Forums

Hello all.

Is there anything in Marat's MZ80 core that would prevent akidd1 from working?

I'm seeing a write to port 0xdf, a write to port 0xde, a read from port 0xde, a write to port 0xed, and a bunch of writes to port 0x44. It appears to loop at some point, and the display is messed. No response to the controller, either.

Thanks in advance.

--Nyef

Quote

> Hello all.
> Is there anything in Marat's MZ80 core that would prevent akidd1 from working?
> I'm seeing a write to port 0xdf, a write to port 0xde, a read from port 0xde, a write to port 0xed, and a bunch of writes to port 0x44. It appears to loop at some point, and the display is messed. No response to the controller, either.

Doesn't seems to be related to Marat's core. Some games does write to 0xDE and 0xDF though.

Quote

> > Hello all.
> > Is there anything in Marat's MZ80 core that would prevent akidd1 from working?
> > I'm seeing a write to port 0xdf, a write to port 0xde, a read from port 0xde, a write to port 0xed, and a bunch of writes to port 0x44. It appears to loop at some point, and the display is messed. No response to the controller, either.

> Doesn't seems to be related to Marat's core. Some games does write to 0xDE and 0xDF though.

Joy. So it's _probably_ not the CPU core, unless I managed to screw it up completely.

That leaves "everything else", although in practice it isn't likely to be the PSG.

Looks like I have quite the bughunt ahead of me, then. :-/

--Nyef

Quote

> Joy. So it's _probably_ not the CPU core, unless I managed to screw it up completely.

There _are_ games that have problems due to Marat's core, but Alex Kidd is not one of them. The primary problem with Marat's engine is that there's a bug in the BIT instruction (which is easily fixible by replacing the BIT macro in his with the BIT macro from Marcel De Kogel's), and the fact that it doesn't emulate the interrupt line, which is less easy to fix.

The BIT bug manifests itself in R-Type on the 3rd stage, Simpsons vs. the Space Mutants, and some others I'm sure, but not on Alex Kidd.

Quote

> That leaves "everything else", although in practice it isn't likely to be the PSG.

Can't be the PSG - it's write-only, and the games have no way of knowing if it's emulated at all. Anyhow I had Alex Kidd running before I added sound as well, so that shouldn't be a problem.

I don't handle writes to ports $DD or $44, and it workes fine in CHASMS... $DC/DD is the controller port, $DE is a pseudo mysterious port that I presume has something to do with lightgun but damned if I know what. Only thing I can think off offhand is that it's writing to $DE maybe and then reading it to check a result, and maybe you're returning 0 instead of 255 for an undefined port ? Idunno. I've never even heard of port $ED being used for something, so Idunno.

Are you just recently switching to the Marat core? Do other games work? If not, maybe you forgot to define the LITTLE_ENDIAN or whatever. That'll easily mess things up.

Seems like it has to be something relatively simple. VDP is for the most part write-only, so it seems like it'd have to be a port or memory-map mishandling or something. Idunno.

- vecna

Quote

> Seems like it has to be something relatively simple. VDP is for the most part write-only, so it seems like it'd have to be a port or memory-map mishandling or something. Idunno.

It turned out to be a 64-byte footer on the image. I'm now trying to think up a suitable warning message to use when I find a ROM with this kind of footer...

How long have these footers been around, anyway?

Quote

> - vecna

--Nyef

Quote

> It turned out to be a 64-byte footer on the image. I'm now trying to think up a suitable warning message to use when I find a ROM with this kind of footer...

> How long have these footers been around, anyway?

.... Heh. After I released the 0.20 version of CHASMS I got an email asking that I add support for ROMs with 64-byte footers. I had never seen one, and the guy couldn't send me one for whatever reason, but he was adamanent they existed so I modified the ROM loader to work with 'em in the 0.30 version. To this day I've never actually seen one, though. Sigh. ^_~

- vecna

All the very first roms appearing on the internet had originally that footer. I could list them if someone is interested.
It contains the full game name, by the way.

Author	Message
Nyef Guest	Alex Kidd 1 problems Posted: Fri Aug 18, 2000 1:18 am
Nyef Guest	Hello all. Is there anything in Marat's MZ80 core that would prevent akidd1 from working? I'm seeing a write to port 0xdf, a write to port 0xde, a read from port 0xde, a write to port 0xed, and a bunch of writes to port 0x44. It appears to loop at some point, and the display is messed. No response to the controller, either. Thanks in advance. --Nyef

Zoop Joined: 24 Jun 1999 Posts: 1732 Location: Paris, France	Posted: Fri Aug 18, 2000 6:41 am
	Quote > Hello all. > Is there anything in Marat's MZ80 core that would prevent akidd1 from working? > I'm seeing a write to port 0xdf, a write to port 0xde, a read from port 0xde, a write to port 0xed, and a bunch of writes to port 0x44. It appears to loop at some point, and the display is messed. No response to the controller, either. Doesn't seems to be related to Marat's core. Some games does write to 0xDE and 0xDF though.

Nyef Guest	Posted: Fri Aug 18, 2000 3:03 pm
Nyef Guest	Quote > > Hello all. > > Is there anything in Marat's MZ80 core that would prevent akidd1 from working? > > I'm seeing a write to port 0xdf, a write to port 0xde, a read from port 0xde, a write to port 0xed, and a bunch of writes to port 0x44. It appears to loop at some point, and the display is messed. No response to the controller, either. > Doesn't seems to be related to Marat's core. Some games does write to 0xDE and 0xDF though. Joy. So it's _probably_ not the CPU core, unless I managed to screw it up completely. That leaves "everything else", although in practice it isn't likely to be the PSG. Looks like I have quite the bughunt ahead of me, then. :-/ --Nyef

vecna Guest	Posted: Sat Aug 19, 2000 2:18 am
vecna Guest	Quote > Joy. So it's _probably_ not the CPU core, unless I managed to screw it up completely. There _are_ games that have problems due to Marat's core, but Alex Kidd is not one of them. The primary problem with Marat's engine is that there's a bug in the BIT instruction (which is easily fixible by replacing the BIT macro in his with the BIT macro from Marcel De Kogel's), and the fact that it doesn't emulate the interrupt line, which is less easy to fix. The BIT bug manifests itself in R-Type on the 3rd stage, Simpsons vs. the Space Mutants, and some others I'm sure, but not on Alex Kidd. Quote > That leaves "everything else", although in practice it isn't likely to be the PSG. Can't be the PSG - it's write-only, and the games have no way of knowing if it's emulated at all. Anyhow I had Alex Kidd running before I added sound as well, so that shouldn't be a problem. I don't handle writes to ports $DD or $44, and it workes fine in CHASMS... $DC/DD is the controller port, $DE is a pseudo mysterious port that I presume has something to do with lightgun but damned if I know what. Only thing I can think off offhand is that it's writing to $DE maybe and then reading it to check a result, and maybe you're returning 0 instead of 255 for an undefined port ? Idunno. I've never even heard of port $ED being used for something, so Idunno. Are you just recently switching to the Marat core? Do other games work? If not, maybe you forgot to define the LITTLE_ENDIAN or whatever. That'll easily mess things up. Seems like it has to be something relatively simple. VDP is for the most part write-only, so it seems like it'd have to be a port or memory-map mishandling or something. Idunno. - vecna

Nyef Guest	Posted: Tue Aug 22, 2000 1:28 am
Nyef Guest	Quote > Seems like it has to be something relatively simple. VDP is for the most part write-only, so it seems like it'd have to be a port or memory-map mishandling or something. Idunno. It turned out to be a 64-byte footer on the image. I'm now trying to think up a suitable warning message to use when I find a ROM with this kind of footer... How long have these footers been around, anyway? Quote > - vecna --Nyef

vecna Guest	Posted: Tue Aug 22, 2000 2:49 am
vecna Guest	Quote > It turned out to be a 64-byte footer on the image. I'm now trying to think up a suitable warning message to use when I find a ROM with this kind of footer... > How long have these footers been around, anyway? .... Heh. After I released the 0.20 version of CHASMS I got an email asking that I add support for ROMs with 64-byte footers. I had never seen one, and the guy couldn't send me one for whatever reason, but he was adamanent they existed so I modified the ROM loader to work with 'em in the 0.30 version. To this day I've never actually seen one, though. Sigh. ^_~ - vecna

Zoop Joined: 24 Jun 1999 Posts: 1732 Location: Paris, France	Footers Posted: Tue Aug 22, 2000 7:39 pm
	All the very first roms appearing on the internet had originally that footer. I could list them if someone is interested. It contains the full game name, by the way.

Forums

View topic - Alex Kidd 1 problems