[CVE-2026-97992][MODERATE REGULAR] vhost-vdpa: protect config_ctx from being freed under the config callback [ Upstream

[CVE-2026-97992][MODERATE REGULAR] vhost-vdpa: protect config_ctx from being freed under the config callback [ Upstream

[AL-KERNEL]

CVE: CVE-2026-97992
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: vhost-vdpa: protect config_ctx from being freed under the config callback [ Upstream
Commit: bf35c8948ab7a3eaf1f724fa1e870e3f73a1749b
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=bf35c8948ab7a3eaf1f724fa1e870e3f73a1749b
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97992
Analysis date: Sat, 26 Sep 2026 16:17:25 -0400
ActionableScore: 4
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-2026-97992 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97992

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-2026-97992	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-97992 CHECK WITH IMPACT FROM ORIG NN 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:H/I:H/A:H';*CWE-416;CWE-362;CWE-667;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'A race in vhost vdpa allows vhost_vdpa_config_cb to dereference a freed eventfd config_ctx while another thread replaces the configuration callback through VHOST_VDPA_SET_CONFIG_CALL. The callback reads config_ctx without a lock or reference, while eventfd_ctx_put can immediately free the old object. This produces a confirmed kernel use after free in eventfd_signal and can at least crash the host kernel. For the CVSS the PR:L value is used for the paranoid score because access to VDUSE and vhost vdpa interfaces can be delegated to a local service account or constrained container root even though default device permissions commonly require root. The issue is local and not directly network reachable. Impact is at least local denial of service via kernel crash and in the worst case may allow confidentiality integrity or privilege impact due to kernel memory corruption. YES REQUIRES MANUAL CHECK';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 3)	YES	READ OOB LOCK DANGER INIT UAF LINUS  INCREASED_TO_MODERATE7_BASED_ON_GUESSCVSS DECREASED_TO_MODERATEREG_BASED_ON_ACTIONABLESCORELESSTHAN4	NO	NO	checked

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

ActionableScore=4
ActionableScoreLower=2
PrimitiveClass=CLASS_B
RealUAF=YES
HostAdminRequired=YES
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES

1. ActionableScore

* Raw conservative score: 2
* Conservative score: 2
* Raw paranoid score: 4
* Paranoid score: 4
* Final recommended bucket: **Borderline / manual review recommended**::

Conservative arithmetic is based on the default VDUSE access model stated in the input, where access to `/dev/vduse` is root restricted. The paranoid score assumes a deployment where a local non-admin service account has been explicitly delegated access to the required VDUSE and vhost-vdpa device interfaces. No floor was applied.

2. Signal breakdown

Positive signals:

* +2 Real UAF / active lifetime violation. `vhost_vdpa_config_cb()` reads `v->config_ctx` without a lock or retained reference. `VHOST_VDPA_SET_CONFIG_CALL` replaces that pointer and `eventfd_ctx_put()` can immediately free the old `eventfd_ctx`. The callback subsequently dereferences the freed object through `eventfd_signal()`.
* +1 Reliable kernel crash or strong DoS. The supplied KASAN trace demonstrates slab use after free in `native_queued_spin_lock_slowpath()` reached through `eventfd_signal_mask()`.
* +1 Attacker influenced active-object lifetime manipulation. The supplied VDUSE sequence allows the same local process or cooperating local threads to repeatedly inject config interrupts while replacing the config eventfd callback. This directly drives both the stale-reader path and the lifetime transition.

Negative signal for the conservative score:

* -2 Host-admin-equivalent access in the default deployment. The input states that access to `/dev/vduse` is root restricted by default. The complete demonstrated trigger also requires control of the relevant vhost-vdpa configuration interface.

No hard-race penalty was applied. Although this is a race, the supplied evidence says the concurrent injection and callback replacement sequence reproduces in seconds, so the timing is repeatable and attacker driven.

Important non-awarded signals:

* Controlled reclaim or reliable replacement of the freed allocation was not demonstrated.
* Arbitrary or directed kernel write was not demonstrated.
* Structured stale-object exploitation leverage was not separately awarded. The stale object is an active `eventfd_ctx`, but the supplied evidence demonstrates a stale `eventfd_signal()` operation and crash, not attacker controlled post-free replacement or consumption of attacker shaped replacement fields.
* Weak LPE concern was not awarded solely because this is a UAF.
* Confidentiality and integrity were not awarded. The patch establishes a stale kernel object dereference, but no disclosure, arbitrary write, credential modification, or security-boundary bypass.
* Remote reachability was not established. The trigger requires local access to VDUSE and vhost-vdpa interfaces.
* Extra Availability was not awarded because the demonstrated kernel crash is already counted once.

3. Reachability analysis

The demonstrated attacker is a local userspace process that can access both the VDUSE device interface and the corresponding vhost-vdpa configuration interface. It creates or controls eventfd objects, injects a VDUSE configuration interrupt, and concurrently invokes `VHOST_VDPA_SET_CONFIG_CALL` to replace the callback eventfd.

The supplied input positively states that `/dev/vduse` is root restricted by default. Therefore, host-admin-equivalent access is required for the conservative/default-deployment interpretation. The input does not establish that an ordinary user can create the VDUSE device, obtain the required vhost-vdpa file descriptor, and perform the complete sequence.

A delegated service account could plausibly be granted access to these device interfaces in a virtualization deployment. That is the basis for the paranoid score of 4, but such delegation is not established as a default or universally reachable condition.

NamespaceScopedPrivilege is UNKNOWN. The supplied evidence does not show that all required VDUSE creation, config IRQ injection, vhost-vdpa configuration, and callback replacement operations are authorized entirely inside an attacker-controlled namespace. No network namespace or container path is sufficient by itself.

No special physical hardware is required by the demonstrated VDUSE path. The race is repeatable because the attacker can repeatedly issue both relevant operations from cooperating threads.

4. Primitive and severity interpretation

The exact primitive is a real active-object use after free of `struct eventfd_ctx`. The old `config_ctx` is removed from `v->config_ctx`, its last reference is dropped through `eventfd_ctx_put()`, and it may be freed immediately. A callback that previously loaded the pointer subsequently invokes `eventfd_signal()` on the stale allocation.

This is an active structured object rather than fixed stale storage. However, the supplied evidence currently supports an ordinary stale dereference during eventfd signaling and a kernel fault. It does not establish controlled reclaim, object substitution, callback hijacking, arbitrary read or write, refcount takeover, or an attacker controlled post-free object layout.

This is not a concrete repeated release or refcount mutation through the same dangling pointer. The first `eventfd_ctx_put()` frees the old context, while the later stale operation is `eventfd_signal()`, not a second `eventfd_ctx_put()` or another explicit lifetime decrement. Therefore, the Concrete repeated release/refcount signal is not awarded, and the related FLOOR 5 condition does not apply.

The bug is CLASS_B because it is a confirmed kernel lifetime violation with a direct stale-object dereference. It is not CLASS_A because no strong exploitation primitive beyond the use after free is established. The demonstrated consequence is kernel crash or denial of service. A stronger consequence from allocator reuse remains speculative and is not scored as confidentiality, integrity, or privilege escalation.

ImportantCandidate is NO because the demonstrated stale operation is constrained to eventfd signaling, the default trigger is root restricted, and no meaningful reuse, post-free structured-field control, repeated stale refcount mutation, or security-boundary primitive is established. AutoDowngradeAllowed is YES because the available evidence supports a constrained crash-capable CLASS_B UAF rather than unresolved strong exploitation leverage. Manual review remains recommended because this is still a confirmed kernel UAF and deployment-specific device delegation can materially change the practical attacker model.

Likely operational severity is a local privileged or delegated-service kernel DoS. It should be reviewed where VDUSE is enabled and exposed to non-root virtualization services, but the patch alone does not establish an Important-class exploitation path.

5. One-sentence report phrase

A race between VDUSE config interrupt delivery and vhost-vdpa config callback replacement can dereference a freed eventfd context and crash the kernel, requiring local access to the relevant VDUSE and vhost-vdpa device interfaces.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Confirmed use after free exists, but the supplied evidence demonstrates a crash path without controlled reclaim, arbitrary write, information disclosure, repeated stale refcount mutation, or privilege escalation. Review deployed device permissions and whether non-admin virtualization services can access the complete VDUSE and vhost-vdpa trigger path.

7. Call-site confidence

Call-site confidence: high

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

Patch: vhost-vdpa: protect config_ctx from being freed under the config callback [ Upstream
Commit: bf35c8948ab7a3eaf1f724fa1e870e3f73a1749b
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=bf35c8948ab7a3eaf1f724fa1e870e3f73a1749b

Changed files:
  drivers/vhost/vdpa.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=bf35c8948ab7a3eaf1f724fa1e870e3f73a1749b

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

The original Linux kernel CVE announcement for CVE-2026-97992 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97992

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:L/I:L/A:H
  The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H
  The Best / paranoid CVSS score: 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: 4

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).