CVE-2026-121876: systemd-gpt-auto-generator credential disclosure in OpenShift Sandboxed Containers

CVE-2026-121876 · Severity: high · CVSS 7.9 · Published 2026-11-12

OpenShift Sandboxed Containers uses lightweight virtual machines to isolate containerized workloads with hardware-enforced confidentiality guarantees. A flaw in systemd-gpt-auto-generator automatically mounts disk partitions without verification, allowing a malicious host to inject a swap partition and force the guest kernel to write encrypted memory to the host, completely bypassing the confidentiality protection that the hardware TEE (AMD SEV-SNP, Intel TDX) is supposed to provide.

Vendors: Red Hat.

Quick answers

Is CVE-2026-121876 exploited in the wild?
No. Junglewise Threat Intelligence has no report of CVE-2026-121876 being exploited in the wild.
How severe is CVE-2026-121876?
High severity, with a CVSS score of 7.9, published 12 November 2026.
Which products and versions are affected?
Red Hat OpenShift Sandboxed Containers confidential peer pods; Red Hat systemd-udev (RHEL) all versions shipped with RHEL.
Is there a fix for CVE-2026-121876?
No fixed version is recorded yet. Add systemd.gpt_auto=0 to the measured kernel command line in the UKI addon to completely disable systemd-gpt-auto-generator. Alternatively, disable swap specifically with systemd.swap=0, mask the generator via ln -sf /dev/null /etc/systemd/system-generators/systemd-gpt-auto-generator, include PCR 5 in PodVM measurement, or for systemd 254+, set restrictive image policy via systemd.image_policy=root=verity+signed+encrypted+unprotected+absent:swap=absent:home=absent:srv=absent:var=absent:tmp=absent. If you run it, Junglewise drafts a mitigation that fits your setup and applies it once you approve, with senior security engineers behind it.
How is CVE-2026-121876 attacked?
It is attacked with local access, in a network appliance.
Am I affected by CVE-2026-121876?
Check whether you run Red Hat OpenShift Sandboxed Containers or Red Hat systemd-udev (RHEL) at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

How to fix

Add systemd.gpt_auto=0 to the measured kernel command line in the UKI addon to completely disable systemd-gpt-auto-generator. Alternatively, disable swap specifically with systemd.swap=0, mask the generator via ln -sf /dev/null /etc/systemd/system-generators/systemd-gpt-auto-generator, include PCR 5 in PodVM measurement, or for systemd 254+, set restrictive image policy via systemd.image_policy=root=verity+signed+encrypted+unprotected+absent:swap=absent:home=absent:srv=absent:var=absent:tmp=absent.

Technical details

systemd-gpt-auto-generator automatically discovers and mounts GPT partitions by type GUID at boot with no user confirmation. In confidential VMs, the host/hypervisor is the adversary; it can inject a swap partition into the guest's virtual disk, causing the kernel to page out TEE-encrypted memory to a host-controlled block device. The dm-verity root partition protection and PCR measurement (3, 9, 11, 12) do not cover non-root partitions or the GPT partition table (PCR 5), leaving the guest vulnerable to partition table modification attacks.

Affected products

  • Red Hat OpenShift Sandboxed Containers confidential peer pods
  • Red Hat systemd-udev (RHEL) all versions shipped with RHEL

Timeline

  • 2026-11-12: disclosed

References

This page as Markdown. Free to reuse under CC BY 4.0 with a link back.