Re: [V2 PATCH 02/10] i2c/of: Allow device node to be passed via i2c_board_info

Re: [V2 PATCH 02/10] i2c/of: Allow device node to be passed via i2c_board_info

[Grant Likely]

Ben & Jean,

I'm getting patches queued up for linux-next.  Any objections to this
patch or to merging it via the devicetree tree?  This patch is part of
the device tree support rework that consolidates most of the
arch-specific code.

Thanks,
g.

On Thu, Mar 18, 2010 at 9:22 AM, Grant Likely <grant.likely@secretlab.ca> wrote:
> The struct device_node *of_node pointer is moving out of dev->archdata
> and into the struct device proper.  of_i2c.c needs to set the of_node
> pointer before the device is registered.  Since the i2c subsystem
> doesn't allow 2 stage allocation and registration of i2c devices, the
> of_node pointer needs to be passed via the i2c_board_info structure
> so that it is set prior to registration.
>
> This patch adds of_node to struct i2c_board_info (conditional on
> CONFIG_OF), sets of_node in i2c_new_device(), and modifies of_i2c.c
> to use the new parameter.  The calling of dev_archdata_set_node()
> from of_i2c will be removed in a subsequent patch when of_node is
> removed from archdata and all users are converted over.
>
> Signed-off-by: Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org>
> ---
>
>  drivers/i2c/i2c-core.c |    3 +++
>  drivers/of/of_i2c.c    |    1 +
>  include/linux/i2c.h    |    4 ++++
>  3 files changed, 8 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
> index 3202a86..4099b2b 100644
> --- a/drivers/i2c/i2c-core.c
> +++ b/drivers/i2c/i2c-core.c
> @@ -387,6 +387,9 @@ i2c_new_device(struct i2c_adapter *adap, struct i2c_board_info const *info)
>        client->dev.parent = &client->adapter->dev;
>        client->dev.bus = &i2c_bus_type;
>        client->dev.type = &i2c_client_type;
> +#ifdef CONFIG_OF
> +       client->dev.of_node = info->of_node;
> +#endif
>
>        dev_set_name(&client->dev, "%d-%04x", i2c_adapter_id(adap),
>                     client->addr);
> diff --git a/drivers/of/of_i2c.c b/drivers/of/of_i2c.c
> index a3a708e..e690a2a 100644
> --- a/drivers/of/of_i2c.c
> +++ b/drivers/of/of_i2c.c
> @@ -43,6 +43,7 @@ void of_register_i2c_devices(struct i2c_adapter *adap,
>                info.addr = be32_to_cpup(addr);
>
>                dev_archdata_set_node(&dev_ad, node);
> +               info.of_node = node;
>                info.archdata = &dev_ad;
>
>                request_module("%s", info.type);
> diff --git a/include/linux/i2c.h b/include/linux/i2c.h
> index 0a5da63..4f37ff1 100644
> --- a/include/linux/i2c.h
> +++ b/include/linux/i2c.h
> @@ -34,6 +34,7 @@
>  #include <linux/device.h>      /* for struct device */
>  #include <linux/sched.h>       /* for completion */
>  #include <linux/mutex.h>
> +#include <linux/of.h>          /* for struct device_node */
>
>  extern struct bus_type i2c_bus_type;
>
> @@ -251,6 +252,9 @@ struct i2c_board_info {
>        unsigned short  addr;
>        void            *platform_data;
>        struct dev_archdata     *archdata;
> +#ifdef CONFIG_OF
> +       struct device_node *of_node;
> +#endif
>        int             irq;
>  };
>
>
>



-- 
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.

Re: [V2 PATCH 02/10] i2c/of: Allow device node to be passed via i2c_board_info

[Jean Delvare]

Hi Grant,

On Tue, 27 Apr 2010 12:19:55 -0600, Grant Likely wrote:
> Ben & Jean,
> 
> I'm getting patches queued up for linux-next.  Any objections to this
> patch or to merging it via the devicetree tree?  This patch is part of
> the device tree support rework that consolidates most of the
> arch-specific code.

I didn't follow closely why it was needed, but I will trust you and the
code contributor. I have no objection.

-- 
Jean Delvare