[CVE-2026-74394][IMPORTANT] RDMA/srpt: fix integer overflow in immediate data length check [ Upstream

[CVE-2026-74394][IMPORTANT] RDMA/srpt: fix integer overflow in immediate data length check [ Upstream

[AL-KERNEL]

CVE: CVE-2026-74394
Priority: IMPORTANT
AL-KERNEL base severity: IMPORTANT
KPANIC flag: NO
Patch: RDMA/srpt: fix integer overflow in immediate data length check [ Upstream
Commit: c82c860f8c8e4f4f454c9f14d0ad0c0466965f7d
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=c82c860f8c8e4f4f454c9f14d0ad0c0466965f7d
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74394
Analysis date: Sat, 15 Aug 2026 06:02:07 -0400
ActionableScore: 8
ActionableScore lower bound: 7
Actionable bucket: Strong Important candidate / Actionable Moderate at minimum
Manual review required: YES

Summary:
A network-adjacent integer overflow in RDMA srpt immediate-data length validation can let a malicious SRP initiator bypass request-size checks and pass a huge length to `sg_init_one()`, causing kernel crash and possible kernel memory exposure through later storage I/O handling.

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

The original Linux kernel CVE announcement for CVE-2026-74394 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74394

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: IMPORTANT
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-2026-74394	IMPORTANT	CHECK WITH IMPACT FROM ORIG NN IMPORTANT	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:L/A:H';CWE-190;CWE-125;*CWE-20;Other CVSS 'AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:H';BEST CVSS score: '8.3';DESCR 'RDMA srpt accepts an immediate data length from the remote initiator and used it in an addition with imm_data_offset without overflow checking. A malicious initiator can make the computed request size wrap to a small value, bypass the request size check, and then cause a very large length to be passed to sg_init_one(). This can make the target treat memory beyond the received SRP command buffer as immediate data, which can crash the kernel and may expose kernel memory through later I/O handling. For the CVSS the PR:N is used because no local user account on the target is required beyond reachability to the RDMA SRP target and the ability to establish the protocol session. The attack vector is AV:A because SRP over RDMA is normally reachable only on an adjacent storage fabric or data center network rather than the public Internet. Impact is at least denial of service and the out-of-bounds read primitive makes confidentiality impact plausible, so this should not be treated as a low severity resource issue.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate / Actionable Moderate at minimum (with actual score 8)	YES	OOB SIMPLEFIX NETWORK LINUS 	NO	NO	checked

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

ActionableScore=8
ActionableScoreLower=7

## 1. ActionableScore

* Conservative score: 7
* Paranoid score: 8
* Final recommended bucket: **Strong Important candidate / Actionable Moderate at minimum**::

## 2. Signal breakdown

Triggered signals:

* Remote reachable / network-triggerable: +2
  The length field is controlled by a remote SRP initiator over RDMA/SRPT. This is not public Internet remote in the usual case, but it is network-triggerable from an RDMA storage fabric.

* Memory corruption, weak/indirect corruption candidate: +1
  The overflow lets a wrapped small `req_size` bypass validation and then passes an attacker-controlled huge length to `sg_init_one()`. This creates an invalid scatterlist length over memory beyond the received SRP command buffer.

* Memory layout invariant restoration: +1
  The patch restores consistency between `imm_data_offset`, `len`, and `req_size` before building the immediate-data scatterlist.

* Reliable kernel crash / strong DoS: +1
  A roughly 4GB scatterlist length from a small command buffer is a credible crash or severe target failure condition.

* Confidentiality impact plausible: +1
  The invalid scatterlist can make later I/O handling treat memory beyond the received command as data. This supports a plausible out-of-bounds read or kernel memory exposure concern.

* Availability impact realistic: +1
  The trigger is simple once the attacker can connect as a malicious SRP initiator, and the impact can be target-wide kernel crash or storage service failure.

* Important filesystem/storage path: +1
  SRPT is a storage target path. Corruption or disclosure in this path can affect storage I/O integrity and availability.

* Rare AND difficult-to-reach subsystem/configuration: -1
  SRPT over RDMA is usually deployed only in specialized storage or data-center fabrics, not on default general-purpose systems.

Paranoid-only signal:

* Integrity impact plausible: +1
  If the oversized scatterlist is consumed as storage payload or affects target I/O state, limited integrity impact is defensible, although the patch does not show an arbitrary write primitive.

## 3. Reachability analysis

A malicious SRP/RDMA initiator can trigger the bug by sending an immediate-data length such as `0xFFFFFFFF`, causing integer wraparound in the request-size calculation. No local account on the target is required if the attacker can reach and authenticate or otherwise connect to the SRPT target endpoint.

This should be treated as adjacent-network reachable rather than broad Internet remote. RDMA/SRPT is normally limited to trusted storage networks, cluster fabrics, or data-center environments. Namespaces and containers are not the main privilege boundary here because the attacker is external to the target host and interacts through the storage protocol.

The vulnerable path is not default-enabled on typical Linux systems. It matters on systems configured as RDMA SRP targets with immediate data support exposed to initiators.

## 4. Severity interpretation

This is not an ordinary Moderate memory leak or local cleanup bug. The patch fixes a network-controlled integer overflow that directly affects scatterlist construction in a storage target.

Realistically, the strongest demonstrated impact is kernel crash or SRPT target DoS. However, because the oversized scatterlist can reference memory beyond the received request buffer, confidentiality impact is plausible and manual review is warranted. Integrity impact is more speculative but defensible in the paranoid score because the corrupted scatterlist is part of storage I/O handling.

Overall this behaves like a Strong Important candidate in affected RDMA storage deployments, and at minimum an Actionable Moderate that should not be auto-closed.

## 5. One-sentence report phrase

A network-adjacent integer overflow in RDMA srpt immediate-data length validation can let a malicious SRP initiator bypass request-size checks and pass a huge length to `sg_init_one()`, causing kernel crash and possible kernel memory exposure through later storage I/O handling.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED

Reason: this is a network-adjacent storage target bug with attacker-controlled length arithmetic feeding scatterlist construction. The likely impact is DoS, but the possible out-of-bounds memory exposure and storage I/O consequences require human review.

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

Patch: RDMA/srpt: fix integer overflow in immediate data length check [ Upstream
Commit: c82c860f8c8e4f4f454c9f14d0ad0c0466965f7d
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=c82c860f8c8e4f4f454c9f14d0ad0c0466965f7d

Changed files:
  drivers/infiniband/ulp/srpt/ib_srpt.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=c82c860f8c8e4f4f454c9f14d0ad0c0466965f7d

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

The original Linux kernel CVE announcement for CVE-2026-74394 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74394

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:H/I:N/A:H
  The Best / paranoid CVSS vector: AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:L/A:H
  The Best / paranoid CVSS score: 8.3

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: 7
  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: IMPORTANT
KPANIC detected for this report: NO
Published priority for this report (same as in Subject): IMPORTANT

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).