Re: [PATCH] clk:sophgo:clk-cv18xx-pll

[Chen Wang]

Pleas improve the patch title and just write "fix..." instead of listing 
the file name changed.

On 2024/11/13 6:58, Ragavendra wrote:
> Initializing the val variable of type u32 as it was not initialized.
Please add "Fixes" tag here, see 
https://www.kernel.org/doc/html/latest/process/submitting-patches.html.
> Signed-off-by: Ragavendra Nagraj <ragavendra.bn@xxxxxxxxx>
> ---
>   drivers/clk/sophgo/clk-cv18xx-pll.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/clk/sophgo/clk-cv18xx-pll.c b/drivers/clk/sophgo/clk-cv18xx-pll.c
> index 29e24098bf5f..04a0419cab4e 100644
> --- a/drivers/clk/sophgo/clk-cv18xx-pll.c
> +++ b/drivers/clk/sophgo/clk-cv18xx-pll.c
> @@ -87,7 +87,7 @@ static int ipll_find_rate(const struct cv1800_clk_pll_limit *limit,
>   
>   static int ipll_determine_rate(struct clk_hw *hw, struct clk_rate_request *req)
>   {
> -	u32 val;
> +	u32 val = 0;
>   	struct cv1800_clk_pll *pll = hw_to_cv1800_clk_pll(hw);
>   
>   	return ipll_find_rate(pll->pll_limit, req->best_parent_rate,

I looked at ipll_find_rate() and found that the parameters 
"rate"/"value" are both input and output parameters, which is really a 
place that needs to be treated with caution.

Seems this patch change is correct, otherwise the value of "detected" in 
the function will be random after calling ipll_find_rate here.  Actually 
I suggest adding some comments for ipll_find_rate() to emphasize it. 
Otherwise, it is easy to think that "rate" and "value" are only output 
parameters at first glance.

But I raised a question here: In ipll_find_rate(), do we really need to 
make "detected" depend on the value of external input? Inochi, can you 
please comment on this, due to you are the author of this code.

Reviewed-by: Chen Wang <unicorn_wang@xxxxxxxxxxx>

Thanks,

Chen