Re: [PATCH v21 3/8] dt-bindings: display: bridge: Add Cadence MHDP8501

[Laurentiu Palcu]

On Wed, Apr 08, 2026 at 09:32:54AM +0200, Krzysztof Kozlowski wrote:
> On 08/04/2026 09:13, Laurentiu Palcu wrote:
> >>> +  phys:
> >>> +    maxItems: 1
> >>> +    description:
> >>> +      phandle to the DP/HDMI PHY
> >>> +
> >>> +  interrupts:
> >>> +    items:
> >>> +      - description: Hotplug cable plugin.
> >>> +      - description: Hotplug cable plugout.
> >>> +
> >>> +  interrupt-names:
> >>> +    items:
> >>> +      - const: plug_in
> >>> +      - const: plug_out
> >>> +
> >>> +  cdns,bridge-type:
> >>
> >> Drop property. Graph defines what is connected on the other side. And if
> >> this is for different devices then compatible tells what bridge is that.
> > 
> > Initially, Sandor did use the remote compatible to decide the bridge
> > type but he assumed the remote is always the connector. However, as
> > Dmitry pointed out [1], this is not reliable as we can have another bridge
> > in-between this one and the connector.
> > 
> > [1] https://lore.kernel.org/all/dpj333mzr5azqhrgw3cxd7x5kiwxms4iomwy74uqfhr2zu4ocr@36rkth27d2jc/
> 
> So you still know what is on the other side, e.g. second bridge for HDMI
> or DP, even if this is DP over USB-C.
> 
> I understand that Dmitry did not want to use that part of code in the
> drivers, but what is located at the end is not really a separate
> property of this bridge, because it already duplicates that information.
> The final endpoint defines the type.

Ok, I'll drop the property and add a routine to traverse the graph and
find the last endpoint.

-- 
Thanks,
Laurentiu