Skip to main content
Vulnerability Database/CVE-2026-98315

CVE-2026-98315: Linux Kernel NTFS Privilege Escalation

CVE-2026-98315 is a privilege escalation vulnerability in the Linux kernel NTFS implementation affecting runlist update operations. This article covers the technical details, affected versions, security impact, and mitigation strategies.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2026-98315 Overview

CVE-2026-98315 is a race condition in the Linux kernel's NTFS file system implementation. The ntfs_non_resident_attr_shrink() function invokes runlist helpers that require the runlist write lock, but the function did not hold the lock while freeing clusters and truncating the runlist. Concurrent operations on the same NTFS attribute can corrupt in-memory runlist state and on-disk mapping pairs. A local user with the ability to trigger NTFS attribute shrink operations can leverage this flaw to cause file system corruption, kernel memory corruption, or denial of service.

Critical Impact

Local attackers can trigger NTFS metadata corruption and kernel-level memory integrity failures through concurrent attribute operations on mounted NTFS volumes.

Affected Products

  • Linux kernel NTFS file system driver
  • Distributions shipping affected mainline kernel versions prior to the stable patch
  • Systems mounting NTFS volumes with write support enabled

Discovery Timeline

  • 2026-10-06 - CVE-2026-98315 published to the National Vulnerability Database
  • 2026-10-07 - Last updated in NVD database

Technical Details for CVE-2026-98315

Vulnerability Analysis

The vulnerability resides in the Linux kernel NTFS driver, specifically in the non-resident attribute shrink path. The ntfs_non_resident_attr_shrink() function calls runlist helpers that are documented to require the runlist write lock. The function failed to acquire that lock before freeing clusters and truncating the runlist. Concurrent threads performing runlist mapping, extension, or resident conversion can observe inconsistent runlist state during the shrink operation.

A related issue affects ntfs_attr_map_cluster(). The function can merge a newly allocated run before updating the on-disk mapping pairs. If the mapping pairs update fails, the kernel must free the newly allocated clusters and restore both the in-memory runlist and on-disk mapping pairs from a saved copy. Without proper locking and rollback, failure paths leave the volume in an inconsistent state.

The attack surface is local. Attackers require the ability to perform file operations against an NTFS-mounted volume, including file truncation and sparse write patterns that trigger runlist shrink and remap paths [CWE-667, CWE-362].

Root Cause

The root cause is improper synchronization. Runlist helpers invoked from ntfs_non_resident_attr_shrink() document a precondition requiring the runlist write lock. The caller did not satisfy that precondition. Cluster freeing, runlist truncation, and resident conversion executed outside the lock, allowing other threads to observe or mutate shared runlist structures mid-update.

Attack Vector

A local, authenticated user with access to an NTFS volume mounted with write support can trigger the race. Workloads that shrink non-resident attributes, such as file truncation, hole punching, and sparse write sequences, can race with concurrent reads or writes that traverse the runlist. Successful races corrupt runlist structures and mapping pairs, resulting in file system damage or kernel memory corruption. The vulnerability does not require user interaction and does not expose a remote attack surface.

The upstream fix serializes the shrink, cluster-free, truncate, and resident conversion operations under the runlist lock. The patched ntfs_attr_map_cluster() path adds rollback logic that frees newly allocated clusters and restores both the in-memory runlist and on-disk mapping pairs from a saved runlist when the mapping pairs update fails. If either rollback step fails, the volume is marked in error to prevent further corruption. See the upstream fixes at Kernel Git Commit 742797432 and Kernel Git Commit 91709ba5d.

Detection Methods for CVE-2026-98315

Indicators of Compromise

  • Kernel log messages from the NTFS driver reporting runlist inconsistencies, mapping pair update failures, or volume error state transitions.
  • Unexpected file system corruption on NTFS volumes, including truncated or unreadable files after concurrent write workloads.
  • Kernel oops or panic traces referencing ntfs_non_resident_attr_shrink, ntfs_attr_map_cluster, or runlist helper functions.

Detection Strategies

  • Audit mounted file systems to identify hosts actively using the in-kernel NTFS driver with write support.
  • Correlate kernel ring buffer messages (dmesg) with NTFS workload timing to identify race-condition artifacts.
  • Monitor process execution for user-space tools that generate concurrent truncate and write operations against NTFS files.

Monitoring Recommendations

  • Centralize kernel log collection and alert on NTFS error strings and volume-marked-in-error events.
  • Track kernel version inventory across the fleet to identify hosts running unpatched kernels with NTFS support compiled in.
  • Monitor crash reporting services for kernel panics with stack frames inside the NTFS runlist code paths.

How to Mitigate CVE-2026-98315

Immediate Actions Required

  • Apply the upstream kernel patches referenced in Kernel Git Commit 742797432 and Kernel Git Commit 91709ba5d once distribution packages are available.
  • Restrict local user access to systems that mount NTFS volumes with write support until patches are deployed.
  • Mount NTFS volumes read-only where write access is not required for the workload.

Patch Information

The fix serializes runlist updates under the runlist write lock inside ntfs_non_resident_attr_shrink(). The patch also adds rollback logic to ntfs_attr_map_cluster() that frees newly allocated clusters and restores the in-memory runlist and on-disk mapping pairs from a saved runlist when the mapping pairs update fails. The volume is marked in error if either rollback step fails. Track your distribution vendor's advisory channel for backported stable kernel updates.

Workarounds

  • Mount NTFS file systems with the ro option to prevent attribute shrink paths from executing.
  • Use the FUSE-based ntfs-3g driver instead of the in-kernel NTFS driver on affected hosts where read-write access is required.
  • Blacklist the ntfs kernel module on systems that do not require NTFS support.
bash
# Example: mount NTFS volume read-only as a temporary mitigation
mount -t ntfs -o ro /dev/sdX1 /mnt/ntfs

# Example: prevent the in-kernel NTFS driver from loading
echo 'blacklist ntfs' | sudo tee /etc/modprobe.d/blacklist-ntfs.conf
sudo update-initramfs -u

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.