CRI-O vulnerable to an arbitrary systemd property injection
### Impact On CRI-O, it looks like an arbitrary systemd property can be injected via a Pod annotation: ``` --- apiVersion: v1 kind: Pod metadata: name: poc-arbitrary-systemd-property-injection annotations: # I believe that ExecStart with an arbitrary command works here too, # but I haven't figured out how to marshalize the ExecStart struct to gvariant string. org.systemd.property.SuccessAction: "'poweroff-force'" spec: containers: - name: hello image: [quay.io/podman/hello](http://quay.io/podman/hello) ``` This means that any user who can create a pod with an arbitrary annotation may perform an arbitrary action on the host system. Tested with CRI-O v1.24 on minikube. I didn't test the latest v1.29 because it is incompatible with minikube: https://github.com/kubernetes/minikube/pull/18367 Thanks to Cédric Clerget (GitHub ID @cclerget) for finding out that CRI-O just passes pod annotations to OCI annotations: https://github.com/opencontainers/runc/pull/3923#discussion_r1532292536 CRI-O has to filter out annotations that have the prefix "org.systemd.property." See also: - https://github.com/opencontainers/runtime-spec/blob/main/features.md#unsafe-annotations-in-configjson - https://github.com/opencontainers/runc/pull/4217 ### Workarounds Unfortunately, the only workarounds would involve an external mutating webhook to disallow these annotations ### References
02 / AFFECTED SOFTWARE
Affected packages
3 explicit affected versions
03 / CONNECTIONS
Connected vulnerabilities
04 / EVIDENCE
Source records
A flaw was found in cri-o, where an arbitrary systemd property can be injected via a Pod annotation. Any user who can create a pod with an arbitrary annotation may perform an arbitrary action on the host system.
### Impact On CRI-O, it looks like an arbitrary systemd property can be injected via a Pod annotation: ``` --- apiVersion: v1 kind: Pod metadata: name: poc-arbitrary-systemd-property-injection annotations: # I believe that ExecStart with an arbitrary command works here too, # but I haven't figured out how to marshalize the ExecStart struct to gvariant string. org.systemd.property.SuccessAction: "'poweroff-force'" spec: containers: - name: hello image: [quay.io/podman/hello](http://quay.io/podman/hello) ``` This means that any user who can create a pod with an arbitrary annotation may perform an arbitrary action on the host system. Tested with CRI-O v1.24 on minikube. I didn't test the latest v1.29 because it is incompatible with minikube: https://github.com/kubernetes/minikube/pull/18367 Thanks to Cédric Clerget (GitHub ID @cclerget) for finding out that CRI-O just passes pod annotations to OCI annotations: https://github.com/opencontainers/runc/pull/3923#discussion_r1532292536 CRI-O has to filter out annotations that have the prefix "org.systemd.property." See also: - https://github.com/opencontainers/runtime-spec/blob/main/features.md#unsafe-annotations-in-configjson - https://github.com/opencontainers/runc/pull/4217 ### Workarounds Unfortunately, the only workarounds would involve an external mutating webhook to disallow these annotations ### References
CRI-O vulnerable to an arbitrary systemd property injection in github.com/cri-o/cri-o
05 / REFERENCES
Further evidence
- https://access.redhat.com/downloads/content/package-browser/
- https://access.redhat.com/errata/RHSA-2024:2669
- https://access.redhat.com/errata/RHSA-2024:2672
- https://access.redhat.com/errata/RHSA-2024:2784
- https://access.redhat.com/errata/RHSA-2024:3496
- https://access.redhat.com/security/cve/CVE-2024-3154
- https://bugzilla.redhat.com/show_bug.cgi?id=2272532
- https://github.com/CVEProject/cvelistV5/tree/main/cves/2024/3xxx/CVE-2024-3154.json
- https://github.com/cri-o/cri-o
- https://github.com/cri-o/cri-o/security/advisories/GHSA-2cgq-h8xw-2v5j
- https://github.com/opencontainers/runc/pull/4217
- https://github.com/opencontainers/runtime-spec/blob/main/features.md#unsafe-annotations-in-configjson
- https://nvd.nist.gov/vuln/detail/CVE-2024-3154