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: <20200428142629.8950-1-peron.clem@gmail.com> <98246e5d-ebef-bcb5-f0b8-d74b3834b835@arm.com>
 <CAJiuCcco0d_UoWeHqh6oc0rFNAMFynXjLRQ6APsT0WBh7m+GQg@mail.gmail.com> <20200428164522.p7ypca7zwocc7alq@gilmour.lan>
In-Reply-To: <20200428164522.p7ypca7zwocc7alq@gilmour.lan>
From:   =?UTF-8?B?Q2zDqW1lbnQgUMOpcm9u?= <peron.clem@gmail.com>
Date:   Thu, 30 Apr 2020 15:48:04 +0200
Message-ID: <CAJiuCce9UDp--XQ=rXPZ5cZyNDMFC3zyq7pnw3ETXkR3=zMWaQ@mail.gmail.com>
Subject: Re: [linux-sunxi] Re: [PATCH v2] arm64: dts: allwinner: h6: Use dummy
 regulator for Tanix TX6
To:     Maxime Ripard <maxime@cerno.tech>
Cc:     Robin Murphy <robin.murphy@arm.com>, Chen-Yu Tsai <wens@csie.org>,
        Rob Herring <robh+dt@kernel.org>,
        devicetree <devicetree@vger.kernel.org>,
        linux-kernel <linux-kernel@vger.kernel.org>,
        linux-sunxi <linux-sunxi@googlegroups.com>,
        Piotr Oniszczuk <warpme@o2.pl>,
        linux-arm-kernel <linux-arm-kernel@lists.infradead.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Maxime,

On Tue, 28 Apr 2020 at 18:45, Maxime Ripard <maxime@cerno.tech> wrote:
>
> On Tue, Apr 28, 2020 at 06:23:35PM +0200, Cl=C3=A9ment P=C3=A9ron wrote:
> > Hi Robin,
> >
> > On Tue, 28 Apr 2020 at 17:21, Robin Murphy <robin.murphy@arm.com> wrote=
:
> > >
> > > On 2020-04-28 3:26 pm, Cl=C3=A9ment P=C3=A9ron wrote:
> > > > Tanix TX6 has a fixed regulator. As DVFS is instructed to change
> > > > voltage to meet OPP table, the DVFS is not working as expected.
> > >
> > > Hmm, isn't that really a bug in the DVFS code? I guess it's just blin=
dly
> > > propagating -EINVAL from the fixed regulators not implementing
> > > set_voltage, but AFAICS it has no real excuse not to be cleverer and
> > > still allow switching frequency as long as the voltage *is* high enou=
gh
> > > for the given OPP. I wonder how well it works if the regulator is
> > > programmable but shared with other consumers... that case probably ca=
n't
> > > be hacked around in DT.
> >
> > Like you, I thought that the DVFS was clever enough to understand this
> > but guess not..
> >
> > Maybe they are some cases where you don't want to leave the voltage hig=
h and
> > reduce the frequency. But I don't know such case.
>
> I assume the intent was to prevent a regulator driver to overshoot and en=
d up
> over-volting the CPU which would be pretty bad.
>
> I guess we could check that the voltage is in the range opp < actual volt=
age <
> max opp voltage ?

As this could take more time than expected,

Could you drop the commit :
add1e27fb703f65f33191ccc70dd9d811254387c
arm64: dts: allwinner: h6: Enable CPU opp tables for Tanix TX6

Thanks,
Clement

>
> Maxime