[1/2] cpufreq-dt: check CPU clock/power domain during initializing

Message ID 1420797291-15972-2-git-send-email-pi-cheng.chen@linaro.org
State New
Headers show

Commit Message

pi-cheng.chen Jan. 9, 2015, 9:54 a.m.
Currently the DT based cpufreq driver is missing some way to check which
CPUs share clocks. In the 1st patch, CPU clock/power domain information is
added to the platform_data of cpufreq-dt so that cpufreq-dt driver could
check which CPUs share clock/power.

Signed-off-by: pi-cheng.chen <pi-cheng.chen@linaro.org>
---
 drivers/cpufreq/cpufreq-dt.c | 15 +++++++++++++++
 include/linux/cpufreq-dt.h   |  6 ++++++
 2 files changed, 21 insertions(+)

Comments

Viresh Kumar Jan. 19, 2015, 8 a.m. | #1
On 9 January 2015 at 15:24, pi-cheng.chen <pi-cheng.chen@linaro.org> wrote:
> Currently the DT based cpufreq driver is missing some way to check which
> CPUs share clocks. In the 1st patch, CPU clock/power domain information is
> added to the platform_data of cpufreq-dt so that cpufreq-dt driver could
> check which CPUs share clock/power.
>
> Signed-off-by: pi-cheng.chen <pi-cheng.chen@linaro.org>
> ---
>  drivers/cpufreq/cpufreq-dt.c | 15 +++++++++++++++
>  include/linux/cpufreq-dt.h   |  6 ++++++
>  2 files changed, 21 insertions(+)
>
> diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
> index fde97d6..ff8c266 100644
> --- a/drivers/cpufreq/cpufreq-dt.c
> +++ b/drivers/cpufreq/cpufreq-dt.c
> @@ -296,6 +296,21 @@ static int cpufreq_init(struct cpufreq_policy *policy)
>         pd = cpufreq_get_driver_data();
>         if (!pd || !pd->independent_clocks)
>                 cpumask_setall(policy->cpus);
> +       else if (pd && !list_empty(&pd->domain_list)) {
> +               struct list_head *domain_node;
> +
> +               list_for_each(domain_node, &pd->domain_list) {
> +                       struct cpufreq_cpu_domain *domain;

Define this with domain_node.

> +
> +                       domain = container_of(domain_node,
> +                                             struct cpufreq_cpu_domain, node);
> +                       if (!cpumask_test_cpu(policy->cpu, &domain->cpus))
> +                               continue;
> +
> +                       cpumask_copy(policy->cpus, &domain->cpus);
> +                       break;
> +               }
> +       }
>
>         of_node_put(np);
>
> diff --git a/include/linux/cpufreq-dt.h b/include/linux/cpufreq-dt.h
> index 0414009..92bffd3 100644
> --- a/include/linux/cpufreq-dt.h
> +++ b/include/linux/cpufreq-dt.h
> @@ -10,6 +10,11 @@
>  #ifndef __CPUFREQ_DT_H__
>  #define __CPUFREQ_DT_H__
>
> +struct cpufreq_cpu_domain {
> +       struct list_head node;
> +       cpumask_t cpus;
> +};
> +
>  struct cpufreq_dt_platform_data {
>         /*
>          * True when each CPU has its own clock to control its
> @@ -17,6 +22,7 @@ struct cpufreq_dt_platform_data {
>          * clock.
>          */
>         bool independent_clocks;
> +       struct list_head domain_list;
>  };

Though we need to keep only one of these two, but I don't think
any of these will stay for long time. So, its okay..

Looks fine. Thanks.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/
pi-cheng.chen Jan. 20, 2015, 7:33 a.m. | #2
Hi Viresh,

Thanks for reviewing.

On 19 January 2015 at 16:00, Viresh Kumar <viresh.kumar@linaro.org> wrote:
> On 9 January 2015 at 15:24, pi-cheng.chen <pi-cheng.chen@linaro.org> wrote:
>> Currently the DT based cpufreq driver is missing some way to check which
>> CPUs share clocks. In the 1st patch, CPU clock/power domain information is
>> added to the platform_data of cpufreq-dt so that cpufreq-dt driver could
>> check which CPUs share clock/power.
>>
>> Signed-off-by: pi-cheng.chen <pi-cheng.chen@linaro.org>
>> ---
>>  drivers/cpufreq/cpufreq-dt.c | 15 +++++++++++++++
>>  include/linux/cpufreq-dt.h   |  6 ++++++
>>  2 files changed, 21 insertions(+)
>>
>> diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
>> index fde97d6..ff8c266 100644
>> --- a/drivers/cpufreq/cpufreq-dt.c
>> +++ b/drivers/cpufreq/cpufreq-dt.c
>> @@ -296,6 +296,21 @@ static int cpufreq_init(struct cpufreq_policy *policy)
>>         pd = cpufreq_get_driver_data();
>>         if (!pd || !pd->independent_clocks)
>>                 cpumask_setall(policy->cpus);
>> +       else if (pd && !list_empty(&pd->domain_list)) {
>> +               struct list_head *domain_node;
>> +
>> +               list_for_each(domain_node, &pd->domain_list) {
>> +                       struct cpufreq_cpu_domain *domain;
>
> Define this with domain_node.
>

Will do it.

>> +
>> +                       domain = container_of(domain_node,
>> +                                             struct cpufreq_cpu_domain, node);
>> +                       if (!cpumask_test_cpu(policy->cpu, &domain->cpus))
>> +                               continue;
>> +
>> +                       cpumask_copy(policy->cpus, &domain->cpus);
>> +                       break;
>> +               }
>> +       }
>>
>>         of_node_put(np);
>>
>> diff --git a/include/linux/cpufreq-dt.h b/include/linux/cpufreq-dt.h
>> index 0414009..92bffd3 100644
>> --- a/include/linux/cpufreq-dt.h
>> +++ b/include/linux/cpufreq-dt.h
>> @@ -10,6 +10,11 @@
>>  #ifndef __CPUFREQ_DT_H__
>>  #define __CPUFREQ_DT_H__
>>
>> +struct cpufreq_cpu_domain {
>> +       struct list_head node;
>> +       cpumask_t cpus;
>> +};
>> +
>>  struct cpufreq_dt_platform_data {
>>         /*
>>          * True when each CPU has its own clock to control its
>> @@ -17,6 +22,7 @@ struct cpufreq_dt_platform_data {
>>          * clock.
>>          */
>>         bool independent_clocks;
>> +       struct list_head domain_list;
>>  };
>
> Though we need to keep only one of these two, but I don't think
> any of these will stay for long time. So, its okay..
>
> Looks fine. Thanks.

Thanks.
I will also try to add intermediate frequency support in next version.
BTW, do you think it's a good idea to add a new device tree binding like
intermediate_clock = <&clksys MAINPLL>
and so that cpufreq-dt could get the intermediate clock?

Best Regards,
Pi-Cheng
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Viresh Kumar Jan. 20, 2015, 7:39 a.m. | #3
On 20 January 2015 at 13:03, Pi-Cheng Chen <pi-cheng.chen@linaro.org> wrote:
> I will also try to add intermediate frequency support in next version.

Sure

> BTW, do you think it's a good idea to add a new device tree binding like
> intermediate_clock = <&clksys MAINPLL>
> and so that cpufreq-dt could get the intermediate clock?

Yeah, we need that. I am trying to implement new opp bindings and will
take care of that. Until that time, use platform data to do that.

--
viresh
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/
pi-cheng.chen Jan. 20, 2015, 8:47 a.m. | #4
On 20 January 2015 at 15:39, Viresh Kumar <viresh.kumar@linaro.org> wrote:
> On 20 January 2015 at 13:03, Pi-Cheng Chen <pi-cheng.chen@linaro.org> wrote:
>> I will also try to add intermediate frequency support in next version.
>
> Sure
>
>> BTW, do you think it's a good idea to add a new device tree binding like
>> intermediate_clock = <&clksys MAINPLL>
>> and so that cpufreq-dt could get the intermediate clock?
>
> Yeah, we need that. I am trying to implement new opp bindings and will
> take care of that. Until that time, use platform data to do that.
>

Got it.
Thanks.

Best Regards,
Pi-Cheng

> --
> viresh
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
index fde97d6..ff8c266 100644
--- a/drivers/cpufreq/cpufreq-dt.c
+++ b/drivers/cpufreq/cpufreq-dt.c
@@ -296,6 +296,21 @@  static int cpufreq_init(struct cpufreq_policy *policy)
 	pd = cpufreq_get_driver_data();
 	if (!pd || !pd->independent_clocks)
 		cpumask_setall(policy->cpus);
+	else if (pd && !list_empty(&pd->domain_list)) {
+		struct list_head *domain_node;
+
+		list_for_each(domain_node, &pd->domain_list) {
+			struct cpufreq_cpu_domain *domain;
+
+			domain = container_of(domain_node,
+					      struct cpufreq_cpu_domain, node);
+			if (!cpumask_test_cpu(policy->cpu, &domain->cpus))
+				continue;
+
+			cpumask_copy(policy->cpus, &domain->cpus);
+			break;
+		}
+	}
 
 	of_node_put(np);
 
diff --git a/include/linux/cpufreq-dt.h b/include/linux/cpufreq-dt.h
index 0414009..92bffd3 100644
--- a/include/linux/cpufreq-dt.h
+++ b/include/linux/cpufreq-dt.h
@@ -10,6 +10,11 @@ 
 #ifndef __CPUFREQ_DT_H__
 #define __CPUFREQ_DT_H__
 
+struct cpufreq_cpu_domain {
+	struct list_head node;
+	cpumask_t cpus;
+};
+
 struct cpufreq_dt_platform_data {
 	/*
 	 * True when each CPU has its own clock to control its
@@ -17,6 +22,7 @@  struct cpufreq_dt_platform_data {
 	 * clock.
 	 */
 	bool independent_clocks;
+	struct list_head domain_list;
 };
 
 #endif /* __CPUFREQ_DT_H__ */