[PATCH] drivers: soc: atmel: Fix refcount leak in atmel_soc_device_init

[PATCH] drivers: soc: atmel: Fix refcount leak in atmel_soc_device_init

[Miaoqian Lin]

of_find_node_by_path() returns a node pointer with refcount incremented,
we should use of_node_put() on it when not need anymore.
Add missing of_node_put() to avoid refcount leak.

Fixes: 960ddf70cc11 ("drivers: soc: atmel: Avoid calling at91_soc_init on non AT91 SoCs")
Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
---
 drivers/soc/atmel/soc.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/soc/atmel/soc.c b/drivers/soc/atmel/soc.c
index b2d365ae0282..46dfa24c52fb 100644
--- a/drivers/soc/atmel/soc.c
+++ b/drivers/soc/atmel/soc.c
@@ -369,10 +369,13 @@ static int __init atmel_soc_device_init(void)
 {
 	struct device_node *np = of_find_node_by_path("/");
 
-	if (!of_match_node(at91_soc_allowed_list, np))
+	if (!of_match_node(at91_soc_allowed_list, np)) {
+		of_node_put(np);
 		return 0;
+	}
 
 	at91_soc_init(socs);
+	of_node_put(np);
 
 	return 0;
 }
-- 
2.25.1

Re: [PATCH] drivers: soc: atmel: Fix refcount leak in atmel_soc_device_init

[Claudiu.Beznea]

On 05.06.2022 11:40, Miaoqian Lin wrote:
> EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> 
> of_find_node_by_path() returns a node pointer with refcount incremented,
> we should use of_node_put() on it when not need anymore.
> Add missing of_node_put() to avoid refcount leak.
> 
> Fixes: 960ddf70cc11 ("drivers: soc: atmel: Avoid calling at91_soc_init on non AT91 SoCs")
> Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
> ---
>  drivers/soc/atmel/soc.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/soc/atmel/soc.c b/drivers/soc/atmel/soc.c
> index b2d365ae0282..46dfa24c52fb 100644
> --- a/drivers/soc/atmel/soc.c
> +++ b/drivers/soc/atmel/soc.c
> @@ -369,10 +369,13 @@ static int __init atmel_soc_device_init(void)
>  {
>         struct device_node *np = of_find_node_by_path("/");
> 

What about having it like this:
	const struct of_device_id *of_id;

	of_id = of_match_node(at91_soc_allowed_list, np);
	of_node_put(np);
	if (!of_id)
		return 0;

And everything else the same?

> -       if (!of_match_node(at91_soc_allowed_list, np))
> +       if (!of_match_node(at91_soc_allowed_list, np)) {
> +               of_node_put(np);
>                 return 0;
> +       }
> 
>         at91_soc_init(socs);
> +       of_node_put(np);
> 
>         return 0;
>  }
> --
> 2.25.1
> 

Re: [PATCH] drivers: soc: atmel: Fix refcount leak in atmel_soc_device_init

[Sudeep Holla]

On Mon, Jun 06, 2022 at 09:00:52AM +0000, Claudiu.Beznea@microchip.com wrote:
> On 05.06.2022 11:40, Miaoqian Lin wrote:
> > EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
> > 
> > of_find_node_by_path() returns a node pointer with refcount incremented,
> > we should use of_node_put() on it when not need anymore.
> > Add missing of_node_put() to avoid refcount leak.
> > 
> > Fixes: 960ddf70cc11 ("drivers: soc: atmel: Avoid calling at91_soc_init on non AT91 SoCs")
> > Signed-off-by: Miaoqian Lin <linmq006@gmail.com>
> > ---
> >  drivers/soc/atmel/soc.c | 5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/soc/atmel/soc.c b/drivers/soc/atmel/soc.c
> > index b2d365ae0282..46dfa24c52fb 100644
> > --- a/drivers/soc/atmel/soc.c
> > +++ b/drivers/soc/atmel/soc.c
> > @@ -369,10 +369,13 @@ static int __init atmel_soc_device_init(void)
> >  {
> >         struct device_node *np = of_find_node_by_path("/");
> > 
> 
> What about having it like this:
> 	const struct of_device_id *of_id;
> 
> 	of_id = of_match_node(at91_soc_allowed_list, np);
> 	of_node_put(np);

+1, I was about to make similar suggestion in couple of other patches making
similar changes.

-- 
Regards,
Sudeep