The Linux kernel’s gfs2 filesystem contains a race condition in the gfs2_logd routine. It calls log‑flushing functions without holding the required sdp->sd_log_flush_lock, allowing concurrent transactions to interleave. This flaw can lead to inconsistent or corrupted log data, potentially cascading into broader filesystem integrity problems. The weakness is a classic improper synchronization (CWE‑362) and race condition.

All Linux kernel releases that ship the gfs2 filesystem are susceptible. The specific kernel versions that compile gfs2 are affected until the issue is mitigated by the commit described in the references. Exact version ranges are not supplied.

The CVSS score is not listed but the EPSS score is not available, so exploitation likelihood is uncertain. Because the flaw resides in kernel code, a successful exploitation requires execution with kernel privileges, which simplifies the attack once an attacker gains elevated access. The issue is not catalogued in CISA KEV. The likely attack path is local privilege escalation or compromising the kernel via a malicious driver that triggers gfs2_logd concurrently with other transactions; based on the description, this is inferred rather than explicitly stated.