Re: [PATCH 1/2] ASoC: ssm2602: add device tree bindings

[Lars-Peter Clausen <lars@metafoo.de>]

On 09/29/2014 04:34 PM, Stefan Kristiansson wrote:
> On Mon, Sep 29, 2014 at 10:50:24AM +0200, Lars-Peter Clausen wrote:
>> On 09/28/2014 09:57 AM, Stefan Kristiansson wrote:
>>> Allow the ssm2602/ssm2603/ssm2604 codec driver to be
>>> instantiated from the device tree.
>>>
>>> Also, add Kconfig prompts to allow manual selection of both the
>>> I2C and SPI configuration versions of the driver.
>>>
>>> Signed-off-by: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
>>
>> Looks mostly good, but you should Cc both the devicetree and the
>> ASoC maintainers.
>>
>
> Ok, I'll make sure to do that in v2.
>
>> [...]
>>> diff --git a/sound/soc/codecs/ssm2602-i2c.c b/sound/soc/codecs/ssm2602-i2c.c
>>> index abd63d5..0d9779d 100644
>>> --- a/sound/soc/codecs/ssm2602-i2c.c
>>> +++ b/sound/soc/codecs/ssm2602-i2c.c
>>> @@ -41,10 +41,19 @@ static const struct i2c_device_id ssm2602_i2c_id[] = {
>>>   };
>>>   MODULE_DEVICE_TABLE(i2c, ssm2602_i2c_id);
>>>
>>> +static const struct of_device_id ssm2602_of_match[] = {
>>> +	{ .compatible = "adi,ssm2602", },
>>> +	{ .compatible = "adi,ssm2603", },
>>> +	{ .compatible = "adi,ssm2604", },
>>
>> Since the driver is supposed to behave different depending on the
>> device you should set the driver data similar to like it is done in
>> the ssm2602_i2c_id table and add code in the probe function to
>> extract the driver data.
>>
>
> As far as I understand, that is already covered for, ssm2602_i2c_probe() get
> called with the i2c_device_id struct set according to the compatible string.
> I.e. if I set it to "adi,ssm2603" in the .dts, ssm2602_i2c_probe() is called
> with id->name = "ssm2603" and id->driver_data = 0 and if I set it to
> "adi,ssm2604" I get id->name = "ssm2603" and id->driver_data = 1.

Ok, looks like the I2C core extracts the string after the "," and than 
matches the I2C device id table against it.

>
> However, when looking into that, I noticed this in ssm2602_probe()
>
> 	 ssm2602->type = SSM2602;
>
> so the value from id->driver_data is never used.
> (it should be ssm2602->type = type;)

Yes, that's clearly a bug.

[...]