Re: [PATCH V2 1/3] arm64: dts: qcom: monaco: Move eMMC CQE support from SoC to board DT

[Monish Chunara <monish.chunara@oss.qualcomm.com>]

On Tue, Jun 16, 2026 at 01:14:04PM +0000, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
> - [High] Incomplete genericization of `sdhc_1` leaves conflicting eMMC-specific properties in the SoC DT, breaking SD card compatibility.
> - [Medium] Removing the `supports-cqe` property from the SoC-level `monaco.dtsi` without explicitly adding it back to all eMMC-based boards causes a silent performance regression by disabling Command Queue Engine (CQE) support.
> --
> 
> commit 05fd7920ab031fea94cd493a589a30892353cf08
> Author: Monish Chunara <monish.chunara@oss.qualcomm.com>
> 
> arm64: dts: qcom: monaco: Move eMMC CQE support from SoC to board DT
> 
> This commit removes the supports-cqe property from the generic Monaco SoC
> device tree and moves it to the qcs8300-ride board device tree. The stated
> goal is to allow SD card compatibility in the base SoC definition, since the
> Command Queue Engine conflicts with SD card operation.
> 
> > diff --git a/arch/arm64/boot/dts/qcom/monaco.dtsi b/arch/arm64/boot/dts/qcom/monaco.dtsi
> > index 7b1d57460f1e6..e5f316897706b 100644
> > --- a/arch/arm64/boot/dts/qcom/monaco.dtsi
> > +++ b/arch/arm64/boot/dts/qcom/monaco.dtsi
> > @@ -4828,7 +4828,6 @@ &mc_virt SLAVE_EBI1 QCOM_ICC_TAG_ALWAYS>,
> >  			qcom,dll-config = <0x000f64ee>;
> >  			qcom,ddr-config = <0x80040868>;
> >  			bus-width = <8>;
> > -			supports-cqe;
> 
> [Severity: Medium]
> Removing supports-cqe from the SoC DT without adding it back to all
> eMMC-based boards appears to cause a performance regression by disabling the
> Command Queue Engine for those boards.
> 
> For example, arch/arm64/boot/dts/qcom/monaco-monza-som.dtsi and
> arch/arm64/boot/dts/qcom/monaco-evk.dts both configure sdhc_1 as an eMMC
> device (using no-sd and non-removable). Since they rely on the SoC DT for
> the supports-cqe property, won't omitting it from this patch silently degrade
> their performance?
> 

Update for monaco-monza-som in v3 series - https://lore.kernel.org/all/20260701081715.2580329-1-monish.chunara@oss.qualcomm.com/

For monaco-evk, it is handled via the overlay monaco-evk-emmc.dtso added as part
of the same patchset.

> [Severity: High]
> The commit message states the goal is to ensure the SoC definition remains
> generic for SD card compatibility by removing supports-cqe. However, does
> leaving the other eMMC-specific properties in the SoC DT defeat this purpose?
> 
> Properties like bus-width = <8>, mmc-hs400-1_8v, mmc-hs400-enhanced-strobe,
> qcom,dll-config, and qcom,ddr-config remain in the base monaco.dtsi. Since
> device tree overlays do not automatically delete base boolean properties or
> override tuning parameters unless explicitly redefined, won't these
> incorrectly inherit to SD card configurations and cause the MMC core to
> misconfigure the interface?
> 
> Should these properties be moved to the eMMC-specific board files as well?
> 

These speed modes are handled in the eMMC driver and won't affect the SD card
functionality. bus-width is overridden in the platform file to match the SD card
requirements.

> >  			dma-coherent;
> >  
> >  			mmc-ddr-1_8v;
> > diff --git a/arch/arm64/boot/dts/qcom/qcs8300-ride.dts b/arch/arm64/boot/dts/qcom/qcs8300-ride.dts
> > index e9a8553a8d821..3090eba0317ac 100644
> > --- a/arch/arm64/boot/dts/qcom/qcs8300-ride.dts
> > +++ b/arch/arm64/boot/dts/qcom/qcs8300-ride.dts
> > @@ -719,6 +719,7 @@ &sdhc_1 {
> >  	vmmc-supply = <&vreg_l8a>;
> >  	vqmmc-supply = <&vreg_s4a>;
> >  
> > +	supports-cqe;
> >  	non-removable;
> >  	no-sd;
> >  	no-sdio;
> 
> -- 
> Sashiko AI review · https://sashiko.dev/#/patchset/20260616130347.3096034-1-monish.chunara@oss.qualcomm.com?part=1