Path: utzoo!attcan!uunet!husc6!bloom-beacon!tut.cis.ohio-state.edu!mailrus!ames!pasteur!ic.Berkeley.EDU!faustus
From: faustus@ic.Berkeley.EDU (Wayne A. Christopher)
Newsgroups: comp.lang.c
Subject: Re: no noalias not negligible
Message-ID: <3630@pasteur.Berkeley.Edu>
Date: 24 May 88 20:42:47 GMT
References: <4152@haddock.ISC.COM> <54080@sun.uucp> <1988May21.030207.25063@light.uucp>
Sender: news@pasteur.Berkeley.Edu
Lines: 28

In article <4152@haddock.ISC.COM>, karl@haddock.ISC.COM (Karl Heuer) writes:
> >How about adding a test before the for loop?  Something like:
> >#define overlap(x,y,n)    (!(x + n <= y || y + n <= x))
> >	if (overlap(dx, dy, n))
> >		return complain("overlapping arrays\n");
> >
> >Now a smart compiler can figure out that dx, dy don't overlap ...
> 
> The information is there, and a human reader can prove it, but I don't think
> they make compilers that smart yet.

The compler doesn't have to see it.  Probably a better way of saying it is
that the compiler will translate

	<conservative, un-optimized code>

into

	if ( <conditions like noalias hold> ) {
		<highly-optimized, vectorized code>
	} else {
		<conservative, un-optimized code>
	}

Are there any constructs that couldn't be figured out at run-time without a
large penalty?  Maybe if you were re-directing through an array of pointers...

	Wayne