CVE-2023-54050

What it actually does

Memory leak of a copied znode in UBIFS when insert_old_idx() fails after dirty_cow_znode calls copy_znode, leaving the allocated structure unreferenced and unreclaimed.

Why it matters to the business

Gradual memory exhaustion on Linux systems using UBIFS filesystems can degrade performance, trigger OOM conditions, or cause service disruptions over time, especially on long-running embedded or storage devices.

How an attack unfolds

An attacker triggers repeated UBIFS copy-on-write operations that force the insert_old_idx error path, accumulating leaked znodes until memory pressure leads to denial of service; no remote code execution or privilege escalation is involved.

How to fix it

Upgrade the Linux kernel to a version containing commit cc29c7216d7f057eb0613b97dc38c7e1962a88d2 or later; for xcaddy packages update to 0.4.4-r4 or newer. Reboot after applying the kernel update and verify UBIFS volumes mount cleanly.

How to confirm the fix worked

Confirm the running kernel contains the fix commit, exercise UBIFS write paths under error injection for insert_old_idx, and monitor /proc/meminfo or slabinfo for absence of growing UBIFS-related znode leaks.

Typical attack path

1. Mount or access a UBIFS volume
2. Induce conditions that cause dirty_cow_znode to invoke copy_znode
3. Force insert_old_idx to return an error
4. Repeat to accumulate unreclaimed znodes
5. Exhaust kernel memory leading to system instability

How to detect it

Review kernel logs for UBIFS error paths around dirty_cow_znode and insert_old_idx failures; correlate with rising slab or kmalloc usage not attributable to normal workload; no public exploit artifacts exist.

What your security team should do

• Prioritize kernel updates on all UBIFS-using systems within standard patch windows
• Add UBIFS memory usage to existing kernel memory monitoring dashboards
• Test the updated kernel on representative embedded and storage hardware before broad deployment