A race condition in the ARM64 KVM implementation allows the nested_mmus array to be reallocated while a reader holds kvm->mmu_lock, causing the array to be freed before the lock is released. The reader can then access freed memory, leading to kernel memory corruption or a kernel panic. The CVE description explicitly notes that the walker can reference the freed array, implying a use‑after‑free that could crash the host or corrupt data. No privilege escalation capability is stated, only the potential for kernel instability.

All ARM64 Linux kernels 7.1rc1 through 7.1rc6 that include KVM with nested paging enabled, as indicated by the contained CPE strings. Any distribution shipping these kernels on ARM64 hardware and running virtual machines with nested paging is affected.

The CVSS score of 8.8 denotes a high severity flaw. The EPSS score of < 1% indicates that exploitation is presently unlikely. The vulnerability is not listed in the CISA KEV catalog. Based on the description, the likely attack vector involves a malicious guest triggering a race condition during nested paging operations; this could lead to a host crash or memory corruption. No publicly documented exploit exists yet, but the conditions of the race suggest that an attacker controlling a guest could attempt to exercise the flaw.