[CVE-2026-98048][MODERATE 7.0] bpf: don't rewrite bpf_fastcall patterns entered by a jump [ Upstream

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

CVE: CVE-2026-98048
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: bpf: don't rewrite bpf_fastcall patterns entered by a jump [ Upstream
Commit: 24adbc2c3bbe3385ce922587e1f8e837a68b3e25
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=24adbc2c3bbe3385ce922587e1f8e837a68b3e25
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98048
Analysis date: Mon, 28 Sep 2026 19:20:07 -0400
ActionableScore: 5
ActionableScore lower bound: unknown
Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review
Manual review required: YES

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-98048 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98048

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-2026-98048	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-98048 CHECK WITH IMPACT FROM ORIG NN LOW	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H';CWE-670;CWE-693;*CWE-20;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7.8';DESCR 'An unsound BPF fastcall optimization can rewrite a verified spill, call, and fill instruction pattern even when a jump enters the middle of that pattern. Removing the spill and fill instructions can change the runtime register state while later instructions remain validated against the original program state. This verifier to runtime semantic mismatch may allow a crafted BPF program to use a scalar or another unexpected value where the verifier expects a safe register value or pointer. Impact is at least a local denial of service and may include kernel memory corruption, privilege escalation, and loss of confidentiality and integrity. For the CVSS the PR:L value is used because a local process must be able to load a crafted BPF program, typically through CAP_BPF, CAP_SYS_ADMIN, delegated BPF access, or enabled unprivileged BPF. The issue is not network reachable. The paranoid score uses CIA:HHH because a post verification semantic mismatch can plausibly undermine BPF memory safety, although a concrete arbitrary read or write primitive is not demonstrated by the supplied diff.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5) 	MAYBE	BPF SIMPLEFIX LINUS  KPANIC INCREASED_FROM_LOW_BASED_ON_REQUIREMANUALCHECK	-	-	checked

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

ActionableScore=5
PrimitiveClass=CLASS_A
RealUAF=NO
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO

1. ActionableScore

* Raw conservative score: 2.
* Conservative score: 5. The manual-review safety floor for concrete attacker-influenced semantic reinterpretation changes 2 to 5.
* Raw paranoid score: 3.
* Paranoid score: 5. The same safety floor changes 3 to 5. The BPF conservative cap of 5 and paranoid cap of 6 do not reduce the result.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

2. Signal breakdown

Conservative signals:

* +2 Strong attacker-influenced semantic reinterpretation. A crafted control-flow edge can enter a spill, call, fill sequence after verification, while the optimizer later removes instructions under the false assumption that the sequence executes as one unit. Runtime register semantics can therefore differ from the state validated by the verifier.
* No negative privilege score. BPF loading is privilege-gated in many deployments, but the supplied evidence does not establish whether the complete trigger requires host-admin-equivalent privilege, a delegated capability, or enabled unprivileged BPF.
* Raw conservative score: 2.
* Manual-review safety floor: 2 to 5 because a concrete verifier-to-runtime semantic mismatch affects active execution and may invalidate the BPF security proof.

Paranoid signals:

* +2 The same semantic reinterpretation primitive.
* +1 The reinterpretation plausibly affects active execution and memory-access semantics. Removed fills can leave runtime register values inconsistent with the values assumed for subsequent verified instructions.
* Raw paranoid score: 3.
* Manual-review safety floor: 3 to 5.

Not awarded:

* No UAF, stale-object lifetime violation, or repeated release is established.
* No controlled reclaim, arbitrary kernel read/write, callback control, or credential modification is demonstrated.
* No crash point is shown, so reliable kernel crash is not separately awarded.
* No confidentiality or integrity point is awarded solely from the theoretical consequences of a verifier mismatch.
* Local unprivileged reachability is not established.
* Remote reachability is not established.
* No race or timing penalty applies because the trigger is a control-flow construction, not a race.
* No extra BPF complexity or subsystem bonus is awarded.

Repeated-release checks:

* First lifetime-ending release: none.
* Surviving exact reference: none.
* Second lifetime/refcount mutation: none.
* Concrete repeated-release bonus: not applicable.
* Paranoid repeated-refcount bonus: not applicable.
* Real-UAF FLOOR 5: not applicable. The applied floor is the semantic-reinterpretation manual-review floor.

3. Reachability analysis

The immediate attacker-controlled input is a BPF program containing a fastcall spill, call, fill pattern and a jump into its interior. Loading the program is the relevant control-plane operation. No attachment, network traffic, special hardware, race, or asynchronous timing requirement is established by the supplied patch.

The complete privilege boundary is unresolved. Modern deployments commonly restrict BPF loading through CAP_BPF, CAP_SYS_ADMIN, related capabilities, or a root-controlled unprivileged-BPF sysctl. The supplied evidence does not establish that host-admin-equivalent privilege is always required, nor that an ordinary user can complete the trigger. It also does not establish that all required operations work with capabilities obtained in an attacker-created user or network namespace.

Therefore:

* HostAdminRequired is UNKNOWN.
* NamespaceScopedPrivilege is UNKNOWN.
* No local-unprivileged bonus or host-admin penalty is applied.
* Prior privileged setup may exist through BPF policy or sysctl configuration, but it is not positively established for the evaluated environment.
* The trigger appears deterministic once a suitable program can be loaded. It is not a race.

4. Primitive and severity interpretation

The exact primitive is an unsound post-verification transformation. The verifier analyzes one control-flow and register-state model, but the optimizer can remove spill and fill instructions even though another branch enters the middle of that sequence. The transformed program may consequently execute with register contents different from those proven safe by the verifier.

This is active executable-state semantic reinterpretation, not stale storage, cleanup behavior, or a lifetime bug. No concrete repeated release or refcount mutation exists.

PrimitiveClass is CLASS_A because the mismatch directly undermines the verifier-to-runtime security invariant and changes attacker-influenced active execution semantics. A concrete arbitrary read, arbitrary write, kernel crash, or privilege escalation is not demonstrated, so those consequences remain unresolved rather than established.

ImportantCandidate is YES because an optimizer-induced verifier bypass can preserve unsafe runtime state after the program has passed verification. AutoDowngradeAllowed is NO because automatically treating this as an ordinary logic bug could discard a concrete BPF security-boundary primitive. Operational severity is Actionable Moderate at minimum, with Important classification dependent on confirming the reachable program types, privilege model, and whether the mismatch can produce a useful pointer or scalar state divergence.

5. One-sentence report phrase

A crafted BPF control-flow edge can make the fastcall optimizer rewrite a spill and fill sequence that is not executed as one unit, creating a verifier-to-runtime register-state mismatch that requires manual exploitation and privilege-boundary review.

6. Manual review recommendation

MANUAL CHECK REQUIRED

Confirm which BPF program types and helpers reach this optimization, whether loading is available below host-admin privilege, and whether the resulting register-state divergence yields an invalid pointer, controlled memory access, or only incorrect program behavior.

7. Call-site confidence

Call-site confidence: medium

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

Patch: bpf: don't rewrite bpf_fastcall patterns entered by a jump [ Upstream
Commit: 24adbc2c3bbe3385ce922587e1f8e837a68b3e25
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=24adbc2c3bbe3385ce922587e1f8e837a68b3e25

Changed files:
  include/linux/bpf_verifier.h
  kernel/bpf/cfg.c
  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=24adbc2c3bbe3385ce922587e1f8e837a68b3e25

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

The original Linux kernel CVE announcement for CVE-2026-98048 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98048

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:L/A:H
  The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
  The Best / paranoid CVSS score: 7.8

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: unknown
  Paranoid ActionableScore: 5

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).