Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!gem.mps.ohio-state.edu!wuarchive!wugate!uunet!crdgw1!crdos1!davidsen
From: davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr)
Newsgroups: comp.sys.ibm.pc
Subject: Re: 32 bit C Comps and Long
Summary: Here's why
Message-ID: <1415@crdos1.crd.ge.COM>
Date: 24 Oct 89 13:05:25 GMT
References: <1224@utkcs2.cs.utk.edu> <1000@kosman.UUCP>
Reply-To: davidsen@crdos1.UUCP (bill davidsen)
Organization: GE Corp R&D Center
Lines: 34


  I think a lot of people are missing the point of the whole question.
On a 386 32 bit arithmetic is a LOT faster. You don't need protected
mode or anything fancy, just use the 32 bit instructions. Here's an
example program:

#include <stdio.h>

	main() {
	  long ary[2000], x = 4, y = 7;
	  int ix, iy;
	
	  for (ix = 0; ix < 2000; ix++) ary[ix] = ix+1200;
	  for (iy = 4000; iy--; ) {
	    for (ix = 0; ix < 2000; ix++) {
	      ary[ix] = (ary[ix] + y) * x / y;
	    }
	  }
	}
	
and the results. It was compiled using 386 and 286 instructions.
	386 runtime:  36.50 sec
	286 runtime: 121.66 sec
Both times were on the same 25MHz machine, just using one set of
instructions or the other.

This is why I tell people not to buy a 286 any more, over the next few
years there will be a lot of programs taking advantage of this 4:1
performance gain. I believe that the ZIP compressor does now.
-- 
bill davidsen	(davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen)
"The world is filled with fools. They blindly follow their so-called
'reason' in the face of the church and common sense. Any fool can see
that the world is flat!" - anon