[v2,0/4] device property: Add fwnode_get_name() helper

Message

Heikki Krogerus Nov. 8, 2018, 4:51 p.m. UTC

Hi,

This is the second version of my proposal for this helper. The
first version can be checked here:
https://lkml.org/lkml/2018/11/5/326

In order to support also ACPI properly, I decided to change the API.
The function fwnode_name() is now fwnode_get_name(), and instead of
returning pointer to the name, the function copies it to a buffer. I
did that because acpica does not offer a way to get a pointer to the
node name, and the name is clearly expected to be accessed only with
the namespace lock held.

I think this is better approach in any case. It will also solve the
problem of getting rid of the unit-address part from DT node names.

Let me know what you guys think.

--
heikki


Heikki Krogerus (4):
  device property: Introduce fwnode_get_name()
  ACPI: property: Add acpi_fwnode_name()
  of/property: Add of_fwnode_name()
  device property: Drop get_named_child_node callback

 drivers/acpi/property.c  | 43 ++++++++++++++++++++++------------------
 drivers/base/property.c  | 34 ++++++++++++++++++++++++++++++-
 drivers/of/property.c    | 26 ++++++++++--------------
 include/linux/fwnode.h   |  6 +++---
 include/linux/property.h |  2 ++
 5 files changed, 73 insertions(+), 38 deletions(-)

Comments

Rafael J. Wysocki Nov. 20, 2018, 10:03 p.m. UTC | #1

On Thursday, November 8, 2018 5:51:52 PM CET Heikki Krogerus wrote:
> Hi,
> 
> This is the second version of my proposal for this helper. The
> first version can be checked here:
> https://lkml.org/lkml/2018/11/5/326
> 
> In order to support also ACPI properly, I decided to change the API.
> The function fwnode_name() is now fwnode_get_name(), and instead of
> returning pointer to the name, the function copies it to a buffer. I
> did that because acpica does not offer a way to get a pointer to the
> node name, and the name is clearly expected to be accessed only with
> the namespace lock held.
> 
> I think this is better approach in any case. It will also solve the
> problem of getting rid of the unit-address part from DT node names.
> 
> Let me know what you guys think.
> 
> --
> heikki
> 
> 
> Heikki Krogerus (4):
>   device property: Introduce fwnode_get_name()
>   ACPI: property: Add acpi_fwnode_name()
>   of/property: Add of_fwnode_name()
>   device property: Drop get_named_child_node callback
> 
>  drivers/acpi/property.c  | 43 ++++++++++++++++++++++------------------
>  drivers/base/property.c  | 34 ++++++++++++++++++++++++++++++-
>  drivers/of/property.c    | 26 ++++++++++--------------
>  include/linux/fwnode.h   |  6 +++---
>  include/linux/property.h |  2 ++
>  5 files changed, 73 insertions(+), 38 deletions(-)
> 
> 

I will be expecting at least one more iteration of this series to address the
Rob's comments.

Thanks,
Rafael

Heikki Krogerus Nov. 21, 2018, 12:36 p.m. UTC | #2

On Tue, Nov 20, 2018 at 11:03:45PM +0100, Rafael J. Wysocki wrote:
> On Thursday, November 8, 2018 5:51:52 PM CET Heikki Krogerus wrote:
> > Hi,
> > 
> > This is the second version of my proposal for this helper. The
> > first version can be checked here:
> > https://lkml.org/lkml/2018/11/5/326
> > 
> > In order to support also ACPI properly, I decided to change the API.
> > The function fwnode_name() is now fwnode_get_name(), and instead of
> > returning pointer to the name, the function copies it to a buffer. I
> > did that because acpica does not offer a way to get a pointer to the
> > node name, and the name is clearly expected to be accessed only with
> > the namespace lock held.
> > 
> > I think this is better approach in any case. It will also solve the
> > problem of getting rid of the unit-address part from DT node names.
> > 
> > Let me know what you guys think.
> > 
> > --
> > heikki
> > 
> > 
> > Heikki Krogerus (4):
> >   device property: Introduce fwnode_get_name()
> >   ACPI: property: Add acpi_fwnode_name()
> >   of/property: Add of_fwnode_name()
> >   device property: Drop get_named_child_node callback
> > 
> >  drivers/acpi/property.c  | 43 ++++++++++++++++++++++------------------
> >  drivers/base/property.c  | 34 ++++++++++++++++++++++++++++++-
> >  drivers/of/property.c    | 26 ++++++++++--------------
> >  include/linux/fwnode.h   |  6 +++---
> >  include/linux/property.h |  2 ++
> >  5 files changed, 73 insertions(+), 38 deletions(-)
> > 
> > 
> 
> I will be expecting at least one more iteration of this series to address the
> Rob's comments.

Yes, I'll send a new version soon.

thanks,