From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-98014][MODERATE 7.0] net/mlx5: E-Switch, prevent mc_list repopulation during vport disable [ Upstream Date: Sat, 26 Sep 2026 14:21:25 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-98014 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: 69904608e25e8ba58111aadd9210892cf3876201 List-Id: CVE: CVE-2026-98014 Priority: MODERATE 7.0 AL-KERNEL base severity: MODERATE KPANIC flag: YES Patch: net/mlx5: E-Switch, prevent mc_list repopulation during vport disable [ Upstream Commit: 69904608e25e8ba58111aadd9210892cf3876201 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=69904608e25e8ba58111aadd9210892cf3876201 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98014 Analysis date: Sat, 26 Sep 2026 14:21:25 -0400 ActionableScore: 5 ActionableScore lower bound: 4 Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review Manual review required: YES Summary: An mlx5 E-Switch recovery ordering flaw leaves dangling FDB flow-rule pointers in mc_list, causing steering-tree refcount corruption, use-after-free, and a host kernel crash during subsequent device reload. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-98014 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98014 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-2026-98014 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-98014 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-672;CWE-911;Other CVSS 'AV:L/AC:H/PR:H/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '7';DESCR 'An ordering flaw in mlx5 E-Switch vport shutdown can repopulate mc_list with FDB flow rules while the device is in INTERNAL_ERROR state. The subsequent FDB teardown leaves dangling flow_rule pointers and corrupts steering tree reference counts. A later recovery or reload can traverse these stale entries and trigger a refcount underflow, use-after-free, page fault, and kernel crash. For the CVSS the PR:L value is used in the paranoid assessment because some SR-IOV environments delegate VF or vport management to a reduced capability tenant or service account that may influence the recovery sequence. The directly documented PCI reset trigger normally requires administrative privileges and therefore the base assessment uses PR:H. The issue is not proven to be directly network reachable because ordinary traffic alone does not initiate the required firmware fatal recovery and device reload sequence. Impact is at least local denial of service. The concrete dangling pointer and use-after-free condition may also permit confidentiality or integrity impact and possible privilege escalation if the freed object can be reused in a controlled manner. This exploitation path is not demonstrated and requires manual review.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5) SKIP CVE-2026-98014 UNKNOWN SKIP No affected files built, so skip this CVE NO - - unknown MAYBE REMOTE OOB DANGER UAF NETWORK KPANIC KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFDH - - 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: 4 * Conservative score: 4 * Raw paranoid score: 5 * Paranoid score: 5 * No floor or cap was applied. * Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**:: ## 2. Signal breakdown Conservative signals: * +2 Real UAF and lifetime violation. FDB destruction leaves dangling `flow_rule` pointers in `vport->mc_list`, and reload subsequently consumes them. * +1 Secondary index or reverse-reference staleness. `mc_list` remains populated with references to rules belonging to the destroyed FDB, and later walkers actively process those obsolete associations. * +1 Reliable kernel crash. The supplied trace establishes refcount underflow, use-after-free, and a page fault during recovery. * Total: 4. Additional paranoid signal: * +1 Privileged device-management memory corruption path. Beyond the terminal stale-pointer fault, the first recovery stage corrupts steering-tree lifetime state and underflows a refcount. This is an established active-state corruption mechanism, although attacker control over its result is unresolved. * Total: 5. Not awarded: * No direct remote-reachability bonus. Network traffic alone is not shown to initiate the required firmware-fatal recovery and reload. * No local-unprivileged bonus. The demonstrated PCI reset method is privileged, and a complete below-host-admin trigger is not established. * No controlled-reclaim, arbitrary-write, callback-control, disclosure, or LPE bonus. * No structured stale-object leverage bonus because attacker influence over reclaimed contents or meaningful replacement fields is not demonstrated. * No concrete repeated-release bonus. Refcount underflow is observed, but the supplied control flow does not prove two lifetime operations on the exact same surviving reference with enough specificity for that dedicated rule. * No extra Availability point. The page fault and crash are the same consequence already counted. * No race penalty. This is an ordered recovery-state failure, not an established narrow timing race. * No hardware-rarity penalty. mlx5 hardware is specialized, but the supplied evidence does not establish that it is sufficiently rare or physically attacker-dependent. ## 3. Reachability analysis The demonstrated trigger writes to the physical PCI device reset sysfs attribute, which normally requires host-administrative authority. However, the vulnerable sequence fundamentally occurs during firmware-fatal recovery, and the evidence does not establish that privileged sysfs reset is the only way to enter it. Therefore `HostAdminRequired` remains `UNKNOWN` rather than assuming either host-admin-only or delegated reachability. No evidence establishes that an SR-IOV VF tenant, container, network namespace, or delegated service can perform the complete sequence of causing the relevant firmware-fatal state, disabling the vport, destroying the FDB, and initiating reload. Consequently, namespace-scoped privilege and the suggested `PR:L` deployment model remain unresolved. The prerequisite is mlx5 E-Switch operation with vports and an `allmulti_rule` present during firmware-fatal recovery. Ordinary packets may influence multicast or promiscuous state, but direct network triggering of the complete failure sequence is not demonstrated. No hard race is required by the documented reproducer. ## 4. Primitive and severity interpretation The exact primitive is an active structured-object lifetime violation. `mc_list` retains `flow_rule` references after their owning FDB and steering subtree are destroyed. Recovery later traverses the stale list and attempts to delete the expired rules, producing a real UAF and page fault. This is stronger than a fixed stale buffer because it involves active flow-steering objects, a stale secondary index, and corrupted refcounted tree state. It is not a proven Class A primitive because controlled reclaim, directed kernel read or write, callback influence, and privilege escalation are absent. The bug is not a verified concrete double-release case under the dedicated rule. Although refcount underflow demonstrates lifetime corruption, the supplied context does not identify the exact first and second operations on one surviving reference sufficiently clearly. The demonstrated consequence is host kernel DoS during device recovery. A stronger consequence is technically possible because stale structured pointers and refcounted tree corruption exist, but exploitation leverage beyond the crash is unresolved. This supports `CLASS_B`, `ImportantCandidate=YES`, and mandatory manual review, while the lack of controlled reuse keeps the numeric recommendation at Actionable Moderate rather than established Important. `AutoDowngradeAllowed=NO` because automatically treating this as an ordinary crash would discard the unresolved significance of active dangling flow-rule references, secondary-index traversal, and steering-tree lifetime corruption. ## 5. One-sentence report phrase An mlx5 E-Switch recovery ordering flaw leaves dangling FDB flow-rule pointers in mc_list, causing steering-tree refcount corruption, use-after-free, and a host kernel crash during subsequent device reload. ## 6. Manual review recommendation MANUAL CHECK REQUIRED Confirm whether an untrusted VF, delegated SR-IOV management service, firmware-mediated input, or network-driven device condition can induce the complete recovery sequence, and inspect whether stale `flow_rule` storage can be predictably reclaimed or reused before reload traversal. ## 7. Call-site confidence Call-site confidence: medium ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: net/mlx5: E-Switch, prevent mc_list repopulation during vport disable [ Upstream Commit: 69904608e25e8ba58111aadd9210892cf3876201 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=69904608e25e8ba58111aadd9210892cf3876201 Changed files: drivers/net/ethernet/mellanox/mlx5/core/eswitch.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=69904608e25e8ba58111aadd9210892cf3876201 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-98014 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98014 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: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).