hi,
On Fri, Feb 10, 2012 at 10:31:11AM +0100, Heiko Stübner wrote:
> > @@ -184,6 +190,36 @@ struct s3c_hsotg_req {
> > unsigned char mapped;
> > };
> >
> > +static void s3c_hsotg_regulator_enable(struct s3c_hsotg *hsotg)
> > +{
> > + mutex_lock(&hsotg->mutex_lock);
> > + if (!hsotg->enabled) {
> > + if (hsotg->reg_vusb_d)
> > + regulator_enable(hsotg->reg_vusb_d);
> > +
> > + if (hsotg->reg_vusb_a)
> > + regulator_enable(hsotg->reg_vusb_a);
> > +
> > + hsotg->enabled = true;
> > + }
> > + mutex_unlock(&hsotg->mutex_lock);
> > +}
> > +
> > +static void s3c_hsotg_regulator_disable(struct s3c_hsotg *hsotg)
> > +{
> > + mutex_lock(&hsotg->mutex_lock);
> > + if (hsotg->enabled) {
> > + if (hsotg->reg_vusb_a)
> > + regulator_disable(hsotg->reg_vusb_a);
> > +
> > + if (hsotg->reg_vusb_d)
> > + regulator_disable(hsotg->reg_vusb_d);
> > +
> > + hsotg->enabled = false;
> > + }
> > + mutex_unlock(&hsotg->mutex_lock);
> > +}
> > +
>
> how about using regulator_bulk_* functions instead of calling each regulator
> separately - i.e. like in s3c_hsudc.c
yup, I agree here. it will make things neater.
> As Felipe will probably also mention, please add another patch to the series
> converting the driver to the newer udc_start / udc_stop callbacks while you're
> at it (essentially also like s3c-hsudc.c)
indeed, thanks for pointing that out ;-)
--
balbi
Attachment:
signature.asc
Description: Digital signature
