let me know if you need me to test on my thinkpad p53

crazii wrote on 2023-04-04, 03:21:

tercio was using the temp release version that fixes Epic Pinball, did you use that too? I tested the game of ver 2.1, using /t5 and SB Pro in setup, and no problems found on my TP T540p, i7 4710MQ Haswell CPU. I'll do more test later.

Epic Pinball worked fine on my HDA. Several versions of SBEMU ago, it had sped-up audio, but that's fixed in the latest betas.

It's OMF:2097 which still gives me headaches, in spite of detecting and configuring well... hard freeze immediately after game executable is run.

th3bar0n wrote on 2023-04-01, 20:43:

th3bar0n wrote on 2023-03-25, 12:19:

I have a Vt8233 in an HP T5520 thin client. I don't get any of the DMA conflict errors, but otherwise similar - MPXPlay works perfectly, but with SBEMU beta 2 no sound at all and with the VT8233FIX build..

crazii wrote on 2023-03-31, 12:20:

Thanks folks! I found no laptop in my local market, I decided to buy a HP T5520 instead. 😁

Excellent choice! 😁

Bad news: the T5520 is out of stock and I have to check other specs. 🤣
There's some T5000 in stock but I cannot find the audio spec for it: https://www.parkytowers.me.uk/thin/hp/t5000/

EDIT: I ordered a T5530 which has VT8233/8235/8237, that identical to T5135, according to this page: https://www.parkytowers.me.uk/thin/hp/t5135/

Baron von Riedesel wrote on 2023-04-02, 08:29:

Anyway, I added an optional numeric argument to hdpmi's -x option, working as "right shift" factor, so -x still restricts mem to 256mb, but -x1 changes that to 128mb, -x2 to 64mb and so on... System Shock is happy with -x2.

I merge the -x[n] code from your repo and I tested with system shock, it still exited on "not enough memory to make email cursor", weird. I then tested with Quake, the quake banner shows exact free memory according to the -x[n] option on each change of the [n] value, that means the options is working. Is there any additional modification needed?

zyzzle wrote on 2023-04-04, 06:31:

crazii wrote on 2023-04-04, 03:21:

tercio was using the temp release version that fixes Epic Pinball, did you use that too? I tested the game of ver 2.1, using /t5 and SB Pro in setup, and no problems found on my TP T540p, i7 4710MQ Haswell CPU. I'll do more test later.

Epic Pinball worked fine on my HDA. Several versions of SBEMU ago, it had sped-up audio, but that's fixed in the latest betas.

It's OMF:2097 which still gives me headaches, in spite of detecting and configuring well... hard freeze immediately after game executable is run.

Yeah, I fixed the spedup problem of epic pinball and also some bugs of the emulation, that might help with the OMF sounds. The games runs on my T540p but games speed is super fast and I got K.O.ed by the computer within seconds. 🤣
there should be unknown compatibility issues for your PC, but it might be hard to find out, maybe disabling some onboard devices or lower the CPU speed helps.

Any chance you can look into Tyrian? It hates a lot of setups (including mine) for sound. It would be nice to have the option to play it natively.

ZellSF wrote on 2023-04-04, 11:06:

Any chance you can look into Tyrian? It hates a lot of setups (including mine) for sound. It would be nice to have the option to play it natively.

As Baron von Riedesel last checked, it is a 16bit protected mode game that currently not easy to support: Re: Possibility to write OPL3 sound driver for protected mode games

EDIT: I think it's possible to write a 16bit protected mode wrapper that calls the 32bit TSR as interrupt service: just trapping the same ports and pass the commands, DMA addresses to the 32bit TSR and let it do the actual job. Thus recompilation of the soundcard drivers and OPL3 emulation can be skipped.

crazii wrote on 2023-04-04, 11:28:

EDIT: I think it's possible to write a 16bit protected mode wrapper that calls the 32bit TSR as interrupt service: just trapping the same ports and pass the commands, DMA addresses to the 32bit TSR and let it do the actual job. Thus recompilation of the soundcard drivers and OPL3 emulation can be skipped.

It is doable but one have to make sure no 16-bit TSS is used anywhere since that leads apart other constraints to lack of I/O trapping capability.
What about Gravis emulation?

crazii wrote on 2023-04-04, 11:06:

Yeah, I fixed the spedup problem of epic pinball and also some bugs of the emulation, that might help with the OMF sounds. The games runs on my T540p but games speed is super fast and I got K.O.ed by the computer within seconds. 🤣
there should be unknown compatibility issues for your PC, but it might be hard to find out, maybe disabling some onboard devices or lower the CPU speed helps.

Thanks for testing OMF. Yes, I already had lowered CPU speed from 3.4 Ghz to 1.6 Ghz, but thanks to your tip, I'll use CPU slowdown utility posted at VOGONS, which can slow (throttle) further from 1/16 to 16/16 max. Perhaps trying 1/16 will make it run with sound. Could also be memory problem with HDPMI32, but game works without sound just fine.

Anyway, I've other laptops (using HDA also) that I can try it on. As is often said "it's good to have choices."

crazii wrote on 2023-04-04, 09:24:

I merge the -x[n] code from your repo and I tested with system shock, it still exited on "not enough memory to make email cursor", weird. I then tested with Quake, the quake banner shows exact free memory according to the -x[n] option on each change of the [n] value, that means the options is working. Is there any additional modification needed?

Not with the version I tried - it's v1.6, located as a bonus on an CD inserted in a computer magazine.

Crazy Stuff, will try it on my Pentium M Laptop

crazii wrote on 2023-04-04, 09:17:

Bad news: the T5520 is out of stock and I have to check other specs. lol There's some T5000 in stock but I cannot find the audio […]
Show full quote

th3bar0n wrote on 2023-04-01, 20:43:

th3bar0n wrote on 2023-03-25, 12:19:

I have a Vt8233 in an HP T5520 thin client. I don't get any of the DMA conflict errors, but otherwise similar - MPXPlay works perfectly, but with SBEMU beta 2 no sound at all and with the VT8233FIX build..

crazii wrote on 2023-03-31, 12:20:

Thanks folks! I found no laptop in my local market, I decided to buy a HP T5520 instead. 😁

Excellent choice! 😁

Bad news: the T5520 is out of stock and I have to check other specs. 🤣
There's some T5000 in stock but I cannot find the audio spec for it: https://www.parkytowers.me.uk/thin/hp/t5000/

EDIT: I ordered a T5530 which has VT8233/8235/8237, that identical to T5135, according to this page: https://www.parkytowers.me.uk/thin/hp/t5135/

Ha! Oh well, that should be ok too according to this,
AC97 DOS drivers, yes, but no

The (Compaq) T5000 is the same as the HP T5710 which has a VT82C686 soundcard - I also have one of these and tested mine this weekend, all worked flawlessly on my set of test games.

@crazii don't you have (or had before) a T5710? You posted about one back in 2021 Re: HP T5710 windows 98 build

crazii wrote on 2023-04-04, 02:23:

It's mainly used for detection of the SB card (not all games uses that), when you send the command to the port, and wait for the IRQ as response, if it happens, then there's a huge possibility that the SB card exist, otherwise there's no SB card exist.
That's why Dynablaster had no sound, it cannot detect a SB card and so stopped doing sound functions at all.

Aha, ok I understand now. It's like INT instruction to trigger interrupt other way than from HW...
BTW I tried to run Dynablaster with Moslo slow-down utility with old SBEMU and it didn't help, any speed derating I tried.

Maybe it's similar reason why also Boost demo doesn't play any sound. It's pmode program (works with dsdma.exe).
https://www.pouet.net/prod.php?which=6485

stamasd wrote on 2023-04-04, 20:31:

@crazii don't you have (or had before) a T5710? You posted about one back in 2021 Re: HP T5710 windows 98 build

Yes I got a T5710 which has VT82C686 chip, but now I need to debug the VT8233, which is reported not working by many people, so I bought another thin client.

Crazii, if you need some money for buying degug hardware, I think it's okay to send me your PayPal email. I want to support in buying this hardware!

MBH wrote on 2023-04-05, 06:58:

Crazii, if you need some money for buying degug hardware, I think it's okay to send me your PayPal email. I want to support in buying this hardware!

I appreciate your kindness but I don't need money, I just bought too many pieces of hardware that I told myself to stop stockpiling. 🤣

Good News!
With the HP T5530, I fixed the bug for VT833 and now it's working - just overlooked a tiny bug.
Tested with a few games all working, except that duke3d will freeze on exit, I don't think it's caused by SBEMU, probably VGA related, but I forgot to verify it without SBEMU, you guys can find it out.

RayeR wrote on 2023-04-04, 22:57:

Aha, ok I understand now. It's like INT instruction to trigger interrupt other way than from HW... BTW I tried to run Dynablast […]
Show full quote

crazii wrote on 2023-04-04, 02:23:

It's mainly used for detection of the SB card (not all games uses that), when you send the command to the port, and wait for the IRQ as response, if it happens, then there's a huge possibility that the SB card exist, otherwise there's no SB card exist.
That's why Dynablaster had no sound, it cannot detect a SB card and so stopped doing sound functions at all.

Aha, ok I understand now. It's like INT instruction to trigger interrupt other way than from HW...
BTW I tried to run Dynablaster with Moslo slow-down utility with old SBEMU and it didn't help, any speed derating I tried.

Maybe it's similar reason why also Boost demo doesn't play any sound. It's pmode program (works with dsdma.exe).
https://www.pouet.net/prod.php?which=6485

I tested the boost demo it but didn't see any SB port IO in the log. weird.

georgel wrote on 2023-04-04, 12:24:

crazii wrote on 2023-04-04, 11:28:

EDIT: I think it's possible to write a 16bit protected mode wrapper that calls the 32bit TSR as interrupt service: just trapping the same ports and pass the commands, DMA addresses to the 32bit TSR and let it do the actual job. Thus recompilation of the soundcard drivers and OPL3 emulation can be skipped.

It is doable but one have to make sure no 16-bit TSS is used anywhere since that leads apart other constraints to lack of I/O trapping capability.
What about Gravis emulation?

I think the 16 bit protected mode support could be done later. When writing SBEMU, I have to search & learn documents about SB to make it work. I think Gravis would be the same, I believe the documents/programming guides can be found on the web, but I'm not ready to to that right now. Maybe later when SBEMU is more stable & compatible. It'll be good if you can help adding the GUS support 😁.