[CVE-2026-89991][MODERATE REGULAR] bpf: Fix infinite loop in pcpu_freelist push with one possible CPU

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

CVE: CVE-2026-89991
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: bpf: Fix infinite loop in pcpu_freelist push with one possible CPU
Commit: a24a146ae2cb5d9faeea96d10010fcb6579abe1f
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=a24a146ae2cb5d9faeea96d10010fcb6579abe1f
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89991
Analysis date: Wed, 16 Sep 2026 11:26:36 -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-2026-89991 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89991

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-2026-89991	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-89991 CHECK WITH IMPACT FROM ORIG NN LOW	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-835;CWE-667;**CWE-400;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '4.7';DESCR 'BPF pcpu_freelist push can enter an infinite loop on systems with only one possible CPU when an NMI reenters pcpu_freelist_push while the interrupted context holds the same CPU freelist lock. The current CPU fast path fails because recursive same CPU lock acquisition is detected, and the fallback loop skips the current CPU but has no other possible CPU to try. For the CVSS the PR:L is used for the paranoid score because some deployments allow BPF use through delegated capabilities or less privileged service accounts, while the base case is closer to PR:H when BPF loading and perf attachment require administrator privileges. The issue is not network reachable and requires local BPF execution paths such as perf event or tracepoint programs. Impact is denial of service through CPU lockup or a stuck kernel execution path. No concrete UAF, OOB access, information leak, or write primitive is indicated by the code change.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 3)  SKIP CVE-2026-89991 UNKNOWN SKIP The Fixes patch not applied yet, so unlikely that actual: url YES  NO NO unknown 	MAYBE	LOCK INIT BPF HARDWARE LINUS  INCREASED_FROM_LOW_BASED_ON_REQUIREMANUALCHECK DECREASED_TO_MODERATEREG_BASED_ON_ACTIONABLESCORELESSTHAN4	-	-	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**::

No floor or cap changed the score.

2. Signal breakdown

Conservative positive signals:

* +1 Reliable kernel DoS. The supplied stack and commit describe an infinite loop in `__pcpu_freelist_push()` that can leave the kernel stuck when the only possible CPU freelist lock is held by the interrupted context.
* +1 Concrete infinite-loop / non-progress primitive. The fallback loop skips the current CPU and has no other CPU candidate when only one CPU is possible, so it can make no lock acquisition attempt and cannot progress.

Conservative negative signals:

* 0 No host-admin penalty applied. BPF loading and perf or tracepoint attachment are privilege-gated in many systems, but the supplied evidence does not fully prove host-admin-equivalent privilege for the complete trigger in all evaluated deployments.
* 0 No race/timing penalty applied in the arithmetic. NMI reentry is a special interleaving, but the provided stack shows a concrete reachable path and repeated BPF/perf activity may make the condition reproducible.
* 0 No rare-subsystem penalty applied. The one-possible-CPU condition is specific, but it can occur on UP kernels or SMP kernels booted with `nr_cpus=1` or `possible_cpus=1`; this is a real supported configuration, not merely hypothetical.

Paranoid additional signal:

* +1 BPF plus NMI execution-context review concern. The issue is still DoS-only, but it occurs in kernel BPF freelist handling from NMI and tracing/perf paths, so manual triage is justified to avoid auto-closing an availability bug in sensitive kernel execution context.

Important non-awarded signals:

* UAF not awarded. No stale object use or post-lifetime access is established.
* Concrete repeated release/refcount not awarded. The patch fixes progress by adding an extra freelist head, not a double put, kref, free, or refcount mutation through a dangling reference.
* Controlled reclaim not demonstrated.
* Arbitrary or directed kernel write not demonstrated.
* Confidentiality and integrity impact not awarded.
* Remote reachability not established.
* Local unprivileged trigger not awarded because complete below-host-admin BPF load, attach, and trigger rights are not established.
* Structured stale-object leverage not awarded because there is no UAF or stale structured object consumption.
* Extra Availability not awarded because the same CPU lockup is already counted as reliable DoS.
* ImportantCandidate not set because the established primitive is infinite loop / DoS-only, not memory corruption, authorization bypass, ownership violation, or unresolved lifetime corruption.

3. Reachability analysis

The trigger requires local execution paths involving BPF programs and perf or tracepoint activity. The observed path includes a perf-event BPF program in NMI context and a raw tracepoint BPF program in the interrupted syscall context.

For BPF/XDP-style privilege analysis:

* Load privilege: unresolved. Loading the required BPF programs is commonly gated by `CAP_BPF`, `CAP_SYS_ADMIN`, `CAP_PERFMON`, sysctls, or product policy.
* Attach privilege: unresolved. Perf-event and raw tracepoint attachment are commonly privileged, but the exact evaluated product policy is not supplied.
* Netdevice privilege: not central to this patch.
* Namespace privilege: not established. The supplied evidence does not show that the complete load, attach, NMI trigger, and freelist operation sequence is achievable using only attacker-created namespaces.
* Root-controlled sysctls: possible in real deployments, but not enough evidence is supplied to score a host-admin-only requirement.

`HostAdminRequired=UNKNOWN` because the complete privilege boundary is not resolved. It should not be treated as ordinary unprivileged, but also should not be forced to host-admin-equivalent without product-specific evidence.

The affected configuration requires only one possible CPU, via `CONFIG_SMP=n` or an SMP kernel limited with `nr_cpus=1` or `possible_cpus=1`. This is a meaningful prerequisite. The triggering interleaving requires NMI reentry while the interrupted context holds the same CPU freelist lock, but the stack trace demonstrates the condition and the involved activity can plausibly be repeated.

There is no direct network trigger.

4. Primitive and severity interpretation

The exact primitive is a kernel infinite loop in `__pcpu_freelist_push()` caused by lack of any fallback lock target when the current CPU is skipped and no other possible CPU exists.

This is not active structured UAF and not fixed stale storage. It is a Class C DoS / invalid-progress bug.

There is no concrete repeated release/refcount operation:

* First release can end lifetime: no such first release is shown for a dangling object.
* Exact same reference survives: not applicable.
* Second lifetime/refcount mutation: not present.
* Dedicated +1 repeated release/refcount: not awarded.
* Paranoid repeated-refcount leverage: not awarded.
* FLOOR 5: not applied.

Demonstrated consequence is CPU lockup or stuck kernel execution path. Speculative consequences beyond DoS are not supported by the patch. There is no evidence of memory corruption, information disclosure, privilege escalation, or authorization bypass.

`ImportantCandidate=NO` because the bug is not a Class A primitive and does not establish Class B corruption or unresolved lifetime leverage. `AutoDowngradeAllowed=YES` because the established issue is constrained to DoS-only behavior with no unresolved corruption or security-boundary primitive. Operationally this is an ordinary to borderline Moderate availability issue, with manual review recommended due to BPF plus NMI context and the supplied prior triage hint.

5. One-sentence report phrase

BPF pcpu_freelist push can loop forever on one-possible-CPU systems when NMI reentry occurs while the interrupted context holds the same CPU freelist lock, causing a local kernel DoS without evidence of UAF, memory corruption, information disclosure, or privilege escalation.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Reason: the issue is DoS-only and auto-downgrade is allowed, but BPF plus NMI reentry and product-specific BPF privilege policy should be checked before final closure.

7. Call-site confidence

Call-site confidence: high

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

Patch: bpf: Fix infinite loop in pcpu_freelist push with one possible CPU
Commit: a24a146ae2cb5d9faeea96d10010fcb6579abe1f
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=a24a146ae2cb5d9faeea96d10010fcb6579abe1f

Commit description:

__pcpu_freelist_push() can loop forever when only one CPU is possible
and an NMI re-enters pcpu_freelist_push() while the interrupted context
holds that CPU's freelist lock.

After the current-CPU fast path fails, the fallback loop walks
cpu_possible_mask while skipping the current CPU. With CONFIG_SMP=n, or
when an SMP kernel is limited to one possible CPU with nr_cpus=1 or
possible_cpus=1, there are no other possible CPUs to examine. The loop
therefore makes no lock acquisition attempt and can never make progress.

The following stack was observed on a UP system:

  NMI context:
    pcpu_freelist_push
    free_htab_elem
    htab_map_delete_elem
    [perf-event BPF program]
    __perf_event_overflow
    perf_event_nmi_handler
    exc_nmi

  Interrupted context:
    __pcpu_freelist_push
    pcpu_freelist_push
    free_htab_elem
    htab_map_delete_elem
    [raw_tp/sys_enter BPF program]
    __bpf_trace_sys_enter
    do_syscall_64

raw_res_spin_lock() detects the same-CPU recursive acquisition and
returns -EDEADLK, but the subsequent fallback loop has no candidate head
on a system with one possible CPU.

Restore the extra fallback head that existed before the rqspinlock
conversion. Keep the current-CPU fast path, then try the other possible
CPUs and finally the extra head. The additional head lets a push, which
cannot fail without losing a preallocated element, make progress when the
only per-CPU head is held by the interrupted context.

Also check the extra head from the pop path so that nodes placed there
can be reused.

Fixes: f2ac0e5 ("bpf: Convert percpu_freelist.c to rqspinlock")
Signed-off-by: Hui Su <sh_def@163.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/bpf/20260806175600.1993595-1-sh_def@163.com
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  kernel/bpf/percpu_freelist.c
  kernel/bpf/percpu_freelist.h

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=a24a146ae2cb5d9faeea96d10010fcb6579abe1f

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

The original Linux kernel CVE announcement for CVE-2026-89991 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89991

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:H/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: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: 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).