Re: [PATCH 1/4] dt-bindings: iio: light: stk33xx: add power supply support

[Jonathan Cameron <jic23@kernel.org>]

On Sat, 07 Nov 2026 22:45:45 +0000
Roman Vivchar <rva333@protonmail.com> wrote:

> Hi Jonathan,
> 
> On Sunday, November 8th, 2026 at 12:36 AM, Jonathan Cameron <jic23@kernel.org> wrote:
> 
> > On Sun, 08 Nov 2026 00:15:03 +0200
> > Roman Vivchar via B4 Relay <devnull+rva333.protonmail.com@kernel.org> wrote:
> >   
> > > From: Roman Vivchar <rva333@protonmail.com>
> > >
> > > On some boards, the chip shares a power rail with other components. If
> > > that rail is turned off, the chip can't function.  
> > Not sure how the sharing is relevant.  Even if shared with nothing
> > else we need to turn it on.  So just mention that on some boards it
> > isn't always on.
> > 
> > The only reason this would be optional in the binding is if there
> > was internal routing in some packaging or similar that meant there
> > was no pin to connect.
> >
> > Upshot, make this required in the binding. Note that doesn't mean
> > the driver must require it - we are simply documenting that the
> > device needs power.
> >   
> 
> I agree this should be required, but that would break dt_binding_check
> for existing boards. I don't have their schematic nor boards themselves
> to verify the regulator for the power supply.

There aren't any upstream (from a quick grep) and so not your problem
unless you have some in a downstream tree that need fixing.
If you do then there is the option of a fixed regulator to close
the gap.  I consider a missing supply a bug that we should fix but maybe
others disagree.

Jonathan


> 
> Best regards,
> Roman