[PATCH 1/2] ASoC: dt-bindings: fsl,audmix: Add support for i.MX952 platform
Shengjiu Wang
shengjiu.wang at gmail.com
Sun Jan 18 23:31:04 PST 2026
On Mon, Jan 19, 2026 at 3:07 PM Krzysztof Kozlowski <krzk at kernel.org> wrote:
>
> On 19/01/2026 07:27, Shengjiu Wang wrote:
> > On Sat, Jan 17, 2026 at 7:44 PM Krzysztof Kozlowski <krzk at kernel.org> wrote:
> >>
> >> On Fri, Jan 16, 2026 at 06:16:47PM +0800, Shengjiu Wang wrote:
> >>> The specific feature for AUDMIX on i.MX952 is that it can be bypassed,
> >>> so add fsl,amix-bypass property for this feature, besides this there is
> >>> no power domain defined on i.MX952, so make power-domains to be
> >>> dedicated to i.MX8QM.
> >>>
> >>> Signed-off-by: Shengjiu Wang <shengjiu.wang at nxp.com>
> >>> ---
> >>> .../devicetree/bindings/sound/fsl,audmix.yaml | 35 +++++++++++++++++--
> >>> 1 file changed, 33 insertions(+), 2 deletions(-)
> >>>
> >>> diff --git a/Documentation/devicetree/bindings/sound/fsl,audmix.yaml b/Documentation/devicetree/bindings/sound/fsl,audmix.yaml
> >>> index 3ad197b3c82c..50fb08460b4f 100644
> >>> --- a/Documentation/devicetree/bindings/sound/fsl,audmix.yaml
> >>> +++ b/Documentation/devicetree/bindings/sound/fsl,audmix.yaml
> >>> @@ -34,7 +34,9 @@ description: |
> >>>
> >>> properties:
> >>> compatible:
> >>> - const: fsl,imx8qm-audmix
> >>> + enum:
> >>> + - fsl,imx8qm-audmix
> >>> + - fsl,imx952-audmix
> >>>
> >>> reg:
> >>> maxItems: 1
> >>> @@ -75,12 +77,41 @@ properties:
> >>> unevaluatedProperties: false
> >>> description: Output port to SAI RX
> >>>
> >>> + fsl,amix-bypass:
> >>> + type: boolean
> >>> + description:
> >>> + The audmix module is bypassed from hardware.
> >>
> >> I don't understand why device AUDMIX would have property saying AUDMIX
> >> (so itself) can by bypassed. If you bypass the device itself, what do
> >> you configure here?
> >
> > The case is that the SAI interface is connected to AUDMIX, but AUDMIX can also
> > be bypassed by hardware configuration.
> >
> > Yes, adding this property in the AUDMIX module looks strange. Another
> > choice is to
> > add this property in the SAI binding document. Is this better?
> >
>
> Not sure, but certainly it does not look like a property of AUDMIX.
>
> I assume you acknowledge all other comments.
I will move the property to the SAI binding document, then please
have a review. As this change, the binding document will be updated, for
the comments to use the example-schema (:false) will be adopted if needed.
Thanks.
Best regards
Shengjiu Wang
>
>
> Best regards,
> Krzysztof
More information about the linux-arm-kernel
mailing list