The sdhci_sprd driver was leaving the runtime power‑management (PM) subsystem in an inconsistent state when a device was removed. The probe path disabled runtime PM but the normal remove path did not call a cleanup routine, directly disabling hardware clocks that are normally controlled by runtime PM callbacks. If the device was in a runtime‑suspended state, those clocks may already have been disabled, leading to an invalid clock state during removal and potentially causing a kernel fault or crash.

All Linux kernel releases that include the sdhci_sprd driver before the fix was applied are affected. The issue exists in the kernel's generic MMC subsystem and applies to any configuration that enables runtime PM for Sprd SDHCI devices.

The vulnerability is not listed in CISA’s KEV catalog, the EPSS score is < 1%, and the CVSS score is 2.3, both indicating low likelihood and low severity. Because the issue requires removal of a device that is in a runtime‑suspended state, the attack vector is likely local or requires privileged access to trigger the removal sequence. The risk is moderate: an unpatched system could experience a kernel panic or device malfunction if the device removal occurs while suspended. The absence of a publicly available high‑volume exploit reduces the immediacy of the threat, but the impact of a kernel crash justifies caution.