You are not logged in.

#1 2012-05-04 09:46:26

moleculecolony
Member
Registered: 2012-05-04
Posts: 3

Kernel freezes at boot - watchdog_overflow_callback

Hello everyone.
I'm using Arch for two months now (am very happy with it's didactic and self-made attitude enhancing quality as opposed to the automatic drivings of Ubuntu and Debian's giganticness I was using before.)

In the last days there was a sporadic incident of the Kernel freezing at boot for 10 or 20 seconds. Happened maybe 2 or 3 times out of dozens of boot processes. (Yes, I had to boot so often because I had troubles installing lirc - didn't notice that the kernel doesn't load the firmware if the card is found in warm state - after rebooting instead of switching the computer off - and looked for a long time in the wrong places. But I'm also grateful to such incidents as they give you the urge to learn about things you otherwise don't ever notice.) (This is not saying you should deliver more problematic packages, mind you.)

Ok, so here's the relevant part of the dmesg:

[    0.054781] CPU0: Intel(R) Core(TM) i7 CPU         860  @ 2.80GHz stepping 05
[    0.160254] Performance Events: PEBS fmt1+, Nehalem events, Intel PMU driver.
[    0.160259] CPU erratum AAJ80 worked around
[    0.160260] CPUID marked event: 'bus cycles' unavailable
[    0.160263] ... version:                3
[    0.160264] ... bit width:              48
[    0.160265] ... generic registers:      4
[    0.160266] ... value mask:             0000ffffffffffff
[    0.160267] ... max period:             000000007fffffff
[    0.160269] ... fixed-purpose events:   3
[    0.160270] ... event mask:             000000070000000f
[    0.180359] NMI watchdog enabled, takes one hw-pmu counter.
[    0.206901] Booting Node   0, Processors  #1
[    0.206906] smpboot cpu 1: start_ip = 8a000
[   16.814734] ------------[ cut here ]------------
[   16.814738] WARNING: at kernel/watchdog.c:241 watchdog_overflow_callback+0x9a/0xc0()
[   16.814740] Hardware name:         
[   16.814741] Watchdog detected hard LOCKUP on cpu 0
[   16.814742] Modules linked in:
[   16.814745] Pid: 1, comm: swapper/0 Not tainted 3.3.4-2-ARCH #1
[   16.814746] Call Trace:
[   16.814748]  <NMI>  [<ffffffff8104f85f>] warn_slowpath_common+0x7f/0xc0
[   16.814753]  [<ffffffff8104f956>] warn_slowpath_fmt+0x46/0x50
[   16.814757]  [<ffffffff8101cf29>] ? sched_clock+0x9/0x10
[   16.814759]  [<ffffffff810cdeb0>] ? touch_nmi_watchdog+0x80/0x80
[   16.814762]  [<ffffffff810cdf4a>] watchdog_overflow_callback+0x9a/0xc0
[   16.814766]  [<ffffffff8110388d>] __perf_event_overflow+0x9d/0x230
[   16.814769]  [<ffffffff81101767>] ? perf_event_update_userpage+0xc7/0x100
[   16.814772]  [<ffffffff81026019>] ? x86_perf_event_set_period+0xd9/0x160
[   16.814775]  [<ffffffff81104304>] perf_event_overflow+0x14/0x20
[   16.814778]  [<ffffffff8102a45a>] intel_pmu_handle_irq+0x16a/0x2e0
[   16.814781]  [<ffffffff81024e2d>] perf_event_nmi_handler+0x1d/0x20
[   16.814785]  [<ffffffff81018629>] nmi_handle.isra.0+0x59/0x90
[   16.814788]  [<ffffffff81018748>] do_nmi+0xe8/0x330
[   16.814792]  [<ffffffff8145faf4>] restart_nmi+0x1a/0x1e
[   16.814796]  [<ffffffff812470af>] ? delay_tsc+0x8f/0xf0
[   16.814798]  [<ffffffff812470af>] ? delay_tsc+0x8f/0xf0
[   16.814801]  [<ffffffff812470af>] ? delay_tsc+0x8f/0xf0
[   16.814802]  <<EOE>>  [<ffffffff8144d9c2>] ? set_cpu_sibling_map+0x2cb/0x2cb
[   16.814807]  [<ffffffff81246f4f>] __delay+0xf/0x20
[   16.814810]  [<ffffffff81246f93>] __const_udelay+0x33/0x40
[   16.814812]  [<ffffffff8144d3de>] do_boot_cpu+0x3a5/0x65e
[   16.814815]  [<ffffffff8144d697>] ? do_boot_cpu+0x65e/0x65e
[   16.814817]  [<ffffffff8144dd5d>] native_cpu_up+0xbe/0x10f
[   16.814819]  [<ffffffff8144f26c>] _cpu_up+0x92/0xff
[   16.814822]  [<ffffffff8144f3af>] cpu_up+0xd6/0xea
[   16.814825]  [<ffffffff818cfe0a>] smp_init+0x41/0x8d
[   16.814828]  [<ffffffff818b4cad>] kernel_init+0x95/0x149
[   16.814830]  [<ffffffff81461224>] kernel_thread_helper+0x4/0x10
[   16.814833]  [<ffffffff818b4c18>] ? start_kernel+0x3d1/0x3d1
[   16.814835]  [<ffffffff81461220>] ? gs_change+0x13/0x13
[   16.814839] ---[ end trace 6d450e935ee1897c ]---
[   17.042399] NMI watchdog enabled, takes one hw-pmu counter.
[   17.060076]  #2
[   17.060079] smpboot cpu 2: start_ip = 8a000
[   17.091176] NMI watchdog enabled, takes one hw-pmu counter.
[   17.109994]  #3
[   17.109996] smpboot cpu 3: start_ip = 8a000
[   17.141192] NMI watchdog enabled, takes one hw-pmu counter.
[   17.159920]  #4
[   17.159923] smpboot cpu 4: start_ip = 8a000
[   17.191121] NMI watchdog enabled, takes one hw-pmu counter.
[   17.209848]  #5
[   17.209851] smpboot cpu 5: start_ip = 8a000
[   17.241050] NMI watchdog enabled, takes one hw-pmu counter.
[   17.259767]  #6
[   17.259770] smpboot cpu 6: start_ip = 8a000
[   17.290968] NMI watchdog enabled, takes one hw-pmu counter.
[   17.309692]  #7 Ok.
[   17.309695] smpboot cpu 7: start_ip = 8a000
[   17.340893] NMI watchdog enabled, takes one hw-pmu counter.
[   17.346283] Brought up 8 CPUs
[   17.346287] Total of 8 processors activated (38416.10 BogoMIPS).
[   17.352227] devtmpfs: initialized
[   17.353398] PM: Registering ACPI NVS region at df6d1000 (974848 bytes)
[   17.354205] NET: Registered protocol family 16

uname -a :

Linux CX 3.3.4-2-ARCH #1 SMP PREEMPT Wed May 2 18:28:42 CEST 2012 x86_64 Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz GenuineIntel GNU/Linux

Board is Intel DP55KG, if it's of interest, and the CPU fan is turned off without any heat problems, CPU is undervolted from 1.2 to 1.1 V, and underclocked from 2.8 to 2.4 GHz.

This is not a problem I need a solution for, but maybe it could be of interest to someone.

Cheers,
Eriq

Offline

Board footer

Powered by FluxBB