[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[Denis Kuzmenko]

Make s3c24xx LEDS driver use gpiolib. Fix error disabling pull during probe.

Signed-off-by: Denis Kuzmenko <linux@solonet.org.ua>
Acked-by: Stephen Warren <swarren@nvidia.com>
---

There is a comment "no point in having a pull-up if we are always driving"
but code was actually enabled pull-resistor for that case probably because of
function parameter misunderstanding (1 - disables, 0 - enables). So this was
changed to another function which takes alphabetically defined constant as
parameter.
Tested on Mini2440 board which implements most complex case where both
S3C24XX_LEDF_ACTLOW and S3C24XX_LEDF_TRISTATE are set.

Changes v1->v2
        Fix typo's and style problems.
Changes v2->v3
        Remove pull-resistor enabling code for case of S3C24XX_LEDF_TRISTATE
flag is set

See https://lkml.org/lkml/2011/11/22/42 for discussion and ack.

diff --git a/drivers/leds/leds-s3c24xx.c b/drivers/leds/leds-s3c24xx.c
index a77771d..a1dbf54 100644
--- a/drivers/leds/leds-s3c24xx.c
+++ b/drivers/leds/leds-s3c24xx.c
@@ -45,23 +45,35 @@ static void s3c24xx_led_set(struct led_classdev *led_cdev,
 	struct s3c24xx_gpio_led *led = to_gpio(led_cdev);
 	struct s3c24xx_led_platdata *pd = led->pdata;
 
-	/* there will be a short delay between setting the output and
-	 * going from output to input when using tristate. */
-
-	s3c2410_gpio_setpin(pd->gpio, (value ? 1 : 0) ^
-			    (pd->flags & S3C24XX_LEDF_ACTLOW));
-
-	if (pd->flags & S3C24XX_LEDF_TRISTATE)
-		s3c2410_gpio_cfgpin(pd->gpio,
-			value ? S3C2410_GPIO_OUTPUT : S3C2410_GPIO_INPUT);
+	/*
+	 * ensure value is 0 or 1 to use it with bitwise XOR (^)
+	 * (only 100% brightness is supported)
+	 */
+	value = value ? 1 : 0;
+
+	if (pd->flags & S3C24XX_LEDF_TRISTATE) {
+		if (value) {
+			/* invert value if S3C24XX_LEDF_ACTLOW is set */
+			value = (pd->flags & S3C24XX_LEDF_ACTLOW) ^ value;
+			gpio_direction_output(pd->gpio, value);
+		} else {
+			gpio_direction_input(pd->gpio);
+		}
+	} else {
+		/* invert value if S3C24XX_LEDF_ACTLOW is set */
+		value = (pd->flags & S3C24XX_LEDF_ACTLOW) ^ value;
+		gpio_set_value(pd->gpio, value);
+	}
 
 }
 
 static int s3c24xx_led_remove(struct platform_device *dev)
 {
+	struct s3c24xx_led_platdata *pdata = dev->dev.platform_data;
 	struct s3c24xx_gpio_led *led = pdev_to_gpio(dev);
 
 	led_classdev_unregister(&led->cdev);
+	gpio_free(pdata->gpio);
 	kfree(led);
 
 	return 0;
@@ -76,7 +88,8 @@ static int s3c24xx_led_probe(struct platform_device *dev)
 	led = kzalloc(sizeof(struct s3c24xx_gpio_led), GFP_KERNEL);
 	if (led == NULL) {
 		dev_err(&dev->dev, "No memory for device\n");
-		return -ENOMEM;
+		ret = -ENOMEM;
+		goto err_kzalloc;
 	}
 
 	platform_set_drvdata(dev, led);
@@ -88,15 +101,24 @@ static int s3c24xx_led_probe(struct platform_device *dev)
 
 	led->pdata = pdata;
 
-	/* no point in having a pull-up if we are always driving */
+	ret = gpio_request(pdata->gpio, pdata->name);
+	if (ret < 0) {
+		dev_err(&dev->dev, "gpio_request failed\n");
+		goto err_gpio_request;
+	}
 
+	/* apply GPIO settings and initially turn off the LED */
 	if (pdata->flags & S3C24XX_LEDF_TRISTATE) {
-		s3c2410_gpio_setpin(pdata->gpio, 0);
-		s3c2410_gpio_cfgpin(pdata->gpio, S3C2410_GPIO_INPUT);
+		ret = gpio_direction_input(pdata->gpio);
 	} else {
-		s3c2410_gpio_pullup(pdata->gpio, 0);
-		s3c2410_gpio_setpin(pdata->gpio, 0);
-		s3c2410_gpio_cfgpin(pdata->gpio, S3C2410_GPIO_OUTPUT);
+		/* no point in having a pull-up as we are always driving */
+		s3c_gpio_setpull(pdata->gpio, S3C_GPIO_PULL_NONE);
+		ret = gpio_direction_output(pdata->gpio,
+				!!(pdata->flags & S3C24XX_LEDF_ACTLOW));
+	}
+	if (ret < 0) {
+		dev_err(&dev->dev, "can't set gpio direction\n");
+		goto err_gpio_set_direction;
 	}
 
 	/* register our new led device */
@@ -104,11 +126,18 @@ static int s3c24xx_led_probe(struct platform_device *dev)
 	ret = led_classdev_register(&dev->dev, &led->cdev);
 	if (ret < 0) {
 		dev_err(&dev->dev, "led_classdev_register failed\n");
-		kfree(led);
-		return ret;
+		goto err_led_classdev_register;
 	}
 
 	return 0;
+
+err_led_classdev_register:
+err_gpio_set_direction:
+		gpio_free(pdata->gpio);
+err_gpio_request:
+		kfree(led);
+err_kzalloc:
+		return ret;
 }
 
 static struct platform_driver s3c24xx_led_driver = {

[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[Joe Perches]

On Tue, 2011-12-13 at 22:22 +0200, Denis Kuzmenko wrote:
> Make s3c24xx LEDS driver use gpiolib. Fix error disabling pull during probe.
[]
> diff --git a/drivers/leds/leds-s3c24xx.c b/drivers/leds/leds-s3c24xx.c
[]
> @@ -45,23 +45,35 @@ static void s3c24xx_led_set(struct led_classdev *led_cdev,
> +	/*
> +	 * ensure value is 0 or 1 to use it with bitwise XOR (^)
> +	 * (only 100% brightness is supported)
> +	 */
> +	value = value ? 1 : 0;

Maybe value = !!value;

> +		/* no point in having a pull-up as we are always driving */
> +		s3c_gpio_setpull(pdata->gpio, S3C_GPIO_PULL_NONE);
> +		ret = gpio_direction_output(pdata->gpio,
> +				!!(pdata->flags & S3C24XX_LEDF_ACTLOW));

akin to this use.

[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[Denis Kuzmenko]

Hi Joe,
Thank you for review.

On 12/13/2011 11:06 PM, Joe Perches wrote:
> On Tue, 2011-12-13 at 22:22 +0200, Denis Kuzmenko wrote:
>> Make s3c24xx LEDS driver use gpiolib. Fix error disabling pull during probe.
> []
>> diff --git a/drivers/leds/leds-s3c24xx.c b/drivers/leds/leds-s3c24xx.c
> []
>> @@ -45,23 +45,35 @@ static void s3c24xx_led_set(struct led_classdev *led_cdev,
>> +	/*
>> +	 * ensure value is 0 or 1 to use it with bitwise XOR (^)
>> +	 * (only 100% brightness is supported)
>> +	 */
>> +	value = value ? 1 : 0;
> 
> Maybe value = !!value;

IMHO my variant is better readable.

>> +		/* no point in having a pull-up as we are always driving */
>> +		s3c_gpio_setpull(pdata->gpio, S3C_GPIO_PULL_NONE);
>> +		ret = gpio_direction_output(pdata->gpio,
>> +				!!(pdata->flags & S3C24XX_LEDF_ACTLOW));
> 
> akin to this use.

There is no so much space for implementation used before so I've used
shorter notation, again, for better readability.

-- 
Best regards, Denis Kuzmenko.

[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[Joe Perches]

On Tue, 2011-12-13 at 23:21 +0200, Denis Kuzmenko wrote:
> Hi Joe,
> Thank you for review.
> 
> On 12/13/2011 11:06 PM, Joe Perches wrote:
> > On Tue, 2011-12-13 at 22:22 +0200, Denis Kuzmenko wrote:
> >> Make s3c24xx LEDS driver use gpiolib. Fix error disabling pull during probe.
> > []
> >> diff --git a/drivers/leds/leds-s3c24xx.c b/drivers/leds/leds-s3c24xx.c
> > []
> >> @@ -45,23 +45,35 @@ static void s3c24xx_led_set(struct led_classdev *led_cdev,
> >> +	/*
> >> +	 * ensure value is 0 or 1 to use it with bitwise XOR (^)
> >> +	 * (only 100% brightness is supported)
> >> +	 */
> >> +	value = value ? 1 : 0;
> > 
> > Maybe value = !!value;
> 
> IMHO my variant is better readable.
> 
> >> +		/* no point in having a pull-up as we are always driving */
> >> +		s3c_gpio_setpull(pdata->gpio, S3C_GPIO_PULL_NONE);
> >> +		ret = gpio_direction_output(pdata->gpio,
> >> +				!!(pdata->flags & S3C24XX_LEDF_ACTLOW));
> > 
> > akin to this use.
> 
> There is no so much space for implementation used before so I've used
> shorter notation, again, for better readability.

Your choice, but I think consistency is better.

The LEDF_ACTLOW use is dependent on it being #defined
to 1 when using ^.  I think that's unintelligible.

cheers, Joe

[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[Denis Kuzmenko]

On 12/13/2011 11:28 PM, Joe Perches wrote:
> On Tue, 2011-12-13 at 23:21 +0200, Denis Kuzmenko wrote:
>> Hi Joe,
>> Thank you for review.
>>
>> On 12/13/2011 11:06 PM, Joe Perches wrote:
>>> On Tue, 2011-12-13 at 22:22 +0200, Denis Kuzmenko wrote:
>>>> Make s3c24xx LEDS driver use gpiolib. Fix error disabling pull during probe.

> Your choice, but I think consistency is better.
> 
> The LEDF_ACTLOW use is dependent on it being #defined
> to 1 when using ^.  I think that's unintelligible.

Sorry, can't understand. Can you please clarify last two sentences?

-- 
Best regards, Denis Kuzmenko.

[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[Joe Perches]

On Tue, 2011-12-13 at 23:33 +0200, Denis Kuzmenko wrote:
> On 12/13/2011 11:28 PM, Joe Perches wrote:
> > The LEDF_ACTLOW use is dependent on it being #defined
> > to 1 when using ^.  I think that's unintelligible.
> Sorry, can't understand. Can you please clarify last two sentences?

	s3c2410_gpio_setpin(pd->gpio, (value ? 1 : 0) ^
			    (pd->flags & S3C24XX_LEDF_ACTLOW));

I think this should be:
	s3c2410_gpio_setpin(pd->gpio,
			    !!value ^ !!(pd->flags & S3C24XX_LEDF_ACTLOW));

[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[Denis Kuzmenko]

On 12/13/2011 11:40 PM, Joe Perches wrote:
> On Tue, 2011-12-13 at 23:33 +0200, Denis Kuzmenko wrote:
>> On 12/13/2011 11:28 PM, Joe Perches wrote:
>>> The LEDF_ACTLOW use is dependent on it being #defined
>>> to 1 when using ^.  I think that's unintelligible.
>> Sorry, can't understand. Can you please clarify last two sentences?
> 
> 	s3c2410_gpio_setpin(pd->gpio, (value ? 1 : 0) ^
> 			    (pd->flags & S3C24XX_LEDF_ACTLOW));
> 
> I think this should be:
> 	s3c2410_gpio_setpin(pd->gpio,
> 			    !!value ^ !!(pd->flags & S3C24XX_LEDF_ACTLOW));
> 

So you've found a BUG. As I understand my code will behave wrong if
S3C24XX_LEDF_ACTLOW value will be changed.
Thank you, I'll make a new version.

-- 
Best regards, Denis Kuzmenko.

[PATCH v3 resend2] s3c/s3c24xx: arm: leds: Make s3c24xx LEDS driver use gpiolib

[Joe Perches]

On Wed, 2011-12-14 at 00:05 +0200, Denis Kuzmenko wrote:
> On 12/13/2011 11:40 PM, Joe Perches wrote:
> > On Tue, 2011-12-13 at 23:33 +0200, Denis Kuzmenko wrote:
> >> On 12/13/2011 11:28 PM, Joe Perches wrote:
> >>> The LEDF_ACTLOW use is dependent on it being #defined
> >>> to 1 when using ^.  I think that's unintelligible.
> >> Sorry, can't understand. Can you please clarify last two sentences?
> > 	s3c2410_gpio_setpin(pd->gpio, (value ? 1 : 0) ^
> > 			    (pd->flags & S3C24XX_LEDF_ACTLOW));
> > I think this should be:
> > 	s3c2410_gpio_setpin(pd->gpio,
> > 			    !!value ^ !!(pd->flags & S3C24XX_LEDF_ACTLOW));
> So you've found a BUG.

It's not really a bug, it's more lack of consistency
style defect.  It's also an example of why I prefer
consistent style to equivalent but different styles.

> As I understand my code will behave wrong if
> S3C24XX_LEDF_ACTLOW value will be changed.

> Thank you, I'll make a new version.

cheers, Joe