Re: [PATCH v6 1/4] dt-bindings: interconnect: qcom,msm8998-cpu-bwmon: add BWMON device

[Rajendra Nayak]

On 6/30/2022 4:53 PM, Krzysztof Kozlowski wrote:
> On 30/06/2022 13:14, Rajendra Nayak wrote:
> > On 6/29/2022 7:32 PM, Krzysztof Kozlowski wrote:
> > > Add bindings for the Qualcomm Bandwidth Monitor device providing
> > > performance data on interconnects.  The bindings describe only BWMON CPU
> > > (version 4), e.g. the instance which appeared for the first on Qualcomm
> > > MSM8998 SoC and is also used on SDM845.  This BWMON device sits between
> > > CPU and Last Level Cache Controller.
> > >
> > > Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx>
> > > Reviewed-by: Rob Herring <robh@xxxxxxxxxx>
> > > Acked-by: Georgi Djakov <djakov@xxxxxxxxxx>
> > > ---
> > >    .../interconnect/qcom,msm8998-llcc-bwmon.yaml | 85 +++++++++++++++++++
> > >    1 file changed, 85 insertions(+)
> > >    create mode 100644 Documentation/devicetree/bindings/interconnect/qcom,msm8998-llcc-bwmon.yaml
> > >
> > > diff --git a/Documentation/devicetree/bindings/interconnect/qcom,msm8998-llcc-bwmon.yaml b/Documentation/devicetree/bindings/interconnect/qcom,msm8998-llcc-bwmon.yaml
> > > new file mode 100644
> > > index 000000000000..76e09658d615
> > > --- /dev/null
> > > +++ b/Documentation/devicetree/bindings/interconnect/qcom,msm8998-llcc-bwmon.yaml
> > > @@ -0,0 +1,85 @@
> > > +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
> > > +%YAML 1.2
> > > +---
> > > +$id: http://devicetree.org/schemas/interconnect/qcom,msm8998-llcc-bwmon.yaml#
> > > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > > +
> > > +title: Qualcomm Interconnect Bandwidth Monitor
> > > +
> > > +maintainers:
> > > +  - Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx>
> > > +
> > > +description: |
> > > +  Bandwidth Monitor measures current throughput on buses between various NoC
> > > +  fabrics and provides information when it crosses configured thresholds.
> > > +
> > > +  Certain SoCs might have more than one Bandwidth Monitors, for example on SDM845::
> > > +   - Measuring the bandwidth between CPUs and Last Level Cache Controller -
> > > +     called LLCC BWMON,
> > > +   - Measuring the bandwidth between Last Level Cache Controller and memory (DDR).
> > > +
> > > +properties:
> > > +  compatible:
> > > +    oneOf:
> > > +      - items:
> > > +          - enum:
> > > +              - qcom,sdm845-llcc-bwmon
> > > +          - const: qcom,msm8998-llcc-bwmon
> > > +      - const: qcom,msm8998-llcc-bwmon       # BWMON v4
> > > +
> > > +  interconnects:
> > > +    maxItems: 1
> > > +
> > > +  interrupts:
> > > +    maxItems: 1
> > > +
> > > +  operating-points-v2: true
> > > +  opp-table: true
> > > +
> > > +  reg:
> > > +    # BWMON v4 (currently described) and BWMON v5 use one register address
> > > +    # space.  BWMON v2 uses two register spaces - not yet described.
> > > +    maxItems: 1
> > > +
> > > +required:
> > > +  - compatible
> > > +  - interconnects
> > > +  - interrupts
> > > +  - operating-points-v2
> > > +  - opp-table
> > > +  - reg
> > > +
> > > +additionalProperties: false
> > > +
> > > +examples:
> > > +  - |
> > > +    #include <dt-bindings/interconnect/qcom,sdm845.h>
> > > +    #include <dt-bindings/interrupt-controller/arm-gic.h>
> > > +
> > > +    pmu@1436400 {
> > > +        compatible = "qcom,sdm845-llcc-bwmon", "qcom,msm8998-llcc-bwmon";
> >
> > so with this compatible fallback scheme, I am trying to understand what
> > do I need to do if I have to add support for another SoC for instance.
> >
> > I just update the binding with the new SoC compatible (lets say qcom,sc7280-llcc-bwmon)
> > and in the device tree node use it as
> > 	compatible = "qcom,sc7280-llcc-bwmon", "qcom,sdm845-llcc-bwmon", "qcom,msm8998-llcc-bwmon";
> > without any updates in the driver?
>
> I expect:
> "qcom,sc7280-llcc-bwmon", "qcom,msm8998-llcc-bwmon";
> and you need to add sc7280 compatible to the driver. The actual proper
> solution in my patch would be to use msm8998 compatible in the driver,
> but I did not test MSM8998.
>
> Maybe we should switch to that anyway?

Right, looks like without it every new SoC compatible added would need a dummy
update in the driver even though you really don't need to do anything different
in the driver.