Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!rutgers!mit-eddie!zrm
From: zrm@eddie.MIT.EDU (Zigurd R. Mednieks)
Newsgroups: comp.arch,comp.sys.nsc.32k
Subject: Re: "Unoptimizing" Dhrystone
Message-ID: <5565@eddie.MIT.EDU>
Date: Tue, 21-Apr-87 12:35:44 EST
Article-I.D.: eddie.5565
Posted: Tue Apr 21 12:35:44 1987
Date-Received: Wed, 22-Apr-87 03:46:28 EST
References: <4190@nsc.nsc.com> <951@moscom.UUCP> <2577@intelca.UUCP> <999@mips.UUCP> <312@gumby.UUCP> <1706@husc6.UUCP>
Reply-To: zrm@eddie.UUCP (Zigurd R. Mednieks)
Organization: MIT, EE/CS Computer Facilities, Cambridge, MA
Lines: 11
Xref: mnetor comp.arch:1030 comp.sys.nsc.32k:99

Just what is optimization? Is Green Hills' compiler optimizing when it
ignores "register" storage class keywords and uses its own intelligent
register allocation? Are RISC compiler postpass optimizers within the
rules of the Dhrystone game?

While it is pretty easy to say that something that radically alters
the traslation of C into machine code is optimizing, what separates
clever (or in the case of RISC, almost required) hacks in code
generation from optimization?

-Zigurd