GHSA-jpmc-7p9c-4rxf
LOWlxd has a restricted TLS certificate privilege escalation when in PKI mode
EPSS Exploitation Probability
EPSS (Exploit Prediction Scoring System) is a daily probability model maintained by FIRST.org. It estimates the likelihood a CVE will be exploited in production environments within the next 30 days, derived from real-world threat intelligence signals.
Blast Radius
github.com/canonical/lxdReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects Go packages — download data is not available via public APIs for these ecosystems.
Description
Summary
If a server.ca file is present in LXD_DIR at LXD start up, LXD is in "PKI mode". In this mode, all clients must have certificates that have been signed by the CA.
The LXD configuration option core.trust_ca_certificates defaults to false. This means that although the client certificate has been signed by the CA, LXD will additionally add the certificate to the trust store and verify it via mTLS.
When a restricted certificate is added to the trust store in this mode, it's restrictions are not honoured, and the client has full access to LXD.
Details
When authorization was refactored to allow for generalisation (at the time for TLS, RBAC, and OpenFGA, see https://github.com/canonical/lxd/pull/12313), PKI mode did not account for the core.trust_ca_certificates configuration option. When this option is enabled, all CA-signed client certificates are given full access to LXD. This cherry-pick from Incus was added to LXD to fix the issue.
The cherry-pick fixed the immediate issue and allowed full access to LXD for CA-signed client certificates when core.trust_ca_certificates is enabled, but did not consider the behaviour of LXD when core.trust_ca_certificates is disabled.
When core.trust_ca_certificates is false, restrictions that are applied to a certificate should be honoured. Instead, they are being ignored due to the presence of a server.ca file in LXD_DIR.
PoC
# Install/initialize LXD
$ snap install lxd --channel 5.21/stable
$ lxd init --auto
$ lxc config set core.https_address=127.0.0.1:8443
# Use easyrsa for configuring CA: https://github.com/OpenVPN/easy-rsa
$ cp -R /usr/share/easy-rsa "/tmp/pki"
$ export EASYRSA_KEY_SIZE=4096
$ cd /tmp/pki
$ ./easyrsa init-pki
$ echo "lxd" | ./easyrsa build-ca nopass
$ ./easyrsa build-client-full lxd-client nopass
$ cp pki/ca.crt /var/snap/lxd/common/lxd/server.ca
$ cp pki/issued/lxd-client.crt ~/snap/lxd/common/config/client.crt
$ cp pki/private/lxd-client.key ~/snap/lxd/common/config/client.key
# Restart daemon.
$ systemctl reload snap.lxd.daemon
# Add a restricted certificate to the trust store.
$ token="$(lxc config trust add --name ca-test --quiet --restricted)"
$ lxc remote add tls "${token}"
# Our client has a CA-signed certificate, but it is restricted, so the client should not be able to view server config.
$ lxc config get tls: core.https_address
127.0.0.1:8443
Impact
I believe this vulnerability is low impact because PKI mode is:
- Not the standard or recommended mode of operation for LXD.
- While
core.trust_ca_certificatesdefaults to false, we believe that users who enable PKI mode will generally havecore.trust_ca_certificatesenabled to allow for passwordless PKI with CRL revocation (see https://github.com/canonical/lxd/issues/3832). When this mode is enabled, all clients with CA-signed certificates have root access* anyway.
*Note: If a restricted certificate is added before core.trust_ca_certificates is enabled, the certificate becomes unrestricted. We believe this was the original intention of the PR, but this should be changed to disallow any unintended permission change.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐹Go | github.com/canonical/lxd | all versions | 0.0.0-20240403103450-0e7f2b5bf4d2 |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for github.com/canonical/lxd. O3's reachability analysis confirms whether the vulnerable code path is actually invoked in your application, so you act on real exposure instead of every transitive match.
Fix
Update github.com/canonical/lxd to 0.0.0-20240403103450-0e7f2b5bf4d2 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms GHSA-jpmc-7p9c-4rxf is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 pinpoints whether GHSA-jpmc-7p9c-4rxf is reachable in your code and exactly where to fix it, then blocks exploitation in production at runtime until the patched version is deployed.
Tailored to GHSA-jpmc-7p9c-4rxf. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is GHSA-jpmc-7p9c-4rxf in your dependencies?
O3 detects GHSA-jpmc-7p9c-4rxf across Go dependencies and uses function-level reachability to confirm whether the vulnerable code path is actually reachable — not just present. No false positives.