[CVE-2026-89510][MODERATE 7.0] RDMA/cxgb4: Cancel reg_work before freeing device on remove

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

CVE: CVE-2026-89510
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: RDMA/cxgb4: Cancel reg_work before freeing device on remove
Commit: fe9c591026c576d8b1f72aab5e4cd67350530763
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=fe9c591026c576d8b1f72aab5e4cd67350530763
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89510
Analysis date: Sat, 12 Sep 2026 03:46:41 -0400
ActionableScore: 4
ActionableScore lower bound: unknown
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-2026-89510 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89510

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-2026-89510	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-89510 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:H/I:H/A:H';*CWE-416;CWE-362;CWE-667;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '7';DESCR 'A race in the RDMA cxgb4 remove path can free ctx dev while reg_work is still pending or running. The queued registration worker can then enter c4iw_register_device and access the freed device object, which makes this a kernel use after free candidate rather than a simple warning or resource leak. The most direct impact is a local denial of service via kernel crash during device remove, driver teardown, or related error recovery paths. For the CVSS the PR:L is used for the paranoid score because some environments delegate device or network administration to a reduced capability root, container root, or service account rather than full host root. The issue is not directly network reachable from packet traffic and normally requires local control over device lifecycle or a driver state transition. Impact is at least availability loss and in the worst still defensible case may include confidentiality and integrity impact due to kernel memory corruption.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4)	YES	LOCK INIT UAF NETWORK HARDWARE DEBUGFS KPANIC 	NO	NO	checked

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

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

1. ActionableScore

* Raw conservative score: 3
* Conservative score: 4
* Raw paranoid score: 3
* Paranoid score: 4
* Applied conservative floor/cap: Real UAF floor 4 applied, 3 -> 4.
* Applied paranoid floor/cap: Real UAF floor 4 applied, 3 -> 4.
* Final recommended bucket: **Borderline / manual review recommended**::

2. Signal breakdown

Positive signals:

* +2 Real UAF / lifetime violation. `c4iw_remove()` can free `ctx->dev` while queued or running `reg_work` may still use it in `c4iw_register_device()`.
* +1 Reliable kernel crash / strong DoS. Accessing a freed RDMA device object from workqueue context can realistically cause kernel fault or crash.

Negative signals:

* No host-admin penalty applied. Device remove is likely privileged or device-lifecycle driven, but the supplied evidence does not fully prove that host-admin-equivalent privilege is required for every possible trigger.
* No race penalty applied. The bug involves asynchronous work and remove ordering, but the evidence does not prove a narrow or externally uncontrollable timing window beyond the missing synchronization.

Important non-awarded signals:

* Controlled reclaim not demonstrated.
* Arbitrary write not demonstrated.
* Remote reachability not established.
* Local unprivileged trigger not established.
* Concrete repeated release/refcount signal not awarded. The commit mentions registration failure can tear down `ctx->dev`, but the supplied diff does not establish a precise first release followed by a second lifetime mutation on the exact same dangling reference.
* Structured stale-object leverage not awarded. The stale object is structured, but attacker-controlled replacement or security-relevant field consumption after reuse is not established.
* Weak LPE not awarded merely because this is a UAF.
* Extra Availability not awarded because the crash/DoS effect is already counted once.
* Confidentiality and Integrity not awarded because no disclosure, arbitrary write, controlled reclaim, or protected-state modification primitive is established.

Repeated release/refcount checklist:

* First release can end lifetime: not concretely established from the supplied context.
* Exact same reference survives to a second release: not concretely established.
* Second operation is a real lifetime/refcount mutation: possible from the prose, but not sufficiently proven from supplied control flow.
* +1 Concrete repeated release/refcount awarded: no.
* Paranoid repeated-refcount leverage +1 awarded: no.
* FLOOR 5 applied: no.

3. Reachability analysis

The vulnerable path is in the RDMA `cxgb4` driver remove path. `c4iw_uld_state_change()` queues `reg_work`, while `c4iw_remove()` tears down the RDMA device. The complete trigger appears tied to device removal, driver teardown, or ULD state transitions rather than ordinary packet traffic.

Below-host-admin reachability is not established. A normal local user is not shown to be able to trigger the complete remove plus pending registration-work sequence. Host-admin-equivalent privilege is also not fully proven from the supplied snippet alone because the trigger may depend on device lifecycle or driver state transitions outside the visible diff.

Namespace-scoped privilege is not supported. No evidence shows that the complete RDMA device remove and registration work sequence can be performed inside attacker-created namespaces.

Prior host setup exists in the practical sense that affected hardware and the `cxgb4` RDMA driver must be present and active. This is hardware-specific, but the supplied evidence does not justify a rare-hardware penalty by itself. No direct network-triggered packet path is established.

Race repeatability is uncertain. It is an async work versus remove synchronization bug, but the patch does not show that an attacker can repeatedly and cheaply drive both sides of the lifetime transition below host-admin privilege.

4. Primitive and severity interpretation

The exact primitive is a real kernel lifetime violation: `ctx->dev` may be freed by `c4iw_remove()` while `reg_work` is still pending or running and may later access the freed object in `c4iw_register_device()`.

This is an active structured object UAF, not a fixed stale DMA buffer. However, the demonstrated context is still largely driver lifecycle cleanup/remove and asynchronous registration handling. The evidence supports crash/DoS but does not establish controlled object replacement, controlled reclaim, callback control, arbitrary read/write, or a concrete privilege-escalation path.

This is PrimitiveClass=CLASS_B because a real UAF exists. It is not CLASS_A because no strong exploitation primitive is demonstrated. It is not CLASS_C because the patch directly establishes post-lifetime use of a freed kernel object, not merely a NULL dereference or passive invalid state.

Demonstrated consequence: local or lifecycle-triggered kernel crash/DoS.

Speculative consequence: memory corruption could theoretically become more serious if an attacker could control reuse of the freed RDMA device object, but that mechanism is not established.

ImportantCandidate=NO because the supplied evidence shows a constrained driver lifecycle UAF with no demonstrated meaningful reuse, write, callback, disclosure, boundary bypass, or concrete repeated lifetime/refcount mutation.

AutoDowngradeAllowed=YES because this appears to be a constrained Class B cleanup/remove-path UAF with demonstrated crash impact only and no unresolved strong exploitation leverage beyond the ordinary UAF.

Likely operational severity: borderline Moderate. It deserves review because it is a real UAF in kernel driver code, but the provided evidence does not justify treating it as a strong Important candidate.

5. One-sentence report phrase

A race in the RDMA cxgb4 remove path can free `ctx->dev` while registration work is still pending or running, allowing `c4iw_register_device()` to access a freed RDMA device object and potentially crash the kernel.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Reason: this is a real kernel UAF in an asynchronous workqueue teardown path, but current evidence supports primarily crash/DoS and does not establish controlled reclaim, arbitrary R/W, repeated refcount mutation, or below-host-admin reachability.

7. Call-site confidence

Call-site confidence: medium

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

Patch: RDMA/cxgb4: Cancel reg_work before freeing device on remove
Commit: fe9c591026c576d8b1f72aab5e4cd67350530763
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=fe9c591026c576d8b1f72aab5e4cd67350530763

Commit description:

c4iw_uld_state_change() queues reg_work to register the RDMA device.
c4iw_remove() can free ctx->dev while this work is pending or running,
leaving c4iw_register_device() accessing the freed device.

Cancel reg_work before removing the device.  The registration work can
tear down ctx->dev when registration fails, so do not unregister or
deallocate it again in that case.

This issue was found by an in-house static analysis tool.

Fixes: 1c8f1da ("iw_cxgb4: Fix possible circular dependency locking warning")
Link: https://patch.msgid.link/r/20260806130128.465460-1-fanwu01@zju.edu.cn
Cc: stable@vger.kernel.org
Assisted-by: Codex:gpt-5.6
Signed-off-by: Fan Wu <fanwu01@zju.edu.cn>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  drivers/infiniband/hw/cxgb4/device.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=fe9c591026c576d8b1f72aab5e4cd67350530763

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

The original Linux kernel CVE announcement for CVE-2026-89510 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89510

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: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: 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).