Make use of the __free() cleanup handler to automatically free nodes
when they get out of scope. Only the probe function is affected by this
modification.
Given that this mechanism requires the node to be initialized, its
initialization and the value check have been moved to the top of the
function.
After removing uses of of_node_put(), the jump to out_put_np is no
longer necessary.
Suggested-by: Julia Lawall <[email protected]>
Signed-off-by: Javier Carrasco <[email protected]>
---
This patch is a proof of concept to remove uses of of_node_put() in
cpufreq, which can be replaced with the clenaup handler introduced with
54da6a092431 ("locking: Introduce __cleanup() based infrastructure").
This change provides a scope-based cleanup mechanism to avoid potential
memory leaks that can appear if of_node_put() is not used correctly.
The patch is based on the latest linux-next tag (next-20240405).
---
drivers/cpufreq/tegra124-cpufreq.c | 19 ++++++-------------
1 file changed, 6 insertions(+), 13 deletions(-)
diff --git a/drivers/cpufreq/tegra124-cpufreq.c b/drivers/cpufreq/tegra124-cpufreq.c
index aae951d4e77c..514146d98bca 100644
--- a/drivers/cpufreq/tegra124-cpufreq.c
+++ b/drivers/cpufreq/tegra124-cpufreq.c
@@ -52,12 +52,15 @@ static int tegra124_cpu_switch_to_dfll(struct tegra124_cpufreq_priv *priv)
static int tegra124_cpufreq_probe(struct platform_device *pdev)
{
+ struct device_node *np __free(device_node) = of_cpu_device_node_get(0);
struct tegra124_cpufreq_priv *priv;
- struct device_node *np;
struct device *cpu_dev;
struct platform_device_info cpufreq_dt_devinfo = {};
int ret;
+ if (!np)
+ return -ENODEV;
+
priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL);
if (!priv)
return -ENOMEM;
@@ -66,15 +69,9 @@ static int tegra124_cpufreq_probe(struct platform_device *pdev)
if (!cpu_dev)
return -ENODEV;
- np = of_cpu_device_node_get(0);
- if (!np)
- return -ENODEV;
-
priv->cpu_clk = of_clk_get_by_name(np, "cpu_g");
- if (IS_ERR(priv->cpu_clk)) {
- ret = PTR_ERR(priv->cpu_clk);
- goto out_put_np;
- }
+ if (IS_ERR(priv->cpu_clk))
+ return PTR_ERR(priv->cpu_clk);
priv->dfll_clk = of_clk_get_by_name(np, "dfll");
if (IS_ERR(priv->dfll_clk)) {
@@ -110,8 +107,6 @@ static int tegra124_cpufreq_probe(struct platform_device *pdev)
platform_set_drvdata(pdev, priv);
- of_node_put(np);
-
return 0;
out_put_pllp_clk:
@@ -122,8 +117,6 @@ static int tegra124_cpufreq_probe(struct platform_device *pdev)
clk_put(priv->dfll_clk);
out_put_cpu_clk:
clk_put(priv->cpu_clk);
-out_put_np:
- of_node_put(np);
return ret;
}
---
base-commit: 8568bb2ccc278f344e6ac44af6ed010a90aa88dc
change-id: 20240407-cpufreq_of_node_put-0d1972a33561
Best regards,
--
Javier Carrasco <[email protected]>
On 07-04-24, 22:15, Javier Carrasco wrote:
> Make use of the __free() cleanup handler to automatically free nodes
> when they get out of scope. Only the probe function is affected by this
> modification.
>
> Given that this mechanism requires the node to be initialized, its
> initialization and the value check have been moved to the top of the
> function.
>
> After removing uses of of_node_put(), the jump to out_put_np is no
> longer necessary.
>
> Suggested-by: Julia Lawall <[email protected]>
> Signed-off-by: Javier Carrasco <[email protected]>
> ---
> This patch is a proof of concept to remove uses of of_node_put() in
> cpufreq, which can be replaced with the clenaup handler introduced with
> 54da6a092431 ("locking: Introduce __cleanup() based infrastructure").
>
> This change provides a scope-based cleanup mechanism to avoid potential
> memory leaks that can appear if of_node_put() is not used correctly.
>
> The patch is based on the latest linux-next tag (next-20240405).
> ---
> drivers/cpufreq/tegra124-cpufreq.c | 19 ++++++-------------
> 1 file changed, 6 insertions(+), 13 deletions(-)
Applied. Thanks.
--
viresh