Implement SUBA

[marhel]

Please implement the instruction for Subtract Address. See implementation of ADDA. Also, please consider implementing related SUB-instructions at the same time. For more information, please read the instruction contribution guidelines

[emoon]

I will take this one

[emoon]

I'm trying to implement this but the test for suba_16_pd is failing. I wonder if Musashi may have a similar bug as ADDA (or was that bug for ADDA only?)

[emoon]

My code might of course be wrong but just want to check

[marhel]

I haven't looked, but if you take a look at the PR fixing adda, the problem should be explained in enough detail for you to check Musashis suba implementation

[marhel]

If pd and pi both fail (sometimes, as it is random based) but never the others, I'd say it's very likely there's a bug in Musashi suba as well

[emoon]

Yeah if you look at the mame code:

https://github.com/mamedev/mame/blob/master/src/devices/cpu/m68000/m68k_in.cpp#L9590

And the original

https://github.com/kstenerud/Musashi/blob/master/m68k_in.c#L9455

it's likely a bug in SUBA also

[emoon]

I'm running all the tests except pd and pi to verify that the current code work as expect with it.

[marhel]

When I have time, I can fix suba in Musashi locally, abd retest

[emoon]

That would be great.. Thanks!

I will hold of with my suba PR right now and do something else in between.

[marhel]

No, please do the PR, it will help me verify the fix to Musashi, without duplicating your work. If there's still something, you can always update the PR before I merge.

[emoon]

Done