The Linux kernel KVM hypervisor fails to synchronize the interrupt shadow state from the guest CPU’s vmcb02 to the cached vmcb12 when executing a VMRUN on a level‑2 (nested) virtual machine. Because the bit controlling the interrupt shadow mask in int_state is not copied, an L2 guest may receive a wake‑up interrupt that should have been masked, causing the guest to hang or experience a timed‑out HLT. This flaw does not allow direct code execution or privilege escalation, but it permits a malicious or mis‑configured nested guest to induce a denial‑of‑service condition on the host or other guests. The primary weakness is a race‑condition in state synchronization.

The vulnerability applies to Linux kernel builds that implement KVM with nested SVM support. No specific kernel version range is listed in the data, so any kernel that includes the affected nested‑VM code paths is potentially affected. The issue is tied to the Linux vendor and the generic Linux kernel product; affected systems include any host running these components with nested virtualization enabled.

The CVSS score is not provided, and the EPSS score is unavailable. The CISA KEV catalog does not list this vulnerability, indicating no known active exploitation. However, the flaw can be triggered by an attacker that controls the ordering of the KVM_SET_VCPU_EVENTS and KVM_SET_NESTED_STATE ioctls on a nested guest; a benign mis‑ordering can also cause a system administrator to suffer a guest hang. Given the lack of publicly documented exploits, the immediate risk is primarily a service disruption scenario, but the potential for significant downtime makes it a high‑priority issue for environments relying on nested virtualization.