The vulnerability is a concurrency flaw in the PowerPC EEH driver. When a device that lacks EEH‑sensitive handlers reports an error, the eeh_rmv_device() function still performs an unlock that is no longer needed. Because the pci_rescan_remove_lock is already held by eeh_handle_normal_event() when eeh_reset_device() is called, the redundant unlock causes the thread to block waiting for a lock it already holds, resulting in a deadlock and a kernel hang. The effect is a loss of kernel responsiveness and service availability, with no evidence of code execution or data disclosure.

All Linux kernel builds for PowerPC platforms that include the EEH subsystem. The vulnerability applies to any kernel where the EEH driver is loaded, regardless of whether the attached device drivers are EEH‑sensitive. No specific kernel version is listed; the issue persists until the commit 1010b4c012b0 is applied.

The bug is triggered by hardware or driver error conditions on the PHB or by drivers that return PCI_ERS_RESULT_NEED_RESET without providing EEH‑sensitive handlers. The likely attack vector is local or system‑level, inferred from the requirement of an error state that activates the EEH error path. Exploitation therefore requires an error state that activates the EEH error path, making the attack vector local or system‑level. The EPSS score is not available and the vulnerability is not included in the CISA KEV catalog, indicating limited public exploit activity. Without a CVSS score, the fault’s severity can be inferred as moderate: it causes a denial of service via a kernel deadlock but offers no pathway to privilege escalation or remote code execution.