[PATCH 2/2] arm64: dts: imx8mp: Add DH i.MX8MP DHCOM SoM on DRC02 carrier board
Marek Vasut
marex at denx.de
Wed Sep 25 15:08:03 PDT 2024
On 9/25/24 11:27 PM, Frank Li wrote:
Hi,
>>>> +/* USB_OTG port is not routed out on DRC02. */
>>>> +&usb3_0 {
>>>> + status = "disabled";
>>>> +};
>>>> +
>>>> +&usb_dwc3_0 {
>>>> + status = "disabled";
>>>> +};
>>>> +
>>>> +/* USB_HOST port has USB Hub connected to it, PWR/OC pins are unused */
>>>> +&usb3_1 {
>>>> + fsl,disable-port-power-control;
>>>> + fsl,permanently-attached;
>>>> +};
>>>
>>> Suggest run https://github.com/lznuaa/dt-format to sort node
>>>
>>> &usb_dwc3_0
>>> &usb_dwc3_1
>>> &usb3_0
>>> &usb3_1
>>
>> I'm afraid the tool (I did not try it, I used plain GNU sort) is not sorting
>> this correctly, look at ASCII table, 0x33 is '3' and 0x5f is '_':
>>
>> $ printf 3_ | hexdump -vC
>> 00000000 33 5f
>
> Sorry, I have not run it for this dts. I make misstake - and _ at ascii
> table.
No worries
>> So if the sorting was correct, it would be:
>>
>> usb3_0
>> usb3_1
>> usb_dwc3_0
>> usb_dwc3_1
>>
>> But I would prefer to keep the controllers close to each other, i.e. the
>> current order which seems more logical in this case:
>>
>> usb3_0
>> usb_dwc3_0
>> usb3_1
>> usb_dwc3_1
>
> next time, we should use better label name in chip.dtsi file. I try to
> reduce trivial node name order issue in patch review. And make contributor
> to be easy to follow.
I suspect the dwc3_0 and co. is somewhat inherited from other DTs which
used the controller before.
More information about the linux-arm-kernel
mailing list