Re: [PATCH v3 1/2] media: i2c: imx412: fix sensor power-on timing

[Sakari Ailus]

Hi Wenmeng,

On Mon, Jan 19, 2026 at 06:15:01PM +0800, Wenmeng Liu wrote:
> Assert the reset GPIO at the start of power-on, enable regulators and the
> clock, then deassert reset. This avoids a mismatch where the first
> power-up (when the reset GPIO defaults deasserted) differs from subsequent
> cycles.
> 
> Signed-off-by: Wenmeng Liu <wenmeng.liu@xxxxxxxxxxxxxxxx>
> ---
>  drivers/media/i2c/imx412.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/media/i2c/imx412.c b/drivers/media/i2c/imx412.c
> index b3826f80354703b17b416dc233854da3f5736e38..ebc2fb706442365f541540bf4ce9d307d42e0061 100644
> --- a/drivers/media/i2c/imx412.c
> +++ b/drivers/media/i2c/imx412.c
> @@ -1022,6 +1022,8 @@ static int imx412_power_on(struct device *dev)
>  	struct imx412 *imx412 = to_imx412(sd);
>  	int ret;
>  
> +	gpiod_set_value_cansleep(imx412->reset_gpio, 1);

I'd assert reset in probe() so you can drop this line.

> +
>  	ret = regulator_bulk_enable(ARRAY_SIZE(imx412_supply_names),
>  				    imx412->supplies);
>  	if (ret < 0) {
> @@ -1029,20 +1031,19 @@ static int imx412_power_on(struct device *dev)
>  		return ret;
>  	}
>  
> -	gpiod_set_value_cansleep(imx412->reset_gpio, 0);
> -
>  	ret = clk_prepare_enable(imx412->inclk);
>  	if (ret) {
>  		dev_err(imx412->dev, "fail to enable inclk\n");
>  		goto error_reset;
>  	}
>  
> +	gpiod_set_value_cansleep(imx412->reset_gpio, 0);
> +
>  	usleep_range(1000, 1200);
>  
>  	return 0;
>  
>  error_reset:
> -	gpiod_set_value_cansleep(imx412->reset_gpio, 1);
>  	regulator_bulk_disable(ARRAY_SIZE(imx412_supply_names),
>  			       imx412->supplies);
>  
> 

-- 
Regards,

Sakari Ailus