6502.org

Hello,

I recently build my own 6502 computer thanks to Ben eater videos and also examples and tutorials from this forum. I also wanted to get EhBasic running so I set up everything like in reference manual compile it with simulator and it worked fine with my computer. So I decided to also add my own code to build simple interface from which I can run EhBasic or my other programs. So since I was using vasm because is pretty simple work with it I wanted to compile EhBasic with vasm. There I got few errors and warnings I don't understand. since I am still begginer and don't understand macros. there are also some garbage characters with I coudlnt get rid of after rewriting line. I know it is noob question but hopefully someone will be able help me. Thank you very much.

I compile with "vasm6502 -dotdir -Fbin min_mon.asm"
source files will be in attachment
Here are errors and warning I got:

error 10 in line 1551 of "basic.asm": number or identifier expected
included from line 8 of "min_mon.asm"
> .IF [* & $FF] == $FD

warning 1007 in line 1551 of "basic.asm": garbage at end of line
included from line 8 of "min_mon.asm"
> .IF [* & $FF] == $FD

error 10 in line 1630 of "basic.asm": number or identifier expected
included from line 8 of "min_mon.asm"
> CMP #[TK_TAB-$80]*2 ; compare normalised token * 2 with TAB

warning 2002 in line 1630 of "basic.asm": trailing garbage in operand
included from line 8 of "min_mon.asm"
> CMP #[TK_TAB-$80]*2 ; compare normalised token * 2 with TAB

error 10 in line 3105 of "basic.asm": number or identifier expected
included from line 8 of "min_mon.asm"
> .IF [* & $FF] == $FD

warning 1007 in line 3105 of "basic.asm": garbage at end of line
included from line 8 of "min_mon.asm"
> .IF [* & $FF] == $FD

error 10 in line 3342 of "basic.asm": number or identifier expected
included from line 8 of "min_mon.asm"
> CPX #<[LAB_224Da+2] ; from concatenate low return address?

warning 2002 in line 3342 of "basic.asm": trailing garbage in operand
included from line 8 of "min_mon.asm"
> CPX #<[LAB_224Da+2] ; from concatenate low return address?

error 10 in line 3344 of "basic.asm": number or identifier expected
included from line 8 of "min_mon.asm"
> CMP #>[LAB_224Da+2] ; from concatenate high return address?

warning 2002 in line 3344 of "basic.asm": trailing garbage in operand
included from line 8 of "min_mon.asm"
> CMP #>[LAB_224Da+2] ; from concatenate high return address?

warning 1007 in line 168 of "min_mon.asm": garbage at end of line
> .end RES_vec

Welcome! The nature of 6502 assembly language is that most assemblers have their own small differences from all the others. If you have a need to choose a different assembler than the one the code is written for, you may have to make substantial changes - in particular, you may need to understand both dialects.

I don't recommend doing this!

My advice would be as follows: when writing your own code, choose your assembler. When assembling someone else's code, stick with the choice they made.

Thank you for reply.

Okay I understand now. So if I want to add some mine code which will be separate from EhBasic I still need to use assembler from simulator? Or there is way to combine them?
because like I said I would like to add some mine programs and let EhBasic stay in memory so I can run it anytime. Since I have 32k of ROM and EhBasic use upper half of ROM.

There's usually more than one way to proceed... perhaps consider how to make a binary from sources, for the two or three parts of your system, and combine the binaries before loading them.

You will need to figure out how to call from one part to the other, by putting the absolute addresses of the routines into the other part's calls.

(I'm assuming you've got some way to get Basic into ROM, and you want your extra code also to be in ROM)

Yep! This definitely appears as a case of operand syntax differences. I agree: for now, assemble with the assembler used to develop the original source code.

Yes just like you said. If I would understand these macros better I could make it work I guess. I have a lot to learn. Anyway Thank you very much for help and advice.

I also tried something with code an now I only got 2 errors after compiling.

expression must be constant
.IF (* & $FF) == $FD

What does * means in this IF statement?

Usually * means the address of the present instruction being assembled - or the next instruction, if the line isn't an opcode line.

The intent of this .IF is to ensure that the return address of the following JSR doesn't end in $FF, or something to that effect. There's a 255 out of 256 chance that you can proceed without it, should it become too difficult to translate.

P.S. I'm proceeding to reverse-engineer WozBASIC, with the goal of getting it to assemble and run in the Kowalski simulator. When I have it fully functional, I'll fix the few bugs, remove the Apple ][ specific graphics and I/O verbs, then add WOZ floats. My ultimate goal is to provide a quicker and smaller alternative to EhBASIC, but I can see a lot of bumps and branches on the path ahead ...

Note that such a check is unnecessary if the target system is running a 65C02 or 65C816, as neither has the JMP (<addr>) bug.

Err ... not so fast, BDD. The return address is incremented incompletely in software inside the subroutine at LAB_1B5B, so it doesn't matter if it's hosted on NMOS or otherwise. Four (or even two, if you want to get fancy) extra bytes of code could do the job "properly", but apparently Lee chose this route of "fixing" it during assembly instead.

Actually, I applied this fix. Lee left it to the comments in the code. Unfortunately vasm6502 cannot handle variables in .if statements. From the vasm manual "Oldstyle Syntax Module" section:

After looking for error I also found vasm can't work with constants in .IF so I commented these block an after compiling code and comparing with binaries from simulator they were same. So I guess I should be okay then right?

As long as the simulator doesn't insert a NOP you're O.K. The chance of hitting a page crosser requiring to insert a NOP instruction in one of the 2 if-blocks is less than 1% (1/256*2).