[CVE-2026-74446][IMPORTANT] drm/amdkfd: hold event_mutex while checkpointing CRIU events

[CVE-2026-74446][IMPORTANT] drm/amdkfd: hold event_mutex while checkpointing CRIU events

[AL-KERNEL]

CVE: CVE-2026-74446
Priority: IMPORTANT
AL-KERNEL base severity: IMPORTANT
KPANIC flag: YES
Patch: drm/amdkfd: hold event_mutex while checkpointing CRIU events
Commit: 2040b7e39027cb83bb8c7b84a4c95c2f6053c32f
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=2040b7e39027cb83bb8c7b84a4c95c2f6053c32f
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74446
Analysis date: Sat, 15 Aug 2026 13:04:45 -0400
ActionableScore: 7
ActionableScore lower bound: 6
Actionable bucket: Strong Important candidate / Actionable Moderate at minimum
Manual review required: YES

Summary:
A race in amdkfd CRIU event checkpointing allows concurrent event create or destroy to change p event_idr between the count and walk, causing an out of bounds write or use after free candidate in kernel memory and requiring manual review for possible local privilege escalation.

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

The original Linux kernel CVE announcement for CVE-2026-74446 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74446

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: IMPORTANT
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-2026-74446	IMPORTANT	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:H/I:H/A:H';CWE-362;*CWE-787;*CWE-416;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'A race in amdkfd CRIU event checkpointing can occur because kfd_criu_checkpoint_events() counts entries in p event_idr, allocates ev_privs for that count, and then walks the same IDR without holding p event_mutex. A second local thread in the same process can create or destroy KFD events between the count and the walk. If an event is inserted, the walk can write past the end of the ev_privs allocation. If an event is removed, the walk can observe an object that is being freed, creating a use-after-free candidate. For the CVSS the PR:L is used because reliable triggering requires local code execution with access to the amdkfd CRIU event checkpoint path, but it does not require full administrative privileges in typical GPU compute environments. The issue is not network reachable. Impact is at least a local denial of service via kernel crash and in worst case may allow confidentiality or integrity impact and privilege escalation due to kernel memory corruption.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate / Actionable Moderate at minimum (with actual score 7)  SKIP CVE-2026-74446 UNKNOWN SKIP No affected files built, so skip this CVE NO  - - unknown 	MAYBE	LOCK DEADLOCK HARDWARE  KPANIC INCREASED_TO_MODERATE_FROM_LOW_BASED_ON_GUESSCVSS KPANIC INCREASED_TO_HIGH_BASED_ON_ACTIONABLESCOREHIGHEREQTHAN6	-	-	checked

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

ActionableScore=7
ActionableScoreLower=6

## 1. ActionableScore

* Conservative score: 6
* Paranoid score: 7
* Final recommended bucket: **Strong Important candidate / Actionable Moderate at minimum**::

## 2. Signal breakdown

Conservative signals:

* Local unprivileged trigger: +1
  A local process with access to the amdkfd KFD event and CRIU checkpoint path can race checkpointing against event create or destroy.

* Memory corruption, strong corruption primitive: +2
  The commit explicitly states that event insertion between count and walk can make the loop write past the end of the ev_privs allocation.

* Real lifetime corruption: +1
  Event removal during the unlocked IDR walk can make the walker observe or dereference an event that is being freed, making this a UAF candidate.

* Weak to moderate LPE concern: +1
  The primitive is local kernel memory corruption, but the overwrite target and payload are not shown to be reliably attacker controlled.

* Reliable kernel crash / strong DoS: +1
  OOB write or UAF in kernel driver code is a credible kernel crash path.

* Hard or unreliable race / special timing required: -1
  The trigger depends on a race between checkpoint count/walk and event create/destroy.

* Rare or specialized path: -1
  The affected path is amdkfd CRIU event checkpointing, not a broad default kernel path.

Paranoid additional interpretation:

* Privilege escalation plausible, stronger interpretation: +1 additional
  Because the patch shows an actual OOB write plus UAF candidate in a local GPU driver object-management path, manual review should treat LPE as plausible even if not demonstrated.

## 3. Reachability analysis

The bug is locally reachable through amdkfd event handling and the CRIU checkpoint path. It is not network reachable. In typical deployments, access depends on /dev/kfd permissions, GPU availability, render group membership, and whether CRIU checkpoint functionality is usable by the local process. This should be treated as PR:L rather than PR:H when an ordinary GPU compute user can access the KFD device and create or destroy events in its own process. Containers or delegated GPU access can make the path relevant for reduced-privilege workloads, but this is still a local attack surface. Exploitation requires racing two threads in the same process, so AC is high, but the race is structurally clear from the patch.

## 4. Severity interpretation

This is more than an ordinary Moderate correctness fix because the patch context explicitly describes an OOB write and a UAF candidate. Realistic exploitation evidence is limited because the race timing, heap layout control, overwrite target, and reclaim behavior are not demonstrated. Conservative handling should classify it as Actionable Moderate. Paranoid triage can treat it as a Strong Important candidate because local kernel memory corruption in a GPU driver object-management path may be exploitable for privilege escalation after manual analysis.

## 5. One-sentence report phrase

A race in amdkfd CRIU event checkpointing allows concurrent event create or destroy to change p event_idr between the count and walk, causing an out of bounds write or use after free candidate in kernel memory and requiring manual review for possible local privilege escalation.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED

Reason: the patch explicitly fixes a race that can produce an OOB write and a UAF candidate, and these are kernel memory corruption primitives even though reliable privilege escalation is not demonstrated.

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

Patch: drm/amdkfd: hold event_mutex while checkpointing CRIU events
Commit: 2040b7e39027cb83bb8c7b84a4c95c2f6053c32f
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=2040b7e39027cb83bb8c7b84a4c95c2f6053c32f

Commit description:

kfd_criu_checkpoint_events() counts the entries in p->event_idr via
kfd_get_num_events(), allocates an array sized to that count, and then
walks the same IDR to fill it. Neither the count nor the walk holds
p->event_mutex.

The CRIU checkpoint caller holds only p->mutex. Event create and destroy
(kfd_event_create()/kfd_event_destroy()) take p->event_mutex and do not
take p->mutex, so a second thread in the same process can insert or remove
events between the count and the walk. If an event is inserted, the walk
iterates more entries than were counted and writes past the end of the
ev_privs allocation; if an event is removed, the walk dereferences an
entry that is being freed.

Hold p->event_mutex across the count and the walk so both observe a
consistent view of p->event_idr. The lock is released before
copy_to_user(), which only touches the local buffer. The caller already
holds p->mutex and the create/destroy paths never take p->mutex, so the
p->mutex -> p->event_mutex order is not inverted and no deadlock is
introduced.

Fixes: 40e8a76 ("drm/amdkfd: CRIU checkpoint and restore events")
Signed-off-by: William Palacek <William.Palacek@amd.com>
Reviewed-by: Alysa Liu <Alysa.Liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(cherry picked from commit ff57e22 )
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Changed files:
  drivers/gpu/drm/amd/amdkfd/kfd_events.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=2040b7e39027cb83bb8c7b84a4c95c2f6053c32f

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

The original Linux kernel CVE announcement for CVE-2026-74446 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-74446

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:L/I:L/A:H
  The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H
  The Best / paranoid CVSS score: 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: 6
  Paranoid ActionableScore: 7

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: IMPORTANT
KPANIC detected for this report: YES
Published priority for this report (same as in Subject): IMPORTANT

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).