• Welcome to OS2World OLD-STATIC-BACKUP Forum.
 

News:

This is an old OS2World backup forum for reference only. IT IS READ ONLY!!!

If you need help with OS/2 - eComStation visit http://www.os2world.com/forum

Main Menu

clock01.sys for os/4 kernel

Started by AAA, 2011.03.01, 17:56:45

Previous topic - Next topic

AAA

Edited by admin (Thomas Klein):

Quote
.../clock01_for_os4.RAR  (with built-in Timer0)

Hi AAA,
the FTP server displays a static "helo" text containing login instructions to access pirated software(z).
Please refrain from posting FTP links to this server! OS/2 World.com does not intend to facilitate access to illegal software. You are putting us in a bad legal situation.

Thanks for your understanding.
Thomas


miturbide

About this clock01.sys

Is it based on some IBM source code or it is a clean replacement without IBM source code on it.

Thanks.
Martín Itúrbide
OS2World.com NewsMaster
Open Source Advocate

Skype - martiniturbide
Google Talk - martiniturbide@gmail.com

AAA

It is not based on IBM code, moreover  it uses extended KEE which was added in OS/4  recently

diver

whats the advantage of using this clock01.sys instead of the IBM one? Soem more infos could be helpfull

miturbide

If it is not based on IBM code, you could upload this file on hobbes too, but it will be good if you can add a readme.txt with some info of the file just like diver says.

martin
Martín Itúrbide
OS2World.com NewsMaster
Open Source Advocate

Skype - martiniturbide
Google Talk - martiniturbide@gmail.com

lpino

#5
Quote from: miturbide on 2011.03.04, 16:37:21
If it is not based on IBM code, you could upload this file on hobbes too, but it will be good if you can add a readme.txt with some info of the file just like diver says.

martin
This is the google translation of the readme file:

The driver of the system timer for OS / 2 Warp 3.0 +.

This driver replaces driver CLOCK01.SYS (c) IBM and
TIMER0.SYS (a) Timur Tabi. All rewritten almost from scratch, but the functionality
Both drivers tried to keep.

Differences from version TIMER $ Timur Tabi:

- Uses interrupt RTC, and not from HRT0, while active TIMER $
 does not interfere with PerfView and requests DosTmrXXXX;
- Periodic event handlers are invoked on a timer to the allowed
 preemption;
- Added ability to request a periodic event with a timer
 period, expressed in milliseconds and fractions for both drivers and
 for problems (using semaphore events);
- Added ability to get the current 64-bit counter HRT0 for drivers
 (With short interruptions are permitted);
- Added ability to appeal to the counter of milliseconds and fractions;
- Tamera resolution does not change, not because that is not implemented, but "by design",
 intervals always supposed to ask only, in milliseconds, and
 their shares;
- The second word in the registration of a periodic interval timer by IDC
 Timur Tabi old query can not be equal to 0, then (ie always)
 it means the share of milliseconds;
- New requests IOCTL codes are more appropriate for the wishes of IBM
 developers;
- Added ability to tinker with the system time to 1 / 100 second and forth
 ago.

Differences on CLOCK $ version IBM:

- RTC registers is read to determine the system time only
 start, rather than every second, it is very important for the possibility of issuing
 system time to within 1 / 100 second;
- The real frequency of RTC interrupts configured only during system boot
 (Specified at compile time), according to the standard requests to change the frequency varies
 only the frequency of the call Scheduler.

Undocumented features (ie, referred to and implemented, but possible
absent in the future final version):

- Request statistics (how many times the timer interrupt occurs during
 processing or to processing the previous one) and some nesting in this
 achieved.

To do:

- Request data synchronization (see the exact correspondence between the counters
 RTC, HRT0 and system time) (unless the request is only reserved);
- Inquiry into the actual frequency setting interrupt RTC at start
 system (which may be an indication in the CONFIG.SYS, for example, the key to
 psevdodrayveru?)
- Write a proper documentation (the README does not count).

New requests for TIMER $:

IDC:

TMR0_REGFIX equ 3

Differs from TMR0_REG word order in the counter-time (milliseconds in older
word, share - in junior).

TMR0_GETFIXPTR equ 4

Returns to an address in Param1 pointer to 32-bit counter of milliseconds and
their shares.

TMR0_GETHRTVAL equ 5

Returns the value of the counter HRT0 in EDX: EAX.

IOCTL:

80h/40h ioBlockUntil

Full analog 80h/05h.

80h/41h ioFreePtr

Releases the pointer to counters milliseconds and fractions.

80h/42h ioStartFixTimer

Registers a periodic event.

80h/43h ioStopFixTimer

Cancels a recurring event.

80h/44h ioShiftTimeHs

Corrects the system time on + -1/100 sec.

80h/60h ioQueryVer

Returns the driver version.

80h/61h ioGetPtr

Returns pointers to counters milliseconds and fractions.

80h/62h ioGetSyncData

Should return the data synchronization RTC and HRT0, yet does nothing.

80h/63h ioGetStat

Returns statistics. Non-standardized format.

It seems to have not forgotten anything, remember what else, I'll write. A detailed description of the query
I'll write soon here or in CLOCK01.IPF, if not in communion pogryaznu
with ipfc.

Bye