The vulnerability originates in the GFS2 filesystem of the Linux kernel. During inline data writes, the buffer head that holds the inode metadata is released prematurely while the iomap structure still references its data. This results in a use‑after‑free that causes the kernel to copy data into a de‑allocated buffer, leading to kernel memory corruption. If an attacker induces this faulty write path, the corrupted memory could be leveraged to alter kernel objects or execute arbitrary code, potentially elevating privileges.

Any Linux kernel that implements the unpatched GFS2 inline data write logic is affected. The CVE entry does not specify a version range, so systems running an unpatched kernel that uses GFS2 are at risk. The affected vendor is Linux itself, through the kernel implementation.

The CVSS score of 7.8 denotes a moderate to high severity, while the EPSS score of < 1% and the absence from the CISA KEV catalog indicate that exploitation is unlikely so far. The likely attack vector requires local filesystem access to trigger the inline write path; this inference is deduced from the description that points to an inline write operation within the GFS2 filesystem. If exploited, the use‑after‑free could permit a local attacker to corrupt kernel memory or achieve privilege escalation.