diff mbox series

of: Add device_type access helper functions

Message ID 20180830185229.20393-1-robh@kernel.org
State Accepted, archived
Headers show
Series of: Add device_type access helper functions | expand

Commit Message

Rob Herring (Arm) Aug. 30, 2018, 6:52 p.m. UTC
In preparation to remove direct access to device_node.type, add
of_node_is_type() and of_node_get_device_type() helpers to check and
retrieve the device type.

Cc: Frank Rowand <frowand.list@gmail.com>
Signed-off-by: Rob Herring <robh@kernel.org>
---
 include/linux/of.h | 12 ++++++++++++
 1 file changed, 12 insertions(+)

Comments

Frank Rowand Sept. 5, 2018, 3:55 a.m. UTC | #1
On 08/30/18 11:52, Rob Herring wrote:
> In preparation to remove direct access to device_node.type, add
> of_node_is_type() and of_node_get_device_type() helpers to check and
> retrieve the device type.
> 
> Cc: Frank Rowand <frowand.list@gmail.com>
> Signed-off-by: Rob Herring <robh@kernel.org>
> ---
>  include/linux/of.h | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/include/linux/of.h b/include/linux/of.h
> index a40f63a36afa..506beca9588d 100644
> --- a/include/linux/of.h
> +++ b/include/linux/of.h
> @@ -980,6 +980,18 @@ static inline struct device_node *of_find_matching_node(
>  	return of_find_matching_node_and_match(from, matches, NULL);
>  }
>  
> +static inline const char *of_node_get_device_type(const struct device_node *np)
> +{
> +	return of_get_property(np, "type", NULL);

I'm a little confused.  The patch comment says "remove direct access to device_node.type.
device_node.type is set to the value of the property "device_type" in populate_node(),
but of_node_get_device_type() is looking for property "type".


> +}> +
> +static inline bool of_node_is_type(const struct device_node *np, const char *type)
> +{
> +	const char *match = of_node_get_device_type(np);
> +
> +	return np && match && type && !of_prop_cmp(match, type);
> +}
> +
>  /**
>   * of_property_count_u8_elems - Count the number of u8 elements in a property
>   *
> 

I'd like to see one or two examples of how you plan to use of_node_get_device_type()
(outside of of_node_is_type()) and examples of planned use of of_node_is_type() so
I can better understand anticipated usage.  Complete, clean patches not needed for
the examples, just enough text for me to see the resulting code change where the
helpers are used.

-Frank
Rob Herring (Arm) Sept. 5, 2018, 1:17 p.m. UTC | #2
On Tue, Sep 4, 2018 at 10:55 PM Frank Rowand <frowand.list@gmail.com> wrote:
>
> On 08/30/18 11:52, Rob Herring wrote:
> > In preparation to remove direct access to device_node.type, add
> > of_node_is_type() and of_node_get_device_type() helpers to check and
> > retrieve the device type.
> >
> > Cc: Frank Rowand <frowand.list@gmail.com>
> > Signed-off-by: Rob Herring <robh@kernel.org>
> > ---
> >  include/linux/of.h | 12 ++++++++++++
> >  1 file changed, 12 insertions(+)
> >
> > diff --git a/include/linux/of.h b/include/linux/of.h
> > index a40f63a36afa..506beca9588d 100644
> > --- a/include/linux/of.h
> > +++ b/include/linux/of.h
> > @@ -980,6 +980,18 @@ static inline struct device_node *of_find_matching_node(
> >       return of_find_matching_node_and_match(from, matches, NULL);
> >  }
> >
> > +static inline const char *of_node_get_device_type(const struct device_node *np)
> > +{
> > +     return of_get_property(np, "type", NULL);
>
> I'm a little confused.  The patch comment says "remove direct access to device_node.type.
> device_node.type is set to the value of the property "device_type" in populate_node(),
> but of_node_get_device_type() is looking for property "type".

Crap. That should be device_type.

>
> > +}> +
> > +static inline bool of_node_is_type(const struct device_node *np, const char *type)
> > +{
> > +     const char *match = of_node_get_device_type(np);
> > +
> > +     return np && match && type && !of_prop_cmp(match, type);
> > +}
> > +
> >  /**
> >   * of_property_count_u8_elems - Count the number of u8 elements in a property
> >   *
> >
>
> I'd like to see one or two examples of how you plan to use of_node_get_device_type()
> (outside of of_node_is_type()) and examples of planned use of of_node_is_type() so
> I can better understand anticipated usage.  Complete, clean patches not needed for
> the examples, just enough text for me to see the resulting code change where the
> helpers are used.

See my dt/testing branch. Most cases are of_node_is_type which
replaces strcmp's of node->type. I'll be sending out the patches soon.

Rob
diff mbox series

Patch

diff --git a/include/linux/of.h b/include/linux/of.h
index a40f63a36afa..506beca9588d 100644
--- a/include/linux/of.h
+++ b/include/linux/of.h
@@ -980,6 +980,18 @@  static inline struct device_node *of_find_matching_node(
 	return of_find_matching_node_and_match(from, matches, NULL);
 }
 
+static inline const char *of_node_get_device_type(const struct device_node *np)
+{
+	return of_get_property(np, "type", NULL);
+}
+
+static inline bool of_node_is_type(const struct device_node *np, const char *type)
+{
+	const char *match = of_node_get_device_type(np);
+
+	return np && match && type && !of_prop_cmp(match, type);
+}
+
 /**
  * of_property_count_u8_elems - Count the number of u8 elements in a property
  *