Re: [PATCH v5 2/2] usb: typec: hd3ss3220: Enable VBUS based on ID pin state

[Dmitry Baryshkov]

On 31/10/2025 08:45, Krishna Kurapati PSSNV wrote:
> On 10/27/2025 6:49 PM, Dmitry Baryshkov wrote:
> > On Mon, Oct 27, 2025 at 11:47:13AM +0200, Heikki Krogerus wrote:
> > > Hi Krishna,
> > >
> > > > +static int hd3ss3220_get_vbus_supply(struct hd3ss3220 *hd3ss3220)
> > > > +{
> > > > +    struct device_node *hd3ss3220_node = hd3ss3220->dev->of_node;
> > > > +    struct device_node *np;
> > > > +
> > > > +    np = of_graph_get_remote_node(hd3ss3220_node, 0, 0);
> > > > +    if (!np) {
> > > > +        dev_err(hd3ss3220->dev, "failed to get device node");
> > > > +        return -ENODEV;
> > > > +    }
> > >
> > > So I guess that's the connector node. Why can't you just place the
> > > regulator reference to the hd3ss3220 controller node instead of the
> > > connector like the port controllers do?
> > >
> > > That would allow us to do a simple devm_regulator_get_optional() call
> > > that's not tied to DT only.
> >
> > But we have devm_of_regulator_get_optional(), it was mentioned in the
> > previous email if I'm not mistaken. If we need, we should add
> > devm_fwnode_regulator_get(_optional).
> >
> > vbus supply is described as a part of the usb-c-connector schema, so
> > it is not that logical to describe it as a part of the Type-C
> > controller.
>
> I tried the following as suggested:
>
> hd3ss3220->vbus =  devm_of_regulator_get_optional(hd3ss3220->dev, 
> to_of_node(connector),
> if (IS_ERR(hd3ss3220->vbus))
>    hd3ss3220->vbus = NULL;
>
> If there is a vbus supply I see its returning proper handle pointer.
> Else it returned ENODEV. (which is fine for our case as there is no vbus 
> in DT).
>
> Can I mark the function as a void one. Instead of returning any int 
> value, would it be fine if to just mark vbus as NULL and proceed ?

You can only ignore -ENODEV. Any other error should be propagated back 
and cause an error in probing the hd3ss3220 driver.

> Regards,
> Krishna,


--
With best wishes
Dmitry