Results 1 to 4 of 4

Thread: Actually looks like a kernel prob at first glance

  1. #1

    Default Actually looks like a kernel prob at first glance

    I've been running openSUsE 11.0 for a while and have kept right up to date on patches. (Though I'm probably a little bit slack at rebooting when a patch asks for a reboot.)

    Last night, my system locked up with the following entry in /var/log/messages:

    May 10 01:16:33 server kernel: invalid opcode: 0000 [1] SMP
    May 10 01:16:33 server kernel: last sysfs file: /sys/devices/system/cpu/cpu3/cache/index2/shared_cpu_map
    May 10 01:16:33 server kernel: CPU 0
    May 10 01:16:33 server kernel: Modules linked in: snd_seq_dummy snd_pcm_oss snd_mixer_oss snd_seq snd_seq_device snd_hda_intel snd_pcm snd_t
    imer snd_page_alloc snd_hwdep snd af_packet it87 hwmon_vid ip6_tables cpufreq_conservative cpufreq_userspace cpufreq_powersave powernow_k8 i
    pt_MASQUERADE iptable_nat nf_nat xt_limit nf_conntrack_ipv4 xt_state nf_conntrack ipt_LOG xt_tcpudp iptable_filter ip_tables x_tables binfmt
    _misc fuse ext2 loop ppdev rtc_cmos rtc_core parport_pc parport floppy i2c_piix4 ohci1394 rtc_lib fglrx(P) button ieee1394 r8169 sr_mod wmi
    i2c_core cdrom sg soundcore dm_mirror dm_log usbhid hid ff_memless ehci_hcd ohci_hcd usbcore sd_mod dm_snapshot dm_mod edd ext3 mbcache jbd
    fan pata_atiixp ahci pata_jmicron libata dock sym53c8xx scsi_transport_spi scsi_mod thermal processor [last unloaded: snd]
    May 10 01:16:33 server kernel: Pid: 6292, comm: Xorg Tainted: P N 2.6.25.20-0.1-default #1
    May 10 01:16:33 server kernel: RIP: 0010:[<ffffffff8020b6f1>] [<ffffffff8020b6f1>] do_notify_resume+0x4de/0x8e1
    May 10 01:16:33 server kernel: RSP: 0000:ffff81016f8d9df8 EFLAGS: 00210246
    May 10 01:16:33 server kernel: RAX: ffffffffffffffff RBX: 0000000000000000 RCX: 0000000000000000
    May 10 01:16:33 server kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000
    May 10 01:16:33 server kernel: RBP: ffff81016f8d9f48 R08: 0000000000000001 R09: ffff81000d800000
    May 10 01:16:33 server kernel: R10: 0000000000000002 R11: ffff81016f8d9d08 R12: 00007fffc921b158
    May 10 01:16:33 server kernel: R13: ffff81016f8d9f58 R14: ffff810179df4700 R15: 00007fffc921b320
    May 10 01:16:33 server kernel: FS: 00007f21c11dd6f0(0000) GS:ffffffff80632000(0000) knlGS:00000000b524db90
    May 10 01:16:33 server kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
    May 10 01:16:33 server kernel: CR2: 00007f551e792000 CR3: 00000001931f3000 CR4: 00000000000006e0
    May 10 01:16:33 server kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
    May 10 01:16:33 server kernel: DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
    May 10 01:16:33 server kernel: Process Xorg (pid: 6292, threadinfo ffff81016f8d8000, task ffff810179df4700)
    May 10 01:16:33 server kernel: Stack: 0000000400000000 ffff810179df4e58 0000000e00000000 0000000000000000
    May 10 01:16:33 server kernel: 0000000000000000 0000000000000000 0000000000000000 ffff810100000000
    May 10 01:16:33 server kernel: 0000000000000020 0000000000000001 ffff81016f8d9f28 ffffffff802a0a2a
    May 10 01:16:33 server kernel: Call Trace:
    May 10 01:16:33 server kernel: [<ffffffff8020c397>] ptregscall_common+0x67/0xb0
    May 10 01:16:33 server kernel: [<ffffffff8020c0a9>] sysret_signal+0x2b/0x45
    May 10 01:16:33 server kernel: [<00007f21bf7bd5d3>]
    May 10 01:16:33 server kernel:
    May 10 01:16:33 server kernel:
    May 10 01:16:33 server kernel: Code: 00 89 8d e4 fe ff ff 49 8b 45 60 89 f7 49 89 84 24 90 00 00 00 89 bd e8 fe ff ff 49 8b 45 58 89 f1 49 8
    9 84 24 a0 00 00 00 89 8d <ec> fe ff ff 49 8b 45 50 89 f7 49 89 84 24 98 00 00 00 89 bd f0
    May 10 01:16:33 server kernel: RIP [<ffffffff8020b6f1>] do_notify_resume+0x4de/0x8e1
    May 10 01:16:33 server kernel: RSP <ffff81016f8d9df8>
    May 10 01:16:33 server kernel: ---[ end trace e3701a04d80856d3 ]---

    I was going to attach the hardware list, but after discovering that it was nearly 18,000 lines, I decided that it might not be a good idea. OTOH, if someone wants to investigate the problem, I'm happy to provide the list.

    Thanks,

    -Don

  2. #2
    Join Date
    Nov 2008
    Location
    N. Wales
    Posts
    1,028

    Default Re: Actually looks like a kernel prob at first glance

    gingrich adjusted his/her AFB on Sunday 10 May 2009 01:36 to write:

    >
    > I've been running openSUsE 11.0 for a while and have kept right up to
    > date on patches. (Though I'm probably a little bit slack at rebooting
    > when a patch asks for a reboot.)


    Do you mean that there might have been a kernel update and you have not
    rebooted?

    The error as far as I can see points to xorg but could be related to your
    graphics which might point to a kernel update an using an outdated driver (
    this is just at a quick glance )

    Have a look in your logs ( YaST) to see what has been updated recently if
    you do auto-updates.

    does the system run OK after the reboot?

    If so then this could be the prob, if it is still unstable then we will need
    more info on the system, hardware, drivers, what you were doing when it went
    belly up etc....


    HTH
    --
    Mark

    Nullus in verba
    Nil illigitimi carborundum


  3. #3

    Default Re: Actually looks like a kernel prob at first glance

    Actually. the lockup only occurred after I had rebooted. (Probably silly, but there's a smidge of old time Linux user in me that tries to see how long the system can run without a reboot -- I was up close to 180 days when I rebooted yesterday.) Yesterday's reboot wasn't forced, I'd just done an update and it mumbled about a reboot, so I decided that it would be a good idea.

    And you're right, it did look like a possible X11 thing, but it's interesting that it locked up the entire system. I must have missed something along the way, but I thought that a 4-core CPU was effectively a 4 CPU system, which should mean that even if one part of the system locks up a CPU, other parts should still be OK. But it was locked up tight. It did not allow an ssh login, and, for that matter, it did not even respond to pings. So, it was thoroughly hosed.

    -Don

  4. #4
    Join Date
    Nov 2008
    Location
    N. Wales
    Posts
    1,028

    Default Re: Actually looks like a kernel prob at first glance

    gingrich adjusted his/her AFB on Sunday 10 May 2009 05:26 to write:

    >
    > Actually. the lockup only occurred after I had rebooted. (Probably
    > silly, but there's a smidge of old time Linux user in me that tries to
    > see how long the system can run without a reboot -- I was up close to
    > 180 days when I rebooted yesterday.) Yesterday's reboot wasn't forced,
    > I'd just done an update and it mumbled about a reboot, so I decided that
    > it would be a good idea.
    >
    > And you're right, it did look like a possible X11 thing, but it's
    > interesting that it locked up the entire system. I must have missed
    > something along the way, but I thought that a 4-core CPU was effectively
    > a 4 CPU system, which should mean that even if one part of the system
    > locks up a CPU, other parts should still be OK.


    Ah but if it is a kernel crash of immense proportions then it will affect
    all CPU`s :-)

    > But it was locked up
    > tight. It did not allow an ssh login, and, for that matter, it did not
    > even respond to pings. So, it was thoroughly hosed.
    >


    Have a look at the SysReQ man pages and enable in sysconfig will get you out
    of most things unless the kernel is really hosed good and proper, at least
    it should reboot without losing data if you can sync the drives with the:

    Alt+SysReQ+s
    Alt=SysReQ+u
    Alt+SysReq+b

    combination ( I normally do a couple of the syncs before the ummount one )


    Is it running OK now?

    HTH

    --
    Mark

    Nullus in verba
    Nil illigitimi carborundum


Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •