Affected versions: Oracle Linux 9

📖 ~4 min read  •  Source: ELSA advisory ELSA-2025-9147

Related CVEs: CVE-2025-22871

Table of contents
  1. Symptom & Impact
  2. Environment & Reproduction
  3. Root Cause Analysis
  4. Quick Triage
  5. Step-by-Step Diagnosis
  6. Solution – Primary Fix
  7. Solution – Alternative Approaches
  8. Verification & Acceptance Criteria
  9. Rollback Plan
  10. Prevention & Hardening
  11. Related Errors & Cross-Refs
  12. References & Further Reading

Symptom & Impact

buildah runs when an image is pulled, unpacked, inspected or a container is started, so a defect here is normally reached through untrusted image content or a registry response rather than inbound traffic to a listening port. On Oracle Linux 9 that shows up as escapes from the container’s namespaces, files written outside the intended root during buildah or podman build, mount and capability handling that grants more than was requested, and images produced before the update carrying the defect onward to every host that pulls them. buildah is container tooling — command-line binaries invoked per operation, though a node runtime such as cri-o does run as a unit — so restart steps apply to that runtime and to containers still running the old code, not to buildah itself.

Environment & Reproduction

Reproduction targets Oracle Linux 9 running either the Red Hat Compatible Kernel (RHCK) or the Unbreakable Enterprise Kernel (UEK). Confirm release, kernel, and installed package:

cat /etc/oracle-release
uname -r
rpm -q buildah
dnf list installed buildah
dnf history list --reverse | head -20

Exercise the workload that uses buildah while collecting:

sudo needs-restarting -u  # --useronly: restrict to the invoking user's processes (useful on shared/app hosts)
sudo lsof -n +L1  # link count < 1 = file deleted but still open/mmap'd; this is the ground truth after an RPM replaces a .so (package: lsof)
sudo lsof -n +L1 2>/dev/null | awk 'NR>1 {print $1, $2}' | sort -u  # condensed COMMAND + PID list of every process still running old code
journalctl -xe --no-pager | tail -200
tail -200 /var/log/dnf.log
tail -200 /var/log/audit/audit.log

Root Cause Analysis

Root cause is documented in ELSA advisory ELSA-2025-9147. Upstream Red Hat / Oracle Linux maintainers shipped a fix in the corresponding buildah errata; running an outdated build leaves the host exposed and may trigger the failure modes described in the advisory. Correlate transaction history with system logs and SELinux audit entries to isolate the originating change:

dnf history list --reverse | head -30
dnf history info $(dnf history list | awk '/buildah/ {print $1; exit}')
ausearch -m AVC,USER_AVC -ts today | tail -100
cat /proc/sys/kernel/tainted   # non-zero = kernel modules / out-of-tree drivers loaded

Quick Triage

Run these checks on Oracle Linux 9 to confirm the failure mode and current state of buildah:

rpm -q buildah                              # installed version
rpm -V buildah                              # verify file integrity
dnf updateinfo info --security buildah     # any security advisories outstanding
systemctl --failed --no-pager
firewall-cmd --list-all 2>/dev/null || echo 'firewalld not running'
getenforce                                 # SELinux mode
# e.g. httpd pkg/unit match, but bind→named, postgresql-server→postgresql):
dnf repoquery -l buildah 2>/dev/null | grep -E '/lib/systemd/system/.*\.(service|socket|timer)$'  # same check WITHOUT installing (EL7/Amazon Linux 2: yum install yum-utils, then: repoquery -l buildah)
systemctl show --no-pager -p Type,RemainAfterExit <UNIT>  # Type=oneshot (often paired with a .timer) means a one-shot job, NOT a daemon -- do not 'restart' it, let the timer fire

Step-by-Step Diagnosis

  1. List failed units.

    systemctl --failed --no-pager
  2. Follow the journal for buildah and the system bus.

    journalctl -xe -f --no-pager
  3. Check firewall posture (skip if firewalld is masked).

    firewall-cmd --list-all-zones --permanent
    nft list ruleset 2>/dev/null | head -50
  4. Surface SELinux denials and translate them to a policy module if needed.

    ausearch -m AVC,USER_AVC -ts today
    ausearch -m AVC -ts today | audit2allow -a -M /tmp/local-fix
    # Inspect /tmp/local-fix.te before applying:
    sudo semodule -i /tmp/local-fix.pp
  5. Verify buildah integrity and dependency closure.

    dnf check
    rpm -V buildah
    rpm -q --requires buildah | xargs -r rpm -q --whatprovides | head
  6. Correlate findings with /var/log/dnf.log, dnf history and ELSA advisory ELSA-2025-9147 to pin the change that introduced the regression.

Solution – Primary Fix

Apply the corrective dnf transaction referenced by ELSA advisory ELSA-2025-9147, reload affected systemd units, and reconcile firewalld / SELinux state:

sudo dnf clean expire-cache
sudo dnf -y update buildah
sudo systemctl daemon-reload
readelf -p .comment /path/to/binary 2>/dev/null | grep -o 'GCC:.*'  # ELF .comment records the exact compiler that produced the file (package: binutils); absent if the build stripped it
annocheck /path/to/binary  # RHEL-native: reports the build-time compiler and hardening flags (package: annobin-annocheck on RHEL/Alma/Rocky 9-10, annobin on RHEL 8)
go version -m /path/to/binary  # prints the Go toolchain and module versions baked into a Go binary -- rebuild anything below the patched go version
rpm -q buildah                       # confirm new NVR

If the advisory says a reboot is required (kernel, glibc, systemd, openssl):

sudo needs-restarting -r           # reports kernel/init/glibc need
sudo systemctl reboot              # or: sudo shutdown -r now

Need help applying this fix at scale? Our IT Solutions & Services team rolls Oracle Linux patches across estates with zero-downtime change windows and Ksplice live-patching. Get in touch for a free consultation.

Solution – Alternative Approaches

If the primary fix is not viable, choose from these alternatives:

  • Roll back the offending dnf transaction:

    sudo dnf history list --reverse
    sudo dnf history undo <id>
  • Pin buildah with the versionlock plugin:

    sudo dnf install -y python3-dnf-plugin-versionlock
    sudo dnf versionlock add buildah
    sudo dnf versionlock list | grep buildah
  • Downgrade to a known-good NVR from the repo cache or vault:

    sudo dnf --showduplicates list buildah
    sudo dnf -y downgrade buildah-<older-NVR>
  • Switch firewalld backend (nftables ↔ iptables) for compatibility:

    sudo sed -i 's/^FirewallBackend=.*/FirewallBackend=iptables/' /etc/firewalld/firewalld.conf
    sudo systemctl restart firewalld
  • If SELinux is suspected, switch to permissive briefly, capture denials, and author a custom module before re-enforcing:

    sudo setenforce 0                                    # do NOT leave permissive
    # reproduce the failure
    sudo ausearch -m AVC -ts recent | audit2allow -a -M mylocal
    sudo semodule -i mylocal.pp
    sudo setenforce 1
  • Where the advisory has Ksplice coverage, live-patch without reboot (Oracle Linux Premier Support):

    sudo uptrack-show                # current live patches
    sudo uptrack-upgrade -y          # apply all available
    uptrack-uname -r                 # effective kernel after live patching

Verification & Acceptance Criteria

All of these should pass after the fix:

rpm -q buildah                                       # shows the expected fixed NVR
dnf updateinfo list --security installed | head    # no security advisories pending for us
firewall-cmd --list-services                        # required services present
getenforce                                          # intended mode (Enforcing/Permissive)

The conditions described in the advisory must no longer be reported for buildah across two consecutive runs.

Rollback Plan

Capture state before any change:

rpm -qa > /root/rpm-pre.txt
dnf history list --reverse > /root/dnf-history-pre.txt
# LVM snapshot of the root LV (size to ~10% of root):
sudo lvcreate -L 4G -s -n root_pre_patch /dev/mapper/$(lvs --noheadings -o lv_path | grep -m1 root | xargs basename)

To revert if the patch is bad:

sudo dnf history undo <id>
# Or downgrade just buildah to the previous NVR:
sudo dnf -y downgrade buildah
sudo systemctl daemon-reload
# For SELinux module additions:
sudo semodule -r mylocal
# Reboot only if kernel/initramfs/glibc were rolled back:
sudo systemctl reboot

For kernel rollbacks, select the previous entry from the GRUB menu or set it as default with grubby --set-default /boot/vmlinuz-<older>.

Prevention & Hardening

Prevent recurrence on Oracle Linux 9 hosts running buildah:

  • Enable scheduled security updates via dnf-automatic:

    sudo dnf install -y dnf-automatic
    # Edit /etc/dnf/automatic.conf:
    # upgrade_type = security
    # apply_updates = yes
    sudo systemctl enable --now dnf-automatic.timer
  • Subscribe to the Oracle Linux Errata RSS / mailing list at linux.oracle.com/security.

  • Mirror through a local yum/dnf repository:

    sudo dnf install -y dnf-utils createrepo_c
    sudo reposync --download-metadata --downloadcomps -p /srv/repos -m --repo=ol9_baseos_latest
    sudo createrepo_c /srv/repos/ol-baseos
  • Pin sensitive packages so they cannot be auto-upgraded:

    sudo dnf install -y python3-dnf-plugin-versionlock
    sudo dnf versionlock add buildah
  • Snapshot the root LV before every upgrade window:

    sudo lvcreate -L 4G -s -n root_pre_$(date +%Y%m%d) /dev/<vg>/<root-lv>
  • Monitor file integrity with AIDE:

    sudo dnf install -y aide
    sudo aide --init && sudo mv /var/lib/aide/aide.db.new.gz /var/lib/aide/aide.db.gz
    sudo aide --check
  • Lock down with SELinux audit rules in /etc/audit/rules.d/:

    # /etc/audit/rules.d/90-cp.rules
    -w /etc/passwd  -p wa -k identity
    -w /etc/shadow  -p wa -k identity
    -w /etc/sudoers -p wa -k privilege
    -a always,exit -F arch=b64 -S execve -k exec
  • Where licensed, enable Oracle Ksplice for live kernel and userspace patching:

    sudo dnf install -y uptrack ksplice-tools
    sudo uptrack-upgrade -y
    sudo systemctl enable --now uptrack

Issues that commonly surface alongside a buildah update: dnf transaction lock contention, systemd unit ordering cycles, SELinux AVC bursts, firewalld zone drift, and kernel taint flags. Triage with:

cat /proc/sys/kernel/tainted
systemd-analyze critical-chain
ausearch -m AVC -ts today | tail
firewall-cmd --get-active-zones
dnf history list --reverse | head

View all oracle-linux-9 tutorials on the Tutorials Hub →

Browse all common problems & solutions on the Tutorials Hub.

References & Further Reading

Primary reference: ELSA advisory ELSA-2025-9147. Useful manual pages on Oracle Linux 9:

man dnf
man dnf.conf
man systemctl
man journalctl
man firewall-cmd
man semanage
man audit2allow
man grubby

Other resources: Oracle Linux 9 Administrator’s Guide at docs.oracle.com, the upstream Red Hat CVE database at access.redhat.com/security/cve, the Oracle Ksplice known-fixes feed, and /usr/share/doc/buildah/ for component-specific notes implicated in buildah.