Re: [patch v8 6/9] sched: compute runnable load avg in cpu_load andcpu_avg_load_per_task

[Gu Zheng]

On 06/10/2013 10:01 AM, Alex Shi wrote:

> On 06/10/2013 09:49 AM, Gu Zheng wrote:
>> On 06/07/2013 03:20 PM, Alex Shi wrote:
>>
>>>> They are the base values in load balance, update them with rq runnable
>>>> load average, then the load balance will consider runnable load avg
>>>> naturally.
>>>>
>>>> We also try to include the blocked_load_avg as cpu load in balancing,
>>>> but that cause kbuild performance drop 6% on every Intel machine, and
>>>> aim7/oltp drop on some of 4 CPU sockets machines.
>> Hi Alex,
>>    Could you explain me why including the blocked_load_avg causes performance drop ?
> 
> 
> Thanks for review!
> 
> the 9th patch has few explanation. like, after the only task got into
> sleep in a CPU, there is only blocked_load_avg left, it looks quite big
> in short time. that, block it get tasks before sleep, drive task to
> other cpu in periodic balance. So, it cause clear load imbalance.
> 

Got it. Thanks very much for your explanation.:)

Best regards,
Gu
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/