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

[Jacek Anaszewski]

Hi Dan,

On 05/14/2018 09:40 PM, Dan Murphy wrote:
> Jacek
>
> On 05/11/2018 06:56 AM, Dan Murphy wrote:
> <snip>
>
> > > > +ÂÂÂ }
> > > > +
> > > > +ÂÂÂ if (led->strobe_node) {
> > > > +ÂÂÂÂÂÂÂ ret = of_property_read_string(led->strobe_node, "label", &name);
> > > > +ÂÂÂÂÂÂÂ if (!ret)
> > > > +ÂÂÂÂÂÂÂÂÂÂÂ snprintf(led->strobe, sizeof(led->strobe),
> > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ "%s:%s", led->strobe_node->name, name);
> > > > +ÂÂÂÂÂÂÂ else
> > > > +ÂÂÂÂÂÂÂÂÂÂÂ snprintf(led->strobe, sizeof(led->strobe),
> > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ "%s::strobe", led->strobe_node->name);
> > > > +
> > > > +ÂÂÂÂÂÂÂ ret = of_property_read_u32(led->strobe_node,
> > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ "flash-max-microamp",
> > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ &led->strobe_current_max);
> > > > +ÂÂÂÂÂÂÂ if (ret < 0) {
> > > > +ÂÂÂÂÂÂÂÂÂÂÂ led->strobe_current_max = LM3601X_MIN_STROBE_I_MA;
> > > > +ÂÂÂÂÂÂÂÂÂÂÂ dev_warn(&led->client->dev,
> > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ "flash-max-microamp DT property missing\n");
> > > > +ÂÂÂÂÂÂÂ }
> > > > +
> > > > +ÂÂÂÂÂÂÂ ret = of_property_read_u32(led->strobe_node,
> > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ "flash-max-timeout-us",
> > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ &led->max_strobe_timeout);
> > > > +ÂÂÂÂÂÂÂ if (ret < 0) {
> > > > +ÂÂÂÂÂÂÂÂÂÂÂ led->max_strobe_timeout = strobe_timeouts[0].reg_val;
> > > > +ÂÂÂÂÂÂÂÂÂÂÂ dev_warn(&led->client->dev,
> > > > +ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ "flash-max-timeout-us DT property missing\n");
> > > > +ÂÂÂÂÂÂÂ }
> > >
> > > 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.

--
Best regards,
Jacek Anaszewski