Re: [PATCH v2] arm64: dts: qcom: sm8550: add Soundwire controllers

[Krzysztof Kozlowski]

On 04/04/2023 19:55, Konrad Dybcio wrote:
> 
> 
> On 4.04.2023 09:39, Krzysztof Kozlowski wrote:
>> Add nodes for LPASS Soundwire v2.0.0 controllers.  Use labels with
>> indices matching downstream DTS, to make any comparisons easier.
>>
>> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx>
>>
>> ---
>>
>> Changes since v1:
>> 1. Correct IO range length.
>>
>> The bindings and driver are here:
>> https://lore.kernel.org/linux-arm-msm/20230403132503.62090-1-krzysztof.kozlowski@xxxxxxxxxx/T/#t
>> ---
>>  arch/arm64/boot/dts/qcom/sm8550.dtsi | 109 +++++++++++++++++++++++++++
>>  1 file changed, 109 insertions(+)
>>
>> diff --git a/arch/arm64/boot/dts/qcom/sm8550.dtsi b/arch/arm64/boot/dts/qcom/sm8550.dtsi
>> index dc6150e97d46..14e4de7a1a0e 100644
>> --- a/arch/arm64/boot/dts/qcom/sm8550.dtsi
>> +++ b/arch/arm64/boot/dts/qcom/sm8550.dtsi
>> @@ -2004,6 +2004,33 @@ lpass_wsa2macro: codec@6aa0000 {
>>  			#sound-dai-cells = <1>;
>>  		};
>>  
>> +		/* WSA2 */
>> +		swr3: soundwire-controller@6ab0000 {
>> +			compatible = "qcom,soundwire-v2.0.0";
>> +			reg = <0 0x06ab0000 0 0x10000>;
>> +			interrupts = <GIC_SPI 171 IRQ_TYPE_LEVEL_HIGH>;
>> +			clocks = <&lpass_wsa2macro>;
>> +			clock-names = "iface";
>> +
> Any reason this and swr0 have comments instead of labels?

Actually no, I can go with label consistently. It's not actually used
for anything...

> 
>> +			qcom,din-ports = <4>;
>> +			qcom,dout-ports = <9>;
>> +
>> +			qcom,ports-sinterval =		<0x07 0x1f 0x3f 0x07 0x1f 0x3f 0x18f 0xff 0xff 0x0f 0x0f 0xff 0x31f>;
>> +			qcom,ports-offset1 =		/bits/ 8 <0x01 0x03 0x05 0x02 0x04 0x15 0x00 0xff 0xff 0x06 0x0d 0xff 0x00>;
>> +			qcom,ports-offset2 =		/bits/ 8 <0xff 0x07 0x1f 0xff 0x07 0x1f 0xff 0xff 0xff 0xff 0xff 0xff 0xff>;
>> +			qcom,ports-hstart =		/bits/ 8 <0xff 0xff 0xff 0xff 0xff 0xff 0x08 0xff 0xff 0xff 0xff 0xff 0x0f>;
>> +			qcom,ports-hstop =		/bits/ 8 <0xff 0xff 0xff 0xff 0xff 0xff 0x08 0xff 0xff 0xff 0xff 0xff 0x0f>;
>> +			qcom,ports-word-length =	/bits/ 8 <0xff 0xff 0xff 0xff 0xff 0xff 0x08 0xff 0xff 0xff 0xff 0xff 0x18>;
>> +			qcom,ports-block-pack-mode =	/bits/ 8 <0x00 0x01 0x01 0x00 0x01 0x01 0x00 0x00 0x00 0x01 0x01 0x00 0x00>;
>> +			qcom,ports-block-group-count =	/bits/ 8 <0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff>;
>> +			qcom,ports-lane-control =	/bits/ 8 <0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff 0xff>;
>> +
>> +			#address-cells = <2>;
>> +			#size-cells = <0>;
>> +			#sound-dai-cells = <1>;
>> +			status = "disabled";
>> +		};
> [...]
> 
> 
>> +		swr2: soundwire-controller@6d30000 {
>> +			compatible = "qcom,soundwire-v2.0.0";
>> +			reg = <0 0x06d30000 0 0x10000>;
>> +			interrupts-extended = <&intc GIC_SPI 496 IRQ_TYPE_LEVEL_HIGH>,
>> +					      <&intc GIC_SPI 520 IRQ_TYPE_LEVEL_HIGH>;
> There's no different interrupt-parent, so -extended seems excessive.
> Unless that was maybe supposed to use some PDC-mapped irq?
> 

Eh, apparently copy-pasta. Thanks,


Best regards,
Krzysztof