By Tyler Healy
Every cloud customer gets the same promise: a bright-line boundary between your workload and your neighbors. Your virtual machine is yours, your neighbors’ are theirs, and the hypervisor keeps it that way.
A hypervisor is large, complex software written by people, and flaws come with that. These boundaries get tested to fix those flaws. What matters is who finds them first and what they do next.
A researcher who finds a flaw in shared infrastructure can publish it, sell it, or tell us. Responsible disclosure partners tell us – that gives open source maintainers who own the code time to build and test a fix before an attacker can use it. It is the difference between a quiet patch cycle and an all-hands emergency.
The gap between a flaw going public and attackers using it is now near-zero, or ahead of patch in some cases. Early notice is how we stay ahead of the issues that could break the day-to-day businesses running on the cloud. We would rather hear about a problem from a partner than from a headline.
Terra Security researcher Oren Yomtov found a flaw in KVM, the hypervisor built into the Linux kernel. On AMD processors with nested virtualization enabled, a customer with root inside their own Droplet could reach and modify the memory of the physical host underneath it. On a shared host, that memory includes other tenants’ VMs.
Terra showed the read on one of our Droplets. The first thing that came back was the host’s kernel version banner. Then came about 127 GB of host RAM, read-only. They showed the write, 32 chosen bytes into a live host kernel page, on a bare-metal AMD server with no other customers attached.
The cause is a stale pointer. When a reset signal pulls a virtual CPU out of a nested VM, KVM skipped one step: refreshing the pointer to the guest’s page table. It kept naming a page KVM had already freed. The processor followed it anyway, so whatever the kernel reused that page for became the guest’s map of memory.
The bug sat in the kernel for just under four years. The 2022 fix for CVE-2022-3344 repaired one problem and missed this pointer. Terra’s PatchSlop research found that 3.2% of kernel CVEs were introduced by the fix for another CVE. Fixes deserve the same scrutiny as the code they patch.
Terra brought this to us through our bug bounty program. We validated it, treated it as critical, and kept it tightly held. We encouraged Terra to report the issue to the KVM maintainers. The community moved fast. The fix was first tagged in v7.4-rc2 on November 9. On November 17 it shipped in stable 7.3.5 and in the 6.18.64 and 6.12.118 long-term kernels.
We validated the exploit on our own hosts and prepared a live patch, which lets us fix a running hypervisor kernel without waiting on a full reboot cycle. When the upstream fix landed, we rebuilt our patch from it and started rolling it across the fleet. Coordination moves the industry from discovery to patch as quickly as possible, minimizing the window of insecure time.
The fix lives in the hypervisor kernel, so for Droplets we have addressed this across our fleet. No action needed on your part… If you run your own KVM hosts on AMD, you have two options:
Read Terra’s full technical write-up for the details.
I’m grateful to Oren and the team at Terra for doing this the right way, and to the KVM maintainers for the speed of the fix. We will never ship perfect software. We can make sure the people who find our problems tell us first, and that the fix is moving before the headline is. If you find something in our platform, tell us through our bug bounty program.

