[PATCH v2] phy: qcom-qusb2: Fix NULL pointer dereference on early suspend

[Dmitry Baryshkov]

On Thu, Dec 11, 2025 at 02:48:36PM +0100, Loic Poulain wrote:
> Enabling runtime PM before attaching the QPHY instance as driver data
> can lead to a NULL pointer dereference in runtime PM callbacks that
> expect valid driver data. There is a small window where the suspend
> callback may run after PM runtime enabling and before runtime forbid.
> This causes a sporadic crash during boot:
> 
> ```
> Unable to handle kernel NULL pointer dereference at virtual address 00000000000000a1
> [...]
> CPU: 0 UID: 0 PID: 11 Comm: kworker/0:1 Not tainted 6.16.7+ #116 PREEMPT
> Workqueue: pm pm_runtime_work
> pstate: 20000005 (nzCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
> pc : qusb2_phy_runtime_suspend+0x14/0x1e0 [phy_qcom_qusb2]
> lr : pm_generic_runtime_suspend+0x2c/0x44
> [...]
> ```
> 
> Enable PM runtime only after attaching the QPHY instance as driver data to
> avoid a NULL pointer dereference in PM runtime callbacks.
> 
> Reorder pm_runtime_enable() and pm_runtime_forbid() to prevent a
> short window where an unnecessary runtime suspend can occur.
> 
> Use the devres-managed version to ensure PM runtime is symmetrically
> disabled during driver removal for proper cleanup.
> 
> Fixes: 891a96f65ac3 ("phy: qcom-qusb2: Add support for runtime PM")
> Signed-off-by: Loic Poulain <loic.poulain at oss.qualcomm.com>
> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov at oss.qualcomm.com>
> ---
>  v2: Move runtime-pm enabling after dev_set_drvdata
>      use devm_pm_runtime_enable() to fix unbalanced enabling on removal
>      reword commit message
> 
>  drivers/phy/qualcomm/phy-qcom-qusb2.c | 19 ++++++++-----------
>  1 file changed, 8 insertions(+), 11 deletions(-)
> 
> diff --git a/drivers/phy/qualcomm/phy-qcom-qusb2.c b/drivers/phy/qualcomm/phy-qcom-qusb2.c
> index b5514a32ff8f..c496acbad0bb 100644
> --- a/drivers/phy/qualcomm/phy-qcom-qusb2.c
> +++ b/drivers/phy/qualcomm/phy-qcom-qusb2.c
> @@ -1093,19 +1093,10 @@ static int qusb2_phy_probe(struct platform_device *pdev)
>  		or->hsdisc_trim.override = true;
>  	}
>  
> -	pm_runtime_set_active(dev);
> -	pm_runtime_enable(dev);
> -	/*
> -	 * Prevent runtime pm from being ON by default. Users can enable
> -	 * it using power/control in sysfs.
> -	 */
> -	pm_runtime_forbid(dev);
> -
>  	generic_phy = devm_phy_create(dev, NULL, &qusb2_phy_gen_ops);

No. phy_create() checks whether the device has runtime PM enabled. By
moving these calls later you've disabled runtime PM for the PHY.

>  	if (IS_ERR(generic_phy)) {
>  		ret = PTR_ERR(generic_phy);
>  		dev_err(dev, "failed to create phy, %d\n", ret);
> -		pm_runtime_disable(dev);
>  		return ret;
>  	}
>  	qphy->phy = generic_phy;
> @@ -1113,9 +1104,15 @@ static int qusb2_phy_probe(struct platform_device *pdev)
>  	dev_set_drvdata(dev, qphy);
>  	phy_set_drvdata(generic_phy, qphy);
>  
> +	/*
> +	 * Enable runtime PM support, but forbid it by default.
> +	 * Users can allow it again via the power/control attribute in sysfs.
> +	 */
> +	pm_runtime_set_active(dev);
> +	pm_runtime_forbid(dev);
> +	devm_pm_runtime_enable(dev);
> +
>  	phy_provider = devm_of_phy_provider_register(dev, of_phy_simple_xlate);
> -	if (IS_ERR(phy_provider))
> -		pm_runtime_disable(dev);
>  
>  	return PTR_ERR_OR_ZERO(phy_provider);
>  }
> -- 
> 2.34.1
> 

-- 
With best wishes
Dmitry