netdev - Re: [PATCH v6 4/4] arm64: dts: qcom: sc8280xp-x13s: Add bluetooth

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAKXuJqh447rZxDZ28aCiRZaL=uj5xDULhyU=HUbFVePYyz7AOw@mail.gmail.com>
Date:   Thu, 16 Mar 2023 14:17:38 -0500
From:   Steev Klimaszewski <steev@...i.org>
To:     Johan Hovold <johan@...nel.org>
Cc:     Bjorn Andersson <andersson@...nel.org>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        Rob Herring <robh+dt@...nel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Andy Gross <agross@...nel.org>,
        Konrad Dybcio <konrad.dybcio@...aro.org>,
        Marcel Holtmann <marcel@...tmann.org>,
        Johan Hedberg <johan.hedberg@...il.com>,
        Luiz Augusto von Dentz <luiz.dentz@...il.com>,
        Sven Peter <sven@...npeter.dev>, netdev@...r.kernel.org,
        devicetree@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-arm-msm@...r.kernel.org, linux-bluetooth@...r.kernel.org,
        Mark Pearson <markpearson@...ovo.com>
Subject: Re: [PATCH v6 4/4] arm64: dts: qcom: sc8280xp-x13s: Add bluetooth

Hi Johan,

On Thu, Mar 16, 2023 at 6:05 AM Johan Hovold <johan@...nel.org> wrote:
>
> On Thu, Mar 16, 2023 at 11:26:12AM +0100, Johan Hovold wrote:
> > On Wed, Mar 15, 2023 at 10:47:58PM -0500, Steev Klimaszewski wrote:
> > > The Lenovo Thinkpad X13s has a WCN6855 Bluetooth controller on uart2,
> > > add this.
> > >
> > > Signed-off-by: Steev Klimaszewski <steev@...i.org>
> > > ---
>
> > > +           vreg_s1c: smps1 {
> > > +                   regulator-name = "vreg_s1c";
> > > +                   regulator-min-microvolt = <1880000>;
> > > +                   regulator-max-microvolt = <1900000>;
> > > +                   regulator-initial-mode = <RPMH_REGULATOR_MODE_HPM>;
> > > +                   regulator-allowed-modes = <RPMH_REGULATOR_MODE_AUTO>,
> > > +                                             <RPMH_REGULATOR_MODE_RET>;
> > > +                   regulator-allow-set-load;
> >
> > So this does not look quite right still as you're specifying an initial
> > mode which is not listed as allowed.
> >
> > Also there are no other in-tree users of RPMH_REGULATOR_MODE_RET and
> > AUTO is used to switch mode automatically which seems odd to use with
> > allow-set-load.
> >
> > This regulator is in fact also used by the wifi part of the chip and as
> > that driver does not set any loads so we may end up with a regulator in
> > retention mode while wifi is in use.
> >
> > Perhaps Bjorn can enlighten us, but my guess is that this should just be
> > "intial-mode = AUTO" (or even HPM, but I have no idea where this came
> > from originally).
>
> This one probably also needs to be marked as always-on as we don't
> currently describe the fact that the wifi part also uses s1c.
>
> Johan

I couldn't remember exactly why I chose HPM, and so I recreated what
I'd done.  I looked to see what modes were available by git grepping
the kernel sources and since they are in
include/dt-bindings/regulator/qcom,rpmh-regulator.h with a comment
explaining what each mode is, I picked HPM since it starts it at the
full rated current.  As to why I chose the others... it was based on
SMPS being mentioned in that comment block.  Since I wasn't sure what
PFM is (and admittedly, did not look it up) I skipped it.

And you are right, we probably don't want to yank that regulator out
from under the wifi...  will add that in v7, so I guess for v7 we want
HPM, LPM, AUTO with AUTO being initial.  I guess I was trying to think
that RET would allow as little power usage as possible when bluetooth
isn't in use.

--steev