Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20210110035846.9155-1-wens@kernel.org> <2241380.NG923GbCHz@diego>
 <CAGb2v67h3v4DdEY68BmHy263=L1tB0DuJnf_exHrLi57SiRjUA@mail.gmail.com> <2633022.BEx9A2HvPv@diego>
In-Reply-To: <2633022.BEx9A2HvPv@diego>
From:   Chen-Yu Tsai <wens@kernel.org>
Date:   Mon, 11 Jan 2021 16:04:56 +0800
Message-ID: <CAGb2v678BgP_23oz+XMzw5B0U2hBy_fW9VK9mvVVhoXqchJmTg@mail.gmail.com>
Subject: Re: [PATCH 3/3] arm64: dts: rockchip: rk3328: Add Radxa ROCK Pi E
To:     =?UTF-8?Q?Heiko_St=C3=BCbner?= <heiko@sntech.de>
Cc:     Chen-Yu Tsai <wens@kernel.org>, Johan Jonker <jbx6244@gmail.com>,
        Rob Herring <robh@kernel.org>,
        "open list:ARM/Rockchip SoC..." <linux-rockchip@lists.infradead.org>,
        linux-kernel <linux-kernel@vger.kernel.org>,
        linux-arm-kernel <linux-arm-kernel@lists.infradead.org>,
        devicetree <devicetree@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Mon, Jan 11, 2021 at 3:50 PM Heiko St=C3=BCbner <heiko@sntech.de> wrote:
>
> Am Montag, 11. Januar 2021, 04:27:47 CET schrieb Chen-Yu Tsai:
> > On Mon, Jan 11, 2021 at 4:06 AM Heiko St=C3=BCbner <heiko@sntech.de> wr=
ote:
> > >
> > > Hi,
> > >
> > > Am Sonntag, 10. Januar 2021, 16:37:15 CET schrieb Chen-Yu Tsai:
> > > > > > +     vcc_sd: sdmmc-regulator {
> > > > > > +             compatible =3D "regulator-fixed";
> > > > > > +             gpio =3D <&gpio0 RK_PD6 GPIO_ACTIVE_LOW>;
> > > > > > +             pinctrl-names =3D "default";
> > > > > > +             pinctrl-0 =3D <&sdmmc0m1_pin>;
> > > > >
> > > > > > +             regulator-boot-on;
> > > > > > +             regulator-name =3D "vcc_sd";
> > > > >
> > > > > regulator-name above other regulator properties
> > > >
> > > > That is actually what I was used to, but some other rockchip dts fi=
les
> > > > have all the properties sorted alphabetically. So I stuck with what=
 I
> > > > saw.
> > >
> > > I try to keep it alphabetical except for the exceptions :-D .
> > >
> > > regulator-name is such an exception. Similar to compatibles, the
> > > regulator-name is an entry needed to see if you're at the right node,
> > > so I really like it being the topmost regulator-foo property - just m=
akes
> > > reading easier.
> > >
> > > (same for the compatible first, then regs, interrupts parts, as well
> > > as "status-last")
> > >
> > > But oftentimes, I just fix the ordering when applying - but seem to h=
ave
> > > missed this somewhere in those "other Rockchip dts files" ;-) .
> >
> > I was slightly confused. I looked again and yes regulator-name is alway=
s the
> > first regulator related property. What's off is that in some cases min/=
max
> > voltage comes before always-on/boot-on, and in others vice versa.
> >
> > For example in the Rock64 and ROC-RK3328-CC device trees, in the fixed
> > regulators, always-on/boot-on come before min/max voltage, while in the
> > PMIC the other order is used.
>
> That's likely undecidednes on my part ;-)
>
> There could be an argument for a "name, voltages, flags" sorting, but on
> the other hand just keeping it alphabetical with the naming on top
> creates less special cases.

And min before max? :D