On Mon, May 06, 2019 at 03:31:01PM +0200, Ondřej Jirman wrote:
> > > > ---
> > > > Changes for v2:
> > > > - emac changes on top of https://patchwork.kernel.org/cover/10899529/
> > > >   series
> > > >
> > > >  .../allwinner/sun50i-h6-orangepi-one-plus.dts |  8 ++++
> > > >  .../dts/allwinner/sun50i-h6-orangepi.dtsi     | 42 +++++++++++++++++++
> > > >  2 files changed, 50 insertions(+)
> > > >
> > > > diff --git a/arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi-one-plus.dts b/arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi-one-plus.dts
> > > > index 12e17567ab56..9e8ed1053715 100644
> > > > --- a/arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi-one-plus.dts
> > > > +++ b/arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi-one-plus.dts
> > > > @@ -9,4 +9,12 @@
> > > >  / {
> > > >       model = "OrangePi One Plus";
> > > >       compatible = "xunlong,orangepi-one-plus", "allwinner,sun50i-h6";
> > > > +
> > > > +     aliases {
> > > > +             ethernet0 = &emac;
> > > > +     };
> > > > +};
> > > > +
> > > > +&emac {
> > > > +     status = "okay";
> > > >  };
> > > > diff --git a/arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi.dtsi b/arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi.dtsi
> > > > index 62e27948a3fa..c48e24acaf8a 100644
> > > > --- a/arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi.dtsi
> > > > +++ b/arch/arm64/boot/dts/allwinner/sun50i-h6-orangepi.dtsi
> > > > @@ -45,6 +45,48 @@
> > > >               regulator-max-microvolt = <5000000>;
> > > >               regulator-always-on;
> > > >       };
> > > > +
> > > > +     /*
> > > > +      * The board uses 2.5V RGMII signalling. Power sequence to enable
> > > > +      * the phy is to enable GMAC-2V5 and GMAC-3V (aldo2) power rails
> > > > +      * at the same time and to wait 100ms.
> > > > +      */
> > > > +     reg_gmac_2v5: gmac-2v5 {
> > > > +             compatible = "regulator-fixed";
> > > > +             regulator-name = "gmac-2v5";
> > > > +             regulator-min-microvolt = <2500000>;
> > > > +             regulator-max-microvolt = <2500000>;
> > > > +             startup-delay-us = <100000>;
> > > > +             enable-active-high;
> > > > +             gpio = <&pio 3 6 GPIO_ACTIVE_HIGH>; /* GMAC_EN: PD6 */
> > > > +
> > > > +             /* The real parent of gmac-2v5 is reg_vcc5v, but we need to
> > > > +              * enable two regulators to power the phy. This is one way
> > > > +              * to achieve that.
> > > > +              */
> > > > +             vin-supply = <&reg_aldo2>; /* VCC3V3-MAC: GMAC-3V */
> > > > +     };
> > > > +};
> > > > +
> > > > +&emac {
> > > > +     pinctrl-names = "default";
> > > > +     pinctrl-0 = <&ext_rgmii_pins>;
> > > > +     phy-mode = "rgmii";
> > > > +     phy-handle = <&ext_rgmii_phy>;
> > > > +     phy-supply = <&reg_gmac_2v5>;
> > > > +     allwinner,rx-delay-ps = <1500>;
> > > > +     allwinner,tx-delay-ps = <700>;
> > > > +};
> > > > +
> > > > +&mdio {
> > > > +     ext_rgmii_phy: ethernet-phy@1 {
> > > > +             compatible = "ethernet-phy-ieee802.3-c22";
> > > > +             reg = <1>;
> > > > +
> > > > +             reset-gpios = <&pio 3 14 GPIO_ACTIVE_LOW>; /* RGMII-RESET: PD14 */
> > > > +             reset-assert-us = <15000>;
> > > > +             reset-deassert-us = <40000>;
> > > > +     };
> > > >  };
> > >
> > > ... however, at no point in time you explain why you made that switch,
> > > and while most of the definition of the EMAC nodes is in the DTSI, you
> > > only enable it in one DTS.
> >
> > The dtsi is shared b/w 1+ and lite2 and 1+ has emac, so I enabled the
> > status directly on dts and keeping the relevant nodes on dtsi just
> > like SoC dtsi does. do I need to mention this in commit log?
>
> Lite 2 doesn't have reg_gmac_2v5 and it also doesn't have the external phy.
> But with this patch, reg_gmac_2v5 will also show up in the Lite 2's final
> DTB.
>
> Comapred to SoC dtsi, the SoC always has things that are in the dtsi, they
> are just not enabled/used by the board, but they are present on the chip.
>
> So this comes down to what the meaning of board-level dtsi should be. I
> doubt we want it to mean "a collection of stuff that may or may not be
> present on the boards that depend on it".

Agreed.

Maxime

--
Maxime Ripard, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com