Re: [PATCH v9 4/4] arm64: dts: cix: sky1: add audss cru

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Double decrement of the PM runtime usage counter in the probe error path.
--

--- Patch [4]: [PATCH v9 4/4] arm64: dts: cix: sky1: add audss cru ---
commit 37987ea88a502f3fe87d70d7d428d9f00deb012d
Author: Joakim Zhang <joakim.zhang@cixtech.com>

arm64: dts: cix: sky1: add audss cru
    
Add the AUDSS CRU device node providing clocks and software resets
for audio subsystem peripherals.

> diff --git a/arch/arm64/boot/dts/cix/sky1.dtsi b/arch/arm64/boot/dts/cix/sky1.dtsi
> index a77c52296ebd3..9b2d969f6bfe7 100644
> --- a/arch/arm64/boot/dts/cix/sky1.dtsi
> +++ b/arch/arm64/boot/dts/cix/sky1.dtsi

[ ... ]

> @@ -558,6 +562,20 @@ mbox_pm2ap: mailbox@65a0080 {
>  			cix,mbox-dir = "rx";
>  		};
>  
> +		audss_cru: clock-controller@7110000 {
> +			compatible = "cix,sky1-audss-cru";
> +			reg = <0x0 0x07110000 0x0 0x10000>;
> +			#clock-cells = <1>;
> +			#reset-cells = <1>;
> +			clocks = <&scmi_clk CLK_TREE_AUDIO_CLK0>,
> +				 <&scmi_clk CLK_TREE_AUDIO_CLK2>,
> +				 <&scmi_clk CLK_TREE_AUDIO_CLK4>,
> +				 <&scmi_clk CLK_TREE_AUDIO_CLK5>;
> +			clock-names = "x8k", "x11k", "sys", "48m";
> +			power-domains = <&smc_devpd SKY1_PD_AUDIO>;
> +			resets = <&s5_syscon SKY1_AUDIO_HIFI5_NOC_RESET_N>;
> +		};
> +

[Severity: High]
While reviewing this device node, I noticed a regression in the corresponding
driver's probe error path in drivers/clk/cix/clk-sky1-audss.c:sky1_audss_clk_probe():

fail_clks_set:
	pm_runtime_put_sync(dev);
err_pm:
	pm_runtime_put_noidle(dev);
	pm_runtime_disable(dev);

Does this sequence result in a double decrement of the PM runtime usage
counter? 

The error path correctly calls pm_runtime_put_sync(dev) but then falls through
to err_pm which calls pm_runtime_put_noidle(dev). 

Decrementing the usage counter twice when it was only incremented once causes
it to underflow to -1, which corrupts the device's PM runtime state machine.

If deferred probing occurs or the driver is re-bound, will the next probe
attempt fail to properly resume the device because a single increment will only
bring the counter back up to 0?

>  		mbox_sfh2ap: mailbox@8090000 {
>  			compatible = "cix,sky1-mbox";
>  			reg = <0x0 0x08090000 0x0 0x10000>;

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260709120125.3997078-1-joakim.zhang@cixtech.com?part=4