Re: [PATCH 01/10] sched: select_task_rq_fair clean up

[Alex Shi <alex.shi@intel.com>]

On 12/07/2012 09:02 AM, Frederic Weisbecker wrote:
> 2012/12/7 Alex Shi <alex.shi@intel.com>:
>> On 12/07/2012 01:50 AM, Frederic Weisbecker wrote:
>>> 2012/12/3 Alex Shi <alex.shi@intel.com>:
>>>> It is impossible to miss a task allowed cpu in a eligible group.
>>>>
>>>> And since find_idlest_group only return a different group which
>>>> excludes old cpu, it's also imporissible to find a new cpu same as old
>>>> cpu.
>>>
>>> Is it possible for weighted_cpuload() to return ULONG_MAX? If so,
>>> find_idlest_cpu() can return -1.
>>>
>>
>> No, non of sched entity can has a ULONG_MAX weight.
> 
> Ok. find_idlest_cpu() can still return -1 but select_task_rq_fair() is
> the only caller. Presumably safe but code evolves. May be add some
> comment to explain why what you're doing is safe. May be a
> WARN_ON_ONCE() could be good to add?
> 

why you think it is possible to be -1?
And there is a WARN_ON_ONCE for cpu = -1 in find_idlest_group when
checking local_group.

Thanks!