From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-97988][MODERATE 7.0] vhost: invalidate vring access on IOTLB transitions [ Upstream Date: Wed, 30 Sep 2026 04:03:10 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-97988 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: 6 X-AL-KERNEL-ActionableScore-Lower: 5 X-AL-KERNEL-Commit: 4a454af8e56c1596ada1421ae5eba90c8896c7c7 List-Id: CVE: CVE-2026-97988 Priority: MODERATE 7.0 AL-KERNEL base severity: MODERATE KPANIC flag: YES Patch: vhost: invalidate vring access on IOTLB transitions [ Upstream Commit: 4a454af8e56c1596ada1421ae5eba90c8896c7c7 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=4a454af8e56c1596ada1421ae5eba90c8896c7c7 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97988 Analysis date: Wed, 30 Sep 2026 04:03:10 -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-2026-97988 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97988 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-2026-97988 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-97988 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-362;*CWE-416;*CWE-672;*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 'A race in vhost IOTLB transitions can leave cached vring pointers and per VQ IOTLB metadata associated with an old address space while a worker continues descriptor processing. Replacing or disabling the device IOTLB without invalidating these cached references can expose a worker to stale mappings after the old IOTLB table is detached or freed. This creates a plausible stale pointer or use after free condition that may cause a host kernel crash and may also direct vring accesses to unintended memory. For the CVSS the PR:L is selected because triggering requires a local process with access to a vhost device or an existing vhost file descriptor, but this process may be a reduced privilege QEMU or service account rather than a fully privileged administrator. The issue is not directly network reachable and guest traffic alone does not normally control VIRTIO_F_ACCESS_PLATFORM transitions. Impact is at least a local denial of service and may include confidentiality, integrity, or privilege escalation impact if stale IOTLB metadata can be reused as a memory corruption primitive.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 6) MAYBE OOB INIT NETWORK LOG HARDWARE LINUS KPANIC KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFKP - - checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=6 ActionableScoreLower=5 PrimitiveClass=CLASS_B RealUAF=YES HostAdminRequired=UNKNOWN NamespaceScopedPrivilege=UNKNOWN ImportantCandidate=YES ManualReview=REQUIRED AutoDowngradeAllowed=NO 1. ActionableScore * Raw conservative score: 5 * Conservative score: 5 * Raw paranoid score: 6 * Paranoid score: 6 * No floor or cap changed either score. * Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**:: 2. Signal breakdown Conservative positive signals: * +2 Real UAF or equivalent lifetime violation. Per-VQ IOTLB metadata can remain associated with the old device IOTLB while workers continue processing descriptors, and the old table can be detached or freed before all worker-visible references are invalidated. * +1 Active semantic or translation-domain transition. Cached vring addresses are retained while `VIRTIO_F_ACCESS_PLATFORM` changes their interpretation between direct userspace addresses and GIOVA/IOTLB addresses. * +1 Secondary index or metadata staleness. `meta_iotlb` and cached vring access remain discoverable by worker processing after the device-wide translation state changes. * +1 Realistic kernel crash or strong DoS. A worker can consume stale translation metadata or translate descriptors through the wrong address domain, supporting an invalid host-kernel access. No additional Availability point is awarded for the same failure. Paranoid additional signal: * +1 Attacker-influenced repeatable active-state transition. A process controlling the vhost file descriptor can configure IOTLB state, vring addresses, and feature transitions repeatedly while virtqueue workers are active. This increases stale-use and reuse opportunities without assuming controlled heap reclaim. Material non-awarded signals: * Controlled reclaim or attacker-shaped replacement of freed IOTLB objects is not demonstrated. * Arbitrary or directed host-kernel read/write is not demonstrated. * Strong LPE leverage and CLASS_A are not awarded. * Confidentiality and integrity points are not separately awarded merely because stale mappings could theoretically be exploitable. * Direct remote reachability is not established. * Guest-only triggering is not established because changing `VIRTIO_F_ACCESS_PLATFORM` is a host userspace or VMM operation. * No race penalty is applied. The transition can be repeated, and the evidence does not establish a narrow, one-shot, externally uncontrollable race. * No privilege penalty is applied because neither host-admin necessity nor a specific reduced-privilege boundary is conclusively established. * No repeated release or refcount signal applies. The patch does not show a first put followed by a second lifetime mutation on the exact same reference. * No extra Availability point is awarded because it would double-count the potential host crash. 3. Reachability analysis The direct trigger is a local process with control of a vhost-net or vhost-vsock file descriptor. That process can change negotiated features and cause direct-to-IOTLB, IOTLB-to-direct, or IOTLB replacement transitions while workers process virtqueue descriptors. The complete path is not shown to be available to an ordinary local user. Access to vhost character devices is commonly restricted or delegated by a privileged launcher, but the supplied evidence does not prove that host-admin-equivalent privilege remains necessary after the descriptor is opened. A reduced-privilege VMM or service process may retain sufficient ioctl authority. Therefore `HostAdminRequired=UNKNOWN`. No namespace capability check governing the complete operation is shown. Merely running a VMM or networking component in a namespace would not prove namespace-scoped reachability, so `NamespaceScopedPrivilege=UNKNOWN`. This is not directly network-triggerable. Guest or network activity may keep a worker active and improve timing, but it does not normally permit the guest to execute the host-side feature transition. The evaluated attacker is therefore a local VMM or vhost-controlling process, not a pure malicious guest and not a remote peer. No special physical hardware is required. The affected functionality requires configured vhost-net or vhost-vsock operation with IOTLB or `VIRTIO_F_ACCESS_PLATFORM`, but this is a runtime virtualization configuration rather than rare hardware. 4. Primitive and severity interpretation The exact primitive is an unsynchronized lifetime and semantic transition involving active vring mappings, per-VQ IOTLB metadata, and the device-wide IOTLB table. Workers can observe a mixed state, including a transient NULL device IOTLB that causes fallback to `umem`, or continue using cached state belonging to the previous address domain. This is active structured state, not fixed stale storage and not a cleanup-only dereference. The stale data participate in descriptor translation and vring memory access during normal worker processing. Freeing the previous IOTLB only after locking all virtqueues and invalidating their caches directly implies that the old lifetime was not previously protected across all worker-visible uses. The patch supports a real stale post-lifetime use or UAF candidate, but it does not establish controlled reclaim, callback control, arbitrary kernel access, or privilege escalation. The correct classification is therefore CLASS_B rather than CLASS_A. The demonstrated mechanism supports mixed-address-space access and host failure. The paranoid interpretation recognizes repeated attacker-controlled transitions but does not assume a complete exploitation chain. Important-grade exploitation remains unresolved because stale translation metadata affects active memory access rather than passive bookkeeping. `ImportantCandidate=YES` and `ManualReview=REQUIRED` are appropriate because this is an active translation-object lifetime violation with meaningful unresolved leverage beyond an immediate cleanup fault. `AutoDowngradeAllowed=NO` because automatic closure would discard the unresolved effect of stale IOTLB metadata on active descriptor translation. 5. One-sentence report phrase A local process controlling a vhost device can race IOTLB address-mode transitions with virtqueue workers, leaving stale ring mappings or freed translation metadata in active use and potentially causing a host crash or more serious memory-access corruption. 6. Manual review recommendation MANUAL CHECK REQUIRED Confirm the old IOTLB free ordering, whether `meta_iotlb` entries directly reference freed IOTLB nodes, the worker operations performed through stale entries, and whether a reduced-privilege VMM with an inherited vhost descriptor can complete the entire trigger. 7. Call-site confidence Call-site confidence: medium ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: vhost: invalidate vring access on IOTLB transitions [ Upstream Commit: 4a454af8e56c1596ada1421ae5eba90c8896c7c7 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=4a454af8e56c1596ada1421ae5eba90c8896c7c7 Changed files: drivers/vhost/net.c drivers/vhost/vhost.c drivers/vhost/vhost.h drivers/vhost/vsock.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=4a454af8e56c1596ada1421ae5eba90c8896c7c7 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-97988 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97988 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 alexanjelausa@gmail.com (and both send reply to CVE record itself too and see "reply" button below for howto reply).