[CVE-2026-98231][MODERATE 7.0] xfrm: serialize state GC with device state flush

[CVE-2026-98231][MODERATE 7.0] xfrm: serialize state GC with device state flush

[AL-KERNEL]

CVE: CVE-2026-98231
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: xfrm: serialize state GC with device state flush
Commit: 937108dc0258d6ac69926b00bc53598c98986ee1
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=937108dc0258d6ac69926b00bc53598c98986ee1
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98231
Analysis date: Fri, 09 Oct 2026 11:25:40 -0400
ActionableScore: 5
ActionableScore lower bound: unknown
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-2026-98231 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98231

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-2026-98231	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-98231 CHECK WITH IMPACT FROM ORIG NN MODERATE | AEGIS IMPACT MODERATE_KPANIC AEGIS_AS=5 LEGACY=MODERATE_KPANIC RESULT=SAME (legacy and Aegis reconciliation agree; DECISION: both MODERATE_KPANIC)	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-667;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 between xfrm_dev_state_flush and the XFRM state GC worker can let both paths process the same xfrm_state because the deferred device GC list does not hold a state reference. After the GC worker destroys and frees the state, the device flush path can continue reading the freed object, invoke a driver callback again, and release the device reference again. KASAN confirms a slab use after free in xfrm_dev_state_free. The demonstrated impact is a local kernel crash or reference state corruption. A more serious outcome is plausible because execution continues through fields of a freed kernel object and includes a driver callback, so controlled slab reuse could potentially affect confidentiality, integrity, and availability. For the CVSS the PR:L value is used for the paranoid score because a process with delegated CAP_NET_ADMIN, container root, or a network management service account may be able to control the required XFRM and device operations without having unrestricted host root access. AC:H applies because successful triggering depends on a race window between device state flushing and asynchronous state destruction. The issue is not directly network reachable because remote traffic alone is not shown to initiate the required control plane operations. Impact is at least a local denial of service via kernel crash and in the worst defensible case may permit privilege escalation through kernel use after free. YES REQUIRES MANUAL CHECK.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5)	MAYBE	READ LOCK DANGER RACE UAF HARDWARE LINUS KPANIC  KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFDH	-	-	checked

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

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

1. ActionableScore

* Raw conservative score: 4
* Conservative score: 5
* Raw paranoid score: 5
* Paranoid score: 5
* Applied conservative REAL UAF FLOOR 5: 4 -> 5. The exact device reference can receive a second lifetime mutation after the first path may have released it.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

2. Signal breakdown

Conservative positive signals:

* +2 Real UAF. The state GC worker can free the `xfrm_state` while `xfrm_dev_state_flush()` retains and continues using the same state.
* +1 Concrete repeated release or refcount operation. Both paths can release the same `xso.dev` reference. The first `netdev_put` can affect or end the referenced device lifetime, while the saved reference remains available to the second path.
* +1 Reliable kernel crash or strong DoS. KASAN confirms a slab use-after-free in `xfrm_dev_state_free()`, and a non-instrumented kernel can Oops or corrupt reference state.

Paranoid-only signal:

* +1 Unresolved post-free lifetime mutation. The second path can invoke the driver cleanup and drop the saved device reference after the owning `xfrm_state` was destroyed. This is a concrete operation, not merely theoretical allocator reuse.

Floor evaluation:

* The first device-reference release can decrement a lifetime reference and potentially participate in final destruction.
* The same saved device reference survives into the concurrent path.
* The second release is a real lifetime mutation rather than a read-only stale dereference.
* Concrete repeated release or refcount +1 was awarded.
* Paranoid repeated-refcount leverage +1 was awarded because the second release operates on a reference whose ownership record has already been destroyed.
* FLOOR 5 was applied to the conservative result.

Important non-awarded signals:

* B5-R was not established. There is no heap OOB write, retry-accumulated index, or later consumption of retry-corrupted descriptors.
* Controlled reclaim or attacker-shaped replacement contents are not demonstrated.
* Arbitrary or directed kernel write is not demonstrated.
* Callback control is not demonstrated. Invoking the legitimate driver callback twice is not equivalent to controlling its function pointer.
* Structured stale-object leverage was not separately awarded because attacker influence over replacement state is missing.
* Weak LPE was not separately awarded for the same repeated-release fact.
* Confidentiality and integrity bonuses were not awarded. The supplied evidence establishes lifetime corruption but not disclosure or unauthorized data modification.
* Remote reachability was not established.
* No extra Availability point was awarded for the same crash.

3. Reachability analysis

The demonstrated path combines XFRM state creation, deferred device-state cleanup, state GC, and a network-device flag change through `devinet_ioctl()`. These operations generally involve network-administration authority, but the supplied evidence does not establish whether the complete trigger requires host or init-namespace `CAP_NET_ADMIN`, or whether it can be completed with delegated authority.

The namespace scope is unresolved. XFRM and device operations can involve network namespaces, but the presence of namespace-aware interfaces does not prove that hardware-offloaded state creation, device flushing, and all required lifecycle operations are available inside an attacker-created namespace.

Hardware offload or an offload-capable driver appears relevant to the deferred-device list, but the supplied material does not prove rare physical hardware or a mandatory host-only setup. No hardware or rarity penalty is therefore applied.

The bug is a race, but no timing penalty is applied. The report includes a working KASAN-triggered sequence, and it does not establish that the interleaving is one-shot, exceptionally narrow, or impractical to repeat. Direct packet traffic alone is not shown to trigger the complete path.

4. Primitive and severity interpretation

The exact primitive is a slab use-after-free of an active `xfrm_state`. The GC worker can destroy and free the state after the device-flush path drops `xfrm_state_dev_gc_lock`, because membership in the deferred-device list does not hold an XFRM state reference.

This is an active structured object rather than fixed stale storage. The stale path can continue through `xfrm_dev_state_free()`, invoke the driver callback again, and release the same device reference again.

This is category B, concrete repeated lifetime corruption, rather than only category A, an ordinary cleanup stale read. The first cleanup can release the device reference and destroy the XFRM state. The concurrent path retains state derived from that object and performs another device-reference release. The second operation therefore mutates lifetime state rather than merely reading a poisoned pointer.

The primitive remains CLASS_B because controlled reclaim, callback hijacking, refcount takeover, arbitrary read or write, and privilege escalation are not established. The demonstrated consequence is kernel crash or reference-state corruption. A stronger consequence remains unresolved because duplicate callback execution and duplicate reference release can affect live driver or device lifetime state.

ImportantCandidate is YES because the confirmed UAF includes a concrete repeated lifetime mutation that should not be discarded as an ordinary crash-only cleanup defect. AutoDowngradeAllowed is NO because automatic downgrade would lose this unresolved post-lifetime corruption mechanism. The likely operational severity is Actionable Moderate, with Important-class treatment dependent on manual confirmation of privilege scope and exploitation leverage.

5. One-sentence report phrase

A race between XFRM device-state flushing and state GC causes a confirmed slab use-after-free with duplicate driver cleanup and device-reference release, requiring manual Important-versus-Moderate review.

6. Manual review recommendation

MANUAL CHECK REQUIRED

The confirmed UAF includes a concrete second lifetime operation on the same device reference, while the complete privilege boundary and post-free exploitation leverage remain unresolved.

7. Call-site confidence

Call-site confidence: high

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

Patch: xfrm: serialize state GC with device state flush
Commit: 937108dc0258d6ac69926b00bc53598c98986ee1
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=937108dc0258d6ac69926b00bc53598c98986ee1

Commit description:

The deferred-device pass in xfrm_dev_state_flush() finds states under
xfrm_state_dev_gc_lock, but drops the lock before calling
xfrm_dev_state_free() because the driver callback may sleep.  The device
GC list does not hold an xfrm_state reference, so the state GC worker can
destroy the same state concurrently.

The race can proceed as follows:

  CPU 0                               CPU 1
  find x on the device GC list
  drop xfrm_state_dev_gc_lock
  read x->xso.dev
                                      xfrm_state_gc_destroy(x)
                                      xfrm_dev_state_free(x)
                                      xfrm_state_free(x)
  continue xfrm_dev_state_free(x)

Both paths can invoke the driver callback and drop the device reference.
CPU 0 can also access the xfrm_state after CPU 1 has freed it.

KASAN reported:

  BUG: KASAN: slab-use-after-free in xfrm_dev_state_free+0x24c/0x2a0
  Read of size 8 at addr ffff88810bbaa960 by task poc/102

  Call Trace:
   xfrm_dev_state_free+0x24c/0x2a0
   xfrm_dev_state_flush+0x353/0x400
   xfrm_dev_event+0x26d/0x3a0
   notifier_call_chain+0xc0/0x280
   __dev_notify_flags+0x169/0x250
   netif_change_flags+0xe7/0x160
   dev_change_flags+0x96/0x220
   devinet_ioctl+0x7f4/0x1880

  Allocated by task 87:
   xfrm_state_alloc+0x1e/0x5c0
   xfrm_add_sa+0xe7f/0x5820
   xfrm_user_rcv_msg+0x4f3/0x940

  Freed by task 57:
   kmem_cache_free+0xcb/0x3d0
   xfrm_state_gc_task+0x4a8/0x650
   process_one_work+0x63a/0x1070

Serialize xfrm_state destruction against the deferred-device pass with a
mutex.  Keep xfrm_state_dev_gc_lock limited to list operations and retain
the existing callback and device-reference release ordering.

Fixes: 07b87f9 ("xfrm: Fix unregister netdevice hang on hardware offload.")
Cc: stable@vger.kernel.org
Signed-off-by: Chengfeng Ye <nicoyip.dev@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  net/xfrm/xfrm_state.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=937108dc0258d6ac69926b00bc53598c98986ee1

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

The original Linux kernel CVE announcement for CVE-2026-98231 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98231

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: unknown
  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).