CVE-2026-99180 in Linux
Summary
by MITRE • 10/28/2026
In the Linux kernel, the following vulnerability has been resolved:
unicode: free the map after a grace period
generic_ci_d_hash() and generic_ci_d_compare() read sb->s_encoding and dereference the unicode map while a pathwalk in rcu mode holds nothing but rcu_read_lock(). The filesystems that use them, ext4, f2fs and tmpfs, drop the map with utf8_unload() from ->put_super(), which runs when the last reference to the mount goes away. A lookup that is still inside such a filesystem at that point reads a freed map.
With a pathwalk held up in utf8_casefold_hash() while the last reference to a casefolded tmpfs goes away:
BUG: KASAN: slab-use-after-free in utf8nlookup+0x74c/0x950 Read of size 8 at addr ff110001030fe658 by task walker/98 Call Trace: <TASK> utf8nlookup+0x74c/0x950 utf8byte+0x160/0x1560 utf8_casefold_hash+0x167/0x230 generic_ci_d_hash+0xdd/0x1c0 link_path_walk+0x11ad/0x1d30 path_lookupat+0x63/0xb00 filename_lookup+0x1f1/0x5b0 vfs_statx+0xd7/0x3b0 [...]
Allocated by task 94: __kmalloc_cache_noprof+0x249/0x560 utf8_load+0x4e/0x570 shmem_parse_opt_casefold.isra.0+0x92/0x270 [...]
Freed by task 97: kfree+0x1db/0x5b0 shmem_put_super+0x72/0x170 generic_shutdown_super+0x14c/0x530 kill_anon_super+0x3a/0x60 deactivate_locked_super+0x9c/0xf0 cleanup_mnt+0x9f/0x130
So far that has been hard to hit. umount(2) waits for a grace period in namespace_unlock() before it drops the mount, so only a mount whose last reference goes away elsewhere, like the close of a file or a chdir out of a lazily unmounted directory, can race a walker. So it is reachable.
Delay the free until a grace period has passed. The map and the data table it holds a reference to then outlive any pathwalk that found the superblock before the mount went away. The reference on the data table module is dropped from the rcu callback as well. utf8-core is built in whenever CONFIG_UNICODE is set, so there is no module exit that would have to wait for the callbacks.
Overlayfs points the s_encoding of an overlay at the map of its first layer without taking a reference of its own. That stays correct. The layer is held by the overlay, so its map is dropped only after the overlay is gone.
If you want to get the best quality for vulnerability data then you always have to consider VulDB.
Analysis
by VulDB Data Team • 10/28/2026
The Linux kernel contains a use-after-free vulnerability within the Unicode casefolding subsystem that affects filesystems relying on RCU-based pathwalks for directory lookups and comparisons. Specifically, functions such as generic_ci_d_hash() and generic_ci_d_compare() access the superblock encoding structure and dereference associated unicode maps while holding only an rcu_read_lock during a pathwalk operation. This mechanism relies on the assumption that the referenced data structures will remain valid until the RCU grace period concludes. However, filesystem implementations like ext4, f2fs, and tmpfs invoke utf8_unload() within their put_super callback to release these maps when the final reference to the mount point is dropped. If a pathwalk operation is still traversing a directory within such a filesystem at the moment the last reference is released, it may attempt to read from memory that has already been freed by kfree(), leading to undefined behavior and potential kernel crashes as evidenced by KASAN slab-use-after-free reports.
The operational impact of this vulnerability stems from the race condition between mount teardown and ongoing file system operations. While standard unmounting via umount(2) waits for a grace period in namespace_unlock() before dropping the mount, thereby preventing most races, specific scenarios can still trigger the flaw. These include situations where the last reference to a casefolded tmpfs is released due to actions such as closing an open file descriptor or changing directory out of a lazily unmounted path. In these cases, a concurrent walker may hold an RCU read lock while attempting to hash or compare filenames using the now-freed unicode map. This can result in kernel panics, data corruption, or potentially exploitable conditions if an attacker can control the memory reallocation patterns following the free operation.
The resolution involves delaying the actual deallocation of the unicode map until after a full RCU grace period has elapsed. By utilizing rcu callbacks to defer the kfree() call and the dropping of references on underlying data table modules, the system ensures that any pathwalk operations initiated before the mount was unmounted will complete safely without accessing freed memory. This approach aligns with standard kernel synchronization practices for managing shared read-mostly structures in concurrent environments. The fix also correctly handles overlayfs scenarios where s_encoding points to a map from an underlying layer; since the overlay holds its own reference, the underlying map remains valid until the overlay itself is destroyed, preserving correctness without additional changes to that specific logic.
From a vulnerability classification perspective, this issue corresponds to CWE-416: Use After Free, as it involves accessing memory after it has been made available for reuse by another process or kernel subsystem. In terms of attack vectors and detection techniques within the MITRE ATT&CK framework, such vulnerabilities are often associated with Tactic 3: Persistence or Tactic 5: Defense Evasion if exploited to maintain access or hide activity, though primarily they fall under initial exploitation phases involving memory corruption. Mitigation strategies include applying the kernel patch that implements RCU-based deferred freeing for unicode maps and ensuring systems are updated to versions containing this fix. Administrators should also monitor for unusual kernel oops messages related to slab-use-after-free in utf8-related functions as an indicator of potential exploitation attempts or system instability resulting from unpatched vulnerabilities.