[CVE-2026-68447][MODERATE REGULAR] drm/amdkfd: clamp v9 CRIU control stack checkpoint copy to BO size

[CVE-2026-68447][MODERATE REGULAR] drm/amdkfd: clamp v9 CRIU control stack checkpoint copy to BO size

[AL-KERNEL]

CVE: CVE-2026-68447
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: drm/amdkfd: clamp v9 CRIU control stack checkpoint copy to BO size
Commit: a0d87beb2660a5098b2b0ecdc1e96810a9074ea9
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=a0d87beb2660a5098b2b0ecdc1e96810a9074ea9
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-68447
Analysis date: Tue, 11 Aug 2026 21:55:23 -0400
ActionableScore: 4
ActionableScore lower bound: 3
Actionable bucket: Borderline Actionable Moderate / manual review required
Manual review required: YES

Summary:
A local out-of-bounds read in the AMD KFD CRIU checkpoint path can copy data past the control stack BO into userspace, causing a kernel or driver memory disclosure when affected GPU checkpoint functionality is reachable.

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

The original Linux kernel CVE announcement for CVE-2026-68447 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-68447

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 REGULAR
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-2026-68447	MODERATE	CHECK WITH IMPACT FROM ORIG NN MODERATE	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:N/A:N';CWE-125;CWE-126;*CWE-200;BEST CVSS score: '5.5';DESCR 'An out-of-bounds read in the AMD KFD CRIU checkpoint path can occur because the v9 MQD control stack copy size is taken from cp_hqd_cntl_stack_size without bounding it to the allocated control stack BO. If this hardware derived size is larger than the queue control stack allocation, memcpy can read past the BO into adjacent GTT memory and copy that data to userspace through the checkpoint output. For the CVSS the PR:L is used because triggering requires local access to the AMD KFD or CRIU checkpoint interface, but such access may be available to non-root GPU compute users on systems where /dev/kfd is exposed to render or video group users. The issue is not network reachable. Impact is a local confidentiality breach due to kernel or driver memory disclosure. No direct integrity impact or privilege escalation primitive is shown by the patch context.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline Actionable Moderate / manual review required (with actual score 4)  SKIP CVE-2026-68447 UNKNOWN SKIP No affected files built, so skip this CVE NO  - - unknown 	MAYBE	LEAK LOG HARDWARE DEBUGFS MEMORY  DECREASED_TO_MODERATEREG_BASED_ON_ACTIONABLESCORELESSTHAN5	-	-	checked

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

ActionableScore=4
ActionableScoreLower=3

## 1. ActionableScore

* Conservative score: 3
* Paranoid score: 4
* Final recommended bucket: **Borderline Actionable Moderate / manual review required**

## 2. Signal breakdown

* Local unprivileged trigger: +1
  Triggering requires local access to AMD KFD / CRIU checkpoint functionality. On GPU compute systems, `/dev/kfd` may be available to non-root users in render/video groups.

* Confidentiality impact plausible: +1
  The patch explicitly describes an out-of-bounds read past the allocated control stack BO into adjacent GTT memory, copied back to userspace.

* Privileged kernel/device-management disclosure path: +1
  The bug is in a GPU kernel driver checkpoint path and leaks data from kernel-managed GPU/GTT memory. This is not just a user-buffer bounds issue.

* Firmware / hardware-mediated size influence: +1 for paranoid only
  The unsafe copy length comes from the hardware-derived `cp_hqd_cntl_stack_size` field. This does not make it remotely reachable, but it supports a higher manual-review score because the kernel trusted a device-provided size without bounding it to the BO allocation.

* Weak corruption / limited read-only OOB: +0
  This is a concrete OOB read and disclosure primitive, but not an OOB write, UAF, double free, refcount bug, or type confusion.

* No LPE bonus: +0
  The patch supports information disclosure, not a direct privilege escalation primitive.

* No remote/network exposure: +0
  The affected path is local GPU/CRIU checkpoint handling, not packet parsing or a network service.

* No rare-config penalty in the main score
  The path is hardware and feature dependent, but still reachable through ordinary local GPU interfaces on affected AMD KFD systems. I would not auto-downgrade it solely as rare hardware.

## 3. Reachability analysis

A local user or local process with access to AMD KFD queue checkpoint functionality can potentially trigger the affected path. In typical desktop or HPC GPU deployments, access to `/dev/kfd` may be delegated to non-root users through GPU-related groups, so PR:L is the practical assumption rather than PR:H.

Namespaces or containers may matter if GPU devices are passed through or exposed into containers. A container user with access to the KFD device node could plausibly reach the same driver path, depending on device cgroup and runtime configuration.

The issue is not network reachable. Exploitation depends on affected AMD GPU generations, KFD enabled, CWSR/checkpoint support, and the relevant v9/v9.4.3 MQD checkpoint path.

Call-site confidence: high. The patch directly shows `memcpy()` using an unbounded hardware-derived size and then replacing it with `min(cp_hqd_cntl_stack_size, mm->ctl_stack_size)`.

## 4. Severity interpretation

This behaves more like an actionable Moderate information disclosure than an Important memory-corruption issue. The realistic impact is local disclosure of adjacent GTT/kernel-driver memory to userspace through checkpoint output.

Theoretical exploit chaining is possible if the leaked data contains useful kernel or driver state, but the patch does not show a write primitive, object lifetime corruption, callback control, type confusion, or direct LPE path. Therefore, Important-class scoring is not justified from this patch alone, but manual review is warranted because the disclosure primitive is concrete.

## 5. One-sentence report phrase

A local out-of-bounds read in the AMD KFD CRIU checkpoint path can copy data past the control stack BO into userspace, causing a kernel or driver memory disclosure when affected GPU checkpoint functionality is reachable.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED

Reason: this is a concrete kernel-driver OOB read with user-visible disclosure, not merely a crash or validation cleanup. It should not be auto-closed even though the current evidence supports confidentiality impact only and no direct privilege escalation primitive.

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

Patch: drm/amdkfd: clamp v9 CRIU control stack checkpoint copy to BO size
Commit: a0d87beb2660a5098b2b0ecdc1e96810a9074ea9
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=a0d87beb2660a5098b2b0ecdc1e96810a9074ea9

Commit description:

CRIU checkpoint copies the MQD control stack using cp_hqd_cntl_stack_size
from hardware without bounding it to the allocated BO region. If the HW
field is larger than the queue's control stack allocation, memcpy reads
past the BO into adjacent GTT memory and can leak kernel data to userspace.

Store the page-aligned control stack BO size in mqd_manager and clamp
checkpoint copies and reported checkpoint sizes to
min(cp_hqd_cntl_stack_size, mm->ctl_stack_size). Apply the same bound
for multi-XCC v9.4.3 checkpoint layout.

Signed-off-by: Yongqiang Sun <Yongqiang.Sun@amd.com>
Reviewed-by: David Francis <David.Francis@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 6c2abd0 )
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.h
  drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.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=a0d87beb2660a5098b2b0ecdc1e96810a9074ea9

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

The original Linux kernel CVE announcement for CVE-2026-68447 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-68447

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: Not available
  The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N
  The Best / paranoid CVSS score: 5.5

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: 4

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 REGULAR

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).