Re: [PATCH v3 2/3] dt-bindings: rockchip: rk809: Document audio codec properties

[Tim Lunn]

On 1/23/24 18:37, Krzysztof Kozlowski wrote:
> On 23/01/2024 05:10, Tim Lunn wrote:
> > On 1/22/24 19:14, Krzysztof Kozlowski wrote:
> > > On 20/01/2024 14:55, Tim Lunn wrote:
> > > > Rockchip RK809 shares the same audio codec block as the rk817 mfd, and
> > > > is compatible with the existing rk817_codec driver.
> > > Please use subject prefixes matching the subsystem. You can get them for
> > > example with `git log --oneline -- DIRECTORY_OR_FILE` on the directory
> > > your patch is touching.
> > Ok I will check this.
> > > > This patch introduces to the binding the standard property #sound-dai-cells
> > > > and also an optional codec child node to hold codec specific properties.
> > > > Currently there is only one property in this node however the downstream
> > > > driver shows a number of other properties that are supported by the codec
> > > > hardware, that could be implemented in the future. This maintains the
> > > > existing driver ABI and keeps consistency with the rk817 bindings.
> > > So you are adding a new node? Just for one property? No, just put it
> > > into parent node.
> > The existing upstream codec driver parses the property from the "codec"
> > sub-node, if I
> > move it to the parent node here, I will need to patch the codec driver
> > to search in both locations,
> > so as to not break the rk817 bindings.  If that is preferred, I can do
> > it that way.
> Your long commit msg has just very short mention about existing driver
> and the rest is not helpful. Please rephrase to explain why and what you
> are doing it.
OK I will rephrase both commit messages for the next version.
> > > Downstream driver does not matter at all in that aspect.
> > The codec hardware supports additional properties but they are not
> > implemented currently in
> > upstream driver.
>
> Again: it does not matter. Bindings are not about drivers.
>
> Best regards,
> Krzysztof