Re: [PATCH v5 3/3] backlight: lm3630a: add firmware node support

[Brian Masney]

On Tue, Apr 23, 2019 at 10:31:41AM -0500, Dan Murphy wrote:
> On 4/23/19 9:01 AM, Brian Masney wrote:
> > On Tue, Apr 23, 2019 at 08:49:20AM -0500, Dan Murphy wrote:
> >>> +static int lm3630a_parse_led_sources(struct fwnode_handle *node,
> >>> +				     int default_led_sources)
> >>> +{
> >>> +	u32 sources[LM3630A_NUM_SINKS];
> >>> +	int ret, num_sources, i;
> >>> +
> >>> +	num_sources = fwnode_property_read_u32_array(node, "led-sources", NULL,
> >>> +						     0);
> >>> +	if (num_sources < 0)
> >>> +		return default_led_sources;
> >>> +	else if (num_sources > ARRAY_SIZE(sources))
> >>> +		return -EINVAL;
> >>> +
> >>> +	ret = fwnode_property_read_u32_array(node, "led-sources", sources,
> >>> +					     num_sources);
> >>> +	if (ret)
> >>> +		return ret;
> >>> +
> >>> +	for (i = 0; i < num_sources; i++) {
> >>> +		if (sources[i] < LM3630A_SINK_0 || sources[i] > LM3630A_SINK_1)
> >>> +			return -EINVAL;
> >>> +
> >>> +		ret |= BIT(sources[i]);
> >>> +	}
> >>> +
> >>> +	return ret;
> >>> +}
> >>> +
> >>> +static int lm3630a_parse_bank(struct lm3630a_platform_data *pdata,
> >>> +			      struct fwnode_handle *node, int *seen_led_sources)
> >>
> >> Why is seen_led_sources passed in here?
> >> It is initialized on the stack in lm3630a_parse_node but the variable is never referenced in that API.
> > 
> > It's to see all of the led-sources that are configured across all of the
> > banks. If it is just in lm3630a_parse_bank(), then it won't catch the
> > following invalid configuration:
> > 
> 
> Ok I see what it is for now.
> 
> Not sure why it is declared as a pointer though.

It's so that lm3630a_parse_bank() can update that value with the
led-sources that have been seen. Otherwise, the changes wouldn't make
their way back out to the outer function.

Brian