From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-98650][MODERATE 7.0] KVM: s390: Fix potential races in dat skey functions [ Upstream Date: Mon, 19 Oct 2026 01:27:29 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-98650 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: c958f0498891203fddf2764b28f8b102c6bfd719 List-Id: CVE: CVE-2026-98650 Priority: MODERATE 7.0 AL-KERNEL base severity: MODERATE KPANIC flag: YES Patch: KVM: s390: Fix potential races in dat skey functions [ Upstream Commit: c958f0498891203fddf2764b28f8b102c6bfd719 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=c958f0498891203fddf2764b28f8b102c6bfd719 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98650 Analysis date: Mon, 19 Oct 2026 01:27:29 -0400 ActionableScore: 5 ActionableScore lower bound: 4 Actionable bucket: Actionable Moderate / Strong Important candidate requiring manual review Manual review required: YES Summary: A race in KVM s390 DAT storage-key handling can reinterpret a concurrently split large-page entry as a leaf, causing operations on an incorrectly derived physical address and resulting in a host crash or constrained corruption of wrong-page storage-key metadata. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-98650 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98650 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-2026-98650 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-98650 CHECK WITH IMPACT FROM ORIG NN MODERATE | 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:L/I:L/A:H';CWE-362;CWE-367;CWE-662;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '5.8';DESCR 'A race in KVM s390 DAT storage key handling can occur when a large page is split after dat_entry_walk but before the caller converts the CRSTE into an absolute physical address. The changed entry may then be interpreted as a large page entry, causing storage key operations to use an incorrect physical address and potentially crash the host. For the CVSS the PR:L value is used because a local user context or a guest workload with access to the affected KVM storage key operations is required. AC:H is used because successful triggering depends on winning a race with concurrent large page splitting. The issue is not directly network reachable, although a remotely controlled workload inside an existing guest could exercise the guest side path. The demonstrated impact is host denial of service. In the paranoid case, limited confidentiality and integrity impact is retained because an incorrect physical page storage key or reference state may be read or modified, but no UAF, OOB write, arbitrary memory write, or reliable privilege escalation primitive is established. Automated impact decision path. The original SimpleNet classifier selected MODERATE. The primary LLM estimated paranoid CVSS 5.8 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 CVSS and initial triage stage MODERATE to IMPORTANT because the CVSS score, vector or bug class threshold rule was applied. KP and FD specialist checks stage IMPORTANT to MODERATE_KPANIC because FD false positive review allowed a downgrade. 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 INCREASED_TO_HIGH_BASED_ON_GUESSCVSS and DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFDH. The independent weaker Aegis shadow evaluated CVSS 4.8 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 DANGER VIRT SIMPLEFIX RACE HARDWARE LINUS KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFDH - - checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=5 ActionableScoreLower=4 PrimitiveClass=CLASS_B RealUAF=NO 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: 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: * +1 constrained active metadata corruption. After a large-page split, a non-leaf CRSTE can be passed to `large_crste_to_phys()`. The set paths may consequently apply a storage key to an incorrectly derived physical page. * +1 active semantic or translation-domain transition. The code can interpret a CRSTE that now describes a lower-level table as if it still described a large-page leaf. * +1 high-value KVM translation-state path. The affected operations read or modify storage-key and reference state associated with absolute host memory. * +1 reliable crash or strong DoS. The commit explicitly states that the race can lead to crashes. Paranoid additional signal: * +1 integrity concern. `page_set_storage_key()` and `page_cond_set_storage_key()` may modify storage-key metadata for the wrong physical page. The target is not sufficiently controlled to establish a directed write, but this is stronger than a harmless invalid read. Not awarded: * No UAF, heap OOB write, arbitrary memory write, controlled reclaim, callback control, or refcount corruption is established. * No confidentiality point was awarded. Reading the wrong storage key is possible in the get path, but no disclosure of protected data or page contents is demonstrated. * No strong guest-to-host exploitation leverage was awarded. The supplied patch does not prove that malicious guest code alone can induce both the storage-key operation and concurrent large-page split. * No local-unprivileged bonus was awarded because complete below-host-admin reachability is unresolved. * No race penalty was applied. A timing window exists, but the evidence does not establish that repeated attempts are prohibitively narrow or externally uncontrollable. * No extra availability point was awarded because it would duplicate the same crash. * NEW52 B5-R does not apply. There is no retry-accumulated heap OOB write. * NEW53 B3-PENDING does not apply. There is no shared file-backed or COW backing write. * NEW54 B4-PENDING does not apply. No authorization decision changes from DENY to ALLOW. V33 assessment: * Write: the set paths can modify storage-key metadata associated with an incorrectly derived absolute address. * Privileged consumer: host KVM and s390 storage-key machinery consume this state during memory protection and reference handling. * Boundary: guest influence is credible for storage-key operations, but control of the concurrent host large-page split and the complete guest-only trigger remain unresolved. * Consequence: host crash is demonstrated. Wrong-page metadata modification is technically credible, while controlled host-memory corruption or VM escape is not demonstrated. 3. Reachability analysis The relevant actors must be separated: * A malicious guest may exercise architectural storage-key operations and influence the requested key or guest frame. * The host VMM or host memory-management path controls the backing mappings and large-page splitting operation. * A host administrator controls VM creation and broad configuration, but the supplied evidence does not establish that host-admin powers are required for every vulnerable invocation. The complete sequence may therefore be a compound guest plus VMM or host-MM trigger. The patch does not show that guest code can independently force the required large-page split at the vulnerable instant. It also does not establish that only host root can trigger the race. Consequently, `HostAdminRequired` and namespace-scoped privilege remain `UNKNOWN`. The issue is not directly network reachable. Remote control of code already running inside a guest could provide one stage, but that does not make ordinary network traffic a complete trigger. It requires s390 KVM, large-page mappings, storage-key operations, and concurrent page splitting. No separate uncommon physical device is required. 4. Primitive and severity interpretation The exact primitive is inconsistent use of a mutable CRSTE across a concurrent large-page split. The walk reports a large-page result, but a later dereference may observe a non-leaf entry. That entry is then semantically misinterpreted by `large_crste_to_phys()`. This is active translation and storage-key state rather than fixed stale storage. It is not a post-lifetime object use and therefore is not a real UAF. There is no repeated release or refcount mutation. The get and reset paths can address the wrong page metadata. The set and conditional-set paths are stronger because they can modify a storage key at an incorrectly calculated absolute address. This supports CLASS_B as a constrained active-state corruption candidate. It does not support CLASS_A because the patch does not establish an attacker-selected target, arbitrary payload, host-memory write, controlled type confusion, or guest escape. The demonstrated consequence is host crash. The plausible unresolved consequence is modification of storage-key or reference metadata for an unintended page. `ImportantCandidate=YES` and `AutoDowngradeAllowed=NO` preserve this active-state and virtualization-boundary issue for review without claiming proven Important severity. 5. One-sentence report phrase A race in KVM s390 DAT storage-key handling can reinterpret a concurrently split large-page entry as a leaf, causing operations on an incorrectly derived physical address and resulting in a host crash or constrained corruption of wrong-page storage-key metadata. 6. Manual review recommendation MANUAL CHECK REQUIRED The patch establishes active translation-state misinterpretation and a possible write to storage-key metadata for the wrong physical page across a credible but unresolved guest-to-host trust boundary. 7. Call-site confidence Call-site confidence: medium ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: KVM: s390: Fix potential races in dat skey functions [ Upstream Commit: c958f0498891203fddf2764b28f8b102c6bfd719 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=c958f0498891203fddf2764b28f8b102c6bfd719 Changed files: arch/s390/kvm/dat.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=c958f0498891203fddf2764b28f8b102c6bfd719 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-98650 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98650 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:N/I:N/A:H The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:L/I:L/A:H The Best / paranoid CVSS score: 5.8 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: 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).