Apple-2-SW/prog8
1
0
Fork 0
mirror of https://github.com/irmen/prog8.git synced 2024-07-10 23:29:02 +00:00
Code Issues Projects Releases Wiki Activity

added the last of the optimized mul_word asm routines

Browse Source
This commit is contained in:
Irmen de Jong 2020-09-14 23:54:01 +02:00
parent aff6b1fca5
commit 48858019b7
2 changed files with 139 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

143
compiler/res/prog8lib/math.asm
View File

@ -786,8 +786,20 @@ mul_byte_3 .proc
.pend
mul_word_3 .proc
.error "todo mul_word_3"
; AY = AY*2 + AY
sta P8ZP_SCRATCH_W1
sty P8ZP_SCRATCH_W1+1
sta P8ZP_SCRATCH_W2
sty P8ZP_SCRATCH_W2+1
asl a
rol P8ZP_SCRATCH_W1+1
clc
adc P8ZP_SCRATCH_W2
sta P8ZP_SCRATCH_W1
lda P8ZP_SCRATCH_W1+1
adc P8ZP_SCRATCH_W2+1
tay
lda P8ZP_SCRATCH_W1
rts
.pend
@ -803,8 +815,22 @@ mul_byte_5 .proc
.pend
mul_word_5 .proc
.error "todo mul_word_5"
; AY = AY*4 + AY
sta P8ZP_SCRATCH_W1
sty P8ZP_SCRATCH_W1+1
sta P8ZP_SCRATCH_W2
sty P8ZP_SCRATCH_W2+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
clc
adc P8ZP_SCRATCH_W2
sta P8ZP_SCRATCH_W1
lda P8ZP_SCRATCH_W1+1
adc P8ZP_SCRATCH_W2+1
tay
lda P8ZP_SCRATCH_W1
rts
.pend
@ -820,8 +846,24 @@ mul_byte_6 .proc
.pend
mul_word_6 .proc
.error "todo mul_word_6"
; AY = (AY*2 + AY)*2
sta P8ZP_SCRATCH_W1
sty P8ZP_SCRATCH_W1+1
sta P8ZP_SCRATCH_W2
sty P8ZP_SCRATCH_W2+1
asl a
rol P8ZP_SCRATCH_W1+1
clc
adc P8ZP_SCRATCH_W2
sta P8ZP_SCRATCH_W1
tay
lda P8ZP_SCRATCH_W1+1
adc P8ZP_SCRATCH_W2+1
sta P8ZP_SCRATCH_W1+1
tya
asl a
rol P8ZP_SCRATCH_W1+1
ldy P8ZP_SCRATCH_W1+1
rts
.pend
@ -837,8 +879,24 @@ mul_byte_7 .proc
.pend
mul_word_7 .proc
.error "todo mul_word_7"
; AY = AY*8 - AY
sta P8ZP_SCRATCH_W1
sty P8ZP_SCRATCH_W1+1
sta P8ZP_SCRATCH_W2
sty P8ZP_SCRATCH_W2+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
sec
sbc P8ZP_SCRATCH_W2
sta P8ZP_SCRATCH_W1
lda P8ZP_SCRATCH_W1+1
sbc P8ZP_SCRATCH_W2+1
tay
lda P8ZP_SCRATCH_W1
rts
.pend
@ -854,8 +912,25 @@ mul_byte_9 .proc
.pend
mul_word_9 .proc
.error "todo mul_word_9"
; AY = AY*8 + AY
sta P8ZP_SCRATCH_W1
sty P8ZP_SCRATCH_W1+1
sta P8ZP_SCRATCH_W2
sty P8ZP_SCRATCH_W2+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
clc
adc P8ZP_SCRATCH_W2
sta P8ZP_SCRATCH_W1
lda P8ZP_SCRATCH_W1+1
adc P8ZP_SCRATCH_W2+1
tay
lda P8ZP_SCRATCH_W1
rts
rts
.pend
@ -871,8 +946,25 @@ mul_byte_10 .proc
.pend
mul_word_10 .proc
.error "todo mul_word_10"
; AY=(AY*4 + AY)*2
sta P8ZP_SCRATCH_W1
sty P8ZP_SCRATCH_W1+1
sta P8ZP_SCRATCH_W2
sty P8ZP_SCRATCH_W2+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
clc
adc P8ZP_SCRATCH_W2
sta P8ZP_SCRATCH_W1
lda P8ZP_SCRATCH_W1+1
adc P8ZP_SCRATCH_W2+1
sta P8ZP_SCRATCH_W1+1
lda P8ZP_SCRATCH_W1
asl a
rol P8ZP_SCRATCH_W1+1
ldy P8ZP_SCRATCH_W1+1
rts
.pend
@ -903,8 +995,25 @@ mul_byte_12 .proc
.pend
mul_word_12 .proc
.error "todo mul_word_12"
; AY=(AY*2 + AY)*4
sta P8ZP_SCRATCH_W1
sty P8ZP_SCRATCH_W1+1
sta P8ZP_SCRATCH_W2
sty P8ZP_SCRATCH_W2+1
asl a
rol P8ZP_SCRATCH_W1+1
clc
adc P8ZP_SCRATCH_W2
sta P8ZP_SCRATCH_W1
lda P8ZP_SCRATCH_W1+1
adc P8ZP_SCRATCH_W2+1
sta P8ZP_SCRATCH_W1+1
lda P8ZP_SCRATCH_W1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
ldy P8ZP_SCRATCH_W1+1
rts
.pend
@ -950,8 +1059,26 @@ mul_byte_15 .proc
.pend
mul_word_15 .proc
.error "todo mul_word_15"
; AY = AY * 16 - AY
sta P8ZP_SCRATCH_W1
sty P8ZP_SCRATCH_W1+1
sta P8ZP_SCRATCH_W2
sty P8ZP_SCRATCH_W2+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
asl a
rol P8ZP_SCRATCH_W1+1
sec
sbc P8ZP_SCRATCH_W2
sta P8ZP_SCRATCH_W1
lda P8ZP_SCRATCH_W1+1
sbc P8ZP_SCRATCH_W2+1
tay
lda P8ZP_SCRATCH_W1
rts
.pend

33
examples/test.p8
View File

@ -18,36 +18,11 @@ main {
; ub1 = ii
; ub1 *= 40 ; TODO implement non-stack optimized muls
; todo a = EXPRESSION * const -> is that optimized?
ub1 = ii * 25
ub1 = ii * 15
txt.print_uw(ub1)
c64.CHROUT(',')
ub1 = 1+(ii * 15)
txt.print_uw(ub1)
; c64.CHROUT(',')
; ub1 = ii * 50
; txt.print_uw(ub1)
; c64.CHROUT(',')
; ub1 = ii * 80
; txt.print_uw(ub1)
; c64.CHROUT(',')
; ub1 = ii * 100
; txt.print_uw(ub1)
c64.CHROUT('\n')
}
c64.CHROUT('\n')
for ii in 0 to 20 {
; ub1 = ii
; ub1 *= 40 ; TODO implement non-stack optimized muls
; todo a = EXPRESSION * const -> is that optimized?
ww1 = (-ii) * 25
txt.print_w(ww1)
; c64.CHROUT(',')
; ub1 = ii * 50
; txt.print_uw(ub1)
; c64.CHROUT(',')
; ub1 = ii * 80
; txt.print_uw(ub1)
; c64.CHROUT(',')
; ub1 = ii * 100
; txt.print_uw(ub1)
c64.CHROUT('\n')
}