The Linux kernel bug arises when CONFIG_PAGE_OWNER is enabled and KASAN shadow pages are freed during vmalloc cleanup. The cleanup path performs stack unwinding while holding an RCU read lock for every page freed, and it does not voluntarily reschedule until the entire purge list is processed. If the list contains hundreds or thousands of vmap_area entries, the task can hold the CPU for many seconds, preventing other CPUs from completing RCU grace periods and potentially causing out‑of‑memory conditions. The effect is a prolonged RCU stall that manifests as a kernel message indicating blocking on level‑0 RCU nodes and can lead to system instability or a service denial.

All Linux kernel builds that enable CONFIG_PAGE_OWNER in combination with KASAN shadow page support are affected. The vulnerability is present regardless of the specific kernel version, but it was patched in a recent commit to the mainline kernel; any user running a kernel prior to that commit faces the risk. The patch is included in downstream distributions that ship a kernel with the kasan_release_vmalloc_node fix applied.

The CVSS score is 5.5 and the EPSS score is not available, indicating a medium severity system impact from this bug. The exploit requires a context that can free vmalloc pages in a large purge list, typically available only to code running with kernel privileges (for example, kernel modules, compromised processes, or development builds with KASAN enabled). No public exploits have been documented, and the acceptance of local kernel code is needed; however, the denial‑of‑service nature of the flaw makes it a severe risk that should be addressed as soon as a patched kernel is available. The lack of a KEV listing indicates no known widespread exploitation, but the bug remains a high‑internal threat to kernel stability.