Re: [PATCH v5 7/7] powerpc/powernv: nest pmu cpumask and cpu hotplug support

[Madhavan Srinivasan <maddy@linux.vnet.ibm.com>]

On Thursday 23 July 2015 12:19 PM, Daniel Axtens wrote:
> On Thu, 2015-07-23 at 12:18 +0530, Madhavan Srinivasan wrote:
>> On Wednesday 22 July 2015 10:33 AM, Daniel Axtens wrote:
>>>> +static void nest_change_cpu_context(int old_cpu, int new_cpu)
>>>> +{
>>>> +	int i;
>>>> +
>>>> +	for (i = 0; per_nest_pmu_arr[i] != NULL; i++)
>>>> +		perf_pmu_migrate_context(&per_nest_pmu_arr[i]->pmu,
>>>> +						old_cpu, new_cpu);
>>> From patch 4, I see per_nest_pmu_arr is defined as:
>>>  +static struct nest_pmu *per_nest_pmu_arr[P8_NEST_MAX_PMUS];
>>>
>>> Therefore, does this loop need to have a check that 
>>> i < P8_NEST_MAX_PMUS?
>> No, that is max possible pmu, but we may have only couple for nest pmus
>> registered.
>>
> What if we have P8_NEST_MAX_PMUS registered? Then we'll check beyond the
> end of the array...

OK, i will add check for P8_NEST_MAX_PMUS also.

>> Thanks for the review comments
>> Maddy