Change gpio-reulator-probe() to use managed resources.

Change gpio-reulator-probe() to use managed resources.

[Rob Jones]

Add new managed resource functions as needed to achieve this.

Amend the following functions:

devm_kstrdup()
gpio_regulator_probe()

Add the following functions:

devm_kmemdup()
devm_gpio_request_array()
devm_gpio_free_array()

Remove the following functions:

gpio_regulator_remove()


>From Rob Jones <rob.jones@codethink.co.uk> # This line is ignored.
From: Rob Jones <rob.jones@codethink.co.uk>
Subject: 
In-Reply-To: 

[PATCH 1/4] drivers/gpio: devres.c: allow gpio array requests for managed devices

[Rob Jones]

Add functions devm_gpio_request_array() and devm_gpio_free_array()
which are exactly analogous to the equivalent non-managed device
functions gpio_request_array() and gpio_free_array(), which can be
found in the module gpiolib.c.

Note that if devm_gpio_request_array() fails, no gpios are obtained.
No indication is provided as to which particular request failed.

Reviewed-by: Ian Molton <ian.molton@codethink.co.uk>
Signed-off-by: Rob Jones <rob.jones@codethink.co.uk>
---
 drivers/gpio/devres.c |   57 +++++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/gpio.h  |    4 ++++
 2 files changed, 61 insertions(+)

diff --git a/drivers/gpio/devres.c b/drivers/gpio/devres.c
index 307464f..d4c2161 100644
--- a/drivers/gpio/devres.c
+++ b/drivers/gpio/devres.c
@@ -186,6 +186,63 @@ int devm_gpio_request_one(struct device *dev, unsigned gpio,
 EXPORT_SYMBOL(devm_gpio_request_one);
 
 /**
+ * devm_gpio_request_array - request multiple GPIOs in a single call
+ *			     for a managed device
+ * @dev:	device requesting the gpio
+ * @array:	array of the 'struct gpio'
+ * @num:	how many GPIOs in the array
+ *
+ * Except for the extra @dev argument, this function takes the
+ * same arguments and performs the same function as
+ * gpio_request().  GPIOs requested with this function will be
+ * automatically freed on driver detach.
+ *
+ * If GPIOs allocated with this function need to be freed separately,
+ * devm_gpio_free_array() or devm_gpio_free() must be used.
+ */
+int devm_gpio_request_array(struct device *dev,
+			    const struct gpio *array,
+			    size_t num)
+{
+	int i, err = 0;
+
+	for (i = 0; i < num; i++, array++) {
+		err = devm_gpio_request_one(dev,
+					    array->gpio,
+					    array->flags,
+					    array->label);
+		if (err) {
+			while (i--)
+				devm_gpio_free(dev, (--array)->gpio);
+		}
+	}
+
+	return err;
+}
+EXPORT_SYMBOL(devm_gpio_request_array);
+
+/**
+ * devm_gpio_free_array - release multiple GPIOs in a single call
+ *			  for a managed device
+ * @dev:	device requesting the gpio
+ * @array:	array of the 'struct gpio'
+ * @num:	how many GPIOs in the array
+ *
+ * Except for the extra @dev argument, this function takes the
+ * same arguments and performs the same function as gpio_free_array().
+ * This function instead of gpio_free_array() should be used to
+ * manually free GPIOs allocated with devm_gpio_request().
+ */
+void devm_gpio_free_array(struct device *dev,
+			  const struct gpio *array,
+			  size_t num)
+{
+	while (num--)
+		devm_gpio_free(dev, (array++)->gpio);
+}
+EXPORT_SYMBOL(devm_gpio_free_array);
+
+/**
  *      devm_gpio_free - free a GPIO
  *      @dev: device to free GPIO for
  *      @gpio: GPIO to free
diff --git a/include/linux/gpio.h b/include/linux/gpio.h
index 85aa5d0..12d5648 100644
--- a/include/linux/gpio.h
+++ b/include/linux/gpio.h
@@ -84,6 +84,10 @@ struct device;
 int devm_gpio_request(struct device *dev, unsigned gpio, const char *label);
 int devm_gpio_request_one(struct device *dev, unsigned gpio,
 			  unsigned long flags, const char *label);
+int devm_gpio_request_array(struct device *dev, const struct gpio *array,
+			    size_t num);
+void devm_gpio_free_array(struct device *dev, const struct gpio *array,
+			 size_t num);
 void devm_gpio_free(struct device *dev, unsigned int gpio);
 
 #else /* ! CONFIG_GPIOLIB */
-- 
1.7.10.4

Re: [Linux-kernel] [PATCH 1/4] drivers/gpio: devres.c: allow gpio array requests for managed devices

[Ben Hutchings]

On Thu, 2014-06-19 at 16:46 +0100, Rob Jones wrote:
[...]
> +int devm_gpio_request_array(struct device *dev,
> +			    const struct gpio *array,
> +			    size_t num)
> +{
> +	int i, err = 0;
> +
> +	for (i = 0; i < num; i++, array++) {
> +		err = devm_gpio_request_one(dev,
> +					    array->gpio,
> +					    array->flags,
> +					    array->label);
> +		if (err) {
> +			while (i--)
> +				devm_gpio_free(dev, (--array)->gpio);

Missing break here.

> +		}
> +	}
> +
> +	return err;
> +}
> +EXPORT_SYMBOL(devm_gpio_request_array);
[...]

[PATCH 2/4] drivers/base: devres.c: Add block copy func. for managed devices

[Rob Jones]

Add function devm_kmemdup().

Reviewed-by: Ian Molton <ian.molton@codethink.co.uk>
Signed-off-by: Rob Jones <rob.jones@codethink.co.uk>
---
 drivers/base/devres.c  |   27 ++++++++++++++++++++++++++-
 include/linux/device.h |    2 ++
 2 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/drivers/base/devres.c b/drivers/base/devres.c
index db4e264..9d0f64c 100644
--- a/drivers/base/devres.c
+++ b/drivers/base/devres.c
@@ -793,7 +793,7 @@ EXPORT_SYMBOL_GPL(devm_kmalloc);
 /**
  * devm_kstrdup - Allocate resource managed space and
  *                copy an existing string into that.
- * @dev: Device to allocate memory for
+ * @dev:Device to allocate memory for
  * @s: the string to duplicate
  * @gfp: the GFP mask used in the devm_kmalloc() call when
  *       allocating memory
@@ -817,6 +817,31 @@ char *devm_kstrdup(struct device *dev, const char *s, gfp_t gfp)
 EXPORT_SYMBOL_GPL(devm_kstrdup);
 
 /**
+ * devm_kmemdup - Allocate resource managed space and
+ *                copy existing data into that.
+ * @dev:	Device to allocate memory for
+ * @s: 		The memory block to duplicate
+ * @gfp: 	The GFP mask used in the devm_kmalloc() call when
+ *		allocating memory
+ * RETURNS:
+ * Pointer to allocated string on success, NULL on failure.
+ */
+void *devm_kmemdup(struct device *dev, const void *s, size_t size, gfp_t gfp)
+{
+	void *buf;
+
+	if (!s || (size == 0))
+		return NULL;
+
+	buf = devm_kmalloc(dev, size, gfp);
+	if (buf)
+		memcpy(buf, s, size);
+
+	return buf;
+}
+EXPORT_SYMBOL_GPL(devm_kmemdup);
+
+/**
  * devm_kfree - Resource-managed kfree
  * @dev: Device this memory belongs to
  * @p: Memory to free
diff --git a/include/linux/device.h b/include/linux/device.h
index d1d1c05..7ace116 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -623,6 +623,8 @@ static inline void *devm_kcalloc(struct device *dev,
 }
 extern void devm_kfree(struct device *dev, void *p);
 extern char *devm_kstrdup(struct device *dev, const char *s, gfp_t gfp);
+extern void *devm_kmemdup(struct device *dev,
+			  const void *s, size_t size, gfp_t gfp);
 
 void __iomem *devm_ioremap_resource(struct device *dev, struct resource *res);
 void __iomem *devm_request_and_ioremap(struct device *dev,
-- 
1.7.10.4

Re: [PATCH 2/4] drivers/base: devres.c: Add block copy func. for managed devices

[Ben Hutchings]

On Thu, 2014-06-19 at 16:46 +0100, Rob Jones wrote:
[...]
> --- a/drivers/base/devres.c
> +++ b/drivers/base/devres.c
> @@ -793,7 +793,7 @@ EXPORT_SYMBOL_GPL(devm_kmalloc);
>  /**
>   * devm_kstrdup - Allocate resource managed space and
>   *                copy an existing string into that.
> - * @dev: Device to allocate memory for
> + * @dev:Device to allocate memory for

You shouldn't be changing this comment...

Ben.

>   * @s: the string to duplicate
>   * @gfp: the GFP mask used in the devm_kmalloc() call when
>   *       allocating memory
[...]

[PATCH 3/4] drivers/base: devres.c: use devm_kmemdup() from with devm_kstrdup()

[Rob Jones]

Avoid code duplication by using devm_kmemdup() to copy data instead
of having a separate loop within devm_kstrdup().

Reviewed-by: Ian Molton <ian.molton@codethink.co.uk>
Signed-off-by: Rob Jones <rob.jones@codethink.co.uk>
---
 drivers/base/devres.c |    9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)

diff --git a/drivers/base/devres.c b/drivers/base/devres.c
index 9d0f64c..cf19664 100644
--- a/drivers/base/devres.c
+++ b/drivers/base/devres.c
@@ -802,17 +802,10 @@ EXPORT_SYMBOL_GPL(devm_kmalloc);
  */
 char *devm_kstrdup(struct device *dev, const char *s, gfp_t gfp)
 {
-	size_t size;
-	char *buf;
-
 	if (!s)
 		return NULL;
 
-	size = strlen(s) + 1;
-	buf = devm_kmalloc(dev, size, gfp);
-	if (buf)
-		memcpy(buf, s, size);
-	return buf;
+	return devm_kmemdup(dev, s, (strlen(s) + 1), gfp);
 }
 EXPORT_SYMBOL_GPL(devm_kstrdup);
 
-- 
1.7.10.4

Re: [PATCH 3/4] drivers/base: devres.c: use devm_kmemdup() from with devm_kstrdup()

[Joe Perches]

On Thu, 2014-06-19 at 16:46 +0100, Rob Jones wrote:
> Avoid code duplication by using devm_kmemdup() to copy data instead
> of having a separate loop within devm_kstrdup().
> 
> Reviewed-by: Ian Molton <ian.molton@codethink.co.uk>
> Signed-off-by: Rob Jones <rob.jones@codethink.co.uk>
[]
> diff --git a/drivers/base/devres.c b/drivers/base/devres.c
[]
> @@ -802,17 +802,10 @@ EXPORT_SYMBOL_GPL(devm_kmalloc);
>   */
>  char *devm_kstrdup(struct device *dev, const char *s, gfp_t gfp)
>  {
[]
> +	return devm_kmemdup(dev, s, (strlen(s) + 1), gfp);
>  }
>  EXPORT_SYMBOL_GPL(devm_kstrdup);

Making this static inline in the header and dropping
EXPORT_SYMBOL_GPL might be smaller code.

static inline
char *devm_kstrdup(struct device *dev, const char *s, gfp_t gfp)
{
	return devm_kmemdup(dev, s, strlen(s) + 1, gfp);
}

Re: [PATCH 3/4] drivers/base: devres.c: use devm_kmemdup() from with devm_kstrdup()

[Rob Jones]

On 19/06/14 16:59, Joe Perches wrote:
> On Thu, 2014-06-19 at 16:46 +0100, Rob Jones wrote:
>> Avoid code duplication by using devm_kmemdup() to copy data instead
>> of having a separate loop within devm_kstrdup().
>>
>> Reviewed-by: Ian Molton <ian.molton@codethink.co.uk>
>> Signed-off-by: Rob Jones <rob.jones@codethink.co.uk>
> []
>> diff --git a/drivers/base/devres.c b/drivers/base/devres.c
> []
>> @@ -802,17 +802,10 @@ EXPORT_SYMBOL_GPL(devm_kmalloc);
>>    */
>>   char *devm_kstrdup(struct device *dev, const char *s, gfp_t gfp)
>>   {
> []
>> +	return devm_kmemdup(dev, s, (strlen(s) + 1), gfp);
>>   }
>>   EXPORT_SYMBOL_GPL(devm_kstrdup);
> Making this static inline in the header and dropping
> EXPORT_SYMBOL_GPL might be smaller code.
>
> static inline
> char *devm_kstrdup(struct device *dev, const char *s, gfp_t gfp)
> {
> 	return devm_kmemdup(dev, s, strlen(s) + 1, gfp);
> }
>

Is it wise not to check for (s == NULL)? Surely the behaviour of 
strlen(NULL) is undefined.

Consequently, I am, on balance, against making the function static 
inline. Given that inline is only a recommendation to the compiler, it 
may well not inline it anyway if the NULL test is present, in which case 
we would end up with a copy of the code in each module that uses it 
rather than a single, global, copy. If it didn't need the NULL test, I 
would agree that it would be best as a static inline but I think it does 
need it.

-- 
Rob Jones
Project Manager
Codethink Ltd
mailto:rob.jones@codethink.co.uk
tel:+44 161 236 5575

[PATCH 4/4] drivers/regulator: gpio-regulator.c: use managed resources for probe.

[Rob Jones]

Use managed resource functions for all resource allocations in
gpio_regulator_probe().

Remove gpio_regulator_remove() as it is now redundant.

Reviewed-by: Ian Molton <ian.molton@codethink.co.uk>
Signed-off-by: Rob Jones <rob.jones@codethink.co.uk>
---
 drivers/regulator/gpio-regulator.c |   70 ++++++++++++------------------------
 1 file changed, 23 insertions(+), 47 deletions(-)

diff --git a/drivers/regulator/gpio-regulator.c b/drivers/regulator/gpio-regulator.c
index 989b23b..6edde12 100644
--- a/drivers/regulator/gpio-regulator.c
+++ b/drivers/regulator/gpio-regulator.c
@@ -254,30 +254,31 @@ static int gpio_regulator_probe(struct platform_device *pdev)
 	if (drvdata == NULL)
 		return -ENOMEM;
 
-	drvdata->desc.name = kstrdup(config->supply_name, GFP_KERNEL);
+	drvdata->desc.name = devm_kstrdup(&pdev->dev,
+					  config->supply_name,
+					  GFP_KERNEL);
 	if (drvdata->desc.name == NULL) {
 		dev_err(&pdev->dev, "Failed to allocate supply name\n");
-		ret = -ENOMEM;
-		goto err;
+		return -ENOMEM;
 	}
 
-	drvdata->gpios = kmemdup(config->gpios,
-				 config->nr_gpios * sizeof(struct gpio),
-				 GFP_KERNEL);
+	drvdata->gpios = devm_kmemdup(&pdev->dev,
+				      config->gpios,
+				      config->nr_gpios * sizeof(struct gpio),
+				      GFP_KERNEL);
 	if (drvdata->gpios == NULL) {
 		dev_err(&pdev->dev, "Failed to allocate gpio data\n");
-		ret = -ENOMEM;
-		goto err_name;
+		return -ENOMEM;
 	}
 
-	drvdata->states = kmemdup(config->states,
-				  config->nr_states *
+	drvdata->states = devm_kmemdup(&pdev->dev,
+				       config->states,
+				       config->nr_states *
 					 sizeof(struct gpio_regulator_state),
-				  GFP_KERNEL);
+				       GFP_KERNEL);
 	if (drvdata->states == NULL) {
 		dev_err(&pdev->dev, "Failed to allocate state data\n");
-		ret = -ENOMEM;
-		goto err_memgpio;
+		return -ENOMEM;
 	}
 	drvdata->nr_states = config->nr_states;
 
@@ -297,16 +298,17 @@ static int gpio_regulator_probe(struct platform_device *pdev)
 		break;
 	default:
 		dev_err(&pdev->dev, "No regulator type set\n");
-		ret = -EINVAL;
-		goto err_memgpio;
+		return -EINVAL;
 	}
 
 	drvdata->nr_gpios = config->nr_gpios;
-	ret = gpio_request_array(drvdata->gpios, drvdata->nr_gpios);
+	ret = devm_gpio_request_array(&pdev->dev,
+				      drvdata->gpios,
+				      drvdata->nr_gpios);
 	if (ret) {
 		dev_err(&pdev->dev,
 		   "Could not obtain regulator setting GPIOs: %d\n", ret);
-		goto err_memstate;
+		return ret;
 	}
 
 	/* build initial state from gpio init data. */
@@ -337,43 +339,18 @@ static int gpio_regulator_probe(struct platform_device *pdev)
 			cfg.ena_gpio_flags |= GPIOF_OUT_INIT_HIGH;
 	}
 
-	drvdata->dev = regulator_register(&drvdata->desc, &cfg);
+	drvdata->dev = devm_regulator_register(&pdev->dev,
+					       &drvdata->desc,
+					       &cfg);
 	if (IS_ERR(drvdata->dev)) {
 		ret = PTR_ERR(drvdata->dev);
 		dev_err(&pdev->dev, "Failed to register regulator: %d\n", ret);
-		goto err_stategpio;
+		return ret;
 	}
 
 	platform_set_drvdata(pdev, drvdata);
 
 	return 0;
-
-err_stategpio:
-	gpio_free_array(drvdata->gpios, drvdata->nr_gpios);
-err_memstate:
-	kfree(drvdata->states);
-err_memgpio:
-	kfree(drvdata->gpios);
-err_name:
-	kfree(drvdata->desc.name);
-err:
-	return ret;
-}
-
-static int gpio_regulator_remove(struct platform_device *pdev)
-{
-	struct gpio_regulator_data *drvdata = platform_get_drvdata(pdev);
-
-	regulator_unregister(drvdata->dev);
-
-	gpio_free_array(drvdata->gpios, drvdata->nr_gpios);
-
-	kfree(drvdata->states);
-	kfree(drvdata->gpios);
-
-	kfree(drvdata->desc.name);
-
-	return 0;
 }
 
 #if defined(CONFIG_OF)
@@ -385,7 +362,6 @@ static const struct of_device_id regulator_gpio_of_match[] = {
 
 static struct platform_driver gpio_regulator_driver = {
 	.probe		= gpio_regulator_probe,
-	.remove		= gpio_regulator_remove,
 	.driver		= {
 		.name		= "gpio-regulator",
 		.owner		= THIS_MODULE,
-- 
1.7.10.4

Re: Change gpio-reulator-probe() to use managed resources.

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 724 bytes --]

On Thu, Jun 19, 2014 at 04:46:01PM +0100, Rob Jones wrote:
> Add new managed resource functions as needed to achieve this.
> 
> Amend the following functions:

You appear to have omitted the actual patch, I'm quoting the entire
mail here.  Please see Documentation/SubmittingPatches for some tips
on how to do things.

> 
> devm_kstrdup()
> gpio_regulator_probe()
> 
> Add the following functions:
> 
> devm_kmemdup()
> devm_gpio_request_array()
> devm_gpio_free_array()
> 
> Remove the following functions:
> 
> gpio_regulator_remove()
> 
> 
> From Rob Jones <rob.jones@codethink.co.uk> # This line is ignored.
> From: Rob Jones <rob.jones@codethink.co.uk>
> Subject: 
> In-Reply-To: 
> 
> 


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: Change gpio-reulator-probe() to use managed resources.

[Rob Jones]

On 19/06/14 18:57, Mark Brown wrote:
> On Thu, Jun 19, 2014 at 04:46:01PM +0100, Rob Jones wrote:
>> Add new managed resource functions as needed to achieve this.
>>
>> Amend the following functions:
> You appear to have omitted the actual patch, I'm quoting the entire
> mail here.  Please see Documentation/SubmittingPatches for some tips
> on how to do things.

I omitted to put "[Patch 0/4][v2]" on the front of the subject line.The 
actual patches followed.

It was a consequence of lack of familiarity with the tools; I'm 
learning, so it shouldn't happen again. Would you like me to resubmit?

>
>> devm_kstrdup()
>> gpio_regulator_probe()
>>
>> Add the following functions:
>>
>> devm_kmemdup()
>> devm_gpio_request_array()
>> devm_gpio_free_array()
>>
>> Remove the following functions:
>>
>> gpio_regulator_remove()
>>
>>
>>  From Rob Jones <rob.jones@codethink.co.uk> # This line is ignored.
>> From: Rob Jones <rob.jones@codethink.co.uk>
>> Subject:
>> In-Reply-To:
>>
>>

-- 
Rob Jones
Project Manager
Codethink Ltd
mailto:rob.jones@codethink.co.uk
tel:+44 161 236 5575

Re: Change gpio-reulator-probe() to use managed resources.

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 641 bytes --]

On Fri, Jun 20, 2014 at 09:38:29AM +0100, Rob Jones wrote:

> I omitted to put "[Patch 0/4][v2]" on the front of the subject line.The
> actual patches followed.

Oh, so this is associated with a patch series?

> It was a consequence of lack of familiarity with the tools; I'm learning, so
> it shouldn't happen again. Would you like me to resubmit?

I do see another patch from you for the regulator subsystem which is
part of a series so I guess no need.

Please, if you are resubmitting try to use a changelog style consistent
with the code you're submitting against - you've got subject lines with
"drivers/" which the kernel doesn't do.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]