[PATCH v2 1/5] arm64: dts: rockchip: Add AP6275P wireless support to Khadas Edge 2

Krzysztof Kozlowski krzk at kernel.org
Mon Jul 1 01:41:30 PDT 2024


On 25/06/2024 10:04, Jacobe Zang wrote:
>>> diff --git a/arch/arm64/boot/dts/rockchip/rk3588s-khadas-edge2.dts b/arch/arm64/boot/dts/rockchip/rk3588s-khadas-edge2.dts
>>> index 3b6286461a746..f674deb6f7da8 100644
>>> --- a/arch/arm64/boot/dts/rockchip/rk3588s-khadas-edge2.dts
>>> +++ b/arch/arm64/boot/dts/rockchip/rk3588s-khadas-edge2.dts
>>> @@ -356,6 +356,22 @@ &pcie2x1l2 {
>>>         reset-gpios = <&gpio3 RK_PD1 GPIO_ACTIVE_HIGH>;
>>>         vpcie3v3-supply = <&vcc3v3_pcie_wl>;
>>>         status = "okay";
>>> +
>>> +     pcie at 0,0 {
>>> +             reg = <0x400000 0 0 0 0>;
>>> +             #address-cells = <3>;
>>> +             #size-cells = <2>;
>>> +             ranges;
>>> +             device_type = "pci";
>>> +             bus-range = <0x40 0x4f>;
>>
>> Isn't bus-range a property of PCI host bridge, so the parent? This is a
>> PCI device, right?
>>
>>> +
>>> +             wifi: wifi at 0,0 {
>>
>> Binding does not say anything about this. Rockchip PCI controller is the
>> PCI host bridge, isn't it? Then the pci at 0,0 is the child, so what is this?
> 
> The host bridge is the parent of pcie at 0,0. And pcie at 0,0 is Bridge1, so the

Do you want to say Rockchip PCI is PCI-PCI bridge? Bindings do not allow it.

> wifi at 0,0 as a device under the Bridge1.
> 
>>
>>> +                     reg = <0x410000 0 0 0 0>;
>>> +                     clocks = <&hym8563>;
>>> +                     clock-names = "32k";
>>
>> 1. Bindings are before the users.
>> 2. Where is the compatible? Are you sure this validates?
> 
> Before, the compatible is "pci14e4,449d", but when I checkpatch the warning
> said that "pci14e4" was not documented, so I remove the compatible which 
> doesn't affect the Wi-Fi function. I have tried to add "pci14e4" to 
> vendor-prefixes.yaml but was refused. So whether should I add the compatible 
> with warning? 

I talk about dtbs_check, not checkpatch. That checkpatch warning does
not matter, obviously.

Best regards,
Krzysztof




More information about the linux-arm-kernel mailing list