Re: [PATCH] interconnect: qcom: icc-rpm: Initialize ret variable to fix smatch error

[Stephan Gerhold]

On Mon, Jul 17, 2023 at 10:34:29AM +0300, djakov@xxxxxxxxxx wrote:
> From: Georgi Djakov <djakov@xxxxxxxxxx>
> 
> Fix the following smatch error:
> drivers/interconnect/qcom/icc-rpm.c:243 qcom_icc_rpm_set() error: uninitialized symbol 'ret'.
> 
> Fixes: 32846c4a8f2a ("interconnect: qcom: icc-rpm: Set bandwidth on both contexts")
> Signed-off-by: Georgi Djakov <djakov@xxxxxxxxxx>

Thanks for spotting this!

> ---
>  drivers/interconnect/qcom/icc-rpm.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/interconnect/qcom/icc-rpm.c b/drivers/interconnect/qcom/icc-rpm.c
> index 612390b9eb18..6718cc648d75 100644
> --- a/drivers/interconnect/qcom/icc-rpm.c
> +++ b/drivers/interconnect/qcom/icc-rpm.c
> @@ -206,7 +206,7 @@ static int qcom_icc_qos_set(struct icc_node *node)
>  
>  static int qcom_icc_rpm_set(struct qcom_icc_node *qn, u64 *bw)
>  {
> -	int ret, rpm_ctx = 0;
> +	int ret = 0, rpm_ctx = 0;
>  	u64 bw_bps;
>  
>  	if (qn->qos.ap_owned)

I think it would be a bit clearer to change the "return ret;" at the end
of the function to "return 0;". The ret variable is only used inside the
for loop and always handled there, so we always want to return 0 when
reaching the end of the function.

Thanks,
Stephan