[CVE-2026-93137][MODERATE 7.0] bpf: Fix use-after-free on mm_struct in bpf_find_vma() [ Upstream

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

CVE: CVE-2026-93137
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: bpf: Fix use-after-free on mm_struct in bpf_find_vma() [ Upstream
Commit: db347840d6b6ee9bb9b8e4a985d4b4419f9f3200
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=db347840d6b6ee9bb9b8e4a985d4b4419f9f3200
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93137
Analysis date: Fri, 18 Sep 2026 03:44:29 -0400
ActionableScore: 5
ActionableScore lower bound: 4
Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review
Manual review required: YES

Summary:
A race in `bpf_find_vma()` can use a foreign task `mm_struct` after concurrent `exit_mm()` frees it, allowing a local BPF-capable attacker to trigger a kernel crash and requiring manual review for possible exploitation beyond DoS.

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

The original Linux kernel CVE announcement for CVE-2026-93137 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93137

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-2026-93137	MODERATE	https://www.kernelcve.org/list/?q=CVE-2026-93137 CHECK WITH IMPACT FROM ORIG NN MODERATE	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-416;CWE-362;*CWE-664;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'A use-after-free can occur in bpf_find_vma() because the helper reads the mm_struct pointer from a foreign task without holding a reference. A concurrent exit_mm() on that task can clear task mm and free the mm_struct before mmap_read_trylock() is called, so the BPF path may operate on freed kernel memory. For the CVSS the PR:L is used because exploitation requires local ability to run a relevant BPF program or a delegated BPF capable context, but it does not require network reachability. The attack complexity is high because the bug depends on winning a race against task exit. Impact is at least a local denial of service via kernel crash and in worst case may allow confidentiality or integrity impact due to use-after-free memory corruption, so this should be reviewed manually.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5)	YES	READ LOCK DEADLOCK INIT BPF RACE UAF HARDWARE LINUS  KPANIC INCREASED_TO_MODERATE7_FROM_MODERATE_BASED_ON_GUESSCVSS	YES	NO	checked

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

ActionableScore=5
ActionableScoreLower=4
PrimitiveClass=CLASS_B
RealUAF=YES
HostAdminRequired=UNKNOWN
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO

## 1. ActionableScore

* Raw conservative score: 2
* Conservative score: 4
* Applied conservative floor/cap: Real UAF floor 4 applied because a concrete post-lifetime `mm_struct` use is established and host-admin-equivalent privilege is not positively established.
* Raw paranoid score: 4
* Paranoid score: 5
* Applied paranoid floor/cap: Real UAF floor 5 applied because the stale object is an active structured `mm_struct`, the stale path performs nontrivial locking and VMA lookup behavior, and the issue is not merely cleanup/destructor-only.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::

## 2. Signal breakdown

Conservative signals:

* +2 Real UAF / active lifetime violation: `bpf_find_vma()` reads `task->mm` for a foreign task without holding an `mm_struct` reference, while concurrent `exit_mm()` can drop the final reference and free it.
* +1 Reliable kernel crash / strong DoS: the stale `mm` is passed to `mmap_read_trylock(mm)`, which can operate on freed kernel memory and fault or corrupt lock state.
* -1 BPF privilege-gated control-plane reachability: triggering requires running a relevant BPF program/helper path, which is normally gated by BPF-related privileges or policy.
* No race penalty awarded: although the bug is a race, the supplied evidence does not prove an externally uncontrollable one-shot condition, and repeated task-exit attempts appear plausible.
* Raw conservative total: 2.
* Floor 4 applied due to real active-object UAF with host-admin not positively established.

Paranoid additional signals:

* +1 Attacker-influenced active-object lifetime: a BPF-capable local attacker can plausibly target a foreign task and repeatedly drive task exit timing to hit the vulnerable window.
* +1 Structured stale-object exploitation leverage: the stale object is `mm_struct`, and the stale path may consume active lock and VMA-management state rather than only passing a pointer to cleanup.
* -1 BPF privilege-gated control-plane reachability retained.
* Raw paranoid total: 4.
* Floor 5 applied because the UAF is active structured use and not a constrained fixed-buffer or cleanup-only stale dereference.

Non-awarded signals:

* Strong LPE leverage +2 not awarded: no controlled reclaim, object replacement, arbitrary write, callback control, or demonstrated privilege escalation is supplied.
* Confidentiality and integrity points not awarded directly: UAF exists, but no concrete information leak or unauthorized write primitive is demonstrated.
* Concrete repeated release/refcount +1 not awarded: the patch does not show first release of the same reference followed by a second lifetime-mutating operation on that exact dangling reference.
* Paranoid repeated-refcount +1 not awarded: no second stale `put` or refcount mutation is present.
* Extra availability +1 not awarded: the same crash/fault is already counted as reliable DoS.
* Remote reachability +2 not awarded: the issue is local BPF/helper reachable, not network-triggered.
* Class A not awarded: no strong controlled corruption or demonstrated exploit chain is established.

Repeated release/refcount check:

* First release can end lifetime: yes, `exit_mm()` can call `mmput(mm)` and free `mm_struct`.
* Exact same reference survives: a stale local `mm` pointer survives in `bpf_find_vma()`.
* Second operation is lifetime/refcount mutation: no, the stale operation is `mmap_read_trylock(mm)` and later VMA lookup behavior, not `mmput()` or another refcount release.
* Concrete repeated release/refcount +1 awarded: no.
* Paranoid repeated-refcount leverage +1 awarded: no.
* FLOOR 5 applied: yes, but due to active structured stale use, not repeated release/refcount.

## 3. Reachability analysis

The immediate trigger is local execution of a BPF program that can call `bpf_find_vma()` on a foreign task while that task concurrently exits. The attacker would need access to a relevant BPF program type and helper context. In typical modern deployments, BPF loading and attachment are restricted by `CAP_BPF`, `CAP_PERFMON`, `CAP_SYS_ADMIN`, LSM policy, sysctl settings, or distribution defaults.

HostAdminRequired is UNKNOWN because the supplied evidence establishes BPF-program reachability but does not prove whether the complete load, attach, and execution chain requires full host-admin-equivalent privilege in the evaluated product. It also does not prove ordinary unprivileged reachability.

NamespaceScopedPrivilege is UNKNOWN. The patch does not establish that all required BPF load, attach, target-task selection, and helper execution operations are available inside attacker-created namespaces.

Prior host setup may exist depending on BPF policy and program type, but it is not fully resolved from the supplied patch. No special hardware is required. `CONFIG_MMU` must be enabled for the foreign-task path, which is normal for most general-purpose Linux systems. The race is against `exit_mm()`, but repeated attempts appear plausible if the attacker can run the BPF path and control or induce foreign task exits.

## 4. Primitive and severity interpretation

The exact primitive is a real use-after-free of `struct mm_struct`. The object is an active structured kernel object, not fixed stale storage. The stale use is not cleanup/destructor-only: `mmap_read_trylock(mm)` operates on the stale object, and if successful the path proceeds toward `find_vma(mm, start)`.

This is PrimitiveClass CLASS_B. It is a real lifetime/corruption primitive that requires exploitation-leverage analysis, but it is not CLASS_A because no controlled reclaim, arbitrary read/write, callback hijack, credential manipulation, or demonstrated LPE chain is shown.

The demonstrated consequence is local kernel crash or host DoS. The speculative consequence is possible confidentiality or integrity impact if the stale `mm_struct` storage can be reclaimed or shaped in a useful way, but that stronger primitive is not established by the patch.

ImportantCandidate is YES because this is an active structured UAF in a BPF helper involving `mm_struct` lifetime, and the impact beyond crash remains unresolved. AutoDowngradeAllowed is NO because automatically closing it as ordinary crash-only would discard a concrete active lifetime primitive that needs manual review. The likely operational severity is actionable Moderate at minimum, with Important-class review justified.

## 5. One-sentence report phrase

A race in `bpf_find_vma()` can use a foreign task `mm_struct` after concurrent `exit_mm()` frees it, allowing a local BPF-capable attacker to trigger a kernel crash and requiring manual review for possible exploitation beyond DoS.

## 6. Manual review recommendation

MANUAL CHECK REQUIRED

Reason: real active-object UAF on `mm_struct` is established, the stale use is not cleanup-only, and BPF/helper reachability plus possible post-free structured processing leave Important-vs-Moderate exploitation leverage unresolved.

## 7. Call-site confidence

Call-site confidence: high

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

Patch: bpf: Fix use-after-free on mm_struct in bpf_find_vma() [ Upstream
Commit: db347840d6b6ee9bb9b8e4a985d4b4419f9f3200
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=db347840d6b6ee9bb9b8e4a985d4b4419f9f3200

Changed files:
  kernel/bpf/task_iter.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=db347840d6b6ee9bb9b8e4a985d4b4419f9f3200

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

The original Linux kernel CVE announcement for CVE-2026-93137 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93137

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: 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).