Re: PROBLEM: network driver causes kernel panic

From: Mikael Pettersson (mikpe_at_csd.uu.se)
Date: 06/05/04

  • Next message: Flavio Stanchina: "Re: 2.6.6 Kernel Oops and ATI 3.9.0 proprietary driver" 
    Date:	Sat, 5 Jun 2004 14:15:04 +0200 (MEST)
To: jgarzik@pobox.com

    On Fri, 04 Jun 2004 16:47:13 -0400, Jeff Garzik wrote:
    >Mikael Pettersson wrote:
    >> This confirms that eth1 is a 21041 driven by the de2104x driver.
    >>
    >> I've seen something very similar to Casey's problem, on a PowerMac
    >> with a built-in 21041. Booting it with no network cable connected
    >> causes a timer to hit a BUG() in de2104x about a second after
    >> the device is ifup:d.
    >>
    >> The 2.4 kernel's tulip driver works just fine.
    >>
    >> I reported this last year, but nothing happened.
    >
    >
    >Well, I'm very interested in debugging it. There were a flurry of
    >de2104x patches in the past year, I thought that took care of the issues.
    >
    >Please email details to netdev@oss.sgi.com and jgarzik@pobox.com...

    Booting 2.6.7-rc1 with the de2104x driver built-in and eth0
    disconnected from the LAN leads to the following oops about
    a second after INIT tried to ifup eth0:

    eth0: timeout expired stopping DMA
    kernel BUG in de_set_media at drivers/net/tulip/de2104x.c:919!
    <register dump omitted>
    Call trace:
    de21041_media_timer
    run_timer_softirq
    __do_softirq
    do_softirq
    timer_interrupt
    ret_from_except
    ppc6xx_idle
    cpu_idle
    rest_init
    start_kernel

    The PowerPC kernel decides to panic() after a brief delay
    at this point, so I can't capture the oops text except by
    typing it down manually. Besides, I doubt the ppc register
    dump would be useful; we know which BUG() was hit.

    /Mikael
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  • Next message: Flavio Stanchina: "Re: 2.6.6 Kernel Oops and ATI 3.9.0 proprietary driver"

    Relevant Pages

    • Re: [GIT PATCH] scsi bug fixes for 2.6.13
      ... Just to make sure everyone agrees on this - there's currently a know bug ... It only was hit by one ... The fix is known - at least revert my patch to dc395x commited to ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: [GIT PATCH] scsi bug fixes for 2.6.13
      ... > non-highmem machines. ... It only was hit by one ... This will fix this bug, leaving another one, which is much ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • 2.6.10-rc2-mm2 crashes in early boot
      ... I enabled crashdump on 2.6.10-rc2-mm2 with CONFIG_SMP=y and hit the bug ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: [PATCH] Fix apm suspend with cs4231 based sound cards
      ... > There was a mutex deadlock in the suspend code, and I guess it hit ... The bug is fixed in 2.6.7-rc3; that'll teach me to test the lastest ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)
    • Re: Opinions on PGI vs. Lahey Fortran
      ... >> it hit a segmentation violation or other similarly fatal error. ... > that kind of problem is common with many compilers. ... > ideal behavior, but not a bug. ... > such as changing the value of i inside of the loop). ...
      (comp.lang.fortran)