From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-99999][IMPORTANT] ipv4: validate checksum_start before completing checksum Date: Fri, 16 Oct 2026 00:38:45 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-99999 X-AL-KERNEL-Priority: IMPORTANT X-AL-KERNEL-Severity: IMPORTANT X-AL-KERNEL-Base-Severity: IMPORTANT X-AL-KERNEL-KPANIC: NO X-AL-KERNEL-ActionableScore: 5 X-AL-KERNEL-ActionableScore-Lower: unknown X-AL-KERNEL-Commit: a464d950f31c89565cbaae038cc60db04fe01956 List-Id: CVE: CVE-2026-99999 Priority: IMPORTANT AL-KERNEL base severity: IMPORTANT KPANIC flag: NO Patch: ipv4: validate checksum_start before completing checksum Commit: a464d950f31c89565cbaae038cc60db04fe01956 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=a464d950f31c89565cbaae038cc60db04fe01956 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99999 Analysis date: Fri, 16 Oct 2026 00:38:45 -0400 ActionableScore: 5 ActionableScore lower bound: unknown 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-2026-99999 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99999 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-2026-99999 IMPORTANT https://www.kernelcve.org/list/?q=CVE-2026-99999 CHECK WITH IMPACT FROM ORIG NN IMPORTANT | AEGIS IMPACT MODERATE_KPANIC AEGIS_AS=1 LEGACY=IMPORTANT RESULT=DIFF INDEPENDENT_LEGACY=IMPORTANT TEMP_KPANIC_MERGE=IMPORTANT_WINS_DISAGREEMENT PRE_MERGE_LEGACY=IMPORTANT (independent Legacy and Aegis differ; FINAL: IMPORTANT, Aegis: MODERATE_KPANIC) Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:A/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:H';*CWE-20;CWE-1285;CWE-823;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:L/A:L';BEST CVSS score: '6.8';DESCR 'An IPv4 CHECKSUM_PARTIAL skb with an invalid csum_start can reach skb_checksum_help, which may write the computed checksum into the IPv4 header instead of the intended transport header. An attacker able to inject packets together with checksum offload metadata through TUN, packet sockets, or a virtual networking path may alter header fields before fragmentation, NFQUEUE, XFRM, routing, or filtering completes. This is not directly reachable from ordinary Internet packets because checksum metadata is local kernel metadata and is not transmitted on the wire. Guest to host or namespace to host reachability is plausible when the relevant virtual interface and offload path accepts attacker controlled metadata, but the exact boundary depends on deployment and device configuration. For the CVSS the PR:L reflects the paranoid scenario where a reduced capability local process, container, or guest can submit crafted checksum metadata without full host administrative control. The demonstrated primitive is corruption of fields inside the crafted skb header rather than a proven out of bounds write or UAF. Impact can include packet handling policy inconsistency and denial of service. Host wide crash or privilege escalation is not proven, so CIA:LLH is retained only as a manual review sensitive worst case. Automated impact decision path. The original SimpleNet classifier selected IMPORTANT. The primary LLM estimated paranoid CVSS 6.8 and the production ActionableScore returned 5 with recommendation Actionable Moderate / Strong Important candidate requiring manual review. The production rules then evaluated CVSS thresholds and ActionableScore thresholds. Recorded rule transitions were KP and FD specialist checks stage IMPORTANT to IMPORTANT because a specific evidence based protection or manual review veto was applied. final reconciliation stage IMPORTANT to IMPORTANT because KP specialist did not establish a qualifying panic. The legacy KP specialist produced a review artifact and its decision was processed by the applicable KP rules. The legacy FD or DH downgrade specialist produced a review artifact and its decision was processed by the applicable downgrade and preservation rules. Exact triggered rule markers were V33_ACTIVE_STATE_BOUNDARY_CANDIDATE and IMPORTANT_CANDIDATE_MANUAL_REVIEW and KP_ONLY_DOWNGRADE_VETO and V34_DOWNSTREAM_METADATA_KP_VETO and MANUAL_REVIEW_REQUIRED and V34_DOWNSTREAM_METADATA_DH_VETO and MANUAL_REVIEW_REQUIRED and TEMP_KPANIC_MERGE_IMPORTANT_WINS and KPANIC_UNSUPPORTED_REMOVED. The independent weaker Aegis shadow evaluated CVSS 2.5 and ActionableScore 1 with primitive class CLASS_B and manual review REQUIRED. Its independently reconciled impact was MODERATE_KPANIC. Cross model aggregation changed the production result under an enabled evidence gated merge rule. The final reported impact is IMPORTANT.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5) MAYBE REMOTE SKB PACKET V33_ACTIVE_STATE_BOUNDARY_CANDIDATE IMPORTANT_CANDIDATE_MANUAL_REVIEW KP_ONLY_DOWNGRADE_VETO V34_DOWNSTREAM_METADATA_KP_VETO MANUAL_REVIEW_REQUIRED V34_DOWNSTREAM_METADATA_DH_VETO MANUAL_REVIEW_REQUIRED TEMP_KPANIC_MERGE_IMPORTANT_WINS KPANIC_UNSUPPORTED_REMOVED - - guess ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=5 PrimitiveClass=CLASS_B RealUAF=NO HostAdminRequired=UNKNOWN NamespaceScopedPrivilege=UNKNOWN ImportantCandidate=YES ManualReview=REQUIRED AutoDowngradeAllowed=NO RequiresHostUID0=NO AuthorizationDecisionBypass=NO ProtectedResourceImpact=NONE B4PendingReview=NO B4PendingConfirmed=NO 1. ActionableScore * Raw conservative score: 3 * Conservative score: 5 * Raw paranoid score: 4 * Paranoid score: 5 * Applied conservative V33 manual-review safety floor: 3 -> 5 * Applied paranoid V33 manual-review safety floor: 4 -> 5 * Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**:: The score preserves this as an Important review candidate, not as a proven Important vulnerability. 2. Signal breakdown Conservative positive signals: * +1 Constrained kernel buffer overwrite. `skb_checksum_help()` performs a two-byte checksum write at `csum_start + csum_offset`. Invalid metadata can place this write inside the active IPv4 header. * +1 Active structured-state consumption. The modified header is subsequently consumed by fragmentation, bridge netfilter, NFQUEUE, XFRM, or continued packet processing. This is stronger than corruption detected and discarded before use. * +1 Common networking packet-processing paths. The affected consumers include IPv4 fragmentation and netfilter-related paths rather than diagnostic-only code. Paranoid additional signal: * +1 Credible but unresolved trust boundary. TUN with `virtio_net_hdr`, packet sockets with checksum or GSO metadata, delegated TUN descriptors, namespaces, and virtual networking provide technically credible sources of attacker-influenced checksum placement. The complete guest-to-host or namespace-to-host chain is not established. Negative signals: * No negative privilege score. Host-admin-equivalent privilege is not established, but complete below-host-admin reachability is also unresolved. * No race penalty. The defect is a validation-to-use problem, not a timing race. * No niche-subsystem penalty. TUN, packet sockets, IPv4, netfilter, and XFRM are established kernel interfaces. Important non-awarded signals: * No UAF, stale lifetime operation, or controlled reclaim is demonstrated. * No heap or slab OOB write is demonstrated. Existing `skb_checksum_help()` semantics constrain the final checksum location to writable linear skb data, while this patch adds the stricter semantic requirement that it follow the IPv4 header. * No arbitrary target or arbitrary payload is established. The write is two bytes and the value is derived from checksum computation, although packet contents can influence it. * No reliable kernel panic is established. The MODERATE_KPANIC marker is retained only as an operational review concern. * No confidentiality or integrity bonus for a protected victim is awarded. Header corruption is concrete, but unauthorized modification of another principal or protected host state is not yet demonstrated. * No remote-reachability bonus is awarded. Ordinary network packets do not carry `csum_start` or `csum_offset` metadata. * NEW52 B5-R does not apply. There is no retry-accumulated heap OOB write. * NEW53 B3-PENDING does not apply. No shared or file-backed page modification is established. * NEW54 B4-PENDING does not apply. The patch restores input validation, not an authorization DENY-to-ALLOW guard. 3. Reachability analysis The immediate input is an skb marked `CHECKSUM_PARTIAL` with attacker-influenced checksum metadata. The commit identifies TUN GSO through `virtio_net_hdr` and packet-socket GSO or checksum offload as originating interfaces. A process using packet sockets generally needs `CAP_NET_RAW` in the governing user namespace. Creating or configuring TUN normally involves `CAP_NET_ADMIN`, although a previously created TUN descriptor can be delegated to a less-privileged process. Numeric host UID 0 is therefore not inherently required. The complete namespace-scoped trigger is unresolved. It is not shown that an attacker-created user and network namespace can make the malformed skb reach a more privileged host networking context. Similarly, a guest boundary is credible because virtual-network headers carry checksum placement metadata, but the supplied evidence does not prove that a malicious guest can bypass all frontend validation and exercise these exact host consumers. No special hardware, boot option, or race timing is required by the vulnerable operation itself. Some prior interface or virtual-network setup may be required, depending on the attack model. 4. Primitive and severity interpretation The concrete primitive is a fixed-width two-byte write into an incorrect location within active skb data. Its destination is calculated from `csum_start + csum_offset`. The new check rejects `csum_start` values before `network_header + ip_hdrlen(skb)`. The exact modified IPv4 field depends on both metadata values. Potentially affected bytes can belong to active length, fragmentation, protocol, checksum, or address fields. This does not establish an OOB kernel-memory write because the existing checksum helper performs final range and writability handling. It does establish semantic corruption of the network header. The corrupted header is used after the write: * `ip_do_fragment()` reads and fragments the packet after checksum completion. * `nf_br_ip_fragment()` continues bridge fragmentation processing. * `nf_queue_checksum_help()` prepares the packet for NFQUEUE handling. * `checksum_tg()` continues netfilter processing after checksum completion. * `xfrm_output()` passes the skb into later XFRM output processing. This satisfies the V33 active-state candidate mechanics: 1. Attacker-influenced checksum placement is credible through identified metadata-capable interfaces. 2. A real write into active structured state is established. 3. Privileged kernel networking consumers use the modified header. 4. A guest, namespace, or delegated-interface boundary remains credible but unresolved. The primitive is CLASS_B rather than CLASS_A. The write and later consumption are concrete, but protected-state modification, host escape, OOB access, arbitrary kernel write, or privilege escalation is not proven. `ImportantCandidate=YES` preserves the unresolved cross-boundary integrity chain for review. `AutoDowngradeAllowed=NO` prevents treating the write as a harmless edit of an attacker-owned packet before the relevant consumers and trust boundaries are checked. This does not assert confirmed host panic or Important severity. 5. One-sentence report phrase Invalid CHECKSUM_PARTIAL metadata can make skb_checksum_help overwrite two bytes of an active IPv4 header before privileged fragmentation, netfilter, NFQUEUE, or XFRM processing, creating a mandatory cross-boundary integrity review candidate without proving OOB access, host panic, or privilege escalation. 6. Manual review recommendation MANUAL CHECK REQUIRED Verify the complete TUN, packet-socket, namespace, and virtio guest-to-host call paths, identify which IPv4 fields can be selected through `csum_start + csum_offset`, and determine whether later consumers can cause policy bypass, host-wide failure, or cross-boundary packet effects. 7. Call-site confidence Call-site confidence: high ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: ipv4: validate checksum_start before completing checksum Commit: a464d950f31c89565cbaae038cc60db04fe01956 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=a464d950f31c89565cbaae038cc60db04fe01956 Commit description: If a packet with bad checksum metadata gets into the ipv4 stack, skb_checksum_help can corrupt the network header and cause a bunch of mischief. This was discovered and reported by Paulos, and has been reporoduced by others independently since. We really shouldn't allow such packets in, but as a defence in depth measure, let's also check before we complete the checksum. A more complete validation at input is forthcoming, but needs more work. Cc: stable@vger.kernel.org Fixes: f43798c ("tun: Allow GSO using virtio_net_hdr") Fixes: bfd5f4a ("packet: Add GSO/csum offload support.") Reported-by: Paulos Yibelo Closes: https://lore.kernel.org/netdev/20260922030310.8684-2-habte.yibelo@gmail.com/ Signed-off-by: Michael S. Tsirkin Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/0a012b4923e189c4c593ef4f471e5ff0edbe9030.1791412497.git.mst@redhat.com Signed-off-by: Jakub Kicinski Signed-off-by: Greg Kroah-Hartman Changed files: include/net/ip.h net/bridge/netfilter/nf_conntrack_bridge.c net/ipv4/ip_output.c net/netfilter/nfnetlink_queue.c net/ip.h net/sock.h net/tcp_states.h net/netfilter/nf_queue.h net/netfilter/xt_CHECKSUM.c net/xfrm/xfrm_output.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=a464d950f31c89565cbaae038cc60db04fe01956 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-99999 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-99999 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:L/PR:L/UI:N/S:U/C:N/I:L/A:L The Best / paranoid CVSS vector: AV:A/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:H The Best / paranoid CVSS score: 6.8 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: 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: IMPORTANT KPANIC detected for this report: NO 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).