[alsa-devel] [PATCH v6] ASoC:Add support for cs42l73 codec

[Mark Brown]

On Wed, Nov 09, 2011 at 10:05:13AM -0600, Brian Austin wrote:
> Signed-off-by:Brian Austin <brian.austin at cirrus.com>
> Signed-off-by:Georgi Vlaev <joe at nucleusys.com>
> 
> This patch adds support for Cirrus Logic CS42L73
> low power stereo codec

Changelogs and signoffs are supposed to go in the other order.

> +static const char * const charge_pump_freq_text[] = {
> +	"0", "1", "2", "3", "4",
> +	"5", "6", "7", "8", "9",
> +	"10", "11", "12", "13", "14", "15" };

> +static const struct soc_enum charge_pump_enum =
> +	SOC_ENUM_SINGLE(CS42L73_CPFCHC, 4,
> +		ARRAY_SIZE(charge_pump_freq_text), charge_pump_freq_text);

Is there no meaningful text you can use here?  I'd expect there must be
as I've no idea how anyone should figure out what to set here, but if
not then making it an enum isn't really worthwhile, just use SOC_SINGLE.  

> +static int cs42l73_probe(struct snd_soc_codec *codec)
> +{
> +	int ret;
> +	unsigned int devid = 0;
> +	struct cs42l73_private *cs42l73 = snd_soc_codec_get_drvdata(codec);
> +
> +	ret = snd_soc_codec_set_cache_io(codec, 8, 8, SND_SOC_I2C);

It'd be better to move this to using SND_SOC_REGMAP and pushing the
cache down into the register map API with the basic device presence
verification done in the I2C probe but that's not essential yet.

> +	if (ret != 0) {
> +		printk(KERN_ERR "%s: can't add i2c driver\n", __func__);

pr_err().