devicetree.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Ezequiel Garcia <ezequiel@collabora.com>
To: Johan Jonker <jbx6244@gmail.com>,
	devicetree@vger.kernel.org, linux-rockchip@lists.infradead.org
Cc: Rob Herring <robh+dt@kernel.org>,
	Heiko Stuebner <heiko@sntech.de>,
	Kever Yang <kever.yang@rock-chips.com>,
	Benjamin Gaignard <benjamin.gaignard@st.com>,
	Peter Geis <pgwipeout@gmail.com>
Subject: Re: [PATCH 3/4] arm64: dts: rockchip: Add GPU node for rk3568
Date: Wed, 04 Aug 2021 23:56:30 -0300	[thread overview]
Message-ID: <c52b31a6e66f46ad2b14ee96868e5c6b212d2b7f.camel@collabora.com> (raw)
In-Reply-To: <b19aa876-f78b-f6af-39e9-ff37fc19f09a@gmail.com>

Hi Johan,

Thanks for the review.

On Fri, 2021-07-30 at 21:38 +0200, Johan Jonker wrote:
> Hi Ezequiel,
> 
> Some comments. Have a look if it's useful.
> 
> On 7/30/21 6:45 PM, Ezequiel Garcia wrote:
> > Rockchip SoCs RK3566 and RK3568 have a Mali Gondul core
> > which is based on the Bifrost architecture. It has
> > one shader core and two execution engines.
> > 
> > Quoting the datasheet:
> > 
> > Mali-G52 1-Core-2EE
> > * Support 1600Mpix/s fill rate when 800MHz clock frequency
> > * Support 38.4GLOPs when 800MHz clock frequency
> > 
> > Signed-off-by: Ezequiel Garcia <ezequiel@collabora.com>
> > ---
> >  arch/arm64/boot/dts/rockchip/rk356x.dtsi | 47 ++++++++++++++++++++++++
> >  1 file changed, 47 insertions(+)
> > 
> > diff --git a/arch/arm64/boot/dts/rockchip/rk356x.dtsi b/arch/arm64/boot/dts/rockchip/rk356x.dtsi
> > index bef747fb1fe2..f4f400792659 100644
> > --- a/arch/arm64/boot/dts/rockchip/rk356x.dtsi
> > +++ b/arch/arm64/boot/dts/rockchip/rk356x.dtsi
> > @@ -121,6 +121,35 @@ opp-1800000000 {
> >                 };
> >         };
> >  
> 
> > +       gpu_opp_table: opp-table2 {
> 
>         gpu_opp_table: gpu-opp-table {
> 

Makes sense.

> > +               compatible = "operating-points-v2";
> > +
> > +               opp-200000000 {
> > +                       opp-hz = /bits/ 64 <200000000>;
> > +                       opp-microvolt = <825000>;
> > +               };
> 
> Similar to cpu0_opp_table keep the same style and add an empty line
> between nodes.
> 

Makes sense.

> > +               opp-300000000 {
> > +                       opp-hz = /bits/ 64 <300000000>;
> > +                       opp-microvolt = <825000>;
> > +               };
> > +               opp-400000000 {
> > +                       opp-hz = /bits/ 64 <400000000>;
> > +                       opp-microvolt = <825000>;
> > +               };
> > +               opp-600000000 {
> > +                       opp-hz = /bits/ 64 <600000000>;
> > +                       opp-microvolt = <825000>;
> > +               };
> > +               opp-700000000 {
> > +                       opp-hz = /bits/ 64 <700000000>;
> > +                       opp-microvolt = <900000>;
> > +               };
> > +               opp-800000000 {
> > +                       opp-hz = /bits/ 64 <800000000>;
> > +                       opp-microvolt = <1000000>;
> > +               };
> > +       };
> > +
> >         firmware {
> >                 scmi: scmi {
> >                         compatible = "arm,scmi-smc";
> > @@ -332,6 +361,24 @@ power-domain@RK3568_PD_RKVENC {
> >                 };
> >         };
> >  
> > +       gpu: gpu@fde60000 {
> > +               compatible = "rockchip,rk3568-mali", "arm,mali-bifrost";
> > +               reg = <0x0 0xfde60000 0x0 0x4000>;
> 
> > +
> 
> remove empty lines
> 

Makes sense.

> > +               interrupts = <GIC_SPI 40 IRQ_TYPE_LEVEL_HIGH>,
> > +                            <GIC_SPI 41 IRQ_TYPE_LEVEL_HIGH>,
> > +                            <GIC_SPI 39 IRQ_TYPE_LEVEL_HIGH>;
> > +               interrupt-names = "job", "mmu", "gpu";
> 
> > +
> 
> dito
> 
> > +               clocks = <&scmi_clk 1>, <&cru CLK_GPU>;
> > +               clock-names = "core", "bus";
> 
> Not sure if it's possible, but could you keep them all a little bit in
> the same order/style as arm,mali-400?
> 
> From arm,mali-utgard.yaml:
> 
>   clock-names:
>     items:
>       - const: bus
>       - const: core
> 

Don't think we can do that in this case, as the first clock
is already expected as the "core" one.

These clocks might need some cleaner specification in the
driver or the dt-binding (or both), but I'd like to defer
that to the driver maintainers.


-- 
Kindly,
Ezequiel


  reply	other threads:[~2021-08-05  2:56 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-30 16:45 [PATCH v2 0/4] RK3568 GPU Ezequiel Garcia
2021-07-30 16:45 ` [PATCH 1/4] dt-bindings: gpu: mali-bifrost: Allow up to two clocks Ezequiel Garcia
2021-07-30 16:45 ` [PATCH 2/4] dt-bindings: gpu: mali-bifrost: Add RK3568 compatible Ezequiel Garcia
2021-07-30 16:45 ` [PATCH 3/4] arm64: dts: rockchip: Add GPU node for rk3568 Ezequiel Garcia
2021-07-30 19:38   ` Johan Jonker
2021-08-05  2:56     ` Ezequiel Garcia [this message]
2021-07-30 16:45 ` [PATCH 4/4] arm64: dts: rockchip: Enable the GPU on Quartz64 Model A Ezequiel Garcia
2021-07-30 18:24   ` Johan Jonker

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=c52b31a6e66f46ad2b14ee96868e5c6b212d2b7f.camel@collabora.com \
    --to=ezequiel@collabora.com \
    --cc=benjamin.gaignard@st.com \
    --cc=devicetree@vger.kernel.org \
    --cc=heiko@sntech.de \
    --cc=jbx6244@gmail.com \
    --cc=kever.yang@rock-chips.com \
    --cc=linux-rockchip@lists.infradead.org \
    --cc=pgwipeout@gmail.com \
    --cc=robh+dt@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).