[CVE-2026-72095][MODERATE 7.0] dma-fence: Make dma_fence_dedup_array() robust against 0-count input commit 77a9298741f8f9e8b963c977f5582ab21c6d3427 upstream.

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

CVE: CVE-2026-72095
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: dma-fence: Make dma_fence_dedup_array() robust against 0-count input commit 77a9298741f8f9e8b963c977f5582ab21c6d3427 upstream.
Commit: 7aa8f3dba53422465dbe1be8dbb7240304462bb2
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=7aa8f3dba53422465dbe1be8dbb7240304462bb2
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72095
Analysis date: Sat, 15 Aug 2026 19:40:04 -0400
ActionableScore: 5
ActionableScore lower bound: 4
Actionable bucket: Actionable Moderate at minimum
Manual review required: YES

Summary:
dma_fence_dedup_array() returned one fence for zero input, allowing amdgpu user queue wait paths to consume an uninitialized dma_fence slot and potentially crash the kernel, with weak but nonzero concern for object misinterpretation in a local GPU interface.

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

The original Linux kernel CVE announcement for CVE-2026-72095 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72095

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-2026-72095	MODERATE	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:L/I:L/A:H';CWE-824;*CWE-457;**CWE-476;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '5.8';DESCR 'dma_fence_dedup_array incorrectly returned 1 for a zero input count, which could make callers believe that one valid fence exists in an array where no fence was initialized. In the amdgpu user queue wait path this can lead to dereferencing an uninitialized dma_fence pointer from a local GPU control interface, causing a kernel crash and denial of service. For the CVSS the PR:L is used for the paranoid score because a local user or process with access to the AMD GPU render or control device may be able to reach the userq wait path without full administrative privileges. The issue is not network reachable and requires local interaction with the GPU driver interface. Impact is at least local denial of service. In the paranoid interpretation, limited confidentiality or integrity impact is kept possible because an uninitialized kernel pointer may be consumed as a dma_fence object, but the patch does not show a reliable arbitrary write primitive or a proven privilege escalation path.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate at minimum (with actual score 5)  SKIP CVE-2026-72095 UNKNOWN SKIP The Fixes patch not applied yet, so unlikely that actual: 575ec9b0c2f11f40535ea737ed5a64792780d1ef YES  NO NO unknown 	MAYBE	HARDWARE KPANIC MEMORY 	-	-	checked

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

ActionableScore=5
ActionableScoreLower=4

## 1. ActionableScore

* Conservative score: 4
* Paranoid score: 5
* Final recommended bucket: **Actionable Moderate at minimum**::

## 2. Signal breakdown

Conservative signals:

* Local unprivileged trigger: +1
  The relevant downstream path is described as amdgpu_userq_wait_*(). On systems exposing AMD GPU render or queue interfaces to local users, a non-root local process may plausibly reach this path.

* Memory corruption, Weak/indirect corruption candidate: +1
  The bug can cause an uninitialized fence slot to be interpreted as a valid struct dma_fence pointer. This is stronger than pure bookkeeping, but the patch does not show controlled reclaim, controlled replacement, arbitrary write, or a reliable type-confusion primitive.

* Reliable kernel crash / strong DoS: +1
  Dereferencing an uninitialized dma_fence slot can realistically cause an invalid pointer dereference and kernel crash.

* Privileged kernel/device-management memory corruption path: +1
  dma_fence is a core synchronization object used by DMA/GPU drivers. Even though attacker control is not proven, the corrupted object interpretation occurs in a trusted GPU synchronization path.

Paranoid additional signal:

* Weak LPE concern: +1
  A bogus dma_fence pointer could in worst case be interpreted as a kernel object with operations and lifetime semantics, but this remains weak because no attacker-controlled reuse, callback control, write primitive, or demonstrated privilege escalation is shown.

Not awarded:

* Remote reachable: +0
  No network path is involved.

* Strong memory corruption primitive: +0
  No UAF reclaim, OOB write, arbitrary write, controlled overwrite, or reliable type confusion is shown.

* Confidentiality or integrity impact as main score: +0
  C/I impact is only a paranoid concern, not demonstrated by the patch.

* Hard race penalty: 0
  This is not described as a timing race.

## 3. Reachability analysis

The likely trigger is local access to AMD GPU user queue wait functionality. In typical desktop or compute deployments, GPU render nodes may be available to non-root users, so PR:L is a reasonable practical assumption. Containers may inherit GPU device access if the host passes through render nodes, which can make the issue relevant in containerized GPU workloads.

The affected helper is generic dma-fence code, but the commit identifies amdgpu_userq_wait_*() as the unsafe caller. Systems without the relevant AMD GPU path or without user-accessible GPU devices have reduced exposure. The path is not remotely reachable.

Call-site confidence: medium. The commit names the vulnerable amdgpu caller, but the full caller code is not included in the supplied diff.

## 4. Severity interpretation

This behaves like an Actionable Moderate issue rather than an ordinary low-priority crash. The realistic impact is local kernel DoS through an uninitialized pointer dereference. The paranoid concern is that a bogus dma_fence pointer may become an invalid object interpretation in a GPU/DMA synchronization path, but the patch does not prove controlled memory corruption or a practical LPE primitive.

Conservative handling should not classify this as Important. However, it should not be auto-closed because the affected path is plausibly local-user reachable and involves kernel object pointer misuse in a device synchronization subsystem.

## 5. One-sentence report phrase

dma_fence_dedup_array() returned one fence for zero input, allowing amdgpu user queue wait paths to consume an uninitialized dma_fence slot and potentially crash the kernel, with weak but nonzero concern for object misinterpretation in a local GPU interface.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED

Reason: this is a local kernel memory-safety issue in a GPU/DMA synchronization path with an uninitialized object pointer, likely DoS impact, and weak but defensible LPE concern.

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

Patch: dma-fence: Make dma_fence_dedup_array() robust against 0-count input commit 77a9298741f8f9e8b963c977f5582ab21c6d3427 upstream.
Commit: 7aa8f3dba53422465dbe1be8dbb7240304462bb2
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=7aa8f3dba53422465dbe1be8dbb7240304462bb2

Commit description:

dma_fence_dedup_array() returns 1 when called with num_fences == 0:
the for-loop body never executes, j stays at 0, and the final
`return ++j` yields 1. This contradicts both the kernel-doc ("Return:
Number of unique fences remaining in the array") and the natural
expectation that 0 input gives 0 output.

The caller __dma_fence_unwrap_merge() bails out via the
`if (count == 0 || count == 1)` fast path and so is save.

But amdgpu_userq_wait_*() could reach the dedup call with a zero local
count and dereference an uninitialized fence slot in the array.

Make the contract match the documentation by returning 0 early. This
also skips an unnecessary sort() call on an empty array.

Cc: stable@vger.kernel.org
Signed-off-by: Baineng Shou <shoubaineng@gmail.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Fixes: 575ec9b0c2f1 ("dma-fence: Add helper to sort and deduplicate dma_fence arrays")
Signed-off-by: Christian König <christian.koenig@amd.com>
Link: https://lore.kernel.org/r/20260629031346.3875683-1-shoubaineng@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Diffstat -rw-r--r-- drivers/dma-buf/dma-fence-unwrap.c 3

Changed files:
  drivers/dma-buf/dma-fence-unwrap.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=7aa8f3dba53422465dbe1be8dbb7240304462bb2

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

The original Linux kernel CVE announcement for CVE-2026-72095 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-72095

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:H
  The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H
  The Best / paranoid CVSS score: 5.8

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).