Difference between revisions of "Z80 Routines:Optimized:PopCountA"

From WikiTI
Jump to: navigation, search
(Created page with "PopCountA PopCountA This is a faster/smaller (and also obfuscated) replacement for the normal popcount a, which r...")
 
Line 8: Line 8:
 
;27 bytes, 108 clock cycles
 
;27 bytes, 108 clock cycles
 
TypicalPopCountA:
 
TypicalPopCountA:
ld c,a
 
 
xor a
 
xor a
ld b,a
+
ld c, a
rrc c
+
rrc b
adc a,b
+
adc a, c
rrc c
+
rrc b
adc a,b
+
adc a, c
rrc c
+
rrc b
adc a,b
+
adc a, c
rrc c
+
rrc b
adc a,b
+
adc a, c
rrc c
+
rrc b
adc a,b
+
adc a, c
rrc c
+
rrc b
adc a,b
+
adc a, c
rrc c
+
rrc b
adc a,b
+
adc a, c
rrc c
+
rrc b
adc a,b</nowiki>
+
adc a, c</nowiki>
  
 
Better routine:
 
Better routine:
Line 36: Line 35:
 
;author: jacobly
 
;author: jacobly
 
PopCountA:
 
PopCountA:
ld c, a
+
ld c, a ; c=(A|B|C|D|E|F|G|H)
and 10101010b
+
and 10101010b ; a=(A|0|C|0|E|0|G|0)
cpl
+
cpl ; a=(~A|1|~C|1|~E|1|~G|1)
rrca
+
rrca ; a=(1|~A|1|~C|1|~E|1|~G), cf=1
adc a, c
+
adc a, c ; a=(A+B|C+D|E+F|G+H)
ld b, a
+
ld b, a ; b=(A+B|C+D|E+F|G+H)
and 00110011b
+
and 00110011b ; a=(00|C+D|00|G+H)
ld c, a
+
ld c, a ; c=(00|C+D|00|G+H)
xor b
+
xor b ; a=(A+B|00|E+F|00)
rrca
+
rrca \ rrca ; a=(0|A+B|00|E+F|0)
rrca
+
add a, c ; a=(A+B+C+D|E+F+G+H)
add a, c
+
ld c, a ; c=(A+B+C+D|E+F+G+H)
ld c, a
+
rrca \ rrca \ rrca \ rrca ; a=(E+F+G+H|A+B+C+D)
rrca
+
add a, c ; a=(A+B+C+D+E+F+G+H|A+B+C+D+E+F+G+H)
rrca
+
and 00001111b ; a=A+B+C+D+E+F+G+H</nowiki>
rrca
+
rrca
+
add a, c
+
and 00001111b</nowiki>
+

Revision as of 20:35, 15 September 2020

This is a faster/smaller (and also obfuscated) replacement for the normal popcount a, which returns the number of set bits in a.

Typical routine:

;input: b
;output: a
;27 bytes, 108 clock cycles
TypicalPopCountA:
	xor	a
	ld	c, a
	rrc	b
	adc	a, c
	rrc	b
	adc	a, c
	rrc	b
	adc	a, c
	rrc	b
	adc	a, c
	rrc	b
	adc	a, c
	rrc	b
	adc	a, c
	rrc	b
	adc	a, c
	rrc	b
	adc	a, c

Better routine:

;- Pop Count A
;input:	byte in A
;output: number of set bits in A
;destroys BC
;22 bytes and 85 clock cycles
;author: jacobly
PopCountA:
	ld	c, a			; c=(A|B|C|D|E|F|G|H)
	and	10101010b		; a=(A|0|C|0|E|0|G|0)
	cpl				; a=(~A|1|~C|1|~E|1|~G|1)
	rrca				; a=(1|~A|1|~C|1|~E|1|~G), cf=1
	adc	a, c			; a=(A+B|C+D|E+F|G+H)
	ld	b, a			; b=(A+B|C+D|E+F|G+H)
	and	00110011b		; a=(00|C+D|00|G+H)
	ld	c, a			; c=(00|C+D|00|G+H)
	xor	b			; a=(A+B|00|E+F|00)
	rrca \ rrca			; a=(0|A+B|00|E+F|0)
	add	a, c			; a=(A+B+C+D|E+F+G+H)
	ld	c, a			; c=(A+B+C+D|E+F+G+H)
	rrca \ rrca \ rrca \ rrca	; a=(E+F+G+H|A+B+C+D)
	add	a, c			; a=(A+B+C+D+E+F+G+H|A+B+C+D+E+F+G+H)
	and	00001111b		; a=A+B+C+D+E+F+G+H