Re: [PATCH 2/2] arm64: dts: amlogic: add the type-c controller on Radxa Zero 2

[Neil Armstrong <neil.armstrong@linaro.org>]

On 1/15/26 14:59, Ricardo Pardini wrote:
> Hi Neil,
> 
> On 15/01/2026 14:00, Neil Armstrong wrote:
> 
>>> The Radxa Zero2 has an FUSB302 controller on i2c3 at address 0x22 and
>>> INT# wired to GPIOA-13.
>>>
>>> +        interrupts = <13 IRQ_TYPE_LEVEL_LOW>;
>>
>> OK it seems I did read too fast, this should be 74 !
>>
>> S922X Datasheet (same as A311D for this)
>>
>> Table 6-32 GPIO Interrupt Sources
>>
>> ---------------------------------------------
>> | Input Mux Location    |    Description |
>> ---------------------------------------------
>> | [76:61]        |    gpioA[15:0] |
>> ---------------------------------------------
>>
>> So gpioA13 ==> 74
> 
> Indeed. Sorry for this.
> 
>>> +
>>> +        vbus-supply = <&ao_5v>;
>>> +
>>> +        status = "okay";
>>
>> And I get:
>>
>>    DTC [C] arch/arm64/boot/dts/amlogic/meson-g12b-radxa-zero2.dtb
>> arch/arm64/boot/dts/amlogic/meson-g12b-radxa-zero2.dtb: fusb302@22 (fcs,fusb302): 'connector' is a required property
>>          from schema $id: http://devicetree.org/schemas/usb/ fcs,fusb302.yaml#
>>
>> Please add a minimal connector, like:
>>
>> +               connector {
>> +                       compatible = "usb-c-connector";
>> +                       label = "USB-C";
>> +                       data-role = "host";
>> +                       power-role = "source";
>> +                       source-pdos = <PDO_FIXED(5000, 2000, PDO_FIXED_USB_COMM)>;
>> +               };
>>
>> I'll drop this change from my tree.
> 
> Again, my bad. My intention here was a "minimal" description, mostly to reserve 0x22 so end-users can run `i2cdetect` or such on the 40-pin exposed i2c3 and not be surprised. Guess I didn't run dtbs_check hard enough and missed the required connector.
> 
>>
>> And you may also add support for the superspeed mux to switch the USB3 polarity.
>>
>> I have a prototype at https://gitlab.com/superna9999/linux/-/tree/topic/ amlogic/radxa-zero2/fusb302?ref_type=heads
> 
> I'll try your prototype. I had no idea it actually had those capabilities.
> 
> Do you think it is worth to resend this with just the fixed 74 pin & the minimal connector as you suggested? Now that I've seen your work, a "add FUSB302" that does nothing (while being capable) feels frivolous and confusing.

Yes please resent with the interrupt fix and minimal connector, I'll rebase my work on top. It still needs some testing, but I'll be happy if you test it on your side aswell.

Neil

> 
> Thanks,
> Ricardo
> 
> 
>