[CVE-2026-98019][MODERATE REGULAR] bpf: mark a NULL call argument precise [ Upstream

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

CVE: CVE-2026-98019
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: bpf: mark a NULL call argument precise [ Upstream
Commit: 69a9ad004ccf5d45e534c7d503f47f643abe64b7
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=69a9ad004ccf5d45e534c7d503f47f643abe64b7
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98019
Analysis date: Fri, 25 Sep 2026 11:04:22 -0400
ActionableScore: 3
ActionableScore lower bound: 2
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-2026-98019 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98019

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-2026-98019	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-98019 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-693;*CWE-20;CWE-284;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '5.8';DESCR 'A BPF verifier precision tracking flaw can cause verifier checkpoints to prune states against an arbitrary scalar value after a nullable helper argument check or a bpf_get_local_storage flags check. This may make verifier validation unsound and could allow a BPF program that should have been rejected to pass verification. For the CVSS the PR:L because an attacker needs local code execution and access to the BPF program loading interface. The issue is not network reachable by itself. The demonstrated patch context supports at least a local kernel denial of service risk. Limited confidentiality or integrity impact is a conservative worst case because verifier unsoundness can potentially weaken BPF memory safety, but no concrete read write or use after free primitive is demonstrated by this patch. Manual review is required to determine whether a practical verifier bypass can reach a memory access primitive on affected kernel configurations.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 3)	YES	INIT BPF SIMPLEFIX LINUS 	NO	NO	checked

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

ActionableScore=3
ActionableScoreLower=2
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES

1. ActionableScore

* Raw conservative score: 2
* Conservative score: 2
* Raw paranoid score: 3
* Paranoid score: 3
* Final recommended bucket: **Borderline / manual review recommended**::

The conservative score does not assume that an unsafe BPF program is actually accepted or that it reaches a memory-access primitive. The paranoid score adds one point for the technically plausible verifier-soundness consequence of accepting a program whose scalar constraints were incorrectly pruned.

2. Signal breakdown

* +1 Weak indirect kernel-safety primitive. The verifier can create a checkpoint after a nullable argument or zero-flags validation without marking the underlying scalar register precise. A later state-pruning decision can therefore treat an arbitrary scalar as equivalent to the validated value.
* +1 Verifier protection-mechanism failure. The affected checks are part of the BPF verifier safety model. The patch establishes that state equivalence can be unsound for helper argument validation.
* +1 Paranoid only. If a crafted BPF program can use the imprecise state to bypass a later verifier constraint, it may be accepted despite requiring rejection. This is a defensible verifier-soundness concern, but the supplied patch does not demonstrate a concrete accepted invalid program, crash, read, write, or privilege-escalation chain.

Important non-awarded signals:

* No real UAF, stale object, refcount misuse, double release, or lifetime mutation is established.
* Controlled reclaim, object replacement, arbitrary kernel write, directed write, callback influence, and type confusion are not demonstrated.
* A concrete kernel crash is not demonstrated by the commit or selftest changes. A DoS consequence is plausible only if the verifier defect can be converted into an unsafe accepted program.
* Confidentiality and integrity impact are not awarded. The patch does not show a user-visible disclosure, arbitrary read, arbitrary write, protected-object modification, or credential manipulation primitive.
* Remote reachability is not established. The affected path is reached through local BPF program verification.
* Local unprivileged reachability is not awarded. The complete BPF program loading and relevant helper-use privilege requirements are not established for the evaluated product.
* No concrete authorization ALLOW-versus-DENY bypass is established. This is verifier state-validation unsoundness, not a demonstrated capability or credential check bypass.

3. Reachability analysis

The triggering actor is a local process attempting to load a crafted BPF program that reaches either a nullable helper-argument check or the bpf_get_local_storage zero-flags check and then relies on verifier state pruning.

The exact privilege boundary is unresolved from the supplied patch. On many kernel configurations, BPF loading or the relevant program type and helper availability require CAP_BPF, CAP_SYS_ADMIN, CAP_NET_ADMIN, LSM authorization, or a root-controlled unprivileged-BPF policy. The patch does not establish that an ordinary unprivileged user can load and execute the complete malicious program.

HostAdminRequired is therefore UNKNOWN, not YES. The supplied evidence does not prove that host-admin-equivalent privilege is required for every affected configuration, but it also does not prove a complete below-host-admin trigger path.

NamespaceScopedPrivilege is UNKNOWN. The patch does not involve a network device, XDP attachment, network namespace, or another namespace-scoped setup path. Even if some BPF-related capabilities can be delegated in a particular environment, the complete load, verifier, helper, and execution sequence is not shown to be namespace-reachable.

No special hardware, firmware, boot option, or race is indicated. The issue is deterministic verifier logic rather than a timing-dependent race. A prior privileged policy configuration may materially affect exploitability through BPF sysctls, capabilities, LSM policy, and allowed program types.

4. Primitive and severity interpretation

The exact primitive is an imprecise scalar-state tracking defect in the BPF verifier. A register checked as NULL for a nullable helper argument, or checked as zero for bpf_get_local_storage flags, is not marked precise before verifier checkpoint creation and state pruning. This can cause the verifier to reuse a state that was valid for one scalar value when evaluating another scalar value.

This is not an active structured-object UAF, fixed stale storage issue, cleanup dereference, double free, or refcount mutation. No object lifetime ends prematurely, and no second release or post-free operation exists.

The patch establishes a verifier soundness defect, but does not establish the next required step: a concrete BPF instruction sequence that is accepted because of the bad pruning and then performs an invalid memory access or another security-sensitive operation. The supplied selftest updates only show changed precision propagation expectations and do not demonstrate arbitrary kernel read, write, disclosure, crash, or privilege escalation.

PrimitiveClass is CLASS_C because the demonstrated behavior is invalid verifier-state handling without an independently useful kernel corruption primitive. ImportantCandidate is NO because meaningful exploitation leverage beyond potential unsafe program acceptance remains unresolved and is not directly demonstrated. AutoDowngradeAllowed is YES because there is no concrete UAF, OOB access, arbitrary write, authorization bypass, controlled corruption, or protected-resource access primitive that would be lost by an automatic lower-priority classification.

Operationally, this should be reviewed as a BPF verifier correctness and possible safety-boundary issue, especially on products that allow non-admin BPF loading. It is not supported by the patch as a strong Important-class vulnerability.

5. One-sentence report phrase

A BPF verifier precision-tracking flaw can incorrectly prune scalar states after nullable helper argument or bpf_get_local_storage flag validation, potentially allowing a BPF program that should be rejected to pass verification, although no concrete memory-access or privilege-escalation primitive is demonstrated.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Review is recommended to determine whether a crafted program can convert the imprecise checkpoint pruning into a verifier acceptance bypass and a concrete kernel memory-safety primitive under the BPF privilege model of the evaluated product.

7. Call-site confidence

Call-site confidence: medium

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

Patch: bpf: mark a NULL call argument precise [ Upstream
Commit: 69a9ad004ccf5d45e534c7d503f47f643abe64b7
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=69a9ad004ccf5d45e534c7d503f47f643abe64b7

Changed files:
  kernel/bpf/verifier.c
  tools/testing/selftests/bpf/progs/verifier_subprog_precision.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=69a9ad004ccf5d45e534c7d503f47f643abe64b7

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

The original Linux kernel CVE announcement for CVE-2026-98019 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98019

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:N/I:N/A:H
  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: 2
  Paranoid ActionableScore: 3

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).