[CVE-2026-98023][MODERATE 7.0] vxlan: reject dynamic fdb entries that reference a nexthop id [ Upstream

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

CVE: CVE-2026-98023
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: vxlan: reject dynamic fdb entries that reference a nexthop id [ Upstream
Commit: 28f86eec851b7406a4bc6143f82aeba2882d9842
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=28f86eec851b7406a4bc6143f82aeba2882d9842
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98023
Analysis date: Fri, 25 Sep 2026 15:06:36 -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-2026-98023 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98023

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-2026-98023	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-98023 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-362;*CWE-416;*CWE-664;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'A race condition in the VXLAN FDB nexthop handling can corrupt a shared per nexthop FDB list when dynamic entries on multiple VXLAN devices are aged or modified concurrently. The race can result in list corruption and a slab use after free in vxlan_fdb_destroy. For the CVSS the PR:L rating is used because reliable triggering requires a local process that can configure VXLAN devices, nexthop objects, and FDB entries through network administration capabilities. The issue is not directly triggered by ordinary VXLAN packet traffic because affected FDB nexthop entries are configured from userspace and cannot be learned or roam. Impact is at least a local denial of service through kernel memory corruption and crash. In the worst case the confirmed use after free and list corruption may permit confidentiality or integrity impact and should be reviewed manually.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5)	YES	RACE UAF NETWORK LOG HARDWARE TEST IPV6 KPANIC  KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFDH	YES	NO	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: 4
* Paranoid score: 5
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

Raw arithmetic is plus 2 for a confirmed UAF and active shared-list corruption, plus 1 for realistic kernel crash or DoS, plus 1 for attacker-influenced repeatable FDB lifetime manipulation. No negative privilege or race penalty is applied because host-admin requirement and race difficulty are not positively established. FLOOR 5 applies because this is an active structured shared-list lifetime violation, the attacker controls creation and update of relevant FDB entries, and repeated configuration operations can drive the vulnerable lifetime transitions.

2. Signal breakdown

* +2 Real UAF and active lifetime corruption. The commit explicitly reports KASAN slab-use-after-free in vxlan_fdb_destroy after shared per-nexthop FDB list corruption.
* +1 Reliable kernel crash or strong DoS. list_del corruption and slab-use-after-free in VXLAN cleanup can produce a kernel warning, Oops, or crash.
* +1 Attacker-influenced active-object state and repeatable lifetime manipulation. Userspace configures the dynamic FDB entries, shared nexthop IDs, and VXLAN objects whose concurrent aging and modification causes the shared-list race.
* +0 Remote reachability. Ordinary VXLAN data-plane traffic does not configure these FDB nexthop entries. The vulnerable configuration path is userspace control-plane netlink input.
* +0 Controlled reclaim or object replacement. The patch does not demonstrate attacker-controlled reuse of the freed slab object.
* +0 Arbitrary or directed write. list corruption can modify list pointers, but the supplied evidence does not establish attacker-selected targets or attacker-controlled write contents.
* +0 Structured stale-object leverage. The stale operation is in FDB destruction and shared-list cleanup. No post-reuse callback, operations table, attacker-controlled replacement fields, or meaningful stale-field consumption is demonstrated.
* +0 Weak LPE concern. A UAF and list corruption alone do not establish privilege escalation leverage.
* +0 Additional availability. The same crash mechanism is already counted by the reliable kernel crash signal.
* +0 Confidentiality or integrity consequence. No read-back, disclosure, protected-object modification, credential corruption, or policy bypass is established.
* No hard-race negative point. The race involves timer aging in softirq context and FDB netlink operations, but the supplied evidence does not prove that the timing window is unusually narrow, one-shot, or difficult to repeat.
* No host-admin privilege penalty. CAP_NET_ADMIN is likely relevant, but the supplied patch does not resolve whether all required VXLAN, nexthop, and FDB operations require host-init-user-namespace authority.

Important non-awarded signals:

* No concrete repeated release or refcount mutation on the same dangling reference is established.
* No concrete double free or second put operation is visible.
* No security-boundary bypass is demonstrated.
* No remote packet-only trigger is established.
* No special hardware prerequisite is present.

3. Reachability analysis

The evaluated attacker is a local userspace control-plane actor, not a remote VXLAN packet sender. Triggering requires configuration of at least two VXLAN devices that reference a common FDB nexthop and creation or update of dynamic FDB entries through rtnetlink or bridge netlink operations.

The relevant operations normally require CAP_NET_ADMIN. VXLAN devices, nexthop objects, and FDB entries are generally network-namespace-aware objects, and many rtnetlink capability checks are evaluated against the target network namespace user namespace. However, the supplied patch does not include the complete capability-check and object-creation call chain. It therefore does not establish that an ordinary unprivileged user can create a user and network namespace, create all required VXLAN and nexthop objects, and execute the complete vulnerable sequence without host-admin-equivalent setup.

HostAdminRequired is UNKNOWN rather than YES because the evidence does not prove that host-init-user-namespace CAP_NET_ADMIN is required. NamespaceScopedPrivilege is also UNKNOWN because only likely namespace-aware portions of the operation are known, while the complete required VXLAN and nexthop configuration chain is not visible.

No special hardware, physical access, firmware influence, boot option, or host-provisioned EVPN deployment is required by the demonstrated mechanics. Although the configuration is intended for userspace control planes such as EVPN, the affected entries are configured through userspace and can be directly manipulated by a sufficiently privileged local networking actor.

The race is plausibly repeatable because the attacker can repeatedly add, update, and remove relevant FDB entries while aging runs asynchronously. The exact reliability of hitting the aging window is not established.

4. Primitive and severity interpretation

The exact primitive is a shared active VXLAN per-nexthop FDB list corruption followed by slab-use-after-free in vxlan_fdb_destroy. Dynamic FDB entries that reference the same nexthop can belong to multiple VXLAN devices, while their list is protected only by a per-device hash lock. Aging executes in softirq context without RTNL, allowing deletion on one device to race with modification or deletion on another device.

This is an active structured kernel-object lifetime violation, not fixed stale storage such as a stale DMA buffer. The affected FDB entries remain members of a shared list that is concurrently traversed and modified. The demonstrated stale use occurs during FDB aging and destruction, but the shared-list mutation is active kernel processing rather than a simple poisoned-pointer dereference.

There is no concrete repeated release, put, refcount decrement, or double-free operation established for the same dangling reference. The patch establishes list corruption and a UAF, but does not establish controlled slab reclaim, attacker-shaped replacement data, arbitrary read or write, callback control, credential corruption, or a privilege-escalation chain.

PrimitiveClass is CLASS_B because a real kernel lifetime and list-corruption primitive exists, but strong exploitation leverage is unresolved. The demonstrated practical consequence is local kernel DoS through warning, Oops, or crash. A stronger consequence remains technically possible because an active shared list is corrupted, but it is not proven from the supplied code and commit context.

ImportantCandidate is YES because this is not a validation-only failure or ordinary stale cleanup fault. It is a confirmed active-object UAF with list corruption, attacker-controlled object configuration, and repeatable lifecycle pressure. AutoDowngradeAllowed is NO because automatically treating it as an ordinary crash-only issue would discard unresolved exploitation analysis for a real shared-list lifetime corruption primitive.

Likely operational severity is Actionable Moderate at minimum. Manual analysis should determine whether all required VXLAN, nexthop, and bridge FDB operations can be performed inside attacker-created user and network namespaces, and whether the corrupted list can be further traversed or modified after stale-object reuse.

5. One-sentence report phrase

A local networking control-plane actor can race VXLAN dynamic FDB aging against FDB updates on multiple devices sharing one nexthop, causing shared-list corruption and a slab-use-after-free in vxlan_fdb_destroy that can crash the kernel.

6. Manual review recommendation

MANUAL CHECK REQUIRED

The patch confirms KASAN slab-use-after-free and list corruption in an active shared VXLAN FDB structure, while privilege scope and post-corruption exploitation leverage remain unresolved.

7. Call-site confidence

Call-site confidence: medium

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

Patch: vxlan: reject dynamic fdb entries that reference a nexthop id [ Upstream
Commit: 28f86eec851b7406a4bc6143f82aeba2882d9842
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=28f86eec851b7406a4bc6143f82aeba2882d9842

Changed files:
  lib/list_debug.c
  drivers/net/vxlan/vxlan_core.c
  tools/testing/selftests/net/fib_nexthops.sh

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=28f86eec851b7406a4bc6143f82aeba2882d9842

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

The original Linux kernel CVE announcement for CVE-2026-98023 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98023

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:L/UI:N/S:U/C:L/I:L/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).