[CVE-2026-52908][IMPORTANT] RDMA: During rereg_mr ensure that REREG_ACCESS is compatible [ Upstream

[CVE-2026-52908][IMPORTANT] RDMA: During rereg_mr ensure that REREG_ACCESS is compatible [ Upstream

[AL-KERNEL]

CVE: CVE-2026-52908
Priority: IMPORTANT
AL-KERNEL base severity: IMPORTANT
KPANIC flag: YES
Patch: RDMA: During rereg_mr ensure that REREG_ACCESS is compatible [ Upstream
Commit: 09dc18894148381d3bfc550083b1236043870dce
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=09dc18894148381d3bfc550083b1236043870dce
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-52908
Analysis date: Fri, 19 Jun 2026 10:58:03 -0400
ActionableScore: 8
ActionableScore lower bound: 6
Actionable bucket: Strong Important candidate / Actionable Moderate at minimum
Manual review required: YES

Summary:
RDMA rereg_mr could change a read-only pinned user memory region to writable RDMA access without revalidating writable pinning, potentially violating GUP and COW protections and allowing unauthorized memory modification from a local RDMA-capable process.

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

The original Linux kernel CVE announcement for CVE-2026-52908 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-52908

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-2026-52908	IMPORTANT	CHECK	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H';CWE-863;CWE-284;CWE-266;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:H';BEST CVSS score: '7.8';DESCR 'RDMA rereg_mr could allow an existing user memory region to change from read only to writable access without rechecking that the underlying umem was pinned as writable. This can violate GUP and COW permission assumptions because the same pinned pages may later be exposed to writable RDMA access even though they were originally pinned read only. For the CVSS the PR:L is used because a local user or process with access to the RDMA uverbs device can request MR reregistration without full administrator privileges in many deployments. The issue is not directly network reachable because the vulnerable operation is a local RDMA control path operation. A remote RDMA peer may only become relevant after a local process exposes the MR and rkey. Impact is at least denial of service or memory corruption through unintended writes and may allow privilege escalation depending on the mapped memory and device configuration.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate / Actionable Moderate at minimum (with actual score 8)	YES	RACE NETWORK KPANIC 	NO	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:

* Local unprivileged trigger: +1
  The vulnerable operation is exposed through RDMA uverbs rereg_mr paths. In many deployments access to RDMA devices is granted to non-root users or service accounts.

* Dirty-Pipe-like ownership/COW violation: +2
  The bug allows an MR originally backed by read-only pinned umem to be changed to writable access without revalidating writable pinning. This can violate GUP/COW assumptions and expose pages to writes that were not authorized as writable.

* Integrity impact plausible: +1
  Writable RDMA access to pages not pinned writable can corrupt memory contents that should have remained read-only or COW-protected.

* Availability impact realistic: +1
  Incorrect RDMA writes into improperly pinned memory can plausibly cause process or kernel instability, data corruption, or crash-like failure modes depending on the mapping and device behavior.

* Privileged device-management / DMA-adjacent permission path: +1
  The affected code controls RDMA memory registration and DMA-visible page permissions, which is a high-value kernel/device boundary.

* Rare/configuration-dependent exposure: -1
  Practical exposure requires RDMA stack availability and access to an RDMA-capable device or rxe-like setup. This is not a universal default desktop/server exposure.

Paranoid additional signal:

* Page-cache or file-backed/COW corruption concern: +2
  If the original read-only umem can reference file-backed or COW-shared pages, later writable RDMA access may become a page ownership or page-cache corruption primitive. This is technically defensible from the RO-to-RW without writable pinning bug class, but depends on mapping details and should be manually verified.

## 3. Reachability analysis

The bug is not directly network-triggerable. The vulnerable operation is a local RDMA control path through MR reregistration. A remote RDMA peer may become relevant only after a local process exposes the MR and rkey, but the unsafe permission transition is initiated locally.

Typical triggering requires access to RDMA uverbs. This is often not full root, because RDMA device access may be delegated to users, HPC jobs, containers, or service accounts. Therefore treating it as PR:L is more realistic than PR:H for triage. Containers or namespaces can matter if RDMA devices are passed through or if rxe is configured, but the exposure is still deployment-dependent.

## 4. Severity interpretation

This is stronger than an ordinary Moderate because it is not just a crash or validation cleanup. The patch prevents a transition from read-only to writable RDMA memory access when the underlying umem was not pinned writable, which is a meaningful memory permission and COW boundary issue.

Realistic exploitation evidence is not equivalent to a demonstrated arbitrary kernel write or reliable LPE chain. However, the primitive is security-relevant because RDMA writes can affect memory outside normal CPU page permission checks. Conservative handling is Actionable Moderate. Paranoid handling is a Strong Important candidate due to possible page ownership or COW bypass impact.

## 5. One-sentence report phrase

RDMA rereg_mr could change a read-only pinned user memory region to writable RDMA access without revalidating writable pinning, potentially violating GUP and COW protections and allowing unauthorized memory modification from a local RDMA-capable process.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED

This needs manual review because the patch affects RDMA memory registration permissions and may create a writable access path to pages that were only pinned read-only. The key question is whether practical mappings can turn this into file-backed page corruption, COW bypass, or privilege escalation rather than only process-local memory corruption.

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

Patch: RDMA: During rereg_mr ensure that REREG_ACCESS is compatible [ Upstream
Commit: 09dc18894148381d3bfc550083b1236043870dce
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=09dc18894148381d3bfc550083b1236043870dce

Changed files:
  drivers/infiniband/core/umem.c
  drivers/infiniband/hw/hns/hns_roce_mr.c
  drivers/infiniband/hw/irdma/verbs.c
  drivers/infiniband/hw/mlx4/mr.c
  drivers/infiniband/hw/mlx5/mr.c
  drivers/infiniband/sw/rxe/rxe_verbs.c
  include/rdma/ib_umem.h

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=09dc18894148381d3bfc550083b1236043870dce

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

The original Linux kernel CVE announcement for CVE-2026-52908 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-52908

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:L/PR:L/UI:N/S:U/C:N/I:H/A:H
  The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
  The Best / paranoid CVSS score: 7.8

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: 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: IMPORTANT
KPANIC detected for this report: YES
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).