Path: utzoo!mnetor!uunet!tektronix!orca!tekecs!frip!andrew
From: andrew@frip.gwd.tek.com (Andrew Klossner)
Newsgroups: comp.arch
Subject: Re: conditional branches
Message-ID: <9782@tekecs.TEK.COM>
Date: 22 Feb 88 16:27:07 GMT
References: <191@telesoft.UUCP> <28200098@ccvaxa>
Sender: nobody@tekecs.TEK.COM
Organization: Tektronix, Wilsonville, Oregon
Lines: 15


	"The test x = 0, and therefore x = y, can almost certainly be
	done with minimum delay in a pipeline, as can x < 0, looking
	just at the sign bit.  A full arithmetic compare might be a
	problem because of carry."

I've seen this assertion a couple of times, but it ain't so.  You need
to worry about carry when doing a subtract, but not when just doing a
compare.  If both operands are positive, a simple bit-for-bit
comparison is sufficient, analogous to the way that C programs compare
strings.  When the operands can be signed, a little extra logic is
required, but nothing as fancy as a full adder/subtractor.

  -=- Andrew Klossner   (decvax!tektronix!tekecs!andrew)       [UUCP]
                        (andrew%tekecs.tek.com@relay.cs.net)   [ARPA]