2019-01-29 17:08:45

by Stefan Wahren

[permalink] [raw]
Subject: [PATCH 1/2] dt-bindings: hwmon: Add optional regulator support to pwm-fan

This adds an optional regulator support (e.g. switchable supply) to the
pwm fan binding.

Signed-off-by: Stefan Wahren <[email protected]>
---
Documentation/devicetree/bindings/hwmon/pwm-fan.txt | 3 +++
1 file changed, 3 insertions(+)

diff --git a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
index c6d5332..49ca5d8 100644
--- a/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
+++ b/Documentation/devicetree/bindings/hwmon/pwm-fan.txt
@@ -6,6 +6,9 @@ Required properties:
- cooling-levels : PWM duty cycle values in a range from 0 to 255
which correspond to thermal cooling states

+Optional properties:
+- fan-supply : phandle to the regulator that provides power to the fan
+
Example:
fan0: pwm-fan {
compatible = "pwm-fan";
--
2.7.4



2019-01-29 17:10:26

by Stefan Wahren

[permalink] [raw]
Subject: [PATCH 2/2] hwmon: pwm-fan: Add optional regulator support

This adds optional regulator support to the pwm-fan driver. This is
necessary for pwm fans which are powered by a switchable supply.

Signed-off-by: Stefan Wahren <[email protected]>
---
drivers/hwmon/pwm-fan.c | 35 +++++++++++++++++++++++++++++++++++
1 file changed, 35 insertions(+)

diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
index 2c94482..9e0591e 100644
--- a/drivers/hwmon/pwm-fan.c
+++ b/drivers/hwmon/pwm-fan.c
@@ -23,6 +23,7 @@
#include <linux/of.h>
#include <linux/platform_device.h>
#include <linux/pwm.h>
+#include <linux/regulator/consumer.h>
#include <linux/sysfs.h>
#include <linux/thermal.h>

@@ -31,6 +32,7 @@
struct pwm_fan_ctx {
struct mutex lock;
struct pwm_device *pwm;
+ struct regulator *reg_en;
unsigned int pwm_value;
unsigned int pwm_fan_state;
unsigned int pwm_fan_max_state;
@@ -244,6 +246,23 @@ static int pwm_fan_probe(struct platform_device *pdev)
return ret;
}

+ ctx->reg_en = devm_regulator_get_optional(&pdev->dev, "fan");
+ if (!IS_ERR(ctx->reg_en)) {
+ ret = regulator_enable(ctx->reg_en);
+ if (ret)
+ dev_err(&pdev->dev,
+ "Failed to enable fan supply: %d\n", ret);
+ } else {
+ switch (PTR_ERR(ctx->reg_en)) {
+ case -ENODEV:
+ case -ENOSYS:
+ ctx->reg_en = NULL;
+ break;
+ default:
+ return PTR_ERR(ctx->reg_en);
+ }
+ }
+
hwmon = devm_hwmon_device_register_with_groups(&pdev->dev, "pwmfan",
ctx, pwm_fan_groups);
if (IS_ERR(hwmon)) {
@@ -287,6 +306,10 @@ static int pwm_fan_remove(struct platform_device *pdev)
thermal_cooling_device_unregister(ctx->cdev);
if (ctx->pwm_value)
pwm_disable(ctx->pwm);
+
+ if (ctx->reg_en)
+ regulator_disable(ctx->reg_en);
+
return 0;
}

@@ -299,6 +322,12 @@ static int pwm_fan_suspend(struct device *dev)

pwm_get_args(ctx->pwm, &args);

+ if (ctx->reg_en) {
+ ret = regulator_disable(ctx->reg_en);
+ if (ret)
+ dev_err(dev, "Failed to disable fan supply: %d\n", ret);
+ }
+
if (ctx->pwm_value) {
ret = pwm_config(ctx->pwm, 0, args.period);
if (ret < 0)
@@ -317,6 +346,12 @@ static int pwm_fan_resume(struct device *dev)
unsigned long duty;
int ret;

+ if (ctx->reg_en) {
+ ret = regulator_enable(ctx->reg_en);
+ if (ret)
+ dev_err(dev, "Failed to enable fan supply: %d\n", ret);
+ }
+
if (ctx->pwm_value == 0)
return 0;

--
2.7.4


2019-01-31 18:23:33

by Guenter Roeck

[permalink] [raw]
Subject: Re: [PATCH 2/2] hwmon: pwm-fan: Add optional regulator support

On Tue, Jan 29, 2019 at 06:07:11PM +0100, Stefan Wahren wrote:
> This adds optional regulator support to the pwm-fan driver. This is
> necessary for pwm fans which are powered by a switchable supply.
>
> Signed-off-by: Stefan Wahren <[email protected]>
> ---
> drivers/hwmon/pwm-fan.c | 35 +++++++++++++++++++++++++++++++++++
> 1 file changed, 35 insertions(+)
>
> diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
> index 2c94482..9e0591e 100644
> --- a/drivers/hwmon/pwm-fan.c
> +++ b/drivers/hwmon/pwm-fan.c
> @@ -23,6 +23,7 @@
> #include <linux/of.h>
> #include <linux/platform_device.h>
> #include <linux/pwm.h>
> +#include <linux/regulator/consumer.h>
> #include <linux/sysfs.h>
> #include <linux/thermal.h>
>
> @@ -31,6 +32,7 @@
> struct pwm_fan_ctx {
> struct mutex lock;
> struct pwm_device *pwm;
> + struct regulator *reg_en;
> unsigned int pwm_value;
> unsigned int pwm_fan_state;
> unsigned int pwm_fan_max_state;
> @@ -244,6 +246,23 @@ static int pwm_fan_probe(struct platform_device *pdev)
> return ret;
> }
>
> + ctx->reg_en = devm_regulator_get_optional(&pdev->dev, "fan");
> + if (!IS_ERR(ctx->reg_en)) {

It is customary to handle the error case first.

> + ret = regulator_enable(ctx->reg_en);
> + if (ret)
> + dev_err(&pdev->dev,
> + "Failed to enable fan supply: %d\n", ret);
> + } else {
> + switch (PTR_ERR(ctx->reg_en)) {
> + case -ENODEV:
> + case -ENOSYS:

Does devm_regulator_get_optional() ever return -ENOSYS ?
I glanced through other calls to devm_regulator_get_optional(),
but I don't see that check anywhere.

> + ctx->reg_en = NULL;
> + break;
> + default:
> + return PTR_ERR(ctx->reg_en);
> + }

This is inconsistent. Provided regulators which fail to enable are
accepted, albeit with an error message, while unexpected errors from
devm_regulator_get_optional() result in a bailout. Please be consistent.

> + }
> +
> hwmon = devm_hwmon_device_register_with_groups(&pdev->dev, "pwmfan",
> ctx, pwm_fan_groups);
> if (IS_ERR(hwmon)) {
> @@ -287,6 +306,10 @@ static int pwm_fan_remove(struct platform_device *pdev)
> thermal_cooling_device_unregister(ctx->cdev);
> if (ctx->pwm_value)
> pwm_disable(ctx->pwm);
> +
> + if (ctx->reg_en)
> + regulator_disable(ctx->reg_en);

As written, this is also called after regulator_enable() failed.
Same below.

> +
> return 0;
> }
>
> @@ -299,6 +322,12 @@ static int pwm_fan_suspend(struct device *dev)
>
> pwm_get_args(ctx->pwm, &args);
>
> + if (ctx->reg_en) {
> + ret = regulator_disable(ctx->reg_en);
> + if (ret)
> + dev_err(dev, "Failed to disable fan supply: %d\n", ret);
> + }
> +
> if (ctx->pwm_value) {
> ret = pwm_config(ctx->pwm, 0, args.period);
> if (ret < 0)
> @@ -317,6 +346,12 @@ static int pwm_fan_resume(struct device *dev)
> unsigned long duty;
> int ret;
>
> + if (ctx->reg_en) {
> + ret = regulator_enable(ctx->reg_en);
> + if (ret)
> + dev_err(dev, "Failed to enable fan supply: %d\n", ret);
> + }
> +
> if (ctx->pwm_value == 0)
> return 0;
>

2019-02-01 10:41:28

by Stefan Wahren

[permalink] [raw]
Subject: Re: [PATCH 2/2] hwmon: pwm-fan: Add optional regulator support

Hi Guenter,

Am 31.01.19 um 18:56 schrieb Guenter Roeck:
> On Tue, Jan 29, 2019 at 06:07:11PM +0100, Stefan Wahren wrote:
>> This adds optional regulator support to the pwm-fan driver. This is
>> necessary for pwm fans which are powered by a switchable supply.
>>
>> Signed-off-by: Stefan Wahren <[email protected]>
>> ---
>> drivers/hwmon/pwm-fan.c | 35 +++++++++++++++++++++++++++++++++++
>> 1 file changed, 35 insertions(+)
>>
>> diff --git a/drivers/hwmon/pwm-fan.c b/drivers/hwmon/pwm-fan.c
>> index 2c94482..9e0591e 100644
>> --- a/drivers/hwmon/pwm-fan.c
>> +++ b/drivers/hwmon/pwm-fan.c
>> @@ -23,6 +23,7 @@
>> #include <linux/of.h>
>> #include <linux/platform_device.h>
>> #include <linux/pwm.h>
>> +#include <linux/regulator/consumer.h>
>> #include <linux/sysfs.h>
>> #include <linux/thermal.h>
>>
>> @@ -31,6 +32,7 @@
>> struct pwm_fan_ctx {
>> struct mutex lock;
>> struct pwm_device *pwm;
>> + struct regulator *reg_en;
>> unsigned int pwm_value;
>> unsigned int pwm_fan_state;
>> unsigned int pwm_fan_max_state;
>> @@ -244,6 +246,23 @@ static int pwm_fan_probe(struct platform_device *pdev)
>> return ret;
>> }
>>
>> + ctx->reg_en = devm_regulator_get_optional(&pdev->dev, "fan");
>> + if (!IS_ERR(ctx->reg_en)) {
> It is customary to handle the error case first.
>
>> + ret = regulator_enable(ctx->reg_en);
>> + if (ret)
>> + dev_err(&pdev->dev,
>> + "Failed to enable fan supply: %d\n", ret);
>> + } else {
>> + switch (PTR_ERR(ctx->reg_en)) {
>> + case -ENODEV:
>> + case -ENOSYS:
> Does devm_regulator_get_optional() ever return -ENOSYS ?
> I glanced through other calls to devm_regulator_get_optional(),
> but I don't see that check anywhere.

sorry, i mixed this up with. In case CONFIG_REGULATOR=n we are getting
-ENODEV.

Thanks Stefan