[CVE-2026-90241][MODERATE 7.0] iommu/vt-d: Tear down scalable-mode context on probe failure [ Upstream

[CVE-2026-90241][MODERATE 7.0] iommu/vt-d: Tear down scalable-mode context on probe failure [ Upstream

[AL-KERNEL]

CVE: CVE-2026-90241
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: iommu/vt-d: Tear down scalable-mode context on probe failure [ Upstream
Commit: 25ac85a9747cd63e1d166ace7b360a2cd9479d9d
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=25ac85a9747cd63e1d166ace7b360a2cd9479d9d
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90241
Analysis date: Fri, 18 Sep 2026 08:16:27 -0400
ActionableScore: 5
ActionableScore lower bound: 4
Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review
Manual review required: YES

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-90241 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90241

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-2026-90241	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-90241 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:H/UI:N/S:U/C:H/I:H/A:H';*CWE-416;*CWE-664;*CWE-672;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '6.4';DESCR 'An Intel VT-d scalable mode probe failure can leave already programmed context entries present while their PASID table is freed. The IOMMU may later walk a present context entry whose PASID table pointer references freed memory, which makes this a use after free candidate in DMA translation state rather than a simple warning. For the CVSS the PR:H is used because reliable triggering normally requires privileged control over PCI device probing, driver binding, hotplug, SR-IOV setup, or similar host device management. The issue is not network reachable and requires local host side device control. Impact is at least denial of service and in the paranoid case may include confidentiality and integrity impact if the stale PASID table pointer causes incorrect DMA translations or memory corruption.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5)	YES	ERRORPATH UAF DMAorINTERRUPT HARDWARE KPANIC MEMORY 	YES	NO	checked

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

ActionableScore=5
ActionableScoreLower=4
PrimitiveClass=CLASS_B
RealUAF=YES
HostAdminRequired=YES
NamespaceScopedPrivilege=NO
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO

1. ActionableScore

* Raw conservative score: 4
* Conservative score: 4
* Raw paranoid score: 5
* Paranoid score: 5
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

No floor or cap was applied.

2. Signal breakdown

Conservative signals:

* +2 Real UAF / lifetime violation. Scalable-mode context entries can remain present and point to a PASID table after intel_pasid_free_table(dev) frees it.
* +1 Active-state retirement bypass. The normal teardown protocol for active scalable-mode context entries is skipped on the probe failure path before freeing the PASID table.
* +1 Secondary translation state staleness. Context entries act as active secondary IOMMU translation references to the PASID table and may later be walked by hardware.
* +1 Privileged kernel/device-management memory corruption path. The stale object is part of Intel VT-d DMA translation state, which is higher value than passive metadata.
* +1 Realistic strong DoS / translation failure concern. A hardware walk of freed PASID-table memory can plausibly cause IOMMU faults, device failure, or host instability.
* -2 Host-admin-equivalent trigger requirement. Reliable triggering appears to require host-side PCI device probing, driver binding, hotplug, SR-IOV setup, or similar privileged device management.

Paranoid additional signal:

* +1 Weak exploitation leverage / unresolved DMA translation consequence. If freed PASID-table memory is reused while context entries remain present, the IOMMU may interpret stale or replacement memory as live translation state. This is not a proven arbitrary read or write primitive, but it is a technically defensible unresolved corruption concern.

Important non-awarded signals:

* CLASS_A not awarded. Controlled reclaim, directed DMA corruption, arbitrary kernel read/write, callback control, or demonstrated privilege escalation is not established.
* Local unprivileged trigger not awarded. No ordinary local user interface is shown for the complete trigger.
* Remote reachability not awarded. This is not network-triggerable from the supplied evidence.
* Concrete repeated release/refcount signal not awarded. The patch shows stale context entries pointing to freed PASID-table memory, not a second put/release/refcount mutation on the same reference.
* Confidentiality and integrity bonuses not awarded conservatively. C/I impact is plausible in the paranoid interpretation because this is DMA translation state, but no concrete disclosure or unauthorized write primitive is demonstrated.
* Extra Availability not awarded beyond the single strong DoS concern.
* Race/timing penalty not applied. The issue is an ordered probe-failure unwind bug, not a demonstrated hard timing race.

3. Reachability analysis

The relevant path is intel_iommu_probe_device() during Intel VT-d device probing. The failure occurs after intel_pasid_setup_sm_context() has programmed scalable-mode context entries for one or more DMA aliases, but before the probe completes successfully.

The realistic attacker model is host-side device-management control. Triggering normally requires influencing PCI probe failure, driver binding, hotplug, SR-IOV, device assignment, or similar host-level hardware management. That is host-admin-equivalent in ordinary deployments.

Namespace-scoped privilege is not established. PCI/IOMMU probing and scalable-mode context programming are host-global device-management operations, not operations that an attacker can complete inside an attacker-created user or network namespace.

No direct network trigger is established. A malicious or faulty device or firmware could materially influence the condition, but the supplied evidence does not establish direct remote reachability. No special timing race is required, but the condition depends on partial failure during DMA alias setup.

4. Primitive and severity interpretation

The exact primitive is a stale active IOMMU translation reference: present scalable-mode context entries may continue to reference a PASID table after that table has been freed.

This is an active structured lifetime bug, not fixed stale storage and not merely passive metadata. The stale object participates in DMA translation. Later hardware or IOMMU activity may consume the freed PASID-table pointer through still-present context entries.

A concrete repeated release/refcount operation is not present. The bug is not a double-put or double-free sequence through the same software reference. It is a use-after-free candidate caused by freeing the PASID table before retiring all active context entries that point to it.

PrimitiveClass is CLASS_B because a real lifetime/UAF primitive in active DMA translation state is established, but no controlled reclaim, arbitrary R/W, callback hijack, demonstrated LPE, or concrete boundary escape is shown.

The demonstrated practical consequence is at least host instability or DoS through stale IOMMU state. The speculative but technically meaningful consequence is incorrect DMA translation or memory corruption if the freed PASID-table memory is reused and later interpreted as translation data.

ImportantCandidate=YES because this is high-value active IOMMU translation state with unresolved exploitation consequences beyond an ordinary cleanup crash. AutoDowngradeAllowed=NO because automatic downgrade would discard a concrete active-state UAF in DMA isolation machinery.

5. One-sentence report phrase

Intel VT-d scalable-mode probe failure can leave present context entries pointing to a freed PASID table, creating a host-side IOMMU use-after-free candidate with at least DoS impact and unresolved DMA translation corruption risk.

6. Manual review recommendation

MANUAL CHECK REQUIRED

The issue affects active IOMMU/PASID translation state and may involve stale hardware walks of freed memory, so it should not be automatically closed as an ordinary privileged cleanup bug without reviewing DMA isolation and reuse consequences.

7. Call-site confidence

Call-site confidence: medium

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

Patch: iommu/vt-d: Tear down scalable-mode context on probe failure [ Upstream
Commit: 25ac85a9747cd63e1d166ace7b360a2cd9479d9d
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=25ac85a9747cd63e1d166ace7b360a2cd9479d9d

Changed files:
  drivers/iommu/intel/iommu.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=25ac85a9747cd63e1d166ace7b360a2cd9479d9d

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

The original Linux kernel CVE announcement for CVE-2026-90241 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90241

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:H/UI:N/S:U/C:H/I:H/A:H
  The Best / paranoid CVSS score: 6.4

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).