Re: [PATCH v1 1/4] dt-bindings: iio: adc: amlogic,meson-saradc: add S4 compatible

[Martin Blumenstingl]

Hi Krzysztof,

On Mon, Mar 2, 2026 at 2:59 PM Krzysztof Kozlowski <krzk@xxxxxxxxxx> wrote:
>
> On 02/03/2026 12:58, Martin Blumenstingl wrote:
> > Hi Krzysztof,
> >
> > On Sat, Feb 28, 2026 at 12:18 PM Krzysztof Kozlowski <krzk@xxxxxxxxxx> wrote:
> >>
> >> On Sat, Feb 28, 2026 at 02:58:37PM +0800, Nick Xie wrote:
> >>> Add the compatible string for the SARADC (Successive Approximation
> >>> Register ADC) IP block found in the Amlogic Meson S4 SoC.
> >>
> >> ... which is fully compatible with g12a?
> >>
> >> Write proper explanations.
> > Would you be fine with:
> > "There are no known differences between the SARADC on S4 and the one
> > on G12A. In the past differences between SARADC on the different SoCs
> > generations have been uncovered late, meaning that a dedicated
> > compatible string has proven to be useful."
>
> No, last sentence is redundant. You do not need to explain in the commit
> msg rules of bindings, because they are obvious/known. You need to
> explain the hardware, e.g. in this case the compatibility between devices.
Understood, thanks.

> >
> > Do you also require Nick to add amlogic,meson-g12a-saradc as fallback
> > compatible string for amlogic,meson-s4-saradc?
> >
> >>>
> >>> Signed-off-by: Nick Xie <nick@xxxxxxxxxx>
> >>> ---
> >>>  .../devicetree/bindings/iio/adc/amlogic,meson-saradc.yaml        | 1 +
> >>>  1 file changed, 1 insertion(+)
> >>>
> >>> diff --git a/Documentation/devicetree/bindings/iio/adc/amlogic,meson-saradc.yaml b/Documentation/devicetree/bindings/iio/adc/amlogic,meson-saradc.yaml
> >>> index bb9825e7346dd..5496a0dc714aa 100644
> >>> --- a/Documentation/devicetree/bindings/iio/adc/amlogic,meson-saradc.yaml
> >>> +++ b/Documentation/devicetree/bindings/iio/adc/amlogic,meson-saradc.yaml
> >>> @@ -27,6 +27,7 @@ properties:
> >>>                - amlogic,meson-gxm-saradc
> >>>                - amlogic,meson-axg-saradc
> >>>                - amlogic,meson-g12a-saradc
> >>> +              - amlogic,meson-s4-saradc
> >>
> >> If this is compatible with other device, you need proper fallback.
> >>
> >>>            - const: amlogic,meson-saradc
> >>
> >> And this one is now kind of fake - how does this device use it?
> > As you have noticed later, it is currently not used.
> > Can you please confirm that my understanding of the next steps is correct:
> > - you're expecting Nick to omit the amlogic,meson-saradc fallback
> > compatible string for amlogic,meson-s4-saradc
>
> Not omit but replace with proper fallback to specific compatible.
So in this case we need to end up with:
compatible = "amlogic,meson-s4-saradc", "amlogic,meson-g12a-saradc"
Stating that the S4 SARADC IP is compatible with the G12A SARADC IP.
Any S4 specific quirks that we need (in case we ever find something)
can then be managed in the driver.

> > - and you would like to see amlogic,meson-saradc removed (from the
> > bindings and .dts files that are already upstream) - I will take care
> > of this once S4 support lands upstream
>
> This no. It's an ABI, you cannot change it. Not worth it, either.
You suggested adding a comment stating that future bindings must not
use the "amlogic,meson-saradc" compatible string anymore.
Since it's part of ABI we need to keep it. So adding a comment is all I'll do.


Best regards,
Martin