Path: utzoo!attcan!uunet!lll-winken!elroy.jpl.nasa.gov!usc!wuarchive!cs.utexas.edu!yale!cmcl2!lanl!jlg
From: jlg@lanl.gov (Jim Giles)
Newsgroups: comp.lang.misc
Subject: Re: Aggressive optimization
Message-ID: <5009@lanl.gov>
Date: 5 Nov 90 22:26:18 GMT
References: <2701@l.cc.purdue.edu>
Organization: Los Alamos Natl Lab, Los Alamos, N.M.
Lines: 19

From article <2701@l.cc.purdue.edu>, by cik@l.cc.purdue.edu (Herman Rubin):
> [...]
> There is a way out of this, which I do not believe has been tried.  That is,
> have the compiler ask the programmer!  It may even be that the programmer
> does not know, but can find out.

And yet, Herman Rubin is one of those who opposed my proposal for
explicit "aliased" attributes together with a feedback mechanism
in the loader for indicating when the user's explicit directives
were mismatched.  The specific purpose of these was to allow
the user do safely direct that optimizations were allowed (by
not aliasing things) or to tell the compiler not to optimize
specific combinations of variables (which are aliased).  Further,
the proposal was made in a specific way so that the user would
_always_ know whether aliasing were needed of not.

Oh well

J. Giles