On 12 November 2019 15:24, Sebastian Reichel wrote:
> > > static const struct snd_soc_dapm_widget da7213_dapm_widgets[] = {
> > > + /*
> > > + * Power Supply
> > > + */
> > > + SND_SOC_DAPM_REGULATOR_SUPPLY("VDDA", 0, 0),
> >
> > Having spoken with our HW team, this will cause a POR in the device so we can't
> > just enable/disable VDD_A supply. Needs to present at all times. How are you
> > verifying this?
>
> Ok. The system, that I used for testing shared a regulator
> for VDDIO and VDDA. I suppose this needs to be moved next
> to enabling VDDIO then.
Yes, and as Mark mentioned you can use the bulk enable/disable calls.
> > > + da7213->vddio = devm_regulator_get(&i2c->dev, "VDDIO");
> > > + if (IS_ERR(da7213->vddio))
> > > + return PTR_ERR(da7213->vddio);
> > > +
> > > + ret = regulator_enable(da7213->vddio);
> > > + if (ret < 0)
> > > + return ret;
> > > +
> > > + ret = devm_add_action_or_reset(&i2c->dev, da7213_power_off,
> > > da7213);
> > > + if (ret < 0)
> > > + return ret;
> >
> > We're seemingly leaving the VDDIO regulator enabled on failure, unless I'm
> > missing some magic somewhere?
>
> If regulator_enable fails, the regulator is not enabled. If
> devm_add_action_or_reset
> fails, it will call da7213_power_off().
Right, didn't spot that as haven't seen that used before. Nice :)
_______________________________________________
Alsa-devel mailing list
Alsa-devel@xxxxxxxxxxxxxxxx
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
