Path: utzoo!attcan!uunet!wuarchive!zaphod.mps.ohio-state.edu!uwm.edu!ux1.cso.uiuc.edu!ux1.cso.uiuc.edu!lhummel
From: lhummel@ux1.cso.uiuc.edu
Newsgroups: comp.sys.amiga
Subject: Re: Fix to PDC 3.33
Message-ID: <15500055@ux1.cso.uiuc.edu>
Date: 20 Aug 90 12:39:00 GMT
References: <392@cancol.oz>
Lines: 28
Nf-ID: #R:cancol.oz:392:ux1.cso.uiuc.edu:15500055:000:1454
Nf-From: ux1.cso.uiuc.edu!lhummel    Aug 20 07:39:00 1990


Written by eyal@cancol.oz in comp.sys.amiga:
/* ---------- "Fix to PDC 3.33" ---------- */
> What is the fix that enables PDC 3.33 to create an executable for A68k 
> (Charlie Gibbs Assembler) that works correctly. A friend and I are working on
> a PD C Compiler derived from the Compiler on Fisk Disk 110 and it creates an
> executable for A68k that fails in exactly the same way as for PDC. I would be 
> very grateful if you could let me know what the error is as I have been unable 
> to trace it in our compiler. 

The problem was fixed quite a while back...long before 3.33.  I'm really
surprised you haven't caught it.  If that blatant mistake in handling logical
and/or expressions has eluded you, then I don't see how you're going to get 
past all of the more insidious problems lurking in the original PDC code.
(HINT:  Try reversing the order of the two operands in the offending A68k 
source, and observe that the problem disappears.  Now do you see the where
the problem is coming from?  The fix is trivial from there.  You might even
find that the fix is more efficient that the original!)

What on Earth are you doing that would move you to start from the first release
of PDC?  That code is NOT maintainable.  Don't waste your time!

< Lionel

P.S.  You can find the fix to that and a host of other problems by looking
in the source to the latest PDC (3.34) in amiga/misc/PDC on:

               128.174.5.59	ux1.cso.uiuc.edu uiucux1 ux1