Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Lars

Pages: [1] 2 3 ... 14
I would love to believe that. However I also have these problems when I don't run Firefox at all.
But again, memory is the problem. When I run EPM.EXE and subsequently close it, I can observe that it leaves an endless sequence of memory "holes" in upper shared memory. That is due to the fact that the DLL that it loads (ETKE603.DLL) specifies trillions of small memory objects (it contains 234 memory objects: who compiled and linked that stuff ?). And for each memory object, 64 kByte of memory is allocated (even though they are considerably smaller).
When EPM is closed the DLL is unloaded but the fragmentation remains.

Are you sure it is a Firefox problem ?

Sometimes I have reboots if I just open the Master volume control (volume.exe). And I had it on other occasions starting applications. But I also had occasions where I would get a "page fault" exception in Ring 0 (for which I generated a system dump).
I think there is a fundamental flaw in some very basic system component. That could be the OS virtual memory manager. But it can also be DOSCALL1.DLL as that is used by about any process that is started.
I think the virtual memory manager is the more likely culprit. And it much looks like it is a serialization issue (access not properly serialized on updating memory structures etc.) as it only happens occasionally.
If you have Theseus: run "getws.exe" and watch your system reboot ...

Comments, Suggestions & Questions / Re: Censorship complaints
« on: February 23, 2018, 09:25:59 am »

I dare to disagree. I am losing more and more interest in this forum because there is so much "white noise". And I am a native german speaker and often enough I don't understand one single word of what Andreas is trying to say (also when he answers on the german forum).
Especially if the real solution to the problem of the original OP is very easy to answer. Andreas tends to flood people with information they have not asked for.

Article Discussions / Re: OS2VOICE RPM Server
« on: February 21, 2018, 10:21:11 pm »
WIC.EXE can do the same from a commandline.

Article Discussions / Re: OS2VOICE RPM Server
« on: February 21, 2018, 03:26:05 pm »
I don't know. But why don't you look here:

? That should give you an idea, at least. I think everybody has read-only access to that directory ...

Article Discussions / Re: OS2VOICE RPM Server
« on: February 21, 2018, 11:12:53 am »
Just to add: everybody can always extract the original WIS installation script file from a WPI archive. It's ASCII plain text and the syntax well explained in the documentation that comes with WarpIn.

That is, even for unmaintained WPI packages:
1) we can fix the installation by extracting the original WIS script from the WPI, modifying it to our liking and putting it back into the WPI
2) we can deduce an uninstallation script (Rich's suggestion) from what the original WIS installation script does

The only thing that has always been a mess is that with WarpIn you can easily screw up your config.sys to a point where after uninstallation of a WPI the system will no longer boot. See the FAT32 WPI.
Short rule: NEVER EVER have a WPI script modify PATH or LIBPATH with subdirectory paths that you can expect to exist (like \OS2 or \OS2\DLL).

Hardware / Re: Limitation on using an USB DVD and CD writer under ArcaOS
« on: February 13, 2018, 10:25:43 am »
1) so, USBMSD.ADD seems to do its job and can therefore be ruled out

2) cdrecord needs a DVD/CD writer device to appear as an SCSI device. That typically requires 2 drivers:
    a) OS2ASPI.DMD (/ALL) which comes with OS/2
    b) a proprietary router driver that routes to OS2ASPI.DMD, normally that would be ASPIROUT.SYS from Paul Ratcliffe. Also there is aspinikk.sys (or something like that) that does about the same. However, for
reasons yet unknown the later sometimes fails to load on bootup (at least on my system).

3) I seem to remember that at least DANIATAP.FLT had a switch /RSJ that would make visible a DCD/CD-ROM drive that would otherwise be claimed exclusively by RSJ (at least while writing). But DANIATAP.FLT does not come into play for USB attached DVD/CD-ROM drives (is DANIATAP.FLT still necessary for CD-ROM/DVD drives attached via AHCI ?).


Hardware / Re: Limitation on using an USB DVD and CD writer under ArcaOS
« on: February 12, 2018, 02:00:39 pm »
Have you tried my USBMSD.ADD instead of the one that AN delivers ?  If not, you could give it a try.
If it still does not work, a trace of USBMSD.ADD would help (it's all explained in the "readusb.txt" file that comes with my package).

In principle the USBMSD.ADD (at least my version) should support SCSI "write" commands for USB DVD/CD drives. It should also support any DVD/CD specific SCSI commands.

Hardware / Re: USB 10.221 is out
« on: February 09, 2018, 04:35:47 pm »
Thank you Lars, I now have sound on several of my computers using a USB adapter.

One question, since you are the expert on USB drivers, is it possible to update the USBKBD.SYS to work with the extra keys on many new USB keyboards?  I have an old no name PS/2 that allows the use of those keys but non of the USB keyboards have any response (the keycodes don't appear in the programs I used to get them for the PS/2 keyboard.

Being called the "expert" is always a minefield  :)

Yes it would be possible for sure. Unfortunately, the HID spec (the specific spec for human input USB devices) is complex compared to the other specs and then I don't have the HW knowledge as to what special keys generate what scancodes. I would have to go through trial and error and I don't have the time for that.
There is some code and definitions currently commented out of USBKBD.SYS (don't know why. I did not touch that). Maybe it's worth putting that back in and giving it a try ...

Hardware / USB 10.221 is out
« on: February 06, 2018, 09:58:53 pm »

It contains fixes to USBD.SYS and more importantly the completely updated USBAUDIO package to support higher sample rates and bit resolutions.

Thanks for the explanation, Lars. I guess I thought conversion was going on, and these are old USB speakers, so makes sense they don't support 24 bit. I actually prefer there is no conversion - now I just need to get some USB speakers that support 24 bit. Is there a tool you know of that can report what speakers are capable of?


There was a good reason to not put conversion into the driver (apart from the stated exceptions) :
The first guy asks: can't you make 24-bit work ?
The second guy asks: can't you downsample from 64000 Hz to 32000 Hz ?
The third guy asks: can't you make it play MIDI ? I thought it was an audio driver !

You could go on forever. I would have no chance to ever get ahead of the game. The correct solution is to move the conversion into either an application or into an MMIO proc (like USBWAV.DLL). It's also more stable if the conversion code traps: a driver is running at Ring0, an application / MMIO proc is running at Ring3 ...

Yes, there is a tool called 'lsusb' and there also exists an OS/2 port of it. I think the latest version can be found on Paul Smedley's website. 'lsusb' will show you detailed USB device info for each device. That also includes the supported bit-depths and sample rates of your USB audio devices.

From practical experience I can say that normally, if a device supports 24-bit (or even 32-bit) it's also sophisticated enough to support 16-bit.

Thanks Lars! This seems to work with regular MP3 and WAV files, but I can't play any 24 bit WAV files (by double-clicking on them). I get an error saying 'The device driver cannot be opened or the required hardware is missing (rc=5006)'. Files are 24bit, 48KHz, 2 channel and are properly recognized if I right click on it and choose 'Open as..' -> 'Wavefile Info'. Far as I can tell the USBWAV codec is installed (using listprocs). Is this a limitation of the OS/2 player? What do you or Wim use to play 24 bit files?

 Still can't switch back and forth from Uniaud to USBWav, probably because there is no AmpMix page for UniAud. Unchecking the 'Default Device' under the USB Ampmix page doesn't stick for some reason. Once checked, USB is all I get until I go into MMPM2.INI and remove it from the [defaultnames] section. Can the Ampmix page for UniAud be added manually by tweaking MMPM2.INI? I'll play around with that...

 One other thing that seems obvious in retrospect... I uninstalled the old USB audio stuff using this package before re-installing, but the USB sequencer stuff was not removed (because it doesn't install it!). Have to use the last package (USBDrv220) to do it right.


1) "cannot play 24-bit": the error messages states exactly what happens: Your USB audio HW does not support replay of 24-bit. I decided to not add additional conversion code into USBAUDIO.SYS (apart from support for "legacy" sample rates and bit sizes of 8 kHz,11.025kHz, 22.050kHz and 8-bit respectively). We let the HW do what the HW can do but no more. You can use PM player which will do 24-bit and 32-bit to 16-bit conversion in SW. I could add that to USBWAV.DLL but currently it's not in there.

2) installation woes: unfortunately, audio installation is a bit tricky. The minstall.exe uses the info in CARDINFO.DLL in order to decide what to install or uninstall (which is ok and the right way to go). As you found out, the new CARDINFO.DLL no longer contains the "sequencer" section because USBAUDIO.SYS has never and probably will never support MIDI. That's the reason why you cannot uninstall it. If everything else fails, you can always copy the "virgin" MMPM2.INI from \MMOS2\INSTUNST\INIBACK directory to \MMOS2 and rerun the installation.
In hindsight I should have never delivered a USBAUDIO installation that installs the sequencer device ...

3) you have to "toggle" the "AMPMIX as default" setting (deselect and then reselect). That will finally create the [defaultnames] section in MMPM2.INI with WMBAMPMIX01 set as default. You do likewise for the WAVEAUDIO device.

As you now have found out, setting the DEVICEFLAG from 2 to 1 in the [UNIAUDAMPMIX] section (and then giving it a sensible ALIASNAME for displaying on the settings tab) will get you the AMPMIX page in the Multimedia Setup object (and the USBAUDIO installation does that  through CARDINFO.DLL for its own installation). I think the UNIAUD installation should be updated but oh well ...


I now uploaded USB 10.221 to Hobbes. I changed the USBAUDIO installation to no longer install a Sequencer (MIDI) device as USBAUDIO.SYS does not support MIDI. But I changed the installation so that the Ampmixer will now get its own page in the Multimedia Setup Object. That allows to not only select default WAVEAUDIO but now also default AMPMIX device and consequently it should now be possible to force WAVE play/record via USB audio.

Article Discussions / Re: AN changing USB driver stack to 32-bit
« on: January 29, 2018, 10:19:44 am »
Did you test the new 32-bit driver with USBECD.SYS? Maybe it still works?

I have not tested. Looking at the USBECD.SYS source code, it expects the "standard" interface to USBD.SYS to register itself with USBD.SYS. Then, it offers the "standard" IDC entry point for being called back by the HC drivers on request termination ("standard" = as defined in the HTML document that is part of the DDK).
That is, it behaves like any other class device driver. Currently, these interfaces have not been touched and therefore it should still work.

I am talking about if AN decides to change these interfaces and making them proprietary.

Article Discussions / Re: AN changing USB driver stack to 32-bit
« on: January 28, 2018, 10:18:34 am »
Guessing the change was made to make future development easier. For AN. Not for the rest of the world. Except David make his changes and his DDK publicly available some day.

If what you say is true I sincerely hope that at least the interface will be published. Because if not, nobody will be able to write a custom USB driver for a specific USB device, for example. Again, think about Wim's USBECD.SYS driver.

Pages: [1] 2 3 ... 14