[CVE-2026-98789][MODERATE REGULAR] bpf: Fix objects stuck in free_by_rcu_ttrace [ Upstream

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

CVE: CVE-2026-98789
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: bpf: Fix objects stuck in free_by_rcu_ttrace [ Upstream
Commit: 33d0a5d2fee06a51f2f056dd901296515912e954
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=33d0a5d2fee06a51f2f056dd901296515912e954
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98789
Analysis date: Mon, 19 Oct 2026 01:15:52 -0400
ActionableScore: 2
ActionableScore lower bound: unknown
Actionable bucket: Ordinary Moderate
Manual review required: NO

Summary:
A race in the BPF memory allocator can leave an unbounded number of objects retained in free_by_rcu_ttrace, potentially causing progressive memory exhaustion, but the supplied evidence does not establish UAF, memory corruption, privilege escalation, or reliable host-wide failure.

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

The original Linux kernel CVE announcement for CVE-2026-98789 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98789

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-2026-98789	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-98789 CHECK WITH IMPACT FROM ORIG NN MODERATE | AEGIS IMPACT MODERATE AEGIS_AS=0 LEGACY=MODERATE RESULT=SAME V39_AEGIS_ONE_STEP_UP=LOW->MODERATE (final reconciliation agrees; DECISION: both 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-362;*CWE-401;*CWE-772;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:L';BEST CVSS score: '4.7';DESCR 'A race in the BPF memory allocator can leave objects indefinitely retained in free_by_rcu_ttrace when they are added while an RCU Tasks Trace grace period is already in progress. The retained objects are released only if another bulk allocation or free operation occurs for the same cache, which may never happen, and the number of retained objects is not bounded. Sustained local BPF activity may therefore cause progressive memory consumption and eventually denial of service through resource exhaustion. For the CVSS the PR:L value is used because triggering requires a local process with access to the required BPF operations, although systems that disable unprivileged BPF may additionally require CAP_BPF or equivalent permissions. AC:H reflects the need to reach the timing window while the grace period is active. The issue is not network reachable. The demonstrated primitive is persistent resource retention, with no supported evidence of information disclosure, data modification, use after free, or privilege escalation. Automated impact decision path. The original SimpleNet classifier selected MODERATE. The primary LLM estimated paranoid CVSS 4.7 and the production ActionableScore returned 2 with recommendation Ordinary Moderate. The production rules then evaluated CVSS thresholds and ActionableScore thresholds. Recorded rule transitions were CVSS and initial triage stage MODERATE to LOW because the CVSS score, vector or bug class threshold rule was applied. final reconciliation stage LOW to MODERATE because KP specialist did not establish a qualifying panic and the cross model evidence reconciliation rule was applied. The legacy KP specialist has no review artifact for this run. The legacy FD or DH downgrade specialist has no review artifact for this run. Exact triggered rule markers were DECREASED_TO_LOW_FROM_MODERATE_BASED_ON_GUESSCVSS_AND_CIANNH and KPANIC_UNSUPPORTED_REMOVED and V39_AEGIS_ONE_STEP_UP and MANUAL_REVIEW_REQUIRED. The independent weaker Aegis shadow evaluated CVSS 4.4 and ActionableScore 0 with primitive class CLASS_C and manual review RECOMMENDED. Its independently reconciled impact was MODERATE. Cross model aggregation changed the production result under an enabled evidence gated merge rule. The final reported impact is MODERATE.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Ordinary Moderate (with actual score 2)	MAYBE	INIT BPF UAF LINUS DECREASED_TO_LOW_FROM_MODERATE_BASED_ON_GUESSCVSS_AND_CIANNH KPANIC_UNSUPPORTED_REMOVED V39_AEGIS_ONE_STEP_UP MANUAL_REVIEW_REQUIRED	-	-	checked

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

ActionableScore=2
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES
RequiresHostUID0=UNKNOWN
AuthorizationDecisionBypass=NO
ProtectedResourceImpact=NONE
B4PendingReview=NO
B4PendingConfirmed=NO

1. ActionableScore

* Raw conservative score: 2
* Conservative score: 2
* Raw paranoid score: 2
* Paranoid score: 2
* No floor or cap changed either score.
* Final recommended bucket: **Ordinary Moderate**::

2. Signal breakdown

* +1 Concrete persistent kernel resource retention. Objects added to `free_by_rcu_ttrace` during an active Tasks Trace grace period can remain there indefinitely because no subsequent grace period is scheduled.
* +1 Meaningful accumulation potential. The commit explicitly states that the retained object count is unbounded and that the later bulk operation needed to recover the objects may never occur.
* +0 Reliable host-wide DoS. Progressive memory consumption is plausible, but actual host-wide exhaustion or panic is not demonstrated.
* +0 Race penalty. The required overlap with an active grace period is established, but it is not shown to be narrow, one-shot, or difficult to repeat.
* +0 Privilege penalty. BPF access is relevant, but the supplied evidence does not establish whether the complete trigger requires host-admin-equivalent privilege, a reduced capability, or an available unprivileged BPF path.
* No UAF, stale post-lifetime dereference, OOB access, arbitrary write, controlled reclaim, authorization bypass, or confidentiality/integrity primitive is established.
* NEW52 B5-R does not apply because there is no heap OOB write.
* NEW53 B3-PENDING does not apply because no shared-backing write exists.
* V33 and V34 do not apply because no attacker-influenced active structured-state corruption or downstream privileged consumer is established.
* No additional paranoid point is awarded because host-wide exhaustion, isolation bypass, or another independent consequence is not demonstrated.

3. Reachability analysis

The vulnerable condition occurs when additional objects are queued while an RCU Tasks Trace grace period is already in progress. BPF map or allocator activity can create this overlap, and repeated activity can leave additional objects retained.

The supplied patch does not identify the complete userspace command sequence or establish ordinary unprivileged reachability. BPF loading and map management may be restricted by CAP_BPF, CAP_SYS_ADMIN, product policy, or the unprivileged BPF sysctl. Therefore, `HostAdminRequired` and `NamespaceScopedPrivilege` remain `UNKNOWN`, with no privilege penalty or unprivileged-trigger bonus.

No network trigger, special hardware dependency, or namespace-complete trigger is shown. The race occurs naturally under concurrent allocator activity, so it is not scored as a proven hard race.

4. Primitive and severity interpretation

The exact primitive is unbounded retention of objects that should pass through a later RCU Tasks Trace grace period and be freed. The objects remain referenced in `free_by_rcu_ttrace`. They are not shown to have been freed and subsequently accessed, so this is not a real UAF or active structured stale-object primitive.

The `synchronize_rcu()` addition is required by the new callback requeue behavior to ensure destruction cannot miss a newly queued callback. It does not independently prove that the original implementation freed `bpf_mem_cache` while an old callback later dereferenced it.

The demonstrated consequence is persistent memory retention. Progressive resource exhaustion is plausible, but a reliable host crash, memcg bypass, cross-namespace accounting bypass, or privilege escalation is not established. This is therefore `CLASS_C`, not Class A or Class B.

`ImportantCandidate=NO` because no corruption or unresolved security-boundary primitive exists. `AutoDowngradeAllowed=YES` because the established issue is a resource-retention DoS candidate without UAF, OOB write, authorization bypass, or meaningful non-DoS exploitation leverage.

5. One-sentence report phrase

A race in the BPF memory allocator can leave an unbounded number of objects retained in free_by_rcu_ttrace, potentially causing progressive memory exhaustion, but the supplied evidence does not establish UAF, memory corruption, privilege escalation, or reliable host-wide failure.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Confirm the exact BPF operation and privilege boundary, and determine whether a practical workload can repeatedly reach the retaining branch quickly enough to cause host-wide memory exhaustion.

7. Call-site confidence

Call-site confidence: medium

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

Patch: bpf: Fix objects stuck in free_by_rcu_ttrace [ Upstream
Commit: 33d0a5d2fee06a51f2f056dd901296515912e954
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=33d0a5d2fee06a51f2f056dd901296515912e954

Changed files:
  kernel/bpf/memalloc.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=33d0a5d2fee06a51f2f056dd901296515912e954

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

The original Linux kernel CVE announcement for CVE-2026-98789 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98789

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: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: 2

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).