Path: utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!unix.cis.pitt.edu!dsinc!bagate!cbmvax!jesup
From: jesup@cbmvax.commodore.com (Randell Jesup)
Newsgroups: comp.sys.amiga.games
Subject: Re: Omega Followup II (Just when you thought it was really port
Message-ID: <16746@cbmvax.commodore.com>
Date: 22 Dec 90 02:41:49 GMT
References: <2790@ux.acs.umn.edu> <36439@cup.portal.com>
Reply-To: jesup@cbmvax.commodore.com (Randell Jesup)
Distribution: na
Organization: Commodore, West Chester, PA
Lines: 28

In article <36439@cup.portal.com> Radagast@cup.portal.com (sullivan - segall) writes:
>>Argh! I got omega 1.0 from the nasa site. And, of course, it was a very
>>nice bug-free version - which will not, under any known circumstances,
>>run on a 3000. ARGH!

>Must be something funny about the A3000, because it works fine on my
>A3001 with all caches enabled, and both slow-fast and 32bit fast ram.
>
>Make sure you are using a reasonable stack size.  Otherwise it may be 
>a problem with 2.0?

	I assume the person tried it under 1.3 on the 3000.  In any case, the
most likely reason is using the upper 8 bits of an address for flags or
some such (since fast ram on an a3000 is at 0x07xxxxxx).  The other likely
reason is writing 1 byte past the end of an allocation (like: 
strcpy(malloc(strlen(src)),src).  On a machine with no mem above 0x00ffffff,
this usually doesn't hit anything, since the word following is often
a pointer (often to the next free block).

	If you have them, mungwall and enforcer are great for tracking down
these sorts of problems.

-- 
Randell Jesup, Keeper of AmigaDos, Commodore Engineering.
{uunet|rutgers}!cbmvax!jesup, jesup@cbmvax.commodore.com  BIX: rjesup  
The compiler runs
Like a swift-flowing river
I wait in silence.  (From "The Zen of Programming")  ;-)