CPU clock throttled prints..

CPU clock throttled prints..

[R, Durgadoss]

Hi All,

I am getting these prints in my Atom based device, running 2.6.35 kernel.

[  350.084005] CPU0: Temperature above threshold, cpu clock throttled (total
events = 26)
[  350.084029] CPU1: Temperature above threshold, cpu clock throttled (total
events = 26)
[  350.085293] CPU0: Temperature/speed normal
[  350.085311] CPU1: Temperature/speed normal

Seems like these are from therm_throt.c and the flow for this,
Starts from smp_thermal_interrupt(...) method.

Can somebody tell me when this method gets called ?
Also, does this actually throttle the CPU ?

Thanks,
Durga

Re: CPU clock throttled prints..

[Chen Gong]

于 2011/7/5 13:36, R, Durgadoss 写道:
> Hi All,
>
> I am getting these prints in my Atom based device, running 2.6.35 kernel.
>
> [  350.084005] CPU0: Temperature above threshold, cpu clock throttled (total
> events = 26)
> [  350.084029] CPU1: Temperature above threshold, cpu clock throttled (total
> events = 26)
> [  350.085293] CPU0: Temperature/speed normal
> [  350.085311] CPU1: Temperature/speed normal
>
> Seems like these are from therm_throt.c and the flow for this,
> Starts from smp_thermal_interrupt(...) method.
>
> Can somebody tell me when this method gets called ?
> Also, does this actually throttle the CPU ?
>

try to apply this patch 29e9bf1841e4f (x86, mce, therm_throt: Don't report
power limit and package level thermal throttle events in mcelog)

Re: CPU clock throttled prints..

[Justin P. Mattock]

On 01/11/2012 06:11 PM, Chen Gong wrote:
> 于 2011/7/5 13:36, R, Durgadoss 写道:
>> Hi All,
>>
>> I am getting these prints in my Atom based device, running 2.6.35 kernel.
>>
>> [ 350.084005] CPU0: Temperature above threshold, cpu clock throttled
>> (total
>> events = 26)
>> [ 350.084029] CPU1: Temperature above threshold, cpu clock throttled
>> (total
>> events = 26)
>> [ 350.085293] CPU0: Temperature/speed normal
>> [ 350.085311] CPU1: Temperature/speed normal
>>
>> Seems like these are from therm_throt.c and the flow for this,
>> Starts from smp_thermal_interrupt(...) method.
>>
>> Can somebody tell me when this method gets called ?
>> Also, does this actually throttle the CPU ?
>>
>
> try to apply this patch 29e9bf1841e4f (x86, mce, therm_throt: Don't report
> power limit and package level thermal throttle events in mcelog)
> --
> To unsubscribe from this list: send the line "unsubscribe linux-x86_64" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>

probably not the best solution, but I just ended up disabling the 
function in the kernel: (either its below or its under a different name)
X86_MCE_INTEL = n

this way when I watch hulu, dmesg is not spammed with something about 
"above threshold" by just watching a movie.

Justin P. Mattock