New Zapscape KVM Flaw Could Let Privileged L1 Guest Code Escape to Linux Hosts
A new Linux kernel vulnerability, tracked as CVE-2026-64561, could allow an attacker with kernel privileges inside an L1 guest virtual machine (VM) to escape KVM isolation and execute code on the host. The flaw affects KVM/x86's shadow memory management unit (MMU) and req…
Intelligence analysis by Llama
A new Linux kernel vulnerability, CVE-2026-64561, allows an attacker with kernel privileges inside an L1 guest VM to escape KVM isolation and execute code on the host. The flaw affects KVM/x86's shadow-MMU bookkeeping and requires L1 kernel privilege, usually guest root, and Intel systems to expose both EPT page-walk length 4 and 5 to the L1 guest.
Imagine you're a guest in a virtual hotel. The hotel has a special system that helps you navigate the rooms. But if someone hacks into the system, they could escape the hotel and run around the entire building. That's what's happening with the Zapscape flaw - it's a way for an attacker to escape the virtual hotel and run code on the host.
Analysis
Zapscape Flaw Overview
The Zapscape flaw is a new Linux kernel vulnerability that could allow an attacker with kernel privileges inside an L1 guest virtual machine (VM) to escape KVM isolation and execute code on the host. The flaw affects KVM/x86's shadow memory management unit (MMU) and requires L1 kernel privilege, usually guest root, and Intel systems to expose both EPT page-walk length 4 and 5 to the L1 guest.
Exploit Path
The demonstrated exploit path can run commands on the host with kernel, or root, privileges. The upstream fix has been merged, and administrators running KVM hosts that expose nested virtualization to untrusted guests should update to a fixed stable kernel or a vendor package that backports the patch.
Technical Details
The issue is a use-after-free in the recursive zap path used when KVM reclaims shadow pages. KVM checked whether the current root was stale before making more MMU pages available. Reclaim could then invalidate that same root, but KVM continued the fault path and created child shadow pages under it. Those child pages inherited the invalid state from the parent and were still placed on KVM's active MMU page list. Later cleanup could attach the same list link to two lists at once, then free the page while stale list references remain, creating a dangling link and post-free write.
Disclosure Timeline
Kim reported the issue to security@kernel.org on July 11, 2026. A patch was posted and merged on July 21, the issue was submitted to the linux-distros list on August 1 under a five-day embargo, and CVE-2026-64561 was assigned on August 4. Public disclosure followed on August 6.
Key points
- A new Linux kernel vulnerability, tracked as CVE-2026-64561, could allow an attacker with kernel privileges inside an L1 guest virtual machine (VM) to escape KVM isolation and execute code on the host.
- The flaw affects KVM/x86's shadow memory management unit (MMU) and requires L1 kernel privilege, usually guest root, and Intel systems to expose both EPT page-walk length 4 and 5 to the L1 guest.
- The upstream fix has been merged, and administrators running KVM hosts that expose nested virtualization to untrusted guests should update to a fixed stable kernel or a vendor package that backports the patch.
If this development plays out positively, it could lead to improved security measures for cloud security and virtualization. This could include better monitoring and detection of potential threats, as well as more robust security protocols to prevent attacks like the Zapscape flaw.
The realistic downside risks or failure modes of the Zapscape flaw include the potential for widespread exploitation, particularly in cloud environments. This could lead to significant security breaches and data losses. Additionally, the flaw could be used to launch targeted attacks on specific systems or organizations.
