Re: [PATCH v5 4/5] arm64: dts: qcom: sm8750: fix traceNoC probe issue

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] Switching the traceNoC device from the AMBA bus to the platform bus actively exposes a PM runtime / devres clock unbalance bug that triggers kernel WARNINGs on device unbind.
--

--- Patch [4]: [PATCH v5 4/5] arm64: dts: qcom: sm8750: fix traceNoC probe issue ---
commit 614093568676102a83b2efb2ba08c1718071ad73
Author: Jie Gan <jie.gan@oss.qualcomm.com>

arm64: dts: qcom: sm8750: fix traceNoC probe issue

This patch drops the "arm,primecell" compatible string from the traceNoC
node to prevent indefinite probe deferral. By using only the standalone
"qcom,coresight-tnoc" compatible, the device binds via the platform
bus instead of the AMBA bus, bypassing the primecell ID check.

> diff --git a/arch/arm64/boot/dts/qcom/sm8750.dtsi b/arch/arm64/boot/dts/qcom/sm8750.dtsi
> index fafed417c66fc..1e25be41aef5c 100644
> --- a/arch/arm64/boot/dts/qcom/sm8750.dtsi
> +++ b/arch/arm64/boot/dts/qcom/sm8750.dtsi
> @@ -4687,7 +4687,7 @@ tpdm_rdpm_cmb2_out: endpoint {
>  		};
>  
>  		tn@109ab000 {
> -			compatible = "qcom,coresight-tnoc", "arm,primecell";
> +			compatible = "qcom,coresight-tnoc";

[Severity: High]
This isn't a bug introduced by this patch, but does changing this to a
platform device expose a clock management unbalance during driver unbind?

When bound as a platform device, tnoc_platform_probe() acquires the clock
via devm_clk_get_optional_enabled() in coresight_get_enable_clocks(). This
queues a devres action to automatically disable the clock upon unbind.

If the device enters runtime suspend, tnoc_runtime_suspend() manually
disables this same clock:

drivers/hwtracing/coresight/coresight-tnoc.c:tnoc_runtime_suspend() {
    ...
    clk_disable_unprepare(drvdata->pclk);
    ...
}

Unlike the AMBA bus framework which automatically wakes devices during
removal, tnoc_platform_remove() does not call pm_runtime_get_sync() before
disabling runtime PM.

If the driver is unbound while suspended, the device stays suspended and the
clock remains disabled. Once tnoc_platform_remove() returns, the devres
unwind process executes the cleanup action, calling clk_disable_unprepare()
a second time.

Will this sequence trigger kernel warnings in clk_core_disable() and
clk_core_unprepare() due to the counts already being zero?

>  			reg = <0x0 0x109ab000 0x0 0x4200>;
>  
>  			clocks = <&aoss_qmp>;

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260707-fix-tracenoc-probe-issue-v5-0-bf733ed9ebd0@oss.qualcomm.com?part=4