This patch adds a inline dummy implementations of_find_matching_node()
in "#ifdef CONFIG_OF" else part. Without this patch,build system can
lead to issues. This was discovered during randconfig testing,in which
ARM_CCI was enabled w/o CONFIG_OF being enabled,leading to the following error:
CC drivers/bus/arm-cci.o
drivers/bus/arm-cci.c: In function ‘cci_probe’:
drivers/bus/arm-cci.c:971:2: error: implicit declaration of function
‘of_find_matching_node’ [-Werror=implicit-function-declaration]
cc1: some warnings being treated as errors
make[2]: *** [drivers/bus/arm-cci.o] Error 1
make[1]: *** [drivers/bus] Error 2
make: *** [drivers] Error 2
Signed-off-by: Manjunath Goudar <[email protected]>
Cc: Grant Likely <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: [email protected]
Cc: [email protected]
---
include/linux/of.h | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/include/linux/of.h b/include/linux/of.h
index f95aee3..2c57aa2 100644
--- a/include/linux/of.h
+++ b/include/linux/of.h
@@ -340,6 +340,13 @@ int of_device_is_stdout_path(struct device_node *dn);
#else /* CONFIG_OF */
+static inline struct device_node *of_find_matching_node(
+ struct device_node *from,
+ const struct of_device_id *matches)
+{
+ return NULL;
+}
+
static inline const char* of_node_full_name(struct device_node *np)
{
return "<no-node>";
--
1.7.9.5
On Mon, Oct 14, 2013 at 03:52:27PM +0530, Majunath Goudar wrote:
> This patch adds a inline dummy implementations of_find_matching_node()
> in "#ifdef CONFIG_OF" else part. Without this patch,build system can
> lead to issues. This was discovered during randconfig testing,in which
> ARM_CCI was enabled w/o CONFIG_OF being enabled,leading to the following error:
Without fetching config info from the DT, the CCI driver is not usable,
so it makes no sense to build this driver without CONFIG_OF.
Your previous patch [1] looked sensible, but I don't see any responses
on lakml. Are you still pursuing that patch, or is this patch intended
as a replacement?
Dummy functions for things like of_find_matching_node() might be useful
for drivers that can fall back to other config mechanisms if OF is
unavailable, but that is a separate discussion.
arm-cci is not such a driver, and any such drivers that exist toady
must be coping without the dummy OF functions.
Cheers
---Dave
[1] http://lists.infradead.org/pipermail/linux-arm-kernel/2013-October/202996.html
([PATCH] BUS: CCI : Fix arm-cci implicit declarations dependence errors.)
>
> CC drivers/bus/arm-cci.o
> drivers/bus/arm-cci.c: In function ‘cci_probe’:
> drivers/bus/arm-cci.c:971:2: error: implicit declaration of function
> ‘of_find_matching_node’ [-Werror=implicit-function-declaration]
> cc1: some warnings being treated as errors
> make[2]: *** [drivers/bus/arm-cci.o] Error 1
> make[1]: *** [drivers/bus] Error 2
> make: *** [drivers] Error 2
>
> Signed-off-by: Manjunath Goudar <[email protected]>
> Cc: Grant Likely <[email protected]>
> Cc: Rob Herring <[email protected]>
> Cc: [email protected]
> Cc: [email protected]
> ---
> include/linux/of.h | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/include/linux/of.h b/include/linux/of.h
> index f95aee3..2c57aa2 100644
> --- a/include/linux/of.h
> +++ b/include/linux/of.h
> @@ -340,6 +340,13 @@ int of_device_is_stdout_path(struct device_node *dn);
>
> #else /* CONFIG_OF */
>
> +static inline struct device_node *of_find_matching_node(
> + struct device_node *from,
> + const struct of_device_id *matches)
> +{
> + return NULL;
> +}
> +
> static inline const char* of_node_full_name(struct device_node *np)
> {
> return "<no-node>";
> --
> 1.7.9.5
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> [email protected]
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel