Re: [PATCH v3 3/3] arm64: dts: qcom: sdm845-oneplus: use generic fuel gauge compatible

[David Heidelberg]

On 11/09/2026 15:37, Andrew Davis wrote:
> On 9/11/26 3:21 AM, Krzysztof Kozlowski wrote:
> > On Thu, Sep 10, 2026 at 02:47:44AM +0300, Rinat Muhamedgaliev wrote:
> > > OnePlus 6 and OnePlus 6T battery packs may contain either a TI
> > > bq27411 or bq27541 fuel gauge. Use the generic ti,bq27xxx compatible
> > > so the driver can select the register map from DeviceType at probe
> > > time.
> > >
> > > Tested on OnePlus 6T (fajita) with a bq27541 (DeviceType 0x0541):
> > > after booting the new kernel, the battery reported 4.360 V, 100% state
> > > of charge, 27.8 C, and Full status.
> > >
> > > Fixes: 12dfb002ca01 ("arm64: dts: qcom: sdm845-oneplus-*: add fuel gauge")
> >
> > What is the bug here? Your commit msg should be specific about it.
>
> Commit msg seems pretty clear about the bug (to me at least), the DT
> says the fuel gauge is a BQ27411 but that isn't the case, some shipped
> with a BQ27541 fuel gauge.
>
> (I do agree though this shouldn't have a fixes tag as back-porting this
> without also back-porting the driver-side change would break users)
>
> Since these fuel gauges often live on the battery pack and not the
> device itself, folks changing their batteries might end up with
> different fuel gauges, so having some detection method is going to
> be helpful.
>
> Rinat,
>
> I'm not familiar with the bootflow on these devices, but is this something
> you could detect in the bootloader and fixup the DT before it is passed
> to kernel? That way the kernel gets a DT with the correct and specific
> compatible to start with?

Hello Andrew:

There are two possible bootflows:
a) from fastboot (we can't do anything)
b) from intermediate u-boot (we could do this)

While your suggestion seems to be the most correct, I'm not sure the 
"complexity" here outweighs the pros of simply reporting

deb_warn your phones has different charging chip

and switching the behavior to the "replacement part" here.

In the future we'll need to add u-boot logic because of camera focus, where the 
focus in some phones is on different i2c addr + from different vendor, where we 
cannot do it so simply as here.

What do you think?

David>
> Andrew
>
> > > Signed-off-by: Rinat Muhamedgaliev <rinat.muhamedgaliev@xxxxxxxxx>
> > > ---
> > >   arch/arm64/boot/dts/qcom/sdm845-oneplus-common.dtsi   | 4 ++--
> > >   arch/arm64/boot/dts/qcom/sdm845-oneplus-enchilada.dts | 2 +-
> > >   arch/arm64/boot/dts/qcom/sdm845-oneplus-fajita.dts    | 2 +-
> > >   3 files changed, 4 insertions(+), 4 deletions(-)
> > >
> > > diff --git a/arch/arm64/boot/dts/qcom/sdm845-oneplus-common.dtsi b/arch/ 
> > > arm64/boot/dts/qcom/sdm845-oneplus-common.dtsi
> > > index 948a0dd..8c5cd94 100644
> > > --- a/arch/arm64/boot/dts/qcom/sdm845-oneplus-common.dtsi
> > > +++ b/arch/arm64/boot/dts/qcom/sdm845-oneplus-common.dtsi
> > > @@ -790,8 +790,8 @@ &i2c10 {
> > >       status = "okay";
> > >       clock-frequency = <100000>;
> > > -    bq27441_fg: bq27441-battery@55 {
> > > -        compatible = "ti,bq27411";
> > > +    bq27xxx_fg: fuel-gauge@55 {
> > > +        compatible = "ti,bq27xxx";
> >
> > And this is just breaking all users without any explanation.
> >
> >
> >
> > >           status = "okay";
> > >           reg = <0x55>;
> > >       };
> > > diff --git a/arch/arm64/boot/dts/qcom/sdm845-oneplus-enchilada.dts b/arch/ 
> > > arm64/boot/dts/qcom/sdm845-oneplus-enchilada.dts
> > > index 5f01115..cf40842 100644
> > > --- a/arch/arm64/boot/dts/qcom/sdm845-oneplus-enchilada.dts
> > > +++ b/arch/arm64/boot/dts/qcom/sdm845-oneplus-enchilada.dts
> > > @@ -44,7 +44,7 @@ main_cam_ois: main-cam-ois-regulator {
> > >       };
> > >   };
> > > -&bq27441_fg {
> > > +&bq27xxx_fg {
> >
> > No, irrelevant to the supposed bug here.
> >
> > Best regards,
> > Krzysztof