[CVE-2026-89810][MODERATE 7.0] drm/amdkfd: Fix error path at svm_migrate_copy_to_ram

[AL-KERNEL <alkernel@kernelcve.org>]

CVE: CVE-2026-89810
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: drm/amdkfd: Fix error path at svm_migrate_copy_to_ram
Commit: da87bcad1f781d822e7ced6d1de9dbc6d381c72e
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=da87bcad1f781d822e7ced6d1de9dbc6d381c72e
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89810
Analysis date: Wed, 16 Sep 2026 10:54:26 -0400
ActionableScore: 6
ActionableScore lower bound: 5
Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review
Manual review required: YES

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-89810 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89810

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-2026-89810	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-89810 CHECK WITH IMPACT FROM ORIG NN LOW	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H';CWE-763;CWE-404;*CWE-416;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'svm_migrate_copy_to_ram can release and unlock the wrong system RAM page on an error path because the old code derived struct page from a DMA address instead of the real page identity. With IOMMU enabled, or in general when DMA addresses differ from physical PFNs, this can corrupt page reference state, leave the real migration page locked or leaked, and create a use-after-free candidate. For the CVSS the PR:L is used because a local user with access to AMD KFD GPU compute or the render device may be able to trigger SVM migration paths without full administrator privileges. AC:H is used because reliable triggering depends on reaching a migration failure path such as dma_map_page() or copy failure. The issue is not network reachable. Impact is at least local denial of service through page state corruption or resource leakage, and in the paranoid case may include confidentiality and integrity impact due to wrong page release and possible memory corruption.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 6)  SKIP CVE-2026-89810 UNKNOWN SKIP No affected files built, so skip this CVE NO  - - unknown 	MAYBE	OOB ERRORPATH DMAorINTERRUPT HARDWARE KPANIC MEMORY  KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFKP	-	-	checked

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

ActionableScore=6
ActionableScoreLower=5
PrimitiveClass=CLASS_B
RealUAF=NO
HostAdminRequired=NO
NamespaceScopedPrivilege=NO
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO

1. ActionableScore

* Raw conservative score: 5
* Conservative score: 5. No floor or cap applied.
* Raw paranoid score: 6
* Paranoid score: 6. No floor or cap applied.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

2. Signal breakdown

Conservative positive signals:

* +1 Local below-host-admin reachability. AMD KFD SVM migration paths are plausibly reachable by a local GPU compute user or render-device user without full host administrator privileges.
* +1 Weak or indirect corruption candidate. The old code converted a DMA address into a PFN and then into struct page, which can select the wrong system RAM page when DMA addresses differ from CPU physical addresses.
* +1 Privileged kernel/device-management memory corruption path. The affected state is page lifetime and lock state, including unlock_page and put_page on a page that may not be the allocated migration page.
* +1 Reliable kernel crash or strong DoS. Wrong page unlock/refcount mutation, leaked locked pages, or leaked DMA mappings can plausibly lead to kernel faults, hangs, or resource exhaustion.
* +1 Concrete persistent kernel resource/reference leak. The patch adds dma_unmap_page and releases the real migration pages, showing that previous error cleanup could lose DMA mappings and page references.

Paranoid additional signal:

* +1 Weak LPE concern. The wrong-page put_page can decrement lifetime state of an unrelated page, which is a plausible page refcount corruption and UAF candidate even though controlled reclaim or arbitrary write is not demonstrated.

Material non-awarded signals:

* +2 Strong exploitation leverage not awarded. Controlled target selection, controlled reclaim, arbitrary kernel read/write, callback control, or refcount takeover is not established.
* RealUAF=YES not set. A UAF candidate is plausible from wrong page refcount corruption, but the supplied patch does not directly establish a concrete stale post-lifetime dereference.
* Concrete repeated release/refcount signal not awarded. This is not a first put followed by the exact same dangling reference reaching a second put. It is a wrong-object lifetime mutation.
* Extra Availability not awarded. Crash/hang/resource impact is already counted once.
* Direct Confidentiality and Integrity signals not awarded. C/I impact is plausible only as unresolved consequence of page lifetime corruption, not as a demonstrated disclosure or unauthorized write primitive.
* Remote reachability not awarded. This is a local DRM/KFD GPU memory-management path, not a network path.
* Hardware rarity penalty not applied. AMD GPU hardware is hardware-specific, but the prompt does not support treating it as rare or physical-only.

3. Reachability analysis

The most relevant attacker is a local user or process with access to AMD KFD GPU compute or the DRM render device. Such access is commonly below full host administrator privilege, so HostAdminRequired=NO.

This is not a namespace-scoped capability case. The vulnerable operation depends on local GPU device access and AMDGPU/KFD SVM migration behavior, not on attacker-created namespaces. NamespaceScopedPrivilege=NO.

The trigger requires entering the svm_migrate_copy_to_ram error path, for example a DMA mapping failure or copy failure during migration from device memory to system RAM. That makes exploitation less straightforward than a normal successful migration path, but it is an ordered error-path condition rather than a race. No race penalty is applied.

Prior host setup may exist in the practical sense that AMDGPU/KFD must be present and the user must have device access, but the supplied evidence does not establish a host-admin-only trigger. No physical-only or rare-hardware penalty is applied.

4. Primitive and severity interpretation

The exact primitive is wrong page lifetime cleanup: the old error path used a DMA address as if it were a CPU physical address, converted it to struct page, then called unlock_page and put_page on that page. With IOMMU enabled, and even generally on systems where DMA addresses differ from PFNs, this can operate on the wrong page.

This is an active kernel page-lifetime and DMA ownership corruption candidate, not fixed stale storage. It is also not merely a passive leak, because put_page mutates page lifetime state and unlock_page mutates page lock state.

A concrete repeated release/refcount operation on the same dangling reference is not established. The stronger issue is wrong-object release/refcount mutation, not double put of the exact same reference.

PrimitiveClass=CLASS_B because there is a real kernel lifetime/ownership corruption primitive, but not enough evidence for CLASS_A. Demonstrated or directly supported consequences include local DoS through page state corruption, locked-page leaks, DMA mapping leaks, or kernel crash. Speculative consequences include UAF and possible privilege escalation if the wrong page refcount mutation can be steered into useful page reuse, but that chain is not demonstrated.

ImportantCandidate=YES because wrong-page put_page against system RAM page state leaves meaningful exploitation leverage unresolved and should not be automatically downgraded as a simple error-path leak.

AutoDowngradeAllowed=NO because automatic closure would discard a concrete page lifetime corruption primitive with plausible UAF implications.

Likely operational severity is actionable Moderate, with Strong Important candidate review warranted.

5. One-sentence report phrase

A local AMD KFD GPU user may trigger an error path in svm_migrate_copy_to_ram where a DMA address is incorrectly treated as a page PFN, causing wrong page unlock or put_page operations, DMA mapping leaks, and possible page lifetime corruption leading at least to DoS and requiring manual review for UAF or privilege escalation potential.

6. Manual review recommendation

MANUAL CHECK REQUIRED

Wrong-page put_page and unlock_page on system RAM page state is a concrete kernel lifetime/ownership corruption primitive. Controlled exploitation is not proven, but the page refcount corruption and UAF candidate are strong enough to prevent automatic downgrade.

7. Call-site confidence

Call-site confidence: medium

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

Patch: drm/amdkfd: Fix error path at svm_migrate_copy_to_ram
Commit: da87bcad1f781d822e7ced6d1de9dbc6d381c72e
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=da87bcad1f781d822e7ced6d1de9dbc6d381c72e

Commit description:

If page migration from device to sys ram fails for some reasons driver needs
release and unlock allocated system pages. To do that driver should use page
physical address, or pfn, then get struct page*. Current driver uses dma
address(for adev) that is not correct with IOMMU enabled, or even in general.

The patch releases and unlocks allocated system pages based on where migration
failed by struct page* of sys ram pages. Also dma_unmap correspodent system
ram pages at error path.

Cc: stable@vger.kernel.org
Signed-off-by: Xiaogang Chen <xiaogang.chen@amd.com>
Reviewed-by: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  drivers/gpu/drm/amd/amdkfd/kfd_migrate.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=da87bcad1f781d822e7ced6d1de9dbc6d381c72e

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

The original Linux kernel CVE announcement for CVE-2026-89810 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89810

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:L/A:H
  The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H
  The Best / paranoid CVSS score: 7

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: 5
  Paranoid ActionableScore: 6

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: YES
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).