Path: utzoo!utgpu!watserv1!watmath!att!pacbell.com!ucsd!hub.ucsb.edu!ucsbuxa!3003jalp
From: 3003jalp@ucsbuxa.ucsb.edu (Applied Magnetics)
Newsgroups: comp.lang.fortran
Subject: Re: strength reduction of array address calculations
Message-ID: <7496@hub.ucsb.edu>
Date: 29 Nov 90 18:41:43 GMT
References: <BURLEY.90Nov28084920@pogo.ai.mit.edu> <BURLEY.90Nov29103957@pogo.ai.mit.edu>
Sender: news@hub.ucsb.edu
Lines: 10

In article <BURLEY.90Nov29103957@pogo.ai.mit.edu> burley@pogo.ai.mit.edu (Craig Burley) writes:

>[Fascinating discussion of compiler internals]

Is it standard practice to translate the high-level language to
low-level construct and _then_ optimize?  Isn't it a case of
reverse-engineering?  (Then again, vectorizing Fortrans do just that:
reverse-engineer DO loops into vector operations.)

--P. Asselin, R&D, Applied Magnetics Corp.  I speak for me.