In rockchip_suspend_init(), of_find_matching_node_and_match() will
return a node pointer with refcount incremented. We should use
of_node_put() in fail path or when it is not used anymore.

Signed-off-by: Liang He <windhl@126.com>
---
 changelog:

 v2: (1) use real name for Sob (2) use goto-label patch style
 v1: add of_node_put() for each return

 arch/arm/mach-rockchip/pm.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/arch/arm/mach-rockchip/pm.c b/arch/arm/mach-rockchip/pm.c
index 87389d9456b9..bb46ac4f65d4 100644
--- a/arch/arm/mach-rockchip/pm.c
+++ b/arch/arm/mach-rockchip/pm.c
@@ -311,7 +311,7 @@ void __init rockchip_suspend_init(void)
 					     &match);
 	if (!match) {
 		pr_err("Failed to find PMU node\n");
-		return;
+		goto out_put;
 	}
 	pm_data = (struct rockchip_pm_data *) match->data;
 
@@ -320,9 +320,13 @@ void __init rockchip_suspend_init(void)
 
 		if (ret) {
 			pr_err("%s: matches init error %d\n", __func__, ret);
-			return;
+			goto out_put;
 		}
 	}
 
 	suspend_set_ops(pm_data->ops);
+	
+out_put:
+	of_node_put(np);
+	
 }
-- 
2.25.1

On 16/06/2022 04:17, Liang He wrote:
> In rockchip_suspend_init(), of_find_matching_node_and_match() will
> return a node pointer with refcount incremented. We should use
> of_node_put() in fail path or when it is not used anymore.
> 
> Signed-off-by: Liang He <windhl@126.com>
> ---
>  changelog:
> 
>  v2: (1) use real name for Sob (2) use goto-label patch style
>  v1: add of_node_put() for each return
> 

Before applying the patch please check it carefully. Previous evidence
[1][2] suggests that not it was not even compiled.



[1] https://lore.kernel.org/all/202206221602.odN70SHs-lkp@intel.com/

[2]
https://lore.kernel.org/all/16f9a971.44e5.1817068ee3c.Coremail.windhl@126.com/


Best regards,
Krzysztof

Am Mittwoch, 22. Juni 2022, 11:00:11 CEST schrieb Krzysztof Kozlowski:
> On 16/06/2022 04:17, Liang He wrote:
> > In rockchip_suspend_init(), of_find_matching_node_and_match() will
> > return a node pointer with refcount incremented. We should use
> > of_node_put() in fail path or when it is not used anymore.
> > 
> > Signed-off-by: Liang He <windhl@126.com>
> > ---
> >  changelog:
> > 
> >  v2: (1) use real name for Sob (2) use goto-label patch style
> >  v1: add of_node_put() for each return
> > 
> 
> Before applying the patch please check it carefully. Previous evidence
> [1][2] suggests that not it was not even compiled.

thanks noticing possible issues and for the heads up :-) .

I re-examined the patch and at least in this case the of_node_put addition
looks correct - the original code was missing this and not putting the node
it got in the error case. And the change also compiles :-) .

Heiko

> [1] https://lore.kernel.org/all/202206221602.odN70SHs-lkp@intel.com/
> 
> [2]
> https://lore.kernel.org/all/16f9a971.44e5.1817068ee3c.Coremail.windhl@126.com/
> 
> 
> Best regards,
> Krzysztof
>

On Thu, 16 Jun 2022 10:17:13 +0800, Liang He wrote:
> In rockchip_suspend_init(), of_find_matching_node_and_match() will
> return a node pointer with refcount incremented. We should use
> of_node_put() in fail path or when it is not used anymore.

Applied, thanks!

[1/1] arm: mach-rockchip: (pm) Add missing of_node_put()
      commit f4470dbfb5ff92804650bc71d115c3f150d430f6

There was an unneeded tab in the blank line above out_put
and a completely unneeded blank line added at the end of
the function. I dropped both.

Best regards,
-- 
Heiko Stuebner <heiko@sntech.de>

At 2022-06-18 01:10:28, "Heiko Stuebner" <heiko@sntech.de> wrote:
>On Thu, 16 Jun 2022 10:17:13 +0800, Liang He wrote:
>> In rockchip_suspend_init(), of_find_matching_node_and_match() will
>> return a node pointer with refcount incremented. We should use
>> of_node_put() in fail path or when it is not used anymore.
>
>Applied, thanks!
>
>[1/1] arm: mach-rockchip: (pm) Add missing of_node_put()
>      commit f4470dbfb5ff92804650bc71d115c3f150d430f6
>
>There was an unneeded tab in the blank line above out_put
>and a completely unneeded blank line added at the end of
>the function. I dropped both.

Thanks very much for your effort.

I am learning the coding style for kernel code.

>
>Best regards,
>-- 
>Heiko Stuebner <heiko@sntech.de>