From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-98796][MODERATE 7.0] net: ethernet: mtk_eth_soc: allocate dummy netdev sooner [ Upstream Date: Fri, 23 Oct 2026 20:24:54 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-98796 X-AL-KERNEL-Priority: MODERATE 7.0 X-AL-KERNEL-Severity: MODERATE 7.0 X-AL-KERNEL-Base-Severity: MODERATE X-AL-KERNEL-KPANIC: YES X-AL-KERNEL-ActionableScore: 5 X-AL-KERNEL-ActionableScore-Lower: 4 X-AL-KERNEL-Commit: 42001839e425b2564b7297be579073a13a8fddf6 List-Id: CVE: CVE-2026-98796 Priority: MODERATE 7.0 AL-KERNEL base severity: MODERATE KPANIC flag: YES Patch: net: ethernet: mtk_eth_soc: allocate dummy netdev sooner [ Upstream Commit: 42001839e425b2564b7297be579073a13a8fddf6 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=42001839e425b2564b7297be579073a13a8fddf6 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98796 Analysis date: Fri, 23 Oct 2026 20:24:54 -0400 ActionableScore: 5 ActionableScore lower bound: 4 Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review Manual review required: YES Summary: A probe-time initialization race in the Mediatek Ethernet driver can expose a netdev before its shared NAPI state exists and can leave a frame-engine reset worker queued across teardown, causing a NULL dereference or potential asynchronous use after free and requiring manual review. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-98796 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98796 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-2026-98796 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-98796 CHECK WITH IMPACT FROM ORIG NN LOW | AEGIS IMPACT MODERATE_KPANIC AEGIS_AS=1 LEGACY=MODERATE_KPANIC RESULT=SAME (final reconciliation agrees; DECISION: both MODERATE_KPANIC) 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-362;CWE-665;**CWE-476;*CWE-416;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '7';DESCR 'A race in the Mediatek Ethernet driver allows a netdev to become visible and be opened before eth dummy_dev and the shared TX and RX NAPI structures are initialized. On NETSYS v2 and later systems this can make the XDP RX queue registration observe a NULL netdev and fail the interface open. On NETSYS v1 systems napi_enable can dereference a NULL napi dev pointer and may crash the kernel. A concurrently opened interface can also queue the frame engine reset worker before probe failure cleanup, creating a plausible use after free risk if the worker runs after shared driver state is released. For the CVSS the PR:L value in the paranoid score reflects environments where a reduced capability service, container root, or a user with delegated CAP_NET_ADMIN can bring the interface up without full host administrative control. AC:H is used because successful triggering depends on winning the short race between netdev registration, automatic interface activation, and completion or failure of driver probe. The issue is not directly network reachable because packet delivery alone does not perform the required local interface open operation. Impact is at least local denial of service through a NULL pointer dereference and kernel crash. In the worst still defensible case the asynchronous worker lifetime issue may permit broader memory corruption and should be reviewed manually. Automated impact decision path. The original SimpleNet classifier selected MODERATE_KPANIC. The primary LLM estimated paranoid CVSS 7 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 CVSS and initial triage stage MODERATE_KPANIC to MODERATE_KPANIC because the CVSS score, vector or bug class threshold rule was applied. 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 has no review artifact for this run. Exact triggered rule markers were INCREASED_TO_HIGH_BASED_ON_GUESSCVSS and LOWERED_FROM_HIGH_BASED_ON_NULLPTR. The independent weaker Aegis shadow evaluated CVSS 4.7 and ActionableScore 1 with primitive class CLASS_C and manual review REQUIRED. Its independently reconciled impact was MODERATE_KPANIC. The shadow agreed with the final production impact and aggregation did not change that impact. The final reported impact is MODERATE_KPANIC.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5) MAYBE REMOTE INIT NOMEMCHK ERRORPATH NETWORK DMAorINTERRUPT HARDWARE LINUS KPANIC KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS KPANIC LOWERED_FROM_HIGH_BASED_ON_NULLPTR - - checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=5 ActionableScoreLower=4 PrimitiveClass=CLASS_B RealUAF=YES HostAdminRequired=UNKNOWN NamespaceScopedPrivilege=UNKNOWN ImportantCandidate=YES ManualReview=REQUIRED AutoDowngradeAllowed=NO RequiresHostUID0=UNKNOWN AuthorizationDecisionBypass=NO ProtectedResourceImpact=NONE B4PendingReview=NO B4PendingConfirmed=NO 1. ActionableScore * Raw conservative score: 3 * Conservative score: 4 * Raw paranoid score: 4 * Paranoid score: 5 * Conservative floor applied: 3 -> 4 under the real UAF floor because queued asynchronous work can outlive the shared driver state. * Paranoid floor applied: 4 -> 5 because the stale reset worker performs nontrivial active processing against structured driver and hardware state rather than an immediate cleanup-only dereference. * Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**:: 2. Signal breakdown Conservative signals: * +2 Real UAF or lifetime violation. A concurrently opened netdev can queue `eth->pending_work`, while the probe failure path subsequently tears down shared state. Adding `cancel_work_sync()` before teardown directly implies that the worker could otherwise execute after its required state loses its lifetime. * +1 Reliable kernel crash or strong DoS. On NETSYS v1, `napi_enable()` can dereference an uninitialized NULL `napi->dev`. The async stale worker also supports a realistic kernel fault, but it is not counted as a second availability point. * Raw conservative score: 2 + 1 = 3. * Floor 4 applies because this is not merely a fixed stale buffer. The queued worker operates on active driver state, and host-admin-equivalent privilege is not positively established. Paranoid signals: * +2 Concrete asynchronous post-teardown lifetime violation. * +1 Kernel crash or strong DoS. * +1 Structured stale-object concern. The reset worker takes RTNL, consumes shared `eth` state, and drives hardware after that state may begin teardown. This is broader than an immediate NULL dereference, although controlled replacement is not demonstrated. * Raw paranoid score: 2 + 1 + 1 = 4. * Floor 5 applies because the stale use continues into active reset and hardware processing, leaving exploitation consequences beyond the demonstrated crash unresolved. Not awarded: * No controlled reclaim, attacker-shaped replacement object, arbitrary write, callback control, refcount takeover, or demonstrated privilege escalation. * No confidentiality or integrity point based only on theoretical reuse. * No additional availability point for the same crash. * No remote-reachability point. Ordinary network packets do not directly perform the required probe-time `ndo_open`. * No local-unprivileged point because the complete below-host-admin trigger is not established. * No race penalty. Although the window is limited to probe, the commit identifies automatic `netifd` activation as a natural trigger and does not establish that the timing is exceptionally improbable or externally uncontrollable. * No rare-hardware penalty. A Mediatek SoC is required, but the supplied evidence does not establish that affected deployed hardware is rare or physical-attacker-only. * NEW52 B5-R does not apply. There is no retry-accumulated heap OOB write. * NEW53 B3-PENDING does not apply. No shared-backing or file-backed data write is present. * V33 and V32 do not apply. There is no attacker-controlled checksum metadata write or established active packet-state corruption across a trust boundary. * B4-PENDING does not apply because no authorization decision changes from DENY to ALLOW. 3. Reachability analysis The immediate race occurs when the first registered MAC netdev is opened before `dummy_dev` and shared NAPI initialization completes. The supplied concrete example is `netifd` automatically opening a WAN interface during driver probe. Opening or administratively configuring a physical interface commonly involves a network-management service or `CAP_NET_ADMIN`, but the supplied evidence does not establish the exact capability check or namespace governing the complete trigger. Therefore `HostAdminRequired=UNKNOWN` rather than YES, and `NamespaceScopedPrivilege=UNKNOWN` rather than YES. A delegated service or constrained network-management process may be able to initiate the open without full host-administrator powers, but this is not established for the complete probe-failure and queued-worker sequence. The attacker must also have affected Mediatek Ethernet hardware and reach the initialization window. No separate boot option, sysctl, BPF setup, or XDP attachment by the attacker is required by the shown path. The issue is not directly network-triggerable. Link activity may participate in phylink or watchdog behavior after the interface is opened, but packet or link influence alone does not establish control over the required probe-time open and cleanup sequence. 4. Primitive and severity interpretation The first primitive is an initialization-order race. A visible netdev can enter `ndo_open` while `eth->dummy_dev` and the shared NAPI structures remain uninitialized. This produces either an XDP registration warning and failed open on newer NETSYS versions or a NULL `napi->dev` dereference on NETSYS v1. The second and stronger lifetime primitive is asynchronous work surviving probe failure. `pending_work` is an active frame-engine reset worker that takes RTNL and drives hardware. The new synchronous cancellation before shared-state teardown directly indicates a potential stale post-lifetime execution path. This is an active structured-object lifetime issue rather than fixed stale storage. It is also broader than a cleanup helper that immediately dereferences a poisoned pointer. However, the supplied patch does not show controlled reclaim, attacker-controlled replacement contents, arbitrary kernel read or write, or function-pointer consumption. It therefore remains CLASS_B rather than CLASS_A. No concrete repeated release or refcount mutation exists: * There is no first `put` or release shown that leaves the exact reference reachable. * There is no second lifetime-mutating operation on that same dangling reference. * The dedicated repeated-release +1 is not awarded. * The repeated-refcount paranoid +1 is not awarded. * FLOOR 5 is based on active asynchronous structured processing, not double release. The demonstrated consequences are failed interface initialization, NULL dereference, and potential host DoS. Broader memory corruption from the stale worker is unresolved rather than demonstrated. `ImportantCandidate=YES` preserves review of that active post-teardown worker path. `AutoDowngradeAllowed=NO` is appropriate because automatically reducing the issue to a simple NULL-dereference bug would discard the independent asynchronous lifetime primitive. Likely operational severity is Actionable Moderate. Important-class treatment remains a review candidate, not a proven privilege-escalation or arbitrary-write result. 5. One-sentence report phrase A probe-time initialization race in the Mediatek Ethernet driver can expose a netdev before its shared NAPI state exists and can leave a frame-engine reset worker queued across teardown, causing a NULL dereference or potential asynchronous use after free and requiring manual review. 6. Manual review recommendation MANUAL CHECK REQUIRED Review the complete `pending_work` callback and all probe-error frees to determine which `eth` fields and hardware resources can be accessed after teardown, and verify whether a below-host-admin network-management actor can drive the complete open, queue, failure, and stale-worker sequence. 7. Call-site confidence Call-site confidence: medium ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: net: ethernet: mtk_eth_soc: allocate dummy netdev sooner [ Upstream Commit: 42001839e425b2564b7297be579073a13a8fddf6 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=42001839e425b2564b7297be579073a13a8fddf6 Changed files: net/core/xdp.c drivers/net/ethernet/mediatek/mtk_eth_soc.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=42001839e425b2564b7297be579073a13a8fddf6 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-98796 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98796 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:H/UI:N/S:U/C:N/I:N/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).