[PATCH v2 11/21] arm64: dts: apple: t8103-j313: Add speaker/headset jack nodes

[James Calligeros]

On Thu, Oct 8, 2026 at 4:33 AM Ajay Kumar Nandam
<ajay.nandam at oss.qualcomm.com> wrote:
>
>
>
> On 10/4/2026 11:33 AM, James Calligeros wrote:
> > +&i2c3 {
> > +     status = "okay";
> > +
> > +     speaker_right: codec at 34 {
> > +             compatible = "ti,tas5770l";
> > +             reg = <0x34>;
> > +             interrupts-extended = <&pinctrl_ap 182 IRQ_TYPE_LEVEL_LOW>;
> > +             shutdown-gpios = <&pinctrl_ap 181 GPIO_ACTIVE_HIGH>;
> > +             #sound-dai-cells = <0>;
>
> Both speaker codec nodes reference the same shutdown GPIO. Is that
> intentional for this board?
>
> The TAS2770 driver requests 'shutdown-gpios' per codec instance, and
> GPIO requests are exclusive by default. So I would expect only one codec
> instance to own this line, while the other appears to fall back to the
> register-only path after the GPIO request fails.
>
> If GPIO181 is a shared physical shutdown line for both amps, how is the
> power/reset ordering expected to work across probe, suspend, and resume
> when only one of the two codec driver instances controls that line?

We use the new(ish) shared GPIO proxy stuff[1] and it all Just Works.

[1] https://lore.kernel.org/asahi/20260831-apple-shared-gpio-v1-0-e855b12e18ce@gmail.com/