Re: [PATCH v3 6/6] ASoC: dt-bindings: mediatek,mt7986-afe: add audio afe document

[Maso Huang (黃加竹)]

On Fri, 2023-08-04 at 11:45 +0200, Krzysztof Kozlowski wrote:
>  	 
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
>  On 01/08/2023 10:25, Maso Huang (黃加竹) wrote:
> > On Fri, 2023-07-28 at 14:51 +0200, Krzysztof Kozlowski wrote:
> >>   
> >> External email : Please do not click links or open attachments
> until
> >> you have verified the sender or the content.
> >>  On 28/07/2023 11:08, Maso Huang wrote:
> >>> Add mt7986 audio afe document.
> >>>
> >>> Signed-off-by: Maso Huang <maso.huang@xxxxxxxxxxxx>
> >>
> >> Thank you for your patch. There is something to discuss/improve.
> >>
> >>> +
> >>> +required:
> >>> +  - compatible
> >>> +  - reg
> >>> +  - interrupts
> >>> +  - clocks
> >>> +  - clock-names
> >>> +  - assigned-clocks
> >>> +  - assigned-clock-parents
> >>
> >> You should constrain your clocks per variants. I doubt that they
> are
> >> really so flexible/optional on each SoC... or maybe missing clocks
> >> are
> >> result of unimplemented parts in the driver? But then this should
> not
> >> really affect bindings. Bindings still should require such clocks.
> >> Your
> >> DTS can always provide a <0>, if needed.
> >>
> >>
> > 
> > Hi Krzysztof,
> > 
> > After internal check, assigned-clocks and assigned-clock-parents
> are
> > not required on this SoC. 
> > Maybe we can just drop these two options?
> 
> It's separate issue, but yes - why requiring them?
> 
> I wrote about missing constraints for your clocks in the bindings.
> 
> Best regards,
> Krzysztof
> 

Hi Krzysztof,

OK, I'll remove assigned-clock and assigned-clock-parents.
And constrain the clocks for different SoC in the binding in v4 patch.

Best regards,
Maso