CVE-2026-99713: KVM: x86/mmu: Conditionally resched when retrying gPTE walk on A/D update failure
In the Linux kernel, the following vulnerability has been resolved:
KVM: x86/mmu: Conditionally resched when retrying gPTE walk on A/D update failure
Conditionally reschedule when retrying a guest PTE walk because the CMPXCHG to update A/D bits failed, e.g. because a different task/vCPU modified the relevant gPTEs. On kernels without involuntary preemption of some kind, a misbehaving guest can "capture" a pCPU effectively indefinitely and trigger soft lockups by continuously modifying its gPTEs such that the result of the walk is remains valid.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.12.118 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.18.64 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 7.3.5
Frequently Asked Questions
Which systems are exposed to this issue?
Systems running KVM on x86 are exposed when they use kernels without some form of involuntary preemption. The issue requires a guest capable of repeatedly modifying its guest page tables while KVM is walking them.
What can a malicious or misbehaving guest do?
A guest can repeatedly cause KVM's compare-and-exchange update of accessed/dirty bits to fail while keeping the page-table walk valid. This can effectively monopolize a physical CPU indefinitely and trigger soft lockups.
Is this a host-local issue or can it be triggered remotely?
The described trigger is a guest workload modifying its own guest PTEs, so exploitation requires the ability to run or control code in a KVM guest. The provided information does not describe a host-local unprivileged trigger or a network-only attack path.
What is the relevant mitigation if patching cannot happen immediately?
Use a kernel configuration that provides involuntary preemption, where applicable. This reduces the condition described as allowing a guest to retain a physical CPU effectively indefinitely.