Re: [PATCH 1/6] dt-bindings: thermal: tsens: Add nvmem cells for calibration data

[Krzysztof Kozlowski]

On 11/07/2023 16:13, Praveenkumar I wrote:
>>>>>              - const: calib
>>>>> @@ -205,6 +209,24 @@ properties:
>>>>>              - const: s9_p2_backup
>>>>>              - const: s10_p1_backup
>>>>>              - const: s10_p2_backup
>>>>> +      - items:
>>>>> +          - const: mode
>>>>> +          - const: base0
>>>>> +          - const: base1
>>>>> +          - const: s0_offset
>>>>> +          - const: s3_offset
>>>>> +          - const: s4_offset
>>>>> +          - const: s5_offset
>>>>> +          - const: s6_offset
>>>>> +          - const: s7_offset
>>>>> +          - const: s8_offset
>>>>> +          - const: s9_offset
>>>>> +          - const: s10_offset
>>>>> +          - const: s11_offset
>>>>> +          - const: s12_offset
>>>>> +          - const: s13_offset
>>>>> +          - const: s14_offset
>>>>> +          - const: s15_offset
>>>> Don't introduce new naming style. Existing uses s[0-9]+, without offset
>>>> suffix. Why this should be different?
>>> As I mentioned above, s[0-9]+_p1 / s[0-9]+p2 is for TSENS V1. TSENS V2
>>> QFPROM layout is different from the existing one.
>> I know, I did not write about p1/p2.
>>
>>> I would like to add mode, base0, base1 and 16 patterns
>>> '^s[0-15]+_offset$'. But DT binding check is failing in oneOf/ anyOf
>>> condintion.
>> This does not explain why you need different style - this "offset" suffix.
> In QFPROM, the BIT field names are s0_offset, s3_offset to s15_offset. 
> s1_offset and s2_offset not present in IPQ5332.
> Hence used the same "offset" suffix here. May I know in this case, which 
> nvmem-cells-names you are suggesting to use?

"Existing uses s[0-9]+"

so s[0-9]+

Best regards,
Krzysztof