[PATCH 2/2] regulator: dbx500: add missing header dependencies

[PATCH 2/2] regulator: dbx500: add missing header dependencies

[Baoyou Xie]

We get 1 warning when building kernel with W=1:
drivers/regulator/dbx500-prcmu.c:158:27: warning: no previous prototype for 'dbx500_regulator_testcase' [-Wmissing-prototypes]

In fact, this function is not declared in any file, but should be
declared in a header file. thus can be recognized in other file.

So this patch adds the declaration into
drivers/regulator/dbx500-prcmu.h.

Signed-off-by: Baoyou Xie <baoyou.xie@linaro.org>
---
 drivers/regulator/dbx500-prcmu.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/regulator/dbx500-prcmu.h b/drivers/regulator/dbx500-prcmu.h
index c8e51ac..644cb9e 100644
--- a/drivers/regulator/dbx500-prcmu.h
+++ b/drivers/regulator/dbx500-prcmu.h
@@ -35,7 +35,8 @@ struct dbx500_regulator_info {
 
 void power_state_active_enable(void);
 int power_state_active_disable(void);
-
+int dbx500_regulator_testcase(struct dbx500_regulator_info *regulator_info,
+			      int num_regulators);
 
 #ifdef CONFIG_REGULATOR_DEBUG
 int ux500_regulator_debug_init(struct platform_device *pdev,
-- 
2.7.4

Re: [PATCH 2/2] regulator: dbx500: add missing header dependencies

[Arnd Bergmann]

On Friday, September 23, 2016 8:50:10 PM CEST Baoyou Xie wrote:
> We get 1 warning when building kernel with W=1:
> drivers/regulator/dbx500-prcmu.c:158:27: warning: no previous prototype for 'dbx500_regulator_testcase' [-Wmissing-prototypes]
> 
> In fact, this function is not declared in any file, but should be
> declared in a header file. thus can be recognized in other file.
> 
> So this patch adds the declaration into
> drivers/regulator/dbx500-prcmu.h.
> 
> Signed-off-by: Baoyou Xie <baoyou.xie@linaro.org>
> 

I'm always suspicious of a patch that only adds a declaration,
as that tends to mean that no other caller ever existed. In this
case, there is an empty "__weak" function definition and it
is called from the driver itself, but there is no other non-weak
definition. If there was another definition, it would probably
be wrong, so I'd just remove the function completely and not
call it.

	Arnd