Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucbvax!decwrl!decvax!ima!haddock!karl
From: karl@haddock.ima.isc.com (Karl Heuer)
Newsgroups: comp.lang.c
Subject: Re: C optimizer
Message-ID: <11782@haddock.ima.isc.com>
Date: 16 Feb 89 01:45:58 GMT
References: <515@larry.UUCP> <3684@arcturus>
Reply-To: karl@haddock.ima.isc.com (Karl Heuer)
Organization: Interactive Systems, Boston
Lines: 12

In article <3684@arcturus> evil@arcturus.UUCP (Wade Guthrie) writes:
>I would expect the following optimizations: ... 3) reducing things like
>intval/2 == intval >> 1 when they are faster.

On most implementations these are not equivalent, unless the compiler can
prove that intval >= 0.

(Ah, many's the time I've wished I could declare a type `nonnegative', which
would be the intersection of signed and unsigned, so the compiler could use
whichever happened to be faster...)

Karl W. Z. Heuer (ima!haddock!karl or karl@haddock.isc.com), The Walking Lint