[CVE-2026-89791][IMPORTANT] perf: Fix use-after-free when perf mmap() revival races with the last munmap() [ Upstream

[CVE-2026-89791][IMPORTANT] perf: Fix use-after-free when perf mmap() revival races with the last munmap() [ Upstream

[AL-KERNEL]

CVE: CVE-2026-89791
Priority: IMPORTANT
AL-KERNEL base severity: IMPORTANT
KPANIC flag: YES
Patch: perf: Fix use-after-free when perf mmap() revival races with the last munmap() [ Upstream
Commit: 929cb3b9dc818dd9fa89d510d4ff2b255e42badd
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=929cb3b9dc818dd9fa89d510d4ff2b255e42badd
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89791
Analysis date: Wed, 16 Sep 2026 05:08:52 -0400
ActionableScore: 8
ActionableScore lower bound: unknown
Actionable bucket: Strong Important candidate
Manual review required: YES

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-89791 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89791

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: IMPORTANT
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-2026-89791	IMPORTANT	https://www.kernelcve.org/list/?q=CVE-2026-89791 CHECK WITH IMPACT FROM ORIG NN MODERATE	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-416;CWE-362;CWE-667;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H';BEST CVSS score: '7.8';DESCR 'A race between perf_mmap_close and perf_mmap_rb can free a newly attached perf ring buffer while another process still has its pages mapped. The stale mapping creates a page level use after free because the released pages may be reused by the kernel while remaining accessible through the existing user mapping. For the CVSS the PR:L is selected because an unprivileged local user must execute code and use the permitted perf interface, while administrative privileges are not required when the default kernel.perf_event_paranoid value is 2. The base score uses AC:H because successful triggering depends on a race window between the final munmap and a concurrent mmap revival. The paranoid score uses AC:L because repeated local attempts may make the race practically reproducible and the commit explicitly identifies local privilege escalation to root. The issue is not network reachable. Impact can include kernel memory disclosure, kernel memory modification, privilege escalation, and denial of service.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate (with actual score 8)  SKIP CVE-2026-89791 UNKNOWN SKIP The Fixes patch not applied yet, so unlikely that actual: url YES  NO NO unknown 	MAYBE	READ LOCK RACE UAF HARDWARE LINUS KPANIC  KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS	-	-	checked

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

ActionableScore=8
PrimitiveClass=CLASS_A
RealUAF=YES
HostAdminRequired=NO
NamespaceScopedPrivilege=NO
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO

1. ActionableScore

* Raw conservative score: 8
* Conservative score: 8
* Raw paranoid score: 8
* Paranoid score: 8
* No floor or cap was applied.
* Final recommended bucket: **Strong Important candidate**::

2. Signal breakdown

* +2 Real UAF. The newly attached perf ring buffer can be freed while its pages remain mapped by another process.
* +2 Strong exploitation leverage. The surviving page mapping exposes storage after its kernel lifetime ends, and the supplied authoritative description explicitly establishes local privilege escalation to root.
* +1 Local unprivileged trigger. The complete trigger is stated to be available to any unprivileged user with the default kernel.perf_event_paranoid value of 2.
* +1 Local unprivileged high-control API. The attacker controls concurrent perf mmap and munmap operations, the vulnerable lifetime transition, and continued access through the stale mapping.
* +1 Confidentiality plausible. Reclaimed pages remain accessible through the stale user mapping, providing a technically supported read-back path to subsequently reused storage.
* +1 Integrity plausible. The page-level stale mapping and explicitly established root escalation support unauthorized modification of kernel-owned state.
* +0 Practical exploit-chain bonus was not separately awarded because it would duplicate the demonstrated LPE and strong-leverage signal.
* +0 Attacker-influenced lifetime manipulation was not separately awarded because the same mmap and munmap control is already represented by the high-control API signal.
* +0 Concrete repeated release/refcount signal was not awarded. This is a refcount-ordering race that frees the replacement buffer, not a demonstrated first release followed by a second lifetime mutation through the exact same dangling reference.
* +0 Reliable crash was not separately awarded because the supplied evidence focuses on page-level UAF and LPE, and no independent crash consequence is needed for the score.
* No race penalty. Concurrent mmap and munmap operations can be attempted repeatedly and are controlled by the local attacker.
* Controlled allocation of the exact replacement object is not separately demonstrated, but it is unnecessary to preserve Class A because page-level post-free access and root LPE are explicitly established.
* Remote reachability is not established and was not scored.

3. Reachability analysis

The evaluated attacker is an ordinary local unprivileged user, not a namespace administrator, delegated service, or host administrator. The supplied evidence states that the vulnerable perf interface is reachable with the default kernel.perf_event_paranoid value of 2 and explicitly identifies privilege escalation to root.

No namespace capability is required or relevant to the complete trigger, so NamespaceScopedPrivilege is NO. No prior host provisioning, special device, unusual hardware, or attacker-created namespace is identified as necessary. A more restrictive administrator-selected perf_event_paranoid setting could reduce exposure, but that does not negate the positively established below-host-admin path under the stated default configuration.

The attacker can repeatedly coordinate mmap revival with the final munmap from concurrent threads or processes. The race requires a particular interleaving, but the operations are inexpensive and repeatable, so a hard-race penalty is unsupported.

4. Primitive and severity interpretation

The exact primitive is a page-level use-after-free. perf_mmap_close first allows rb->mmap_count to reach zero without event->mmap_mutex. A concurrent perf_mmap_rb then allocates and attaches a replacement ring buffer and establishes a mapping. The closing path subsequently treats event->mmap_count as reaching zero, detaches the replacement buffer, and drops the reference that frees it while another process retains its mapped pages.

This is not fixed asynchronous stale storage and not an ordinary cleanup-only stale dereference. The stale VMA remains usable after the backing pages return to kernel allocation, creating post-lifetime access to reusable pages. The supplied evidence directly connects this primitive to local root escalation.

There is no separate concrete repeated-release primitive under the dedicated rule. Although refcounts and release operations participate in the race, the important mechanism is incorrect serialization and freeing of the revived buffer, not a deterministic second refcount mutation through an already dangling reference.

PrimitiveClass is CLASS_A because a strong page-level UAF and local privilege escalation are established. The demonstrated consequence is privilege escalation to root. Kernel memory disclosure, modification, and denial of service are technically consistent with the same stale page mapping. Arbitrary kernel read or write is not independently claimed beyond this page-reuse primitive.

ImportantCandidate is YES because the issue is reachable by an ordinary local user and has an explicitly established root escalation consequence. AutoDowngradeAllowed is NO because automatic Moderate classification would discard a concrete page-lifetime primitive and established privilege escalation. The likely operational severity is Important, near the upper end of that category, but the evidence does not require classification as Critical.

5. One-sentence report phrase

A local unprivileged user can race perf ring-buffer mmap revival against the final munmap to retain access to freed and reusable kernel pages, creating a page-level use-after-free that can escalate privileges to root.

6. Manual review recommendation

MANUAL CHECK REQUIRED

The vulnerability combines ordinary-user reachability, attacker-driven repeatable lifetime operations, a page-level UAF, and an explicitly established local root escalation path.

7. Call-site confidence

Call-site confidence: high

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

Patch: perf: Fix use-after-free when perf mmap() revival races with the last munmap() [ Upstream
Commit: 929cb3b9dc818dd9fa89d510d4ff2b255e42badd
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=929cb3b9dc818dd9fa89d510d4ff2b255e42badd

Changed files:
  kernel/events/core.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=929cb3b9dc818dd9fa89d510d4ff2b255e42badd

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

The original Linux kernel CVE announcement for CVE-2026-89791 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89791

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:H/I:H/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: 8

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: IMPORTANT
KPANIC detected for this report: YES
Published priority for this report (same as in Subject): IMPORTANT

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).