* [CVE-2026-93105][MODERATE 7.0] esp: do not unref managed frag pages in esp_ssg_unref() [ Upstream
@ 2026-09-18 0:56 21% AL-KERNEL
0 siblings, 0 replies; 1+ results
From: AL-KERNEL @ 2026-09-18 0:56 UTC (permalink / raw)
To: kernel-cve
CVE: CVE-2026-93105
Priority: MODERATE 7.0
AL-KERNEL base severity: MODERATE
KPANIC flag: YES
Patch: esp: do not unref managed frag pages in esp_ssg_unref() [ Upstream
Commit: 26b6b14c7a0368e317a1e9fb5144ebe6f8d495cf
Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=26b6b14c7a0368e317a1e9fb5144ebe6f8d495cf
Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93105
Analysis date: Thu, 17 Sep 2026 20:56:41 -0400
ActionableScore: 6
ActionableScore lower bound: 5
Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review
Manual review required: YES
Summary:
No one-sentence report phrase was found.
======================================================================
ABOUT THIS REPORT
======================================================================
The original Linux kernel CVE announcement for CVE-2026-93105 is available here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93105
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-2026-93105 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-93105 CHECK WITH IMPACT FROM ORIG NN IMPORTANT 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-911;*CWE-416;*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 'ESP output can incorrectly unreference page fragments when an skb uses managed zerocopy frags. The skb does not own per-frag page references in this mode, but esp_ssg_unref can still call skb_page_unref on those pages after the AEAD operation. A local user may be able to reach this path with io_uring SEND_ZC and registered buffers when outbound traffic is processed by an IPsec ESP policy. For the CVSS the PR:L is used because local code execution is needed, while full administrator privileges may not be required if IPsec is already configured by the system. The issue is not directly network reachable from a remote peer. Impact is at least local denial of service and in the paranoid case may include confidentiality and integrity impact due to page refcount underflow and possible use-after-free.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 6) YES REMOTE OOB DANGER ERRORPATH NETWORK SKB LINUS IPV6 IO_URING KPANIC DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFKP NO NO checked
======================================================================
ACTIONABLESCORE ANALYSIS
======================================================================
ActionableScore=6
ActionableScoreLower=5
PrimitiveClass=CLASS_B
RealUAF=YES
HostAdminRequired=NO
NamespaceScopedPrivilege=UNKNOWN
ImportantCandidate=YES
ManualReview=REQUIRED
AutoDowngradeAllowed=NO
1. ActionableScore
* Raw conservative score: 5
* Conservative score: 5. No floor or cap changed the score.
* Raw paranoid score: 6
* Paranoid score: 6. No floor or cap changed the score.
* Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**::
2. Signal breakdown
Conservative positive signals:
* +2 Real UAF / lifetime violation. esp_ssg_unref() may call skb_page_unref() on managed skb frags for which the skb does not own a per-frag page reference. The commit explicitly says this can underflow the page reference count and free a page that is still in use.
* +1 Reliable kernel crash / strong DoS. Premature page free or page refcount underflow in the networking ESP output path is a realistic kernel-fault and host-DoS condition.
* +1 Local below-host-admin trigger. The described path can be reached by a local user using io_uring SEND_ZC with registered buffers when outbound packets are processed by an existing IPsec ESP policy.
* +1 Local high-control kernel data-plane API. Registered zerocopy buffers give the local process meaningful influence over the pages attached as managed frags and over the skb/page ownership state that reaches ESP.
Conservative total: 2 + 1 + 1 + 1 = 5.
Paranoid additional signal:
* +1 Attacker-influenced repeatable lifetime manipulation. A local process can plausibly submit repeated zerocopy sends using attacker-chosen registered buffers through the ESP output path, repeatedly exercising the incorrect page-refcount transition. This is stronger than a one-shot cleanup crash but does not prove controlled reclaim.
Paranoid total: 5 + 1 = 6.
Material non-awarded signals:
* Remote reachable +2 not awarded. The vulnerable operation is not directly triggered by remote inbound traffic. It is reached from local outbound zerocopy send activity through ESP.
* Strong exploitation leverage +2 not awarded. Controlled reclaim, arbitrary kernel read/write, callback control, credential overwrite, or demonstrated LPE is not established.
* Concrete repeated release/refcount +1 not awarded. The patch proves an unbalanced page-reference drop, but the supplied context does not show the exact sequence of first lifetime-ending release, same surviving dangling reference, and a concrete second lifetime mutation required by the dedicated repeated-release rule.
* Weak LPE +1 not separately awarded. The unresolved exploitation concern is already represented by the page lifetime violation and repeatable controlled data-plane signal.
* Confidentiality and integrity impact are not separately awarded. The patch supports page lifetime corruption and possible UAF, but not a concrete disclosure path or unauthorized write primitive.
* Extra availability +1 not awarded. The same crash or page-refcount failure is already counted as reliable DoS.
* Race/timing penalty not applied. The bug is not primarily a narrow race. It is an ownership/refcounting error on a reachable completion/error path.
* Rare subsystem penalty not applied. IPsec ESP and io_uring zerocopy are configuration-dependent, but the supplied evidence does not prove a rare or physical-only prerequisite.
Repeated-release checklist:
* First release can end lifetime: not established as a visible first release in the dedicated repeated-release pattern.
* Exact same reference survives: not established in the required first-release then second-release form.
* Second operation mutates lifetime/refcount: esp_ssg_unref() does mutate page refcount, but as an unbalanced release rather than a proven second release on the same dangling reference.
* +1 Concrete repeated release/refcount awarded: no.
* Paranoid repeated-refcount leverage +1 awarded: no, because the dedicated repeated-release mechanics are not fully shown.
* FLOOR 5 applied: not needed. Raw conservative score is already 5 due to real page lifetime corruption and local high-control reachability.
3. Reachability analysis
The evaluated attacker is a local user or local process on the host. The likely trigger is io_uring SEND_ZC with registered buffers, producing managed skb frags that later pass through IPsec ESP output handling.
Host-admin-equivalent privilege is not required for the send operation itself when an ESP policy is already configured for the relevant traffic. However, creating or changing the IPsec/XFRM policy would normally require administrative network privileges. This means there is a prior-environment dependency, but not necessarily host-admin privilege for the triggering local process.
Namespace-scoped privilege is unresolved. XFRM and networking objects can involve network namespaces, but the supplied context does not prove that the complete setup and trigger chain is attacker-creatable inside an unprivileged namespace.
The issue is not directly network reachable from a remote peer. A remote peer may receive or influence traffic patterns only after the local host has generated outbound ESP traffic, but the vulnerable page ownership state is created locally.
No special hardware requirement is established. No hard race is required. Repeated attempts appear plausible because a local process can repeatedly submit zerocopy sends, assuming the required ESP path is active.
4. Primitive and severity interpretation
The exact primitive is an incorrect page-reference release on managed zerocopy skb fragments. Managed frags are owned by the zerocopy ubuf, and the skb does not own a separate page reference. esp_ssg_unref() nevertheless drops a page reference unless guarded by skb_zcopy_managed().
This is an active page lifetime and ownership violation rather than a pure NULL dereference or passive bookkeeping error. The commit explicitly states that the result can be page refcount underflow and freeing a page that is still in use, which directly implies possible post-lifetime page use.
This is not merely ordinary stale cleanup dereference. The problematic operation actively mutates page lifetime/refcount state. At the same time, the dedicated repeated-release exception is not fully established because the supplied context does not show a concrete first-release then surviving exact dangling reference then second-release sequence.
PrimitiveClass is CLASS_B. The patch establishes a real potentially exploitable kernel lifetime/refcount primitive, but it does not establish CLASS_A evidence such as controlled reclaim, attacker-shaped replacement contents, arbitrary write, protected shared-page overwrite, or demonstrated local privilege escalation.
Demonstrated consequence is local kernel memory lifetime corruption with likely crash/DoS. Speculative consequence is broader exploitation through premature page reuse while still referenced by zerocopy state. That concern is technically meaningful and should be manually reviewed, but it is not enough by itself to claim proven LPE or arbitrary kernel R/W.
ImportantCandidate is YES because this is page refcount underflow with possible free-while-in-use in a user-influenced networking zerocopy path. AutoDowngradeAllowed is NO because automatic closure would discard unresolved exploitation leverage from a concrete kernel page lifetime corruption primitive.
Likely operational severity is actionable Moderate at minimum, with Strong Important candidate review warranted if affected deployments combine untrusted local users, io_uring zerocopy, and active IPsec ESP policies.
5. One-sentence report phrase
A local user may trigger a page refcount underflow and possible use-after-free in ESP output by sending io_uring zerocopy traffic with managed skb frags through an IPsec ESP policy, causing at least kernel DoS and requiring manual review for broader page-lifetime corruption impact.
6. Manual review recommendation
MANUAL CHECK REQUIRED
Reason: the patch establishes a real page ownership and lifetime violation where ESP drops a page reference that the skb never owned, potentially freeing a page still in use. Controlled reclaim or LPE is not proven, but the primitive is too security-relevant for automatic downgrade.
7. Call-site confidence
Call-site confidence: medium
======================================================================
UPSTREAM PATCH SUMMARY
======================================================================
Patch: esp: do not unref managed frag pages in esp_ssg_unref() [ Upstream
Commit: 26b6b14c7a0368e317a1e9fb5144ebe6f8d495cf
Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=26b6b14c7a0368e317a1e9fb5144ebe6f8d495cf
Changed files:
net/ipv4/esp4.c
net/ipv6/esp6.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=26b6b14c7a0368e317a1e9fb5144ebe6f8d495cf
======================================================================
DETAILED REPORT METHODOLOGY
======================================================================
The original Linux kernel CVE announcement for CVE-2026-93105 can be found here:
https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93105
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: 5
Paranoid ActionableScore: 6
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 [email protected] (and both
send reply to CVE record itself too and see "reply" button below for howto reply).
^ permalink raw reply [relevance 21%]
Results 1-1 of 1 | reverse | sort options + mbox downloads above
-- links below jump to the message on this page --
2026-09-18 0:56 21% [CVE-2026-93105][MODERATE 7.0] esp: do not unref managed frag pages in esp_ssg_unref() [ Upstream AL-KERNEL
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox