Trying Out ACK on MC6800 with emu6800

[zu2]

I'm porting ACK to the 8-bit MC6800 CPU for use with the emu6800 emulator. I'm using the cg code generator, but I'm having trouble adding the library, so runnable code isn't working yet. My current work is here: https://github.com/zu2/ack-6800
Any advice would be appreciated!

[zu2]

In the emu6800 environment (-me6800), support for stdin, stdout, and stderr has been added.
Currently, examples/hilo.c and paranoia.c are working.
On emu6800, no file system, so FILE I/O is unavailable.

$ ack -me6800 hilo.c
$ emu6800 6800 e6800.img /dev/null

Hi there! I'm written in C. Before we start, what is your name?
> zu2

Hello, zu2! See if you can guess my number.

> 50

Try a bit higher.

> 75

Try a bit lower.

> 67

You got it right in only 3 goes!

Would you like another go?
> 

[davidgiven]

Looks good so far! Is emu6800 small enough to be able to embed into the ACK source code? That way, it can be used to run the compiler tests.

What's the code quality like? I don't know a great deal about cg.

[zu2]

emu6800 source is at https://github.com/EtchedPixels/Fuzix-Compiler-Kit/tree/main/test.

Sizes:

  4853  20150 140642 6800.c
   306    913   7265 6800.h
   193    485   3491 emu6800.c
  5352  21548 151398 total

Performance matches typical MC6800 stack machine speed, which is decent for a first try.

cg/table pattern tweaks may improve speed by 10-20%.
Further improvements call for EM code reordering suited for MC6800, plus more reuse of register-held values.

I'll check if cg can support the latter.

[davidgiven]

Hello,

I was just reminded of this thread --- someone else asked me about a 6801 code generator... is your code in good enough state to merge? It'd be really nice to have, even with just cg.

Thanks!

[KristaMHill]

Hi, thanks for your latest note. I'm not yet up to speed wth the ACK, but I have a keen interest in the 6801 processor. I'll be paying attention to any posted comments.

[zu2]

Hi, @KristaMHill

If you're working with MC6801 or MC6803, the following compilers may be of interest:

• Fuzix-Compiler-Kit: features a working MC6803 compiler and emulator. It should also run on MC6801 without major issues.
• chibicc-6800-v1: an MC6800-targeted compiler supporting floating-point operations.

Hope this information is helpful for your project.

[davidgiven]

The order of the values on the stack shouldn't matter; the code which pops them off the stack again is in the control of the code generator and runtime library, so it can just be written to do it in the right order. Unless there's helper code which is trying to index specific bytes by offset, but it ought not to be doing that if the order matters. I'll try to fork your code and take a look, but unfortunately I don't know when I'll get time (too many projects, not enough time).

[zu2]

For example, in tests/plat/core/cii,e.e, pushing a 4-byte value (1) uses loc 0; loc 1 instead of ldc 1.

zu2@6ded164

Ah, I just remembered—I've seen similar stack byte position assumptions in Pascal set operations code generation. I'll investigate this further and report back.

[davidgiven]

Could you send me a PR, please? That will allow me to try it and make modifications while keeping your commits in the log.

[KristaMHill]

Hello, thanks for the references. This upcoming semester I'm teaching a computer devices lab course and I'm considering the 6801. I'll follow up and do some more digging. Krista

…

________________________________ From: zu2 ***@***.***> Sent: Friday, December 5, 2025 12:30 AM To: davidgiven/ack ***@***.***> Cc: Hill, Krista ***@***.***>; Mention ***@***.***> Subject: [External-email]: Re: [davidgiven/ack] Trying Out ACK on MC6800 with emu6800 (Discussion #331) **Caution: This email was sent from outside the university. Please exercise caution when clicking links, opening attachments, or replying. When in doubt, contact the helpdesk.** Hi, @KristaMHill<https://github.com/KristaMHill> If you're working with MC6801 or MC6803, the following compilers may be of interest: * Fuzix-Compiler-Kit<https://github.com/EtchedPixels/Fuzix-Compiler-Kit>: features a working MC6803 compiler and emulator. It should also run on MC6801 without major issues. * chibicc-6800-v1<https://github.com/zu2/chibicc-6800-v1>: an MC6800-targeted compiler supporting floating-point operations. Hope this information is helpful for your project. — Reply to this email directly, view it on GitHub<#331 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAYSHRUVFHQJX34D64FTJVD4AEKAHAVCNFSM6AAAAACBOQSZNSVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTKMJWHAZTINA>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[KristaMHill]

Please help me, what's a PR?

…

________________________________ From: David Given ***@***.***> Sent: Thursday, December 11, 2025 6:18 AM To: davidgiven/ack ***@***.***> Cc: Hill, Krista ***@***.***>; Mention ***@***.***> Subject: [External-email]: Re: [davidgiven/ack] Trying Out ACK on MC6800 with emu6800 (Discussion #331) **Caution: This email was sent from outside the university. Please exercise caution when clicking links, opening attachments, or replying. When in doubt, contact the helpdesk.** Could you send me a PR, please? That will allow me to try it and make modifications while keeping your commits in the log. — Reply to this email directly, view it on GitHub<#331 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAYSHRTQ6ZT3LULMBFYJSN34BFHJVAVCNFSM6AAAAACBOQSZNSVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTKMRTGAYTGOI>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[davidgiven]

Sorry, that was aimed at @zu2. It's a mechanism for allowing me to merge their changes into my repository via github. I want to see if I have any luck getting their 6800 backend upstreamed.

[zu2]

I’ve sent a PR. Please let me know if there are any issues.

[KristaMHill]

Thanks for clarifying that, not a problem 👌 Krista

…

________________________________ From: David Given ***@***.***> Sent: Thursday, December 11, 2025 10:38 AM To: davidgiven/ack ***@***.***> Cc: Hill, Krista ***@***.***>; Mention ***@***.***> Subject: [External-email]: Re: [davidgiven/ack] Trying Out ACK on MC6800 with emu6800 (Discussion #331) **Caution: This email was sent from outside the university. Please exercise caution when clicking links, opening attachments, or replying. When in doubt, contact the helpdesk.** Sorry, that was aimed at @zu2<https://github.com/zu2>. It's a mechanism for allowing me to merge their changes into my repository via github. I want to see if I have any luck getting their 6800 backend upstreamed. — Reply to this email directly, view it on GitHub<#331 (reply in thread)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAYSHRUBAIZ47LEJBEEXVSD4BGFX7AVCNFSM6AAAAACBOQSZNSVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTKMRTGIZTQMA>. You are receiving this because you were mentioned.Message ID: ***@***.***>