[PATCH 2/2] leds: lp50xx: remove unused regulator

The regulator for vled-supply is unused in the driver.
It is just assigned from DT and disabled in lp50xx_remove.
So the code can be removed from the driver.

Part 1 updates the documentation
Part 2 removes the code

Signed-off-by: Sven Schuchmann <redacted>

---
 drivers/leds/leds-lp50xx.c | 14 --------------
 1 file changed, 14 deletions(-)

diff --git a/drivers/leds/leds-lp50xx.c b/drivers/leds/leds-lp50xx.c
index f13117eed976..b0871495bae3 100644
--- a/drivers/leds/leds-lp50xx.c
+++ b/drivers/leds/leds-lp50xx.c

@@ -11,7 +11,6 @@
 #include <linux/of.h>
 #include <linux/of_gpio.h>
 #include <linux/regmap.h>
-#include <linux/regulator/consumer.h>
 #include <linux/slab.h>
 #include <uapi/linux/uleds.h>
 

@@ -275,7 +274,6 @@ struct lp50xx_led {
 /**
  * struct lp50xx -
  * @enable_gpio: hardware enable gpio
- * @regulator: LED supply regulator pointer
  * @client: pointer to the I2C client
  * @regmap: device register map
  * @dev: pointer to the devices device struct

@@ -286,7 +284,6 @@ struct lp50xx_led {
  */
 struct lp50xx {
 	struct gpio_desc *enable_gpio;
-	struct regulator *regulator;
 	struct i2c_client *client;
 	struct regmap *regmap;
 	struct device *dev;

@@ -462,10 +459,6 @@ static int lp50xx_probe_dt(struct lp50xx *priv)
 		return ret;
 	}
 
-	priv->regulator = devm_regulator_get(priv->dev, "vled");
-	if (IS_ERR(priv->regulator))
-		priv->regulator = NULL;
-
 	device_for_each_child_node(priv->dev, child) {
 		led = &priv->leds[i];
 		ret = fwnode_property_count_u32(child, "reg");

@@ -583,13 +576,6 @@ static int lp50xx_remove(struct i2c_client *client)
 		return ret;
 	}
 
-	if (led->regulator) {
-		ret = regulator_disable(led->regulator);
-		if (ret)
-			dev_err(&led->client->dev,
-				"Failed to disable regulator\n");
-	}
-
 	mutex_destroy(&led->lock);
 
 	return 0;

-- 
2.17.1

On Wed 2021-02-03 08:34:08, Sven Schuchmann wrote:

The regulator for vled-supply is unused in the driver.
It is just assigned from DT and disabled in lp50xx_remove.
So the code can be removed from the driver.

Dan, what is going on here? Do we need to also enable the regulator,
or is the removal correct thing to do?

Best regards,
							Pavel

-- 
http://www.livejournal.com/~pavelmachek

Pavel

On 2/3/21 3:02 AM, Pavel Machek wrote:

On Wed 2021-02-03 08:34:08, Sven Schuchmann wrote:

quoted

The regulator for vled-supply is unused in the driver.
It is just assigned from DT and disabled in lp50xx_remove.
So the code can be removed from the driver.

Dan, what is going on here? Do we need to also enable the regulator,
or is the removal correct thing to do?

I think it would be better to do an enable as opposed to removing the code.

This would be needed especially in applications that have to meet strict 
power management requirements.

Users may want to disable or enable the regulator during suspend/resume. 
Otherwise it would be considered always-on and the regulator does not 
need to be populated.

Dan

On Wed 2021-02-03 07:49:35, Dan Murphy wrote:

Pavel

On 2/3/21 3:02 AM, Pavel Machek wrote:

quoted

On Wed 2021-02-03 08:34:08, Sven Schuchmann wrote:

quoted

The regulator for vled-supply is unused in the driver.
It is just assigned from DT and disabled in lp50xx_remove.
So the code can be removed from the driver.

Dan, what is going on here? Do we need to also enable the regulator,
or is the removal correct thing to do?

I think it would be better to do an enable as opposed to removing the code.

This would be needed especially in applications that have to meet strict
power management requirements.

Users may want to disable or enable the regulator during suspend/resume.
Otherwise it would be considered always-on and the regulator does not need
to be populated.

Do you have set up where this is needed and you can test this? Will
you submit the fixes?

Best regards,

									Pavel
-- 
http://www.livejournal.com/~pavelmachek

Pavel

On 2/3/21 8:23 AM, Pavel Machek wrote:

On Wed 2021-02-03 07:49:35, Dan Murphy wrote:

quoted

Pavel

On 2/3/21 3:02 AM, Pavel Machek wrote:

quoted

On Wed 2021-02-03 08:34:08, Sven Schuchmann wrote:

quoted

The regulator for vled-supply is unused in the driver.
It is just assigned from DT and disabled in lp50xx_remove.
So the code can be removed from the driver.

Dan, what is going on here? Do we need to also enable the regulator,
or is the removal correct thing to do?

I think it would be better to do an enable as opposed to removing the code.

This would be needed especially in applications that have to meet strict
power management requirements.

Users may want to disable or enable the regulator during suspend/resume.
Otherwise it would be considered always-on and the regulator does not need
to be populated.

Do you have set up where this is needed and you can test this? Will
you submit the fixes?

No I use an always on regulator in my setup. I have no managed supplies 
exposed.

Dan

Best regards,

									Pavel

Hello Dan, hello Pavel,

quoted

Do you have set up where this is needed and you can test this? Will
you submit the fixes?

No I use an always on regulator in my setup. I have no managed supplies
exposed.

I am also sorry I do not have a setup ready for testing this.
I think we should ignore this patch then?`

For now I am working on the enable pin in the driver.
The enablement seems it does not work. The order of I2C-Reset
and Enable is wrong and the pin is used before initialized.
But I do not have any patches ready.
Is the enable pin working on your side?

Best Regards,

   Sven

On Wed 2021-02-03 10:02:00, Dan Murphy wrote:

Pavel

On 2/3/21 8:23 AM, Pavel Machek wrote:

quoted

On Wed 2021-02-03 07:49:35, Dan Murphy wrote:

quoted

Pavel

On 2/3/21 3:02 AM, Pavel Machek wrote:

quoted

On Wed 2021-02-03 08:34:08, Sven Schuchmann wrote:

quoted

The regulator for vled-supply is unused in the driver.
It is just assigned from DT and disabled in lp50xx_remove.
So the code can be removed from the driver.

Dan, what is going on here? Do we need to also enable the regulator,
or is the removal correct thing to do?

I think it would be better to do an enable as opposed to removing the code.

This would be needed especially in applications that have to meet strict
power management requirements.

Users may want to disable or enable the regulator during suspend/resume.
Otherwise it would be considered always-on and the regulator does not need
to be populated.

Do you have set up where this is needed and you can test this? Will
you submit the fixes?

No I use an always on regulator in my setup. I have no managed supplies
exposed.

Hmm, that is not ideal :-(. Can you try to provide patch, anyway?

Best regards,
								Pavel
-- 
http://www.livejournal.com/~pavelmachek