Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!njin!princeton!phoenix!blackman
From: blackman@phoenix.Princeton.EDU (Scott Michael Blackman)
Newsgroups: comp.sys.apple
Subject: Re: Hyper C
Message-ID: <10237@phoenix.Princeton.EDU>
Date: 4 Sep 89 03:04:25 GMT
Reply-To: blackman@phoenix.Princeton.EDU (Scott Michael Blackman)
Organization: Princeton University, NJ
Lines: 49

Here's a technical tidbit.  Working a little randomly tonight, I decided
to try to fix Hyper C to work on the IIgs.  As some of us know, it won't
boot on the IIgs, although it works fine on an enhanced //e.  Here's the
progress I made, and if anybody wants to pick up where I left off, fine.
I've identified the problem:

The Hyper C boot routine immediately loads $D000-FFFF into
memory.  If you can interrupt the boot process at any point
in the three-track load (go into the control panel, use a
WildCard, etc.), then $F8CD holds the JMP instruction to
coldstart Hyper C after the boot.  Redirecting this to a
routine that sets ROM and jumps to the monitor, you can then
alter the BRK vector and routine until it suits you.

The problem lies with the IIgs BRK handlers.  Hyper C uses a
subroutine called with a BRK, and the vector it uses is
$FFFE-FFFF.  (It is initialized to $D0D4, and the BRK
handler itself is at $D0E1)  In the IIgs, this vector is
fixed in ROM.  Changing this to another vector has been
fruitless, for one reason or another.  The easiest method,
it seems, would be to fix the IIgs BRK handler to emulate
the old //e handlers and JMP to $D0E1.  Unfortunately, this
hasn't worked so far, and I'm about burned out about it.

At $D0E1, Hyper C expects the stack as follows:

(top)     Status Register
          Low byte of BRK address plus two (?)
          High byte
          .
          .
(bottom)  .

My attempts have been to use the vector at $3F0, and set up
the stack to imitate the $FFFE-FFFF stage.  The calling
address is in $3A.3B at that point.

Another strategy would be to use the IIgs BRK vector at
$E1/0070.74 and to try to imitate from there.


That's it...Of course, if anybody has distributed the ProDOS version of
it, that is, if we've decided that the copyright has expired 8-), then I
would be delighted if somebody would fill me in.  Thanks for your
attention...

Scott Blackman
blackman@phoenix.princeton.edu
blackman@pucc.bitnet