RE: [PATCH] dt-bindings: usb: Add ITE IT5271/IT5272/IT5273/IT5274 support

[amber.kao]

On Mon, Sep 07, 2026 at 05:03:00PM +0800, Krzysztof Kozlowski wrote:
> On Thu, Sep 03, 2026 at 06:15:43PM +0800, Amber Kao wrote:
> > This binding is submitted independently so it can be reviewed on
> 
> No, it cannot be reviewed on its own, because splitting it means you
> have undocumented compatibles in the other patchset.
> 
> And if you do not, then there is no user for this binding.
> 
> > its own; once the driver series lands, IT527x support will be wired
> > into it there.
> 
> Then what is the point of this binding now?

Understood, that's a fair point — reviewing a binding without a driver
user in tree is a waste of everyone's time. I'll drop this standalone
binding patch.

> > +    - IT5271: single-port variant, SPR
> > +    - IT5272: two-port variant, SPR
> > +    - IT5273: single-port variant, EPR
> > +    - IT5274: two-port variant, EPR
> > +
> 
> Why isn't this entire binding part of the other file?

IT527x and IT885x are different hardware, which is why we described
them in separate DT bindings. However, since their UCSI operational
flow is similar, they share the same driver (ucsi_itepd.c).

> > +properties:
> > +  compatible:
> > +    oneOf:
> > +      - const: ite,it5271
> > +      - items:
> > +          - const: ite,it5272
> 
> enum

Noted, will fix in the next submission.

> I finished actually here, becausr I still find reviewing outside of
> driver a waste of our time.

Instead of resubmitting this binding alone, I'm planning to send a new
series containing both the binding (with the enum fix) and the
corresponding compatible strings added to ucsi_itepd.c, so there's
always a driver user for the binding under review.

One thing I'd like to check first: the base IT8851/IT8853 series this
depends on has already picked up Reviewed-by tags from Heikki Krogerus
and Rob Herring, but hasn't landed in mainline yet. Heikki previously
suggested it's fine to send a follow-up series before the base lands,
as long as the dependency is noted in the cover letter. Given that,
would you be okay with me sending the IT527x driver+binding series now
with that dependency called out, or would you prefer I wait until the
IT8851/IT8853 series is actually merged first?

Best regards,
Amber Kao