Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1561106 > unrolled thread

[PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name()

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name() Florian Fainelli <f.fainelli@gmail.com> - 2017-01-18 00:30 +0100
    Re: [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name() Andy Shevchenko <andy.shevchenko@gmail.com> - 2017-01-18 00:40 +0100
      Re: [PATCH net-next v4 05/10] drivers: base: Add  device_find_in_class_name() Florian Fainelli <f.fainelli@gmail.com> - 2017-01-18 01:00 +0100
        Re: [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name() Andy Shevchenko <andy.shevchenko@gmail.com> - 2017-01-18 01:10 +0100
          Re: [PATCH net-next v4 05/10] drivers: base: Add  device_find_in_class_name() Florian Fainelli <f.fainelli@gmail.com> - 2017-01-18 01:20 +0100
        Re: [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name() Andy Shevchenko <andy.shevchenko@gmail.com> - 2017-01-18 01:10 +0100
          Re: [PATCH net-next v4 05/10] drivers: base: Add  device_find_in_class_name() Florian Fainelli <f.fainelli@gmail.com> - 2017-01-18 01:10 +0100
    Re: [PATCH net-next v4 05/10] drivers: base: Add  device_find_in_class_name() Greg Kroah-Hartman <gregkh@linuxfoundation.org> - 2017-01-18 08:20 +0100

#1561106 — [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name()

Add a helper function to lookup a device reference given a class name.
This is a preliminary patch to remove adhoc code from net/dsa/dsa.c and
make it more generic.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/base/core.c    | 31 +++++++++++++++++++++++++++++++
 include/linux/device.h |  2 ++
 2 files changed, 33 insertions(+)

diff --git a/drivers/base/core.c b/drivers/base/core.c
index 8c25e68e67d7..fb9fced38634 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -2058,6 +2058,37 @@ struct device *device_find_child(struct device *parent, void *data,
 }
 EXPORT_SYMBOL_GPL(device_find_child);
 
+static int device_class_name_match(struct device *dev, void *class)
+{
+	if (dev->class != NULL && !strcmp(dev->class->name, class))
+		return 1;
+
+	return 0;
+}
+
+/**
+ * device_find_in_class_name - device iterator for locating a particular device
+ * within the specified class name
+ * @parent: parent struct device
+ * @class_name: Class name to match against
+ *
+ * This function returns 1 if the device (specified by @parent), or one of its child
+ * is in the class whose name is specified by @class_name. Returns 0 otherwise.
+ *
+ * NOTE: you will need to drop the reference with put_device() after use.
+ */
+struct device *device_find_in_class_name(struct device *parent,
+					 char *class_name)
+{
+	if (device_class_name_match(parent, class_name)) {
+		get_device(parent);
+		return parent;
+	}
+
+	return device_find_child(parent, class_name, device_class_name_match);
+}
+EXPORT_SYMBOL_GPL(device_find_in_class_name);
+
 int __init devices_init(void)
 {
 	devices_kset = kset_create_and_add("devices", &device_uevent_ops, NULL);
diff --git a/include/linux/device.h b/include/linux/device.h
index 491b4c0ca633..fbc2a255f92e 100644
--- a/include/linux/device.h
+++ b/include/linux/device.h
@@ -1120,6 +1120,8 @@ extern int device_for_each_child_reverse(struct device *dev, void *data,
 		     int (*fn)(struct device *dev, void *data));
 extern struct device *device_find_child(struct device *dev, void *data,
 				int (*match)(struct device *dev, void *data));
+extern struct device *device_find_in_class_name(struct device *parent,
+						char *class_name);
 extern int device_rename(struct device *dev, const char *new_name);
 extern int device_move(struct device *dev, struct device *new_parent,
 		       enum dpm_order dpm_order);
-- 
2.9.3

[toc] | [next] | [standalone]

#1561126

On Wed, Jan 18, 2017 at 1:21 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
> Add a helper function to lookup a device reference given a class name.
> This is a preliminary patch to remove adhoc code from net/dsa/dsa.c and
> make it more generic.


> +static int device_class_name_match(struct device *dev, void *class)

And why not const char *class?

> +{
> +       if (dev->class != NULL && !strcmp(dev->class->name, class))

if (dev->class && ...)

> +               return 1;
> +
> +       return 0;

Perhaps even one line:

return dev->class && ...;

> +}
> +
> +/**
> + * device_find_in_class_name - device iterator for locating a particular device
> + * within the specified class name
> + * @parent: parent struct device
> + * @class_name: Class name to match against
> + *
> + * This function returns 1 if the device (specified by @parent), or one of its child
> + * is in the class whose name is specified by @class_name. Returns 0 otherwise.
> + *
> + * NOTE: you will need to drop the reference with put_device() after use.
> + */
> +struct device *device_find_in_class_name(struct device *parent,
> +                                        char *class_name)

const char *class_name

> +{
> +       if (device_class_name_match(parent, class_name)) {
> +               get_device(parent);
> +               return parent;
> +       }
> +
> +       return device_find_child(parent, class_name, device_class_name_match);
> +}
> +EXPORT_SYMBOL_GPL(device_find_in_class_name);

> +extern struct device *device_find_in_class_name(struct device *parent,
> +                                               char *class_name);

Ditto.

-- 
With Best Regards,
Andy Shevchenko

[toc] | [prev] | [next] | [standalone]

#1561136 — Re: [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name()

On 01/17/2017 03:34 PM, Andy Shevchenko wrote:
> On Wed, Jan 18, 2017 at 1:21 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
>> Add a helper function to lookup a device reference given a class name.
>> This is a preliminary patch to remove adhoc code from net/dsa/dsa.c and
>> make it more generic.
> 
> 
>> +static int device_class_name_match(struct device *dev, void *class)
> 
> And why not const char *class?

This was raised back in v2, and the same response applies:

https://www.mail-archive.com/netdev@vger.kernel.org/msg147559.html

Changing the signature of a callback is out of the scope of this patch
series.
-- 
Florian

[toc] | [prev] | [next] | [standalone]

#1561159

On Wed, Jan 18, 2017 at 2:04 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
> On 01/17/2017 04:00 PM, Andy Shevchenko wrote:
>> On Wed, Jan 18, 2017 at 1:43 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
>>> On 01/17/2017 03:34 PM, Andy Shevchenko wrote:
>>>> On Wed, Jan 18, 2017 at 1:21 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:

>> But why not to use void *class_name to be consistent with callback and
>> device_find_child()?
>
> The top-level function: device_find_in_class_name() should have a
> stronger typing of its argument even if it internally uses
> device_find_child() and a callback that takes a void * argument, that's
> how I see it.

Fair enough.

>> Btw,
>> return get_device(parent);
>
> Not sure I follow what that means here?

Missed remark. Instead of

get_device(parent);
return parent;

you can use

return get_device(parent);

-- 
With Best Regards,
Andy Shevchenko

[toc] | [prev] | [next] | [standalone]

#1561166 — Re: [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name()

On 01/17/2017 04:07 PM, Andy Shevchenko wrote:
> On Wed, Jan 18, 2017 at 2:04 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
>> On 01/17/2017 04:00 PM, Andy Shevchenko wrote:
>>> On Wed, Jan 18, 2017 at 1:43 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
>>>> On 01/17/2017 03:34 PM, Andy Shevchenko wrote:
>>>>> On Wed, Jan 18, 2017 at 1:21 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
> 
>>> But why not to use void *class_name to be consistent with callback and
>>> device_find_child()?
>>
>> The top-level function: device_find_in_class_name() should have a
>> stronger typing of its argument even if it internally uses
>> device_find_child() and a callback that takes a void * argument, that's
>> how I see it.
> 
> Fair enough.
> 
>>> Btw,
>>> return get_device(parent);
>>
>> Not sure I follow what that means here?
> 
> Missed remark. Instead of
> 
> get_device(parent);
> return parent;
> 
> you can use
> 
> return get_device(parent);

Seems reasonable, if I have to respin a v5, will add that, thanks!
-- 
Florian

[toc] | [prev] | [next] | [standalone]

#1561163

On Wed, Jan 18, 2017 at 1:43 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
> On 01/17/2017 03:34 PM, Andy Shevchenko wrote:
>> On Wed, Jan 18, 2017 at 1:21 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:

>>> +static int device_class_name_match(struct device *dev, void *class)
>>
>> And why not const char *class?
>
> This was raised back in v2, and the same response applies:
>
> https://www.mail-archive.com/netdev@vger.kernel.org/msg147559.html
>
> Changing the signature of a callback is out of the scope of this patch
> series.

Ah, right.

But why not to use void *class_name to be consistent with callback and
device_find_child()?

Btw,

return get_device(parent);

-- 
With Best Regards,
Andy Shevchenko

[toc] | [prev] | [next] | [standalone]

#1561164 — Re: [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name()

On 01/17/2017 04:00 PM, Andy Shevchenko wrote:
> On Wed, Jan 18, 2017 at 1:43 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
>> On 01/17/2017 03:34 PM, Andy Shevchenko wrote:
>>> On Wed, Jan 18, 2017 at 1:21 AM, Florian Fainelli <f.fainelli@gmail.com> wrote:
> 
>>>> +static int device_class_name_match(struct device *dev, void *class)
>>>
>>> And why not const char *class?
>>
>> This was raised back in v2, and the same response applies:
>>
>> https://www.mail-archive.com/netdev@vger.kernel.org/msg147559.html
>>
>> Changing the signature of a callback is out of the scope of this patch
>> series.
> 
> Ah, right.
> 
> But why not to use void *class_name to be consistent with callback and
> device_find_child()?

The top-level function: device_find_in_class_name() should have a
stronger typing of its argument even if it internally uses
device_find_child() and a callback that takes a void * argument, that's
how I see it.

> 
> Btw,
> 
> return get_device(parent);

Not sure I follow what that means here?
-- 
Florian

[toc] | [prev] | [next] | [standalone]

#1561327 — Re: [PATCH net-next v4 05/10] drivers: base: Add device_find_in_class_name()

On Tue, Jan 17, 2017 at 03:21:47PM -0800, Florian Fainelli wrote:
> Add a helper function to lookup a device reference given a class name.
> This is a preliminary patch to remove adhoc code from net/dsa/dsa.c and
> make it more generic.
> 
> Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
> ---
>  drivers/base/core.c    | 31 +++++++++++++++++++++++++++++++
>  include/linux/device.h |  2 ++
>  2 files changed, 33 insertions(+)

My NAK still stands here, please give me a day or so to respond to the
other thread about this...

thanks,

greg k-h

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web