Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site brl-tgr.ARPA
Path: utzoo!linus!philabs!cmcl2!seismo!brl-tgr!tgr!cottrell@nbs-vms.arpa
From: cottrell@nbs-vms.arpa (COTTRELL, JAMES)
Newsgroups: net.lang.c
Subject: Division With Shifts
Message-ID: <1395@brl-tgr.ARPA>
Date: Thu, 9-Jan-86 13:07:24 EST
Article-I.D.: brl-tgr.1395
Posted: Thu Jan  9 13:07:24 1986
Date-Received: Sat, 11-Jan-86 05:44:49 EST
Sender: news@brl-tgr.ARPA
Lines: 30

/*
> Just shifting using an arithmetic shift may give the wrong answer, but
> you could correct the rounding and still get much faster execution like:
> 
> 	shiftRightArithmetic	register
> 	branchIfPositive	label
> 	increment		register
> label:
> 
> so -23/2 = -23>>1 + 1 = 11101001>>2 + 1 = 11110100 + 1 = 11110101 = -11
> which is the right answer...

For -23, but not in the general case. Try an even number. The code must
be like so (for the general shift case):

	Test			Target
	BranchPositive		label
	Add			2**Shiftcount - 1,Target
label:	ShiftRightArithmatic	Target,Shiftcount

So in order to divide by 2, add 1 & shift once; by 4 add three & shift 
twice; by 8 add 7 & shift thrice; etc

> 					-Miles

Miles from nowhere?

	jim		cottrell@nbs
*/
------