[CVE-2026-90312][MODERATE 7.0] bpf: Check load-acquire src ptr type before the load [ Upstream

[CVE-2026-90312][MODERATE 7.0] bpf: Check load-acquire src ptr type before the load [ Upstream

[AL-KERNEL]

CVE: CVE-2026-90312
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: bpf: Check load-acquire src ptr type before the load [ Upstream
Commit: 6ee7b00888498cf387dd30729e18a05328b94709
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=6ee7b00888498cf387dd30729e18a05328b94709
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90312
Analysis date: Thu, 17 Sep 2026 23:54:35 -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-2026-90312 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90312

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-2026-90312	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-90312 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-843;*CWE-20;CWE-693;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7.8';DESCR 'An incorrect validation order in check_atomic_load can allow a BPF load acquire instruction with identical source and destination registers to bypass source pointer type restrictions. check_load_mem replaces the source register type before atomic_ptr_type_ok validates it, so prohibited ctx, packet, flow keys, or socket pointers may no longer be recognized. The resulting program can retain a raw access to an underlying kernel object while the verifier assigns the register a trusted socket pointer type. After a NULL check this becomes a kernel pointer type confusion and may permit invalid kernel memory access, information disclosure, memory corruption, a kernel crash, or local privilege escalation. For the CVSS the PR:L is selected because an attacker must execute local code and load a crafted BPF program, while unprivileged BPF access may be available to an ordinary local user on affected configurations. Systems that disable unprivileged BPF normally restrict this path to users with CAP_BPF, CAP_SYS_ADMIN, or equivalent delegated capabilities. The issue is not remotely reachable without a separate mechanism that allows an attacker to submit BPF programs. The base impact is CIA:LLH because the patch establishes type confusion but does not provide a complete arbitrary read or write primitive. The paranoid impact is CIA:HHH because a verifier bypass involving a trusted kernel pointer type provides a plausible path to kernel memory corruption and privilege escalation.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5)  SKIP CVE-2026-90312 UNKNOWN SKIP The Fixes patch not applied yet, so unlikely that actual: url YES  NO NO unknown 	MAYBE	OOB BPF LEAK IMPROVEONLY  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: 3
* Conservative score: 5
* Raw paranoid score: 4
* Paranoid score: 5
* Applied conservative manual-review safety floor: 3 -> 5 because the patch establishes a Class A verifier boundary bypass with strong semantic reinterpretation and kernel pointer type confusion.
* Applied paranoid manual-review safety floor: 4 -> 5. The BPF privileged-control-plane cap was checked but did not modify the result.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

2. Signal breakdown

Conservative positive signals:

* +2 Strong attacker-influenced semantic reinterpretation and corruption primitive. An attacker-selected atomic load is validated while the source has pointer semantics, but the resulting unconverted bytes are subsequently treated as a trusted socket pointer type. This is an explicit verifier type confusion, not merely an invalid value.
* +1 Confidentiality plausible. The commit explicitly distinguishes the defect from a simple kernel-data leak, and the raw access places underlying kernel-object bytes into a BPF register. This supports potential disclosure, although the complete userspace extraction path is not shown.

Paranoid additional signal:

* +1 Semantic reinterpretation affecting active memory access. After the NULL check, the verifier accepts the falsely typed value as a socket pointer, making further pointer-based kernel-object processing technically plausible. This does not assume arbitrary read, arbitrary write, or LPE.

Not awarded:

* No UAF or stale post-lifetime object use is established.
* No controlled reclaim, arbitrary write, callback control, credential modification, or demonstrated privilege escalation is established.
* Integrity impact is not separately awarded because no concrete write sink is visible.
* Reliable kernel crash is plausible but not demonstrated by the supplied evidence, so no DoS point is awarded.
* Local-unprivileged reachability is not awarded because the complete BPF load and execution path below host-admin privilege is unresolved.
* No remote-reachability point is awarded.
* The verifier bypass, semantic reinterpretation, and type confusion are treated as one primary +2 primitive and are not double-counted.
* No repeated release or refcount operation exists. FLOOR 5 was not applied under the UAF rules.

3. Reachability analysis

Triggering requires submitting a crafted BPF program containing a load-acquire instruction whose source and destination registers are identical, using a prohibited source pointer type such as context, packet, flow-keys, or socket state.

The complete privilege boundary is unresolved. Systems permitting unprivileged BPF may expose the path to an ordinary local process, while systems with unprivileged BPF disabled normally require CAP_BPF, CAP_SYS_ADMIN, or related privileges. The supplied material does not establish which policy applies to the evaluated product, so HostAdminRequired remains UNKNOWN and no privilege penalty or unprivileged bonus is applied.

NamespaceScopedPrivilege is UNKNOWN because no evidence establishes that all required loading, program-type, attachment, and execution stages can be completed with attacker-obtainable namespace-scoped capabilities. A root-controlled unprivileged-BPF sysctl may be a prerequisite on some systems. No special hardware, race, timing window, or network trigger is required by the verifier defect itself.

4. Primitive and severity interpretation

The exact primitive is a BPF verifier security-boundary bypass producing strong semantic reinterpretation and kernel pointer type confusion. `check_load_mem()` previously replaced the source register state before `atomic_ptr_type_ok()` checked whether the original pointer type was allowed. Consequently, raw bytes read from an underlying kernel object could be represented as `PTR_TO_SOCK_COMMON_OR_NULL` and accepted as a socket pointer after a NULL check.

This is broader active use rather than cleanup or destructor behavior. It is not a UAF, fixed stale-storage case, or repeated lifetime mutation.

PrimitiveClass is CLASS_A because the patch directly establishes attacker-influenced bytes crossing from raw kernel-object representation into a trusted active pointer semantic domain. The demonstrated consequence is verifier acceptance of an invalidly typed kernel pointer and potential kernel-data exposure. Invalid memory access, memory corruption, crash, and privilege escalation are technically plausible consequences, but arbitrary read/write and a complete LPE chain are not established.

ImportantCandidate is YES because verifier-approved kernel pointer type confusion is a strong security-boundary primitive whose downstream exploitation leverage requires review. AutoDowngradeAllowed is NO because automatic closure would discard an established Class A semantic reinterpretation primitive. The likely operational severity is Actionable Moderate at minimum, with Important severity dependent mainly on BPF reachability and the operations permitted on the confused pointer.

5. One-sentence report phrase

A BPF verifier validation-order flaw allows raw kernel-object bytes to be accepted as a trusted socket pointer, creating a local kernel pointer type confusion that requires manual exploitation and privilege-boundary review.

6. Manual review recommendation

MANUAL CHECK REQUIRED

The patch establishes a Class A verifier type-confusion primitive, while the complete BPF privilege boundary and the downstream read, write, crash, or privilege-escalation capabilities remain unresolved.

7. Call-site confidence

Call-site confidence: medium

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

Patch: bpf: Check load-acquire src ptr type before the load [ Upstream
Commit: 6ee7b00888498cf387dd30729e18a05328b94709
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=6ee7b00888498cf387dd30729e18a05328b94709

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=6ee7b00888498cf387dd30729e18a05328b94709

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

The original Linux kernel CVE announcement for CVE-2026-90312 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90312

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:L/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).