[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <285d7b38-da08-e263-d664-1cdf73d7c0d1@somainline.org>
Date: Thu, 29 Jul 2021 15:26:08 +0200
From: AngeloGioacchino Del Regno
<angelogioacchino.delregno@...ainline.org>
To: Rob Herring <robh@...nel.org>
Cc: bjorn.andersson@...aro.org, agross@...nel.org,
daniel.lezcano@...aro.org, rjw@...ysocki.net,
linux-kernel@...r.kernel.org, linux-pm@...r.kernel.org,
linux-arm-msm@...r.kernel.org, phone-devel@...r.kernel.org,
konrad.dybcio@...ainline.org, marijn.suijten@...ainline.org,
martin.botka@...ainline.org, jeffrey.l.hugo@...il.com,
jami.kettunen@...ainline.org,
~postmarketos/upstreaming@...ts.sr.ht, devicetree@...r.kernel.org,
stephan@...hold.net
Subject: Re: [PATCH v7 5/5] dt-bindings: soc: qcom: spm: Document SDM660 and
MSM8998 compatibles
Il 14/07/21 00:21, Rob Herring ha scritto:
> On Tue, Jun 22, 2021 at 04:11:17PM +0200, AngeloGioacchino Del Regno wrote:
>> The driver was updated to add SAW2 v4.1 support for new SoCs: document
>> the new compatibles.
>
> Can't take patches without a S-o-b. Run checkpatch.pl, it points this
> out for you.
>
I am truly sorry for missing my S-o-b.
>> ---
>> .../bindings/soc/qcom/qcom,spm.yaml | 21 +++++++++++++++++++
>> 1 file changed, 21 insertions(+)
>>
>> diff --git a/Documentation/devicetree/bindings/soc/qcom/qcom,spm.yaml b/Documentation/devicetree/bindings/soc/qcom/qcom,spm.yaml
>> index 4aaa319b2932..0faf52700dec 100644
>> --- a/Documentation/devicetree/bindings/soc/qcom/qcom,spm.yaml
>> +++ b/Documentation/devicetree/bindings/soc/qcom/qcom,spm.yaml
>> @@ -17,6 +17,10 @@ description: |
>> properties:
>> compatible:
>> enum:
>> + - qcom,sdm660-gold-saw2-v4.1-l2
>> + - qcom,sdm660-silver-saw2-v4.1-l2
>> + - qcom,msm8998-gold-saw2-v4.1-l2
>> + - qcom,msm8998-silver-saw2-v4.1-l2
>
> What's the difference between gold and silver? Are the h/w instances
> different (I realize the CPUs are) in some way? How does the OS use the
> different compatible strings?
>
They have different configuration parameters and the HW instances should
be different indeed (at least from what I remember), plus they're always
at different iostart.
The driver is using the different compatible strings to choose which
configuration gets written. You can also avoid writing the configuration
to one of them or both (if you wish to lose capabilities given by this
driver, perhaps also if you want to simply never use the gold cluster, for
example).
>> - qcom,msm8974-saw2-v2.1-cpu
>> - qcom,apq8084-saw2-v2.1-cpu
>> - qcom,apq8064-saw2-v1.1-cpu
>> @@ -33,6 +37,8 @@ additionalProperties: false
>>
>> examples:
>> - |
>> +
>> + /* Example 1: SoC using SAW2 and kpss-acc-v2 CPUIdle */
>> cpus {
>> #address-cells = <1>;
>> #size-cells = <0>;
>> @@ -52,4 +58,19 @@ examples:
>> reg = <0xf9089000 0x1000>;
>> };
>>
>> + - |
>> +
>> + /* Example 2: New-gen multi cluster SoC using SAW only for L2;
>> + * This does not require any cpuidle driver, nor any cpu phandle.
>> + */
>> + power-controller@...12000 {
>> + compatible = "qcom,msm8998-gold-saw2-v4.1-l2", "qcom,saw2";
>> + reg = <0x17812000 0x1000>;
>> + };
>> +
>> + power-controller@...12000 {
>> + compatible = "qcom,msm8998-silver-saw2-v4.1-l2", "qcom,saw2";
>> + reg = <0x17912000 0x1000>;
>> + };
>> +
>> ...
>> --
>> 2.32.0
>>
>>
Powered by blists - more mailing lists