Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!rutgers!seismo!rochester!cornell!batcomputer!braner
From: braner@batcomputer.UUCP
Newsgroups: comp.lang.c
Subject: Re: fabs(x) vs. (x) < 0 ? -(x) : (x)
Message-ID: <160@batcomputer.tn.cornell.edu>
Date: Wed, 11-Feb-87 01:32:46 EST
Article-I.D.: batcompu.160
Posted: Wed Feb 11 01:32:46 1987
Date-Received: Thu, 12-Feb-87 04:06:03 EST
References: <4943@mimsy.UUCP> <2550005@hpisod2.HP> <756@unc.unc.UUCP> <2181@batcomputer.tn.cornell.edu> <1315@ho95e.ATT.COM>
Reply-To: braner@batcomputer.UUCP (braner)
Distribution: na
Organization: Theory Center, Cornell University, Ithaca NY
Lines: 21
Summary: bit-clearing via #define: wrong

[]

eeexcuse me, but:

#define fabs(x) ((x) & 0x7FFFFF...)

will NOT work, and for several reasons:

The two operands ((x) and 0x7F...) are not of the same type, so one
of them has to be converted.  Then it won't work!!  Actually, it is
illegal to do '&' on a real, if I remember right.  And also, doubles
are usually longer than longs (until we get 64-bit longs :-).

You MIGHT get away with ((*(long *)&x) & 0x7FFFFFFF), but only in cases
where &x is legal, and only on machines where the sign bit of x is where
you expect it...

- Moshe Braner

THE COMPLETE FP LIB MUST BE WRITTEN IN HAND-OPTIMIZED AL!!!
ESPECIALLY SO IF YOU HAVE FP HARDWARE!!!