[PATCH] regulator: gpio: Don't require a regulator-type property

[PATCH] regulator: gpio: Don't require a regulator-type property

[Mark Brown]

From: Mark Brown <broonie@linaro.org>

Since essentially all GPIO controlled regulators are voltage regulators
make the regulator-type property optional, defaulting to voltage.

Signed-off-by: Mark Brown <broonie@linaro.org>
---

Compile tested only.

 drivers/regulator/gpio-regulator.c | 14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

diff --git a/drivers/regulator/gpio-regulator.c b/drivers/regulator/gpio-regulator.c
index 234960dc9607..13ec94d984ca 100644
--- a/drivers/regulator/gpio-regulator.c
+++ b/drivers/regulator/gpio-regulator.c
@@ -203,17 +203,15 @@ of_get_gpio_regulator_config(struct device *dev, struct device_node *np)
 	}
 	config->nr_states = i;
 
+	config->type = REGULATOR_VOLTAGE;
 	ret = of_property_read_string(np, "regulator-type", &regtype);
-	if (ret < 0) {
-		dev_err(dev, "Missing 'regulator-type' property\n");
-		return ERR_PTR(-EINVAL);
+	if (ret >= 0) {
+		if (!strncmp("voltage", regtype, 7))
+			config->type = REGULATOR_VOLTAGE;
+		else if (!strncmp("current", regtype, 7))
+			config->type = REGULATOR_CURRENT;
 	}
 
-	if (!strncmp("voltage", regtype, 7))
-		config->type = REGULATOR_VOLTAGE;
-	else if (!strncmp("current", regtype, 7))
-		config->type = REGULATOR_CURRENT;
-
 	return config;
 }
 
-- 
1.8.5.1

Re: [PATCH] regulator: gpio: Don't require a regulator-type property

[Laurent Pinchart]

Hi Mark,

Thanks for the patch.

On Thursday 05 December 2013 00:31:55 Mark Brown wrote:
> From: Mark Brown <broonie@linaro.org>
> 
> Since essentially all GPIO controlled regulators are voltage regulators
> make the regulator-type property optional, defaulting to voltage.
> 
> Signed-off-by: Mark Brown <broonie@linaro.org>
> ---
> 
> Compile tested only.
> 
>  drivers/regulator/gpio-regulator.c | 14 ++++++--------
>  1 file changed, 6 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/regulator/gpio-regulator.c
> b/drivers/regulator/gpio-regulator.c index 234960dc9607..13ec94d984ca
> 100644
> --- a/drivers/regulator/gpio-regulator.c
> +++ b/drivers/regulator/gpio-regulator.c
> @@ -203,17 +203,15 @@ of_get_gpio_regulator_config(struct device *dev,
> struct device_node *np) }
>  	config->nr_states = i;
> 
> +	config->type = REGULATOR_VOLTAGE;
>  	ret = of_property_read_string(np, "regulator-type", &regtype);
> -	if (ret < 0) {
> -		dev_err(dev, "Missing 'regulator-type' property\n");
> -		return ERR_PTR(-EINVAL);
> +	if (ret >= 0) {
> +		if (!strncmp("voltage", regtype, 7))
> +			config->type = REGULATOR_VOLTAGE;
> +		else if (!strncmp("current", regtype, 7))
> +			config->type = REGULATOR_CURRENT;

Should we warn or error in case of an unsupported value ?

>  	}
> 
> -	if (!strncmp("voltage", regtype, 7))
> -		config->type = REGULATOR_VOLTAGE;
> -	else if (!strncmp("current", regtype, 7))
> -		config->type = REGULATOR_CURRENT;
> -
>  	return config;
>  }
-- 
Regards,

Laurent Pinchart

Re: [PATCH] regulator: gpio: Don't require a regulator-type property

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 611 bytes --]

On Thu, Dec 05, 2013 at 01:44:42AM +0100, Laurent Pinchart wrote:
> On Thursday 05 December 2013 00:31:55 Mark Brown wrote:

> > +	if (ret >= 0) {
> > +		if (!strncmp("voltage", regtype, 7))
> > +			config->type = REGULATOR_VOLTAGE;
> > +		else if (!strncmp("current", regtype, 7))
> > +			config->type = REGULATOR_CURRENT;

> Should we warn or error in case of an unsupported value ?

That's sensible, yes - this was just a straight change to make the
property optional.  Given that we've always accepted anything a warning
seems more appropriate otherwise we might break existing DTs with typos
or something.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]