Updated Uniaud32 build

[Sigurd Fastenrath]

20210126 works fine with the Yoga 12.

thumbs up!!

[Alexander]

I have sblive CT4830 4-channels card. It supports multistream and it really works with old uniaud.

Check mixer controls with pmunimix.

I doubt that mixer controls can lead to msi error

Ok, now i found, whats wrong, its very old uniaud16
I take this...
::@@Universal Audio MMPM/2 Driver for OS/2
Vendor:          Netlabs
Revision:        1.9.7
Date/Time:       25 Jun 2017 17:11:40
Build Machine:   DAZAR1
File Version:    1.9.7
Description:     Universal Audio MMPM/2 Driver for OS/2
...and with it sound works ok with uniaud32-linux-5.10.10-20210126.zip

But mutistream is not working
I attach several logs:
uniaud0 - after boot, no wps sounds
uniaud1 - warpvision playing
uniaud2 - now i started pm123 and in says mci err 5613: insufficient buffer space specified
uniaud3 - warpvision closed and pm123 playing ok

Ps With old uniaud warpvision and pm123 playing simultaneously without any errors

[Paul Smedley]

Hi Alexander,

But mutistream is not working
I attach several logs:
uniaud0 - after boot, no wps sounds
uniaud1 - warpvision playing
uniaud2 - now i started pm123 and in says mci err 5613: insufficient buffer space specified
uniaud3 - warpvision closed and pm123 playing ok

Ps With old uniaud warpvision and pm123 playing simultaneously without any errors

Can you repeat the above test with https://smedley.id.au/tmp/uniaud32-20210128-Alexander.zip ? It reverts a change which allowed HDA hardware to produce sound, and which (I think) might be the cause of the issues with stream interruptions.

NB: for anyone else following along, the above build will NOT work with HDA hardware.

Cheers,

Paul

[Alexander]

Can you repeat the above test with https://smedley.id.au/tmp/uniaud32-20210128-Alexander.zip ? It reverts a change which allowed HDA hardware to produce sound, and which (I think) might be the cause of the issues with stream interruptions.

No, the same result

[Alexander]

This old uniaud version 1.1
uniaud16 says 1.1.2
uniaud32 - 1.1.0
29.12.04   2:20      61554             0  uniaud16.sys
29.12.04  19:52     479694            54  uniaud32.sys
ftp://icn@79.135.223.10/os2soft/uniaud/uniaud_latest.zi6

[Ian B Manners]

I see what Alexander means about mutistream not working.
I normally only play multimedia one at a time
If I play an mp3, or a movie via mplayer, then try to play something via pm123 I get "MCI Error 5613: Insufficient buffer size specified."
If I stop mplayer then pm123 works fine.
If I start playing music via pm123, then start something in mplayer, mplayer doesnt complain it just doesnt make any sound.

Tried it with uniaud32-20210128-Alexander.zip and there is no difference.
Log attached.

PS, I'm using uniaud16.sys (dated 25th Jun 2017, revision 1.09.7) which I assumed is the latest 16 driver.
PPS, thought I would try system sounds again. They work fine until I use pm123 or another mm application, then they nolonger work - ever. If I warm boot then test the system sounds (works), then start pm123, click on the system sounds Test button, I get no system sound. If I click a second time on the Test button, the audio from pm123 stops. To get audio back for pm123 I need to restart pm123. <== with uniaud32-20210128-Alexander.zip

[Martin Iturbide]

Hi Ian, Alexander.

Do you think this "multistream issue" is related (or the same) as this ticket? "Shared mode doesn't work"
Or are different things?

Regards

[Ian B Manners]

Hi Martin,

Speaking only for me, yes, I would say the issue is related if not the same. Though I'm not angry about it
I think the system sounds issue is also part of the same issue, ie, there is no input mixing happening.
As to a MM application resuming audio, or having to restart it I think would depend on the programming behind that application.

PS, same issue, installed Daniel's Mahjongg - also 100% CPU after pm123 audio stops about 5 seconds after starting Mahjongg. When I quit Mahjongg, take about another 5 seconds then pm123 audio starts again. Tried it a second time, pm123 is still playing ok start Mahjongg but there is no sound from Mahjongg.

[Paul Smedley]

Hi Guys,

Do you think this "multistream issue" is related (or the same) as this ticket? "Shared mode doesn't work"
Or are different things?

Not to contradict Ian, but I think they're related. Whilst Ian and Alexander have *hardware* that is capable of multiple streams, uniaud16 doesn't currently support it (noting that uniaud 1.1.2 seems to have supported multiple streams, and the code was commented out.

Shared mode is the alternative when multiple streams are NOT supported.

Cheers,

Paul

[Ian B Manners]

Hi Paul,

I think I'm looking at it from an Audio path perspective, your looking at it from a programming perspective

Programming how multiple streams are used and allocated and/or selecting a shared mode for inputs could very easily get complex in programming. Might be why the code was commented out as I suspect it would be easier to understand using a mixing console as a representation, a bit like how its done under Linux with Jack, Pulse Audio, and the MIDI connection PATCH panel which is neat. In the end it comes down to audio inputs from different physical and/or virtual sources, and what to do with them.

As Paul says, streams are to do with the inputs from the hardware, shared mode is the software side of it which I assume simply throws it altogether into the one L or R audio 'stream' as a basic software mixer. From what you've said this means that the shared mode is what is not working, and I've made the assumption that its all in the same code location.

I was thinking of also remming out Timidity and seeing how it worked with the MIDI drivers but haven't gone that far yet, now I think about the software, not sure if that's even important.

[Paul Smedley]

Actually, I think I disagree with myself now, after re-reading the ticket.

To me, shared mode was describing an brief interruption to an MP3 stream as a result of a system sound being played. There are reports that this is broken currently - ie the MP3 doesn't resume.

When I re-read the ticket, I think they're asking for sound mixing where (per the ticket) the Klondike sounds would play over the MP3 without any interruption a - ie a software version of hardware like the SB live that supports multiple streams. Afaik no version of uniaud has ever supported this on hardware that only supports a single stream (ie 99% of hardware).

I'll look into a uniaud16 build that re-enables the code for multiple streams over the weekend.

[Paul Smedley]

Hi All,

https://smedley.id.au/tmp/uniaud32-linux-5.10.11-20210131.zip

Fixed a problem in hda_codec.c that allowed me to remove a dirty hack from lib32/sound.c

This *might* help fix a reported problem with system sounds interrupting a MP3 stream and not resuming correctly, which is a regression from the 2.02 driver.

Cheers,

Paul

[Ian B Manners]

Hi Paul,

uniaud32-linux-5.10.11-20210131.zip
Installed, logs attached, one taken directly after a fresh boot, the other after the fault condition.

If I'm playing an MP3 file using pm123, then play a system sound (test was moving emails in pmmail folders) or via the Sound panel in System Setup, the mp3 will keep playing for x number of seconds before going silent. After this event there is no mp3 or other audio output. I can get the mp3 audio back by restarting play (without restarting pm123) but the system sounds are gone for good. Before playing the mp3, system sounds worked fine.

Bit like the pm123 buffer can no longer pull data because OS/2 has control of the queue after attempting to play a system sound.

PS, Apart from that, it works fine

[Doug Bissett]

This *might* help fix a reported problem with system sounds interrupting a MP3 stream and not resuming correctly, which is a regression from the 2.02 driver.

I find that it takes more system sounds to trigger the problem, but it still happens.

Run PM123, and play something. then open a window that has scroll bars. Click to move out of the window. It makes a sound. try that a few more times, allowing the sound to finish. That seems to work, at least for a few tries. Now, click fast enough to start a new sound, before the first sound has finished playing, and it will lock the desktop. The mouse won't even move. Reset is required.

Now, an interesting thing: If I restart the desktop (without locking it), sounds come back. Sometimes, it seems to be okay (until the next restart). Sometimes, it is worse than before.

When it locks, there is a POPUPLOG associated:

Code: [Select]

01-31-2021  12:51:57  SYS3175  PID 01b1  TID 0002  Slot 0060
C:\OS2\PMSHELL.EXE
c0000005
1f0aa7a5
P1=00000001  P2=00000008  P3=XXXXXXXX  P4=XXXXXXXX 
EAX=00000002  EBX=00000004  ECX=00000000  EDX=004bb82c
ESI=00000004  EDI=00381be8 
DS=0053  DSACC=d0f3  DSLIM=9fffffff 
ES=0053  ESACC=d0f3  ESLIM=9fffffff 
FS=150b  FSACC=00f3  FSLIM=00000030
GS=0000  GSACC=****  GSLIM=********
CS:EIP=005b:1f0aa7a5  CSACC=d0df  CSLIM=9fffffff
SS:ESP=0053:004bb8d8  SSACC=d0f3  SSLIM=9fffffff
EBP=004bb910  FLG=00010206

MMPM.DLL 0001:0007a7a5

There is also a process dump, which shows this:

Code: [Select]

IBM OS/2 Dump Formatter for a retail or an hstrict SMP kernel.
Formatter is --> Internal revision 14.106_SMP
Dump file is --> Internal revision 14.202_SMP (process dump)
 
Symbol (c:\os2\pdpsi\pmdf\warp45_s\os2krnlr.sym) linked
Symbol (c:\os2\pdpsi\pmdf\warp45_s\pmctls.sym) linked
Symbol (c:\os2\pdpsi\pmdf\warp45_s\pmwp.sym) linked
Symbol (c:\os2\pdpsi\pmdf\warp45_s\pmspl.sym) linked
Symbol (c:\os2\pdpsi\pmdf\warp45_s\pmviop.sym) linked
Symbol (c:\os2\pdpsi\pmdf\warp45_s\pmmerge.sym) linked
Symbol (c:\os2\pdpsi\pmdf\warp45_s\doscall1.sym) linked
 
Current slot number: 0060
 
 Slot  Pid  Ppid Csid Ord  Sta Pri  pTSD     pPTDA    pTCB     Disp SG Name
*0060# 01b1 001d 01b1 0002 run 0800 f705f000 fcf477d0 fce91b98 06c4 1a PMSHELL
 
eax=00000002 ebx=00000004 ecx=00000000 edx=004bb82c esi=00000004 edi=00381be8
eip=1f0aa7a5 esp=004bb8d8 ebp=004bb910 iopl=0 rf -- -- nv up ei pl nz na pe nc
cs=005b ss=0053 ds=0053 es=0053 fs=150b gs=0000 cr2=00000000 cr3=00000000 p=00
005b:1f0aa7a5 ff7108         push      dword ptr [ecx+08]   ds:00000008=invalid
#
Loading version data ... please wait
Loading Process Information ... please wait


Loading structure info for WSeB/ACP SMP ... please wait.
....Structure definition file (warp45_s\kernel.sdf) loaded.
....Structure definition file (warp45_s\shell.sdf) loaded.
Done.
 
#
                        TRAP SCREEN INFORMATION


------------------------------------------------------------
Trap screen 1 found at address #70:989c
This screen is most likely related to a previous trap.

P1=00000001  P2=00000008  P3=XXXXXXXX  P4=XXXXXXXX   
CS:EIP=005b:1f0aa7a5  CSACC=d0df  CSLIM=9fffffff
SS:ESP=0053:004bb8d8  SSACC=d0f3  SSLIM=9fffffff
EBP=004bb910  FLG=00010206
EAX=00000002  EBX=00000004  ECX=00000000  EDX=004bb82c
ESI=00000004  EDI=00381be8   
DS=0053  DSACC=d0f3  DSLIM=9fffffff   
ES=0053  ESACC=d0f3  ESLIM=9fffffff   
FS=150b  FSACC=00f3  FSLIM=00000030
GS=0000  GSACC=****  GSLIM=*******
 
No Symbols Found
#

------------------------------------------------------------
Trap screen 2 found at address #70:9c84

P1=00000001  P2=00000008  P3=XXXXXXXX  P4=XXXXXXXX   
EAX=00000002  EBX=00000004  ECX=00000000  EDX=004bb82c
ESI=00000004  EDI=00381be8   
DS=0053  DSACC=d0f3  DSLIM=9fffffff   
ES=0053  ESACC=d0f3  ESLIM=9fffffff   
FS=150b  FSACC=00f3  FSLIM=00000030
GS=0000  GSACC=****  GSLIM=********
CS:EIP=005b:1f0aa7a5  CSACC=d0df  CSLIM=9fffffff
SS:ESP=0053:004bb8d8  SSACC=d0f3  SSLIM=9fffffff
EBP=004bb910  FLG=0001020

ASCII found at #70:9c84-12 is c0000005 (Exception Code?)
c0000005:  System Encountered an access violation

ASCII found at #70:9c84-9 is 1f0aa7a5. (Linear Address?)

ASCII found after trap screen: (Module name Object:Offset?)
MMPM.DLL 0001:0007a7a5
No Symbols Found
# 
 
#
The dump is about 2.5 GB. I can get it to you, if you think it will help.

I am doing this on my Asus A88XN-A machine (AMD A6 processor), but I get similar results on my Lenovo ThinkPad L530 (Intel based).

[Paul Smedley]

Hi Doug,

This *might* help fix a reported problem with system sounds interrupting a MP3 stream and not resuming correctly, which is a regression from the 2.02 driver.

I find that it takes more system sounds to trigger the problem, but it still happens.

Thanks for the response and the notes on how to reproduce. I'll try and do some testing locally this week - and also see if I can reproduce with the 2.02.06 driver.

No need for a dump right now - I'm not sure I'd know what to do with it anyway

Cheers,

Paul