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

[Matti Vaittinen <mazziesaccount@gmail.com>]

On 18/05/2026 18:22, Jonathan Cameron wrote:
> On Mon, 18 May 2026 09:48:39 -0500
> David Lechner <dlechner@baylibre.com> wrote:
> 
>> On 5/18/26 8:12 AM, Matti Vaittinen wrote:
>>> On 13/05/2026 07:49, 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.
>>>>
>>>> When at it, fix minor checkpatch code style warning (alignment).
>>>>
>>>> Signed-off-by: Javier Carrasco <javier.carrasco.cruz@gmail.com>
>>>> ---
>>>>    drivers/iio/light/veml6030.c | 8 +++++---
>>>>    1 file changed, 5 insertions(+), 3 deletions(-)
>>>>
>>>> diff --git a/drivers/iio/light/veml6030.c b/drivers/iio/light/veml6030.c
>>>> index 745cf3ad7092..855f052b60c2 100644
>>>> --- a/drivers/iio/light/veml6030.c
>>>> +++ b/drivers/iio/light/veml6030.c
>>>> @@ -871,9 +871,11 @@ static irqreturn_t veml6030_event_handler(int irq, void *private)
>>>>        else
>>>>            evtdir = IIO_EV_DIR_FALLING;
>>>>    -    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));
>>>>          return IRQ_HANDLED;
>>>>    }
>>>>   
>>>
>>> Is this change reflected to the user-space? If so, is it safe? We shouldn't break the ABI, right?
>>
>> On things like this, we will sometime risk breaking userspace in hopes
>> that no one was actually depending on it. If no one notices that we
>> broke it, did we really break it?
> 
> If it's a bug we often just break ABI compatibility anyway.  If we get a report
> we'll make it spit out both events (and someone doesn't handle an extra one
> we point them at the docs - that hopefully exist ;) - that say you must handle
> ones you don't expect).

Well, you guys are the ones being grilled if the brown thing hits the 
fan. ;) So, if you're willing to take the risk, then I am not complaining :)

Yours,
	-- Matti

-- 
---
Matti Vaittinen
Linux kernel developer at ROHM Semiconductors
Oulu Finland

~~ When things go utterly wrong vim users can always type :help! ~~