[CVE-2026-98178][LOW] drm/amdgpu: Skip KFD mapping clear before initialization

[CVE-2026-98178][LOW] drm/amdgpu: Skip KFD mapping clear before initialization

[AL-KERNEL]

CVE: CVE-2026-98178
Priority: LOW
AL-KERNEL base severity: LOW
KPANIC flag: NO
Patch: drm/amdgpu: Skip KFD mapping clear before initialization
Commit: 157d3f1db7e7e6f320daa221fae84a4c13555b6f
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=157d3f1db7e7e6f320daa221fae84a4c13555b6f
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98178
Analysis date: Thu, 08 Oct 2026 07:57:29 -0400
ActionableScore: 1
ActionableScore lower bound: unknown
Actionable bucket: Ordinary Moderate
Manual review required: NO

Summary:
An AMDGPU KFD initialization failure can leave a partially initialized device state that causes a NULL-pointer dereference during a later GPU reset, resulting in a local denial of service without demonstrated memory corruption or privilege escalation.

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

The original Linux kernel CVE announcement for CVE-2026-98178 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98178

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: LOW
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-2026-98178	LOW	https://www.kernelcve.org/list/?q=CVE-2026-98178 CHECK WITH IMPACT FROM ORIG NN LOW | AEGIS IMPACT LOW AEGIS_AS=0 LEGACY=LOW RESULT=SAME (legacy and Aegis reconciliation agree; DECISION: both LOW)	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-476;CWE-665;CWE-754;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '5.5';DESCR 'The supplied source contains only the abbreviated stable commit ID 7f9caa7. The analysis is based on the shown commit message and diff. Reassess manually if a report demonstrates GPU reset triggering without local render or compute access, or any memory corruption beyond the shown NULL dereference.';NO MANUAL CHECK; ,and ActionableScore result is Ordinary Moderate (with actual score 1)	MAYBE	INIT SIMPLEFIX HARDWARE  INCREASED_FROM_LOW_BASED_ON_HIGHCVSSSCORE DECREASED_TO_MODERATEREG_BASED_ON_ACTIONABLESCORELESSTHAN4 DECREASED_TO_LOW_FROM_MODERATE_BASED_ON_ACTIONABLESCORELESSTHAN3	-	-	checked

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

ActionableScore=1
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=NO
ManualReview=NO
AutoDowngradeAllowed=YES

## 1. ActionableScore

* Raw conservative score: 1
* Conservative score: 1
* Raw paranoid score: 1
* Paranoid score: 1
* Final recommended bucket: **Ordinary Moderate**::

No floor or cap was applied. The paranoid interpretation cannot defensibly add a reachability or corruption point because the supplied patch establishes only an uninitialized pointer dereference after a failed KFD initialization and later GPU reset.

## 2. Signal breakdown

* +1 Reliable kernel fault or strong DoS. The commit directly describes a later GPU reset dereferencing `nodes[0]->id` when `nodes[0]` was never allocated. This can produce a kernel NULL-pointer dereference and an Oops or crash.

Important non-awarded signals:

* No real UAF. `kfd_dev` remains allocated and attached. The invalid object state is that its `nodes[0]` member was never initialized. There is no freed object, dangling reference, or post-lifetime use.
* No repeated release or refcount mutation. The vulnerable path walks the node array and dereferences an uninitialized entry. It does not perform a second `put`, `free`, or refcount operation on a potentially freed object.
* No controlled reclaim, object replacement, arbitrary write, OOB access, callback control, or attacker-shaped post-free data is established.
* No confidentiality or integrity signal. The patch supports a NULL-pointer dereference, not information disclosure or unauthorized modification.
* No local-unprivileged-trigger point. The supplied evidence does not establish that an ordinary local user can cause the required GPU reset after the failed KFD initialization.
* No host-admin penalty. Host-admin-equivalent privilege is not positively established as required for the complete reset trigger.
* No rare-hardware penalty. Polaris hardware and missing PCIe atomics support are configuration and platform conditions, but the evidence does not establish genuinely rare physical-only access requirements.
* No remote-reachability point. Network input does not reach this path.

## 3. Reachability analysis

The vulnerable state is created when KFD initialization fails after probe. The described example is a Polaris GPU without required PCIe atomics support. `num_nodes` is already set, but `nodes[0]` is not allocated, while `kfd_dev` remains attached to the AMDGPU device.

A later GPU reset reaches `amdgpu_amdkfd_clear_kfd_mapping()` and dereferences the uninitialized node pointer. The patch does not show what actor can reliably initiate that reset. It may occur through driver recovery, GPU hang handling, display operations, or privileged device management, but none of those complete paths is established in the supplied context.

A logged-in graphical user, DRM render client, DRM master, compositor, delegated graphics service, and host administrator have materially different permissions. The supplied patch does not prove that any below-host-admin actor can trigger the complete sequence. Therefore:

* HostAdminRequired is `UNKNOWN`.
* NamespaceScopedPrivilege is `UNKNOWN`.
* No ordinary-user or delegated-capability reachability bonus is applied.
* No host-admin penalty is applied either, because host-admin privilege is not positively proven.

The issue is not network reachable. No namespace-specific capability check or attacker-created namespace path is relevant in the supplied evidence.

## 4. Primitive and severity interpretation

The exact primitive is a NULL-pointer dereference caused by an initialization-state mismatch:

`kfd_dev exists`
`num_nodes is nonzero`
`nodes[0] was never allocated`
`GPU reset walks nodes[0]`
`kernel dereferences nodes[0]->id`

This is not an active-object lifetime violation. `kfd_dev` has not been freed, and the pointer is not stale after object lifetime termination. The node array entry is simply absent because initialization failed before node allocation.

The vulnerable operation is neither a destructor UAF nor a repeated lifetime/refcount mutation. It is a direct invalid-state dereference during KFD mapping cleanup before KFD initialization completed.

The demonstrated consequence is local denial of service through a kernel fault during GPU reset. A GPU-only disruption may also be possible depending on kernel Oops and panic policy, but no stronger host integrity, confidentiality, or privilege-escalation primitive is supported.

The primitive is therefore `CLASS_C`. It is not a meaningful Important candidate because no memory corruption, controlled reuse, active-object lifetime corruption, security-boundary bypass, or privilege-escalation mechanism is established. Automatic downgrade is allowed because the available evidence supports a constrained crash/DoS condition only.

## 5. One-sentence report phrase

An AMDGPU KFD initialization failure can leave a partially initialized device state that causes a NULL-pointer dereference during a later GPU reset, resulting in a local denial of service without demonstrated memory corruption or privilege escalation.

## 6. Manual review recommendation

NO MANUAL CHECK NEEDED

The patch and commit message establish a deterministic NULL-pointer dereference in a partially initialized KFD state, but do not establish a UAF, arbitrary memory access, privilege-boundary bypass, or below-host-admin trigger path requiring escalation review.

## 7. Call-site confidence

Call-site confidence: medium

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

Patch: drm/amdgpu: Skip KFD mapping clear before initialization
Commit: 157d3f1db7e7e6f320daa221fae84a4c13555b6f
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=157d3f1db7e7e6f320daa221fae84a4c13555b6f

Commit description:

amdgpu_amdkfd_clear_kfd_mapping() assumes that a non-NULL kfd_dev
has a fully populated node array. This is not true when KFD device
initialization fails after probe.

For example, kgd2kfd_device_init() sets num_nodes before checking
PCIe atomics support. On Polaris systems without the required atomics,
it returns before allocating nodes[0], but the kfd_dev remains attached
to the amdgpu device. A later GPU reset then dereferences nodes[0]->id.

Require the authoritative KFD initialization flag before walking the
node array, matching the existing KFD reset and teardown paths.

Fixes: 70cadef ("drm/amdgpu: unmap all user mappings of framebuffer and doorbell before mode1 reset")
Closes: https://gitlab.freedesktop.org/drm/amd/-/work_items/5833
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit 4ac1835 )
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.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=157d3f1db7e7e6f320daa221fae84a4c13555b6f

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

The original Linux kernel CVE announcement for CVE-2026-98178 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98178

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:N/I:N/A:H
  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: 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: LOW
KPANIC detected for this report: NO
Published priority for this report (same as in Subject): LOW

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).