Re: [PATCH v2] counter: ti-eqep: fix runtime PM leak in probe error path

Re: [PATCH v2] counter: ti-eqep: fix runtime PM leak in probe error path

[David Lechner]

On 4/17/26 11:08 AM, Felix Gu wrote:
> In ti_eqep_probe(), if devm_clk_get_enabled() fails, the function
> returns without cleaning up the runtime PM state.
> 
> Switch to devm functions to fix the problem and remove unneeded
> remove() callback.
> 
> Fixes: 0cf81c73e4c6 ("counter: ti-eqep: enable clock at probe")
> Signed-off-by: Felix Gu <ustc.gu@gmail.com>
> ---
> Changes in v2:
> - Switch to devm functions to fix David's comment.
> - Link to v1: https://lore.kernel.org/r/20260417-ti-eqep-v1-1-a7db4642d9d0@gmail.com
> ---
>  drivers/counter/ti-eqep.c | 33 +++++++++++++--------------------
>  1 file changed, 13 insertions(+), 20 deletions(-)
> 
> diff --git a/drivers/counter/ti-eqep.c b/drivers/counter/ti-eqep.c
> index d21c157e531a..08dc6d0ad5e4 100644
> --- a/drivers/counter/ti-eqep.c
> +++ b/drivers/counter/ti-eqep.c
> @@ -492,6 +492,11 @@ static const struct regmap_config ti_eqep_regmap16_config = {
>  	.max_register = QCPRDLAT,
>  };
>  
> +static void pm_put_sync(void *data)

This should have ti_eqep_ prefix in function name.

> +{
> +	pm_runtime_put_sync(data);
> +}
> +
>  static int ti_eqep_probe(struct platform_device *pdev)
>  {
>  	struct device *dev = &pdev->dev;
> @@ -544,31 +549,20 @@ static int ti_eqep_probe(struct platform_device *pdev)
>  	 * parent PWMSS bus driver. On AM17xx, this comes from the PSC power
>  	 * domain.
>  	 */
> -	pm_runtime_enable(dev);
> +	err = devm_pm_runtime_enable(dev);
> +	if (err)
> +		return err;
> +
>  	pm_runtime_get_sync(dev);
> +	err = devm_add_action_or_reset(dev, pm_put_sync, dev);

I have no idea if this is OK or not. It seems like devm_pm_runtime_get_sync()
would exist already if it was OK. So that makes me think it is not OK.

I added PM mailing list to see if anyone can advise.

> +	if (err)
> +		return err;
>  
>  	clk = devm_clk_get_enabled(dev, NULL);
>  	if (IS_ERR(clk))
>  		return dev_err_probe(dev, PTR_ERR(clk), "failed to enable clock\n");
>  
> -	err = counter_add(counter);
> -	if (err < 0) {
> -		pm_runtime_put_sync(dev);
> -		pm_runtime_disable(dev);
> -		return err;
> -	}
> -
> -	return 0;
> -}
> -
> -static void ti_eqep_remove(struct platform_device *pdev)
> -{
> -	struct counter_device *counter = platform_get_drvdata(pdev);
> -	struct device *dev = &pdev->dev;
> -
> -	counter_unregister(counter);
> -	pm_runtime_put_sync(dev);
> -	pm_runtime_disable(dev);
> +	return devm_counter_add(dev, counter);

Using devm_counter_add() is not part of the fix, so should be in
a separate patch.

>  }
>  
>  static const struct of_device_id ti_eqep_of_match[] = {
> @@ -580,7 +574,6 @@ MODULE_DEVICE_TABLE(of, ti_eqep_of_match);
>  
>  static struct platform_driver ti_eqep_driver = {
>  	.probe = ti_eqep_probe,
> -	.remove = ti_eqep_remove,
>  	.driver = {
>  		.name = "ti-eqep-cnt",
>  		.of_match_table = ti_eqep_of_match,
> 
> ---
> base-commit: 452c3b1ea875276105ac90ba474f72b4cd9b77a2
> change-id: 20260417-ti-eqep-8efb9cc713ea
> 
> Best regards,