A flaw in the Linux kernel scheduler causes a task that is enqueued with SCX_ENQ_LAST to stall when it returns to its CPU’s local dispatch queue. This results in the CPU idling without rescheduling, which keeps the task queued until an unrelated wake‑up or the watchdog fires. In mass cgroup teardown scenarios, this can freeze exiting tasks within the exit_mmap() phase for over 40 seconds, effectively degrading system responsiveness and preventing normal CPU utilization. The weakness enables an attacker with sufficient privileges to cause substantial delays or unavailability for the affected tasks, presenting a denial‑of‑service scenario at the kernel level.

The vulnerability affects the Linux kernel on all platforms and architectures. Vendor and product information lists the operating system as Linux:Linux, and no specific kernel version is excluded; the fix applies to any release that incorporates the reported changes.

Official CVSS and EPSS scores are not provided, and the vulnerability is not listed in the CISA KEV catalog, which indicates that there is no public evidence of exploitation yet. Nonetheless, the bug can be triggered by a privileged process that enqueues tasks with SCX_OPS_ENQ_LAST and can lead to significant performance degradation. Because the attack requires kernel‑level access, the potential impact remains high for affected systems until the kernel is updated.