[CVE-2026-98639][MODERATE 7.0] KVM: riscv: Fix NACL hfence entry update order [ Upstream

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

CVE: CVE-2026-98639
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: KVM: riscv: Fix NACL hfence entry update order [ Upstream
Commit: d6de766d8d26d71b714f548192099112def50a9c
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=d6de766d8d26d71b714f548192099112def50a9c
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98639
Analysis date: Mon, 19 Oct 2026 03:02:32 -0400
ActionableScore: 5
ActionableScore lower bound: 4
Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review
Manual review required: YES

Summary:
KVM RISC-V can publish a NACL HFENCE entry before its page range is visible, allowing the privileged SBI or hardware consumer to flush stale ranges and potentially leave security-relevant translations active across an unresolved virtualization boundary.

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

The original Linux kernel CVE announcement for CVE-2026-98639 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98639

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-2026-98639	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-98639 CHECK WITH IMPACT FROM ORIG NN NONE | AEGIS IMPACT MODERATE AEGIS_AS=0 LEGACY=MODERATE_KPANIC RESULT=DIFF (final Legacy and Aegis differ; FINAL: MODERATE_KPANIC, Aegis: 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-362;CWE-662;CWE-696;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'The KVM RISC-V NACL HFENCE producer can set Config.Pending before updating Page_Number and Page_Count. The SBI implementation or NACL hardware may therefore consume stale range parameters and flush the wrong TLB entries while the intended translations remain active. A stale translation can potentially preserve access to a page after its mapping has changed, which creates a plausible confidentiality and integrity risk across a virtualization boundary and can also cause crashes or guest corruption. For the CVSS the PR:L value is used because an attacker needs controlled execution in a guest or nested virtualization context but does not need administrative privileges on the host. AC:H is used because successful triggering depends on the external consumer observing the entry during a narrow store ordering window. The issue is not directly network reachable. The paranoid CIA:H/H/H interpretation is justified as a manual review boundary because stale stage translation may affect isolation, although cross guest or guest to host exploitation is not proven by the supplied code. Automated impact decision path. The original SimpleNet classifier selected LOW. The primary LLM estimated paranoid CVSS 7 and the production ActionableScore returned 5 with recommendation Actionable Moderate / Strong Important candidate requiring manual review. The production rules then evaluated CVSS thresholds and ActionableScore thresholds. Recorded rule transitions were CVSS and initial triage stage LOW to MODERATE_KPANIC because the ActionableScore threshold rule was applied. The legacy KP specialist produced a review artifact and its decision was processed by the applicable KP rules. The legacy FD or DH downgrade specialist has no review artifact for this run. Exact triggered rule markers were INCREASED_TO_MODERATE7_FROM_LOW_BASED_ON_GUESSCVSS_AND_ACTIONABLESCOREHIGHER4. The independent weaker Aegis shadow evaluated CVSS 0.0 and ActionableScore 0 with primitive class CLASS_C and manual review RECOMMENDED. Its independently reconciled impact was MODERATE. The models disagreed but the aggregation rules retained the production result. The final reported impact is MODERATE_KPANIC.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5)	MAYBE	VIRT SIMPLEFIX SKIP HARDWARE LINUS  KPANIC INCREASED_TO_MODERATE7_FROM_LOW_BASED_ON_GUESSCVSS_AND_ACTIONABLESCOREHIGHER4	-	-	checked

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

ActionableScore=5
ActionableScoreLower=4
PrimitiveClass=CLASS_B
RealUAF=NO
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=NO
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO
RequiresHostUID0=UNKNOWN
AuthorizationDecisionBypass=NO
ProtectedResourceImpact=NONE
B4PendingReview=NO
B4PendingConfirmed=NO

1. ActionableScore

* Raw conservative score: 4.
* Conservative score: 4. No floor or cap applied.
* Raw paranoid score: 5.
* Paranoid score: 5. No floor or cap applied.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

2. Signal breakdown

Conservative signals:

* +2: Strong Class B active-state synchronization violation. `Config.Pending` makes an HFENCE descriptor active before its new page range is globally visible, so a consumer can process stale structured parameters.
* +1: Concrete privileged downstream consumption. The SBI implementation or NACL hardware consumes the prematurely published entry and performs an HFENCE for the wrong range.
* +1: Active translation synchronization failure. The intended TLB entries can remain operational after an invalidation was expected, extending the effect beyond passive metadata.
* Total: 4.

Additional paranoid signal:

* +1: Credible but unresolved virtualization trust boundary. The defect is in host KVM nested-virtualization handling and stale translations may affect memory isolation, but the supplied call sites do not prove guest control over the complete sequence or unauthorized host-memory access.
* Total: 5.

Not awarded:

* No UAF, OOB access, arbitrary write, controlled reclaim, or refcount corruption is established.
* No reliable host panic or kernel crash is documented, so no DoS point is awarded.
* No confidentiality or integrity point is awarded because access to a protected page or modification of protected state is not demonstrated.
* No direct network reachability is established.
* No hard-race penalty is applied. A producer-consumer observation window exists, but the evidence does not show that attempts are one-shot or impractical to repeat.
* No privilege penalty is applied because host-admin-equivalent privilege is not positively established.
* B5-R does not apply because there is no retry-accumulated heap OOB write.
* B3-PENDING does not apply because no shared file-backed or COW storage write is shown.
* B4-PENDING does not apply because no authorization decision changes from DENY to ALLOW.

3. Reachability analysis

The likely security principals are a nested guest or guest hypervisor issuing virtualization operations, a local VMM using KVM, and the host administrator configuring the VM. The patch confirms that host KVM publishes nested HFENCE requests to an external SBI or NACL consumer, but it does not show the callers needed to prove whether a malicious guest alone controls `page_num`, `page_count`, entry reuse, and the complete vulnerable timing sequence.

The trigger may therefore be guest-driven, VMM-driven, or compound. Host-admin privilege is not proven necessary, but complete below-host-admin reachability is also not sufficiently visible, so `HostAdminRequired=UNKNOWN`. User and network namespaces are not relevant to this RISC-V KVM interface. RISC-V NACL support, nested virtualization, coherent shared memory, and an SBI or hardware consumer are required. The ordering race can recur whenever HFENCE entries are reused, but practical attacker control over repetition remains unresolved.

4. Primitive and severity interpretation

The exact primitive is premature publication of an active shared-memory HFENCE descriptor. The consumer can observe `Pending` with stale `Page_Number` and `Page_Count`, flush an obsolete range, and leave the intended TLB translations active.

This is active structured state rather than fixed stale storage. It is not a lifetime bug, and no repeated release or refcount mutation exists. It is classified as Class B because an explicit synchronization protocol for retiring active translations is violated and the incorrect descriptor is consumed by a privileged external agent.

The demonstrated consequence is an incorrect TLB flush range. Continued use of stale translations is directly implied. A guest crash, guest corruption, isolation failure, cross-guest access, or guest-to-host memory access remains unresolved rather than demonstrated.

`ImportantCandidate=YES` follows from the concrete active-state publication error, privileged downstream consumer, and credible unresolved KVM trust boundary. This is a review candidate and not proof of Important severity. `AutoDowngradeAllowed=NO` because automatic closure would discard the unresolved effect of operationally active stale translations across a virtualization boundary.

5. One-sentence report phrase

KVM RISC-V can publish a NACL HFENCE entry before its page range is visible, allowing the privileged SBI or hardware consumer to flush stale ranges and potentially leave security-relevant translations active across an unresolved virtualization boundary.

6. Manual review recommendation

MANUAL CHECK REQUIRED

Review the callers of `__kvm_riscv_nacl_hfence()`, guest control over the range and entry reuse, the affected translation stage, and whether stale entries can reference reassigned or host-owned pages.

7. Call-site confidence

Call-site confidence: medium

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

Patch: KVM: riscv: Fix NACL hfence entry update order [ Upstream
Commit: d6de766d8d26d71b714f548192099112def50a9c
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=d6de766d8d26d71b714f548192099112def50a9c

Changed files:
  arch/riscv/kvm/nacl.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=d6de766d8d26d71b714f548192099112def50a9c

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

The original Linux kernel CVE announcement for CVE-2026-98639 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98639

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: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: 4
  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).