The vulnerability arises from a race condition in the Linux kernel’s SCSI core where the error handler thread may never be woken when all commands time out or become inactive. This occurs because there is no enforced memory ordering between placing the host into an error recovery state and decrementing the busy command count. If a CPU observes that no commands are active while the host is in error state, the wakeup signal can be missed, leaving the error handler thread stalled. The result is a denial‑of‑service condition that could prevent the kernel from effectively handling further SCSI operations. The weakness is a classic race condition (CWE‑665).

All Linux kernel versions where the scsi_schedule_eh function is used in SCSI drivers are affected. No specific version range is supplied in the advisory, so any kernel that relies on scsi_schedule_eh and does not include the patch will be vulnerable.

The CVSS score is not listed, and EPSS is not available, but the issue can be exploited locally with kernel privileges or through a vulnerability that grants such access. Because it involves a kernel race condition that can cause a persistence hang, the risk is high for affected systems. The vulnerability is not listed in CISA’s KEV catalog, suggesting no known active exploitation at the time of this analysis. Nonetheless, administrators should treat the risk as significant. The likely attack vector would involve an attacker initiating a flood of SCSI commands that time out or otherwise trigger the error handler path in a controlled way, leading the kernel to hang until a reboot or manual intervention.