Weird USB problem - OS2

This is a discussion on Weird USB problem - OS2 ; I have a very strange problem with my main OS/2 system. It's a Tyan Tiger MPX (S2466N-4M) motherboard with 2 x AMD Athlon MP 2200+ processors and 1GB of RAM. It has an Adaptec 2940UW and 2906 installed and its ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: Weird USB problem

  1. Weird USB problem

    I have a very strange problem with my main OS/2 system. It's a Tyan Tiger MPX
    (S2466N-4M) motherboard with 2 x AMD Athlon MP 2200+ processors and 1GB of RAM.
    It has an Adaptec 2940UW and 2906 installed and its main boot drive is a RAID5
    array hung off an LSiLogic MegaRAID Express 500.

    The embedded USB controller on the motherboard is USB 1.1 and when I first
    migrated to this motherboard I had to disable it in the BIOS to get OS/2 to boot
    at all - with it enabled the monitor would go into powerdown mode just after the
    OS/2 splash screen cleared off the screen and before the normal "2 processors
    initialized" message came out. I decided I could live with this since that got
    it to boot at all!

    Over the weekend I found an NEC chipset based USB 2.0 PCI card and plugged that
    in. Now I have exactly the same problem again but with this beast installed
    instead. I can boot with a config.sys that has no USB drivers in it at all. I
    can boot if I enable BASEDEV=USBEHCD.SYS and BASEDEV=USBD.SYS only and I see
    initialisation messages for EHCD.SYS telling me it has found one controller.
    However, if I unREM the BASEDEV=USBMSD.ADD line I get the monitor powerdown
    problem again.

    What have I tried so far. I've tried REMming out the drivers in config.sys for
    the Adaptec controllers, os2scsi.dmd, os2aspi.dmd, aspirout.sys, print01.sys,
    xdfloppy.flt, ibm1flpy.add, vaspi.sys and BA/2's BTSSCSI.SYS. I've tried adding
    /prec=lid to the OS2APIC.PSD line. APM is already disabled both in config.sys
    and in the BIOS. I've updated os2ldr and os2krnl to the one from the beta
    XR_C006 (also used 14.103a from XR_C005). I'm using the SNAP drivers though I'm
    pretty sure that I also tried booting in VGA mode when I was experimenting with
    the embedded USB controller.

    I'm not using the very latest USB drivers but I think the ones I do have should
    be 'good enough'. USBD.SYS is 10.158, USBEHCD.SYS is 10.137 and USBMSD.ADD is
    10.145. The rest of the system is using XR_C005.

    Does anyone have any bright ideas as to anything else I can try or anything else
    I can use to find out where the problem resides?

    --
    Trevor Hemsley, Brighton, UK
    Trevor dot Hemsley at ntlworld dot com

  2. Re: Weird USB problem

    Hi Trevor

    Trevor Hemsley wrote:
    > I have a very strange problem with my main OS/2 system. It's a Tyan Tiger MPX
    > (S2466N-4M) motherboard with 2 x AMD Athlon MP 2200+ processors and 1GB of RAM.
    > It has an Adaptec 2940UW and 2906 installed and its main boot drive is a RAID5
    > array hung off an LSiLogic MegaRAID Express 500.
    >
    > The embedded USB controller on the motherboard is USB 1.1 and when I first
    > migrated to this motherboard I had to disable it in the BIOS to get OS/2 to boot
    > at all - with it enabled the monitor would go into powerdown mode just after the
    > OS/2 splash screen cleared off the screen and before the normal "2 processors
    > initialized" message came out. I decided I could live with this since that got
    > it to boot at all!
    >
    > Over the weekend I found an NEC chipset based USB 2.0 PCI card and plugged that
    > in. Now I have exactly the same problem again but with this beast installed
    > instead. I can boot with a config.sys that has no USB drivers in it at all. I
    > can boot if I enable BASEDEV=USBEHCD.SYS and BASEDEV=USBD.SYS only and I see
    > initialisation messages for EHCD.SYS telling me it has found one controller.
    > However, if I unREM the BASEDEV=USBMSD.ADD line I get the monitor powerdown
    > problem again.
    >
    > What have I tried so far. I've tried REMming out the drivers in config.sys for
    > the Adaptec controllers, os2scsi.dmd, os2aspi.dmd, aspirout.sys, print01.sys,
    > xdfloppy.flt, ibm1flpy.add, vaspi.sys and BA/2's BTSSCSI.SYS. I've tried adding
    > /prec=lid to the OS2APIC.PSD line. APM is already disabled both in config.sys
    > and in the BIOS. I've updated os2ldr and os2krnl to the one from the beta
    > XR_C006 (also used 14.103a from XR_C005). I'm using the SNAP drivers though I'm
    > pretty sure that I also tried booting in VGA mode when I was experimenting with
    > the embedded USB controller.
    >
    > I'm not using the very latest USB drivers but I think the ones I do have should
    > be 'good enough'.



    But maybe not...

    USBD.SYS is 10.158, USBEHCD.SYS is 10.137 and USBMSD.ADD is
    > 10.145.



    I suggest updating those as a 1st step to the current (last) release -
    all of the above should be at 10.162.


    The rest of the system is using XR_C005.


    Does that include os2dasd.dmd buildlevel 14.104? - If not I think it is
    in the IBM IDE 10161 update.


    >
    > Does anyone have any bright ideas as to anything else I can try or anything else
    > I can use to find out where the problem resides?
    >



    Hopefully updating above drivers will help.

    If not then maybe you need to get hold of the current acpi driver
    package from ecomstation Betazone and see if that helps - requires an
    eCS Software Subscription though.

    Regards

    Pete

  3. Re: Weird USB problem

    Trevor Hemsley wrote:

    > I have a very strange problem with my main OS/2 system. It's a Tyan Tiger
    > MPX (S2466N-4M) motherboard with 2 x AMD Athlon MP 2200+ processors and
    > 1GB of RAM. It has an Adaptec 2940UW and 2906 installed and its main boot
    > drive is a RAID5 array hung off an LSiLogic MegaRAID Express 500.
    >
    > The embedded USB controller on the motherboard is USB 1.1 and when I first
    > migrated to this motherboard I had to disable it in the BIOS to get OS/2
    > to boot at all - with it enabled the monitor would go into powerdown mode
    > just after the OS/2 splash screen cleared off the screen and before the
    > normal "2 processors initialized" message came out. I decided I could live
    > with this since that got it to boot at all!
    >
    > Over the weekend I found an NEC chipset based USB 2.0 PCI card and plugged
    > that in. Now I have exactly the same problem again but with this beast
    > installed instead. I can boot with a config.sys that has no USB drivers in
    > it at all. I can boot if I enable BASEDEV=USBEHCD.SYS and BASEDEV=USBD.SYS
    > only and I see initialisation messages for EHCD.SYS telling me it has
    > found one controller. However, if I unREM the BASEDEV=USBMSD.ADD line I
    > get the monitor powerdown problem again.
    >
    > What have I tried so far. I've tried REMming out the drivers in config.sys
    > for the Adaptec controllers, os2scsi.dmd, os2aspi.dmd, aspirout.sys,
    > print01.sys, xdfloppy.flt, ibm1flpy.add, vaspi.sys and BA/2's BTSSCSI.SYS.
    > I've tried adding /prec=lid to the OS2APIC.PSD line. APM is already
    > disabled both in config.sys and in the BIOS. I've updated os2ldr and
    > os2krnl to the one from the beta XR_C006 (also used 14.103a from XR_C005).
    > I'm using the SNAP drivers though I'm pretty sure that I also tried
    > booting in VGA mode when I was experimenting with the embedded USB
    > controller.
    >
    > I'm not using the very latest USB drivers but I think the ones I do have
    > should be 'good enough'. USBD.SYS is 10.158, USBEHCD.SYS is 10.137 and
    > USBMSD.ADD is 10.145. The rest of the system is using XR_C005.
    >
    > Does anyone have any bright ideas as to anything else I can try or
    > anything else I can use to find out where the problem resides?



    Trevor, this sounds an awful lot like the problems I've had with a couple of
    different motherboards. What Video are you using?

    What was biting me was the IRQ assignments and the scanner (snoop). The
    system would assign PCI interrupts then scan - hitting just exactly the
    wrong port so that somebody else was grabbing the interrupt signal and
    processing what it thought was a message for it when it was actually the
    result of a totally unrelated event. With those two PCI SCSI controllers
    thrown in, you are prime for this interrupt problem. I don't know what
    your mb does, but the old cure was to swap slots with the PCI cards to get
    a different IRQ pin assignment until I (hopefully) finally cleared the
    interference. Black magic, at best. I finally found the offending driver
    and pulled it from the snoop list - don't recall exactly which one it was,
    but it was one of the SCSI boards which I never had anyway. Best way to
    start for me was to take a clean partition and play with an install from
    boot floppies so that I could controll the config.sys and snoop.lst files
    until the crashes stopped. The problems usually showed up pretty early in
    the install, so it was reasonably painless.
    --
    Will Honea

  4. Re: Weird USB problem (Solved)

    On Sun, 22 Apr 2007 20:31:37 UTC in comp.os.os2.setup.misc, Will Honea
    wrote:

    > With those two PCI SCSI controllers
    > thrown in, you are prime for this interrupt problem.


    This does seem to be what it was. I also found that there was a jumper on the
    motherboard to physically disable the integrated 3Com 3C905C LAN controller
    which I haven't used under OS/2 (there are drivers but I got strange hangs and
    dropped packets so I first installed an old Compaq Netelligent card and later a
    DLink DGE-530T gigabit card). Disabling the 3Com freed up IRQ 10 which was in
    use exclusively for this and also seemed to allow it to use IRQ 3 as well.

    In addition to having my USB card working I am now also able to enable the
    floppy controller in the BIOS which has been disabled since I first tried to
    boot OS/2 on this motherboard and found exactly the same symptoms with that too.

    It probably helped that the jumper to disable the 3Com was underneath every
    single PCI card in the system so I had to take them all out and re-arrange them
    to get at it anyway!

    --
    Trevor Hemsley, Brighton, UK
    Trevor dot Hemsley at ntlworld dot com

  5. Re: Weird USB problem (Solved)

    Trevor Hemsley wrote:

    > It probably helped that the jumper to disable the 3Com was underneath
    > every single PCI card in the system so I had to take them all out and
    > re-arrange them to get at it anyway!


    Seems that everyone likes to hide the jumpers - and either forget to put
    them in the docs, mis-label them, or not mark pin 1 when it's a
    multi-position selection.. Funny about the 3C905C port - I have 3 of those
    working like champs. Whoever it is that designs motherboards gives lie to
    the idea of computer setup as a science - it's still a matter of black
    magic and holding your mouth just right!

    --
    Will Honea

  6. Re: Weird USB problem

    I have a VIA mini-itx that I'm trying to setup os2 warp 3 on. It
    *only* gets anywhere at all if I go into the CMOS and reset the
    "OS2>memory" option. But when I do, I notice that drdos boots dont
    work right any more. ram disk maxes at 15 meg. graphic paints roll
    down the screen like a 486.

    Right now, I'm trying to get the floppy boot to see the cd-rom. None
    of the three ( hitachi & Optorite DVD * Toshiba) CD drives are seen by
    the install software.... see other post.


+ Reply to Thread