CVE-2026-98507

A flaw was found in the Linux kernel's memory management subsystem. Due to a missing synchronization barrier when preparing anonymous virtual memory areas, concurrent memory operations can be reordered. A local attacker can exploit this race condition by running concurrent threads on adjacent memory regions, causing lock mismatches or improper memory reuse. This flaw primarily results in a kernel hang or deadlock, causing a Denial of Service (DoS).

A race in __anon_vma_prepare can publish vma anon_vma before initialization of the anon_vma root field becomes visible to another CPU. A concurrent page fault can then obtain a stale root from reused SLAB_TYPESAFE_BY_RCU memory, lock the old root, and unlock a different root after the initialization becomes visible. This creates a use after free candidate and can corrupt rwsem state, permanently lock an anon_vma, and hang tasks. For the CVSS the PR:L is selected because an unprivileged local process can normally create adjacent VMAs, change VMA access flags with mprotect, and trigger concurrent page faults. AC:H is selected because successful triggering depends on a narrow race window and weak memory ordering. The issue is not network reachable.

The CVSS score(s) listed for this vulnerability do not reflect the associated product's status, and are included for informational purposes to better understand the severity of this vulnerability.

Mitigation for this issue is either not available or the currently available options don't meet the Red Hat Product Security criteria comprising ease of use and deployment, applicability to widespread installation base or stability.