Re: [PATCH] ASoC: cs42888: Add codec driver support

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




On Mon, Feb 24, 2014 at 02:55:29PM +0800, Nicolin Chen wrote:
> This patch adds support for the Cirrus Logic CS42888 Audio CODEC that
> has four 24-bit A/D and eight 24-bit D/A converters.

Looks generally good, some fairly small nits below.

> [ CS42888 supports both I2C and SPI control ports. As initial patch,
>   this patch only adds the support for I2C. ]

>  5 files changed, 795 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/sound/cs42888.txt
>  create mode 100644 sound/soc/codecs/cs42888.c
>  create mode 100644 sound/soc/codecs/cs42888.h

Given that we're starting to split out separate bus drivers for the I2C
and SPI CODECs (look at the recent submissions from Lars-Peter) it'd be
good to start this off with a separate bus driver for I2C even if the
SPI one is still to be done - that way the Kconfig stuff for machine
drivers is all in place and doesn't need updating.

> +  - clocks : phandle to the clock source for MCLK
> +
> +  - clock-names : must contain "mclk".

These should really be lists though there's only one documented element
so it's purely a documentation update.

> +	/* Disable auto-mute */
> +	regmap_update_bits(cs42888->regmap, CS42888_TXCTL,
> +			   CS42888_TXCTL_AMUTE | CS42888_TXCTL_DAC_SZC_MASK,
> +			   CS42888_TXCTL_DAC_SZC_SR);

Does this interfere with the manual mute controls or is it a separate
thing?  If it plays nicely with the manual controls it's probably better
to leave it enabled since it improves performance in some benchmarks
(that's why hardware tends to have the feature).

> +	/*
> +	 * We haven't marked the chip revision as volatile due to
> +	 * sharing a register with the right input volume; explicitly
> +	 * bypass the cache to read it.
> +	 */
> +	regcache_cache_bypass(cs42888->regmap, true);

The other option here is to just not provide a default so that the first
time it's read it goes to hardware.  It doesn't make much difference
either way though.

> +static int cs42888_i2c_remove(struct i2c_client *i2c_client)
> +{
> +	snd_soc_unregister_codec(&i2c_client->dev);
> +	return 0;
> +}

The driver ought to disable runtime PM, the clock and the regulators here.

> +	/*
> +	 * In case the device was put to hard reset during sleep,
> +	 * we need to wait 500ns here before any I2C communication
> +	 */
> +	mdelay(5);

Do we need 500ns or 5ms?  

> +	regcache_sync(cs42888->regmap);

Should really check the return value here.

> +	if (!IS_ERR(cs42888->clk))
> +		clk_disable_unprepare(cs42888->clk);

Does the device work without MCLK?

Attachment: signature.asc
Description: Digital signature


[Index of Archives]     [Device Tree Compilter]     [Device Tree Spec]     [Linux Driver Backports]     [Video for Linux]     [Linux USB Devel]     [Linux PCI Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Yosemite Backpacking]
  Powered by Linux