Re: [PATCH v6 8/9] arm64: dts: qcom: kodiak: Describe EUD UTMI path using child node

[Elson Serrao]

On 8/24/2026 6:48 AM, Konrad Dybcio wrote:
> On 8/11/26 11:36 PM, Elson Serrao wrote:
>> The existing EUD description uses a legacy single-path representation
>> that assumes uniform UTMI routing. This was sufficient for minimal
>> configurations where the USB port operated only in device mode and role
>> switching was not considered.
> 
> [...]
> 
>> +				ports {
>> +					#address-cells = <1>;
>> +					#size-cells = <0>;
>> +
>> +					port@0 {
>> +						reg = <0>;
>> +						eud_usb: endpoint {
>> +							remote-endpoint = <&usb_1_dwc3_hs>;
> 
> This change alone will trip "connection is not bi-directional" warnings,
> as board DTs override that endpoint already
> 
Patch 8 on its own would indeed trigger the "connection is not
bi-directional" warnings. Patch 9 updates the corresponding
board-level endpoint overrides and resolves those warnings.
 
Would you prefer that I squash patch 9 into patch 8 so that the
graph remains valid within a single patch? Alternatively, we could
move the endpoint assignment into the board DTS. However, since the
EUD-to-controller mapping is on-SoC rather than board-specific, keeping
it in the SoC DTSI appears to be the more appropriate placement.
> 
>> +						};
>> +					};
>> +
>> +					port@1 {
>> +						reg = <1>;
>> +						eud_con: endpoint {
> 
> Let's keep an \n between the last property and the following subnodes
> (like you did in the below hunk)
> 
Ack

Thanks
Elson

> 
>> +						};
>>  					};
>>  				};
>>  			};
>> @@ -4989,6 +4995,7 @@ port@0 {
>>  					reg = <0>;
>>  
>>  					usb_1_dwc3_hs: endpoint {
>> +						remote-endpoint = <&eud_usb>;
>>  					};
>>  				};
>>