Quoting Srinivasa Rao Mandadapu (2022-02-09 06:26:58)
>
> On 2/9/2022 2:42 AM, Stephen Boyd wrote:
> > Quoting Srinivasa Rao Mandadapu (2022-02-08 07:34:14)
> >
> >> + pins = "gpio83";
> >> + function = "gpio";
> >> + drive-strength = <16>;
> >> + output-high;
> >> + };
> >> +
> >> + wcd938x_reset_sleep: wcd938x_reset_sleep {
> >> + pins = "gpio83";
> >> + function = "gpio";
> >> + drive-strength = <16>;
> >> + bias-disable;
> >> + output-low;
> > Why doesn't the device drive the reset gpio by requesting the gpio and
> > asserting and deasserting it? We shouldn't need to use pinctrl settings
> > to toggle reset gpios.
> Okay. Verified without these nodes and didn't see any impact. But
> similar way it's mentioned in sm8250-mtp.dts. Could You please suggest
> on it how to go ahead on this?.
I'd expect the wcd938x codec device node to have a 'reset-gpios'
property like
reset-gpios = <&tlmm 83 GPIO_ACTIVE_LOW>
and then the driver to request that gpio via
reset_gpio = devm_gpiod_get(dev, "reset", GPIOD_OUT_HIGH);
so it gets the gpio during driver probe. Then the gpio can be deasserted
during suspend and reasserted on resume, if that's even important?
