[CVE-2026-93151][MODERATE REGULAR] nvmet-rdma: fix response resource leak on queue teardown [ Upstream

[CVE-2026-93151][MODERATE REGULAR] nvmet-rdma: fix response resource leak on queue teardown [ Upstream

[AL-KERNEL]

CVE: CVE-2026-93151
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: nvmet-rdma: fix response resource leak on queue teardown [ Upstream
Commit: 91cfa74f1f3772e8c3662c064a45d8633e00c78e
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=91cfa74f1f3772e8c3662c064a45d8633e00c78e
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93151
Analysis date: Thu, 17 Sep 2026 15:00:15 -0400
ActionableScore: 4
ActionableScore lower bound: 3
Actionable bucket: Borderline / manual review recommended
Manual review required: YES

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-93151 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93151

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 REGULAR
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-2026-93151	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-93151 CHECK WITH IMPACT FROM ORIG NN LOW	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H';*CWE-401;*CWE-772;**CWE-400;Other CVSS 'AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:L';BEST CVSS score: '5.3';DESCR 'An RDMA response resource leak in the NVMe/RDMA target path can occur when a queue is torn down while I/O is still in flight. A response may have been posted but its send completion is not delivered before the connection is destroyed, so nvmet_rdma_send_done and nvmet_rdma_release_rsp are not reached and the RDMA read/write context and request SGL allocations are leaked. For the CVSS the PR:N is used because an allowed remote initiator can trigger queue teardown through the NVMe/RDMA fabric without local code execution on the target. The issue is adjacent network reachable rather than general Internet reachable because NVMe/RDMA is normally deployed on RDMA capable storage or data center networks. Impact is denial of service through gradual memory exhaustion after repeated disconnect or teardown cycles, with no concrete UAF or write primitive indicated by the patch context.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4)	YES	INIT NVME LEAK LINUS  INCREASED_TO_MODERATE7_FROM_LOW_BASED_ON_GUESSCVSS_AND_ACTIONABLESCOREHIGHER4 DECREASED_TO_MODERATEREG_BASED_ON_FALSEPOSCHECKOFKP	NO	NO	checked

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

ActionableScore=4
ActionableScoreLower=3
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=NO
NamespaceScopedPrivilege=NO
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES

1. ActionableScore

* Raw conservative score: 3
* Conservative score: 3. No floor or cap applied.
* Raw paranoid score: 4
* Paranoid score: 4. No floor or cap applied.
* Final recommended bucket: **Borderline / manual review recommended**::

2. Signal breakdown

Conservative signals:

* +2 Remote or fabric reachable trigger. An NVMe/RDMA initiator with access to the target fabric can have I/O in flight and cause connection or queue teardown without local code execution on the target.
* +1 Concrete persistent kernel resource/reference leak. The patch directly establishes that RDMA read/write context and request SGL allocations can be lost when send completion is not delivered before teardown.

Paranoid additional signal:

* +1 Sustained resource exhaustion concern. Repeated disconnect or teardown cycles with in-flight I/O can realistically accumulate leaked kernel memory and cause availability degradation.

Non-awarded signals:

* RealUAF not awarded. The patch describes missing cleanup of still-tagged responses, not post-free use of a stale object.
* Concrete repeated release/refcount not awarded. There is no second put/release/free through the same potentially dangling reference.
* Controlled reclaim not demonstrated.
* Arbitrary write not demonstrated.
* Confidentiality and integrity not awarded. The established primitive is resource leakage, not disclosure or modification of protected state.
* Reliable kernel crash not awarded. Kmemleak reports persistent leaked objects, but the patch does not establish immediate panic or fatal fault.
* Hard race penalty not applied. The condition requires in-flight I/O and teardown, but this is a repeatable lifecycle sequence rather than a narrow one-shot race.
* Host-admin penalty not applied. The supplied context supports a remote initiator/fabric-triggered queue teardown path and does not positively establish host-admin as required for the complete trigger.
* Rare hardware penalty not applied. RDMA/NVMe-oF is specialized, but the scoring input concerns deployments where the RDMA transport is present and reachable.

3. Reachability analysis

The relevant attacker is an NVMe/RDMA initiator with network or fabric reachability to the target. The trigger requires I/O to be in flight while the RDMA queue or QP is torn down, for example by forcefully disconnecting the connection before send completion is delivered. This is adjacent-network or storage-fabric reachable rather than ordinary public Internet exposure.

The issue does not require local code execution on the target in the evaluated attacker model. Prior target configuration is a deployment prerequisite, but the attacker does not need to perform host-side administrative setup once the target endpoint is exposed to them. No namespace-scoped privilege is involved. No sysctl, boot option, or rare debug-only kernel mode is required by the supplied evidence. The trigger appears repeatable through connection lifecycle operations, not dependent on an externally uncontrollable race.

4. Primitive and severity interpretation

The exact primitive is a missing release of response-associated resources during RDMA queue teardown. A response can remain marked in rsp_tags while its send completion is never delivered, so nvmet_rdma_send_done and nvmet_rdma_release_rsp are not reached. The leaked objects include RDMA read/write context and request SGL allocations.

This is not an active structured UAF and not fixed stale storage being used after free. It is a persistent kernel resource leak with potential resource exhaustion. There is no concrete repeated release/refcount operation, because the bug is failure to release resources rather than releasing the same object twice. The demonstrated consequence is memory leakage and possible DoS after repeated cycles. A stronger consequence such as privilege escalation, arbitrary write, or information disclosure is speculative and unsupported by the supplied patch context.

PrimitiveClass is CLASS_C because the demonstrated behavior is DoS/resource-exhaustion oriented without an independently useful corruption primitive. ImportantCandidate is NO because no Class A or strong unresolved Class B primitive is established. AutoDowngradeAllowed is YES because the issue is a resource leak without UAF, authorization bypass, controlled corruption, or unresolved lifetime/refcount mutation.

5. One-sentence report phrase

An NVMe/RDMA target can leak RDMA read/write context and request SGL resources when a queue is torn down with I/O in flight, allowing a fabric-reachable initiator to cause gradual memory exhaustion and denial of service through repeated disconnect or teardown cycles.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Manual review is recommended because the issue is reachable over an NVMe/RDMA storage fabric and may be operationally relevant in exposed storage networks, but the supplied evidence supports resource-leak DoS only and does not establish memory corruption or privilege escalation.

7. Call-site confidence

Call-site confidence: medium

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

Patch: nvmet-rdma: fix response resource leak on queue teardown [ Upstream
Commit: 91cfa74f1f3772e8c3662c064a45d8633e00c78e
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=91cfa74f1f3772e8c3662c064a45d8633e00c78e

Changed files:
  drivers/nvme/target/rdma.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=91cfa74f1f3772e8c3662c064a45d8633e00c78e

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

The original Linux kernel CVE announcement for CVE-2026-93151 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93151

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:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:L
  The Best / paranoid CVSS vector: AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H
  The Best / paranoid CVSS score: 5.3

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: 3
  Paranoid ActionableScore: 4

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: NO
Published priority for this report (same as in Subject): MODERATE REGULAR

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).