Re: [resend, PATCH v1 1/1] platform/x86: int3472: Don't leak reference on error

[Rafael J. Wysocki]

On Tue, Oct 4, 2022 at 2:26 PM Andy Shevchenko
<andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
>
> The for_each_acpi_consumer_dev() takes a reference to the iterator
> and if we break a loop we must drop that reference. This usually
> happens when error handling is involved. However it's not the case
> for skl_int3472_fill_clk_pdata().
>
> Don't leak reference on error by dropping it properly.
>
> Fixes: 43cf36974d76 ("platform/x86: int3472: Support multiple clock consumers")
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
> ---
>
> resent to include Rafael and linux-acpi@ to the Cc list
>
>  drivers/platform/x86/intel/int3472/tps68470.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/platform/x86/intel/int3472/tps68470.c b/drivers/platform/x86/intel/int3472/tps68470.c
> index f83e9c393f31..5b8d1a9620a5 100644
> --- a/drivers/platform/x86/intel/int3472/tps68470.c
> +++ b/drivers/platform/x86/intel/int3472/tps68470.c
> @@ -128,15 +128,15 @@ skl_int3472_fill_clk_pdata(struct device *dev, struct tps68470_clk_platform_data
>         for_each_acpi_consumer_dev(adev, consumer) {
>                 sensor_name = devm_kasprintf(dev, GFP_KERNEL, I2C_DEV_NAME_FORMAT,
>                                              acpi_dev_name(consumer));
> -               if (!sensor_name)
> +               if (!sensor_name) {
> +                       acpi_dev_put(consumer);
>                         return -ENOMEM;
> +               }
>
>                 (*clk_pdata)->consumers[i].consumer_dev_name = sensor_name;
>                 i++;
>         }
>
> -       acpi_dev_put(consumer);
> -
>         return n_consumers;
>  }
>
> --

Applied, thanks!