my case
with the Raster Ops 264) that they work at all!

    This must drive the kernel hackers nuts!  I mean, when Paul 
and I got into trying to trouble-shoot via email, the differences
between these two REVISIONS of the same model card; the one I had
sent to him, he had working in short order, yet the one I had 
would not successfully boot using the same special kernel he had
built that fixed it.  We learned it came down to the ROM Rev 
between the two; 1.1 and 1.3.

    For those that may be specifically interested in  this, the 
Raster Ops 264 now works with SLOTMAN (as of 12/25/97 #175 I think
it is) IF the ROM's show Rev 1.3.

Paul