brooke/mandel-6502
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

hmm

Browse source
This commit is contained in:
Brooke Vibber 2023-01-04 20:21:51 -08:00
parent 58519381bf
commit 38091e535f
1 changed files with 10 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
mandel.s
View file

@ -30,8 +30,8 @@ FRX = $ec
.endmacro .endmacro
; inner loop for imul16 ; inner loop for imul16
; bitnum < 8: 25 or 30 cycles ; bitnum < 8: 25 or 41 cycles
; bitnum >= 8: 39 or 44 cycles ; bitnum >= 8: 30 or 46 cycles
.macro bitmul16 arg1, arg2, result, bitnum .macro bitmul16 arg1, arg2, result, bitnum
.local zero .local zero
.local one .local one
@ -54,7 +54,7 @@ FRX = $ec
.endif .endif
bne one ; 2 cyc bne one ; 2 cyc
zero: ; 18 cyc / 23 cyc zero: ; 18 cyc, 23 cyc
lsr result + 3 ; 5 cyc lsr result + 3 ; 5 cyc
ror result + 2 ; 5 cyc ror result + 2 ; 5 cyc
ror result + 1 ; 5 cyc ror result + 1 ; 5 cyc
@ -65,7 +65,7 @@ zero: ; 18 cyc / 23 cyc
.endif .endif
jmp next ; 3 cyc jmp next ; 3 cyc
one: ; 32 cyc / 37 cyc one: ; 32 cyc, 37 cyc
; 16-bit add on the top bits ; 16-bit add on the top bits
clc ; 2 cyc clc ; 2 cyc
lda result + 2 ; 3 cyc lda result + 2 ; 3 cyc
@ -73,13 +73,16 @@ one: ; 32 cyc / 37 cyc
sta result + 2 ; 3 cyc sta result + 2 ; 3 cyc
lda result + 3 ; 3 cyc lda result + 3 ; 3 cyc
adc arg2 + 1 ; 3 cyc adc arg2 + 1 ; 3 cyc
ror a ; 2 cyc - get a jump on the shift ror a ; 2 cyc
sta result + 3 ; 3 cyc sta result + 3 ; 3 cyc
ror result + 2 ; 5 cyc ror result + 2 ; 5 cyc
ror result + 1 ; 5 cyc ror result + 1 ; 5 cyc
.if bitnum >= 8 .if bitnum >= 8
; we can save 5 cycles * 8 bits = 40 cycles total by skipping this byte
; when it's all uninitialized data
ror result ; 5 cyc ror result ; 5 cyc
.endif .endif
next: next:
.endmacro .endmacro
@ -118,8 +121,8 @@ positive:
; of a larger routine ; of a larger routine
; 424 to 672 cycles ; 424 to 672 cycles
.repeat 16, bitnum .repeat 16, bitnum
; first half: 24 to 40 cycles ; first half: 22 to 40 cycles
; second half: 29 to 44 cycles ; second half: 29 to 47 cycles
bitmul16 arg1, arg2, result, bitnum bitmul16 arg1, arg2, result, bitnum
.endrepeat .endrepeat