Changes between Version 15 and Version 16 of KnownProblems


Ignore:
Timestamp:
Mar 30, 2012, 10:53:35 PM (12 years ago)
Author:
David Azarewicz
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • KnownProblems

    v15 v16  
    4747== Booting slowness ==
    4848
    49 Because of some recent changes to make the PSD compatible with an even larger number of vendor's ACPI, the boot time may have increased by a few seconds. On some systems, the boot time may have increased significantly (30-40 seconds). The slowness occurs when the Kernel is loading the basedev drivers during boot. The basedev drivers are loaded in between the time the OS/2 IBM copyright message is displayed on the screen and the time the PSD sign on banner is displayed.  This is not an ACPI.PSD problem. This is a peculiarity in the way certain BIOS operate (typically BIOS from Award). The OS/2 kernel uses the BIOS to load the basedev drivers, In order for this to work in Mode 2, the PSD must tell the BIOS to switch interrupt modes and reassign certain interrupts each time the Kernel calls the BIOS. Most BIOS do this very quickly. A few take a long time to make the switch.
     49Starting with version 3.20.01, changes were made to make the PSD compatible with an even larger number of vendor's ACPI. This may have increased the boot time by a few seconds on some systems. On a few systems, the boot time may have increased significantly (30-40 seconds). The slowness occurs when the Kernel is loading the basedev drivers during boot. The basedev drivers are loaded in between the time the OS/2 IBM copyright message is displayed on the screen and the time the PSD sign on banner is displayed.  This is not an ACPI.PSD problem. This is a peculiarity in the way certain BIOS operate (typically BIOS from Award). The OS/2 kernel uses the BIOS to load the basedev drivers, In order for this to work in Mode 2, the PSD must tell the BIOS to switch interrupt modes and reassign certain interrupts each time the Kernel calls the BIOS. Most BIOS do this very quickly. A few take a long time to make the switch.
    5050
    51 Since this is a BIOS issue and not an ACPI.PSD issue, it is extremely unlikely that I will be able to do anything about this slowness. However, I will gather more information in the hope that there may be something I can do -- at lease for some of the systems. if you have this slowness issue on your system you can help me gather data by installing the debug version of the PSD, creating a testlog, and attaching it to ticket #536. If the increase in boot time bothers you, your only solution is to run your system with the /VW switch which does not require the interrupt mode switching.
     51Since this is a BIOS issue and not an ACPI.PSD issue, it is extremely unlikely that I will be able to do anything about this slowness. However, I will gather more information in the hope that there may be something I can do -- at least for some of the systems. if you have this slowness issue on your system you can help me gather data by installing the debug version of the PSD, creating a testlog, and attaching it to ticket #536. If the increase in boot time bothers you, your only solution is to run your system with the /VW switch which does not require the interrupt mode switching.
    5252
    5353== Immediate suspend due to misbehaving SLPB device ==
    5454
     55Some systems enter suspend immediately after !AcpiDaemon is started. This is due to a misbehaving SLPB device which generates repeating suspend requests.
     56
    5557NOTE: In version 3.19.16 and later, a workaround was implemented to ignore SLPB events if one occurs within 10 seconds of starting !AcpiDaemon.exe.  The SLPB events are still generated by the hardware every 2 seconds, but they are simply ignored. If !AcpiDaemon logging is enabled, the log file could grow very large logging these events. An actual fix is still being investigated.
    56 
    57 Some systems enter suspend immediately after !AcpiDaemon is started. This is due to a misbehaving SLPB device which generates repeating suspend requests. If you have this problem, don't run the !AcpiDaemon until a version with a fix or workaround is released.
    5858
    5959== SCSI drivers ==