Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!usc!pollux.usc.edu!addison
From: addison@pollux.usc.edu (Richard Addison)
Newsgroups: comp.sys.amiga.tech
Subject: Re: MutualExclude for Gadgets
Keywords: Amiga MutualExclude
Message-ID: <20076@usc.edu>
Date: 23 Sep 89 21:29:55 GMT
References: <1849@cbnewsd.ATT.COM> <125013@sun.Eng.Sun.COM> <7964@cbmvax.UUCP> <125083@sun.Eng.Sun.COM> <7980@cbmvax.UUCP>
Sender: news@usc.edu
Reply-To: addison@pollux.usc.edu (Richard Addison)
Distribution: usa
Organization: University of Southern California, Los Angeles, CA
Lines: 28

In article <7980@cbmvax.UUCP> jesup@cbmvax.UUCP (Randell Jesup) writes:
>	Looking at *0?  Must we add 0-detect circuits in order to not
>"support" it?

Could do this with the MMU (wish I had one).  Trap all access to that page
of memory, and continue if it is an allowed read (like, oh I don't know, um,
maybe location 4?) but complain if it is a read from location 0.

Yeah, I know it slows things down, but it's a debugging tool.

>	Sorry if I sound testy, I just found some major stuff that was playing
>evil games and calling internal BCPL routines that shouldn't have been. Grrrr.

Gee, I was wondering when you guys would notice.  (-;

I have a question that is unrelated to the subject line, but this remark
reminded me:

I've noticed that some of the ROM libraries (at least those routines I've
traced to find out what is really going on) push their parameters from the
registers onto the stack.  Considering that at least one C compiler for
the Amiga eliminates the need to have stub routines, wouldn't it be nice
if the ROM routines themselves could use the registerized parameters
directly?  Like maybe in 1.4?

Just a thought.

Richard Addison