Re: [PATCH 0/4] CPUFreq: Implement per policy instances of governors

On Wed, Feb 6, 2013 at 3:28 PM, Viresh Kumar viresh.kumar@linaro.org wrote:

On 5 February 2013 21:51, Viresh Kumar viresh.kumar@linaro.org wrote:

...

commit 15b5548c9ccfb8088270f7574710d9d67edfe33b Author: Viresh Kumar viresh.kumar@linaro.org Date: Tue Feb 5 21:29:05 2013 +0530

cpufreq: Make governors directory sysfs location based on

have_multiple_policies

Until now directory for governors tunables was getting created in
cpu/cpufreq/<gov-name>. With the introduction of following patch:
"cpufreq: governor: Implement per policy instances of governors"

this directory would be created in

cpu/cpu<num>/cpufreq/<gov-name>. This might break userspace of existing platforms. Lets do this change only for platforms which need support for multiple policies and thus above mentioned patch.

From now on, such platforms would be require to do following from

their init() routines:

    policy->have_multiple_policies = true;

Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>

---

drivers/cpufreq/cpufreq_governor.c | 2 +- include/linux/cpufreq.h | 14 ++++++++++++++ 2 files changed, 15 insertions(+), 1 deletion(-)

Hi Rafael,

Because this patch was quite big (317 insertions(+), 238 deletions(-)), i was planning a detailed self review to capture any mistakes and luckily i found one for above patch :)

diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c index 41ee86f..fe037c0 100644 --- a/drivers/cpufreq/cpufreq_governor.c +++ b/drivers/cpufreq/cpufreq_governor.c @@ -342,7 +342,8 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy, mutex_lock(&dbs_data->mutex); mutex_destroy(&cpu_cdbs->timer_mutex);

•           sysfs_remove_group(&policy->kobj, dbs_data->cdata->attr_group);
  

•           sysfs_remove_group(get_governor_parent_kobj(policy),
  

•                           dbs_data->cdata->attr_group);
            if (dbs_data->cdata->governor == GOV_CONSERVATIVE)
                    cpufreq_unregister_notifier(cs_ops->notifier_block,
                                    CPUFREQ_TRANSITION_NOTIFIER);
  
  

I have pushed the complete patchset here:

http://git.linaro.org/gitweb?p=people/vireshk/linux.git%3Ba=shortlog%3Bh=ref...

Viresh, perhaps you should ask Stephen Rothwell to pull in your tree to get some more testing before Rafael pulls it in for 3.10?