RE: [PATCH 1/2] cpufreq: ppc-corenet: remove duplicate update of cpu_data

[Yuantian Tang <Yuantian.Tang@freescale.com>]

> -----Original Message-----
> From: Viresh Kumar [mailto:viresh.kumar@linaro.org]
> Sent: Wednesday, September 10, 2014 12:33 PM
> To: Rafael Wysocki; Tang Yuantian-B29983
> Cc: Lists linaro-kernel; linux-pm@vger.kernel.org; Zhang Hongbo-B45939; Li
> Yang-Leo-R58472; Viresh Kumar
> Subject: Re: [PATCH 1/2] cpufreq: ppc-corenet: remove duplicate update of
> cpu_data
> 
> On 2 September 2014 09:11, Viresh Kumar <viresh.kumar@linaro.org> wrote:
> > 'cpu_data' is updated for policy->cpu first and then for all CPUs in
> > policy->cpus. policy->cpus is guaranteed to contain policy->cpu as
> > policy->well and so
> > the first write to 'cpu_data' for policy->cpu is redundant. Remove it.
> >
> > Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> > ---
> > Hi Yuantian,
> >
> > I was looking into this driver due to issues reported by Hongtao
> > (cc'd) and found that we can live without some code. These aren't
> > fixing any bugs and are just cleanups.
> >
> > I didn't had a compiler for this and so this isn't even compiled. It
> > would be great if you can please review/test these patches.
> >
> >  drivers/cpufreq/ppc-corenet-cpufreq.c | 1 -
> >  1 file changed, 1 deletion(-)
> >
> > diff --git a/drivers/cpufreq/ppc-corenet-cpufreq.c
> > b/drivers/cpufreq/ppc-corenet-cpufreq.c
> > index 3607070..bee5df7 100644
> > --- a/drivers/cpufreq/ppc-corenet-cpufreq.c
> > +++ b/drivers/cpufreq/ppc-corenet-cpufreq.c
> > @@ -199,7 +199,6 @@ static int corenet_cpufreq_cpu_init(struct
> cpufreq_policy *policy)
> >         }
> >
> >         data->table = table;
> > -       per_cpu(cpu_data, cpu) = data;
> >
> >         /* update ->cpus if we have cluster, no harm if not */
> >         cpumask_copy(policy->cpus, per_cpu(cpu_mask, cpu));
> 
> Hi Yuantian,
> 
> Can we get your Ack for this patch atleast so that Rafael can apply it ?
Yeah, your patch should be ok, but is it the best solution?
I was thinking about a final solution just like ARM did.

Thanks,
Yuantian