On 12/11/16 16:59, Brian Masney wrote:
> On Sat, Nov 12, 2016 at 04:36:37PM +0000, Jonathan Cameron wrote:
>> On 10/11/16 09:25, Brian Masney wrote:
>>> If channel 0 does not have any data, then the code sets the lux to zero.
>>> The corresponding comment says that the last value is returned. This
>>> updates the comment to correctly reflect what the code does.
>>>
>>> Signed-off-by: Brian Masney <masn...@onstation.org>
>> Better perhaps to just return an error, -EAGAIN perhaps?
>> I'm not sure why it would not give a value.
>
> This check is to avoid a division by zero. Here is the relevant code
> that wasn't shown in the diff:
>
> if (!ch0) {
> /* have no data, so return 0 */
> ret = 0;
> chip->als_cur_info.lux = 0;
> goto done;
> }
>
> /* calculate ratio */
> ratio = (ch1 << 15) / ch0;
>
> Channel 0 is sensitive to both infrared and visible light. In total
> darkness, the sensor should return 0. Correct me if I am wrong, but
> I believe that returning 0 here is more correct than -EAGAIN.
>
> Brian
>
Fair enough I hadn't understood that. Maybe expand the comment
to cover that?
