Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!sundc!pitstop!sun!amdcad!ames!ptsfa!ihnp4!homxb!mtuxo!mtune!codas!killer!jfh
From: jfh@killer.UUCP (The Beach Bum)
Newsgroups: comp.unix.questions
Subject: Re: Wierd Compilers
Message-ID: <1972@killer.UUCP>
Date: Mon, 2-Nov-87 17:10:00 EST
Article-I.D.: killer.1972
Posted: Mon Nov  2 17:10:00 1987
Date-Received: Sat, 7-Nov-87 04:21:23 EST
References: <367@white.gcm>
Organization: Big "D" Home for Wayward Hackers
Lines: 54
Keywords: Sun
Summary: Optimizer is broke? Or you didn't use it? Or it's just that way?

The answer is in how the compiler is going to generate the code.  Pre-
and Post- decrement require different code to be generated.

In article <367@white.gcm>, dc@gcm (Dave Caswell) writes:

> 
> /* Program 1 */
> main()
> {
> 	int 	i = 500000;
> 
> 	while(--i);
> }

[ code done by hand ... ]

main:
	link	#-4,a6
	movl	#500000,-4(a6)
1$:	subl	#1,-4(a6)
	bne	1$		| condition codes set as a side effect.
	rts
> 
> /* Program 2 */
> main()
> {
> 	int 	i = 500000;
> 
> 	while(i--);
> }

[ coded by hand again ... ]

main:
	link	#-4,a6
	movl	#50000,-4(a6)
1$:	movl	-4(a6),d0
	subl	#1,-4(a6)
	tstl	d0		| codes got trashed after the subl.

> Does anyone have a guess why program one runs in half the time of program two?
> BTW this is a SUN 3.

Sure, look at the assembler I am suggesting and the reason is obvious.  You
might want to look at the code your compiler actually generates.  In general,
don't post-anything when evaluating for side effects only.  _Unless_ you
really have to.  Micro-optimizations are a loser though.

- John.
-- 
John F. Haugh II		HECI Exploration Co. Inc.
UUCP:	...!ihnp4!killer!jfh	11910 Greenville Ave, Suite 600
"Don't Have an Oil Well?"	Dallas, TX. 75243
" ... Then Buy One!"		(214) 231-0993