[CVE-2026-98151][MODERATE REGULAR] bpf: Fix REG INVARIANTS VIOLATION on speculative pointer arithmetic [ Upstream

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

CVE: CVE-2026-98151
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: bpf: Fix REG INVARIANTS VIOLATION on speculative pointer arithmetic [ Upstream
Commit: 21a681526c715aebb4f918b8c131583442d1b3d8
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=21a681526c715aebb4f918b8c131583442d1b3d8
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98151
Analysis date: Sat, 26 Sep 2026 16:44:16 -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-2026-98151 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98151

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-2026-98151	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-98151 CHECK WITH IMPACT FROM ORIG NN IMPORTANT	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:L';CWE-682;*CWE-20;CWE-703;BEST CVSS score: '3.3';DESCR 'A BPF verifier invariant violation can occur during speculative pointer arithmetic because sanitize_ptr_alu may snapshot a register after its 32-bit bounds were cleared but before var_off and r32 are synchronized. An unprivileged BPF program can trigger a verifier warning when a bounded scalar is added to a map value pointer and the resulting pointer is copied on the speculative path. For the CVSS the PR:L is used because a local process must execute code and load a BPF program. The issue is not network reachable. The demonstrated impact is limited to a verifier warning and possible kernel log noise. No out of bounds access memory corruption information disclosure integrity impact or privilege escalation primitive is demonstrated by the patch. Systems that disable unprivileged BPF or restrict bpf syscall access are not exposed to the unprivileged trigger.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Ordinary Moderate (with actual score 1)	YES	WARNONLY BPF SPECTRE LINUS  LOWERED_FROM_HIGH_BASED_ON_GUESSCVSS  DECREASED_TO_MODERATEREG_BASED_ON_ACTIONABLESCORELESSTHAN4	YES	NO	checked

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

ActionableScore=1
PrimitiveClass=NONE
RealUAF=NO
HostAdminRequired=NO
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**::

No floor or cap was applied.

2. Signal breakdown

* +1 Local unprivileged trigger. The supplied context explicitly provides an unprivileged BPF program that reaches the affected verifier path when unprivileged BPF loading is enabled.
* +0 Verifier warning. The demonstrated result is `reg_bounds_sanity_check()` reporting an invariant violation. A WARN alone is not a kernel crash, memory-corruption primitive, or availability-impact signal.
* +0 No real UAF, stale reference, refcount misuse, OOB access, arbitrary write, disclosure, authorization bypass, or security-boundary bypass is established.
* +0 No reliable kernel crash or strong DoS is demonstrated.
* +0 No remote reachability is established. The path is reached through a local BPF program load.
* +0 No local unprivileged high-control data-plane bonus. The program controls verifier input, but the patch does not establish control of kernel object lifetime, memory layout, post-free state, or an exploitation-relevant kernel data structure.
* +0 No C or I impact. The inconsistent verifier snapshot reaches a sanity check, but there is no demonstrated unsafe program acceptance, out-of-bounds access, information disclosure, or integrity effect.
* +0 No paranoid escalation. The patch explicitly describes the speculative masking sequence as harmless and does not show that the temporary inconsistent state bypasses verifier safety checks.

3. Reachability analysis

The triggering actor is an ordinary local process that can load an unprivileged BPF program. The supplied reproducer uses pointer arithmetic on a `PTR_TO_MAP_VALUE` register with a bounded scalar and then copies the resulting pointer on the speculative verifier path.

Host-admin-equivalent privilege is not required by the supplied unprivileged trigger model. Therefore `HostAdminRequired=NO`.

Actual exposure depends on product configuration. Many distributions disable unprivileged BPF through `kernel.unprivileged_bpf_disabled`, LSM policy, seccomp filtering, container policy, or other restrictions on the `bpf()` syscall. Such restrictions prevent the described unprivileged trigger.

No namespace-specific capability chain is provided. The issue is not dependent on a network namespace, BPF attachment, XDP setup, netdevice management, or a privileged BPF control-plane operation. Therefore `NamespaceScopedPrivilege=UNKNOWN`.

There is no hardware dependency, race condition, device prerequisite, prior host-admin setup requirement, or network trigger in the supplied evidence.

4. Primitive and severity interpretation

The patch fixes a temporary internal BPF verifier-state inconsistency. `__mark_reg32_unbounded()` clears the 32-bit register bounds before `sanitize_ptr_alu()` may snapshot the destination register for speculative verification. At that point, `r32` is unbounded while `var_off` still reflects the earlier constant offset. The speculative verifier state therefore violates the invariant checked by `reg_bounds_sanity_check()`.

This is not an active-object lifetime bug, UAF, refcount mutation, stale-object dereference, OOB primitive, or semantic reinterpretation of live kernel state. No object is freed, replaced, or reused. No attacker-controlled kernel memory is consumed after a lifetime transition.

The demonstrated consequence is a verifier warning during program verification. The patch does not establish that the verifier accepts an unsafe program, emits unsafe JIT code, permits an out-of-bounds map access, or allows BPF execution to access unintended kernel memory.

The primitive is therefore `NONE`, not Class A, B, or C. A theoretical verifier-unsoundness consequence is insufficient to assign corruption, confidentiality, integrity, LPE, or Important-class signals without evidence that an unsafe program is accepted or executed.

`ImportantCandidate=NO` because no meaningful unresolved memory-corruption, ownership, lifetime, authorization, or boundary-bypass primitive remains. `AutoDowngradeAllowed=YES` because the available evidence supports a warning-only verifier consistency defect rather than an actionable corruption primitive.

Likely operational severity is Low-like. The issue can create warning noise or fail debugging assertions, but does not establish a security-impacting runtime consequence.

5. One-sentence report phrase

A local unprivileged BPF program can trigger a BPF verifier invariant warning during speculative pointer arithmetic because the speculative register snapshot temporarily contains inconsistent 32-bit bounds and offset state, but no unsafe program acceptance or memory-corruption primitive is demonstrated.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

The patch affects the BPF verifier security boundary and an unprivileged trigger is explicitly shown, so a reviewer should confirm that no verifier-bypass path accepts or executes a program using the inconsistent speculative state. Based on the supplied patch alone, the issue remains warning-only and may be automatically deprioritized after that verification.

7. Call-site confidence

Call-site confidence: medium

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

Patch: bpf: Fix REG INVARIANTS VIOLATION on speculative pointer arithmetic [ Upstream
Commit: 21a681526c715aebb4f918b8c131583442d1b3d8
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=21a681526c715aebb4f918b8c131583442d1b3d8

Changed files:
  kernel/bpf/verifier.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=21a681526c715aebb4f918b8c131583442d1b3d8

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

The original Linux kernel CVE announcement for CVE-2026-98151 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98151

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: Not available
  The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:L
  The Best / paranoid CVSS score: 3.3

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).