On Mon, Jul 10, 2023 at 12:20 PM Conor Dooley <conor@xxxxxxxxxx> wrote:
>
> On Mon, Jul 10, 2023 at 12:03:38PM -0600, Rob Herring wrote:
> > On Mon, Jul 10, 2023 at 11:57 AM Conor Dooley <conor@xxxxxxxxxx> wrote:
> > >
> > > On Fri, Jul 07, 2023 at 03:06:46PM -0600, Rob Herring wrote:
> > > > Convert the OmniVision OV7251 Image Sensor binding to DT schema format.
> > > >
> > > > vddd-supply was listed as required, but the example and actual user
> > > > don't have it. Also, the data brief says it has an internal regulator,
> > > > so perhaps it is truly optional.
> > >
> > > ov7251.c:
> > > ov7251->core_regulator = devm_regulator_get(dev, "vddd");
> > > if (IS_ERR(ov7251->core_regulator)) {
> > > dev_err(dev, "cannot get core regulator\n");
> > > return PTR_ERR(ov7251->core_regulator);
> > > }
> > >
> > > Looks like the driver's probe function disagrees?
> >
> > Doesn't the regulator framework return a dummy regulator if missing?
>
> Huh, I think I misunderstood how _regulator_get() worked. Apologies for
> the noise, I thought it was returning an error when it used a dummy.
>
> > > I was going to ask how it worked, but the one user has
> > > status = "disabled"...
> >
> > Saw that too, but figured there's some other include with that
> > overridden. We should really add a built .dts output target to avoid
> > trying to manually walk includes.
>
> Probably gonna sound like an eejit, but the user is in a dts - can an
> include overwrite that status?
Yes, sometimes .dts files are both built into a dtb and included. Say
you have a board with and without an optional camera sensor.
Rob
