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 ... 24 25 [26] 27 28 ... 85
376
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 14, 2022, 11:10:20 pm »
So what are you trying to prove? That hidapi is not working properly? Or libusb? Or USBCALL version 12.11?
Just to make things clear: I am not actively using OS/2 any more. And I am not going to put any effort into something that is not approached systematically.

377
Hardware / Re: New ArcaOS USB driver and Floppy drive
« on: January 14, 2022, 10:24:17 am »
You will need to write an AN bug report.

378
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 14, 2022, 10:17:32 am »
Should have looked in popuplog.os2. Problem is the pthread library being newer in netlabs-exp.
Code: [Select]
01-13-2022  20:34:18  SYS2070  PID 0065  TID 0001  Slot 0081
K:\WORK\HIDAPITESTER\HIDAPITESTER.EXE
USB100->PTHR01._pthread_condattr_init
127
Threw me off as I expected something like,
Code: [Select]
[K:\work\hidapitester]hidapitester  --list
A non-recoverable error occurred.  The process ended.
Which is the result after I upgraded USBCALLS.DLL to 12.11. No trp, no entry in popuplog.

Edit: Above was with netlabs-rel. Upgrading usbcalls.dll on a netlabs-exp install things do work. The only USBCALLS.DLL installed here is in \os2\dll.

  • SYS2070: this almost always means that some DLL is missing an entry point. If that is the case, then you have an outdated pthread installed. Sorry but you will need to upgrade to the one in the Netlabs EXP repo. Please note that I have not knowingly used some pthread functionality that is "too new" but rather the rest of "libusb" that I did not touch uses that functionality. In short: we don't have a choice.
  • USBCALLS 12.11 ? I did not even know that version 12.11 of the USB stack already exists (I cannot see it in my AN subscription). Maybe it is not a good idea to test with something nobody else has available  ;)

379
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 13, 2022, 08:24:01 pm »
Just to state: I just pulled the latest hidapi stuff from rpm.netlabs.org and installed the RPMs manually.
It now works exactly as the static build from Dave. I rebuilt "hidapitester" with some changes to the makefile to link to hidapi dynamically. I also built with "-Zhigh-mem" as I know that the whole chain (libusb, usbcalls) works just well with high memory.

Regards,
Lars

380
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 13, 2022, 07:31:17 pm »
https://github.com/josch1710/hidapi
I'll try booting to my netlabs-exp partition later and test there. Maybe a reboot is needed as I can't think of why things would be different here then there.
What architecture are you using? Here, libusb1 is Pentium4 and the Yum installed usbcalls-dev is i686. It shouldn't matter but you never know.

libusb: pentium 4
usbcalls-devel: n/a as I am using my stuff from the USB stack build tree ("leftovers", so to say)

But at least for usbcalls-devel, it should not matter. It's just a bunch of text files and 2 import libraries whose format is not dependant on processor architecture.

By the way: the "usbcalls" package might be the culprit. Unless you are using AN, the dependencies try to install a USBCALLS (20140416-1) that is by far older than what comes with AN or with my USB stack. Just stumbled over this problem. What I did was to just delete \usr\lib\usbcalls.dll which will reinstate the proper file in \os2\dll.

381
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 13, 2022, 10:40:42 am »
Whoops, committed patches are attached. Jochen commited my first patch.

Where did Jochen commit your first patch?

382
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 13, 2022, 09:05:01 am »
About usb100.dll: I don't understand why pulling the stuff from the BWW repo and building yourself works better than what is put in Netlabs. The only thing that Silvan did was to build it, likely with optimization.
Remember the NEW libusb is in Netlabs EXP and not (yet) in Netlabs REL !

Edit: sorry, did not read all postings, you are using the correct DLLs.

Still don't understand. My own built libusb DLL works as good as the one from Netlabs repo. I cannot spot any functional difference.

383
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 13, 2022, 08:59:55 am »
Ok, then it is most likely hidapi itself (the BWW repo is pulled straight from the official repo) or you using the AN USB stack (Wim and me have been testing with my USB stack).
I tried to be compatible with the AN USB stack but that one is closed source, I therefore don't have a clue of what they are doing.

384
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 13, 2022, 07:53:35 am »
Wait a sec, this is about hidapi: can you send your patched hid.c file? I can then merge that into the BWW repo.

385
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 13, 2022, 07:32:29 am »
As to my USB stack: I need to release a new stack first...

386
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 13, 2022, 07:29:40 am »
That's odd because BWW has been building from exactly "my" version (the BWW repo) and I have used your hidapitester and that worked just fine.
Can you make sure that you have usb100.dll and USBCALLS.dll on your system only once and that they are the expected version?
Please note that I am using my USB drivers. That might make a difference as a few things had to be fixed in USBCALLS.dll as well as in the HC drivers.
As a test, can you test with my stack (after checking that the rest is ok) ?
If that works I will try and convince AN to pick up the changes that I did for USBCALLS and the HC drivers.

387
Games / Re: DOS VDM - USB and Game port Gamepad support
« on: January 12, 2022, 09:03:49 pm »
By now, I finally finished libusb (and also libuvc). Additionally, BWW has now built hidapi.
All 3 packages can now be installed from netlabs-exp.

Dave, can you build hidapitester.exe and link to hidapi dynamically ? That would allow me to test the hidapi->libusb chain.
I did test with your static build of "hidapitester.exe" but I don't know how many "hacks" you needed to get hidapitester running properly.

Cheers,
Lars

388
Hardware / Re: New ArcaOS USB driver and Floppy drive
« on: January 11, 2022, 11:30:01 am »
Ivan's question is not about broken floppies.
It's about "kicking" the drive to read the disk when you insert a new disk.
I remember that I had a hell of a time to get that to work correctly. I wish AN much fun with it :-).

389
Applications / Re: LIBC panic errors anyone?
« on: January 10, 2022, 10:12:22 am »
Yes, I have also seen this trap report:

Don't shoot the messenger. It looks like there is an actual improvement in libc to properly report if you call "free" on the very same pointer more than once (which is forbidden).

Consequently, that likely means that the affected applications actually have a problem managing their dynamic memory. These problems can go unnoticed if for example a memory page is not directly assigned to another allocation and therefore remains accessible for some time. That's why it is so important to report this error.

About lsusb: if you see them in LIBUSB, then that's the reason why I am still working on it :-)

Cheers,
Lars

390
Well, IJFW should never hang the whole system.
But if the load becomes too massive, FXWRAP.SYS will start overflowing its internal work queues. That will lead to connections timing out because packets are being rejected (from being added to the work queue) but it should not lead to system hangs. If it does, there is still something not quite right.
But it should be clear, that in this case, you will never be able to achieve a high throughput. FXWRAP.SYS will then act like a "throttle".

Pages: 1 ... 24 25 [26] 27 28 ... 85