meaning of events in /var/log/messages

[john wendel]

On 05/16/2011 03:01 PM, Mike Wright wrote:
> On 05/16/2011 02:11 PM, Michael Cronenworth wrote:
>> Mike Wright wrote:
>>> I've noticed that the messages log file has*many*  of these.  Does
>>> anybody know what they mean?  Do they mean something good/bad/ugly/etc?
>>>
>>> It's always a "mce_notify_irq" followed by four "Machine check events
>>> logged".
>>>
>>> May 16 10:03:37 lan51 kernel: [73338.698074] mce_notify_irq: 28
>>> callbacks suppressed
>>> May 16 10:03:37 lan51 kernel: [73338.698082] Machine check events logged
>>> May 16 10:06:07 lan51 kernel: [73488.698064] Machine check events logged
>>> May 16 10:07:22 lan51 kernel: [73563.698071] Machine check events logged
>>> May 16 10:07:59 lan51 kernel: [73601.198071] Machine check events logged
>>
>> What CPU do you have?
>>
>> Does /var/log/mcelog contain anything interesting?
>
> Awesome, Michael.  I really appreciate your help here.
>
> CPU is AMD Phenom(tm) II X4 B60 Processor (64bit quad-core 3300mhz)
>
> The results of looking through /var/log/mcelog don't look so good...
>
> MCE 0 (through MCE 31, then repeating)
> Mon May 16 06:02:18 2011
> HARDWARE ERROR. This is *NOT* a software problem!
> Please contact your hardware vendor
> CPU 2 2 bus unit ADDR 2000
>     L2 cache ECC error
>     Cache tag array error
>          bit46 = corrected ecc error
>          bit62 = error overflow (multiple errors)
>     memory/cache error 'evict mem transaction, generic transaction, level 2'
> STATUS d40040000002017a MCGSTATUS 0
> CPUID Vendor AMD Family 16 Model 4
>
> Not sure where this points.  CPU 2 (AMD problem?) or L2 cache (mobo
> problem?)
>

Not a motherboard problem, your level2 cache is on your CPU.

Regards,

John