[CVE-2026-92500][MODERATE 7.0] ext4: use fsdata to track inline data write state and fix race [ Upstream

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

CVE: CVE-2026-92500
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: ext4: use fsdata to track inline data write state and fix race [ Upstream
Commit: bd8d74bd46d09905164255b8635fa58b45f41068
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=bd8d74bd46d09905164255b8635fa58b45f41068
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-92500
Analysis date: Fri, 18 Sep 2026 05:03:40 -0400
ActionableScore: 4
ActionableScore lower bound: unknown
Actionable bucket: Borderline / manual review recommended
Manual review required: YES

Summary:
No one-sentence report phrase was found.

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

The original Linux kernel CVE announcement for CVE-2026-92500 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-92500

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-2026-92500	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-92500 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-362;**CWE-476;CWE-367;*CWE-664;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:H';BEST CVSS score: '5.8';DESCR 'An ext4 inline data race can occur when write_begin() prepares an inline write but a concurrent path such as ext4_page_mkwrite() converts the inode to extent based storage before write_end() completes. The old write_end logic checked the live inode state instead of the state prepared by write_begin(), so it could miss the inline write_end path and execute extent based logic without allocated block buffers. This can lead to a NULL pointer dereference through a NULL folio_buffers(folio), a BUG_ON panic in ext4_write_inline_data_end(), or possible file data loss. For the CVSS the PR:L is used because a local user or process with write access to files on an affected ext4 filesystem may be able to trigger the race without administrative privileges. The issue is not network reachable and requires local filesystem activity plus timing of concurrent write and page fault or conversion paths. Impact is at least local denial of service via kernel crash and low integrity impact due to possible data loss. For the paranoid score, limited confidentiality impact is kept only as a manual review sensitivity margin, because the patch shows a race and NULL dereference but does not demonstrate a strong arbitrary write or privilege escalation primitive.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4)	YES	WRITE OOB LOCK DANGER NULLPTR DISK RACE IMPROVEONLY LINUS SYZBOT KPANIC 	NO	NO	checked

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

ActionableScore=4
PrimitiveClass=CLASS_C
RealUAF=NO
HostAdminRequired=NO
NamespaceScopedPrivilege=NO
ImportantCandidate=NO
ManualReview=RECOMMENDED
AutoDowngradeAllowed=YES

1. ActionableScore

* Raw conservative score: 4
* Conservative score: 4
* Raw paranoid score: 4
* Paranoid score: 4
* Final recommended bucket: **Borderline / manual review recommended**::

No floor or cap changed the score.

2. Signal breakdown

Positive signals:

* +1 Local unprivileged trigger. A local user with write access to an affected ext4 filesystem can plausibly reach the buffered write path and race it with page fault or inline-data conversion activity.
* +1 Reliable kernel crash / strong DoS. The patch description explicitly identifies NULL pointer dereference through NULL folio_buffers(folio) and BUG_ON panic in ext4_write_inline_data_end().
* +1 Important filesystem/storage path. The bug is in ext4 write_begin/write_end and inline data handling, which is a security-relevant core filesystem write path.
* +1 Distinct persistent storage/data-loss consequence. The commit explicitly says the race could result in data loss, which is separate from the immediate kernel crash symptom.

Negative signals:

* No race/timing penalty awarded. This is a race, but the supplied evidence does not prove a one-shot or externally uncontrollable timing window. Local concurrent filesystem activity can likely be attempted repeatedly.
* No rare configuration penalty awarded. The issue depends on ext4 inline data, but the supplied evidence is not enough to treat the reachable path as rare and attacker-unreachable for scoring purposes.
* No host-admin penalty awarded. The vulnerable operation is normal file write/page-fault activity, not mounting or filesystem administration.

Important non-awarded signals:

* RealUAF not awarded. The patch establishes stale write-state use and a NULL pointer dereference, not object use after free.
* Controlled reclaim not demonstrated.
* Arbitrary or directed kernel write not demonstrated.
* Confidentiality impact not awarded. There is no disclosure path.
* Strong integrity impact not awarded. Data loss is indicated, but no cross-user file overwrite, COW ownership violation, protected-file corruption, or privilege boundary bypass is established.
* Structured stale-object leverage not awarded because no post-free structured object consumption is established.
* Concrete repeated release/refcount not awarded. There is no first release ending object lifetime followed by a second lifetime mutation on the same dangling reference.
* Extra availability was not double-counted from the same crash. The additional point is for possible persistent data loss, not for crash reliability.

3. Reachability analysis

The evaluated attacker is a local user or local process with permission to write to files on an affected ext4 filesystem using inline data. The trigger does not require host administrator privileges once such a filesystem and file are available.

The issue is not network reachable. It is exercised through local filesystem operations, specifically a write path where write_begin prepares inline data and a concurrent path such as ext4_page_mkwrite converts the inode to extent-based storage before write_end completes.

Namespace-scoped privilege is not relevant here. No CAP_NET_ADMIN or namespace-created capability is required for the ordinary file-write trigger. Prior system setup may be required in the sense that the filesystem must be ext4 with inline data support in use, but the supplied patch does not establish that the attacker must perform privileged setup as part of the trigger.

No special hardware is required. Race repeatability is likely practical under local concurrent workload, but exact exploitability beyond crash/data loss is not established.

4. Primitive and severity interpretation

The exact primitive is a race-induced state mismatch between the inline write state prepared by write_begin and the live inode state observed by write_end. The old logic checked ext4_has_inline_data(inode) and EXT4_STATE_MAY_INLINE_DATA at write_end time instead of using the state prepared by write_begin. If inline data was converted to extents in between, write_end could enter extent-based logic even though block buffers had not been allocated, causing NULL folio_buffers(folio) dereference.

This is not an active structured UAF and not fixed stale storage after free. It is a filesystem state/race bug with invalid write-path selection. It is also not cleanup/destructor-only because the vulnerable path is the normal write completion path.

A concrete repeated release/refcount operation does not exist. No first lifetime operation is shown to free an object whose exact same reference later receives another put/free/refcount mutation.

PrimitiveClass=CLASS_C because the demonstrated behavior is primarily invalid-state behavior leading to NULL dereference, BUG_ON panic, and possible data loss. The demonstrated consequence is local DoS and possible filesystem data loss. The speculative consequence would be stronger filesystem corruption, but the patch does not establish a protected-object overwrite, page-cache ownership violation, arbitrary write, or privilege escalation primitive.

ImportantCandidate=NO because no Class A primitive and no strong unresolved Class B lifetime/corruption primitive is established. ManualReview=RECOMMENDED because this is an ext4 race in a core write path with kernel crash and possible data loss. AutoDowngradeAllowed=YES because the issue remains crash/data-loss oriented and lacks unresolved memory-corruption, ownership, authorization, or privilege-escalation leverage.

Likely operational severity is Moderate-like, but not safe to ignore without review in environments using ext4 inline data.

5. One-sentence report phrase

An ext4 inline data race can cause write_end to use extent-based write logic after write_begin prepared an inline write, leading to NULL pointer dereference, BUG_ON panic, or possible data loss for a local user with file write access.

6. Manual review recommendation

MANUAL CHECK RECOMMENDED

Manual review is recommended because this is a local ext4 filesystem write-path race with kernel crash and possible data loss, although the supplied evidence does not establish UAF, arbitrary write, COW ownership violation, or privilege escalation.

7. Call-site confidence

Call-site confidence: high

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

Patch: ext4: use fsdata to track inline data write state and fix race [ Upstream
Commit: bd8d74bd46d09905164255b8635fa58b45f41068
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=bd8d74bd46d09905164255b8635fa58b45f41068

Changed files:
  fs/ext4/ext4.h
  fs/ext4/inline.c
  fs/ext4/inode.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=bd8d74bd46d09905164255b8635fa58b45f41068

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

The original Linux kernel CVE announcement for CVE-2026-92500 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-92500

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:L/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: unknown
  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).