The Linux kernel KVM module for SVM contains a race condition between the capabilities KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM and KVM_CAP_VM_COPY_ENC_CONTEXT_FROM. When both operations are invoked concurrently, the owner and mirror lists are updated without acquiring the appropriate owner lock, which can corrupt the list and lead to a use‑after‑free when an incorrect kvm_put_kvm() is called. This flaw stems from improper synchronization of these data structures, potentially allowing an attacker to trigger kernel memory corruption and host crash.

All Linux kernels that implement KVM SVM and have not incorporated the commits referenced in the advisory are affected. No explicit version range is supplied, so any kernel lacking the patch is vulnerable.

The EPSS score, which is less than 1%, indicates a very low probability of exploitation, and the vulnerability is not listed in CISA KEV, indicating no public exploitation has been observed. The CVSS score of 8.8 classifies this vulnerability as high severity. Based on the description, it is inferred that exploitation would require a local attacker to control or influence VM migration or context copy operations on a host running KVM. Such an attack could result in a kernel panic and denial of service. The opportunity for privilege escalation is not confirmed by the description, so an attack is limited to loss of availability and potential memory corruption, though further exploitation might be possible if the bug allows arbitrary writes.