Re: [PATCH] extcon: Fix return value in extcon_register_interest()

[Chanwoo Choi <cw00.choi@samsung.com>]

On 09/25/2012 03:58 PM, Sachin Kamat wrote:

> Return the value obtained from extcon_find_cable_index()
> instead of -ENODEV.
> 
> Fixes the following smatch info:
> drivers/extcon/extcon-class.c:478 extcon_register_interest() info:
> why not propagate 'obj->cable_index' from extcon_find_cable_index()
> instead of -19?
> 
> Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
> ---
>  drivers/extcon/extcon-class.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/extcon/extcon-class.c b/drivers/extcon/extcon-class.c
> index 946a318..e996800 100644
> --- a/drivers/extcon/extcon-class.c
> +++ b/drivers/extcon/extcon-class.c
> @@ -475,7 +475,7 @@ int extcon_register_interest(struct extcon_specific_cable_nb *obj,
>  
>  	obj->cable_index = extcon_find_cable_index(obj->edev, cable_name);
>  	if (obj->cable_index < 0)
> -		return -ENODEV;
> +		return obj->cable_index;
>  
>  	obj->user_nb = nb;
>  


I agree.
But, if extcon_register_interest() return directly 'obj-cable_index' value
when extcon_find_cable_index() return -EINVAL, it would spoil readability
of extcon_register_interest() function. So, I think we can make it as
following
patch  a little better.

---
diff --git a/drivers/extcon/extcon-class.c b/drivers/extcon/extcon-class.c
index 936580b..078e6a5 100644
--- a/drivers/extcon/extcon-class.c
+++ b/drivers/extcon/extcon-class.c
@@ -468,7 +468,7 @@ int extcon_register_interest(struct
extcon_specific_cable_nb *obj,

        obj->cable_index = extcon_find_cable_index(obj->edev, cable_name);
        if (obj->cable_index < 0)
-               return -ENODEV;
+               return -EINVAL;

        obj->user_nb = nb;



Thanks,
Chanwoo Choi