Technical impact
The Linux kernel omits proper error handling in the autofs_notify_daemon() function, allowing a replacement pipe to be installed while an old pipe still holds the wq_mutex. If the old pipe’s reader closes, the delayed write fails with –EPIPE and the generic error path re‑enters catatonic mode. This transition mistakenly acts on the current superblock state, closes the replacement pipe, and releases new daemon generation’s wait queues, effectively freezing the autofs service until the next valid request. The underlying weakness is a race condition identified as CWE-366, leading to a denial of shared autofs functionality during legitimate daemon restarts.
Affected Systems
All Linux kernel implementations are affected, including any distribution kernel that has not yet incorporated the autofs fix. The CVE data does not provide specific version ranges. The bug was reproduced in kernel v7.3‑rc5‑337‑gff47652a4b66c on a QEMU guest, indicating that the issue existed in development releases and likely persisted through later stable kernels until patched.
Risk and Exploitability
Based on the description, it is inferred that an unprivileged user could trigger the denial by performing autofs lookups that cause the daemon to restart. The exploit requires the ability to trigger an autofs lookup that causes the daemon to restart; an unprivileged user can often do this if they have permission to access autofs mounts. The EPSS score is < 1% and the CVSS score is 4.7, indicating moderate severity, and the vulnerability is not listed in the CISA KEV catalog. Because the attack path is low in complexity and the impact is the loss of a shared service, the overall risk is moderate to high on affected systems, though no active exploitation has been reported.
Debian DLA