[PATCH 1/3] cpufreq: dt: Check OPP count before marking them shared

[PATCH 1/3] cpufreq: dt: Check OPP count before marking them shared

[Viresh Kumar]

We need to explicitly mark OPPs as shared, when they are not defined
with OPP-v2 bindings. But this isn't required to be done if we failed to
initialize OPP table.

Reorder code to verify OPP count before marking them shared.

Fixes: 2e02d8723edf ("cpufreq: dt: Add support for operating-points-v2 bindings")
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 drivers/cpufreq/cpufreq-dt.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
index c3583cdfadbd..8c38b5192baa 100644
--- a/drivers/cpufreq/cpufreq-dt.c
+++ b/drivers/cpufreq/cpufreq-dt.c
@@ -239,6 +239,17 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 	 */
 	of_cpumask_init_opp_table(policy->cpus);
 
+	/*
+	 * But we need OPP table to function so if it is not there let's
+	 * give platform code chance to provide it for us.
+	 */
+	ret = dev_pm_opp_get_opp_count(cpu_dev);
+	if (ret <= 0) {
+		pr_debug("OPP table is not ready, deferring probe\n");
+		ret = -EPROBE_DEFER;
+		goto out_free_opp;
+	}
+
 	if (need_update) {
 		struct cpufreq_dt_platform_data *pd = cpufreq_get_driver_data();
 
@@ -256,17 +267,6 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 		transition_latency = dev_pm_opp_get_max_clock_latency(cpu_dev);
 	}
 
-	/*
-	 * But we need OPP table to function so if it is not there let's
-	 * give platform code chance to provide it for us.
-	 */
-	ret = dev_pm_opp_get_opp_count(cpu_dev);
-	if (ret <= 0) {
-		pr_debug("OPP table is not ready, deferring probe\n");
-		ret = -EPROBE_DEFER;
-		goto out_free_opp;
-	}
-
 	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
 	if (!priv) {
 		ret = -ENOMEM;
-- 
2.4.0

[PATCH 2/3] cpufreq: dt: Print error on failing to mark OPPs as shared

[Viresh Kumar]

We need to explicitly mark OPPs as shared, when they are not defined
with OPP-v2 bindings. This operation can potentially fail, and in that
case we should at least print an error message.

Fixes: 2e02d8723edf ("cpufreq: dt: Add support for operating-points-v2 bindings")
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 drivers/cpufreq/cpufreq-dt.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
index 8c38b5192baa..b1131cf89757 100644
--- a/drivers/cpufreq/cpufreq-dt.c
+++ b/drivers/cpufreq/cpufreq-dt.c
@@ -260,7 +260,10 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 		 * OPP tables are initialized only for policy->cpu, do it for
 		 * others as well.
 		 */
-		set_cpus_sharing_opps(cpu_dev, policy->cpus);
+		ret = set_cpus_sharing_opps(cpu_dev, policy->cpus);
+		if (ret)
+			dev_err(cpu_dev, "%s: failed to mark OPPs as shared: %d\n",
+				__func__, ret);
 
 		of_property_read_u32(np, "clock-latency", &transition_latency);
 	} else {
-- 
2.4.0

[PATCH 3/3] cpufreq: dt: Tolerance applies on both sides of target voltage

[Viresh Kumar]

Tolerance applies on both sides of the target voltage, i.e. both min and
max sides. But while checking if a voltage is supported by the regulator
or not, we haven't taken care of tolerance on the lower side. Fix that.

Cc: Lucas Stach <l.stach@pengutronix.de>
Fixes: 045ee45c4ff2 ("cpufreq: cpufreq-dt: disable unsupported OPPs")
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 drivers/cpufreq/cpufreq-dt.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
index b1131cf89757..3b64c203bf99 100644
--- a/drivers/cpufreq/cpufreq-dt.c
+++ b/drivers/cpufreq/cpufreq-dt.c
@@ -303,7 +303,8 @@ static int cpufreq_init(struct cpufreq_policy *policy)
 			rcu_read_unlock();
 
 			tol_uV = opp_uV * priv->voltage_tolerance / 100;
-			if (regulator_is_supported_voltage(cpu_reg, opp_uV,
+			if (regulator_is_supported_voltage(cpu_reg,
+							   opp_uV - tol_uV,
 							   opp_uV + tol_uV)) {
 				if (opp_uV < min_uV)
 					min_uV = opp_uV;
-- 
2.4.0

Re: [PATCH 3/3] cpufreq: dt: Tolerance applies on both sides of target voltage

[Lucas Stach]

Am Mittwoch, den 02.09.2015, 14:36 +0530 schrieb Viresh Kumar:
> Tolerance applies on both sides of the target voltage, i.e. both min and
> max sides.

While I'm not really comfortable with the above, it is exactly how OPPv1
and the voltage tolerance property are specified.

Reviewed-by: Lucas Stach <l.stach@pengutronix.de>

> But while checking if a voltage is supported by the regulator
> or not, we haven't taken care of tolerance on the lower side. Fix that.
> 
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Fixes: 045ee45c4ff2 ("cpufreq: cpufreq-dt: disable unsupported OPPs")
> Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
> ---
>  drivers/cpufreq/cpufreq-dt.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/cpufreq/cpufreq-dt.c b/drivers/cpufreq/cpufreq-dt.c
> index b1131cf89757..3b64c203bf99 100644
> --- a/drivers/cpufreq/cpufreq-dt.c
> +++ b/drivers/cpufreq/cpufreq-dt.c
> @@ -303,7 +303,8 @@ static int cpufreq_init(struct cpufreq_policy *policy)
>  			rcu_read_unlock();
>  
>  			tol_uV = opp_uV * priv->voltage_tolerance / 100;
> -			if (regulator_is_supported_voltage(cpu_reg, opp_uV,
> +			if (regulator_is_supported_voltage(cpu_reg,
> +							   opp_uV - tol_uV,
>  							   opp_uV + tol_uV)) {
>  				if (opp_uV < min_uV)
>  					min_uV = opp_uV;

-- 
Pengutronix e.K.             | Lucas Stach                 |
Industrial Linux Solutions   | http://www.pengutronix.de/  |

Re: [PATCH 3/3] cpufreq: dt: Tolerance applies on both sides of target voltage

[Viresh Kumar]

On 02-09-15, 11:39, Lucas Stach wrote:
> Am Mittwoch, den 02.09.2015, 14:36 +0530 schrieb Viresh Kumar:
> > Tolerance applies on both sides of the target voltage, i.e. both min and
> > max sides.
> 
> While I'm not really comfortable with the above, it is exactly how OPPv1
> and the voltage tolerance property are specified.
> 
> Reviewed-by: Lucas Stach <l.stach@pengutronix.de>

Thanks.

And this is exactly how the regulator API work:

static inline int regulator_set_voltage_tol(struct regulator *regulator,
					    int new_uV, int tol_uV)
{
	if (regulator_set_voltage(regulator, new_uV, new_uV + tol_uV) == 0)
		return 0;
	else
		return regulator_set_voltage(regulator,
					     new_uV - tol_uV, new_uV + tol_uV);
}



But yeah, OPP-v2 is going to take care of this with target/min/max
values..

-- 
viresh