Re: [PATCH v2 14/24] arm64: dts: qcom: sm4450: Reduce OS PDC DRV span to 0x10000

[Rob Herring <robh@kernel.org>]

On Wed, May 27, 2026 at 09:53:42AM +0000, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [Medium] Reducing the `reg` span to 0x10000 in the device tree is futile without an accompanying driver change and will intentionally trigger a boot warning.
> --
> 
> commit befb5e3c3aa5e0f54a7fff3d0f4a5e96013f4a7e
> Author: Mukesh Ojha <mukesh.ojha@oss.qualcomm.com>
> 
> arm64: dts: qcom: sm4450: Reduce OS PDC DRV span to 0x10000
> 
> The OS PDC DRV register window on sm4450 spans 0x10000 bytes, so this
> commit reduces the size of the first reg entry from 0x30000 to 0x10000.
> 
> > diff --git a/arch/arm64/boot/dts/qcom/sm4450.dtsi b/arch/arm64/boot/dts/qcom/sm4450.dtsi
> > --- a/arch/arm64/boot/dts/qcom/sm4450.dtsi
> > +++ b/arch/arm64/boot/dts/qcom/sm4450.dtsi
> > @@ -464,7 +464,7 @@ dispcc: clock-controller@af00000 {
> >  
> >  		pdc: interrupt-controller@b220000 {
> >  			compatible = "qcom,sm4450-pdc", "qcom,pdc";
> > -			reg = <0 0x0b220000 0 0x30000>, <0 0x174000f0 0 0x64>;
> > +			reg = <0 0x0b220000 0 0x10000>, <0 0x174000f0 0 0x64>;
> 
> [Severity: Medium]
> Will this change cause a boot warning and fail to actually reduce the mapping
> size?
> 
> Looking at the qcom-pdc driver in drivers/irqchip/qcom-pdc.c, it enforces a
> minimum mapping size of QCOM_PDC_SIZE (0x30000) for all devices to maintain
> backward compatibility. In qcom_pdc_probe(), it calculates the size as:
> 
> 	res_size = max_t(resource_size_t, resource_size(&res), QCOM_PDC_SIZE);
> 
> If the device tree provides a size smaller than 0x30000 (like 0x10000), the
> driver automatically expands the mapping back to 0x30000 and logs a warning:
> "%pOF: invalid reg size, please fix DT".
> 
> Since there are no accompanying driver changes to gracefully accept the smaller
> size for newer compatibles like "qcom,sm4450-pdc", does this patch introduce a
> spurious warning on every boot without achieving the intended mapping
> reduction?

Seems valid to me. Even if you did change the driver, it would be an ABI 
break (though if only introducing a spurious warning for existing OS).

Please reply to sashiko reports because the maintainers are not going to 
analyze them for you.

Rob