Received-SPF: pass (google.com: domain of linux-kernel+bounces-23829-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
Precedence: bulk
MIME-Version: 1.0
References: <20240110112059.2498-1-quic_luoj@quicinc.com> <20240110112059.2498-2-quic_luoj@quicinc.com>
 <a42718a9-d0f9-47d9-9ee8-fb520ed2a7a8@linaro.org> <de0ad768-05fa-4bb1-bcbc-0adb28cb2257@quicinc.com>
In-Reply-To: <de0ad768-05fa-4bb1-bcbc-0adb28cb2257@quicinc.com>
From: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Date: Thu, 11 Jan 2024 18:06:42 +0200
Message-ID: <CAA8EJppeQdB4W8u0ux16pxBBwF_fpt1j-5aC0f849n9_iaaYtQ@mail.gmail.com>
Subject: Re: [PATCH 1/6] arm64: dts: qcom: ipq9574: Add PPE device tree node
To: Jie Luo <quic_luoj@quicinc.com>
Cc: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>, andersson@kernel.org, 
	konrad.dybcio@linaro.org, robh+dt@kernel.org, 
	krzysztof.kozlowski+dt@linaro.org, conor+dt@kernel.org, 
	linux-arm-msm@vger.kernel.org, devicetree@vger.kernel.org, 
	linux-kernel@vger.kernel.org, netdev@vger.kernel.org, 
	quic_kkumarcs@quicinc.com, quic_suruchia@quicinc.com, quic_soni@quicinc.com, 
	quic_pavir@quicinc.com, quic_souravp@quicinc.com, quic_linchen@quicinc.com, 
	quic_leiwei@quicinc.com
Content-Type: text/plain; charset="UTF-8"

On Thu, 11 Jan 2024 at 17:31, Jie Luo <quic_luoj@quicinc.com> wrote:
>
>
>
> On 1/10/2024 7:40 PM, Krzysztof Kozlowski wrote:
> > On 10/01/2024 12:20, Luo Jie wrote:
> >> The PPE device tree node includes the PPE initialization configurations
> >> and UNIPHY instance configuration.
> >>
> >> Ther are 3 UNIPHYs(PCS) on the platform ipq9574, which register the
> >> clock provider to output the clock for PPE port to work on the different
> >> link speed.
> >>
> >> Signed-off-by: Luo Jie <quic_luoj@quicinc.com>
> >> ---
> >>   arch/arm64/boot/dts/qcom/ipq9574.dtsi | 730 +++++++++++++++++++++++++-
> >>   1 file changed, 724 insertions(+), 6 deletions(-)
> >>
> >> diff --git a/arch/arm64/boot/dts/qcom/ipq9574.dtsi b/arch/arm64/boot/dts/qcom/ipq9574.dtsi
> >> index 810cda4a850f..5fa241e27c8b 100644
> >> --- a/arch/arm64/boot/dts/qcom/ipq9574.dtsi
> >> +++ b/arch/arm64/boot/dts/qcom/ipq9574.dtsi
> >> @@ -775,16 +775,734 @@ nsscc: nsscc@39b00000 {
> >>                               <&bias_pll_nss_noc_clk>,
> >>                               <&bias_pll_ubi_nc_clk>,
> >>                               <&gcc_gpll0_out_aux>,
> >> -                             <0>,
> >> -                             <0>,
> >> -                             <0>,
> >> -                             <0>,
> >> -                             <0>,
> >> -                             <0>,
> >> +                             <&uniphys 0>,
> >> +                             <&uniphys 1>,
> >> +                             <&uniphys 2>,
> >> +                             <&uniphys 3>,
> >> +                             <&uniphys 4>,
> >> +                             <&uniphys 5>,
> >>                               <&xo_board_clk>;
> >>                      #clock-cells = <1>;
> >>                      #reset-cells = <1>;
> >>              };
> >> +
> >> +            qcom_ppe: qcom-ppe@3a000000 {
> >
> > qcom is definitely not a generic name.
> >
> > Node names should be generic. See also an explanation and list of
> > examples (not exhaustive) in DT specification:
> > https://devicetree-specification.readthedocs.io/en/latest/chapter2-devicetree-basics.html#generic-names-recommendation
>
> Ok, will update to use a generic name in the link, Thanks for the
> guidance and the link.
> >
> >
> >> +                    compatible = "qcom,ipq9574-ppe";
> >
> > I don't see this documented. I don't see reference to posted bindings.
>
> The DT bindings patch was part of the driver series as below. This
> property was documented in the DT bindings patch. Attaching it to DTSI
> series should make it more clear. If this is fine, I will update the
> DTSI series with the DT bindings patch.
> https://lore.kernel.org/netdev/20240110142428.52026d9e@kernel.org/
>
> >
> > Please run scripts/checkpatch.pl and fix reported warnings. Some
> > warnings can be ignored, but the code here looks like it needs a fix.
> > Feel free to get in touch if the warning is not clear.
> >
> > Ignoring this warning is a sign you don't really check your patches
> > before sending.
>
> We have run the checkpatch.pl on the whole patch series including this
> device tree patch set together with PPE driver patch set.
> As mentioned above, I will add the DT bindings patch into the DTS
> series. This should help with the checkpatch issue.

This will cause even more confusion, as there will be two instances of
the dt-bindings patch. One in the driver patchset, another one in the
DT changes. You just have to specify the dependencies in the cover
letter. Another option is to wait for the bindings + driver to be
accepted, then send the DTSI changes (and again, specify the
dependency).

>
> >
> >> +                    reg = <0x3a000000 0xb00000>;
> >> +                    #address-cells = <1>;
> >> +                    #size-cells = <1>;
> >> +                    ranges;
> >
> > Put after reg.
> Ok.
>
> >
> >> +                    status = "okay";
> >
> > Drop
> Ok.
>
> >
> > All of above comments apply to your entire patchset and all places.
> >
> > Looking at code further, it does not look like suitable for mainline,
> > but copy of downstream code. That's not what we expect upstream. Please
> > go back to your bindings first. Also, I really insist you reaching out
> > to other folks to help you in this process.
> >
> > Best regards,
> > Krzysztof
> >
> We will do internal review of the gaps and update the patches as per
> your comments.
>
> Thanks for the review comments.

From the first glance, the bindings do not follow upstream principles.
You have all the settings (tdm, port config, etc) in the DT, while
they should instead go to the driver. Well, unless you expect that the
board might need to override them.

-- 
With best wishes
Dmitry