[CVE-2026-72250][MODERATE REGULAR] netfilter: nf_conntrack_reasm: guard mac_header adjustment after IPv6 defrag

[CVE-2026-72250][MODERATE REGULAR] netfilter: nf_conntrack_reasm: guard mac_header adjustment after IPv6 defrag

[AL-KERNEL]

CVE: CVE-2026-72250
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: netfilter: nf_conntrack_reasm: guard mac_header adjustment after IPv6 defrag
Commit: 6e8cd710ca35c576f5f2e5a396047c9ac61f75e5
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=6e8cd710ca35c576f5f2e5a396047c9ac61f75e5
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72250
Analysis date: Sat, 15 Aug 2026 19:46:00 -0400
ActionableScore: 4
ActionableScore lower bound: 3
Actionable bucket: Borderline manual review recommended / below Important
Manual review required: YES

Summary:
An IPv6 netfilter defragmentation bug can incorrectly convert an unset skb MAC header sentinel into a valid-looking offset on the LOCAL_OUT path, causing later packet-processing code to interpret headroom as a MAC header and potentially leading to local DoS.

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

The original Linux kernel CVE announcement for CVE-2026-72250 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72250

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-2026-72250	MODERATE	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:L/I:L/A:H';CWE-190;CWE-682;CWE-125;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:L';BEST CVSS score: '5.8';DESCR 'An IPv6 netfilter defragmentation metadata bug can corrupt the logical state of skb mac_header on the NF_INET_LOCAL_OUT path. The skb has no link layer header yet, so mac_header may still contain the unset sentinel value, and adding the fragment header size can wrap it into a small offset that later appears valid. Later code may treat bytes in skb headroom as a real MAC header, which can cause incorrect packet metadata interpretation and possible availability impact in dependent paths. For the CVSS the PR:L is used because reliable triggering likely requires local code or a reduced capability process able to generate suitable IPv6 fragmented output on the victim, but not necessarily full administrator privileges in all deployments. The issue is not directly reachable from remote inbound packets because the described problematic path is LOCAL_OUT rather than normal receive processing. Impact is at least low confidentiality risk from incorrect header reads and availability risk, while the paranoid interpretation keeps limited integrity impact for manual review without converting the bug into a proven memory corruption primitive.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline manual review recommended / below Important (with actual score 4)	YES	REMOTE SIMPLEFIX NETFILTER NETWORK SKB LINUS TEST IPV6 PACKET  DECREASED_TO_MODERATEREG_BASED_ON_ACTIONABLESCORELESSTHAN5	NO	NO	checked

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

ActionableScore=4
ActionableScoreLower=3

## 1. ActionableScore

* Conservative score: 3
* Paranoid score: 4
* Final recommended bucket: **Borderline manual review recommended / below Important**::

## 2. Signal breakdown

Conservative signals:

* Local unprivileged trigger: +1
  The problematic path is NF_INET_LOCAL_OUT. A local process may be able to generate IPv6 output that reaches conntrack defragmentation, depending on configuration.
* Memory corruption, weak/indirect corruption candidate: +1
  The bug corrupts skb metadata state by converting the unset mac_header sentinel into a small apparently valid offset. This is a bogus header pointer / invalid metadata interpretation issue, not a proven UAF or write primitive.
* Memory layout invariant restoration: +1
  The patch restores consistency between skb header-offset metadata after IPv6 fragment-header removal.
* Common networking core path or packet-processing path: +1
  This is in netfilter IPv6 conntrack defragmentation, a packet-processing path that deserves manual scrutiny.
* Hard or specific triggering conditions: -1
  The issue depends on the LOCAL_OUT defrag path, unset MAC header state, IPv6 fragmentation, and later code using skb_mac_header().

Paranoid additional signal:

* Availability impact realistic: +1
  Later consumers may misinterpret headroom bytes as a MAC header, plausibly causing packet-processing faults, warnings, or crashes in dependent paths, though no concrete crash primitive is shown.

Not awarded:

* No remote reachable signal. The described vulnerable path is LOCAL_OUT, not inbound remote packet processing.
* No strong memory corruption. The patch does not show UAF, double free, OOB write, arbitrary write, type confusion, or refcount takeover.
* No LPE signal. There is no demonstrated reclaim, overwrite, callback control, or privilege boundary bypass.
* No confidentiality or integrity impact in the conservative score. There is no clear read-back or write primitive.

## 3. Reachability analysis

The most realistic trigger is local code that causes IPv6 fragmented output to pass through netfilter conntrack defragmentation on the LOCAL_OUT path. This may be reachable by an unprivileged local user in some configurations, but reliable triggering depends on IPv6, conntrack defrag being active, and packet construction conditions that leave skb->mac_header unset.

Namespaces may lower practical privilege requirements if an attacker can generate suitable traffic inside a namespace whose packets still traverse the affected host netfilter path. However, this is not equivalent to a normal remote network attack, because the commit specifically identifies NF_INET_LOCAL_OUT.

The path is security-relevant because netfilter and skb metadata are common kernel networking infrastructure, but the affected condition is narrower than a generic IPv6 receive-path parser bug.

Call-site confidence: medium. The changed code and immediate skb metadata effects are visible, but downstream consumers of the corrupted mac_header are not shown in the provided patch.

## 4. Severity interpretation

This behaves more like a borderline actionable Moderate than an Important-class vulnerability. The theoretical concern is that corrupted skb header metadata can make later code treat headroom bytes as a valid MAC header. The realistic evidence supports metadata corruption and possible DoS, but not a strong memory corruption primitive, privilege escalation, or remote exploitation.

The paranoid score is kept at 4 because this is a netfilter skb metadata bug in a packet path and should not be auto-closed without review. It should not be raised to Important based only on the current patch because no UAF, OOB write, arbitrary write, or ownership-boundary bypass is demonstrated.

## 5. One-sentence report phrase

An IPv6 netfilter defragmentation bug can incorrectly convert an unset skb MAC header sentinel into a valid-looking offset on the LOCAL_OUT path, causing later packet-processing code to interpret headroom as a MAC header and potentially leading to local DoS.

## 6. Manual review recommendation

MANUAL CHECK RECOMMENDED

This should receive manual review because it is a networking skb metadata corruption issue in netfilter, but the available evidence supports only weak/indirect corruption and possible DoS rather than a proven LPE or remote primitive.

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

Patch: netfilter: nf_conntrack_reasm: guard mac_header adjustment after IPv6 defrag
Commit: 6e8cd710ca35c576f5f2e5a396047c9ac61f75e5
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=6e8cd710ca35c576f5f2e5a396047c9ac61f75e5

Commit description:

nf_ct_frag6_reasm() slides the packet head forward to drop the IPv6
fragment header and then unconditionally advances skb->mac_header:

	skb->mac_header += sizeof(struct frag_hdr);

On the NF_INET_LOCAL_OUT defrag path the skb has no link-layer header
yet, so skb->mac_header is still the "not set" sentinel (u16)~0U. Adding
sizeof(struct frag_hdr) wraps it to a small value (0xffff + 8 == 7),
after which skb_mac_header_was_set() wrongly reports a MAC header is
present and skb_mac_header() points into the headroom.

The reassembler has done this unconditional add since it was introduced;
it was harmless while mac_header was a bare pointer, but wrong once
mac_header became a u16 offset whose unset state is the ~0U sentinel
tested by skb_mac_header_was_set(). The sibling net/ipv6/reassembly.c
does the same relocation and does guard the adjustment; mirror the
guard here.

Fixes: 9fb9cbb ("[NETFILTER]: Add nf_conntrack subsystem.")
Cc: stable@vger.kernel.org
Reported-by: Weiming Shi <bestswngs@gmail.com>
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  net/ipv6/reassembly.c
  net/ipv6/netfilter/nf_conntrack_reasm.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=6e8cd710ca35c576f5f2e5a396047c9ac61f75e5

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

The original Linux kernel CVE announcement for CVE-2026-72250 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72250

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:N/A:L
  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: 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: 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).