Benchmark Programs

;Latest Revision — September 30, 1986
;
move	#points/2,n0			;initialize butterﬂies per group
move	#1,n2			;initialize groups per pass
move	#points/4,n6			;initialize C pointer offset
move	#-1,m0			;initialize A and B address modiﬁers
move	m0,m1			;for linear addressing
move	m0,m4
move	m0,m5
move	#0,m6			;initialize C address modiﬁer for
				;reverse carry (bit-reversed) addressing
;
;Perform all FFT passes with triple nested DO loop
;
do	#@cvi (@log(points)/@log(2)+0.5),_end_pass
move	#data,r0			;initialize A input pointer
move	r0,r4			;initialize A output pointer
lua	(r0)+n0,r1			;initialize B input pointer
move	#coef,r6			;initialize C input pointer
lua	(r1)-,r5			;initialize B output pointer
move	n0,n1			;initialize pointer offsets
move	n0,n4
move	n0,n5
do	n2,_end_grp
move	x:(r1),X1		y:(r6),y0		;lookup -sine and
					; -cosine values
move	x:(r5),a		y:(r0),b		;preload data
move	x:(r6)+n6,x0				;update C pointer
do	n0,_end_bfy
mac	x1,y0,b		y:(r1)+,y1		;Radx 2 DIT
					;butterﬂy kernel
macr	-x0,y1,b		a,x:(r5)+	y:(r0),a
subl	b,a		x:(r0),b	b,y:(r4)
mac	-x1,x0,b		x:(r0)+,a	a,y:(r5)
macr	-y1,y0,b		x:(r1),x1
subl	b,a		b,x:(r4)+	y:(r0),b
_end_bfy
move	a,x:(r5)+n5		y:(r1)+n1,y1		;update A and B pointers
move	x:(r0)+n0,x1		y:(r4)+n4,y1
_end_grp
move	n0,b1			;divide butterﬂies per group by two
Isr	b	n2,a1		;multiply groups per pass by two
IsI	a	b1,n0
move	a1,n2
_end_pass
endm

Figure B-2 Radix 2, In-Place, Decimation-In-Time FFT (Sheet 2 of 2) Figure B-5 Real Input FFT Based on Glenn Bergland Algorithm (Sheet 1 of 8)

B - 12