[CVE-2026-68329][IMPORTANT] iommu/amd: Wait for completion instead of returning early in iommu_completion_wait() [ Upstream

[CVE-2026-68329][IMPORTANT] iommu/amd: Wait for completion instead of returning early in iommu_completion_wait() [ Upstream

[AL-KERNEL]

CVE: CVE-2026-68329
Priority: IMPORTANT
AL-KERNEL base severity: IMPORTANT
KPANIC flag: YES
Patch: iommu/amd: Wait for completion instead of returning early in iommu_completion_wait() [ Upstream
Commit: ab7faf5a172ebfdc423ebb3eea4d472740de82f9
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=ab7faf5a172ebfdc423ebb3eea4d472740de82f9
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-68329
Analysis date: Mon, 10 Aug 2026 19:21:20 -0400
ActionableScore: 7
ActionableScore lower bound: 5
Actionable bucket: Strong Important candidate / Actionable Moderate at minimum
Manual review required: YES

Summary:
AMD IOMMU completion waiting could return before prior invalidation commands completed, allowing page-table pages to be freed while stale IOMMU translations may still be used, creating a race-dependent UAF window with at least DoS impact and possible DMA isolation consequences.

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

The original Linux kernel CVE announcement for CVE-2026-68329 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-68329

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-2026-68329	IMPORTANT	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-362;*CWE-416;CWE-667;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'AMD IOMMU completion wait could return before previously queued invalidation commands completed because need_sync was treated as completion state instead of only CWAIT queued state. A local attacker who can cause IOMMU mapping teardown or device DMA unmap races may trigger a window where page table pages are freed while the IOMMU can still walk stale translations. For the CVSS the PR:L is used for the paranoid score because practical triggering may be possible from a local user or service with access to VFIO, device passthrough, or DMA heavy device operations, not necessarily full host root. The issue is not directly network reachable, but network devices using DMA could be part of a local or adjacent attack setup if their IOMMU domain is being torn down. Impact is at least denial of service and stale DMA translation risk. In the paranoid case it may allow confidentiality or integrity impact due to IOMMU page table use after free, so manual review is needed.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate / Actionable Moderate at minimum (with actual score 7)	YES	LOCK RACE UAF HARDWARE TEST  KPANIC INCREASED_TO_MODERATE7_BASED_ON_GUESSCVSS KPANIC INCREASED_TO_HIGH_BASED_ON_ACTIONABLESCOREHIGHEREQTHAN7	NO	NO	checked

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

ActionableScore=7
ActionableScoreLower=5

## 1. ActionableScore

* Conservative score: 5
* Paranoid score: 7
* Final recommended bucket: **Strong Important candidate / Actionable Moderate at minimum**::

## 2. Signal breakdown

Conservative signals:

* Local unprivileged or low-privileged trigger: +1. Practical triggering may be possible by a local user or service with access to VFIO, passthrough devices, or DMA mapping teardown paths.
* Memory corruption, weak or indirect corruption candidate: +1. The patch explicitly describes a stale IOMMU translation window where page-table pages may be freed while still walkable by hardware.
* Real lifetime corruption: +1. This is a lifetime ordering bug involving invalidation completion and freeing page-table pages.
* Privileged kernel/device-management memory corruption path: +1. The affected code is AMD IOMMU command completion handling, which protects DMA isolation and page-table lifetime.
* Availability impact realistic: +1. A stale IOMMU walk over freed page tables can plausibly cause device faults, host instability, or kernel crash.
* Hard race / special timing required: -1. The issue depends on a multi-CPU ordering race between invalidation, CWAIT queuing, and page-table freeing.
* Weak privilege escalation concern: +1. Stale DMA translations in an IOMMU page-table context can plausibly affect confidentiality or integrity, but the patch does not demonstrate controlled reclaim or a reliable write primitive.

Conservative total: 5.

Paranoid adjustment:

* Privilege escalation plausible: +2 instead of weak +1. In the paranoid interpretation, stale IOMMU translations after freed page-table pages could become a DMA isolation bypass if memory is reused in an attacker-relevant way.
* Confidentiality impact plausible: +1. Stale DMA access may expose memory that should no longer be mapped to the device.
* Integrity impact plausible: +1. Stale DMA access may permit writes through obsolete translations if the device can DMA through the stale mapping.
* Keep hard race penalty: -1. Exploitation still depends on timing and device or IOMMU-domain activity.

Paranoid total: 7.

Call-site confidence: medium. The commit message gives the relevant consequence directly, including freeing page-table pages while the IOMMU can still walk stale translations, but the full caller tree is not present in the provided patch.

## 3. Reachability analysis

This is not directly network reachable. A remote packet alone does not call this path.

The realistic trigger is local or delegated device-management activity involving AMD IOMMU domains, DMA mappings, VFIO, passthrough devices, or similar map/unmap teardown paths. In typical systems, full exploitation may require access to a device or an IOMMU-backed interface configured by an administrator. However, that is not necessarily equivalent to full host root, because VFIO or device access may be delegated to a service account, VM process, container-like environment, or reduced-privilege user.

Namespaces do not automatically make this reachable, but delegated device access can reduce the effective privilege requirement from host admin to PR:L in practical CVSS-style reasoning. The path is hardware and configuration dependent because it requires AMD IOMMU usage and concurrent command activity, but AMD IOMMU is a normal production subsystem rather than an obscure debug-only path.

## 4. Severity interpretation

This should not be treated as an ordinary Moderate auto-close candidate. The bug is a race in IOMMU completion semantics, and the commit explicitly identifies a use-after-free window involving page-table pages and stale hardware translations.

Realistically, the most likely impact may be host instability or DoS under a difficult race. The theoretical but technically defensible concern is stronger: stale IOMMU translations can undermine DMA isolation and may allow confidentiality or integrity impact if freed page-table memory is reused in a useful way. There is no demonstrated arbitrary write or reliable LPE primitive in the patch, so the conservative score stays at Actionable Moderate. The paranoid score reaches Strong Important candidate because the affected object class is IOMMU page-table lifetime, not ordinary bookkeeping.

## 5. One-sentence report phrase

AMD IOMMU completion waiting could return before prior invalidation commands completed, allowing page-table pages to be freed while stale IOMMU translations may still be used, creating a race-dependent UAF window with at least DoS impact and possible DMA isolation consequences.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED

Reason: the patch explicitly describes a page-table use-after-free window in AMD IOMMU completion handling. Even though the race is timing-dependent and no reliable privilege-escalation primitive is shown, stale DMA translations in an IOMMU isolation path are security-sensitive enough to require manual review.

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

Patch: iommu/amd: Wait for completion instead of returning early in iommu_completion_wait() [ Upstream
Commit: ab7faf5a172ebfdc423ebb3eea4d472740de82f9
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=ab7faf5a172ebfdc423ebb3eea4d472740de82f9

Changed files:
  drivers/iommu/amd/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=ab7faf5a172ebfdc423ebb3eea4d472740de82f9

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

The original Linux kernel CVE announcement for CVE-2026-68329 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-68329

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: 5
  Paranoid ActionableScore: 7

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).