Re: [PATCH 4/4] arm64: dts: qcom: glymur: Add CoreSight nodes for APSS debug block

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] dts: empty port node left after endpoint deletion triggers dtbs_check warning
--

commit b9b7f9e54234f4ce9f0f624d710b7719ad97a035
Author: Jie Gan <jie.gan@oss.qualcomm.com>

arm64: dts: qcom: glymur: Add CoreSight nodes for APSS debug block

This patch adds CoreSight devices such as ETM, replicator, funnel, and
TMC ETF to the APSS debug block on Glymur. It also deletes the nodes
corresponding to cluster 2 on Mahua, since that platform lacks the third
CPU cluster.

> diff --git a/arch/arm64/boot/dts/qcom/glymur.dtsi b/arch/arm64/boot/dts/qcom/glymur.dtsi
> index de7c395d9c353..049bebbb80f92 100644
> --- a/arch/arm64/boot/dts/qcom/glymur.dtsi
> +++ b/arch/arm64/boot/dts/qcom/glymur.dtsi
[ ... ]
> @@ -7625,6 +7939,51 @@ pcie_rscc_tpda_out: endpoint {
>  			};
>  		};
>  
> +		funnel@12080000 {
> +			compatible = "arm,coresight-dynamic-funnel", "arm,primecell";
> +			reg = <0x0 0x12080000 0x0 0x1000>;
> +
> +			clocks = <&aoss_qmp>;
> +			clock-names = "apb_pclk";
> +
> +			in-ports {
> +				#address-cells = <1>;
> +				#size-cells = <0>;
> +
> +				port@0 {
> +					reg = <0>;
> +
> +					apss_funnel_in0: endpoint {
> +						remote-endpoint = <&ncc0_etf_out>;
> +					};
> +				};
> +
> +				port@1 {
> +					reg = <1>;
> +
> +					apss_funnel_in1: endpoint {
> +						remote-endpoint = <&ncc1_etf_out>;
> +					};
> +				};
> +
> +				port@2 {
> +					reg = <2>;
> +
> +					apss_funnel_in2: endpoint {
> +						remote-endpoint = <&ncc2_etf_out>;
> +					};
> +				};
> +			};
[ ... ]
> diff --git a/arch/arm64/boot/dts/qcom/mahua.dtsi b/arch/arm64/boot/dts/qcom/mahua.dtsi
> index a59c5ed7b48e7..77a456fb0772b 100644
> --- a/arch/arm64/boot/dts/qcom/mahua.dtsi
> +++ b/arch/arm64/boot/dts/qcom/mahua.dtsi
> @@ -6,8 +6,18 @@
>  /* Mahua is heavily based on Glymur, with some meaningful differences */
>  #include "glymur.dtsi"
>  
> +/delete-node/ &apss_funnel_in2;

[Severity: Low]
Does this deletion of the endpoint leave the parent port@2 node empty?

The device tree schema for graph ports strictly requires a port node to
contain at least one endpoint child. Leaving the port node empty violates
this schema and causes a dtbs_check validation regression. 

Is there a way to delete the entire parent port node instead of just the
endpoint, to avoid leaving an empty port@2 behind in the device tree?

>  /delete-node/ &bwmon_cluster2;
> +/delete-node/ &cluster2_etf;
> +/delete-node/ &cluster2_funnel_l1;

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260922-add-etm-devices-for-multiple-qcom-platforms-v1-0-c2a8c3225a2e@oss.qualcomm.com?part=4