Looks good to me, yet two small comments inline.
Please add this to this patch in the next version:
Acked-by: Nicolin Chen <nicoleots...@gmail.com>
On Wed, Jul 03, 2019 at 02:42:04PM +0800, shengjiu.w...@nxp.com wrote:
> +static int fsl_esai_register_restore(struct fsl_esai *esai_priv)
> +{
> + int ret;
> + /* FIFO reset for safety */
> + regmap_update_bits(esai_priv->regmap, REG_ESAI_TFCR,
Checkpatch script would probably warn this. Usually we add a blank
line after variable declarations.
> @@ -866,22 +935,9 @@ static int fsl_esai_probe(struct platform_device *pdev)
>
> dev_set_drvdata(&pdev->dev, esai_priv);
>
> - /* Reset ESAI unit */
> - ret = regmap_write(esai_priv->regmap, REG_ESAI_ECR, ESAI_ECR_ERST);
> - if (ret) {
> - dev_err(&pdev->dev, "failed to reset ESAI: %d\n", ret);
> + ret = fsl_esai_init(esai_priv);
Could we rename this function to fsl_easi_hw_init() or something
clear like fsl_esai_register_init? fsl_easi_init() feels like a
driver init() function to me.
Thank you
Nicolin
