TX unit hang a UBUNTU only sickness? - Ubuntu

This is a discussion on TX unit hang a UBUNTU only sickness? - Ubuntu ; OK : Subject: TX unit hang a UBUNTU only sickness? Is kind of a strong wording. BUT I'M encountering this problem since Ubuntu 7.10 (before that I used 7.04 with a different NIC) and this NIC. 02:09.0 Ethernet controller: Intel ...

+ Reply to Thread
Results 1 to 3 of 3

Thread: TX unit hang a UBUNTU only sickness?

  1. TX unit hang a UBUNTU only sickness?


    OK :


    Subject: TX unit hang a UBUNTU only sickness?
    Is kind of a strong wording.

    BUT I'M encountering this problem since Ubuntu 7.10 (before
    that I used 7.04 with a different NIC) and this NIC.

    02:09.0 Ethernet controller: Intel Corporation 82541PI
    Gigabit Ethernet Controller (rev 05)

    I've (about the same time as I installed Ubuntu) had
    Mandriva installed, but I didn't have this problem. Not
    having the time to customize and set up two OS, I choose
    Ubuntu and kept working on it. Came from Debian and got used
    to their package-management, lost my working OS through a
    upgrade of the wrong installation (instead of testing I
    upgraded stable to unstable. Ubuntu was the best choice in
    this situation... :-)

    OK I went to the website:

    "http://e1000.sourceforge.net/doku.php"

    where I got a couple of tips:
    *Upgraded to a newest driver (08.0x.0y something)
    *run a script if the eeprom needed to be changed

    The script told me that the eeprom was fine wasn't surprised
    as the Mandriva 2008 "official" didn't have any issue with
    this.

    All of that didn't solve the issue. I can unload the older
    module and insert the new one it's the same. "ethtool -K tso
    off eth0" with either module didn't help either.

    Very early in the search I found that mem=4096mb to the
    grub-boot line would help. While in fact the "Tx unit hang"
    messages disappeared, I have now instead of 4GB RAM

    total used free shared buffers cached
    Mem: 3031564 852252 2179312 0 18776 406956
    -/+ buffers/cache: 426520 2605044
    Swap: 1020116 0 1020116

    For the heck of it, this afternoon I toyed with Mandriva and
    sure enough, it sees all the 4GB RAM, and no TX unit hang
    message, but then I have X-Windows problems, that would need
    more research.

    Kernel is now self-compiled: "2.6.25.9-custom-02" and yet
    still the same issue, so I wonder if this is something else
    with the Ubuntu code itself.

    All this on a 64bit AMD64 box running the 64bit
    installation, now I've got the 32bit version too, here
    installed, can't recall if I had the same issue with it. I
    just might fire it up to check.

    so my rant is off my chest.....

    Dragomir Kollaric
    --
    This signature is licensed under the GPL and may be
    freely distributed as long as a copy of the GPL is included... :-)


  2. Re: TX unit hang a UBUNTU only sickness?

    Ubuntu does not write NIC card drivers.It is a package of standard
    kernels with a bunch of standard programs, done in a nice way and
    supplied with updates.

    You can try to compile a newer kernel just to try it out to isolate
    the problem.

    i

    On 2008-08-02, Dragomir Kollaric wrote:
    >
    > OK :
    >
    >
    > Subject: TX unit hang a UBUNTU only sickness?
    > Is kind of a strong wording.
    >
    > BUT I'M encountering this problem since Ubuntu 7.10 (before
    > that I used 7.04 with a different NIC) and this NIC.
    >
    > 02:09.0 Ethernet controller: Intel Corporation 82541PI
    > Gigabit Ethernet Controller (rev 05)
    >
    > I've (about the same time as I installed Ubuntu) had
    > Mandriva installed, but I didn't have this problem. Not
    > having the time to customize and set up two OS, I choose
    > Ubuntu and kept working on it. Came from Debian and got used
    > to their package-management, lost my working OS through a
    > upgrade of the wrong installation (instead of testing I
    > upgraded stable to unstable. Ubuntu was the best choice in
    > this situation... :-)
    >
    > OK I went to the website:
    >
    > "http://e1000.sourceforge.net/doku.php"
    >
    > where I got a couple of tips:
    > *Upgraded to a newest driver (08.0x.0y something)
    > *run a script if the eeprom needed to be changed
    >
    > The script told me that the eeprom was fine wasn't surprised
    > as the Mandriva 2008 "official" didn't have any issue with
    > this.
    >
    > All of that didn't solve the issue. I can unload the older
    > module and insert the new one it's the same. "ethtool -K tso
    > off eth0" with either module didn't help either.
    >
    > Very early in the search I found that mem=4096mb to the
    > grub-boot line would help. While in fact the "Tx unit hang"
    > messages disappeared, I have now instead of 4GB RAM
    >
    > total used free shared buffers cached
    > Mem: 3031564 852252 2179312 0 18776 406956
    > -/+ buffers/cache: 426520 2605044
    > Swap: 1020116 0 1020116
    >
    > For the heck of it, this afternoon I toyed with Mandriva and
    > sure enough, it sees all the 4GB RAM, and no TX unit hang
    > message, but then I have X-Windows problems, that would need
    > more research.
    >
    > Kernel is now self-compiled: "2.6.25.9-custom-02" and yet
    > still the same issue, so I wonder if this is something else
    > with the Ubuntu code itself.
    >
    > All this on a 64bit AMD64 box running the 64bit
    > installation, now I've got the 32bit version too, here
    > installed, can't recall if I had the same issue with it. I
    > just might fire it up to check.
    >
    > so my rant is off my chest.....
    >
    > Dragomir Kollaric


    --
    Due to extreme spam originating from Google Groups, and their inattention
    to spammers, I and many others block all articles originating
    from Google Groups. If you want your postings to be seen by
    more readers you will need to find a different means of
    posting on Usenet.
    http://improve-usenet.org/

  3. Re: TX unit hang a UBUNTU only sickness?

    On 2008-08-02, Ignoramus23772 hit the keyboard and wrote:
    > Ubuntu does not write NIC card drivers.It is a package of standard
    > kernels with a bunch of standard programs, done in a nice way and
    > supplied with updates.
    >
    > You can try to compile a newer kernel just to try it out to isolate
    > the problem.


    I did, I posted also the version I just compiled. Same
    thing, if I don't leave the "mem=xyz" line in menu.lst.

    In any even, this problem seems to be related to the 64bit
    version of Ubuntu. I do have a (somewhat working) 32bit
    Installation of the "Studio Ubuntu" and there I don't get
    this error at all. Besides in this (32bit) I have also all
    4GB RAM to work with.

    I'm just going to leave it for now, as so far I hadn't have
    the need to use more then 3GB RAM. And if I do need it, I'll
    just boot the 32bit version. I've got some work to do as
    well.


    >
    > i
    >
    > On 2008-08-02, Dragomir Kollaric wrote:
    >>
    >> OK :
    >>
    >>
    >> Subject: TX unit hang a UBUNTU only sickness?
    >> Is kind of a strong wording.
    >>
    >> BUT I'M encountering this problem since Ubuntu 7.10 (before
    >> that I used 7.04 with a different NIC) and this NIC.
    >>
    >> 02:09.0 Ethernet controller: Intel Corporation 82541PI
    >> Gigabit Ethernet Controller (rev 05)
    >>
    >> I've (about the same time as I installed Ubuntu) had
    >> Mandriva installed, but I didn't have this problem. Not
    >> having the time to customize and set up two OS, I choose
    >> Ubuntu and kept working on it. Came from Debian and got used
    >> to their package-management, lost my working OS through a
    >> upgrade of the wrong installation (instead of testing I
    >> upgraded stable to unstable. Ubuntu was the best choice in
    >> this situation... :-)
    >>
    >> OK I went to the website:
    >>
    >> "http://e1000.sourceforge.net/doku.php"
    >>
    >> where I got a couple of tips:
    >> *Upgraded to a newest driver (08.0x.0y something)
    >> *run a script if the eeprom needed to be changed
    >>
    >> The script told me that the eeprom was fine wasn't surprised
    >> as the Mandriva 2008 "official" didn't have any issue with
    >> this.
    >>
    >> All of that didn't solve the issue. I can unload the older
    >> module and insert the new one it's the same. "ethtool -K tso
    >> off eth0" with either module didn't help either.
    >>
    >> Very early in the search I found that mem=4096mb to the
    >> grub-boot line would help. While in fact the "Tx unit hang"
    >> messages disappeared, I have now instead of 4GB RAM
    >>
    >> total used free shared buffers cached
    >> Mem: 3031564 852252 2179312 0 18776 406956
    >> -/+ buffers/cache: 426520 2605044
    >> Swap: 1020116 0 1020116
    >>
    >> For the heck of it, this afternoon I toyed with Mandriva and
    >> sure enough, it sees all the 4GB RAM, and no TX unit hang
    >> message, but then I have X-Windows problems, that would need
    >> more research.
    >>
    >> Kernel is now self-compiled: "2.6.25.9-custom-02" and yet
    >> still the same issue, so I wonder if this is something else
    >> with the Ubuntu code itself.
    >>
    >> All this on a 64bit AMD64 box running the 64bit
    >> installation, now I've got the 32bit version too, here
    >> installed, can't recall if I had the same issue with it. I
    >> just might fire it up to check.
    >>
    >> so my rant is off my chest.....
    >>
    >> Dragomir Kollaric

    >




    Dragomir Kollaric
    --
    This signature is licensed under the GPL and may be
    freely distributed as long as a copy of the GPL is included... :-)


+ Reply to Thread