I want to do hello world in CP/M (Yes, this isn't strictly z80, but the 8080 is close enough). My code is:


Code:


But my Hello.prn output is:


Code:


Which means that the label isn't being assembled. Anyone know why it isn't working?

Moved by Merth to General Programming. The only thing "close enough" to z80 to be in the z80 forum is z80.

I don't see how you can assume the label isn't being assembled. The "LXI D, HELLO" hex is 110801, which is opcode 11, argument 0108 (assuming this is little-endian, or 0801 if it's big-endian). You should have "ret" or whatever the CP/M equivalent is after the call to BDOS, but other than that I see nothing wrong. Presumably whatever a .prn output is doesn't show constants/data fields?

I'm saying that it isn't being assembled because that value points to nothing. What the .prn output is is the assembled output. There is nothing else in hello.hex but 0s.

Which assembler are you using?

A number of tools require that each line in a text document ends in a newline (this is the reason for the double .end in TASM source code; you only need one, but it needs to have a newline after it). Your assembler may be ignoring the final line in your program if it doesn't have a newline after it. You may also need an END statement in there.

You need to jump back to $0000 after your program has executed - stick a RST 0 after your call to BDOS.

Here's a version that works in M80 (in superior Z80 syntax):


Code:

Hello, I know this is a bit dated of a question, but I wanted to add a quick bit to it, just for anyone trying to learn assembler programming for CP/M systems. I'd ran into this page while doing so, and it caused a bit of confusion, so here's my take on the issue (with a few side notes as well)

Assuming that you are using the built in ASM.COM assembler that comes packaged with CP/M, it may be due to using double quotes (") for your string, with the assembler it requires you to use single quotes ('), otherwise it throws a fit. I believe that would explain why HELLO, the label, is pointing to the correct address (0108h) while the string itself isn't present in the assembled binary/hex.


Code:


Then the PRN output file should be something along the lines of this, mine may not be 100% accurate as I don't have a system to check it on currently

Code:



Here are some non-essential side notes, which may be helpful to those just starting off-

▪ ASM.COM will only work for 8080 syntax; it does NOT accept the z80 style syntax. the z80 is a the newer version of the 8080, therefore all binaries written for the 8080/8085 should run fine on the z80. being that the z80 has more op codes than the 8080 does, this only is guaranteed one way.
▪ Generally, its a good idea to add "RST 0" to the end of you code block, as without it the system will try to execute your data as code, and on the off chance that no error is found there, then it will continue falling through memory until it finds a "CALL 0", "JMP 0", or encounters an error. which is not ideal-
▪ Though optional with the ASM assembler, it is good practice to have an "END 100h" at the end of your program. This tells the assembler where the starting address should be, in most cases, for CP/M, you will want to start at the base of the TPA (i.e. 100h)

Please note that the most of this is verry specific to CP/M and the supplied ASM.COM assembler.

Slight necropost but I was browsing the forums and found this - thanks for stopping by and providing clarification!

It may be worth adding that CP/M 3 comes with MAC as its supplied assembler in place of CP/M 2's ASM. It's a macro assembler and does support the Z80 via the supplied Z80.LIB which defines the appropriate macros but this is a particularly cursed combination of Z80 features with Intel-esque mnemonics and syntax, e.g. LD (IX+D),NN would need to be written MVIX NN,D.

If you're going to be writing Z80 code on CP/M I still recommend Microsoft's M80 as that has a Zilog-compatible Z80 mode.