|
|
Joined: May 2010
Posts: 92 Likes: 2
Member
|
Member
Joined: May 2010
Posts: 92 Likes: 2 |
Yeah, if you're on Windows you can even output to the built-in synth with -midiout "Microsoft GS Wavetable Synth". It sounds like crap, especially if the software expects an MT-32, but it should be enough for testing.
And, yes, other systems work. I just tried the ISA MPU-401 on one of the PC drivers and it's fine.
|
|
|
|
Joined: Mar 2001
Posts: 17,234 Likes: 259
Very Senior Member
|
Very Senior Member
Joined: Mar 2001
Posts: 17,234 Likes: 259 |
That 0,0 should indeed be 0,3. If there's some easy software to test (ideally something where I just boot it with the card attached and it's supposed to play MIDI) I can take a look.
|
|
|
|
Joined: Jan 2012
Posts: 891 Likes: 17
Senior Member
|
Senior Member
Joined: Jan 2012
Posts: 891 Likes: 17 |
It seems the problem is the ports are at even addresses. I tried uses 0x00ff for the mask but that doesn't work with the 32bit address space of the 9821 and 0x00ff00ff fails with the 16bit address space of the 9801rs. It'll probably need a trampoline.
Edit: oops, everything okay, the rs was crashing because it was mapping a missing device.
Last edited by crazyc; 06/09/18 06:38 PM.
|
|
|
|
Joined: Mar 2001
Posts: 17,234 Likes: 259
Very Senior Member
|
Very Senior Member
Joined: Mar 2001
Posts: 17,234 Likes: 259 |
Ahh, ok. That makes sense.
|
|
|
|
Joined: Jun 2001
Posts: 526 Likes: 37
Senior Member
|
Senior Member
Joined: Jun 2001
Posts: 526 Likes: 37 |
That 0,0 should indeed be 0,3. If there's some easy software to test (ideally something where I just boot it with the card attached and it's supposed to play MIDI) I can take a look. Why should it be 0,3 ? It was 32-bits mask 0x00ff little-endian before, that means address 0 only in my book. Hint: that was converted through a lua program. When a conversion program fucks up while you have 35K drivers, it shows :-)
|
|
|
|
Joined: Mar 2001
Posts: 17,234 Likes: 259
Very Senior Member
|
Very Senior Member
Joined: Mar 2001
Posts: 17,234 Likes: 259 |
It's (supposed to be) a 16 bit mask 0x00ff as per crazyc's fix, there are two addresses in use.
|
|
|
|
Joined: Jun 2001
Posts: 526 Likes: 37
Senior Member
|
Senior Member
Joined: Jun 2001
Posts: 526 Likes: 37 |
It's (supposed to be) a 16 bit mask 0x00ff as per crazyc's fix, there are two addresses in use. Oh cute. then a map(0, 3).umask16(0x00ff).(...) would work for both 16 and 32 bits. OG.
|
|
|
|
Joined: May 2010
Posts: 92 Likes: 2
Member
|
Member
Joined: May 2010
Posts: 92 Likes: 2 |
Another regression I've noticed after updating to the latest build... the floppy drives seem to have stopped working entirely. After some tests I think it's probably related to this commit (I guess because of the DMAC changes): https://github.com/mamedev/mame/commit/fbcc6c1b37266e328cb87028e380fb29841ba590I don't know if this affects other drivers, but this is a pretty common device, so it's likely.
|
|
|
|
Joined: Mar 2001
Posts: 17,234 Likes: 259
Very Senior Member
|
Very Senior Member
Joined: Mar 2001
Posts: 17,234 Likes: 259 |
I'd see if it's possible to fix PC98 to work with the changes, because they seem pretty reasonable hardware-wise.
|
|
|
|
Joined: Jan 2012
Posts: 891 Likes: 17
Senior Member
|
Senior Member
Joined: Jan 2012
Posts: 891 Likes: 17 |
That change also breaks the qx10.
Edit: Both use active low dreq. He changed the meaning of the request bits in the status register from logical request status to line level (but only partially) which breaks other things.
Last edited by crazyc; 10/28/18 10:13 PM.
|
|
|
1 members (Dorando),
196
guests, and
1
robot. |
Key:
Admin,
Global Mod,
Mod
|
|
Forums9
Topics9,328
Posts122,128
Members5,074
|
Most Online1,283 Dec 21st, 2022
|
|
These forums are sponsored by Superior Solitaire, an ad-free card game collection for macOS and iOS. Download it today!
|
|
|
|
|