Steve Fosdick

These things can be interesting. I don't have a real Z80 machine either but perhaps we can compare with one of the mainstream CP/M emulators. A couple spring to mind...

That's very interesting about not getting the same executable at the end. I did have this previously with an assembler in a similar situation. I went looking for bug in...

Ok, the executable from the CPM.zip file from JGH's website runs fine on YAZE.

I did try looking at some individual instructions but was not getting far. I have come up with a plan which is to trace the execution of the offending tests....

I have fixed at least one of the errors - see https://github.com/stardot/b-em/commit/c3b2ec58ffd5f7893e9e1349482536bf18f98794 For some reason the actual logical operation for AND, OR and XOR was not being done but only...

The YAZE version of the flag setting for 8bit INC/DEC looks strange, but passes: ``` case 0x04: /* INC B */ BC += 0x100; temp = hreg(BC); AF = (AF...

So commenting the YAZE version: ``` case 0x04: /* INC B */ BC += 0x100; temp = hreg(BC); /* just B */ AF = (AF & ~0xfe) | (temp &...

So I think I can see one bug in the b-em version - znptable should be znptablenv, i.e. we don't want the parity in the V/P flag as we're using...

Continuing on INC/DEC, you're right. I had that thought after I went to bed. Comparing a run of the test suite is proving to generate too much data - even...

Thanks, David. SO my simple test case for INC/DEC presumably didn't find the 'N' flag issue because it was already clear. I have got as far as CCF/SCF, just working...