[CVE-2026-97949][MODERATE 7.0] configfs: unhash the dentry before dropping the item in rmdir [ Upstream

[AL-KERNEL <alkernel@kernelcve.org>]

CVE: CVE-2026-97949
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: configfs: unhash the dentry before dropping the item in rmdir [ Upstream
Commit: 5d7cbfb65e47cc71d5d94d87d5d0d1679080f3eb
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=5d7cbfb65e47cc71d5d94d87d5d0d1679080f3eb
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97949
Analysis date: Fri, 25 Sep 2026 11:52:36 -0400
ActionableScore: 5
ActionableScore lower bound: 4
Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review
Manual review required: YES

Summary:
No one-sentence report phrase was found.

======================================================================
ABOUT THIS REPORT
======================================================================

The original Linux kernel CVE announcement for CVE-2026-97949 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97949

The original announcement does not normally provide a security severity
estimate, CVSS assessment, or enough information to determine whether the
reported kernel bug represents a practically relevant security issue.

This report was generated by AL-KERNEL, an AI-assisted Linux kernel
vulnerability analysis system developed by Alexander Larkin. It combines
an autonomous classifier with LLM-assisted technical analysis and a
separate ActionableScore mechanism.

The purpose of this report is to prioritize Linux kernel CVEs before
manual review, identify cases that require prompt investigation, and
support automatic closure of issues that are unlikely to have meaningful
security impact.

Published priority for this report: MODERATE 7.0
Manual review required: YES

A detailed explanation of the methodology and priority rules is included
at the end of this message.

======================================================================
AL-KERNEL CLASSIFICATION RESULT
======================================================================

CVE-2026-97949	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-97949 CHECK WITH IMPACT FROM ORIG NN MODERATE	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H';*CWE-416;CWE-362;*CWE-664;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '7';DESCR 'A race in configfs can result in a slab use-after-free because configfs_rmdir drops the final config_item reference while the removed dentry remains hashed until the VFS later unhashes it. A concurrent configfs_symlink target lookup can treat the hashed dentry as proof that the config_item is live and call config_item_get on freed memory. This is a local race and is not directly network reachable. For the CVSS the PR:L paranoid value is used because write access to an already mounted configfs subsystem can be delegated to a local service account or container root even when mounting and global configuration require higher privileges. Impact is at least local denial of service through a kernel crash. The confirmed slab use-after-free also creates a plausible worst case risk of confidentiality integrity and privilege impact if freed memory is reused in a controllable way. ';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5)	YES	DANGER DISK UAF LINUS SYZBOT KPANIC  KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFDH	NO	NO	checked

======================================================================
ACTIONABLESCORE ANALYSIS
======================================================================

ActionableScore=5
ActionableScoreLower=4
PrimitiveClass=CLASS_B
RealUAF=YES
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO

1. ActionableScore

* Raw conservative score: 3
  * +2 Real UAF or active lifetime violation.
  * +1 Plausible kernel crash or strong DoS from the KASAN-confirmed slab use-after-free.
* Conservative score: 4
  * FLOOR 4 applied. This is a real stale config_item use in an active configfs lookup path. Host-admin-equivalent privilege is not positively established by the supplied evidence.
* Raw paranoid score: 4
  * +2 Real UAF or active lifetime violation.
  * +1 Plausible kernel crash or strong DoS.
  * +1 Attacker-influenced active-object lifetime manipulation. A principal with write access to the relevant configfs subsystem can drive rmdir and symlink operations that create the stale-reference window.
* Paranoid score: 5
  * No additional floor was needed.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

2. Signal breakdown

Positive signals:

* +2 Real UAF or active lifetime violation.
  * `configfs_rmdir()` can drop the final `config_item` reference while its dentry remains hashed.
  * Concurrent `configfs_symlink()` resolves the target through `get_target()`.
  * `configfs_get_config_item()` interprets the still-hashed dentry as proof that `sd->s_element` is alive and reaches `config_item_get()` after object lifetime may have ended.
  * The commit includes a KASAN slab-use-after-free report in `config_item_get()`.

* +1 Reliable kernel crash or strong DoS.
  * The demonstrated consequence is a slab UAF in kernel configfs processing. A KASAN failure is shown, and an uninstrumented kernel can plausibly Oops or panic depending on the stale object state and fault handling configuration.
  * No separate Availability point is awarded because it would double-count the same crash outcome.

* +1 Attacker-influenced active-object lifetime manipulation. Paranoid only.
  * Configfs is a user-driven configuration filesystem.
  * A principal able to create or remove relevant configfs objects and create symlinks can repeatedly influence the target object lifetime transition and stale lookup timing.
  * This point is not awarded conservatively because the supplied material does not establish that an ordinary unprivileged user can write to the affected configfs subsystem.

Negative signals:

* No host-admin penalty applied.
  * The supplied patch does not prove that the complete trigger requires host root, init user namespace CAP_SYS_ADMIN, or another host-admin-equivalent privilege.
  * It also does not prove below-host-admin reachability.

* No hard-race penalty applied.
  * The bug is concurrent, but the supplied evidence does not establish that the race is rare, one-shot, externally uncontrollable, or difficult to repeat.
  * The syzbot report demonstrates that the window is reachable.

Important non-awarded signals:

* Controlled reclaim or controlled replacement of the freed `config_item` is not demonstrated.
* Arbitrary or directed kernel write is not demonstrated.
* Remote reachability is not established.
* Confidentiality and integrity points are not awarded. A UAF alone does not demonstrate disclosure, modification of protected state, or privilege escalation.
* Strong exploitation leverage is not awarded because no callback control, object substitution, post-free attacker-shaped contents, refcount takeover, or arbitrary read/write primitive is shown.
* No concrete repeated release or repeated `put()` operation is established.
  * The stale operation is `config_item_get()`, not a second release or a second lifetime-ending `put()`.
  * Therefore the dedicated concrete repeated release/refcount signal is not awarded.
  * FLOOR 5 is not applied under the repeated-release exception.
* Structured stale-object leverage is not separately awarded.
  * The stale object is an active `config_item`, but the patch only establishes a stale `config_item_get()` operation and does not show meaningful attacker-controlled post-reuse field consumption.

3. Reachability analysis

The evaluated attacker is a local principal that can manipulate the relevant configfs subsystem and concurrently perform object removal and symlink creation. The vulnerable sequence requires a configfs item to be removed through `rmdir` while another path resolves it as a symlink target.

The supplied evidence does not establish the exact permission model of the affected configfs subsystem. Mounting configfs and registering many configfs subsystems are commonly administrative operations, but that does not prove that every operation on an already mounted subsystem requires host-admin-equivalent privilege. Individual configfs subsystems can expose writable objects to delegated services or constrained management principals.

HostAdminRequired is therefore UNKNOWN. NamespaceScopedPrivilege is also UNKNOWN. The patch does not identify the relevant capability checks, user namespace ownership, or whether the complete create, remove, and symlink sequence can be performed inside an attacker-controlled namespace.

No network packet path is involved. No special hardware, boot option, firmware behavior, or root-controlled sysctl prerequisite is identified by the patch. The race is between local configfs operations and can potentially be retried by a principal with access to the affected directory hierarchy.

4. Primitive and severity interpretation

The exact primitive is a real slab use-after-free involving a `config_item`. The object lifetime may end after the final reference is dropped by the rmdir path, while a still-hashed dentry permits a concurrent symlink target lookup to obtain and use `sd->s_element`.

This is an active structured-object lifetime bug rather than fixed stale DMA or fixed stale storage. The stale path calls `config_item_get()`, which normally performs a refcount operation on the object. If the slab storage has already been freed or reused, this operation can modify memory outside the original object lifetime.

The demonstrated consequence is kernel memory safety failure and potential local DoS through an Oops, KASAN failure, or panic. The patch does not demonstrate controlled reallocation, attacker-selected replacement contents, arbitrary kernel read/write, callback dispatch, information disclosure, integrity impact, or privilege escalation.

This is CLASS_B rather than CLASS_A. The UAF is concrete and the stale operation is not merely passive metadata inspection, but there is insufficient evidence for controlled reclaim, active type confusion, directed corruption, or a demonstrated security-boundary consequence.

ImportantCandidate is YES because this is a confirmed slab UAF in an active refcounted kernel object and the stale `config_item_get()` mutation occurs after the original lifetime may have ended. The practical exploitability boundary depends substantially on which configfs subsystem is writable by an attacker and whether freed-object reuse can be influenced.

AutoDowngradeAllowed is NO because automatically treating this as an ordinary crash-only bug would discard a concrete active-object UAF with unresolved post-free refcount mutation behavior. This does not claim a proven LPE or arbitrary kernel write primitive.

Likely operational severity is Actionable Moderate pending manual review of reachable configfs subsystems, directory permissions, and allocator reuse behavior.

5. One-sentence report phrase

A local race between configfs rmdir and symlink target lookup can call config_item_get on a freed config_item while its dentry remains hashed, causing a slab use-after-free and potential kernel crash.

6. Manual review recommendation

MANUAL CHECK REQUIRED

The patch confirms a real slab UAF and a stale refcount acquisition path. Review which configfs subsystems are enabled, mounted, writable by non-host-admin principals, and capable of supporting repeatable rmdir and symlink races.

7. Call-site confidence

Call-site confidence: medium

======================================================================
UPSTREAM PATCH SUMMARY
======================================================================

Patch: configfs: unhash the dentry before dropping the item in rmdir [ Upstream
Commit: 5d7cbfb65e47cc71d5d94d87d5d0d1679080f3eb
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=5d7cbfb65e47cc71d5d94d87d5d0d1679080f3eb

Changed files:
  fs/configfs/symlink.c
  fs/configfs/dir.c

Diff excerpt:

Not included in this email. See the upstream URL for the full patch.

Full patch:
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=5d7cbfb65e47cc71d5d94d87d5d0d1679080f3eb

======================================================================
DETAILED REPORT METHODOLOGY
======================================================================

The original Linux kernel CVE announcement for CVE-2026-97949 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97949

The original CVE announcement normally does not include a security-level
estimate. In particular, it may not contain a CVSS assessment, an impact
level, or enough information to determine whether the reported bug is a
practically relevant security issue. One purpose of this parallel CVE list
is to provide that missing technical and prioritization information.

The original goal of the AL-KERNEL project was to prioritize Linux kernel
CVE analysis automatically before manual review. The system can also help
identify non-security issues that may be suitable for automatic closure.

This report was generated by AL-KERNEL, an AI-assisted Linux kernel
vulnerability analysis system developed by Alexander Larkin.

The first analysis stage combines an autonomous classifier with additional
LLM-based analysis. The autonomous classifier runs locally on a CPU and is
based on a backpropagation neural network. Together, these mechanisms
produce a technical vulnerability description, identify likely weakness
types, estimate CVSS severity, and provide input for ActionableScore.

Two CVSS estimates are retained because incomplete kernel vulnerability
information often permits more than one defensible interpretation:

  Conservative CVSS vector: AV:L/AC:H/PR:H/UI:N/S:U/C:N/I:N/A:H
  The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H
  The Best / paranoid CVSS score: 7

The conservative vector represents a lower-impact interpretation.
The Best/paranoid vector intentionally represents a plausible upper-bound
interpretation and should not automatically be treated as demonstrated
real-world impact.

CVSS may also need to be adjusted for a particular Linux deployment,
because actual reachability, privileges, enabled kernel configuration,
hardware, namespaces, exposed device nodes, and other environmental
conditions can differ significantly between systems.

A separate ActionableScore mechanism evaluates practical remediation
urgency. Its analysis may include reachability, attack prerequisites,
subsystem exposure, memory-corruption characteristics, denial-of-service
reliability, and possible confidentiality, integrity, or
privilege-escalation impact.

  Conservative ActionableScore: 4
  Paranoid ActionableScore: 5

The final base severity is taken directly from the second tab-separated
field of the AL-KERNEL classification result. ActionableScore does not
replace or independently override that final AL-KERNEL decision, and
if ActionableScore adjusted impact level of ALKERNEL, then you would see
self-readable flags above like INCREASED_TO_HIGH_BASED_ON_ACTIONABLESCOREHIGHEREQTHAN7.

For an AL-KERNEL result of MODERATE, this report uses the following
additional presentation split:

  ActionableScore below 5   -> MODERATE REGULAR
  ActionableScore 5 or more -> MODERATE 7.0

The distinction between MODERATE REGULAR and MODERATE 7.0 makes it
possible to identify Moderate issues that should receive manual analysis
and fixes before lower-priority MODERATE REGULAR issues. In many cases,
MODERATE REGULAR fixes may wait for a later rebase or routine update.

There is one override in which MODERATE REGULAR becomes MODERATE 7.0
even when the ActionableScore is below 5. When the AL-KERNEL result
contains the KPANIC flag, a MODERATE result is always presented as
MODERATE 7.0. The KPANIC flag selected with few regexps without
usage of AI at all, so it helps to detect cases when Kernel Crash happens
and similar (to filter False-Negative results from the LLM usage).

KPANIC indicates that a reliable kernel crash, kernel panic, or similarly
serious kernel availability impact was identified by the classification
workflow.

AL-KERNEL base severity for this report: MODERATE
KPANIC detected for this report: YES
Published priority for this report (same as in Subject): MODERATE 7.0

These results are intended to support engineering triage. They are
machine-generated estimates, and cases marked for manual review should
be validated by a human security engineer before final disposition.
For more info read docs linked from here: https://kernelcve.org/
(and you can submit you own patch there to generate such a report
for non-existant CVE-id yet).

Note that in many cases this AI tool selects higher severity, than
real is (means you can expect Importants instead of Moderate 7.0 or
Moderates 7.0 instead of regular Moderates). If you see such cases,
please use reply email interface to add additional manual analyses
info to this particular CVE.
And please, please, let me know when you see Lows instead of Importants
or Important instead of Low (because particular for such cases I
need to tune this AI tool to make it better for this one and next similar).
My contact email for such notifications is alexanjelausa@gmail.com (and both
send reply to CVE record itself too and see "reply" button below for howto reply).