[CVE-2026-46229][MODERATE 7.0] drm/amdkfd: Clear VRAM on allocation to prevent stale data exposure

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

CVE: CVE-2026-46229
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: drm/amdkfd: Clear VRAM on allocation to prevent stale data exposure
Commit: 1db431380879fd9d28b763a88a0c0431be5be8df
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=1db431380879fd9d28b763a88a0c0431be5be8df
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-46229
Analysis date: Sat, 01 Aug 2026 10:00:45 -0400
ActionableScore: 5
ActionableScore lower bound: 4
Actionable bucket: Strong Important candidate / Actionable Moderate at minimum**:: Actionable Moderate at minimum. Not an ordinary auto-close Moderate because the bug is a real cross-workload stale VRAM disclosure.
Manual review required: YES

Summary:
AMDKFD VRAM allocations were not cleared before reuse, allowing a local GPU compute user to observe stale VRAM contents from prior allocations and potentially causing limited RCCL P2P crashes due to stale non-zero protocol fields.

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

The original Linux kernel CVE announcement for CVE-2026-46229 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-46229

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-2026-46229	MODERATE	CHECK WITH IMPACT FROM ORIG NN 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:H/I:N/A:L';CWE-226;*CWE-200;CWE-665;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N';BEST CVSS score: '6.1';DESCR 'AMDKFD VRAM allocations could be returned to compute kernels without being cleared, allowing stale data from previous VRAM use to become visible in newly allocated user buffers. The exposed data may include stale page table remnants or prior user buffer contents, so the main impact is confidentiality exposure across GPU workloads or users sharing the same device. For the CVSS the PR:L is used because a local process with access to KFD GPU allocation interfaces can trigger the exposure without full administrator privileges. The issue is not directly network reachable and requires local GPU access. Impact is primarily information disclosure, with limited availability impact possible because stale non zero fields were observed to corrupt RCCL P2P transport state and cause crashes.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate / Actionable Moderate at minimum (with actual score 5)  SKIP CVE-2026-46229  SKIP No affected files built, so skip this CVE NO  - - unknown 	MAYBE	INIT SIMPLEFIX LEAK HARDWARE  KPANIC INCREASED_FROM_LOW_BASED_ON_REQUIREMANUALCHECK	-	-	checked

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

ActionableScore=5
ActionableScoreLower=4

## 1. ActionableScore

* Conservative score: 4
* Paranoid score: 5
* Final recommended bucket: **Strong Important candidate / Actionable Moderate at minimum**:: Actionable Moderate at minimum. Not an ordinary auto-close Moderate because the bug is a real cross-workload stale VRAM disclosure.

## 2. Signal breakdown

Conservative signals:

* Local unprivileged trigger: +1
  A local process with access to AMD KFD or GPU compute allocation interfaces can request VRAM allocations without full root privileges in typical GPU compute deployments.

* Confidentiality impact plausible: +1
  Fresh KFD VRAM allocations may expose stale data from prior VRAM use, including prior user buffers or page table remnants.

* Cross-boundary isolation or policy bypass: +1
  The issue can expose data across GPU workloads or users sharing the same device, bypassing the expected zero-on-allocation isolation property.

* Broad or practically relevant subsystem exposure: +1
  AMDGPU KFD is a relevant production GPU compute path, especially on ROCm or multi-user compute systems.

Paranoid additional signal:

* Availability impact realistic: +1
  The commit states that stale non-zero data can corrupt RCCL P2P transport handshake fields and cause crashes. This is not a strong kernel DoS primitive, but it supports limited practical availability impact.

Not counted:

* No generic memory corruption. The patch does not show UAF, OOB write, double free, arbitrary write, or type confusion.
* No privilege escalation signal. The primitive is stale data disclosure, not control-flow or kernel object corruption.
* No remote reachability. Triggering requires local GPU/KFD access.

## 3. Reachability analysis

The bug is locally reachable through the AMDKFD GPU memory allocation path. A user or process that can run compute workloads and allocate VRAM through KFD may observe stale VRAM contents in newly allocated buffers. This is typically PR:L rather than PR:H because many GPU compute systems grant non-root users access to render or KFD devices.

Namespaces and containers may matter if GPU devices are passed through to containers. In such deployments, the issue can become a cross-container or cross-workload confidentiality problem. The path is not network-triggerable directly and requires AMD GPU hardware plus KFD/ROCm style compute usage.

Call-site confidence: medium. The changed function is the direct KFD GPU memory allocation path, and the commit explicitly states that missing AMDGPU_GEM_CREATE_VRAM_CLEARED exposes stale VRAM to compute kernels.

## 4. Severity interpretation

This behaves more like an actionable Moderate than a low-risk correctness bug. The realistic impact is information disclosure of stale VRAM contents across GPU allocations, not kernel memory corruption or LPE. The availability impact is secondary and limited to application or transport crashes described in RCCL P2P usage.

Conservative handling should not classify it as Important solely from the crash symptom. Paranoid handling should not auto-close it either, because stale GPU memory can contain sensitive workload data and the bug affects an isolation expectation between users or compute jobs.

## 5. One-sentence report phrase

AMDKFD VRAM allocations were not cleared before reuse, allowing a local GPU compute user to observe stale VRAM contents from prior allocations and potentially causing limited RCCL P2P crashes due to stale non-zero protocol fields.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED. YES REQUIRES MANUAL CHECK.

Reason: this is a real information disclosure issue across GPU memory allocations. It is not proven LPE or kernel memory corruption, but stale VRAM exposure can leak sensitive workload data in multi-user GPU environments and should not be auto-closed.

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

Patch: drm/amdkfd: Clear VRAM on allocation to prevent stale data exposure
Commit: 1db431380879fd9d28b763a88a0c0431be5be8df
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=1db431380879fd9d28b763a88a0c0431be5be8df

Commit description:

KFD VRAM allocations set AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE
but not AMDGPU_GEM_CREATE_VRAM_CLEARED, leaving freshly allocated
VRAM with stale data from prior use observable by compute kernels.

The GEM ioctl path already sets VRAM_CLEARED for all userspace
allocations via amdgpu_gem_create_ioctl() and
amdgpu_mode_dumb_create(). The KFD path was missing this flag,
allowing stale page table remnants to leak into user buffers.

This causes crashes in RCCL P2P transport where non-zero data in
ptrExchange/head/tail fields corrupts the protocol handshake.

Signed-off-by: Amir Shetaia <Amir.Shetaia@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.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=1db431380879fd9d28b763a88a0c0431be5be8df

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

The original Linux kernel CVE announcement for CVE-2026-46229 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-46229

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:H/I:N/A:N
  The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:L
  The Best / paranoid CVSS score: 6.1

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: 4
  Paranoid ActionableScore: 5

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).