Re: [PATCH v7 1/5] leds: populate the device's of_node when possible

[Jean-Jacques Hiblot <jjhiblot@ti.com>]

Hi Jacek,

On 24/09/2019 23:03, Jacek Anaszewski wrote:
> Hi Jean,
>
> Thank you for rebasing the set
>
> On 9/18/19 4:57 PM, Jean-Jacques Hiblot wrote:
>> If initialization data is available and its fwnode is actually a of_node,
>> store this information in the led device's structure. This will allow the
>> device to use or provide OF-based API such (devm_xxx).
>>
>> Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
>> ---
>>   drivers/leds/led-class.c | 5 ++++-
>>   1 file changed, 4 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/leds/led-class.c b/drivers/leds/led-class.c
>> index 647b1263c579..c2167b66b61f 100644
>> --- a/drivers/leds/led-class.c
>> +++ b/drivers/leds/led-class.c
>> @@ -276,8 +276,11 @@ int led_classdev_register_ext(struct device *parent,
>>   		mutex_unlock(&led_cdev->led_access);
>>   		return PTR_ERR(led_cdev->dev);
>>   	}
>> -	if (init_data && init_data->fwnode)
>> +	if (init_data && init_data->fwnode) {
>>   		led_cdev->dev->fwnode = init_data->fwnode;
>> +		if (is_of_node(init_data->fwnode))
>> +			led_cdev->dev->of_node = to_of_node(init_data->fwnode);
> It would be step backwards. You can do the conversion in the place of
> use i.e. in devm_led_get().

Could be done. But it would break another use case I have. I'm also 
working on the regulator support and for this one, of_node needs to be 
populated.

Is there a problem populating of_node if the LED is indeed described in 
the DT ?

JJ

>
>> +	}
>>   
>>   	if (ret)
>>   		dev_warn(parent, "Led %s renamed to %s due to name collision",
>>