[CVE-2026-97965][MODERATE 7.0] vxlan: initialize _md in vxlan_xmit_one() [ Upstream

[CVE-2026-97965][MODERATE 7.0] vxlan: initialize _md in vxlan_xmit_one() [ Upstream

[AL-KERNEL]

CVE: CVE-2026-97965
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: vxlan: initialize _md in vxlan_xmit_one() [ Upstream
Commit: 0d13b5a413bffc8718b3821b78537ccc6596c233
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=0d13b5a413bffc8718b3821b78537ccc6596c233
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97965
Analysis date: Fri, 25 Sep 2026 14:24:16 -0400
ActionableScore: 4
ActionableScore lower bound: 3
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-2026-97965 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97965

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-2026-97965	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-97965 CHECK WITH IMPACT FROM ORIG NN LOW	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:N';CWE-908;*CWE-200;*CWE-457;Other CVSS 'AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N';BEST CVSS score: '4.8';DESCR 'An uninitialized stack vxlan_metadata structure can be used when a VXLAN device enables both collect metadata and GBP, but the external ip_tunnel_info does not include the VXLAN option flag. A packet sent through the affected VXLAN egress path can then carry stack derived bits in the GBP header, including a 16 bit policy_id field. This is network reachable only in deployments using this specific flow based VXLAN configuration. For the CVSS the PR:N value reflects that an attacker may only need to send traffic through an already configured affected VXLAN forwarding path and does not necessarily need local access to the host. The confirmed impact is limited kernel stack information disclosure. The paranoid score also assigns limited integrity impact because corrupted GBP flags or policy metadata may affect handling at the remote VXLAN endpoint. No memory corruption, privilege escalation, or denial of service primitive is demonstrated by the patch.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4)	YES	LEAK NETWORK HARDWARE KPANIC PACKET 	NO	NO	checked

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

ActionableScore=4
ActionableScoreLower=3
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=NO
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES

1. ActionableScore

* Raw conservative score: 3
* Conservative score: 3
* Raw paranoid score: 4
* Paranoid score: 4
* Final recommended bucket: **Borderline / manual review recommended**::

No floor or cap was applied.

2. Signal breakdown

* +2 Remote reachable packet path. In a deployment already using a VXLAN device with both collect metadata and GBP enabled, an attacker able to send traffic through the affected VXLAN egress path can cause a packet with externally visible uninitialized GBP metadata. This is conditional on a specific existing VXLAN control plane configuration.
* +1 Confidentiality plausible. The patch explicitly establishes that 16 bits of uninitialized kernel stack data can be placed in the outgoing GBP policy_id field and leaked onto the wire.
* +1 Paranoid only. Incorrect GBP flags, dont learn, policy applied, and policy_id values can plausibly influence packet treatment by the remote VXLAN endpoint. This supports limited integrity impact, but no protected-object modification or policy bypass is demonstrated.

Important non-awarded signals:

* No UAF, stale post-free object use, refcount misuse, repeated release, or lifetime corruption is established.
* No attacker-controlled kernel write, arbitrary write, controlled reclaim, type confusion, callback control, or privilege escalation primitive is established.
* No availability point is awarded. The patch does not establish a crash, hang, panic, or resource exhaustion consequence.
* No strong semantic reinterpretation bonus is awarded. The GBP header receives unintended metadata, but the supplied evidence does not establish an active ownership, resource, pointer, or authorization reinterpretation primitive.
* No broad common networking-path bonus is awarded. VXLAN is broadly available, but the affected collect metadata plus GBP combination is not shown to be broadly exercised.
* No niche subsystem penalty is applied. The configuration is specialized, but the supplied evidence does not prove that an attacker cannot reach an already configured affected forwarding path.

3. Reachability analysis

The relevant attacker is a network peer that can cause packets to traverse an already configured affected VXLAN egress path. The attacker does not need local code execution or local host administrative rights merely to provide packet traffic once the VXLAN forwarding setup exists. Therefore HostAdminRequired is NO for the triggering attacker model.

The vulnerable path requires a VXLAN device configured with both VXLAN_F_COLLECT_METADATA and VXLAN_F_GBP. It also requires an external ip_tunnel_info without IP_TUNNEL_VXLAN_OPT_BIT. These are deployment and control-plane prerequisites, commonly associated with flow-based VXLAN setups such as virtual switching or policy-managed overlay networking.

NamespaceScopedPrivilege is UNKNOWN. The affected device may exist in a network namespace, but the supplied patch does not establish that an attacker can create the complete collect metadata and GBP configuration, attach the required datapath, and trigger the affected external metadata transmission entirely from an attacker-controlled namespace.

No special hardware, boot option, race, or timing window is required. The issue is deterministic once the required configuration and packet path exist.

4. Primitive and severity interpretation

The primitive is an uninitialized stack read followed by direct disclosure of a limited value in an outgoing VXLAN GBP header. The disclosed data is specifically the 16 bit GBP policy_id field derived from uninitialized vxlan_metadata stack contents.

This is not an active structured-object lifetime issue and not fixed stale storage. No object is freed, reused, replaced, or dereferenced after lifetime end. There is no concrete repeated release or refcount operation.

The unintended stack-derived GBP bits can also spuriously set VXLAN_HF_GBP, dont learn, and policy applied metadata. This may alter how a remote VXLAN endpoint interprets the packet. However, the patch does not establish that a remote endpoint treats those metadata bits as an authorization bypass, that protected state is modified, or that an attacker gains elevated privileges.

PrimitiveClass is CLASS_C because the established behavior is invalid metadata construction with a bounded information disclosure and possible incorrect packet metadata handling. It is not CLASS_A because the disclosed value is limited and no strong corruption, boundary bypass, protected-resource access, or privilege-escalation consequence is established.

ImportantCandidate is NO. The patch establishes a real network-visible information leak, but only a small stack-derived field is exposed and there is no demonstrated memory corruption or exploitation chain. AutoDowngradeAllowed is YES because the concrete primitive is bounded, does not involve lifetime corruption, and does not leave a strong unresolved kernel corruption mechanism.

Likely operational severity is ordinary Moderate-like conditional information disclosure. Manual review remains recommended because exposure depends on actual use of flow-based VXLAN with GBP and because remote peers may receive the leaked field.

5. One-sentence report phrase

A VXLAN device using collect metadata and GBP can transmit 16 bits of uninitialized kernel stack data in the GBP policy_id field when external tunnel metadata lacks the VXLAN option flag.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Review whether affected systems use VXLAN_F_COLLECT_METADATA together with VXLAN_F_GBP and whether untrusted peers can send traffic through the affected VXLAN egress path.

7. Call-site confidence

Call-site confidence: medium

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

Patch: vxlan: initialize _md in vxlan_xmit_one() [ Upstream
Commit: 0d13b5a413bffc8718b3821b78537ccc6596c233
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=0d13b5a413bffc8718b3821b78537ccc6596c233

Changed files:
  drivers/net/vxlan/vxlan_core.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=0d13b5a413bffc8718b3821b78537ccc6596c233

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

The original Linux kernel CVE announcement for CVE-2026-97965 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97965

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:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N
  The Best / paranoid CVSS vector: AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:N
  The Best / paranoid CVSS score: 4.8

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: 3
  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: 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).