[PATCH] i2c: tegra: Fix compile when !CONFIG_OF
Grant Likely
grant.likely at secretlab.ca
Thu Jul 14 08:48:25 EST 2011
On Thu, Jul 14, 2011 at 5:36 AM, Stephen Warren <swarren at nvidia.com> wrote:
> When !CONFIG_OF, <linux/of.h> doesn't prototype of_get_property(), which
> causes i2c-tegra.c to fail to compile. Make the references to this
> function conditional to solve this.
>
> Signed-off-by: Stephen Warren <swarren at nvidia.com>
> ---
> Grant, is this the correct way to solve this? Other solutions I can see are:
> a) Make ARM or Tegra select CONFIG_OF, now that ARM is moving to Device Tree.
> b) Make <linux/of.h> provide dummy inline versions of its functions when
> !CONFIG_OF.
The solution you chose is sufficient for now, but turning on CONFIG_OF
unconditionally for Tegra would be fine too since I hope to get the
tegra board-dt patch merged for 3.1 (I hope to have the patch cleaned
up and posted in the next 24 hrs). I've been avoiding creating a
dummy inline of of_get_property(), but not for any particularly strong
reason other than to avoid bloating linux/of.h. I probably would
accept a patch that added it.
Acked-by: Grant Likely <grant.likely at secretlab.ca>
g.
>
> drivers/i2c/busses/i2c-tegra.c | 5 ++++-
> 1 files changed, 4 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/i2c/busses/i2c-tegra.c b/drivers/i2c/busses/i2c-tegra.c
> index 2440b74..5a18d6d 100644
> --- a/drivers/i2c/busses/i2c-tegra.c
> +++ b/drivers/i2c/busses/i2c-tegra.c
> @@ -610,12 +610,15 @@ static int tegra_i2c_probe(struct platform_device *pdev)
> if (pdata) {
> i2c_dev->bus_clk_rate = pdata->bus_clk_rate;
>
> - } else if (i2c_dev->dev->of_node) { /* if there is a device tree node ... */
> + }
> +#ifdef CONFIG_OF
> + else if (i2c_dev->dev->of_node) { /* if there is a device tree node ... */
> prop = of_get_property(i2c_dev->dev->of_node,
> "clock-frequency", NULL);
> if (prop)
> i2c_dev->bus_clk_rate = be32_to_cpup(prop);
> }
> +#endif
>
> if (pdev->id == 3)
> i2c_dev->is_dvc = 1;
> --
> 1.7.0.4
>
>
--
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
More information about the devicetree-discuss
mailing list