On 23/08/2022 16:34, Bryan O'Donoghue wrote:
> On 23/08/2022 13:23, Krzysztof Kozlowski wrote:
>> On 23/08/2022 15:12, Krzysztof Kozlowski wrote:
>>> On 22/08/2022 15:03, Bryan O'Donoghue wrote:
>>>> We need to define pwm@bc000 to stop dtbs_check from making the following
>>>> complaint, text pruned.
>>>>
>>>> pmic@1: 'pwm' does not match any of the regexes: 'pwm@[0-9a-f]+$'
>>>>
>>>> Fixes: e79a1385ab74 ("arm64: dts: qcom: Add LPG to pm8916, pm8994, pmi8994 and pmi8998")
>>>> Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@xxxxxxxxxx>
>>>> ---
>>>> arch/arm64/boot/dts/qcom/pm8916.dtsi | 2 +-
>>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/arch/arm64/boot/dts/qcom/pm8916.dtsi b/arch/arm64/boot/dts/qcom/pm8916.dtsi
>>>> index 606c2a6d1f0fc..d6922379729cb 100644
>>>> --- a/arch/arm64/boot/dts/qcom/pm8916.dtsi
>>>> +++ b/arch/arm64/boot/dts/qcom/pm8916.dtsi
>>>> @@ -124,7 +124,7 @@ pm8916_1: pmic@1 {
>>>> #address-cells = <1>;
>>>> #size-cells = <0>;
>>>>
>>>> - pm8916_pwm: pwm {
>>>> + pm8916_pwm: pwm@bc00 {
>>>> compatible = "qcom,pm8916-pwm";
>>>
>>> This does not look like proper fix.
>>> 1. It requires a reg.
>>> 2. reg is not allowed by pwm bindings.
>>>
>>> See also other wrong commit:
>>> https://lore.kernel.org/all/20220719205058.1004942-1-bhupesh.sharma@xxxxxxxxxx/
>>>
>>
>> and this:
>> https://lore.kernel.org/all/20220721195502.1525214-1-bhupesh.sharma@xxxxxxxxxx/
>>
>>
>> Best regards,
>> Krzysztof
>
> Hmm OK.
>
> So if I've understood this conversation properly instead of converting
> "label: nodename" to "label: nodename@xxx" and adding reg, we would just
> revert to "nodename {};"
One of two:
1. Yes and fix the bindings.
2. Change the bindings to accept multiple reg, add unit addresses and
multiple reg's to match real HW.
In any case, all in one patchset, so we see the impact. It seems we all
are fine with both approaches, so maybe let's choose less work?
Best regards,
Krzysztof
