Assembly: Print to Screen & EOUTCH

[+Ripdubski]

Aside from being XL/XE specific, whats wrong with using this vector (EOUTCH) to print to the screen - if anything - rather than using IOCB?

 

Test Assembler:

10    *= $0600
20 EOUTCH=$F2B0  ; F6A4 for 800
30    LDA #65
40    JSR EOUTCH
50    RTS
60    .END

Output:

A

[+JAC!]

There are many different OS variants/replacements for Atari computers (QMEG, HI-Speed, LLE, ...). Therefore you should always use the correct vectors. The shortest way for your use case is something like:

 

.proc print_char;IN: <A> = char, changes <A>, <X>, <Y>

sta temp
lda $e407 ;Use PUT_CHAR from E: handler

pha

lda $e406

pha

temp = *+1

lda #$00

rts

.endp

Edited April 24, 2014 by JAC!

[+Ripdubski]

I had not thought about non atari os's. Good advice. Thats the why I was looking for.

 

Since im learning assembly help me understand that code.

 

Store acc in temp. (The value A from my code)

Load acc with value in e407

Push acc to stack

Load acc with value in e406

Push acc to stack

Temp storage space (where the a is stored)

Load acc with 0

Return

 

So my code loads acc with 65 (A), then jumps to this routine. Which instruction is actually causing the a to print via the e: put vector? I guess im not seeing something. And Why E and not S? Noob questions i know, but im learning. Thx.

[kenjennings]

The code pushes the PUT_CHAR vector address on the stack.

The RTS causes this routine to "return" to the address on the stack... which is now the PUT_CHAR vector address.

The PUT CHAR routine will return by RTS which will end up using the original address -- that is, returning to the point in the code that called this print_char routine in the first place.

[phaeron]

Better to use ICPTL/H from IOCB #0 ($0346/$0347) rather than hardcoding the OS's E: vector. It's automatically set by CIO when the IOCB is opened and that'll still work if the E: device is replaced in HATABS.

[+Ripdubski]

The code pushes the PUT_CHAR vector address on the stack.

The RTS causes this routine to "return" to the address on the stack... which is now the PUT_CHAR vector address.

The PUT CHAR routine will return by RTS which will end up using the original address -- that is, returning to the point in the code that called this print_char routine in the first place.

Ok, that makes sense. Thx! And thsnks phaeron for the additional why. Edited April 25, 2014 by Ripdubski

[+Ripdubski]

Using CIO and IOCB 0 I was finally able to get Hello World displayed. Thx for the help! Feels like a huge accomplishment though I know it isn't.

Edited April 25, 2014 by Ripdubski