[CVE-2026-98103][MODERATE REGULAR] igmp: convert struct ip_sf_list to RCU [ Upstream

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

CVE: CVE-2026-98103
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: igmp: convert struct ip_sf_list to RCU [ Upstream
Commit: f5182dfad54277267a8cb0c004a9cd4cf35aeebb
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=f5182dfad54277267a8cb0c004a9cd4cf35aeebb
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98103
Analysis date: Sat, 26 Sep 2026 14:09:11 -0400
ActionableScore: 1
ActionableScore lower bound: unknown
Actionable bucket: Ordinary Moderate
Manual review required: NO

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-98103 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98103

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: NO

A detailed explanation of the methodology and priority rules is included
at the end of this message.

======================================================================
AL-KERNEL CLASSIFICATION RESULT
======================================================================

CVE-2026-98103	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-98103 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:N/I:N/A:H';CWE-667;CWE-362;*CWE-833;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:N/I:N/A:L';BEST CVSS score: '4.7';DESCR 'An IPv4 IGMP multicast source filter update can trigger recursive locking in ip_check_mc_rcu when the packet receive or route lookup path runs while the IGMPv3 reporting path holds pmc->lock. A matching XFRM policy for a multicast destination can cause the same CPU to attempt to acquire im->lock recursively, resulting in a lockdep warning, CPU stall, or spinlock deadlock. For the CVSS the PR:L paranoid score assumes a preconfigured multicast and XFRM environment where a local unprivileged process can influence multicast membership or source filter activity. The base score uses PR:H because reliably configuring the required XFRM policy normally requires CAP_NET_ADMIN. The issue is not directly reachable by network traffic alone because local multicast state changes and configuration preconditions are required. Impact is denial of service through a stalled CPU or networking path. No confidentiality or integrity impact is supported by the patched code path.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Ordinary Moderate (with actual score 1)	YES	LOCK DEADLOCK UAF HARDWARE IPV6 SYZBOT PACKET  DECREASED_TO_MODERATE_REGULAR_FROM_MODERATE7_BASED_ON_GUESSCVSS_AND_CIANNH	NO	NO	checked

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

ActionableScore=1
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES

1. ActionableScore

* Raw conservative score: 1
* Conservative score: 1
* Raw paranoid score: 1
* Paranoid score: 1
* Final recommended bucket: **Ordinary Moderate**::

The conservative and paranoid scores are identical because the supplied patch establishes a recursive-locking deadlock and possible CPU stall, but does not establish an active use-after-free in the vulnerable post-23d2b94 code state.

2. Signal breakdown

* +1 Reliable kernel DoS. The commit explicitly describes recursive acquisition of im->lock on the same CPU while pmc->lock is held, resulting in a lockdep recursive locking warning or spinlock deadlock. A deadlocked CPU or blocked networking path is a realistic availability effect.
* +0 Real UAF. Not awarded. The earlier commit 23d2b94 added im->lock specifically to prevent the previously existing concurrent-deletion UAF. The code fixed by this patch is therefore protected from that UAF, but introduces a deadlock through the added locking.
* +0 RCU lifetime corruption. Not awarded. The RCU conversion is the solution used to avoid both the former UAF risk and the new lock recursion. The supplied vulnerable code does not establish an RCU reader dereferencing already-freed ip_sf_list storage.
* +0 Remote reachability. Not awarded. Packet receive and route lookup are involved, but the problematic path additionally requires local multicast state and an XFRM policy matching a multicast destination. Network traffic alone is not shown to trigger the complete condition.
* +0 Local unprivileged trigger. Not awarded. An ordinary process can often influence multicast socket membership, but the supplied material does not establish that one below-host-admin attacker can perform the full source-filter, IGMPv3 reporting, and matching XFRM-policy sequence.
* +0 Common networking path. Not awarded. The affected helpers are in packet receive and route lookup paths, but the vulnerable condition requires the comparatively specific conjunction of IGMPv3 reporting, multicast source filters, and matching XFRM policy.
* +0 Strong exploitation leverage. Not awarded. No controlled reclaim, stale-object reuse, arbitrary write, callback dispatch, refcount corruption, disclosure, authorization bypass, or privilege escalation primitive is established.
* +0 Additional Availability impact. Not awarded. The CPU stall or deadlock is the same demonstrated DoS consequence already counted above.
* No race penalty. The reported issue is an ordered recursive locking path, not an externally uncontrollable narrow timing race.
* No host-admin penalty. Host-admin-equivalent privilege is not positively established as necessary for the complete trigger. XFRM policy setup commonly requires CAP_NET_ADMIN, but a pre-existing policy may be sufficient for a lower-privileged process to reach the remaining steps.

3. Reachability analysis

The evaluated attacker is a local process that can influence IPv4 multicast membership or multicast source-filter activity. The problematic path occurs when IGMPv3 report generation holds pmc->lock and route resolution reaches ip_check_mc_rcu while an XFRM policy matching the multicast destination is active.

A matching XFRM policy is a material prerequisite. Creating or modifying host XFRM policy normally involves CAP_NET_ADMIN, but the supplied evidence does not show whether the same attacker must create that policy or whether it can already exist as part of host networking configuration. Therefore HostAdminRequired is UNKNOWN rather than YES or NO.

NamespaceScopedPrivilege is UNKNOWN. The supplied patch does not identify the precise capability checks, target network namespace, or whether the complete XFRM and multicast policy configuration sequence can be performed inside an attacker-controlled namespace. The presence of networking and multicast state alone is insufficient to establish namespace-scoped reachability.

No special hardware, boot parameter, or sysctl requirement is shown. The triggering sequence is configuration-dependent but is not established as a narrow timing race. Once the required lock ordering and XFRM route-resolution path occur, recursive lock acquisition is deterministic.

4. Primitive and severity interpretation

The exact primitive is recursive spinlock acquisition in the IPv4 IGMP multicast path. The vulnerable execution holds pmc->lock during IGMPv3 report generation and then reaches ip_check_mc_rcu through XFRM-assisted route lookup, where it attempts to acquire im->lock again. The demonstrated result is lockdep detection, CPU stall, or deadlock.

This is not an active structured-object UAF in the code state fixed by this patch. The previous locking change was introduced to stop a UAF during concurrent deletion of ip_sf_list entries. The current patch replaces that lock-based protection with RCU lifetime management, using rcu_assign_pointer and kfree_rcu so readers can traverse the source list without acquiring im->lock.

There is no concrete repeated release, put, free, or refcount mutation through a dangling reference. There is no demonstrated controlled replacement, stale field consumption after reuse, arbitrary read or write, information disclosure, integrity violation, privilege escalation, or security-boundary bypass.

PrimitiveClass is CLASS_C because the established issue is invalid lock state causing DoS behavior. ImportantCandidate is NO because meaningful exploitation leverage beyond the deadlock is not unresolved in the supplied evidence. AutoDowngradeAllowed is YES because there is no current concrete lifetime-corruption primitive requiring Important-versus-Moderate preservation.

Likely operational severity is Low to ordinary Moderate. It can affect networking responsiveness or a CPU executing the recursive locking path, but it is not supported as a host-wide crash, memory-corruption, or privilege-escalation vulnerability.

5. One-sentence report phrase

A recursive locking bug in the IPv4 IGMP multicast source-filter path can deadlock a CPU when IGMPv3 reporting and XFRM-assisted multicast route lookup attempt to acquire the same lock, causing local denial of service.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

The patch performs a substantial RCU lifetime conversion in a networking fast path, and the exact practical reachability depends on multicast source-filter activity and XFRM policy configuration. However, the supplied evidence supports a deadlock rather than an active UAF or stronger corruption primitive.

7. Call-site confidence

Call-site confidence: medium

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

Patch: igmp: convert struct ip_sf_list to RCU [ Upstream
Commit: f5182dfad54277267a8cb0c004a9cd4cf35aeebb
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=f5182dfad54277267a8cb0c004a9cd4cf35aeebb

Changed files:
  net/ipv6/mcast.c
  include/linux/igmp.h

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=f5182dfad54277267a8cb0c004a9cd4cf35aeebb

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

The original Linux kernel CVE announcement for CVE-2026-98103 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98103

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:L
  The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:H
  The Best / paranoid CVSS score: 4.7

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: 1

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).