A flaw in KVM, the hypervisor built into the Linux kernel, let researchers read about 127 GB of a cloud server’s memory from inside a virtual machine of their own on DigitalOcean.
On a shared server, that memory can include other customers’ virtual machines. The flaw, tracked as CVE-2026-99312, affects only hosts with AMD processors and nested virtualization enabled.
Fixed kernels shipped upstream on Tuesday, the day before the research went public. The provider patched its own fleet with a live patch, telling Droplet customers: “No action needed on your part.” Operators who run their own KVM hosts on AMD hardware need to update the host kernel or turn nested virtualization off.
Oren Yomtov, a researcher at the security firm Terra Security, found the flaw and reported it through the provider’s bug bounty program. The provider confirmed the exploit worked on its own hosts and treated it as critical, and the Linux KVM maintainers, told on October 21, merged a fix on November 2.
In the test on one of its Droplets, the first thing to come back was the host’s kernel version banner, followed by about 127 GB of the host’s RAM, mapped read-only, the provider said in a disclosure post yesterday.
The write test ran on the researchers’ own bare-metal AMD server, with no other customers attached. There, the guest wrote 32 chosen bytes into a live page of the host kernel’s memory, and a witness module loaded into that kernel confirmed the change.
Both tests started with root access inside a VM controlled by the researchers. An attacker would need the same position: control of a guest on a vulnerable host, not just network access to it.
Intel-based KVM hosts are not affected, because the vulnerable code exists only in KVM’s support for nested virtualization on AMD. Many cloud providers the researchers surveyed were not vulnerable because they had not enabled nested virtualization on AMD.
There are no reports of the flaw being used outside research, according to VirtFusion, a virtualization platform vendor that is tracking the issue for its operators. No exploit code has been published either, and no guidance exists for detecting a past attempt.
Severity ratings disagree. The Linux kernel’s CVE team scores the flaw 8.8 out of 10, with high impact on confidentiality, integrity, and availability, while Red Hat’s preliminary security assessment rates it as Moderate at 5.6 and treats the impact as denial-of-service only. The provider treated it as critical for its own shared cloud.
Terra’s technical analysis, published on Wednesday, describes a pointer left pointing at freed memory. KVM confines each VM with a table, the nested page table, that translates the VM’s memory addresses into the host’s real memory, and the processor finds that table through a pointer kept for each virtual CPU.
Some hosts let a VM run a VM of its own, a nested VM. When a reset signal called INIT stops the nested VM, KVM skips restoring the outer VM’s pointer, leaving it pointing to a page of memory the kernel has already freed.
Once the kernel reused that freed page, the guest could fill it with entries of its own choosing, and the processor followed them as if they were the real table. The forged entries could name any physical page on the host, which is how about 127 GB of the server’s RAM came into reach of a guest that was never given it.
The fix makes KVM reset that memory-management state on every INIT.
The flaw was created by a security fix. A 2022 patch for a different nested-virtualization bug, CVE-2022-3344, switched a reset path back to the outer VM’s control block but missed the page-table pointer, the researchers say, and the error then sat in the kernel for just under four years.
In earlier work called PatchSlop, the team found that the fix for another one introduced 3.2% of kernel vulnerabilities.
What to Do
The fix lives in the host kernel. Updating the operating systems inside the VMs changes nothing, and a host keeps running its old kernel until it is rebooted.
Upstream, the fix shipped on Tuesday in stable kernel 7.3.5 and in the long-term kernels 6.18.64 and 6.12.118, the kernel team’s CVE announcement shows. The affected code goes back years: it entered the 6.1 kernel in 2022 and was carried into the 5.15.81 and 6.0.11 releases, so older kernel lines and the distribution kernels built on them are in range.
Distribution kernels mostly do not have the fix yet. As of publication, Debian’s security tracker lists the kernels in all of its releases as vulnerable, with a fix only in unstable, and Ubuntu’s CVE tracker rates the flaw High and lists its kernels as vulnerable with no fixed package released. Red Hat lists the Enterprise Linux 9 and 10 kernels as affected with the fix deferred, and the Enterprise Linux 8 kernel as not affected.
Until a fixed host kernel is installed and the host rebooted, the only available protection is to turn nested virtualization off. That closes this path completely, but VMs can then no longer run VMs of their own, so warn tenants who depend on that before making the change.
Nested virtualization is on by default in the kernel’s kvm_amd module, according to VirtFusion’s operator guide, so an unpatched host is exposed unless the setting was deliberately changed. To check and change it:
- Check the setting with
cat /sys/module/kvm_amd/parameters/nested. A value of 1 or Y means nested virtualization is on. - If the host cannot be updated yet, add
options kvm_amd nested=0to the host’s module configuration. - Reboot the host. The setting cannot change while VMs are running, and a newly installed kernel does not protect the host until the reboot either. Confirm with
uname -rand by reading the nested setting again.
The flaw is the latest in a series of KVM isolation failures disclosed this year, after a 16-year-old flaw in KVM’s shadow memory management that affected both Intel and AMD systems, and the PwnEscape flaw in Linux’s networking code, which let a guest read host memory in Vercel’s live sandboxes in October.






