USB sub-system: more observations and some conclusions - OS2

This is a discussion on USB sub-system: more observations and some conclusions - OS2 ; My "phantom" drives that appear relate to USBCDROM.add. This confirms a discussion that occurred here back in Jan 2003 (found by searching on usbcdrom.add in Google and clicking "groups" on the page that appears and then clicking on the 3rd ...

+ Reply to Thread
Results 1 to 16 of 16

Thread: USB sub-system: more observations and some conclusions

  1. USB sub-system: more observations and some conclusions

    My "phantom" drives that appear relate to USBCDROM.add. This
    confirms a discussion that occurred here back in Jan 2003 (found by
    searching on usbcdrom.add in Google and clicking "groups" on the page
    that appears and then clicking on the 3rd article on the page that
    follows). The initiator of the thread called them "extra drives". The
    conclusion was that there is a bug in the IBM software.

    The extra drives are not related to the presence of a "3rd" IDE
    channel in the form of SATA ports.

    My extra drives are designated logical volume names [ v1 ] and [
    CDROM ]) by LVM. The respective drive letters are I: thru to K: (my
    hard disk drives letters are C: thru to H.

    Drive J: was I: without USBCDROM.add loaded and K: was J: These are
    respectively a DVD reader master on the 2ndary IDE port and a CD burner
    as slave.

    Trying to access I: from an OS/2 prompt gives a Sys0021 error.

    When LVM is invoked it first reports "Physical Disk ( 3) The
    partition table on this disk may be corrupt." One hits [Enter} and the
    next message is, "Physical Disk ( 3) The system is not able to use this
    disk". Touch [Enter} LVM then presents its usual information.

    LVM is not giving sensible information because I don't have a
    Physical Disk ( 3).

    When I Rem out BASEDEV=USBCDROM.ADD from D:\OS2\BOOT\Config.x and
    boot to the command line LVM presents the drive letter allocation like
    without the USB lines.

    However, when I try the same in the main config.sys boot up of the
    WPS hangs after the welcoming sound. It hangs before the Warp Centre
    tries to come up.

    So for now I have to live with one phantom drive I: by leaving the
    BASEDEV=USBCDROM.ADD line active.

    Is this an unfixable problem due to a bug in the USB sub-system?

    To William Hartzell I don't think there is any point in using
    RESERVEDRIVELETTER=R because I think ALL the CDROM drives will shift
    including the phantom one.

    Another question: what was the logic behind creating USBCDROM.add in
    relation to the MSD concept? I mean isn't a CDROM a mass storage device
    of sorts? Certainly a CD writer IS a MSD. However, from the Google
    references I noticed that RSJ doesn't talk to USB CD writers.

    This whole area therfore looks a bit of a mess!?

    --
    To reply by e-mail edit this address to the correct form: vbien at
    attglobal dot net
    

  2. Re: USB sub-system: more observations and some conclusions

    Sir:

    Victor Bien wrote:
    > My "phantom" drives that appear relate to USBCDROM.add. This confirms
    > a discussion that occurred here back in Jan 2003 (found by searching on
    > usbcdrom.add in Google and clicking "groups" on the page that appears
    > and then clicking on the 3rd article on the page that follows). The
    > initiator of the thread called them "extra drives". The conclusion was
    > that there is a bug in the IBM software.
    >
    > The extra drives are not related to the presence of a "3rd" IDE
    > channel in the form of SATA ports.
    >
    > My extra drives are designated logical volume names [ v1 ] and [ CDROM
    > ]) by LVM. The respective drive letters are I: thru to K: (my hard disk
    > drives letters are C: thru to H.
    >
    > Drive J: was I: without USBCDROM.add loaded and K: was J: These are
    > respectively a DVD reader master on the 2ndary IDE port and a CD burner
    > as slave.
    >
    > Trying to access I: from an OS/2 prompt gives a Sys0021 error.
    >
    > When LVM is invoked it first reports "Physical Disk ( 3) The partition
    > table on this disk may be corrupt." One hits [Enter} and the next
    > message is, "Physical Disk ( 3) The system is not able to use this
    > disk". Touch [Enter} LVM then presents its usual information.
    >
    > LVM is not giving sensible information because I don't have a Physical
    > Disk ( 3).
    >
    > When I Rem out BASEDEV=USBCDROM.ADD from D:\OS2\BOOT\Config.x and boot
    > to the command line LVM presents the drive letter allocation like
    > without the USB lines.
    >
    > However, when I try the same in the main config.sys boot up of the WPS
    > hangs after the welcoming sound. It hangs before the Warp Centre tries
    > to come up.
    >
    > So for now I have to live with one phantom drive I: by leaving the
    > BASEDEV=USBCDROM.ADD line active.
    >
    > Is this an unfixable problem due to a bug in the USB sub-system?
    >
    > To William Hartzell I don't think there is any point in using
    > RESERVEDRIVELETTER=R because I think ALL the CDROM drives will shift
    > including the phantom one.
    >
    > Another question: what was the logic behind creating USBCDROM.add in
    > relation to the MSD concept? I mean isn't a CDROM a mass storage device
    > of sorts? Certainly a CD writer IS a MSD. However, from the Google
    > references I noticed that RSJ doesn't talk to USB CD writers.
    >
    > This whole area therfore looks a bit of a mess!?
    >

    IF you don't have a USB CDROM drive, put a REM in front of the statement
    that has USBCDROM.ADD. If you have an USB CDROM drive, use the /csd:x
    set x to a value like 1, the default. The phantom CDROM only appears if
    you do not have an USB CDROM attached. Use the RESERVERDRIVELETTER=R to
    move the phantom USB CDROM drive up out of the way with the other CDROM
    drives. On my system, CDROMs don't show up in LVM. Suggest that you
    download the Drive Manager code on Hobbes and substitute it for the IBM
    version. We all have
    been assuming that you have a recent version of OS/2 or even eCS. Look
    into Klass & Gordan's System tool-Configtool database for info on the
    statement, if you are running an older system.

    BTW, this is a different problem than the one you describe beforehand.
    You should never install drivers for stuff that you don't have. That
    can cause problems or make another problem worse. You may have a
    problem with multiple IDE channels, but until you disable some of them,
    you won't know. I don't understand your reluctance to do this. Have
    you looked for an updated BIOS for this main board? Others are using
    main boards with multiple SATA channels and PATA channels without
    problems. That just means that the problem is local with your main
    board, configuration and maybe you have fixated upon the wrong symptom.


    --
    Bill
    Thanks a Million!

  3. Re: USB sub-system: more observations and some conclusions

    Top Post Alert!

    Hi Victor

    Would you *Please* try an application available from hobbes
    http://hobbes.nmsu.edu/pub/incoming/usbcfgb3.zip

    I think the application, a USB settings notebook, will help sort out
    your usb problems. I do suggest a read through of the Help file which
    may explain some, if not all, of the problems you have with usb.

    I would appreciate some feedback on whether the above application helps
    or not - and if Not, Why not? That way the app can be further developed
    to meet users needs :-)

    Regards

    Pete



    Victor Bien wrote:
    > My "phantom" drives that appear relate to USBCDROM.add. This confirms
    > a discussion that occurred here back in Jan 2003 (found by searching on
    > usbcdrom.add in Google and clicking "groups" on the page that appears
    > and then clicking on the 3rd article on the page that follows). The
    > initiator of the thread called them "extra drives". The conclusion was
    > that there is a bug in the IBM software.
    >
    > The extra drives are not related to the presence of a "3rd" IDE
    > channel in the form of SATA ports.
    >
    > My extra drives are designated logical volume names [ v1 ] and [ CDROM
    > ]) by LVM. The respective drive letters are I: thru to K: (my hard disk
    > drives letters are C: thru to H.
    >
    > Drive J: was I: without USBCDROM.add loaded and K: was J: These are
    > respectively a DVD reader master on the 2ndary IDE port and a CD burner
    > as slave.
    >
    > Trying to access I: from an OS/2 prompt gives a Sys0021 error.
    >
    > When LVM is invoked it first reports "Physical Disk ( 3) The partition
    > table on this disk may be corrupt." One hits [Enter} and the next
    > message is, "Physical Disk ( 3) The system is not able to use this
    > disk". Touch [Enter} LVM then presents its usual information.
    >
    > LVM is not giving sensible information because I don't have a Physical
    > Disk ( 3).
    >
    > When I Rem out BASEDEV=USBCDROM.ADD from D:\OS2\BOOT\Config.x and boot
    > to the command line LVM presents the drive letter allocation like
    > without the USB lines.
    >
    > However, when I try the same in the main config.sys boot up of the WPS
    > hangs after the welcoming sound. It hangs before the Warp Centre tries
    > to come up.
    >
    > So for now I have to live with one phantom drive I: by leaving the
    > BASEDEV=USBCDROM.ADD line active.
    >
    > Is this an unfixable problem due to a bug in the USB sub-system?
    >
    > To William Hartzell I don't think there is any point in using
    > RESERVEDRIVELETTER=R because I think ALL the CDROM drives will shift
    > including the phantom one.
    >
    > Another question: what was the logic behind creating USBCDROM.add in
    > relation to the MSD concept? I mean isn't a CDROM a mass storage device
    > of sorts? Certainly a CD writer IS a MSD. However, from the Google
    > references I noticed that RSJ doesn't talk to USB CD writers.
    >
    > This whole area therfore looks a bit of a mess!?
    >


  4. Re: USB sub-system: more observations and some conclusions

    Peter Brown wrote:
    > Top Post Alert!
    >
    > Hi Victor
    >
    > Would you *Please* try an application available from hobbes
    > http://hobbes.nmsu.edu/pub/incoming/usbcfgb3.zip
    >
    > I think the application, a USB settings notebook, will help sort out
    > your usb problems. I do suggest a read through of the Help file which
    > may explain some, if not all, of the problems you have with usb.
    >
    > I would appreciate some feedback on whether the above application helps
    > or not - and if Not, Why not? That way the app can be further developed
    > to meet users needs :-)


    Well I downloaded and installed it. It was only marginally helpful
    because I having become familiar with how to handle config.sys don't
    really need it. I didn't like the impact it had on the
    layout/readability of config.sys (although it saved my original). Your
    help doc was more useful than the program itself in that it pulls
    together all the facts associated with USB in the one place.
    Particularly 1. having as many lines of BASEDEV=USBUHCD.sys as HCIMONIT
    says there are such devices; 2. leaving out the
    /REQ:USBUHCD$,USBOHCD$,USBEHCD$ specs following BASEDEV=USBD.SYS and 3.
    a spec in BASEDEV=USBMSD.ADD to make atached MSDs behave like "large
    floppies".

    Having DEVICE=D:\OS2\BOOT\USBCOM.SYS CDS:0 still resulted in the WPS
    hanging before the Warp Centre appears. Then I noticed that my
    USBCOM.sys was dated 2002 and I had one dated 2004 which I had failed to
    copy over. Now I've done that but can't yet report on the effect of
    that because...

    I deferred to Bill H's urging to update the BIOS. Well having done
    that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    site was ambigous in having two files to apply. I tried one and the
    QFlash utility said this was the wrong update so I tried the other one.
    It went thru and said the update was successful and said to restart.
    That was the last time I saw anything on the screen. I removed every
    card on the board and tried again. Same - nothing. I've done flash
    bios updates many times before (on customer's machines!) know it is
    absolutely essential to have the correct update, this time I tried as
    usual to be sure I have the correct update and still it fails! I think
    Gigabyte has let me down. There is no chance of rolling back. The
    board doesn't dual BIOSs. I will have to go thru the fuss and bother of
    getting a warranty replacement.

    This is posted from my Windoze machine. Don't know when OS/2 will be
    back up.

    _Backtracking_ the story:
    When the system came backup with the settings your USBCfg setup I got
    2 phantom drives instead of 1 and even more strangely my LAN connection
    stopped! After remming out all the four BASEDEV=USBUHCD.sys lines and
    restarting my LAN was back. There were still 4 phantom drives.

    The NIC is on IRQ3 whereas the board reports the USBUHCD devices on
    IRQ11. I had a lot of trouble getting this board to allocate different
    irqs to its 3 PCI slots and cut and tried a lot with the location of
    this card and that in the slots.

    How update is my system?
    I think it is fairly up-to-date (except for slip-ups). It is a MCP
    4.52 with FP5 plus all the other fixes applied as far as I can tell. I
    can post the Syslevel data here if you like.

    Victor Bien

  5. Re: USB sub-system: more observations and some conclusions

    Sir:

    Victor Bien wrote:

    > I deferred to Bill H's urging to update the BIOS. Well having done
    > that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    > site was ambigous in having two files to apply. I tried one and the
    > QFlash utility said this was the wrong update so I tried the other one.
    > It went thru and said the update was successful and said to restart.
    > That was the last time I saw anything on the screen. I removed every
    > card on the board and tried again. Same - nothing. I've done flash
    > bios updates many times before (on customer's machines!) know it is
    > absolutely essential to have the correct update, this time I tried as
    > usual to be sure I have the correct update and still it fails! I think
    > Gigabyte has let me down. There is no chance of rolling back. The
    > board doesn't dual BIOSs. I will have to go thru the fuss and bother of
    > getting a warranty replacement.
    >

    Hey, I am sorry. IF the machine won't come up, then there is nothing
    you can do. That BIOS chip is surface mounted, so there is no pulling
    it and finding an EEPROM burner to put back the old BIOS, right?

    --
    Bill
    Thanks a Million!

  6. Re: USB sub-system: more observations and some conclusions

    Hi Victor

    Victor Bien wrote:
    > Peter Brown wrote:
    >
    >> Top Post Alert!
    >>
    >> Hi Victor
    >>
    >> Would you *Please* try an application available from hobbes
    >> http://hobbes.nmsu.edu/pub/incoming/usbcfgb3.zip
    >>
    >> I think the application, a USB settings notebook, will help sort out
    >> your usb problems. I do suggest a read through of the Help file which
    >> may explain some, if not all, of the problems you have with usb.
    >>
    >> I would appreciate some feedback on whether the above application
    >> helps or not - and if Not, Why not? That way the app can be further
    >> developed to meet users needs :-)

    >
    >
    > Well I downloaded and installed it. It was only marginally helpful
    > because I having become familiar with how to handle config.sys don't
    > really need it. I didn't like the impact it had on the
    > layout/readability of config.sys (although it saved my original). Your
    > help doc was more useful than the program itself in that it pulls
    > together all the facts associated with USB in the one place.
    > Particularly 1. having as many lines of BASEDEV=USBUHCD.sys as HCIMONIT
    > says there are such devices; 2. leaving out the
    > /REQ:USBUHCD$,USBOHCD$,USBEHCD$ specs following BASEDEV=USBD.SYS and 3.
    > a spec in BASEDEV=USBMSD.ADD to make atached MSDs behave like "large
    > floppies".



    Glad the Help file came in handy at least :-)


    >
    > Having DEVICE=D:\OS2\BOOT\USBCOM.SYS CDS:0 still resulted in the WPS
    > hanging before the Warp Centre appears. Then I noticed that my
    > USBCOM.sys was dated 2002 and I had one dated 2004 which I had failed to
    > copy over. Now I've done that but can't yet report on the effect of
    > that because...



    Err... Is that some sort of copy'n'paste error?

    USBCOM.SYS (USB Modem) does not have a /CDS: parameter.

    I'm guessing that you mean USBCDROM.ADD.

    Using the USB settings notebook you could just uncheck the USBCD/DVD
    support and that would resolve that problem by rewriting the config.sys
    without any USBCDROM.ADD line at all - or edit config.sys manually and
    REM the line eg:-

    REM ***** BASEDEV=USBCDROM.ADD


    >
    > I deferred to Bill H's urging to update the BIOS. Well having done
    > that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    > site was ambigous in having two files to apply. I tried one and the
    > QFlash utility said this was the wrong update so I tried the other one.
    > It went thru and said the update was successful and said to restart.
    > That was the last time I saw anything on the screen. I removed every
    > card on the board and tried again. Same - nothing. I've done flash
    > bios updates many times before (on customer's machines!) know it is
    > absolutely essential to have the correct update, this time I tried as
    > usual to be sure I have the correct update and still it fails! I think
    > Gigabyte has let me down. There is no chance of rolling back. The
    > board doesn't dual BIOSs. I will have to go thru the fuss and bother of
    > getting a warranty replacement.



    I have had that happen after flashing a mainboard BIOS. I was very lucky
    in that the system, although not displaying anything onscreen, seemed to
    be trying to boot from the floppy drive.

    I put together a floppy disk with the previous BIOS and flashing tool
    and a batch file to kick it all off and "rescued" that mainboard as the
    floppy drive was working and booted OK.


    >
    > This is posted from my Windoze machine. Don't know when OS/2 will be
    > back up.
    >
    > _Backtracking_ the story:
    > When the system came backup with the settings your USBCfg setup I got
    > 2 phantom drives instead of 1 and even more strangely my LAN connection
    > stopped! After remming out all the four BASEDEV=USBUHCD.sys lines and
    > restarting my LAN was back. There were still 4 phantom drives.



    Do you have a requirement for 4 * usbuhcd.sys lines?

    What does the Controllers Available (in Red) show about how many of what
    type of Controller you have available? You should set Controllers
    Installed to match those quantities - having said that rereading a
    section above it seems that is what you have done.

    What nic is in use - could be an IRQ sharing problem with the nic drivers.

    "phantom drives" are caused by incorrect parameters applied to the
    usbmsd.add and usbcdrom.add drivers - both of these drivers have default
    values where if there are no drive quantity parameters then you
    automatically get 1 USB CDROM, 1 USB Floppy and 1 USB Removable ie:--

    BASEDEV=USBCDROM.ADD is treated the same as BASEDEV=USBCDROM.ADD /CDS:1
    BASEDEV=USBMSD.ADD is treated the same as BASEDEV=USBMSD.ADD FLOPPIES:1
    /REMOVABLES:1


    >
    > The NIC is on IRQ3 whereas the board reports the USBUHCD devices on
    > IRQ11. I had a lot of trouble getting this board to allocate different
    > irqs to its 3 PCI slots and cut and tried a lot with the location of
    > this card and that in the slots.



    Sounds like you may be using some drivers that do not "share" IRQs very
    well...


    >
    > How update is my system?
    > I think it is fairly up-to-date (except for slip-ups). It is a MCP
    > 4.52 with FP5 plus all the other fixes applied as far as I can tell. I
    > can post the Syslevel data here if you like.
    >



    The USB base driver/controller is currently at buildlevel 10.158, both
    the MSD and CDROM are at 10.145.

    Also of relevence is the buildlevel of OS2DASD.DMD which should be kept
    current - I have 14.104 here but it may not be the very latest.


    Hope the above is of some help/use.

    Happy New Year

    Pete

  7. Re: USB sub-system: more observations and some conclusions

    William L. Hartzell wrote:
    > Sir:
    >
    > Victor Bien wrote:
    >
    >
    >> I deferred to Bill H's urging to update the BIOS. Well having done
    >> that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    >> site was ambigous in having two files to apply. I tried one and the
    >> QFlash utility said this was the wrong update so I tried the other
    >> one. It went thru and said the update was successful and said to
    >> restart. That was the last time I saw anything on the screen. I
    >> removed every card on the board and tried again. Same - nothing.
    >> I've done flash bios updates many times before (on customer's
    >> machines!) know it is absolutely essential to have the correct update,
    >> this time I tried as usual to be sure I have the correct update and
    >> still it fails! I think Gigabyte has let me down. There is no chance
    >> of rolling back. The board doesn't dual BIOSs. I will have to go
    >> thru the fuss and bother of getting a warranty replacement.
    >>

    > Hey, I am sorry. IF the machine won't come up, then there is nothing
    > you can do. That BIOS chip is surface mounted, so there is no pulling
    > it and finding an EEPROM burner to put back the old BIOS, right?
    >


    Yes 'fraid so. The box says the board has a dual bios but the one I
    have only has one surface mounted and not in sockets like in the
    picture. I think Gigabyte is being overrun by tech change like everyone
    else and can't keep everything together like their printed box and
    what's actually inside it, and whether their flash bios is any good it
    would seem!

    Anyway I have now installed another board this time a Gigabyte K8U.
    This has no built-in video and 5 PCI slots which I thought may give me
    more room to manoeuvre but so far that isn't proving true. Like all my
    experience OS/2 always has bucked around like a wild horse whenever you
    do something new. I'm back to working out which slots to plug the
    cards into. It has come up phantom drives and all but only stays up for
    about a minute or two and freezes!!! Now troubleshooting - irq clashes?
    rem out all USB stuff and see what happens, and so and so forth...

    Wait for the next exciting episode!

    Victor Bien

  8. Re: USB sub-system: more observations and some conclusions

    Victor Bien wrote:

    > Yes 'fraid so. The box says the board has a dual bios but the one I
    > have only has one surface mounted and not in sockets like in the
    > picture. I think Gigabyte is being overrun by tech change like everyone
    > else and can't keep everything together like their printed box and
    > what's actually inside it, and whether their flash bios is any good it
    > would seem!
    >
    > Anyway I have now installed another board this time a Gigabyte K8U.
    > This has no built-in video and 5 PCI slots which I thought may give me
    > more room to manoeuvre but so far that isn't proving true. Like all my
    > experience OS/2 always has bucked around like a wild horse whenever you
    > do something new. I'm back to working out which slots to plug the
    > cards into. It has come up phantom drives and all but only stays up for
    > about a minute or two and freezes!!! Now troubleshooting - irq clashes?
    > rem out all USB stuff and see what happens, and so and so forth...
    >
    > Wait for the next exciting episode!


    It sure caused bad excitement! After several freezes (with different
    combinations of cards in slots, irq etc.)it wouldn't boot! After
    getting at it from a floppy boot up (in which device=testcfg.sys had to
    be remmed out to stop crashes even here) LVM showed things were OK -
    volumes were still there! Chkdsk after cleanup put more than a dozen
    thing into \Found2 Amongst these things were \tcpip, my customised
    \usr, \PSFONTS, startup.cmd etc. All these had to be manually put back
    into place. I think this GA-K8U is hopeless! The board looks appealing
    enough but its hopeless for OS/2! Its going to be back to the Asus
    A7N266 with Duron CPU for now.

  9. Re: USB sub-system: more observations and some conclusions

    Victor Bien wrote:
    > Peter Brown wrote:
    >
    > I deferred to Bill H's urging to update the BIOS. Well having done
    > that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    > site was ambigous in having two files to apply. I tried one and the
    > QFlash utility said this was the wrong update so I tried the other one.
    > It went thru and said the update was successful and said to restart.
    > That was the last time I saw anything on the screen. I removed every
    > card on the board and tried again. Same - nothing. I've done flash
    > bios updates many times before (on customer's machines!) know it is
    > absolutely essential to have the correct update, this time I tried as
    > usual to be sure I have the correct update and still it fails! I think
    > Gigabyte has let me down. There is no chance of rolling back. The
    > board doesn't dual BIOSs. I will have to go thru the fuss and bother of
    > getting a warranty replacement.
    >


    Victor,

    If you want to try one last thing (if its not to late).
    I have had similar problems in the past with one bios update, and with
    memory or card changes. After the change the system was dead. Even
    taking out the hardware did not help.
    I found by taking the power plug out and the c-mos battery on the system
    board the system would power up.
    I also had one bios update on a Compaq that finished OK, but did not
    recover and it was toast. It was a new system & I got a board to swap
    in from them. The newer Compaq's had the emergency recover then I never
    had to try.

    Long shot anyway
    Rich W

  10. Re: USB sub-system: more observations and some conclusions

    Hi Victor

    Victor Bien wrote:
    > William L. Hartzell wrote:
    >
    >> Sir:
    >>
    >> Victor Bien wrote:
    >>
    >>
    >>> I deferred to Bill H's urging to update the BIOS. Well having done
    >>> that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    >>> site was ambigous in having two files to apply. I tried one and the
    >>> QFlash utility said this was the wrong update so I tried the other
    >>> one. It went thru and said the update was successful and said to
    >>> restart. That was the last time I saw anything on the screen. I
    >>> removed every card on the board and tried again. Same - nothing.
    >>> I've done flash bios updates many times before (on customer's
    >>> machines!) know it is absolutely essential to have the correct
    >>> update, this time I tried as usual to be sure I have the correct
    >>> update and still it fails! I think Gigabyte has let me down. There
    >>> is no chance of rolling back. The board doesn't dual BIOSs. I will
    >>> have to go thru the fuss and bother of getting a warranty replacement.
    >>>

    >> Hey, I am sorry. IF the machine won't come up, then there is nothing
    >> you can do. That BIOS chip is surface mounted, so there is no pulling
    >> it and finding an EEPROM burner to put back the old BIOS, right?
    >>

    >
    >
    > Yes 'fraid so. The box says the board has a dual bios but the one I
    > have only has one surface mounted and not in sockets like in the
    > picture. I think Gigabyte is being overrun by tech change like everyone
    > else and can't keep everything together like their printed box and
    > what's actually inside it, and whether their flash bios is any good it
    > would seem!
    >
    > Anyway I have now installed another board this time a Gigabyte K8U.
    > This has no built-in video and 5 PCI slots which I thought may give me
    > more room to manoeuvre but so far that isn't proving true. Like all my
    > experience OS/2 always has bucked around like a wild horse whenever you
    > do something new. I'm back to working out which slots to plug the
    > cards into. It has come up phantom drives and all but only stays up for
    > about a minute or two and freezes!!! Now troubleshooting - irq clashes?
    > rem out all USB stuff and see what happens, and so and so forth...



    The phantom drives are almost definitely unattached usb drives - cdroms,
    floppies and removables.

    The rest of the problems sounds like a couple of (I hate to say it) BIOS
    settings may need changing.

    Things like:-

    Play and Play OS: No
    Legacy USB (may talk in terms of USB Mouse and Keyboard) : No
    OS select for DRAM > 64Mb RAM: Non-OS2

    Above settings should work.

    I have seen a fairly recent BIOS setup guide - may have been for eCS but
    I think it could well apply to OS/2 :-)

    Sorry, not sure where to locate it at the moment... Might be as easy to
    try to get the author to post it in this ng.

    Regards

    Pete

  11. Re: USB sub-system: more observations and some conclusions

    Peter Brown wrote:
    > Hi Victor


    [cuts]

    >> Having DEVICE=D:\OS2\BOOT\USBCOM.SYS CDS:0 still resulted in the WPS
    >> hanging before the Warp Centre appears. Then I noticed that my
    >> USBCOM.sys was dated 2002 and I had one dated 2004 which I had failed
    >> to copy over. Now I've done that but can't yet report on the effect
    >> of that because...

    >
    > Err... Is that some sort of copy'n'paste error?


    Oops yes.

    >
    > USBCOM.SYS (USB Modem) does not have a /CDS: parameter.
    >
    > I'm guessing that you mean USBCDROM.ADD.


    Yes

    >
    > Using the USB settings notebook you could just uncheck the USBCD/DVD
    > support and that would resolve that problem by rewriting the config.sys
    > without any USBCDROM.ADD line at all - or edit config.sys manually and
    > REM the line eg:-
    >
    > REM ***** BASEDEV=USBCDROM.ADD


    As I mentioned a couple of times I can't do this. When remmed out,
    the WPS hangs after the sound emits and before the Warpcenter tries to
    come up. Removing the Warpcenter from the SET AUTOSTART spec has no effect.

    >>
    >> I deferred to Bill H's urging to update the BIOS. Well having done
    >> that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    >> site was ambigous in having two files to apply. I tried one and the
    >> QFlash utility said this was the wrong update so I tried the other
    >> one. It went thru and said the update was successful and said to
    >> restart. That was the last time I saw anything on the screen. I
    >> removed every card on the board and tried again. Same - nothing.
    >> I've done flash bios updates many times before (on customer's
    >> machines!) know it is absolutely essential to have the correct update,
    >> this time I tried as usual to be sure I have the correct update and
    >> still it fails! I think Gigabyte has let me down. There is no chance
    >> of rolling back. The board doesn't dual BIOSs. I will have to go
    >> thru the fuss and bother of getting a warranty replacement.

    >
    >
    >
    > I have had that happen after flashing a mainboard BIOS. I was very lucky
    > in that the system, although not displaying anything onscreen, seemed to
    > be trying to boot from the floppy drive.
    >
    > I put together a floppy disk with the previous BIOS and flashing tool
    > and a batch file to kick it all off and "rescued" that mainboard as the
    > floppy drive was working and booted OK.


    Hmm interesting. However I didn't bother to save a copy of the
    original bios.

    >>
    >> This is posted from my Windoze machine. Don't know when OS/2 will
    >> be back up.
    >>
    >> _Backtracking_ the story:
    >> When the system came backup with the settings your USBCfg setup I
    >> got 2 phantom drives instead of 1 and even more strangely my LAN
    >> connection stopped! After remming out all the four
    >> BASEDEV=USBUHCD.sys lines and restarting my LAN was back. There were
    >> still 4 phantom drives.

    >
    >
    >
    > Do you have a requirement for 4 * usbuhcd.sys lines?
    >
    > What does the Controllers Available (in Red) show about how many of what
    > type of Controller you have available? You should set Controllers
    > Installed to match those quantities - having said that rereading a
    > section above it seems that is what you have done.
    >
    > What nic is in use - could be an IRQ sharing problem with the nic drivers.


    A D-Link 528 which uses a RTL8029AS chip.

    >
    > "phantom drives" are caused by incorrect parameters applied to the
    > usbmsd.add and usbcdrom.add drivers - both of these drivers have default
    > values where if there are no drive quantity parameters then you
    > automatically get 1 USB CDROM, 1 USB Floppy and 1 USB Removable ie:--
    >
    > BASEDEV=USBCDROM.ADD is treated the same as BASEDEV=USBCDROM.ADD /CDS:1
    > BASEDEV=USBMSD.ADD is treated the same as BASEDEV=USBMSD.ADD FLOPPIES:1
    > /REMOVABLES:1
    >
    >>
    >> The NIC is on IRQ3 whereas the board reports the USBUHCD devices on
    >> IRQ11. I had a lot of trouble getting this board to allocate
    >> different irqs to its 3 PCI slots and cut and tried a lot with the
    >> location of this card and that in the slots.

    >
    >
    > Sounds like you may be using some drivers that do not "share" IRQs very
    > well...


    I don't think it's a driver issue. I think it's a mainboard issue.
    I'll be posting something under that heading soon.

    >>
    >> How update is my system?
    >> I think it is fairly up-to-date (except for slip-ups). It is a MCP
    >> 4.52 with FP5 plus all the other fixes applied as far as I can tell.
    >> I can post the Syslevel data here if you like.
    >>

    >
    >
    > The USB base driver/controller is currently at buildlevel 10.158, both
    > the MSD and CDROM are at 10.145.
    >
    > Also of relevence is the buildlevel of OS2DASD.DMD which should be kept
    > current - I have 14.104 here but it may not be the very latest.


    usbuhcd.sys 10.58
    usbohcd.sys 10.123
    usbehcd.sys 10.123
    usbd.sys 10.117
    usbhid.sys 10.97
    usmsd.add 10.58
    usbcdrom.add 10.145
    os2dasd,dmd 14.100

    >
    >
    > Hope the above is of some help/use.


    Too right it helps! I don't think trying to self support an OS/2
    system would be possible without people like you around! :-)

    >
    > Happy New Year
    >
    > Pete


    Return the good wishes.

    Big thanks has been implicit now I say it! If I don't say it, it is
    because I don't like being "too treacly". :-)

    --
    To reply by e-mail edit this address to the correct form: vbien at
    attglobal dot net
    

  12. Re: USB sub-system: more observations and some conclusions

    Sir:

    Victor Bien wrote:
    > Peter Brown wrote:
    >> Hi Victor

    >
    > [cuts]
    >
    >>> Having DEVICE=D:\OS2\BOOT\USBCOM.SYS CDS:0 still resulted in the
    >>> WPS hanging before the Warp Centre appears. Then I noticed that my
    >>> USBCOM.sys was dated 2002 and I had one dated 2004 which I had failed
    >>> to copy over. Now I've done that but can't yet report on the effect
    >>> of that because...

    >>
    >> Err... Is that some sort of copy'n'paste error?

    >
    > Oops yes.
    >
    >>
    >> USBCOM.SYS (USB Modem) does not have a /CDS: parameter.
    >>
    >> I'm guessing that you mean USBCDROM.ADD.

    >
    > Yes
    >
    >>
    >> Using the USB settings notebook you could just uncheck the USBCD/DVD
    >> support and that would resolve that problem by rewriting the
    >> config.sys without any USBCDROM.ADD line at all - or edit config.sys
    >> manually and REM the line eg:-
    >>
    >> REM ***** BASEDEV=USBCDROM.ADD

    >
    > As I mentioned a couple of times I can't do this. When remmed out,
    > the WPS hangs after the sound emits and before the Warpcenter tries to
    > come up. Removing the Warpcenter from the SET AUTOSTART spec has no
    > effect.
    >



    This is a MMOS2 error. You got something messed up in the multimedia
    department, like the USBCDROM checked as the default CDrom device
    (Multimedia Setup object). Uncheck that and try rem the basedev line.
    --
    Bill
    Thanks a Million!

  13. Re: USB sub-system: more observations and some conclusions

    William L. Hartzell wrote:
    > Sir:
    >
    > Victor Bien wrote:
    >
    >> Peter Brown wrote:
    >>
    >>> Hi Victor

    >>
    >>
    >> [cuts]
    >>
    >>>> Having DEVICE=D:\OS2\BOOT\USBCOM.SYS CDS:0 still resulted in the
    >>>> WPS hanging before the Warp Centre appears. Then I noticed that my
    >>>> USBCOM.sys was dated 2002 and I had one dated 2004 which I had
    >>>> failed to copy over. Now I've done that but can't yet report on the
    >>>> effect of that because...
    >>>
    >>>
    >>> Err... Is that some sort of copy'n'paste error?

    >>
    >>
    >> Oops yes.
    >>
    >>>
    >>> USBCOM.SYS (USB Modem) does not have a /CDS: parameter.
    >>>
    >>> I'm guessing that you mean USBCDROM.ADD.

    >>
    >>
    >> Yes
    >>
    >>>
    >>> Using the USB settings notebook you could just uncheck the USBCD/DVD
    >>> support and that would resolve that problem by rewriting the
    >>> config.sys without any USBCDROM.ADD line at all - or edit config.sys
    >>> manually and REM the line eg:-
    >>>
    >>> REM ***** BASEDEV=USBCDROM.ADD

    >>
    >>
    >> As I mentioned a couple of times I can't do this. When remmed out,
    >> the WPS hangs after the sound emits and before the Warpcenter tries to
    >> come up. Removing the Warpcenter from the SET AUTOSTART spec has no
    >> effect.
    >>

    >
    >
    > This is a MMOS2 error. You got something messed up in the multimedia
    > department, like the USBCDROM checked as the default CDrom device
    > (Multimedia Setup object). Uncheck that and try rem the basedev line.


    Tried that - still no go. The default CDROM device was of the form
    CDaudio01. I notice that in my Compact Disc objects in my Multimedia
    folder has three objects: one for CDaudio01, the 2nd for CDaudio02 and
    the third has a properties notebook that is different from the first
    two. Is that significant? I'm not sure I'd want to delete the
    parameter spec from the first two though.

    --
    To reply by e-mail edit this address to the correct form: vbien at
    attglobal dot net
    

  14. Re: USB sub-system: more observations and some conclusions

    Victor Bien wrote:

    > I deferred to Bill H's urging to update the BIOS. Well having done
    > that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    > site was ambigous in having two files to apply. I tried one and the
    > QFlash utility said this was the wrong update so I tried the other one.
    > It went thru and said the update was successful and said to restart.
    > That was the last time I saw anything on the screen. I removed every
    > card on the board and tried again. Same - nothing. I've done flash
    > bios updates many times before (on customer's machines!) know it is
    > absolutely essential to have the correct update, this time I tried as
    > usual to be sure I have the correct update and still it fails! I think
    > Gigabyte has let me down. There is no chance of rolling back. The


    Well it probably isn't Gigabyte's fault. Only NOW I've discovered
    that there is something wrong with the floppy drive system on my Windoze
    machine. The copy of the bios file was corrupt!

    I have checked and I don't think it's the drive itself. The problem
    seems to be some corruption in the floppy disk controller system. I
    wonder now whether a virus has crept in despite all the protection and
    corrupted the flash bios on the doze machine???

    The symptoms are that a freshly formatted diskette done by the OS/2
    machine gives no errors with chkdsk (as you'd expect) but after copying
    a file to that diskette in the doze machine and applying chkdsk again I
    now always get cross linked files!!

    I will be trying to flash the doze bios to try to fix the problem
    there but will have to make a bootable CD to do it since with a sick
    floppy system I might finish up with the doze bios getting wrecked as well!

    Such is computing life!

    --
    To reply by e-mail edit this address to the correct form: vbien at
    attglobal dot net
    

  15. Re: USB sub-system: more observations and some conclusions

    Sir:

    Victor Bien wrote:
    > Victor Bien wrote:
    >
    >> I deferred to Bill H's urging to update the BIOS. Well having done
    >> that the damn #@%$&* mainboard has stopped working!!!! The Gigabyte
    >> site was ambigous in having two files to apply. I tried one and the
    >> QFlash utility said this was the wrong update so I tried the other
    >> one. It went thru and said the update was successful and said to
    >> restart. That was the last time I saw anything on the screen. I
    >> removed every card on the board and tried again. Same - nothing.
    >> I've done flash bios updates many times before (on customer's
    >> machines!) know it is absolutely essential to have the correct update,
    >> this time I tried as usual to be sure I have the correct update and
    >> still it fails! I think Gigabyte has let me down. There is no chance
    >> of rolling back. The

    >
    > Well it probably isn't Gigabyte's fault. Only NOW I've discovered
    > that there is something wrong with the floppy drive system on my Windoze
    > machine. The copy of the bios file was corrupt!
    >
    > I have checked and I don't think it's the drive itself. The problem
    > seems to be some corruption in the floppy disk controller system. I
    > wonder now whether a virus has crept in despite all the protection and
    > corrupted the flash bios on the doze machine???
    >
    > The symptoms are that a freshly formatted diskette done by the OS/2
    > machine gives no errors with chkdsk (as you'd expect) but after copying
    > a file to that diskette in the doze machine and applying chkdsk again I
    > now always get cross linked files!!
    >
    > I will be trying to flash the doze bios to try to fix the problem
    > there but will have to make a bootable CD to do it since with a sick
    > floppy system I might finish up with the doze bios getting wrecked as well!
    >
    > Such is computing life!
    >

    Before you run off to flash the bios of the windows machine, go to
    Sourceforge and get the ClamAV program for Windows and update it with
    the latest AV file. Burn that to a Win bootable CD and start the
    windows machine with it and scan that windows machine. This could fix
    the problem without running the risk of having two dead computers.
    --
    Bill
    Thanks a Million!

  16. Re: OT: USB sub-system: more observations and some conclusions

    William L. Hartzell wrote:

    [big cuts]

    >>
    >> I will be trying to flash the doze bios to try to fix the problem
    >> there but will have to make a bootable CD to do it since with a sick
    >> floppy system I might finish up with the doze bios getting wrecked as
    >> well!
    >>
    >> Such is computing life!
    >>

    > Before you run off to flash the bios of the windows machine, go to
    > Sourceforge and get the ClamAV program for Windows and update it with
    > the latest AV file. Burn that to a Win bootable CD and start the
    > windows machine with it and scan that windows machine. This could fix
    > the problem without running the risk of having two dead computers.


    As the following posting to another group shows I think the problem
    is not in the BIOS of the W computer so I won't be trying to flash it
    with attendant risks...

    I'm not sure that ClamAV would show things which PCCillin I use could
    miss.

    --- Begin OT quote ---
    My system is a WinXP Pro SP2 with all patches up-to-date. It is
    protected by PCCillin which is up-to-date. Scans do not detect any viruses.

    Recently I used it to copy a flashbios file onto a floppy to flash
    up-date another computer. The flash process apparently went thru fine,
    the Gigabyte QFlash system congratulated me that the process was
    "successful" but when the mainboard tried to reboot it beeped indicating
    that there was a problem with it and refused to come up again. The mb
    was toast!

    Poking around in the days following chkdsk found that the floppy used
    to transfer the BIOS file was corrupt! It had cross linked files!!
    Doing an FC on the bios file on the diskette with the one on the hard
    disk from which it was copied showed screen-scrollfulls of
    "differences"!!! Bugger, hell, damn and all that!

    While the computer was in this state it could not format any
    diskettes. It said that every one of them had a bad sector 0. A
    freshly formatted diskette from another computer which chkdsk OK would
    come up with cross linked files after a file was copied in the problem
    computer and checked again.

    Sometime previously I had strange behaviour with the floppy drive
    when upon clicking on a filename in the Windows Explorer window would
    cause the filename to get corrupted. In some cases non-ascii characters
    in the filename would appear in place of the character previously there.

    At that time I found the problem was not the floppy drive itself
    established by substituting it with a couple of other ones. At that
    time I got the problem to go away by switching the CMOS setting for the
    FD drive to none, saving it, going back into the CMOS screen and
    resetting it to 1.44 MB etc.

    Now I got the above disaster!

    I then did the following checks:
    * put an old HD in and booted up to DOS (6.2). Formatting and copying
    worked OK. This seems to show that my problem is not due to faulty
    hardware or a problem in the BIOS;

    * I reconnected the working HD and booted to Safe Mode and tried
    formatting and copying. Result OK. Did (does) that suggest the problem
    was (is) in WinXP normal mode?

    * I rebooted to normal mode and tried formatting and copying.
    Strangely, the problem has gone away!!!!!! Will it creep back in when
    I'm not watching????

    I looked up Google about this and did not find anything. Has anyone
    else experienced this?

    --- end OT quote ---

    --
    To reply by e-mail edit this address to the correct form: vbien at
    attglobal dot net

+ Reply to Thread