Path: utzoo!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!uunet!tut.cis.ohio-state.edu!ucbvax!SUN.COM!wmb
From: wmb@SUN.COM (Mitch Bradley)
Newsgroups: comp.lang.forth
Subject: double precision arithmetic in C
Message-ID: <8912301506.AA00638@jade.berkeley.edu>
Date: 29 Dec 89 21:21:38 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Reply-To: Forth Interest Group International List <FIGI-L%SCFVM.bitnet@jade.berkeley.edu>
Organization: The Internet
Lines: 19

I just discovered an easy way to develop carry and borrow for implementing
multiple precision arithmetic in C.  This way is easier and faster than
my previous solution (posted some time ago).

#define CARRY(a,b)   ((unsigned)a >= (unsigned)(-b))
#define BORROW(a,b)  ((unsigned)a <  (unsigned)b)

This method assumes 2's complement arithmetic.

My previous solution for CARRY computed a logical function involving
the sign bits of the arguments and the result, similar to the way it
is usually done in hardware.  I discovered this new solution
while implementing D- ; the definition of BORROW is pretty obvious in
retrospect, and then I got to thinking that there must be a symmetrical
solution for CARRY.  Proof of the correctness is left to the reader as
an exercise (Hint: draw a number circle and consider negation as reflection
about the axis).

Mitch