Re: false nr_running check in load balance?

[Lei Wen]

Paul,

On Tue, Aug 13, 2013 at 5:25 PM, Paul Turner <pjt@xxxxxxxxxx> wrote:
> On Tue, Aug 13, 2013 at 1:18 AM, Lei Wen <adrian.wenl@xxxxxxxxx> wrote:
>> Hi Paul,
>>
>> On Tue, Aug 13, 2013 at 4:08 PM, Paul Turner <pjt@xxxxxxxxxx> wrote:
>>> On Tue, Aug 13, 2013 at 12:38 AM, Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
>>>> On Tue, Aug 13, 2013 at 12:45:12PM +0800, Lei Wen wrote:
>>>>> > Not quite right; I think you need busiest->cfs.h_nr_running.
>>>>> > cfs.nr_running is the number of entries running in this 'group'. If
>>>>> > you've got nested groups like:
>>>>> >
>>>>> >  'root'
>>>>> >    \
>>>>> >    'A'
>>>>> >    / \
>>>>> >   t1 t2
>>>>> >
>>>>> > root.nr_running := 1 'A', even though you've got multiple running tasks.
>>>>> >
>>>>>
>>>>> You're absolutely right for this. :)
>>>>> I miss it for not considering the group case...
>>>>>
>>>>> Then do you think it is necessary to do below change in load_balance() code?
>>>>>  -       if (busiest->nr_running > 1) {
>>>>>  +       if (busiest->cfs.h_nr_running > 1) {
>>>>>
>>>>
>>>> Yes I think that would be fine.
>>>
>>> If we pivot to use h_nr_running we should probably also update
>>> call-sites such as cpu_load_avg_per_task() for consistency.
>>
>> I didn't find cpu_load_avg_per_task in the latest linux git...
>> Is it a new patch pending while not being submitted?
>
> Transposition typo: cpu_avg_load_per_task()
> More generally: Most things that examine ->nr_running in the fair
> load-balance path.
>

I see...

I have submitted several patches, which covers cpu_avg_load_per_task.
Please help to check them.

Thanks,
Lei
--
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/