Re: [PATCH v5 2/2] leds: lm3601x: Introduce the lm3601x LED driver

[Dan Murphy <dmurphy@ti.com>]

On 05/14/2018 03:05 PM, Jacek Anaszewski wrote:
> Hi Dan,
> 
> On 05/14/2018 09:40 PM, Dan Murphy wrote:
>> Jacek
>>
>> On 05/11/2018 06:56 AM, Dan Murphy wrote:
>> <snip>
>>
>>>>> +    }
>>>>> +
>>>>> +    if (led->strobe_node) {
>>>>> +        ret = of_property_read_string(led->strobe_node, "label", &name);
>>>>> +        if (!ret)
>>>>> +            snprintf(led->strobe, sizeof(led->strobe),
>>>>> +                "%s:%s", led->strobe_node->name, name);
>>>>> +        else
>>>>> +            snprintf(led->strobe, sizeof(led->strobe),
>>>>> +                "%s::strobe", led->strobe_node->name);
>>>>> +
>>>>> +        ret = of_property_read_u32(led->strobe_node,
>>>>> +                    "flash-max-microamp",
>>>>> +                    &led->strobe_current_max);
>>>>> +        if (ret < 0) {
>>>>> +            led->strobe_current_max = LM3601X_MIN_STROBE_I_MA;
>>>>> +            dev_warn(&led->client->dev,
>>>>> +                 "flash-max-microamp DT property missing\n");
>>>>> +        }
>>>>> +
>>>>> +        ret = of_property_read_u32(led->strobe_node,
>>>>> +                    "flash-max-timeout-us",
>>>>> +                    &led->max_strobe_timeout);
>>>>> +        if (ret < 0) {
>>>>> +            led->max_strobe_timeout = strobe_timeouts[0].reg_val;
>>>>> +            dev_warn(&led->client->dev,
>>>>> +                 "flash-max-timeout-us DT property missing\n");
>>>>> +        }
>>>>
>>>> Common LED bindings state that flash-max-microamp and
>>>> flash-max-timeout-us properties are mandatory.
>>>
>>> OK.
>>
>> OK I looked at the max776973 driver and well if the flash-max-microamp and
>> flash-max-timeout-us nodes are missing it sets a default value for each if the
>> node is not present.
> 
> Ah, yes, this driver was being introduced as the first LED flash class driver and we were being iteratively adjusting LED common bindings
> according to the new findings, so some details could have been left
> out of sync.
> 
>> So should we remove this code from the Max77693 driver too and fail probe as being asked
>> in this driver?
> 
> Yes, that would match what the bindings require.

Did you want me to remove it and submit?  I don't have a board to verify but I can definitely test out the probe and parse dt functionality.
Don't need HW for that.

Dan

> 


-- 
------------------
Dan Murphy