Re: [PATCH 2/2] arm64: dts: qcom: Add support for Acer Swift Go Pro AI (SFA14-11)

[Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>]

On 7/5/26 5:29 PM, Kaipeng Zeng wrote:
> 
> Hi Konrad,
> 
> On 7/2/26 19:45, Konrad Dybcio wrote:
>> On 7/2/26 8:31 AM, Kaipeng Zeng wrote:
>>> +    wcd938x: audio-codec {
>>
>> Nodes should be sorted - by 'reg' and unit address if there's one,
>> by name if there isn't (minus some special cases like GPIOs), see:
>>
>> https://docs.kernel.org/devicetree/bindings/dts-coding-style.html

[...]

>>> +&apps_rsc {
>>> +    regulators-0 {
>>> +        compatible = "qcom,pm8550-rpmh-regulators";
>>> +        qcom,pmic-id = "b";
>>
>> Have you verified this against the DSDT of your device? Incorrect
>> settings may lead to hw damage
>>
>> [...]
>>
> 
> Thank you for pointing this out, I had not verified this before.
> Looking inside the DSDT, I found some "PMICVREGVOTE" packages that list
> a value as the required voltage.
> And I need to check if any required voltage falls outside the
> range set by DTS. Is that correct?
> However, some PPP_RESOURCE_ID_* seem to be mentioned only and lack a specific voltage.
> For these, should I retain the original value (from x1e80100-crd.dtsi),
> or is there another way to determine their correct ranges?

You can only reasonably-safely copy settings for L12B and L15B (they
power always-on components) and regulators that supply the on-SoC
components (like PHYs). Everything else may be board specific, so avoid
if unsure.

> 
>>> +&iris {
>>> +    status = "okay";
>>
>> You need to specify a firmare path
>>
> 
> I will add a firmware path.
> But since the firmware of this device is not archived by linux-firmware, is it proper to use "qcom/x1e80100/ACER/SFA14-11/*.mbn" as the firmware path?

Yes, it's still useful if you e.g. keep a local firmware package, since
it allows for the same image to be reused across machines (you don't have
to overwrite the existing files)

Konrad