[CVE-2026-89968][MODERATE 7.0] nvmet-tcp: reject unsolicited H2CData PDUs

[CVE-2026-89968][MODERATE 7.0] nvmet-tcp: reject unsolicited H2CData PDUs

[AL-KERNEL]

CVE: CVE-2026-89968
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: nvmet-tcp: reject unsolicited H2CData PDUs
Commit: 02341ee424fb2eed16b5b8a9d247492e49335c21
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=02341ee424fb2eed16b5b8a9d247492e49335c21
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89968
Analysis date: Wed, 16 Sep 2026 14:58:26 -0400
ActionableScore: 4
ActionableScore lower bound: unknown
Actionable bucket: Borderline / manual review recommended
Manual review required: YES

Summary:
An unauthenticated remote NVMe/TCP host can send unsolicited H2CData before R2T, causing double command completion and corrupting response list state, which can crash the kernel or wedge the nvmet_tcp workqueue.

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

The original Linux kernel CVE announcement for CVE-2026-89968 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89968

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-2026-89968	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-89968 CHECK WITH IMPACT FROM ORIG NN MODERATE	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:A/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:H';*CWE-20;CWE-841;*CWE-835;*CWE-672;Other CVSS 'AV:A/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '7.6';DESCR 'An unsolicited H2CData PDU in the NVMe/TCP target path can be accepted before the target has sent an R2T for the command. A remote host can submit a write command that takes the R2T path and then send H2CData for the same tag before R2T is transmitted, causing early command completion and a later second completion on an error path. This can add the same command to the response list twice and create a self referential list node, resulting in a kernel BUG with CONFIG_DEBUG_LIST or a wedged nvmet_tcp workqueue and soft lockup without it. For the CVSS the PR:N is used for the paranoid score because an allow_any_host subsystem can be reached without authentication once the attacker has access to the NVMe/TCP target network. The issue is network reachable inside the storage fabric, but NVMe/TCP targets are usually deployed on isolated storage or data center networks rather than the public Internet. Impact is at least denial of service via kernel crash or workqueue lockup and in worst case may allow limited confidentiality or integrity impact due to kernel list state corruption and should be reviewed manually.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4)	YES	REMOTE DEADLOCK DANGER NVME NETWORK LINUS KPANIC  KPANIC INCREASED_TO_MODERATE7_BASED_ON_GUESSCVSS	NO	NO	checked

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

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

## 1. ActionableScore

* Raw conservative score: 4
* Conservative score: 4. No floor or cap applied.
* Raw paranoid score: 4
* Paranoid score: 4. No additional paranoid point awarded because no useful primitive beyond crash or workqueue wedge is established.
* Final recommended bucket: **Borderline / manual review recommended**::

## 2. Signal breakdown

Positive signals:

* +2 Remote reachable / directly network-triggerable: a remote NVMe/TCP host can send an unsolicited H2CData PDU before R2T is sent.
* +1 Reliable kernel crash / strong DoS: CONFIG_DEBUG_LIST can trigger a kernel BUG, and without it the nvmet_tcp workqueue can wedge in a soft lockup.
* +1 Weak / indirect corruption candidate: the same command can be added twice to response list state, creating a self-referential llist node. This is kernel list state corruption, but the demonstrated consequence is DoS.

Negative signals:

* No negative privilege penalty: host-admin-equivalent privilege is not required for the remote host when an allow_any_host subsystem is reachable.
* No rare-subsystem penalty: NVMe/TCP targets are often isolated storage-fabric services, but isolation is deployment-dependent and not a concrete scoring penalty.

Important non-awarded signals:

* RealUAF not awarded: no object free followed by stale use is established.
* Controlled reclaim not demonstrated.
* Arbitrary write not demonstrated.
* Confidentiality and integrity not awarded: list state corruption is shown, but no disclosure, protected-object modification, or attacker-directed write primitive is established.
* Strong LPE / strong exploitation leverage not awarded: no controlled object replacement, callback control, refcount takeover, or privilege escalation path is shown.
* Extra Availability point not awarded: the BUG and workqueue wedge are the same availability consequence already counted as strong DoS.
* Concrete repeated release/refcount not awarded: the bug is double completion and double list insertion, not a second put/release/refcount mutation on a potentially freed object.
* FLOOR 5 not applied: no UAF, no repeated lifetime mutation, and no meaningful exploitation surface beyond crash/list wedge is established.

## 3. Reachability analysis

The evaluated attacker is a remote NVMe/TCP host that can connect to the target subsystem. The supplied text explicitly states that the issue is remotely triggerable and needs no authentication on an allow_any_host subsystem.

HostAdminRequired=NO because the complete trigger does not require host administrative privileges on the victim once the NVMe/TCP target is exposed to the attacker. NamespaceScopedPrivilege=NO because the trigger is network protocol access, not a namespace-scoped local capability path.

There is a prior deployment prerequisite: the victim must run an NVMe/TCP target and expose a subsystem reachable by the attacker. In practice this is commonly an isolated storage or data-center network rather than the public Internet, but that does not remove remote reachability for triage.

No timing race is required. The attack is an ordered protocol-state violation: submit a write command that takes the R2T path, send H2CData before R2T, and reach double completion on the error path.

## 4. Primitive and severity interpretation

The exact primitive is unsolicited H2CData acceptance before the target has sent R2T. This can cause early command completion and later second completion, resulting in the same command being queued twice on response list state.

This is not a real UAF and not a repeated release/refcount bug. It is active kernel protocol/list state corruption, but the demonstrated outcome is kernel BUG or workqueue soft lockup. After excluding the crash/wedge, no independently useful attacker-controlled corruption primitive is established.

PrimitiveClass=CLASS_C because the behavior is primarily remote DoS through invalid protocol state and list corruption without demonstrated confidentiality, integrity, UAF, arbitrary write, controlled reclaim, or privilege-escalation leverage.

ImportantCandidate=NO because the evidence does not show Important-grade exploitation leverage beyond DoS. ManualReview=RECOMMENDED because the issue is remotely reachable and unauthenticated in allow_any_host configurations. AutoDowngradeAllowed=YES because the established primitive remains crash/DoS-only and no unresolved Class A or strong Class B primitive is present.

Likely operational severity: actionable as a remote storage-fabric DoS, but not an Important-class kernel corruption candidate based on supplied evidence.

## 5. One-sentence report phrase

An unauthenticated remote NVMe/TCP host can send unsolicited H2CData before R2T, causing double command completion and corrupting response list state, which can crash the kernel or wedge the nvmet_tcp workqueue.

## 6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Reason: the issue is remotely reachable and unauthenticated in allow_any_host NVMe/TCP deployments, but the demonstrated primitive is DoS-only list-state corruption without UAF, controlled reclaim, arbitrary write, or privilege-escalation evidence.

## 7. Call-site confidence

Call-site confidence: high

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

Patch: nvmet-tcp: reject unsolicited H2CData PDUs
Commit: 02341ee424fb2eed16b5b8a9d247492e49335c21
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=02341ee424fb2eed16b5b8a9d247492e49335c21

Commit description:

nvmet_tcp_handle_h2c_data_pdu() accepts an H2CData PDU after only checking
that its TTAG is a valid in-range command index and that the command's
data buffers are mapped. It never checks that the target has actually
solicited that data by sending an R2T for the command.

A remote host can abuse this. It submits a write command that takes the
R2T path and, before the target transmits the R2T, sends an H2CData PDU
for that command's tag. The data completes the command early, and when
the command then fails synchronously (e.g. a length mismatch caught by
nvmet_check_transfer_len()), it is completed a second time. Each
completion calls nvmet_tcp_queue_response(), so the same command is added
to queue->resp_list twice while it is still linked; the second llist_add()
makes the node point to itself (lentry->next == lentry).

nvmet_tcp_process_resp_list() then walks that self-referential node and
adds the command to resp_send_list twice. With CONFIG_DEBUG_LIST this
trips the "list_add double add" check (kernel BUG); without it the loop
never terminates and the nvmet_tcp workqueue wedges (soft-lockup). It is
remotely triggerable and needs no authentication on an allow_any_host
subsystem.

Track whether an R2T has been transmitted for a command and reject an
H2CData PDU that arrives before it. The flag is cleared on command reuse
(nvmet_tcp_get_cmd() zeroes cmd->flags) and stays set across the multiple
H2CData PDUs of a single solicited transfer.

Fixes: 872d26a ("nvmet-tcp: add NVMe over TCP target driver")
Cc: stable@vger.kernel.org
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Shivam Kumar <kumar.shivam43666@gmail.com>
Signed-off-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  drivers/nvme/target/tcp.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=02341ee424fb2eed16b5b8a9d247492e49335c21

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

The original Linux kernel CVE announcement for CVE-2026-89968 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89968

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:L/PR:N/UI:N/S:U/C:N/I:N/A:H
  The Best / paranoid CVSS vector: AV:A/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:H
  The Best / paranoid CVSS score: 7.6

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: unknown
  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: 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).