Re: [PATCH] cpuidle: Fix reference count on CPU node

From: Miquel Sabaté Solà
Date: Tue Oct 01 2024 - 16:13:47 EST


On dt., de set. 17 2024, Miquel Sabaté Solà wrote:

> For the qcom-spm driver, an early return was not calling the proper
> of_node_put call for a previously acquired device node.
>
> Signed-off-by: Miquel Sabaté Solà <mikisabate@xxxxxxxxx>
> ---
> drivers/cpuidle/cpuidle-qcom-spm.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/cpuidle/cpuidle-qcom-spm.c b/drivers/cpuidle/cpuidle-qcom-spm.c
> index 1fc9968eae19..d3608f47d02b 100644
> --- a/drivers/cpuidle/cpuidle-qcom-spm.c
> +++ b/drivers/cpuidle/cpuidle-qcom-spm.c
> @@ -96,8 +96,10 @@ static int spm_cpuidle_register(struct device *cpuidle_dev, int cpu)
> return -ENODEV;
>
> saw_node = of_parse_phandle(cpu_node, "qcom,saw", 0);
> - if (!saw_node)
> + if (!saw_node) {
> + of_node_put(cpu_node);
> return -ENODEV;
> + }
>
> pdev = of_find_device_by_node(saw_node);
> of_node_put(saw_node);

Gently ping for a fix in the same spirit as [1].

Could you take a look whenever you have some time?

Thanks!
Miquel

[1] https://lore.kernel.org/all/20240917134246.584026-1-mikisabate@xxxxxxxxx/

Attachment: signature.asc
Description: PGP signature