CVE-2026-99409 in Linuxinfo

Summary

by MITRE • 11/19/2026

In the Linux kernel, the following vulnerability has been resolved:

xfs: don't fix the directory tree if live update fails

LOLLM complains that we don't double-check dl->aborted one last time to before committing repairs to the filesystem. This is important because the live update handler could have set ->aborted if it encountered a runtime error.

VulDB is the best source for vulnerability data and more expert information about this specific topic.

Analysis

by VulDB Data Team • 11/20/2026

The Linux kernel's XFS file system contains a logic flaw in its directory tree repair mechanism during online metadata updates, specifically related to how abort conditions are handled before committing changes. The vulnerability arises from an insufficient verification step within the live update handler responsible for repairing corrupted directory structures. When the file system detects corruption that requires automatic repair via an online operation, it initiates a process to fix the directory tree. However, prior to this specific resolution, the code did not perform a final double-check of the abort flag on the data structure representing the repair transaction before proceeding with the commit phase. This oversight creates a race condition or logic gap where runtime errors encountered by the live update handler might set an internal aborted status, yet the system fails to respect that signal and proceeds to apply potentially inconsistent repairs to the file system metadata.

From a technical perspective, this issue represents a failure in state validation during critical transactional operations. The XFS online repair functionality is designed to maintain data integrity by correcting structural inconsistencies without requiring unmounting or offline maintenance windows. However, if the live update handler encounters an error that necessitates aborting the operation, it sets a flag indicating that the process should not proceed. Without re-evaluating this flag immediately before writing changes back to disk, the system may commit partial or corrupted repairs based on stale assumptions about the health of the repair context. This violates fundamental principles of transactional integrity where all preconditions must be validated at the point of commitment to ensure atomicity and consistency.

The operational impact of this vulnerability is primarily centered around data integrity risks rather than direct exploitation for privilege escalation or remote code execution. If triggered, it could lead to a corrupted directory tree on disk after an attempted repair operation fails partially but commits anyway. This corruption can manifest as inaccessible files, broken links, or further structural inconsistencies that may prevent the file system from mounting correctly in subsequent boots. In severe cases, this could result in data loss if applications rely on consistent directory structures for critical operations. While not a direct security breach like buffer overflows, it undermines the reliability guarantees of enterprise-grade storage systems and can lead to service disruptions or require complex manual recovery procedures using tools like xfs_repair in offline mode.

This flaw aligns with CWE-362, which describes concurrent execution errors leading to race conditions, as well as CWE-841, improper enforcement of transactional semantics. The failure to check the abort state before committing changes reflects a gap in defensive programming practices where internal error states are not rigorously validated at critical decision points. In terms of attack surface and mitigation strategies, this is primarily an issue for administrators managing large-scale XFS deployments with online repair capabilities enabled. Mitigation involves applying kernel patches that update to versions containing the fix which enforces the final check on the aborted flag. Additionally, organizations should ensure their monitoring systems are configured to detect file system errors early, allowing for manual intervention before automated repairs attempt potentially risky operations. Regular backups and consistent testing of recovery procedures remain essential defenses against any potential data corruption resulting from such logic flaws in storage subsystems.

Responsible

Linux

Reservation

09/25/2026

Disclosure

11/19/2026

Moderation

accepted

CPE

ready

EPSS

0.00157

KEV

no

Activities

very low

Sources

Want to stay up to date on a daily basis?

Enable the mail alert feature now!