Re: [PATCH v2 2/4] iio: light: veml6030: fix channel type when pushing events

["Javier Carrasco" <javier.carrasco.cruz@gmail.com>]

On Thu May 14, 2026 at 9:02 AM +13, Andy Shevchenko wrote:
> On Thu, May 14, 2026 at 07:13:41AM +1300, Javier Carrasco wrote:
>> On Thu May 14, 2026 at 6:48 AM +13, Andy Shevchenko wrote:
>> > On Wed, May 13, 2026 at 05:49:42PM +1300, Javier Carrasco wrote:
>> >> The events are registered for IIO_LIGHT and not for IIO_INTENSITY.
>> >> Use the correct channel type.
>> >
>> >> This bug was introduced in the first version of the driver.
>> >
>> > Unneeded detail, if it's a bug, use Fixes tag.
>>
>> >> When at it, fix minor checkpatch code style warning (alignment).
>
> ...
>
>> >> -	iio_push_event(indio_dev, IIO_UNMOD_EVENT_CODE(IIO_INTENSITY,
>> >> -					0, IIO_EV_TYPE_THRESH, evtdir),
>> >> -					iio_get_time_ns(indio_dev));
>> >> +	iio_push_event(indio_dev, IIO_UNMOD_EVENT_CODE(IIO_LIGHT,
>> >> +						       0,
>> >> +						       IIO_EV_TYPE_THRESH,
>> >> +						       evtdir),
>> >> +			iio_get_time_ns(indio_dev));
>> >
>> > AFAICS the indentation is still broken. Why not doing like this:
>> >
>> > 	iio_push_event(indio_dev,
>> > 		       IIO_UNMOD_EVENT_CODE(IIO_LIGHT, 0, IIO_EV_TYPE_THRESH, evtdir),
>> > 		       iio_get_time_ns(indio_dev));
>>
>> Thank you for your feedback. According to checkpatch.pl, both variants
>> are fine. Mine takes into account the indentation within
>> IIO_UNMOD_EVENT_CODE(),
>
> And still have broken indentation with the last parameter. So it's not fine.
>

I am sorry to insist on this, but I beg to differ. The last paramter
(iio_get_time_ns()) is properly aligned as an argument of
iio_push_event() and not IIO_UNMOD_EVENT_CODE(). That is exactly my
point: with my indentation it is clear that iio_get_time_ns() is an
argument of iio_push_event() and not IIO_UNMOD_EVENT_CODE() because of
the alignment. Moreove, my proposed alignment (which again, is fine with
checkpatch --strict and the original one for example wasn't) is
consistent with many usages of iio_push_event() in existing drivers. I
just checked that there are dozens like mine, being the majority when it
comes to this kind of indentation.


>> and yours only accounts for the indentation for
>> the arguments of iio_push_event(). Moreover, your suggestion goes beyond
>> 80 characters and mine does not,
>
> When it's about readability the 80 characters is not a strict limit.
>
>> so I would prefer sticking to mine if
>> possible.
>
> I recommend to reconsider. Mine has no indentation issues, the only subtle
> "problem" is 86 character line. And looking at the result I find mine better
> to read (hence the exception may apply and we are fine with the length of
> the line).
>
>> As I said, it passes checkpatch --strict without warnings
>> in both cases.
>
>> I will send a new version adding the Fixes tag and removing the comment.
>
> Make it the first patch as the currently first one does not sound like a fix
> to me.

Ok, I will make this one the first patch of a smaller series with the
right Fixes tag added to it and removed from the other patch that
affects veml6030. I will split the new driver in smaller chunks and send
it as a dedicated series but continuing with the current versioning.

Best regards,
Javier