[CVE-2026-98715][IMPORTANT] KVM: SVM: Preserve TLB control (i.e. pending TLB flush) on failed VMRUN

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

CVE: CVE-2026-98715
Priority: IMPORTANT
AL-KERNEL base severity: IMPORTANT
KPANIC flag: NO
Patch: KVM: SVM: Preserve TLB control (i.e. pending TLB flush) on failed VMRUN
Commit: fd873c3d1b5240e80365f09e758e8c96b88a7cdb
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=fd873c3d1b5240e80365f09e758e8c96b88a7cdb
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98715
Analysis date: Sun, 18 Oct 2026 23:48:43 -0400
ActionableScore: 5
ActionableScore lower bound: 4
Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review
Manual review required: YES

Summary:
A failed SVM VMRUN could discard a pending all-ASID TLB flush, allowing stale translations to survive ASID reuse and requiring manual review for possible cross-vCPU or cross-VM memory-isolation impact.

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

The original Linux kernel CVE announcement for CVE-2026-98715 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98715

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-2026-98715	IMPORTANT	https://www.kernelcve.org/list/?q=CVE-2026-98715 CHECK WITH IMPACT FROM ORIG NN MODERATE | AEGIS IMPACT MODERATE_KPANIC AEGIS_AS=2 LEGACY=IMPORTANT RESULT=DIFF INDEPENDENT_LEGACY=IMPORTANT INDEPENDENT_RESULT=DIFF TEMP_KPANIC_MERGE=IMPORTANT_WINS_DISAGREEMENT PRE_MERGE_LEGACY=IMPORTANT (final Legacy and Aegis differ; FINAL: IMPORTANT, Aegis: MODERATE_KPANIC)	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:H/PR:L/UI:N/S:C/C:H/I:H/A:H';CWE-226;CWE-284;*CWE-672;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:C/C:L/I:L/A:H';BEST CVSS score: '7.8';DESCR 'KVM SVM could clear a pending TLB flush after a failed VMRUN even though the processor performs the requested flush only after a successful VMRUN. This may leave stale TLB translations available when an ASID is reused and a later VMRUN succeeds. A malicious guest or nested guest may be able to influence the required sequence, although reliable triggering requires an ASID rollover, a failed VMRUN, and later execution on the affected physical CPU. For the CVSS the PR:L value is used because attacker controlled code must run in a guest context, but host administrative privileges are not inherently required. The issue is not directly network reachable. Impact is at least a denial of service and may include cross vCPU or cross VM confidentiality and integrity impact if stale translations resolve to memory associated with a previous ASID owner. The paranoid score uses CIA:HHH because failure to invalidate translations before ASID reuse can plausibly violate virtualization memory isolation, but direct host code execution is not established. Automated impact decision path. The original SimpleNet classifier selected MODERATE. The primary LLM estimated paranoid CVSS 7.8 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 MODERATE to IMPORTANT because the CVSS score, vector or bug class threshold rule was applied. final reconciliation stage IMPORTANT to IMPORTANT because KP specialist did not establish a qualifying panic. 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 produced a review artifact and its decision was processed by the applicable downgrade and preservation rules. Exact triggered rule markers were INCREASED_TO_HIGH_BASED_ON_GUESSCVSS and TEMP_KPANIC_MERGE_IMPORTANT_WINS and KPANIC_UNSUPPORTED_REMOVED. The independent weaker Aegis shadow evaluated CVSS 3.6 and ActionableScore 2 with primitive class CLASS_B and manual review REQUIRED. Its independently reconciled impact was MODERATE_KPANIC. Cross model aggregation changed the production result under an enabled evidence gated merge rule. The final reported impact is IMPORTANT.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5)	MAYBE	VIRT INIT HARDWARE TEST INCREASED_TO_HIGH_BASED_ON_GUESSCVSS TEMP_KPANIC_MERGE_IMPORTANT_WINS KPANIC_UNSUPPORTED_REMOVED	-	-	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
* Raw paranoid score: 5
* Paranoid score: 5
* No floor or cap was applied.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

2. Signal breakdown

Conservative signals:

* +1 Weak or indirect corruption candidate. A required TLB invalidation can be discarded, leaving an operationally stale hardware translation after ASID reuse.
* +1 Active-state retirement bypass. `TLB_CONTROL_FLUSH_ALL_ASID` is the explicit retirement mechanism, but the vulnerable path clears it after a failed VMRUN even though hardware did not perform the flush.
* +1 Active semantic or translation-domain transition. An ASID is reused in a new generation while translations associated with its previous use may remain active.
* +1 Active translation synchronization failure. The missing TLB flush means stale state can survive outside the VMCB software structure and be consumed by a later successful VMRUN.

Paranoid additional signal:

* +1 Credible unresolved trust-boundary consumption. The stale translation is consumed by privileged CPU virtualization machinery after guest-triggerable ASID activity. Cross-vCPU or cross-VM impact is technically credible but not demonstrated.

Not awarded:

* No UAF, OOB access, heap overwrite, arbitrary write, controlled reclaim, callback control, or refcount corruption is established.
* No reliable host panic or host-wide DoS is shown.
* No confidentiality or integrity point is awarded as a demonstrated consequence because access to another VM or host-owned memory is unresolved.
* No direct remote-reachability point is awarded.
* No strong guest-to-host exploitation leverage is awarded merely because the defect is in host KVM state.
* No race penalty is applied. This is an ordered failed-VMRUN state transition, not an evidenced narrow timing race.
* NEW52 B5-R does not apply because there is no heap OOB write.
* NEW53 B3-PENDING does not apply because there is no shared-backing write.
* NEW54 B4-PENDING does not apply because no authorization DENY-to-ALLOW decision is involved.

3. Reachability analysis

The commit explicitly states that `new_asid()` is guest-triggerable through KVM behavior around emulated INIT. This establishes guest influence over ASID generation rollover, but the supplied context does not fully establish that the same malicious guest can reliably force the required failed hardware VMRUN and subsequent successful run without VMM participation.

A local VMM process with `/dev/kvm` access can run the vCPU and has broader control over guest state, but the patch does not establish whether ordinary `/dev/kvm` access alone guarantees the complete trigger. Host-administrator privileges and host UID 0 are therefore not positively established, but complete below-host-admin reachability is also not proven.

This is not a namespace-capability path. No memslot deletion, host mapping replacement, MMU-notifier operation, or host administrative reconfiguration is identified as a required stage. AMD SVM with ASID support is required, but this is not sufficient evidence for a rare-hardware penalty. The trigger is a state sequence involving ASID rollover, failed VMRUN, and a later successful VMRUN rather than a timing race.

4. Primitive and severity interpretation

The exact primitive is continued availability of a stale CPU TLB translation after the logical ASID generation has changed. The pending flush is active retirement state. Clearing it after failed VMRUN falsely records retirement as complete even though the processor performs the flush only on successful VMRUN.

This is active translation state rather than fixed stale storage. A later successful VMRUN can consume the stale translation. Reuse of the ASID can reinterpret that translation under a different vCPU or ownership context, which creates a concrete active-state consistency primitive.

There is no repeated release or refcount operation. There is no literal UAF. The issue is CLASS_B because a live translation bypasses its mandatory retirement and synchronization protocol. It is not CLASS_A because the patch does not establish directed memory access, a chosen victim, host memory corruption, cross-VM disclosure, privilege escalation, or VM escape.

The demonstrated consequence is a dropped TLB flush. Cross-vCPU or cross-VM memory-isolation impact is an unresolved but technically defensible consequence. A host panic is not demonstrated. ImportantCandidate is YES because stale translations are consumed by privileged virtualization hardware across a credible security boundary and the ultimate translation ownership remains unresolved. AutoDowngradeAllowed is NO because reducing this to passive bookkeeping would discard the active-state and trust-boundary primitive.

5. One-sentence report phrase

A failed SVM VMRUN could discard a pending all-ASID TLB flush, allowing stale translations to survive ASID reuse and requiring manual review for possible cross-vCPU or cross-VM memory-isolation impact.

6. Manual review recommendation

MANUAL CHECK REQUIRED

Confirm whether a guest alone can force the complete failed-VMRUN sequence and whether a stale translation can resolve to memory associated with a different vCPU, VM, or host-owned mapping.

7. Call-site confidence

Call-site confidence: medium

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

Patch: KVM: SVM: Preserve TLB control (i.e. pending TLB flush) on failed VMRUN
Commit: fd873c3d1b5240e80365f09e758e8c96b88a7cdb
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=fd873c3d1b5240e80365f09e758e8c96b88a7cdb

Commit description:

Don't reset the VMCB's TLB control back to "do nothing" on a failed VMRUN,
as empirical testing shows that the CPU performs the requested TLB flush if
and only if VMRUN is successful, i.e. clearing TLB control on a failed
VMRUN effectively drops a TLB flush.

Explicitly track the need to flush all ASIDs on a per-CPU basis, as the
ASID reuse condition is tied to the pCPU, not to the vCPU.  As a bonus,
this also obviates the need to avoid clobbering FLUSH_ALL_ASID with
TLB_CONTROL_FLUSH_ASID, e.g. in svm_flush_tlb_asid().

Deliberately don't bother saving/restoring the "old" tlb_ctl on failure,
in quotes because it's not exactly the old tlb_ctl, it's the tlb_ctl from
after pre_svm_run(), but before updating tlb_ctl for flush_all_asids.  If
VMRUN fails and TLB_CONTROL_FLUSH_ALL_ASID is forced, then the next
successful run of the VMCB *may* unnecessarily flush all ASIDs, which
strictly speaking could result in noisy neighbor issues.  However, the
fact that new_asid() is already guest-triggerable, because of KVM's flawed
behavior of clearing the ASID on emulated INIT, means that a guest can
already trigger a flush of all ASIDs at roughly the same rate.  And once
KVM stops clobbering the ASID on emulated INIT, *or* assigns a static ASID
to each vCPU, this flaw goes away.

Fixes: 38e5e92 ("KVM: SVM: Implement Flush-By-Asid feature")
Cc: stable@vger.kernel.org
Reported-by: Stefan Teodorescu <fane@google.com>
Suggested-by: Yosry Ahmed <yosry@kernel.org>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Jim Mattson <jmattson@google.com>
Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-ID: <20260904170642.3291466-2-seanjc@google.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  arch/x86/kvm/svm/svm.c
  arch/x86/kvm/svm/svm.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=fd873c3d1b5240e80365f09e758e8c96b88a7cdb

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

The original Linux kernel CVE announcement for CVE-2026-98715 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98715

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:C/C:L/I:L/A:H
  The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:C/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: 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: IMPORTANT
KPANIC detected for this report: NO
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).