Re: [PATCH] hwmon: ltc2947: Properly handle errors when looking for the external clock

[Guenter Roeck <linux@roeck-us.net>]

On Thu, Sep 23, 2021 at 10:11:13PM +0200, Uwe Kleine-König wrote:
> The return value of devm_clk_get should in general be propagated to
> upper layer. In this case the clk is optional, use the appropriate
> wrapper instead of interpreting all errors as "The optional clk is not
> available".
> 
> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>

Applied.

Thanks,
Guenter

> ---
>  drivers/hwmon/ltc2947-core.c | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
> 
> 
> base-commit: 7d2a07b769330c34b4deabeed939325c77a7ec2f
> 
> diff --git a/drivers/hwmon/ltc2947-core.c b/drivers/hwmon/ltc2947-core.c
> index bb3f7749a0b0..5423466de697 100644
> --- a/drivers/hwmon/ltc2947-core.c
> +++ b/drivers/hwmon/ltc2947-core.c
> @@ -989,8 +989,12 @@ static int ltc2947_setup(struct ltc2947_data *st)
>  		return ret;
>  
>  	/* check external clock presence */
> -	extclk = devm_clk_get(st->dev, NULL);
> -	if (!IS_ERR(extclk)) {
> +	extclk = devm_clk_get_optional(st->dev, NULL);
> +	if (IS_ERR(extclk))
> +		return dev_err_probe(st->dev, PTR_ERR(extclk),
> +				     "Failed to get external clock\n");
> +
> +	if (extclk) {
>  		unsigned long rate_hz;
>  		u8 pre = 0, div, tbctl;
>  		u64 aux;