[CVE-2026-72499][MODERATE 7.0] RDMA/bnxt_re: Free CQ toggle page after firmware teardown [ Upstream

[CVE-2026-72499][MODERATE 7.0] RDMA/bnxt_re: Free CQ toggle page after firmware teardown [ Upstream

[AL-KERNEL]

CVE: CVE-2026-72499
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: RDMA/bnxt_re: Free CQ toggle page after firmware teardown [ Upstream
Commit: b193854675ecad43b4d69c304c1b6a90b206cc99
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=b193854675ecad43b4d69c304c1b6a90b206cc99
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72499
Analysis date: Sat, 15 Aug 2026 13:35:40 -0400
ActionableScore: 8
ActionableScore lower bound: 6
Actionable bucket: Strong Important candidate / Actionable Moderate at minimum
Manual review required: YES

Summary:
RDMA bnxt_re frees a CQ toggle page before firmware CQ teardown is complete, allowing an NQ interrupt to write a toggle value into an already freed page and creating a local race dependent kernel memory corruption risk.

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

The original Linux kernel CVE announcement for CVE-2026-72499 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72499

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-2026-72499	MODERATE	CHECK WITH IMPACT FROM ORIG NN IMPORTANT	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-664;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'RDMA bnxt_re can hit a use-after-free style write because the CQ toggle page was freed before firmware CQ teardown completed. An NQ interrupt arriving during bnxt_qplib_destroy_cq can still write the toggle value into the already freed page, so a freed page could be corrupted or the kernel could crash. For the CVSS the PR:L is used because a local user or service with access to RDMA uverbs device nodes may be able to create and destroy CQs without full host administrator privileges. The issue is not network reachable in the generic CVSS sense and requires local access to the RDMA device interface, although the affected hardware is commonly used for RoCE or data center RDMA workloads. Impact is at least local denial of service and the freed page write makes limited confidentiality or integrity impact defensible for base scoring. For the paranoid score, choose the highest still-defensible interpretation supported by the bug class and patch context, with preference for manual-review sensitivity over autoclosed false negatives.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate / Actionable Moderate at minimum (with actual score 8)	YES	UAF NETWORK KPANIC  DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFKP	YES	NO	checked

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

ActionableScore=8
ActionableScoreLower=6

## 1. ActionableScore

* Conservative score: **6**
* Paranoid score: **8**
* Final recommended bucket: **Strong Important candidate / Actionable Moderate at minimum**::

## 2. Signal breakdown

Conservative signals:

* **Firmware-mediated external influence: +1**. The bad write is caused by an NQ interrupt arriving while firmware teardown is still in progress.
* **Local unprivileged trigger: +1**. A local user or service with access to RDMA uverbs device nodes may be able to create and destroy CQs without full host root.
* **Weak LPE concern: +1**. The bug is a freed-page write, but the written value appears constrained and no controlled reclaim or object replacement is demonstrated.
* **Constrained / indirect memory corruption candidate: +1**. The interrupt path may write a toggle value into a page already returned to the allocator.
* **Real lifetime corruption: +1**. This is a clear lifetime ordering bug: `free_page()` occurs before firmware can no longer write to the page.
* **Privileged kernel/device-management memory corruption path: +1**. The corruption happens in an RDMA driver teardown path involving firmware and interrupt handling.
* **Availability impact realistic: +1**. A write into freed memory can plausibly corrupt kernel state or crash the host.
* **Hard or unreliable race / special timing required: -1**. The issue requires an NQ interrupt during the CQ destruction window.
* **Hardware / configuration dependent exposure: -1**. The issue requires affected Broadcom RDMA hardware and the CQ toggle page mode.

Paranoid additions:

* **Memory corruption, strong corruption primitive: +2 instead of +1**. In the paranoid interpretation, a write-after-free into a page allocator object is treated as strong memory corruption.
* **Strong LPE plausibility: +2 instead of +1**. If a local RDMA user can influence timing and page reuse, privilege escalation is technically defensible enough for manual review, though not demonstrated.

Not awarded:

* **Remote reachable / network-triggerable: +0**. RoCE/RDMA hardware is network-related, but this bug is not directly triggered by packets in the generic CVSS/network sense.
* **Local high-control kernel data-plane API: +0**. RDMA uverbs object creation alone is not enough for this bonus, although the patch does show a lifetime bug.
* **Dirty-Pipe-like / page-cache ownership violation: +0**. No shared file-backed page or COW ownership bypass is shown.

## 3. Reachability analysis

The realistic trigger is local access to the RDMA device interface, especially the ability to create and destroy completion queues using the bnxt_re driver. This does not normally require public network reachability, but it may be available to non-root users or service accounts on systems configured for RDMA workloads.

Namespaces and containers can matter if RDMA device nodes are delegated into containers or assigned to workloads. In a strictly locked-down environment where only trusted administrators can access `/dev/infiniband/uverbs*`, practical exploitability is lower. In HPC, storage, virtualization, or data-center RDMA environments, delegated device access is common enough that PR:L is a reasonable actionable assumption.

Call-site confidence: **high**, because the patch directly shows the relevant ordering in `bnxt_re_destroy_cq()`: `free_page()` was before `bnxt_qplib_destroy_cq()`, and the commit states that an NQ interrupt may write to the freed page.

## 4. Severity interpretation

This behaves more like an **Actionable Moderate at minimum**, with a **Strong Important candidate** paranoid interpretation.

The realistic evidence supports a local, race-dependent freed-page write with constrained payload and firmware/interrupt timing requirements. That is stronger than ordinary DoS-only because the bug is not merely a NULL dereference or WARN. It is a lifetime violation where memory is freed while another execution context can still write to it.

The conservative interpretation should avoid claiming reliable privilege escalation because no controlled reclaim, object replacement, or arbitrary write is shown. The paranoid interpretation is still defensible because kernel write-after-free in a device driver teardown path can become security-relevant depending on allocator reuse and RDMA device access policy.

## 5. One-sentence report phrase

RDMA bnxt_re frees a CQ toggle page before firmware CQ teardown is complete, allowing an NQ interrupt to write a toggle value into an already freed page and creating a local race dependent kernel memory corruption risk.

## 6. Manual review recommendation

**MANUAL CHECK REQUIRED**

This should not be auto-closed because the patch fixes a real freed-page write in a kernel RDMA driver. Even though the race and payload are constrained, the combination of local RDMA user access, firmware-mediated interrupt behavior, and write-after-free semantics warrants manual security review.

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

Patch: RDMA/bnxt_re: Free CQ toggle page after firmware teardown [ Upstream
Commit: b193854675ecad43b4d69c304c1b6a90b206cc99
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=b193854675ecad43b4d69c304c1b6a90b206cc99

Changed files:
  drivers/infiniband/hw/bnxt_re/ib_verbs.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=b193854675ecad43b4d69c304c1b6a90b206cc99

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

The original Linux kernel CVE announcement for CVE-2026-72499 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72499

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: 6
  Paranoid ActionableScore: 8

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).