[CVE-2026-90128][MODERATE 7.0] vdpa/mlx5: fix wrong list iterated in add_direct_chain error path [ Upstream

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

CVE: CVE-2026-90128
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: vdpa/mlx5: fix wrong list iterated in add_direct_chain error path [ Upstream
Commit: c93defccf5eb0a92bdafa43487be6ce0221a2477
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=c93defccf5eb0a92bdafa43487be6ce0221a2477
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90128
Analysis date: Fri, 18 Sep 2026 08:43:24 -0400
ActionableScore: 5
ActionableScore lower bound: 4
Actionable bucket: Actionable Moderate
Manual review required: NO

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-90128 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90128

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: NO

A detailed explanation of the methodology and priority rules is included
at the end of this message.

======================================================================
AL-KERNEL CLASSIFICATION RESULT
======================================================================

CVE-2026-90128	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-90128 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:L/UI:N/S:U/C:L/I:L/A:H';*CWE-664;CWE-763;*CWE-401;*CWE-672;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '5.8';DESCR 'add_direct_chain in vdpa mlx5 used the wrong list in the allocation error cleanup path. Newly allocated direct MR entries were stored on a temporary list, but the error path iterated mr head instead of tmp. This can incorrectly unmap and free existing MR entries while leaving the newly allocated temporary entries uncleared, resulting in corrupted memory registration state and possible resource leakage. For the CVSS the PR:L is used in the paranoid score because some environments delegate vDPA or device management to reduced capability root, container root, or a service account rather than full host administration. The issue is local and is not directly network reachable. Impact is mainly denial of service against the vDPA device or dependent guest workloads, with limited confidentiality or integrity impact possible only from corrupted MR state. YES REQUIRES MANUAL CHECK.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate (with actual score 5)	YES	SIMPLEFIX ERRORPATH LINUS  KPANIC INCREASED_FROM_LOW_BASED_ON_REQUIREMANUALCHECK	NO	NO	checked

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

ActionableScore=5
ActionableScoreLower=4
PrimitiveClass=CLASS_B
RealUAF=NO
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=NO

1. ActionableScore

* Raw conservative score: 4
* Conservative score: 4
* Raw paranoid score: 5
* Paranoid score: 5
* Final recommended bucket: **Actionable Moderate**::

No floor or cap was applied.

2. Signal breakdown

Conservative signals:

* +1 Weak / indirect kernel object corruption candidate. The error path iterates mr->head instead of tmp, so cleanup can operate on existing MR entries rather than only newly allocated entries.
* +1 Privileged device-management memory-registration state corruption. The affected objects are mlx5 vDPA direct MR entries used for memory registration and unmapping, which are high-value DMA/device state even though arbitrary host memory corruption is not demonstrated.
* +1 Concrete persistent resource leak. Newly allocated entries on tmp are not cleaned by the buggy loop, so mappings or allocations from the failed operation can be leaked.
* +1 Availability impact. Incorrectly unmapping/freeing existing MR entries and leaking failed temporary entries can break vDPA memory registration and dependent guest or datapath operation.

Paranoid additional signal:

* +1 Weak exploitation concern from active MR state corruption. The affected state is not passive metadata. It controls device memory registration and unmap behavior, so a manual review should confirm whether stale mappings, wrong unmaps, or guest memory exposure can occur.

Non-awarded signals:

* Controlled reclaim not demonstrated.
* Arbitrary write not demonstrated.
* Real UAF not established from the supplied diff.
* Concrete repeated release/refcount signal not awarded. The patch shows wrong-list cleanup, not a first release followed by a second lifetime mutation on the exact same dangling reference.
* Structured stale-object leverage not separately awarded because no later stale field consumption is shown.
* Confidentiality and integrity points are not awarded conservatively because the patch does not establish unauthorized data read/write, only corrupted MR state risk.
* Remote reachability not awarded. This is not a packet-triggered path.
* Race/timing penalty not applied. This is an ordered error-path bug, not a demonstrated hard race.
* Host-admin penalty not applied because host-admin-equivalent privilege is not positively established from the supplied evidence.

3. Reachability analysis

The likely trigger is local control over vDPA/mlx5 device setup or a VMM/service path that causes add_direct_chain() to allocate direct MR entries and then hit the err_alloc path. Ordinary unprivileged local reachability is not established. Full host administrator privilege is also not positively established, because vDPA device access may be delegated to a VMM process or management service.

Namespace-scoped privilege is unresolved. The supplied patch does not show that the complete operation can be performed inside an attacker-created user or network namespace.

Prior privileged setup may exist in real deployments, for example provisioning a vDPA device to a VM or management service, but the supplied evidence does not fully define that boundary. Special mlx5-capable hardware is required, but this is a normal data-center device class rather than evidence for a rare-hardware penalty.

4. Primitive and severity interpretation

The exact primitive is wrong-object cleanup in a DMA/device memory-registration error path. New direct MR entries are placed on tmp and should be cleaned from tmp if allocation fails. The buggy code instead walks mr->head, which can remove, unmap, and free existing MR entries while leaving the failed temporary entries uncleared.

This is active structured device state, not fixed stale storage. However, the supplied evidence does not establish a real UAF, controlled reclaim, arbitrary read/write, callback control, refcount takeover, or privilege escalation.

PrimitiveClass is CLASS_B because the bug corrupts active lifetime/ownership state in a kernel device memory-registration path. It is not CLASS_A because no strong exploitation primitive is demonstrated. It is not CLASS_C because the effect is more than a simple NULL dereference or passive warning.

ImportantCandidate is NO because the current evidence does not show controlled corruption, LPE, guest-to-host escape, or sensitive data access. AutoDowngradeAllowed is NO because active MR/DMA state corruption plus resource leakage should not be automatically dismissed without checking whether wrong unmaps or stale mappings can affect isolation or guest memory semantics.

Likely operational severity is Actionable Moderate.

5. One-sentence report phrase

A wrong-list cleanup bug in vdpa mlx5 add_direct_chain() can unmap and free existing direct MR entries while leaking newly allocated temporary entries on the allocation error path, causing corrupted vDPA memory-registration state and possible denial of service for dependent device or guest workloads.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Manual review is recommended because the bug affects active vDPA/mlx5 memory-registration and unmap state. The patch does not prove privilege escalation or data exposure, but wrong cleanup of MR entries can have device isolation and guest memory consequences that are not fully resolved from the one-line diff.

7. Call-site confidence

Call-site confidence: medium

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

Patch: vdpa/mlx5: fix wrong list iterated in add_direct_chain error path [ Upstream
Commit: c93defccf5eb0a92bdafa43487be6ce0221a2477
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=c93defccf5eb0a92bdafa43487be6ce0221a2477

Changed files:
  drivers/vdpa/mlx5/core/mr.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=c93defccf5eb0a92bdafa43487be6ce0221a2477

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

The original Linux kernel CVE announcement for CVE-2026-90128 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90128

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:N/I:N/A:H
  The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H
  The Best / paranoid CVSS score: 5.8

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: 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).