Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20220721221212.18491-1-prabhakar.mahadev-lad.rj@bp.renesas.com>
 <20220721221212.18491-2-prabhakar.mahadev-lad.rj@bp.renesas.com> <CAMuHMdVy0w8Y-+MegHnjdFkQ7vB1Y=jPY-9atTFT_WTVYOiGLg@mail.gmail.com>
In-Reply-To: <CAMuHMdVy0w8Y-+MegHnjdFkQ7vB1Y=jPY-9atTFT_WTVYOiGLg@mail.gmail.com>
From:   "Lad, Prabhakar" <prabhakar.csengg@gmail.com>
Date:   Fri, 22 Jul 2022 11:21:12 +0100
Message-ID: <CA+V-a8tddFVQnJjKSg9MTKB-Hh3X=2OwsU+THis+KFEfz7KB=g@mail.gmail.com>
Subject: Re: [PATCH 1/2] dt-bindings: soc: renesas: renesas,rzg2l-sysc:
 Document RZ/Five SoC
To:     Geert Uytterhoeven <geert@linux-m68k.org>
Cc:     Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>,
        Magnus Damm <magnus.damm@gmail.com>,
        Rob Herring <robh+dt@kernel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
        Linux-Renesas <linux-renesas-soc@vger.kernel.org>,
        "open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS" 
        <devicetree@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Biju Das <biju.das.jz@bp.renesas.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi Geert,

Thank you for the review.

On Fri, Jul 22, 2022 at 10:31 AM Geert Uytterhoeven
<geert@linux-m68k.org> wrote:
>
> Hi Prabhakar,
>
> On Fri, Jul 22, 2022 at 12:15 AM Lad Prabhakar
> <prabhakar.mahadev-lad.rj@bp.renesas.com> wrote:
> > Document RZ/Five (R9A07G043) SYSC bindings. SYSC block found on the
> > RZ/Five SoC is almost identical to one found on the RZ/G2L (and alike)
> > SoC's. To differentiate RZ/G2UL from RZ/Five, "-rzfive" is included in
> > the compatible string for the RZ/Five SoC as there are no interrupts
> > from the SYSC block to the core.
> >
> > Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>
>
> Thanks for your patch!
>
> > ---
> >  .../soc/renesas/renesas,rzg2l-sysc.yaml       | 56 +++++++++++++------
> >  1 file changed, 39 insertions(+), 17 deletions(-)
> >
> > diff --git a/Documentation/devicetree/bindings/soc/renesas/renesas,rzg2l-sysc.yaml b/Documentation/devicetree/bindings/soc/renesas/renesas,rzg2l-sysc.yaml
> > index ce2875c89329..bdaf05f8b29b 100644
> > --- a/Documentation/devicetree/bindings/soc/renesas/renesas,rzg2l-sysc.yaml
> > +++ b/Documentation/devicetree/bindings/soc/renesas/renesas,rzg2l-sysc.yaml
> > @@ -20,35 +20,57 @@ description:
> >  properties:
> >    compatible:
> >      enum:
> > -      - renesas,r9a07g043-sysc # RZ/G2UL
> > -      - renesas,r9a07g044-sysc # RZ/G2{L,LC}
> > -      - renesas,r9a07g054-sysc # RZ/V2L
> > +      - renesas,r9a07g043-rzfive-sysc # RZ/Five
>
> renesas,r9a07g043f-sysc?
>
Agreed.

> But I'm wondering if we really need a different compatible value?
> It looks like both blocks differ only in external wiring, so if
> anything needs to be handled differently (the removed/added registers
> are related to CPU topology), that can be inferred from the system
> topology (or even #ifdef CONFIG_{ARM64,RISCV} ;-)
>
Good point, but I wonder if we would end up in too many #ifdef
CONFIG_{ARM64,RISCV} checks. If thats OK I will stick with
"renesas,r9a07g043-sysc"

> > +      - renesas,r9a07g043-sysc        # RZ/G2UL
> > +      - renesas,r9a07g044-sysc        # RZ/G2{L,LC}
> > +      - renesas,r9a07g054-sysc        # RZ/V2L
> >
> >    reg:
> >      maxItems: 1
> >
> > -  interrupts:
> > -    items:
> > -      - description: CA55/CM33 Sleep/Software Standby Mode request interrupt
> > -      - description: CA55 Software Standby Mode release request interrupt
> > -      - description: CM33 Software Standby Mode release request interrupt
> > -      - description: CA55 ACE Asynchronous Bridge Master/Slave interface deny request interrupt
> > +  interrupts: true
> >
> > -  interrupt-names:
> > -    items:
> > -      - const: lpm_int
> > -      - const: ca55stbydone_int
> > -      - const: cm33stbyr_int
> > -      - const: ca55_deny
> > +  interrupt-names: true
> >
> >  required:
> >    - compatible
> >    - reg
> > -  - interrupts
> > -  - interrupt-names
> >
> >  additionalProperties: false
> >
> > +allOf:
> > +  - if:
> > +      not:
> > +        properties:
> > +          compatible:
> > +            contains:
> > +              enum:
> > +                - renesas,r9a07g043-rzfive-sysc
> > +    then:
> > +      properties:
> > +        interrupts:
> > +          items:
> > +            - description: CA55/CM33 Sleep/Software Standby Mode request interrupt
> > +            - description: CA55 Software Standby Mode release request interrupt
> > +            - description: CM33 Software Standby Mode release request interrupt
> > +            - description: CA55 ACE Asynchronous Bridge Master/Slave interface deny request interrupt
> > +
> > +        interrupt-names:
> > +          items:
> > +            - const: lpm_int
> > +            - const: ca55stbydone_int
> > +            - const: cm33stbyr_int
> > +            - const: ca55_deny
> > +
> > +      required:
> > +        - interrupts
> > +        - interrupt-names
> > +
> > +    else:
> > +      properties:
> > +        interrupts: false
> > +        interrupt-names: false
>
> Do all interrupts{,-names} have to be moved?
> Wouldn't it be sufficient to just have
>
Agreed.

>     if [...]
>     then:
>         required:
>           - interrupts
>           - interrupt-names
>     else:
>         properties:
>             interrupts: false
>             interrupt-names: false
>
> ?
>
> But again, without a new compatible value, you could just make
> interrupts{,-names} not required?
>
You mean we just make it optional for all the SoC's?

Cheers,
Prabhakar