[CVE-2026-98323][IMPORTANT] RDMA/siw: Bound fragmented header copies by the remaining length [ Upstream

[CVE-2026-98323][IMPORTANT] RDMA/siw: Bound fragmented header copies by the remaining length [ Upstream

[AL-KERNEL]

CVE: CVE-2026-98323
Priority: IMPORTANT
AL-KERNEL base severity: IMPORTANT
KPANIC flag: NO
Patch: RDMA/siw: Bound fragmented header copies by the remaining length [ Upstream
Commit: 2c6fbcf4bfac0b2b186acc8d91154c0fa24468a7
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=2c6fbcf4bfac0b2b186acc8d91154c0fa24468a7
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98323
Analysis date: Sat, 10 Oct 2026 03:05:53 -0400
ActionableScore: 7
ActionableScore lower bound: unknown
Actionable bucket: Strong Important candidate
Manual review required: YES

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-98323 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98323

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-2026-98323	IMPORTANT	https://www.kernelcve.org/list/?q=CVE-2026-98323 CHECK WITH IMPACT FROM ORIG NN IMPORTANT | AEGIS IMPACT IMPORTANT AEGIS_AS=4 LEGACY=IMPORTANT RESULT=SAME (legacy and Aegis reconciliation agree; DECISION: both IMPORTANT)	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H';*CWE-787;CWE-131;*CWE-20;Other CVSS 'AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '9.8';DESCR 'A memory corruption flaw in the RDMA siw receive path occurs because the copy length for a fragmented extended DDP or RDMAP header is calculated without subtracting all header bytes already received. A second TCP callback can copy past the end of the header and overwrite receive state including fpdu_part_rcvd. A later callback can then use a negative fpdu_part_rcvd value as a destination offset and perform another out of bounds write. For the CVSS the PR:N value is used because no account or local execution is required when an attacker can reach a configured siw iWARP endpoint and send crafted protocol traffic. The base AC:H reflects the need to obtain a suitable header split across TCP callbacks. The paranoid AC:L reflects the possibility that an attacker can reliably control fragmentation through staged TCP writes. Impact is at least remote denial of service through kernel memory corruption and may include confidentiality loss, integrity loss, or code execution if the out of bounds write can be shaped into a stronger primitive. In practice exposure requires RDMA siw to be enabled and reachable, which is more typical in internal RDMA or data center networks than on the public Internet.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate (with actual score 7)	MAYBE	WRITE OOB SIMPLEFIX NETWORK SKB LINUS KPANIC_UNSUPPORTED_REMOVED	-	-	checked

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

ActionableScore=7
PrimitiveClass=CLASS_A
RealUAF=NO
HostAdminRequired=NO
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO
RequiresHostUID0=NO
AuthorizationDecisionBypass=NO
ProtectedResourceImpact=NONE
B4PendingReview=NO
B4PendingConfirmed=NO

1. ActionableScore

* Raw conservative score: 6
* Conservative score: 7
* Raw paranoid score: 6
* Paranoid score: 7
* Applied B5-R Important-preservation calibration: 6 -> 7. All B5-R gates are supported, including a confirmed heap OOB write, accumulated receive offset across callback re-entry, attacker-influenced fragmentation, and later consumption of corrupted state in another OOB write.
* Final recommended bucket: **Strong Important candidate**::

2. Signal breakdown

* +2 Confirmed attacker-influenced heap OOB write. `skb_copy_bits()` copies network-supplied bytes beyond the extended header because the old bound does not subtract `fpdu_part_rcvd`.
* +2 Direct network trigger. Traffic reaching a configured SIW iWARP endpoint enters the vulnerable TCP receive callback without requiring attacker credentials or local execution.
* +1 Concrete multi-stage corruption chain. The initial overwrite can corrupt `fpdu_part_rcvd`, and a later callback consumes its negative value as a copy offset, producing another OOB write.
* +1 Strong DoS. Corruption of receive state and an out-of-bounds destination make a kernel fault or crash realistic.
* Total: 2 + 2 + 1 + 1 = 6.
* B5-R preservation raises the result to 7 because automatic Moderate downgrade would discard a confirmed retry or re-entry accumulated write primitive with subsequent corrupted-state consumption.
* No extra availability point was awarded because it would double-count the same likely crash.
* No confidentiality or integrity impact point was awarded. No protected victim, disclosure path, directed target, credential overwrite, or persistent unauthorized modification is demonstrated.
* Controlled reclaim, arbitrary target selection, arbitrary kernel write, callback control, and LPE are not demonstrated.
* No race penalty was applied. The issue depends on staged TCP callback processing, but the attacker can materially influence fragmentation through staged writes.
* No niche-subsystem penalty was applied merely because SIW is specialized.
* B4-PENDING does not apply because no authorization decision or omitted security guard is involved.

3. Reachability analysis

A remote peer can trigger the vulnerable receive path by sending a fragmented extended DDP or RDMAP header to a configured SIW iWARP endpoint. The attacker controls the transmitted header bytes and can materially influence callback re-entry by dividing transmission across TCP writes.

SIW must already be enabled, associated with a network interface, and reachable. That is prior administrative deployment setup, but the complete post-deployment trigger itself does not require host UID 0, host capabilities, an account, or local code execution. Therefore `HostAdminRequired=NO` and `RequiresHostUID0=NO`.

No namespace capability is required by the remote trigger, but the supplied context does not establish whether all SIW provisioning is namespace-scoped. Therefore `NamespaceScopedPrivilege=UNKNOWN`.

No special physical hardware is required because SIW is a software iWARP implementation. The callback split may affect reliability, but it is not shown to be a narrow uncontrollable race.

4. Primitive and severity interpretation

The exact primitive is an attacker-influenced heap OOB write followed by consumption of corrupted structured receive state. The first oversized copy advances beyond the header and can overwrite `fpdu_part_rcvd`. A subsequent callback then uses the corrupted negative value in destination pointer arithmetic and performs another OOB write.

This is active receive-state corruption, not a read-only OOB access, passive metadata discrepancy, validation-only issue, or UAF. There is no repeated release or refcount operation.

`PrimitiveClass=CLASS_A` is used because the patch directly establishes an attacker-influenced OOB write with network bytes and a second corruption sink driven by the overwritten offset. The destination is not arbitrary, so this does not prove arbitrary kernel write or code execution.

The demonstrated practical consequence is kernel memory corruption with credible host DoS. Confidentiality loss, privilege escalation, and code execution remain possible areas for investigation but are not established.

The B5-R profile is satisfied:

1. A heap OOB write is confirmed.
2. `fpdu_part_rcvd` accumulates across TCP callback re-entry, and the old bound fails to account for it.
3. A network peer can influence the fragmented receive sequence.
4. Corrupted `fpdu_part_rcvd` is later consumed as a copy offset in another kernel write.
5. No harmless containment or host-admin-only complete trigger is demonstrated.

`ImportantCandidate=YES` and `AutoDowngradeAllowed=NO` follow from the concrete multi-stage kernel write primitive. Manual analysis is needed to determine overwrite range, adjacent slab layout, copied-byte control, fault reliability, and whether the second write can reach security-sensitive state.

5. One-sentence report phrase

A remote peer reaching a configured SIW iWARP endpoint can fragment an extended DDP or RDMAP header so that receive state is overwritten and later consumed as a negative copy offset, producing a second heap OOB write and requiring Important-class manual review.

6. Manual review recommendation

MANUAL CHECK REQUIRED

The B5-R retry or re-entry accumulated OOB profile is established, and the second write may provide exploitation leverage beyond the demonstrated crash even though arbitrary write and LPE are not proven.

7. Call-site confidence

Call-site confidence: medium

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

Patch: RDMA/siw: Bound fragmented header copies by the remaining length [ Upstream
Commit: 2c6fbcf4bfac0b2b186acc8d91154c0fa24468a7
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=2c6fbcf4bfac0b2b186acc8d91154c0fa24468a7

Changed files:
  drivers/infiniband/sw/siw/siw_qp_rx.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=2c6fbcf4bfac0b2b186acc8d91154c0fa24468a7

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

The original Linux kernel CVE announcement for CVE-2026-98323 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98323

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:N/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:H
  The Best / paranoid CVSS vector: AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
  The Best / paranoid CVSS score: 9.8

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: unknown
  Paranoid ActionableScore: 7

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).