On 03/21/2014 04:24 PM, Viresh Kumar wrote:
> On 21 March 2014 16:13, Gautham R Shenoy <e...@linux.vnet.ibm.com> wrote:
>> On Fri, Mar 21, 2014 at 02:11:32PM +0530, Viresh Kumar wrote:
>
>>>> + pr_debug("PState id %d freq %d MHz\n", id, freq);
>>>> + powernv_freqs[i].driver_data = i;
>>>
[...]
>>>> +static int powernv_cpufreq_cpu_init(struct cpufreq_policy *policy)
>>>> +{
>>>> + int base, i;
>>>> +
>>>> +#ifdef CONFIG_SMP
>>>
>>> What will break if you don't have this ifdef here? Without that as well
>>> below code should work?
>
> Missed this one?
>
Nothing will break, its just that the code size will be a tiny bit
lesser on UP configs, that's all :-) Anyway, I think removing the ifdef
improves the readability (and doesn't add any noticeable overhead on UP
kernels), so let's get rid of it.
Regards,
Srivatsa S. Bhat
>>>> + base = cpu_first_thread_sibling(policy->cpu);
>>>> +
>>>> + for (i = 0; i < threads_per_core; i++)
>>>> + cpumask_set_cpu(base + i, policy->cpus);
>>>> +#endif
>>>> + policy->cpuinfo.transition_latency = 25000;
>>>> +
>>>> + policy->cur = powernv_freqs[0].frequency;
>>>> + cpufreq_frequency_table_get_attr(powernv_freqs, policy->cpu);
>>>
_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev
