* [CVE-2026-98106][MODERATE REGULAR] drm/pagemap: Prevent double migration of device pages [ Upstream
@ 2026-09-25 11:47 21% AL-KERNEL
0 siblings, 0 replies; 1+ results
From: AL-KERNEL @ 2026-09-25 11:47 UTC (permalink / raw)
To: kernel-cve
CVE: CVE-2026-98106
Priority: MODERATE REGULAR
AL-KERNEL base severity: MODERATE
KPANIC flag: NO
Patch: drm/pagemap: Prevent double migration of device pages [ Upstream
Commit: 62451f540c46bf5c51143a1c8681b472608bd54d
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=62451f540c46bf5c51143a1c8681b472608bd54d
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98106
Analysis date: Fri, 25 Sep 2026 07:47:33 -0400
ActionableScore: 1
ActionableScore lower bound: unknown
Actionable bucket: Ordinary Moderate
Manual review required: NO
Summary:
No one-sentence report phrase was found.
======================================================================
ABOUT THIS REPORT
======================================================================
The original Linux kernel CVE announcement for CVE-2026-98106 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98106
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: NO
A detailed explanation of the methodology and priority rules is included
at the end of this message.
======================================================================
AL-KERNEL CLASSIFICATION RESULT
======================================================================
CVE-2026-98106 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-98106 CHECK WITH IMPACT FROM ORIG NN MODERATE Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:H';CWE-362;*CWE-835;**CWE-400;BEST CVSS score: '4.7';DESCR 'A race between CPU fault migration and raw PFN eviction can migrate the same DRM device private folio twice before migration finalization drops the source reference. The second migration can leave an uncharged folio on an LRU list and cause folio_lruvec_lock_irqsave to retry indefinitely, resulting in a CPU soft lockup and RCU stall. For the CVSS the PR:L is used because a local process needs access to the affected GPU render interface and Xe GPU SVM functionality, but does not necessarily need administrative privileges. The issue is not network reachable. Impact is denial of service through system unresponsiveness. No use after free, out of bounds access, stale rmap traversal, information disclosure, or write primitive is demonstrated by the patch. YES REQUIRES MANUAL CHECK';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Ordinary Moderate (with actual score 1) YES READ OOB DEADLOCK DANGER HARDWARE MEMORY DECREASED_TO_MODERATE_REGULAR_FROM_MODERATE7_BASED_ON_GUESSCVSS_AND_CIANNH YES NO checked
======================================================================
ACTIONABLESCORE ANALYSIS
======================================================================
ActionableScore=1
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES
1. ActionableScore
* Raw conservative score: 1
* Conservative score: 1
* Raw paranoid score: 1
* Paranoid score: 1
* Final recommended bucket: **Ordinary Moderate**::
The supplied CVSS assessment of 4.7 reflects a realistic local availability impact under an assumed GPU SVM attacker model. The ActionableScore is lower because the patch establishes a DoS-only migration-state inconsistency and does not establish a corruption, lifetime, ownership, authorization, or disclosure primitive.
2. Signal breakdown
* +1 Reliable kernel DoS: The commit documents an observed CPU soft lockup and RCU stall. The retry loop in folio_lruvec_lock_irqsave can indefinitely stall a kernel worker after the same device-private folio is migrated twice.
* +0 Local unprivileged trigger: Not awarded. Xe render-node access may be available to non-administrative users in some deployments, but the supplied patch does not establish that the complete GPU SVM allocation, CPU fault, eviction, and double-migration sequence is reachable by an ordinary local user.
* +0 Attacker-influenced active-object lifetime manipulation: Not awarded. The affected state is a migration operation on a device-private folio, not a stale object lifetime or refcount transition controlled by the attacker.
* +0 Real UAF: Not awarded. The source device-private folio remains reachable during migration finalization, but the evidence describes duplicate migration, not dereference after free or use after lifetime end.
* +0 Strong or weak exploitation leverage: Not awarded. No controlled reclaim, object replacement, arbitrary write, directed write, callback influence, refcount takeover, structured stale-object consumption, or security-boundary bypass is shown.
* +0 Confidentiality and integrity: Not awarded. The uncharged LRU folio and retry loop demonstrate an availability failure only. There is no demonstrated read-back, disclosure, cross-owner page modification, or unauthorized state change.
* +0 Additional availability: Not awarded. The observed soft lockup and RCU stall are the same DoS consequence already counted above.
* -0 Hard race penalty: Not applied. The patch describes a concurrent window between CPU-fault migration and raw-PFN eviction, but does not establish that it is a one-shot, externally uncontrollable, or prohibitively difficult race.
* -0 rare hardware penalty: Not applied. Intel Xe GPU SVM is hardware- and configuration-dependent, but the supplied evidence does not establish rare physical-only hardware or a restriction that would justify a negative score.
Important non-awarded signals:
* Controlled reclaim is not demonstrated.
* Arbitrary or directed kernel write is not demonstrated.
* A real UAF or repeated release/refcount mutation is not demonstrated.
* Remote reachability is not established.
* A security-boundary bypass or privilege escalation path is not established.
* The patch does not establish stale rmap traversal, active secondary-index corruption, or an ownership/COW violation.
* The low-bit migrated marker prevents duplicate migration. It does not indicate that stale pointer storage is dereferenced after object destruction.
3. Reachability analysis
The affected path requires an Intel Xe DRM environment using GPU Shared Virtual Memory and device-private memory migration. The observed execution involves a CPU fault path and a raw-PFN eviction path racing before migration finalization drops the source reference.
The likely triggering actor is a local GPU client that can create or use the relevant Xe GPU SVM memory objects and generate memory pressure, faults, or eviction activity. However, the supplied patch does not expose the exact DRM ioctl permission checks, render-node policy, required Xe SVM setup, or whether all required operations are available to an ordinary unprivileged user in the evaluated product.
Therefore:
* HostAdminRequired is UNKNOWN. Host-admin privilege is not positively established, but below-host-admin reachability of the complete trigger is also not fully proven.
* NamespaceScopedPrivilege is UNKNOWN. No namespace capability path is shown, and GPU device access is not established as fully namespace-scoped.
* No prior host-admin setup is proven beyond the general requirement that the system has a supported Intel Xe GPU, the relevant driver and GPU SVM support enabled, and an accessible DRM device.
* The issue is not network reachable.
* The race may be repeatable under workload and memory pressure, but repeatability sufficient for an unprivileged attacker is not established by the supplied context.
4. Primitive and severity interpretation
The concrete primitive is duplicate migration of the same device-private folio before migration finalization removes the source reference. The duplicate migration can leave an uncharged folio on an LRU list. A later LRU operation retries indefinitely, producing a soft lockup and RCU stall.
This is not a use-after-free. The source reference has not yet been dropped when the duplicate migration becomes possible. The patch introduces a migration-retired marker in zone_device_data and makes both migration paths skip already migrated device-private folios.
The affected state is active migration bookkeeping, but no active stale structured object is consumed after free. There is no second put, release, free, refcount decrement, callback invocation, reclaimed-object interpretation, or post-free write. This is therefore CLASS_C rather than CLASS_B.
The demonstrated consequence is host availability degradation through CPU soft lockup and RCU stall. A system may become substantially unresponsive while the worker loops. A kernel panic, data corruption, information disclosure, integrity violation, or privilege escalation is not established.
ImportantCandidate is NO because the patch supports a DoS-only state-consistency failure without a meaningful unresolved memory-corruption or security-boundary primitive. AutoDowngradeAllowed is YES because automatic closure as non-Important is technically defensible after verifying whether the affected Xe GPU SVM configuration is deployed.
5. One-sentence report phrase
A race between CPU-fault migration and raw-PFN eviction can migrate the same Xe GPU device-private folio twice, causing an LRU retry loop, CPU soft lockup, and local denial of service without demonstrated memory corruption or privilege escalation.
6. Manual review recommendation
MANUAL CHECK RECOMMENDED
Confirm whether the affected product enables Intel Xe GPU SVM and whether non-administrative users can access the complete DRM render-node and memory-migration trigger path. The available evidence supports DoS only, so this does not require Important-class escalation unless additional reachability or corruption evidence is found.
7. Call-site confidence
Call-site confidence: medium
======================================================================
UPSTREAM PATCH SUMMARY
======================================================================
Patch: drm/pagemap: Prevent double migration of device pages [ Upstream
Commit: 62451f540c46bf5c51143a1c8681b472608bd54d
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=62451f540c46bf5c51143a1c8681b472608bd54d
Changed files:
drivers/gpu/drm/drm_pagemap.c
include/drm/drm_pagemap.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=62451f540c46bf5c51143a1c8681b472608bd54d
======================================================================
DETAILED REPORT METHODOLOGY
======================================================================
The original Linux kernel CVE announcement for CVE-2026-98106 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98106
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:H/PR:L/UI:N/S:U/C:N/I:N/A:H
The Best / paranoid CVSS score: 4.7
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: unknown
Paranoid ActionableScore: 1
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 [email protected] (and both
send reply to CVE record itself too and see "reply" button below for howto reply).
^ permalink raw reply [relevance 21%]
Results 1-1 of 1 | reverse | sort options + mbox downloads above
-- links below jump to the message on this page --
2026-09-25 11:47 21% [CVE-2026-98106][MODERATE REGULAR] drm/pagemap: Prevent double migration of device pages [ Upstream AL-KERNEL
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox