On 19/09/2023 14:48, Sricharan Ramabadhran wrote:
>
>
> On 9/19/2023 6:02 PM, Krzysztof Kozlowski wrote:
>> On 19/09/2023 09:22, Sricharan Ramabadhran wrote:
>>>
>>>
>>> On 9/15/2023 6:15 PM, Krzysztof Kozlowski wrote:
>>>> On 15/09/2023 14:43, Krzysztof Kozlowski wrote:
>>>>> On 15/09/2023 14:15, Sricharan Ramabadhran wrote:
>>>>>> IPQ5018 has tsens v1.0 block with 4 sensors and 1 interrupt.
>>>>>>
>>>>>> Signed-off-by: Sricharan Ramabadhran <quic_srichara@xxxxxxxxxxx>
>>>>>> ---
>>>>>> [v2] Sorted the compatible and removed example
>>>>>>
>>>>>
>>>>> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx>
>>>>
>>>> No, unreviewed. Your driver says it is not compatible with
>>>> qcom,tsens-v1. This does not look right :/
>>>>
>>>
>>> Yes it is V1 IP, but since there is no RPM, to enable the IP/SENSORS
>>> have to do those steps after calling init_common. Similar reason
>>> added a new feat as well in patch #2 as well. Hence for this,
>>> new compatible was required.
>>
>> I dud not write about new or old compatible ("compatible" as noun). I
>> wrote that it is not compatible ("compatible" as adjective) with v1.
>>
>
> Ho, in that case, yes it is not compatible with V1 init and features
> because of 'no rpm'. So in that case, should this be documented
> as a separate version of 'V1 without rpm' ?
It should not be mixed with regular v1, just as new entry there. I don't
think fallback is needed - just use SoC specific compatible.
Best regards,
Krzysztof
