[CVE-2026-99106][MODERATE 7.0] drm/amdgpu: free the userptr ranges when a CS fails after parser_bos [ Upstream

[CVE-2026-99106][MODERATE 7.0] drm/amdgpu: free the userptr ranges when a CS fails after parser_bos [ Upstream

[AL-KERNEL]

CVE: CVE-2026-99106
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: drm/amdgpu: free the userptr ranges when a CS fails after parser_bos [ Upstream
Commit: 6ac46ae5ac561d4fc7d4246ad83d71bed0d9af9a
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=6ac46ae5ac561d4fc7d4246ad83d71bed0d9af9a
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99106
Analysis date: Wed, 28 Oct 2026 23:05:11 -0400
ActionableScore: 4
ActionableScore lower bound: 3
Actionable bucket: Borderline / manual review recommended
Manual review required: YES

Summary:
A render-node client can repeatedly fail AMDGPU command submission after userptr setup, leaking HMM ranges, per-page arrays, and BO references and potentially causing sustained memory pressure.

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

The original Linux kernel CVE announcement for CVE-2026-99106 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99106

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-2026-99106	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-99106 CHECK WITH IMPACT FROM ORIG NN LOW | AEGIS IMPACT MODERATE_KPANIC AEGIS_AS=1 LEGACY=MODERATE_KPANIC RESULT=SAME V39_AEGIS_ONE_STEP_UP=MODERATE->MODERATE_KPANIC (final reconciliation agrees; DECISION: both MODERATE_KPANIC)	Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H';*CWE-401;*CWE-772;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:L';BEST CVSS score: '5.5';DESCR 'A failed AMDGPU command submission can leak every HMM range allocated for userptr BO entries after parser_bos succeeds. Each leaked range retains its hmm_pfns array with 8 bytes per userptr page and a BO reference. A local process with access to the DRM render device can reach the demonstrated path by submitting an IB on a ring such as VCN with an unmapped va_start that fails during CS patching. For the CVSS the PR:L value reflects the need for local access to the AMDGPU render node and the ability to submit GPU command streams. The issue is not directly network reachable. The demonstrated primitive is persistent resource retention rather than UAF, OOB access, or arbitrary memory modification. Base impact is limited availability degradation. Sustained repeatable submissions with large userptr ranges could potentially exhaust host memory and justify A:H in the paranoid vector, but immediate host wide failure is not established. Automated impact decision path. The original SimpleNet classifier selected LOW. The primary LLM estimated paranoid CVSS 5.5 and the production ActionableScore returned 4 with recommendation Borderline / manual review recommended. The production rules then evaluated CVSS thresholds and ActionableScore thresholds. Recorded rule transitions were CVSS and initial triage stage LOW to MODERATE_KPANIC because the CVSS score, vector or bug class threshold rule was applied. KP and FD specialist checks stage MODERATE_KPANIC to MODERATE because KP specialist did not establish a qualifying panic. final reconciliation stage MODERATE to MODERATE_KPANIC because the cross model evidence reconciliation rule was applied. The legacy KP specialist produced a review artifact and its decision was processed by the applicable KP rules. The legacy FD or DH downgrade specialist has no review artifact for this run. Exact triggered rule markers were INCREASED_FROM_LOW_BASED_ON_HIGHCVSSSCORE and DECREASED_TO_MODERATEREG_BASED_ON_FALSEPOSCHECKOFKP and V39_AEGIS_ONE_STEP_UP and MANUAL_REVIEW_REQUIRED. The independent weaker Aegis shadow evaluated CVSS 3.3 and ActionableScore 1 with primitive class CLASS_C and manual review REQUIRED. Its independently reconciled impact was MODERATE_KPANIC. Cross model aggregation changed the production result under an enabled evidence gated merge rule. The final reported impact is MODERATE_KPANIC.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4)	MAYBE	INIT LEAK ERRORPATH UAF HARDWARE  INCREASED_FROM_LOW_BASED_ON_HIGHCVSSSCORE DECREASED_TO_MODERATEREG_BASED_ON_FALSEPOSCHECKOFKP KPANIC V39_AEGIS_ONE_STEP_UP MANUAL_REVIEW_REQUIRED	-	-	checked

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

ActionableScore=4
ActionableScoreLower=3
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=NO
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES
RequiresHostUID0=NO
AuthorizationDecisionBypass=NO
ProtectedResourceImpact=NONE
B4PendingReview=NO
B4PendingConfirmed=NO

1. ActionableScore

* Raw base score: 3.
* Conservative score: 3; no floor or cap applied.
* Raw paranoid score: 4.
* Paranoid score: 4; no floor or cap applied.
* Final recommended bucket: **Borderline / manual review recommended**::

2. Signal breakdown

* +1 Concrete persistent kernel resource/reference leak: failed command submissions omit `amdgpu_hmm_range_free()`, retaining each range, its `hmm_pfns` allocation, and its BO reference.
* +1 Local below-host-admin trigger: supplied evidence grounds the path in a process with AMDGPU render-node access.
* +1 Local high-control data-plane API: the submitter can repeatedly create large userptr-backed ranges and deliberately reach CS patch failure.
* Paranoid +1 Availability/resource exhaustion: repeated failures accumulate allocations proportional to userptr size and can plausibly create substantial host-memory pressure. Immediate host-wide failure is not established, so this is omitted from the conservative score.
* No UAF, OOB read/write, arbitrary write, authorization bypass, persistent protected-data corruption, or reliable kernel crash is shown.
* No B5-R retry-accumulated OOB applies: the patch fixes missing cleanup, not an accumulating write index.
* No extra availability point is awarded beyond the paranoid resource-exhaustion concern.

3. Reachability analysis

The grounded actor is a local process permitted to open the AMDGPU render node and submit command streams. Host UID 0 or effective host-administrator powers are not required by the supplied evidence. The process can submit a patch-in-place IB, use an unmapped `va_start`, and repeat the resulting failure. Namespace capability scope and any container/device-delegation path are not established. AMDGPU hardware and render-node exposure are required, but the evidence does not support a rare-hardware penalty or a timing-race penalty.

4. Primitive and severity reading

The exact primitive is persistent leakage of HMM range allocations, per-page `hmm_pfns` arrays, and BO references on several post-`parser_bos` error paths. This is resource-retention DoS behavior, classified CLASS_C, not stale-object use or memory corruption. The shown effect is accumulating kernel memory and references; host-wide OOM or crash remains plausible but unproven. There is no meaningful Important-grade corruption or boundary primitive, so ImportantCandidate is NO and automatic downgrade remains allowed.

5. One-sentence report phrase

A render-node client can repeatedly fail AMDGPU command submission after userptr setup, leaking HMM ranges, per-page arrays, and BO references and potentially causing sustained memory pressure.

6. Manual review advice

MANUAL CHECK RECOMMENDED

Confirm practical leak rate, allocation limits, and whether normal GPU isolation or process teardown eventually bounds retention.

7. Call-site confidence

Call-site confidence: high

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

Patch: drm/amdgpu: free the userptr ranges when a CS fails after parser_bos [ Upstream
Commit: 6ac46ae5ac561d4fc7d4246ad83d71bed0d9af9a
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=6ac46ae5ac561d4fc7d4246ad83d71bed0d9af9a

Changed files:
  drivers/gpu/drm/amd/amdgpu/amdgpu_cs.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=6ac46ae5ac561d4fc7d4246ad83d71bed0d9af9a

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

The original Linux kernel CVE announcement for CVE-2026-99106 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99106

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