On Sat, Jan 21, 2006 at 05:09:51AM +0000, Stuart Brady wrote:
> It seems to work, but needs more testing. There was one snapshot that
> I couldn't get working, but I've no idea why... perhaps a timing issue?
No, it turned out to be my buggy handling of the im and iff1/2. It now
works on everything I've tested so far, but I still need help testing.
> pop af
> ld r,a ; restore r
> jp po,0933 ; check bit 2 of f, skip if zero
> ei
I thought I'd got this wrong at first, but it seems I was right. PV is
reset iff parity is odd.
> + /* Byte 18 not used */
> + i = regs[19];
> + if (i == 00 || i == 0x3f) {
> + libspectrum_snap_set_im ( snap, 0 );
I think this should have been IM1, not IM0. Sound right?
> + } else {
> + libspectrum_snap_set_im ( snap, 2 );
> + }
> + libspectrum_snap_set_i ( snap, i );
> +
> + sp = (regs[20] + regs[21] * 0x100) & 0xffff;
> + libspectrum_snap_set_sp ( snap, (sp + 6) & 0xffff );
> +
> + libspectrum_snap_set_iff1 ( snap, buffer[sp + 0] & 0x02 ? 0 : 1 );
0x04 for the PV flag, not 0x02. I got the semantics reversed here, as
interrupts should be enabled iff PV is set.
> + libspectrum_snap_set_iff2 ( snap, 0 );
I think this should be set equal to iff1?
Is there anything that's wrong?
BTW, if anyone can understand this jump, I'd love to know:
out (fa),a ; lmpr
jp 0928 ; ???
0928: ld a,05 ; bcd 5
(I just realised, I got the jp and ld instructions the wrong way wround
in the disassembly. Sorry about that.)
Cheers,
--
Stuart Brady
