Re: [RFC PATCH] arm64: dts: qcom: Use labels with generic node names for ADC channels

[Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>]

On 14/12/2022 21:55, Marijn Suijten wrote:
> On 2022-12-12 09:48:26, Krzysztof Kozlowski wrote:
>> On 10/12/2022 17:54, Marijn Suijten wrote:
>>> On 2022-12-10 12:02:03, Krzysztof Kozlowski wrote:
>>>> On 09/12/2022 22:53, Marijn Suijten wrote:
>>>>> As discussed in [1] the DT should use labels to describe ADC
>>>>> channels, with generic node names, since the IIO drivers now
>>>>> moved to the fwnode API where node names include the `@xx`
>>>>> address suffix.
>>>>>
>>>>> Especially for the ADC5 driver that uses extend_name - which
>>>>> cannot be removed for compatibility reasons - this results in
>>>>> sysfs files with the @xx name that wasn't previously present, and
>>>>> leads to an unpleasant file-browsing experience.
>>>>>
>>>>> Also remove all the unused channel labels in pm660.dtsi.
>>>>>
>>>>> [1]:
>>>>> https://lore.kernel.org/linux-arm-msm/20221106193018.270106-1-marijn.suijten@somainline.org/T/#u
>>>>>
>>>>>
>>>>>
>> Signed-off-by: Marijn Suijten <marijn.suijten@somainline.org>
>>>>
>>>> The talk was in context of bindings, not about changing all
>>>> existing users thus affecting DTS.
>>>
>>> And as a consequence, DTS.  The already-merged transition from OF to 
>>> fwnode resulted in `@xx` to be included in the ADC channel name - and
>>> in the case of ADC5 even in sysfs filenames - so this seems like a 
>>> necessary change to make.
>>>
>>> At the very least I would have changed the bindings submitted or 
>>> co-authored /by myself/ since I initially decided to rely on this
>>> (now obviously) wrong behaviour, and should have used labels from the
>>> get go.
>>>
>>>> What's more, to me "skin-temp-thermistor" is quite generic name,
>>>> maybe "thermistor" would be more and reflects the purpose of the
>>>> node, so it was more or less fine.
>>>
>>> Are you suggesting to not use "adc-chan", but "thermistor" as node
>>> name (and still use skin_temp as label)?
>>
>> No, I am just saying that some of the names were correct, so the
>> reasoning in commit msg is not entirely accurate.
> 
> The commit msg doesn't state that any of these names was wrong, rather
> that the label property should be used to convey a name to the driver.

It stated: "with generic node names" and I said that some of the names
are generic. Anyway, I am fine with changing the node names if this is
coded in the bindings.

> 
> If you think multiple /node names/ are acceptable, please discuss in
> that direction.  Do we use a predetermined set of names (adc-chan@xx,
> thermistor@xx, ...) or do you have another suggestion?
> 
>>> Or to keep the fully-written-out "thermistor" word in the label?
>>
>> No, I don't refer to labels. Labels don't matter, they are being removed
>> entirely during DTS build.
> 
> Ugh, I knew this was becoming a problem, even in an earlier IIO
> discussions with Jonathan it was sometimes hard to keep apart:
> 
> I am /not/ referring to DTS labels, but to the `label =` node
> attribute/property, which is read by the driver (which isn't removed,
> how else would the driver fish this value out at runtime...).
> 
> After all none of this patch has any DTS labels, in fact a bunch of
> unused ones in pm660 have been removed... but this patch /does add/ the
> `label =` property to many nodes.
> 
>>>> Anyway I am against such changes without expressing it in the
>>>> bindings.
>>>
>>> As expressed in [1] I suggested and am all for locking this change
>>> in via bindings, and you are right to expect that to have gone paired
>>> with this patch.
>>
>> Yes, I expect such changes to have both binding and DTS change together.
>>
>>>
>>> I'll submit that as the leading patch to this in v2, with the
>>> wildcard pattern changed to adc-chan (or something else pending the
>>> discussion above), and should I then also require the label property
>>> via `label: true`?
>>
>> I don't think label is required.
> 
> My commit message explains why the ADC5/VADC driver should really
> receive a label property instead of using the node name.

The reason "unpleasant file-browsing experience" is usually OS specific,
so it does not justify requiring a label property. Label is just a
helper for the users.

Best regards,
Krzysztof