[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#255903: kernel-image-2.6.6-1-k7: Machine check exception



Package: kernel-image-2.6.6-1-k7
Version: 2.6.6-1
Severity: important

Every once in a while I get a message similar to the following on the console
and in /var/log/kern.log:

Jun 23 16:15:35 localhost kernel: CPU 0: Machine Check Exception:
0000000000000004
Jun 23 16:15:35 localhost kernel: Bank 1: f600200000000152 at
0000000000165500
Jun 23 16:15:35 localhost kernel: Kernel panic: CPU context corrupt

Using the 'parsemce' program that I found available for analysing such
problems, this is decoded into:

CPU 0
Status: (4) Machine Check in progress.
Restart IP invalid.
parsebank(1): f600200000000152 @ 165500
        External tag parity error
        Uncorrectable ECC error
        CPU state corrupt. Restart not possible
        Address in addr register valid
        Error enabled in control register
        Error not corrected.
        Error overflow
        Memory heirarchy error
        Request: Generic error
        Transaction type : Instruction
        Memory/IO : I/O

However, this doesn't really mean much to me. I assume that the 'I/O' at the
end might imply that the problem is not actually with the memory, but some
other part of the system?

Normally when this happens, it is a threaded app, such as firefox or rhythmbox,
which just hangs. The system is normally usable for a while afterwards, though
the app in question is unkillable. Once I've ignored it, and the machine
rebooted some time after, but only once.

Are MCEs always a hardware problem, since I never used to have these problems?
Is there any indication of what the problem might be, from the above? If such
checks were only added in recent kernel versions, is it really dangerous to use
the 'nomce' boot option? Could it be a bug in my hardware (ie. no bug-detection
in the kernel)? OTOH, I don't see any other reports of something like this bug
:(

Cheers,

-- 
Neil


-- System Information:
Debian Release: testing/unstable
  APT prefers unstable
  APT policy: (500, 'unstable')
Architecture: i386 (i686)
Kernel: Linux 2.6.6-1-k7
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8

Versions of packages kernel-image-2.6.6-1-k7 depends on:
ii  coreutils [fileutils]        5.0.91-2    The GNU core utilities
ii  fileutils                    5.0.91-2    The GNU file management utilities 
ii  initrd-tools                 0.1.70      tools to create initrd image for p
ii  module-init-tools            3.0-pre10-4 tools for managing Linux kernel mo

-- no debconf information




Reply to: