Author Topic: New OS/4 kernel :: SVN4227  (Read 148954 times)

dbanet

  • Guest
Re: New OS/4 kernel :: SVN4188 :: New SNAP helper driver (SDDHELP.SYS)
« Reply #75 on: May 24, 2014, 01:33:07 pm »
Up! 4188!

Dave Yeo

  • Hero Member
  • *****
  • Posts: 4787
  • Karma: +99/-1
    • View Profile
Re: New OS/4 kernel :: SVN4183 :: New SNAP helper driver (SDDHELP.SYS)
« Reply #76 on: May 24, 2014, 06:21:31 pm »
Finally got around to testing the newest OS/4 kernel on my Warp V4+latest free fixes install. It hangs when it should go graphical. The new Doscall1.dll did seem to make the system more stable though, it's a somewhat crufty install from '96 or so and crashes if I enable more then one core but runs fine with one core and doesn't have the FPU problems that newer versions of OS/2 have

os/4 is SMP kernel  and  cannot work with UNI doscall1 from Warp V4
you should use dc1 from os/4 package or from Warp 4.5 SMP

interesting to see comport log with that crash

Missed this post until now  :)

I have the SMP kernel and SMP doscall1 as I've tried to enable both cores previously. No comport on this box. Wish there was if only for the modem.

OS4User

  • Sr. Member
  • ****
  • Posts: 406
  • Karma: +10/-0
    • View Profile
Re: New OS/4 kernel :: SVN4183 :: New SNAP helper driver (SDDHELP.SYS)
« Reply #77 on: May 24, 2014, 06:49:55 pm »
Finally got around to testing the newest OS/4 kernel on my Warp V4+latest free fixes install. It hangs when it should go graphical. The new Doscall1.dll did seem to make the system more stable though, it's a somewhat crufty install from '96 or so and crashes if I enable more then one core but runs fine with one core and doesn't have the FPU problems that newer versions of OS/2 have

os/4 is SMP kernel  and  cannot work with UNI doscall1 from Warp V4
you should use dc1 from os/4 package or from Warp 4.5 SMP

interesting to see comport log with that crash

Missed this post until now  :)

I have the SMP kernel and SMP doscall1 as I've tried to enable both cores previously. No comport on this box. Wish there was if only for the modem.

it is hard  to say what is wrong in your case  -  try some other hardware if possible
and do not forget use proper com.sys and other drivers  for SMP

OS4User

  • Sr. Member
  • ****
  • Posts: 406
  • Karma: +10/-0
    • View Profile
Re: New OS/4 kernel :: SVN4183 :: New SNAP helper driver (SDDHELP.SYS)
« Reply #78 on: June 04, 2014, 05:52:01 pm »
4195

dbanet

  • Guest
Re: New OS/4 kernel :: SVN4198 :: New SNAP helper driver (SDDHELP.SYS)
« Reply #79 on: June 06, 2014, 01:23:30 pm »
Up! 4198!

OS4User

  • Sr. Member
  • ****
  • Posts: 406
  • Karma: +10/-0
    • View Profile
Re: New OS/4 kernel :: SVN4200
« Reply #80 on: June 07, 2014, 05:07:33 pm »
4200

to set VIRTUALADDRESSLIMIT   use  VALIMIT  statement  in os2ldr.ini
for example  VALIMIT = 3072
« Last Edit: June 07, 2014, 07:47:43 pm by OS4User »

dbanet

  • Guest
Re: New OS/4 kernel :: SVN4208
« Reply #81 on: June 25, 2014, 02:54:09 am »
Up! 4202, 4203, 4204, 4205, 4206, 4207 and 4208!

OS4User

  • Sr. Member
  • ****
  • Posts: 406
  • Karma: +10/-0
    • View Profile
New OS/4 kernel :: SVN4221
« Reply #82 on: August 03, 2014, 09:28:38 pm »
http://gus.biysk.ru/os4/os2krnlSVN4221_unoff.zip

lanserver/requester (NETWKSTA.200)  test is  needed

Sergey Posokhov

  • Full Member
  • ***
  • Posts: 169
  • Karma: +8/-6
    • View Profile
    • OS/2 API Research
Re: New OS/4 kernel :: SVN4221
« Reply #83 on: August 12, 2014, 01:43:31 pm »
I've got a bug on ThinkPad 510.

Something hangs when PM starts on, so I can only see a mouse pointer.
Keyboard monitors (such as KEYBMON) works, NumLock and PC speaker works too.
WPS starts only in 10% of times, and then, OS/4 works fine.
Screen driver is faster that Panorama (seems that it has a bitmap cache).

There are OEMHLP$ and ___HLP$ log files.
« Last Edit: August 12, 2014, 03:44:20 pm by Sergey Posokhov »

OS4User

  • Sr. Member
  • ****
  • Posts: 406
  • Karma: +10/-0
    • View Profile
Re: New OS/4 kernel :: SVN4221
« Reply #84 on: August 19, 2014, 08:52:53 am »
Something hangs when PM starts on, so I can only see a mouse pointer.
Keyboard monitors (such as KEYBMON) works, NumLock and PC speaker works too.
WPS starts only in 10% of times, and then, OS/4 works fine.

I also noticed such behavior but rarely.
It also happened on IBM kernels.
Until now I don't have an explanation for it.

Screen driver is faster that Panorama (seems that it has a bitmap cache).

Screen driver can't be faster than Panorama as they work on different levels. Screen driver is just a set of helpers that are used also by Panorama. Screen03 differs from Screen01 by mapping LFB using new kernel function and which sets a type of caching WC through PAT.

Panorama sets WC through MTRR which doesn't work always correctly. Thus you may see that with Screen03 Panorama works faster.

Of course Screen doesn't have anything in common with caching bitmap.
« Last Edit: August 19, 2014, 08:58:55 am by OS4User »

Silvan Scherrer

  • Full Member
  • ***
  • Posts: 200
  • Karma: +1/-0
    • View Profile
Re: New OS/4 kernel :: SVN4222
« Reply #85 on: August 19, 2014, 01:03:59 pm »
* The pirates don't have access to Panorama VESA source code
I really wonder if such comments are either necessary or constructive. I personally have my doubts about both.

regards
Silvan
kind regards
Silvan
CTO bww bitwise works GmbH

Please help us with donations, so we can further work on OS/2 based projects. Our Shop is at https://www.bitwiseworks.com/shop/index.php

Doug Bissett

  • Hero Member
  • *****
  • Posts: 1593
  • Karma: +4/-2
    • View Profile
Re: New OS/4 kernel :: SVN4222
« Reply #86 on: August 20, 2014, 12:33:50 am »
I've got a bug on ThinkPad 510.

Something hangs when PM starts on, so I can only see a mouse pointer.
Keyboard monitors (such as KEYBMON) works, NumLock and PC speaker works too.
WPS starts only in 10% of times, and then, OS/4 works fine.

That is an OLD OS/2 bug. It has been around, for years (long before OS4 was started), known as a "startup conflict". I rarely see it on my T510 after adjusting the boot sequence (use the LCSS package from HOBBES, with the documented update - be sure to READ the instructions). It also helps to put a file  called AltF2on.$$$ into the root of the boot drive (The contents don't matter, but it displays the drivers as they load, which slows down booting a little). I also use REXX Autostart so that all of the contents of the Startup folder don't try to start at the same time. From what I can tell, OS4 boots a little faster than the IBM kernel, which changes the timing when things load. If you get a bad combination, you will see exactly what you are describing.

Lars

  • Hero Member
  • *****
  • Posts: 1271
  • Karma: +65/-0
    • View Profile
Re: New OS/4 kernel :: SVN4222
« Reply #87 on: August 21, 2014, 11:25:29 am »
The latest  versions of Panorama also support PAT where MTRRs cannot be set up properly.
I'd appreciate a less biased information flow.
To be more precise: PAT support is in GRADD.SYS which is used by Panorama.
« Last Edit: August 21, 2014, 11:30:45 am by Lars »

Eugene Gorbunoff

  • Full Member
  • ***
  • Posts: 131
  • Karma: +1/-0
    • View Profile
Re: New OS/4 kernel :: SVN4222
« Reply #88 on: August 21, 2014, 12:33:38 pm »
My message aimed to protect the developers of Panorama disappeared:

Support of PAT and MTRR is implemented in Panorama VESA and was improved in 2012:
http://ecomstation.ru/commentnews.php?id=2233

OS4User

  • Sr. Member
  • ****
  • Posts: 406
  • Karma: +10/-0
    • View Profile
Re: New OS/4 kernel :: SVN4222
« Reply #89 on: August 21, 2014, 05:08:06 pm »
The latest  versions of Panorama also support PAT where MTRRs cannot be set up properly.
I'd appreciate a less biased information flow.
To be more precise: PAT support is in GRADD.SYS which is used by Panorama.

want  to draw your attention to a few points:
yes, the latest Panorama can use PAT to set WC for LFB,  but Panorama tries to use PAT when it is not able to use MTRR  - (I can hardly imagine when such a scenario may happen).

to use PAT or MTRR  it is necessary to program corresponding MSRs and it is very important to have the same value in these registers between CPU cores, unfortunately, original kernel never offers an API to do proper MSR programing, that is why Panorama uses chain Panorama -> GRADD.SYS -> ACPI.PSD,  this path does not give 100% guaranty that MSRs will be coherent (fortunately this missing coherency happens not very often,  but I saw it a few times).

to use PAT Panorama (GRADD.SYS)  hacks the PTEs corresponding to LFB (by the way SNAP is doing similar hack but for other reason). Yes it works  and there is no other way to use PAT with original kernel  -  but me personally, I do not like such hacks.

os/4 kernel provides KernVMAlloc2 and a programer is able to set any type of memory caching without hacking PTEs,  PAT MSRs are set during CPU start  -  this way coherency is provided.