[CVE-2026-97950][MODERATE REGULAR] configfs: pin the symlink target's dirent instead of chasing ->ci_dentry

[CVE-2026-97950][MODERATE REGULAR] configfs: pin the symlink target's dirent instead of chasing ->ci_dentry

[AL-KERNEL]

CVE: CVE-2026-97950
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: configfs: pin the symlink target's dirent instead of chasing ->ci_dentry
Commit: 4f54beb2e7f6d399396466682fba3539bcdcb414
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=4f54beb2e7f6d399396466682fba3539bcdcb414
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97950
Analysis date: Fri, 25 Sep 2026 14:04:36 -0400
ActionableScore: 4
ActionableScore lower bound: unknown
Actionable bucket: Borderline / manual review recommended
Manual review required: YES

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-97950 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97950

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 REGULAR
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-2026-97950	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-97950 CHECK WITH IMPACT FROM ORIG NN LOW	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/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: '5.8';DESCR 'A race in configfs symlink creation can dereference a freed target dentry or configfs_dirent when the symlink target is removed concurrently. The item reference acquired by get_target does not keep item ci_dentry alive. A concurrent rmdir can unhash and release the dentry before create_link accesses its d_fsdata field. This is at least a local denial of service through a WARN or possible kernel failure. The use after free dereference also creates a limited but plausible memory safety risk if freed memory is reused. For the CVSS the PR:L is used for the paranoid score because delegated configfs access may be available to a restricted service account or container root without full host administrator privileges. The race requires a timing window and is not directly network reachable. Impact is at least local denial of service and in the worst case may allow limited confidentiality or integrity impact due to the use after free condition.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4)	YES	LOCK DISK UAF LINUS  DECREASED_TO_MODERATEREG_BASED_ON_FALSEPOSCHECKOFKP	NO	NO	checked

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

ActionableScore=4
PrimitiveClass=CLASS_B
RealUAF=YES
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES

1. ActionableScore

* Raw conservative score: 2
* Conservative score: 4
* Raw paranoid score: 2
* Paranoid score: 4
* Conservative floor applied: 2 -> 4. A concrete stale dentry and configfs_dirent use exists, host-admin-equivalent privilege is not positively established, and this is not clearly fixed async stale storage.
* Paranoid floor applied: 2 -> 4.
* Final recommended bucket: **Borderline / manual review recommended**::

2. Signal breakdown

* +2 Real UAF or lifetime violation. `get_target()` retains a `config_item` reference but does not retain the target dentry or its `configfs_dirent`. Concurrent `rmdir` may release the dentry and dirent before `create_link()` accesses `item->ci_dentry->d_fsdata` and calls the dirent lifetime helper.
* +0 Reliable kernel crash or strong DoS. The supplied evidence establishes `WARN_ON(!atomic_read(&sd->s_count))`, but does not establish a reliable panic or host-wide crash.
* +0 Local unprivileged trigger. Configfs is user-driven, but the supplied patch does not establish that an ordinary local user can create the relevant configfs objects, create the symlink, and remove the target without delegated or host administrative rights.
* +0 Remote reachability. The vulnerable operation is a local configfs filesystem operation. No network trigger is established.
* +0 Attacker-influenced active-object lifetime manipulation. The attacker may drive symlink and removal operations if authorized, but there is no supplied evidence of meaningful control over post-free reuse, replacement contents, or allocator placement.
* +0 Structured stale-object exploitation leverage. The stale use is a dentry-to-dirent lifetime access and a refcount-related helper call. No callback, operations table, attacker-controlled field consumption, directed memory access, or meaningful reused-object interpretation is established.
* +0 Concrete repeated release/refcount operation. The patch establishes a release followed by a stale `configfs_get()`-style access, not an exact same-reference second `put`, `release`, `free`, or decrement operation after a first lifetime-ending release. The dedicated repeated-release signal is therefore not awarded.
* +0 Paranoid repeated-refcount leverage. The stale access may affect a released dirent reference count, but the supplied control flow does not establish a concrete post-free second release/decrement mutation or controlled reuse.
* +0 Confidentiality and integrity. No disclosure path, arbitrary write, controlled replacement, protected-object modification, or privilege escalation is established.
* +0 Additional availability. The observed warning and any possible fault arise from the same stale-object condition. No distinct persistent availability consequence is shown.
* -0 Hard race penalty. This is a race, but the supplied evidence does not prove a rare one-shot or externally uncontrollable timing condition. Symlink creation and target removal may be repeatedly attempted by an authorized actor.
* -0 Host-admin privilege penalty. Host-admin-equivalent privilege is not positively established. Configfs subtree permissions and subsystem-specific object controls are not supplied.

Important non-awarded signals:

* Controlled reclaim is not demonstrated.
* Arbitrary or directed kernel write is not demonstrated.
* A callback or function-pointer path from stale memory is not demonstrated.
* Security-boundary bypass is not demonstrated.
* The previous CVSS hint of C:L/I:L is a conservative worst-case CVSS interpretation, but it does not independently establish actionable confidentiality or integrity impact.
* No concrete repeated release/refcount mutation on the same dangling reference is shown.
* No privilege escalation chain is supplied.

3. Reachability analysis

The triggering sequence requires access to a relevant configfs hierarchy and permission to perform both symlink creation and removal of the target object concurrently. The patch itself does not show the mode bits, subsystem-specific `allow_link` policy, mount permissions, or configuration ownership rules for the affected configfs consumer.

Configfs mounting and many configuration operations are commonly administrator-controlled, but that deployment fact is not sufficient evidence that the complete vulnerable sequence requires host-admin-equivalent privilege. Conversely, the patch does not prove that an ordinary local user can reach the complete sequence. Therefore `HostAdminRequired=UNKNOWN`.

No network namespace, user namespace, or capability scope analysis is present in the supplied material. Configfs access may be delegated by a specific kernel subsystem, but the complete operation is not shown to be namespace-scoped. Therefore `NamespaceScopedPrivilege=UNKNOWN`.

No special hardware, boot parameter, sysctl, or prior privileged device setup is established. The race is between a local symlink operation and `rmdir`. It may be repeatable for an actor with the necessary configfs permissions, but the evidence does not establish attacker-controlled post-free allocation or replacement.

4. Primitive and severity interpretation

The exact primitive is a real lifetime violation: `config_item_get()` keeps the config item alive but does not pin `item->ci_dentry`. Concurrent removal can release the dentry and its `configfs_dirent`, after which `create_link()` follows the stale `ci_dentry` pointer and uses the stale `d_fsdata` dirent.

This is an active structured kernel lifetime object rather than fixed DMA or fixed stale buffer storage. However, the demonstrated stale use is narrow: it obtains the dirent and reaches `configfs_get()`, with the supplied symptom being a reference-count warning. There is no shown controlled reclaim, attacker-shaped replacement object, callback dispatch, disclosure, arbitrary write, or security-boundary bypass.

This is not a concrete repeated-release/refcount case under the dedicated rule. The first removal may end dentry and dirent lifetime, but the supplied patch does not show the exact same surviving reference later receiving a second `put`, `release`, `free`, or decrement operation. The stale operation is instead an attempted get/use of the released dirent.

The correct primitive classification is `CLASS_B`: a concrete use-after-free and lifetime-management defect with possible kernel fault behavior, but without demonstrated exploitation leverage beyond the stale access. The practical demonstrated impact is local warning, potential kernel fault, or denial of service. A stronger memory-corruption or privilege-escalation consequence remains speculative.

`ImportantCandidate=NO` because the evidence does not establish meaningful unresolved reuse, write, callback, boundary, or repeated-lifetime-mutation leverage. `AutoDowngradeAllowed=YES` because this is currently a constrained Class B lifetime bug with no demonstrated controlled reclamation or active post-free corruption primitive. Manual review remains recommended to check real permissions for affected configfs subsystems and whether stale dirent reuse can proceed beyond the observed warning.

5. One-sentence report phrase

A race between configfs symlink creation and target removal can dereference a released dentry or configfs dirent, causing a local kernel warning or possible denial of service without a demonstrated controlled memory-corruption primitive.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Verify permissions and deployment exposure of the affected configfs subsystem, and determine whether `configfs_get()` can mutate reclaimed or attacker-reused dirent storage rather than stopping at the documented warning path.

7. Call-site confidence

Call-site confidence: medium

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

Patch: configfs: pin the symlink target's dirent instead of chasing ->ci_dentry
Commit: 4f54beb2e7f6d399396466682fba3539bcdcb414
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=4f54beb2e7f6d399396466682fba3539bcdcb414

Commit description:

create_link() reads the target's configfs_dirent from
item->ci_dentry->d_fsdata, relying on the item reference taken by
get_target().  That reference pins the item, not its dentry: the dentry is
pinned by DCACHE_PERSISTENT, which configfs_remove_dir() releases via
simple_rmdir() while the item is still alive.  A symlink racing with rmdir
of its target can therefore find ->ci_dentry freed and its dirent
released, triggering WARN_ON(!atomic_read(&sd->s_count)) in configfs_get().

Take the dirent in get_target() as well, under ->d_lock and atomically
with the item reference, and pass it down to create_link().  A hashed
dentry has not been killed yet, so its ->d_fsdata reference keeps the
dirent alive there.

Cc: stable@vger.kernel.org
Fixes: 7063fbf ("[PATCH] configfs: User-driven configuration filesystem")
Signed-off-by: Vasileios Almpanis <vasilisalmpanis@gmail.com>
Tested-by: Breno Leitao <leitao@debian.org>
Reviewed-by: Breno Leitao <leitao@debian.org>
Link: https://patch.msgid.link/20260730093435.195441-2-vasilisalmpanis@gmail.com
Signed-off-by: Breno Leitao <leitao@debian.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  fs/configfs/symlink.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=4f54beb2e7f6d399396466682fba3539bcdcb414

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

The original Linux kernel CVE announcement for CVE-2026-97950 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97950

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:L/I:L/A:H
  The Best / paranoid CVSS score: 5.8

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: unknown
  Paranoid ActionableScore: 4

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: NO
Published priority for this report (same as in Subject): MODERATE REGULAR

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).