[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ZHhHhk/4BX53R1p/@matsya>
Date: Thu, 1 Jun 2023 12:53:50 +0530
From: Vinod Koul <vkoul@...nel.org>
To: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
Cc: Bjorn Andersson <andersson@...nel.org>,
linux-arm-msm@...r.kernel.org,
Konrad Dybcio <konrad.dybcio@...aro.org>,
Georgi Djakov <djakov@...nel.org>,
Rob Herring <robh+dt@...nel.org>,
Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
Conor Dooley <conor+dt@...nel.org>, linux-pm@...r.kernel.org,
devicetree@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 03/15] dt-bindings: interconnect: split SC8180x to own
schema
On 31-05-23, 10:22, Krzysztof Kozlowski wrote:
> On 30/05/2023 18:24, Vinod Koul wrote:
> > SC8180x comes with interconnects with missing IO address space and
> > variable number of clocks, so split it from common file for easier
> > maintenance and to fix warnings like:
> >
> > sc8180x-lenovo-flex-5g.dtb: interconnect-0: 'reg' is a required property
> >
> > Signed-off-by: Vinod Koul <vkoul@...nel.org>
> > ---
> > .../bindings/interconnect/qcom,rpmh.yaml | 11 -----
> > .../interconnect/qcom,sc8180x-rpmh.yaml | 49 +++++++++++++++++++
> > 2 files changed, 49 insertions(+), 11 deletions(-)
> > create mode 100644 Documentation/devicetree/bindings/interconnect/qcom,sc8180x-rpmh.yaml
> >
> > diff --git a/Documentation/devicetree/bindings/interconnect/qcom,rpmh.yaml b/Documentation/devicetree/bindings/interconnect/qcom,rpmh.yaml
> > index 4d93ad415e0b..5cbc3be49e99 100644
> > --- a/Documentation/devicetree/bindings/interconnect/qcom,rpmh.yaml
> > +++ b/Documentation/devicetree/bindings/interconnect/qcom,rpmh.yaml
> > @@ -39,17 +39,6 @@ properties:
> > - qcom,sc7180-npu-noc
> > - qcom,sc7180-qup-virt
> > - qcom,sc7180-system-noc
> > - - qcom,sc8180x-aggre1-noc
> > - - qcom,sc8180x-aggre2-noc
> > - - qcom,sc8180x-camnoc-virt
> > - - qcom,sc8180x-compute-noc
> > - - qcom,sc8180x-config-noc
> > - - qcom,sc8180x-dc-noc
> > - - qcom,sc8180x-gem-noc
> > - - qcom,sc8180x-mc-virt
> > - - qcom,sc8180x-mmss-noc
> > - - qcom,sc8180x-qup-virt
> > - - qcom,sc8180x-system-noc
> > - qcom,sdm670-aggre1-noc
> > - qcom,sdm670-aggre2-noc
> > - qcom,sdm670-config-noc
> > diff --git a/Documentation/devicetree/bindings/interconnect/qcom,sc8180x-rpmh.yaml b/Documentation/devicetree/bindings/interconnect/qcom,sc8180x-rpmh.yaml
> > new file mode 100644
> > index 000000000000..b182c2c5addc
> > --- /dev/null
> > +++ b/Documentation/devicetree/bindings/interconnect/qcom,sc8180x-rpmh.yaml
> > @@ -0,0 +1,49 @@
> > +# SPDX-License-Identifier: GPL-2.0-only OR BSD-2-Clause
> > +%YAML 1.2
> > +---
> > +$id: http://devicetree.org/schemas/interconnect/qcom,sc8180x-rpmh.yaml#
> > +$schema: http://devicetree.org/meta-schemas/core.yaml#
> > +
> > +title: Qualcomm RPMh Network-On-Chip Interconnect on SC8180X
> > +
> > +maintainers:
> > + - Bjorn Andersson <andersson@...nel.org>
> > + - Konrad Dybcio <konrad.dybcio@...aro.org>
> > +
> > +description: |
> > + RPMh interconnect providers support system bandwidth requirements through
> > + RPMh hardware accelerators known as Bus Clock Manager (BCM).
> > +
> > + See also:: include/dt-bindings/interconnect/qcom,sc8180x.h
> > +
> > +properties:
> > + compatible:
> > + enum:
> > + - qcom,sc8180x-aggre1-noc
> > + - qcom,sc8180x-aggre2-noc
> > + - qcom,sc8180x-camnoc-virt
> > + - qcom,sc8180x-compute-noc
> > + - qcom,sc8180x-config-noc
> > + - qcom,sc8180x-dc-noc
> > + - qcom,sc8180x-gem-noc
> > + - qcom,sc8180x-ipa-virt
> > + - qcom,sc8180x-mc-virt
> > + - qcom,sc8180x-mmss-noc
> > + - qcom,sc8180x-qup-virt
> > + - qcom,sc8180x-system-noc
> > +
> > +required:
> > + - compatible
> > +
> > +allOf:
> > + - $ref: qcom,rpmh-common.yaml#
>
> This should be based on sc7280. You need reg for some of the entries.
Yes missed reg part here.
Question is should it be added sc7280 or have a different file, I think
doing former is better
>
> > +
> > +unevaluatedProperties: false
> > +
> > +examples:
> > + - |
> > + interconnect-0 {
>
> Just "interconnect"
Nope. We have multiple virtual nodes here so interconnect-N would
need to be done
--
~Vinod
Powered by blists - more mailing lists