From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-90105][MODERATE REGULAR] vxlan: fix reading neigh ha [ Upstream Date: Fri, 18 Sep 2026 09:17:00 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-90105 X-AL-KERNEL-Priority: MODERATE REGULAR X-AL-KERNEL-Severity: MODERATE REGULAR X-AL-KERNEL-Base-Severity: MODERATE X-AL-KERNEL-KPANIC: NO X-AL-KERNEL-ActionableScore: 4 X-AL-KERNEL-ActionableScore-Lower: 3 X-AL-KERNEL-Commit: ecc8bac11a101e5f6989a14b92e17b51f9761630 List-Id: CVE: CVE-2026-90105 Priority: MODERATE REGULAR AL-KERNEL base severity: MODERATE KPANIC flag: NO Patch: vxlan: fix reading neigh ha [ Upstream Commit: ecc8bac11a101e5f6989a14b92e17b51f9761630 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=ecc8bac11a101e5f6989a14b92e17b51f9761630 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90105 Analysis date: Fri, 18 Sep 2026 09:17:00 -0400 ActionableScore: 4 ActionableScore lower bound: 3 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-2026-90105 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90105 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 REGULAR 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-2026-90105 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-90105 CHECK WITH IMPACT FROM ORIG NN MODERATE Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:A/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:L';CWE-362;CWE-662;CWE-667;Other CVSS 'AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:L';BEST CVSS score: '5';DESCR 'VXLAN ARP and IPv6 neighbour suppression paths read the neighbour hardware address directly while the neighbour entry may be updated concurrently. A remote peer on the same VXLAN or adjacent L2 domain could trigger ARP or ND processing during such an update and cause a partially read MAC address to be used in FDB lookup or in generated ARP and NA replies. For the CVSS the PR:N is used because no local account on the target is required if the attacker already has network reachability to the affected VXLAN or adjacent segment. The issue is not a generic Internet remote bug and normally requires access to the overlay or local data center network where VXLAN neighbour suppression is active. Impact is mainly denial of service or limited integrity impact due to incorrect L2 resolution and traffic disruption. In the paranoid interpretation limited confidentiality impact is also possible if inconsistent address resolution causes traffic to be misdirected, but the patch does not indicate UAF, OOB access, or arbitrary write.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4) YES OOB SIMPLEFIX NETWORK SKB IPV6 DECREASED_TO_MODERATEREG_BASED_ON_FALSEPOSCHECKOFKP NO NO checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=4 ActionableScoreLower=3 PrimitiveClass=CLASS_C RealUAF=NO HostAdminRequired=NO NamespaceScopedPrivilege=NO ImportantCandidate=NO ManualReview=RECOMMENDED AutoDowngradeAllowed=YES 1. ActionableScore * Raw conservative score: 3 * Conservative score: 3 * Raw paranoid score: 4 * Paranoid score: 4 * Final recommended bucket: **Borderline / manual review recommended**:: No floor or cap was applied. 2. Signal breakdown Conservative signals: * +2 Remote or network-adjacent trigger. Attacker-controlled ARP or IPv6 ND traffic can reach the vulnerable VXLAN ARP and neighbour suppression paths when the VXLAN overlay or adjacent L2 domain is reachable. * +1 Weak unsafe object selection / invalid state. A partially read neighbour hardware address can be used as an FDB lookup key and as the source link-layer address in generated ARP or NA replies. * +1 Availability impact. Incorrect L2 resolution can disrupt overlay connectivity or cause wrong ARP/ND responses. * -1 Hard timing. The partial read requires concurrent update of the neighbour hardware address while arp_reduce() or neigh_reduce() reads it, and the attacker may not fully control both sides of that timing window. Paranoid delta: * +1 by withholding the conservative timing penalty. Repeated ARP/ND traffic and neighbour updates in an active VXLAN environment can make the race more reachable than a one-shot teardown race. Important non-awarded signals: * No UAF or lifetime violation is established. * No OOB access is established. * No arbitrary write or controlled kernel memory corruption is established. * No controlled reclaim or object replacement is established. * No authorization or security-boundary bypass is established. * Confidentiality is not awarded as an ActionableScore signal because traffic misdirection is only speculative and no concrete disclosure path is demonstrated. * Integrity is not awarded as a protected-state modification signal because the established effect is incorrect L2 resolution rather than unauthorized modification of protected kernel state. * Extra availability is not awarded because there is no kernel crash, panic, persistent resource exhaustion, or lasting host-wide failure shown. 3. Reachability analysis The most plausible attacker is a peer that can send ARP or IPv6 ND traffic into the affected VXLAN overlay or adjacent L2 environment. This is not a generic public Internet remote issue. It normally requires that VXLAN neighbour suppression or proxy behaviour is already configured and reachable. Host administrator privileges are not required for the packet-level trigger once the VXLAN environment is exposed to the attacker. However, prior privileged setup of the VXLAN device and its policy is normally required by the environment. That prior setup does not make the packet sender host-admin-equivalent. Namespace-scoped privilege is not the main reachability model here. No complete namespace-capability trigger chain is established or needed for the network-peer model. The race requires a neighbour hardware address update to overlap with direct reads of n->ha. The attacker can likely generate repeated ARP/ND events, but the update side may depend on normal neighbour state changes or another actor, so conservative analysis keeps a timing penalty. 4. Primitive and severity interpretation The exact primitive is a non-atomic or unstable read of a neighbour hardware address while the neighbour entry may be updated. The stale or partial value is then used for VXLAN FDB lookup and for generated ARP or IPv6 Neighbor Advertisement packets. This is not an active structured UAF and not fixed stale storage. It is a race-induced invalid value / inconsistent state bug in packet processing. There is no concrete repeated release or refcount mutation. There is no evidence that the first operation can free an object or that a second operation mutates lifetime state through a dangling reference. PrimitiveClass is CLASS_C because the demonstrated effect is invalid network state and potential DoS-like traffic disruption without an independently useful kernel corruption primitive. ImportantCandidate is NO because the patch does not establish memory corruption, ownership violation, authorization bypass, or a practical path to LPE. AutoDowngradeAllowed is YES because the unresolved risk is limited to network correctness and availability impact, with no concrete Important-preserving primitive. Likely operational severity is ordinary to borderline Moderate. It is worth manual review mainly because it is network-adjacent and can affect VXLAN overlay forwarding correctness, not because it shows kernel memory corruption. 5. One-sentence report phrase VXLAN ARP and IPv6 neighbour suppression could read a partially updated neighbour hardware address and use it for FDB lookup or generated ARP/NA replies, allowing a peer on the affected overlay or adjacent L2 domain to cause incorrect L2 resolution and limited DoS-like traffic disruption. 6. Manual review recommendation MANUAL CHECK RECOMMENDED Reason: network-adjacent packet-triggered race in VXLAN forwarding logic, but no UAF, OOB access, arbitrary write, authorization bypass, or demonstrated privilege-escalation primitive is established. 7. Call-site confidence Call-site confidence: high ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: vxlan: fix reading neigh ha [ Upstream Commit: ecc8bac11a101e5f6989a14b92e17b51f9761630 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=ecc8bac11a101e5f6989a14b92e17b51f9761630 Changed files: drivers/net/vxlan/vxlan_core.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=ecc8bac11a101e5f6989a14b92e17b51f9761630 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-90105 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90105 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:A/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:L The Best / paranoid CVSS vector: AV:A/AC:H/PR:N/UI:N/S:U/C:L/I:L/A:L The Best / paranoid CVSS score: 5 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: 3 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: NO Published priority for this report (same as in Subject): MODERATE REGULAR 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).