backwards multiply (Re: A86: Sorry Dux. =-|)

To: assembly-86@lists.ticalc.org
Subject: backwards multiply (Re: A86: Sorry Dux. =-|)
From: rabidcow@juno.com (Joshua J Seagoe)
Date: Thu, 13 Aug 1998 00:10:22 -0700
Delivered-To: assembly-86-outgoing@towerguard.unix.edu.sollentuna.se
References: <35CE0E8D.26BC@busprod.com> <35CFA57F.3DB9A99D@eden.rutgers.edu>
Reply-To: assembly-86@lists.ticalc.org



On Mon, 10 Aug 1998 19:59:27 -0600 Dux Gregis <assets@eden.rutgers.edu>
writes:
>
>It can't be done backwards :-)  It wouldn't be faster if you did, 
>because
>you would end up looping more times than needed.  You always need one 
>value
>shifting right and one value shifting left.  The value shifting left 
>must
>be 16-bit and preferably the larger of the two (for speed).
>

noooooooooooo!!  it has to work!

;hl=de*b
;destroys af,bc

	xor a
	ld h,a
	ld l,a
	cp b
	jp z,quit
	ld c,b		;save for postshift
	ld h,d
	ld l,e
preshift:
	sla b		;find leftmost bit
	jp nc,preshift
mainloop:
	add hl,hl	;multiply :)
	sla b
	jp nc,mainloop	;it'll only become 0 when
	add hl,de	;a 1 is shifted out
	jp nz,mainloop
	srl c		;start postshift...
	jp c,quit	;this fixes the problem that
postshift:		;you'd get working backwards
	add hl,hl	;with even numbers
	srl c
	jp nc,postshift
quit:	ret		;jp z falls through faster that ret z

-josh


_____________________________________________________________________
You don't need to buy Internet access to use free Internet e-mail.
Get completely free e-mail from Juno at http://www.juno.com
Or call Juno at (800) 654-JUNO [654-5866]

References:

A86: Sorry Dux. =-|

From: Tercero <tercero@busprod.com>

Re: A86: Sorry Dux. =-|

From: Dux Gregis <assets@eden.rutgers.edu>