Re: [PATCH V3 4/5] arm64: dts: qcom: x1-crd: Add Embedded controller node

[Konrad Dybcio]

On 3/9/26 10:13 PM, Dmitry Baryshkov wrote:
> On Mon, Mar 09, 2026 at 04:21:50PM +0530, Sibi Sankar wrote:
>>
>> On 3/9/2026 2:39 PM, Krzysztof Kozlowski wrote:
>>> On 09/03/2026 10:03, Sibi Sankar wrote:
>>>> On 3/9/2026 12:55 PM, Krzysztof Kozlowski wrote:
>>>>> On Mon, Mar 09, 2026 at 05:06:45AM +0530, Sibi Sankar wrote:
>>>>>> Add embedded controller node for Hamoa/Purwa CRDs which adds fan control,
>>>>>> temperature sensors, access to EC internal state changes and suspend
>>>>>> entry/exit notifications to the EC.
>>>>>>
>>>>>> Signed-off-by: Sibi Sankar <sibi.sankar@xxxxxxxxxxxxxxxx>
>>>>>> ---
>>>>>>    arch/arm64/boot/dts/qcom/x1-crd.dtsi | 16 ++++++++++++++++
>>>>>>    1 file changed, 16 insertions(+)
>>>>>>
>>>>>> diff --git a/arch/arm64/boot/dts/qcom/x1-crd.dtsi b/arch/arm64/boot/dts/qcom/x1-crd.dtsi
>>>>>> index ded96fb43489..29a1aeb98353 100644
>>>>>> --- a/arch/arm64/boot/dts/qcom/x1-crd.dtsi
>>>>>> +++ b/arch/arm64/boot/dts/qcom/x1-crd.dtsi
>>>>>> @@ -1042,6 +1042,16 @@ eusb6_repeater: redriver@4f {
>>>>>>    		#phy-cells = <0>;
>>>>>>    	};
>>>>>> +
>>>>>> +	embedded-controller@76 {
>>>>>> +		compatible = "qcom,hamoa-it8987-ec", "qcom,hamoa-ec";
>>>>> I don't see updates to other x1e boards, thus my arguments from v2 stay
>>>>> valid. It's wrong to call it "hamoa-ec" since only one Hamoa board has
>>>>> it. All of other Hamoa boards apparently do not have it.
>>>>
>>>> Hey Krzysztof,
>>>> Thanks for taking time to review the series :)
>>>>
>>>> What other Hamoa boards are you referring to? The series
>>>> mentions that the driver and bindings is meant for Qualcomm
>>>> Hamoa/Purwa/Glymur "reference" devices, so it only covers
>>>> CRD and IOT-EVK. It definitely does not cover all Hamoa boards
>>>> boards like you are assuming.
>>> hamoa-ec compatible implies that and that's something I raised in v2
>>> already. You need a specific compatible.
>>
>>
>> Hamoa/Glymur reference devices can have different EC MCUs
>> depending on the SKU. This introduces the need to deal with
>> possibility of quirks and bugs introduced by these differences.
>> Hamoa/Purwa CRDs and IOT EVK runs on IT8987, while Glymur
>> reference devices run on NPCX498/488. This pretty much was
>> the rationale to make the MCU part of the compatible. Anyway
>> I can keep it as qcom,hamoa-ec and qcom,glymur-ec for now
>> and add specific compatibles when we upstream those boards?
> 
> Why do you need a generic compat at all? Glancing at the database, at
> least the following laptops seem to have similar protocol (I might be
> wrong, this is based on a very quick glance):
> 
> acer-sfa14-11
> asus-vivobook-s-15
> asus-vivobook-s15-q5507
> asus-vivobook-s15-s5507
> honor-magicbook-art-14
> honor-mro-521
> hp-elitebook-6-g1q
> hp-omnibook-5-14-he0xxx
> lenovo-ideacentre-01q8x10
> lenovo-ideapad-slim3x-15q8x10-WCN7850
> lenovo-thinkpad-t14s-120hz-64gb
> lenovo-thinkpad-t14s
> lenovo-yoga-slim-7x
> medion-14-s1-elite-sprchrgd
> medion-14-s1-sprchrgd
> qualcomm-snapdragon-dev-kit
> tuxedo-elite-14-gen1
> 
> I assume that some of them might be false positives and others will use
> vendor (or device) extensions to your protocol.
> 
> I'd suggest implicitly using "qcom,hamoa-crd-ec" / "qcom,glymyr-crd-ec",
> because then we can use something like "asus,vivobook-s15-ec" to
> identify the EC on VivoBook S15.

+1

Konrad