Re: [PATCH v5 15/20] dt-bindings: clock: imx8m-clock: support spread spectrum clocking

[Krzysztof Kozlowski <krzk@kernel.org>]

On 08/12/2024 17:47, Dario Binacchi wrote:
>>>    '#clock-cells':
>>>      const: 1
>>> @@ -43,6 +43,34 @@ properties:
>>>        ID in its "clocks" phandle cell. See include/dt-bindings/clock/imx8m-clock.h
>>>        for the full list of i.MX8M clock IDs.
>>>
>>> +  fsl,ssc-modfreq-hz:
>>> +    description:
>>> +      The values of modulation frequency (Hz unit) for each clock
>>> +      supporting spread spectrum.
>>> +    minItems: 7
>>> +    maxItems: 10
>>
>> Why all cloks receive now spread spectrum? I had impression - and all
>> your previous versions were doing this - that you have only three or
>> four clocks with SSC.
> 
> Exactly. Indeed, the first six values are not valid as SSC properties but are
> only used to reach the point where the first PLL with SSC (i.e., audio_pll1)
> can be indexed, which is in position 7 in the clocks list.
> This was the rationale I followed.
> And it is explicitly outlined in the example section.
> The "" for the fsl,ssc-method property is precisely aimed at specifying a
> "no SSC" method, which also fixes the warning:
> 
> fsl,ssc-method:0: '' is not one of ['down-spread', 'up-spread', 'center-spread']
> 
> raised by
> make dt_binding_check DT_SCHEMA_FILES=imx8m-clock.yaml
> 
> Or would it be acceptable to specify a list of SSC values that applies only to
> the last 4 PLLs in the clocks list?

Lists are strictly ordered, so their items are known. You take some
clocks, of which last clocks are "foo" and "bar". Then you have new
property for configuring SSC for clocks - again list - "foo" and "bar".

> 
> I feel like I might be missing something.
> 
> Could you kindly suggest what to do or provide a DTS example to show me
> what you expect?

Best regards,
Krzysztof