[CVE-2026-74377][MODERATE 7.0] RDMA/rxe: Copy WQE to local buffer in non-SRQ receive path [ Upstream

[CVE-2026-74377][MODERATE 7.0] RDMA/rxe: Copy WQE to local buffer in non-SRQ receive path [ Upstream

[AL-KERNEL]

CVE: CVE-2026-74377
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: RDMA/rxe: Copy WQE to local buffer in non-SRQ receive path [ Upstream
Commit: 2e60378fb3c8b51c94103bb40014c4fe38fa5033
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=2e60378fb3c8b51c94103bb40014c4fe38fa5033
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74377
Analysis date: Sun, 16 Aug 2026 08:25:52 -0400
ActionableScore: 5
ActionableScore lower bound: 3
Actionable bucket: Actionable Moderate at minimum
Manual review required: YES

Summary:
A TOCTOU flaw in the RDMA RXE non-SRQ receive path lets a local user modify userspace mapped WQE fields while the kernel responder processes them, causing OOB reads and likely kernel crash, with limited confidentiality or integrity concern requiring manual review.

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

The original Linux kernel CVE announcement for CVE-2026-74377 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74377

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-2026-74377	MODERATE	CHECK WITH IMPACT FROM ORIG NN MODERATE	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-367;CWE-125;*CWE-20;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:H';BEST CVSS score: '5.8';DESCR 'A TOCTOU flaw in the RDMA rxe non SRQ receive path allows a local malicious user to modify a receive WQE in the userspace mapped queue while the kernel responder is processing it. Fields such as num_sge or SGE entries can change between validation and use, which can cause out of bounds reads in rxe_resp_check_length() and copy_data(). For the CVSS the PR:L is used because triggering requires local access to RDMA rxe user objects and the shared receive queue, but does not normally require full administrative privileges. AC:H is used because reliable triggering depends on a race window while the kernel processes the WQE. The issue is not directly network reachable by itself, although RDMA traffic may be used to drive responder processing after a local user has prepared the QP. Impact is at least local denial of service via kernel crash and in the paranoid case may allow limited confidentiality or integrity impact due to kernel OOB reads and unstable descriptor processing. YES REQUIRES MANUAL CHECK';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate at minimum (with actual score 4)	YES	OOB DANGER NETWORK  DECREASED_TO_MODERATEREG_BASED_ON_ACTIONABLESCORELESSTHAN5	NO	NO	checked

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

ActionableScore=5
ActionableScoreLower=3

## 1. ActionableScore

* Conservative score: 3
* Paranoid score: 5
* Final recommended bucket: **Actionable Moderate at minimum**

## 2. Signal breakdown

Conservative signals:

* Local unprivileged trigger: +1. A local malicious user with access to RDMA RXE user objects and the userspace mapped receive queue can modify WQE fields.
* Memory corruption, weak/indirect corruption candidate: +1. The patch describes attacker-controlled TOCTOU changes to `num_sge` and SGE entries causing kernel out-of-bounds reads, but not an OOB write or proven reclaim primitive.
* Reliable kernel crash / strong DoS: +1. OOB reads in `rxe_resp_check_length()` and `copy_data()` can plausibly crash the kernel.
* Hard or unreliable race / special timing required: -1. Exploitation depends on modifying the WQE while the responder is processing it.
* Confidentiality impact plausible: +1. Kernel OOB reads from adjacent WQE or descriptor data may expose limited kernel or queue memory depending on call path and error handling.

Paranoid additional signals:

* Integrity impact plausible: +1. Unstable SGE descriptor processing may influence which descriptors are consumed by the responder, though the patch does not prove an arbitrary write primitive.
* RDMA device-object API with exploitation-relevant shared descriptor control: +1. This is not counted merely because RDMA objects exist. It is counted because the patch shows userspace can modify shared WQE/SGE descriptors while kernel code consumes them.

Not counted:

* Remote reachable: +0. RDMA traffic may drive responder processing, but the vulnerable setup requires local control of the QP and receive queue.
* Strong LPE plausibility: +0. No attacker-controlled reclaim, callback control, arbitrary write, type confusion, or refcount takeover is shown.
* Generic strong memory corruption: +0. The described primitive is OOB read, not OOB write/UAF/double-free.

Call-site confidence: high. The patch shows the affected non-SRQ receive path and the exact replacement with validated kernel-local WQE copying.

## 3. Reachability analysis

The realistic trigger is local access to RXE or RDMA userspace objects, typically through `/dev/infiniband/uverbs*` when an RXE device is configured and accessible to the user or group. Full host root is not inherently required once the device is available, so PR:L is the practical model. Namespaces or containers may matter if RDMA device nodes are passed through or if a container has access to uverbs devices. The issue is not directly network-triggerable by an unauthenticated remote peer because the attacker must control the local userspace mapped receive queue.

RXE Soft RoCE is not universally enabled by default, but it is reachable through ordinary RDMA userspace interfaces when configured. The race requirement reduces reliability, but the attacker has relatively high control over the shared WQE fields and can repeatedly drive responder processing.

## 4. Severity interpretation

This is stronger than an ordinary Moderate because it combines a local user controlled shared descriptor race with kernel OOB reads in an RDMA data path. The realistic demonstrated impact is local DoS through kernel crash. The theoretical impact includes limited confidentiality or integrity concern due to unstable SGE descriptor processing, but the patch does not support treating it as a strong privilege escalation primitive.

Overall this should be treated as Actionable Moderate at minimum, not auto-closed. It is not a Strong Important candidate unless further analysis shows an OOB write, attacker-controlled object replacement, or a reliable LPE chain.

## 5. One-sentence report phrase

A TOCTOU flaw in the RDMA RXE non-SRQ receive path lets a local user modify userspace mapped WQE fields while the kernel responder processes them, causing OOB reads and likely kernel crash, with limited confidentiality or integrity concern requiring manual review.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED

Reason: the bug is a race involving attacker-controlled RDMA WQE/SGE descriptors and kernel OOB reads. The current evidence supports DoS and limited C/I concern, but manual review is needed to determine whether `copy_data()` or related RXE paths can turn unstable SGE interpretation into a stronger write or privilege escalation primitive.

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

Patch: RDMA/rxe: Copy WQE to local buffer in non-SRQ receive path [ Upstream
Commit: 2e60378fb3c8b51c94103bb40014c4fe38fa5033
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=2e60378fb3c8b51c94103bb40014c4fe38fa5033

Changed files:
  drivers/infiniband/sw/rxe/rxe_resp.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=2e60378fb3c8b51c94103bb40014c4fe38fa5033

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

The original Linux kernel CVE announcement for CVE-2026-74377 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74377

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: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: 3
  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: NO
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).