hello,

i have a big Problem. I would like to program modulo.
which I can do in Michael Kowalski of 6502
macroassembler & simulator tests.
for sample i will test.

B modulo 3 = 2


*= $200

io_puth = io_area + 3
count1 = $41
count2 = $42

LDA #$0B
STA io_puth
STA count1

LDA #$03
STA io_puth
STA count2

; ????????? count1 modulo count2 = 2
; OR
; B modulo 3 =2


as I can make that

thanks ahead
lee

How about:

LDA count1
SEC

label:
SBC count2
BCS label

ADC count1

IOW, you just subtract count2 from the first number (which started out as the contents of count1) until the subtraction results in a borrow. Then you add count1 back in to get back to the last number you had before the borrow occurred. You don't need CLC before the ADC because it's already clear by the time we get there. Unless I'm forgetting something, this should work for any pair of unsiged 8-bit numbers, even if the subtraction result causing the carry looks positive. (For example, for $4E mod $F3, $4E minus $F3 yields $5B.) This still works because we're testing the C flag and not the N flag. It is not necessary to test V.

Garth

When the numbers get bigger, there are more complex but much faster methods. This one works for all sorts of problems, including converting binary to decimal.

To copy one number to another, we can repeatedly halve the input while doubling the output, adding one whenver we halved an odd number. To convert from one representation to another, just do the appropriate "halve" and "double" operations for those representations.

In this case we want the output modulo count2, so the doubling has to be done modulo count2. That's easy - double, and if the result is greater than count2, subtract count2.

If you wanted to convert to decimal, you'd do the doubling in decimal mode. To convert from decimal, you need a decimal-mode halve. You can even convert to or from a mixed-radix representation like BCD hours:minutes:seconds.

LDA #0
loop:
LSR count1
ROL
CMP count2
BCS skip
SEC
SBC count2
skip:
LDX count1
BNE loop

Untested, and I might have got the branch condition wrong (I always do that)