From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-89758][MODERATE 7.0] mm/mempolicy: skip non-present PMDs when queueing folios Date: Sat, 12 Sep 2026 02:32:46 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-89758 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: 3d927093b2add7ced588a926c3dc5d7a83f18596 List-Id: CVE: CVE-2026-89758 Priority: MODERATE 7.0 AL-KERNEL base severity: MODERATE KPANIC flag: YES Patch: mm/mempolicy: skip non-present PMDs when queueing folios Commit: 3d927093b2add7ced588a926c3dc5d7a83f18596 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=3d927093b2add7ced588a926c3dc5d7a83f18596 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89758 Analysis date: Sat, 12 Sep 2026 02:32:46 -0400 ActionableScore: 5 ActionableScore lower bound: 4 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-2026-89758 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89758 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-2026-89758 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-89758 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-20;CWE-843;CWE-822;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:H';BEST CVSS score: '7';DESCR 'A device-private non-present PMD can reach queue_folios_pmd() and be passed to pmd_folio(), which interprets the software entry as a hardware PFN and produces a bogus folio pointer. The immediate result may be a VM_BUG_ON, an invalid vmemmap dereference, or a kernel oops. If the encoded PFN aliases a valid page, the MM code may instead isolate an unrelated live folio from the LRU and pass it into later page migration processing. For the CVSS the PR:L value is used because an unprivileged local user may operate on memory in its own process, although an HMM capable GPU driver, an anonymous THP migrated to device memory, and a matching mempolicy operation are required. The issue is not network reachable. Impact is at least a local denial of service via kernel crash and in the worst case may affect confidentiality and integrity through wrong folio manipulation. The aliasing case is a memory corruption candidate and should be reviewed for refcount, LRU, and migration side effects before privilege escalation is ruled out.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5) SKIP CVE-2026-89758 UNKNOWN SKIP The Fixes patch not applied yet, so unlikely that actual: url YES NO NO unknown MAYBE READ HARDWARE KPANIC MEMORY 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=NO 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 positive signals: * +1 Local unprivileged trigger. A process can operate on its own address range using `mbind()`, `migrate_pages()`, or `set_mempolicy_home_node()` after an HMM driver has migrated its anonymous THP to device memory. * +1 Weak or indirect corruption candidate. A device-private PMD software encoding is interpreted as a hardware PFN, producing a bogus `struct folio` pointer. * +1 Reliable kernel crash or strong DoS. The documented outcomes include `VM_BUG_ON()`, invalid vmemmap dereference, and kernel Oops. * +1 Integrity impact. In the aliasing case, the bogus pointer resolves to an unrelated live folio and the kernel silently changes its LRU state. This is an operation on the wrong active MM object, not only an internal invalid read. Paranoid additional signal: * +1 Weak exploitation concern. The wrong live folio can proceed beyond pointer lookup into folio isolation and subsequent migration-related processing. This leaves meaningful page lifetime, refcount, ownership, and migration effects unresolved, although no controlled target or arbitrary write is established. Negative signals: * No negative points. HMM-capable GPU support is a prerequisite, but common GPU hardware is not shown to be rare or physical-only. * No race penalty. The trigger is a state sequence, not a demonstrated narrow timing race. * No privilege penalty. The relevant memory-policy operations on a process-owned mapping do not require host-admin-equivalent privilege. Important non-awarded signals: * No real UAF is established. * No controlled reclaim, arbitrary write, disclosure, callback control, or credential modification is demonstrated. * Strong exploitation leverage +2 is not awarded because the attacker is not shown to control which unrelated folio the encoded PFN aliases. * Remote reachability is not established. * No extra Availability point is awarded for the same Oops or crash. * No confidentiality point is awarded because no read-back or disclosure path is shown. * No repeated release or refcount signal exists. There is no first lifetime-ending release followed by a second mutation through the same reference. * REAL UAF FLOOR 5 is not applicable. 3. Reachability analysis The evaluated attacker is an ordinary local process with access to an HMM-capable GPU or comparable device. The process must cause an anonymous THP belonging to its address space to migrate into device-private memory and then invoke a memory-policy operation over that range. The documented syscalls can operate on process-owned memory without host-admin-equivalent privilege. Device-node access or normal graphics-compute authorization may still be required, but this is below host administrator authority in ordinary delegated GPU deployments. Therefore `HostAdminRequired=NO`. No namespace capability forms part of the documented trigger, so complete namespace-scoped reachability is neither required nor established. `NamespaceScopedPrivilege=UNKNOWN` is used rather than inferring namespace behavior. Prior system support for HMM and device-private THP migration is required. The attacker need not configure the kernel or load the driver, but the suitable GPU path must already be available. There is no demonstrated timing race, and the trigger can be attempted whenever the mapping remains represented by a device-private PMD. 4. Primitive and severity interpretation The exact primitive is semantic misinterpretation of a non-present device-private PMD as a present hardware PFN. `pmd_folio()` consequently returns either an invalid vmemmap-derived pointer or a pointer that aliases an unrelated valid folio. This is an active structured-object error rather than fixed stale storage. The vulnerable path can pass the wrong `struct folio` into LRU isolation and migration-related processing. It is not cleanup-only behavior. There is no post-lifetime use proving a UAF and no repeated release or refcount mutation. The issue therefore remains CLASS_B rather than being classified as a lifetime UAF or controlled CLASS_A corruption. The demonstrated consequences are kernel crash and mutation of an unrelated folio state. Potentially broader page ownership, refcount, or migration corruption is technically plausible because processing continues with the wrong active folio, but controlled object selection, arbitrary read or write, and privilege escalation are not established. `ImportantCandidate=YES` because the aliasing case reaches an unrelated live memory-management object and requires exploitation-leverage review beyond the immediate crash. `AutoDowngradeAllowed=NO` because automatically reducing the issue to ordinary DoS would discard the unresolved wrong-folio processing primitive. The likely operational severity is Actionable Moderate. Important severity remains possible if call-site review shows that the unrelated folio can be migrated, freed, remapped, or otherwise subjected to attacker-useful lifetime or ownership changes. 5. One-sentence report phrase A local user with access to HMM device-private memory can cause a non-present PMD to be interpreted as a hardware PFN, resulting in a kernel crash or operations on an unrelated live folio and requiring manual review of downstream migration and lifetime effects. 6. Manual review recommendation MANUAL CHECK REQUIRED Review the complete `queue_folios_pmd()` consumer path to determine whether an aliased unrelated folio can undergo migration, refcount changes, ownership transitions, or release beyond LRU isolation. 7. Call-site confidence Call-site confidence: medium ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: mm/mempolicy: skip non-present PMDs when queueing folios Commit: 3d927093b2add7ced588a926c3dc5d7a83f18596 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=3d927093b2add7ced588a926c3dc5d7a83f18596 Commit description: Patch series "mm: handle device-private PMDs in walk callbacks", v3. Since commit 368076f ("mm/huge_memory: add device-private THP support to PMD operations") a PMD may hold a device-private swap entry whenever an HMM-based GPU driver migrates an anonymous THP folio to device memory via migrate_vma_pages(). pmd_trans_huge_lock() succeeds for such PMDs (pmd_is_huge() returns true for any non-present, non-none huge PMD), so several MM walk callbacks that used to assume present THP or migration entry are now reachable with a device-private PMD. The results range from a VM_BUG_ON() firing on debug kernels, to an oops on a bogus vmemmap dereference, to silently isolating an unrelated live folio from LRU in the aliasing case. This patch (of 3): queue_folios_pmd() is called under pmd_trans_huge_lock(), whose pmd_is_huge() check returns true for any non-present, non-none PMD softleaf. Passing such a PMD to pmd_folio() treats the softleaf encoding as a hardware PFN and can return a bogus folio pointer. Mirror queue_folios_pte_range(): handle non-present entries before looking up a folio. Keep migration entries counted as failures, but skip other non-present PMDs such as device-private entries. Potential trigger: an HMM-based GPU driver migrates an anonymous THP folio to device memory via migrate_vma_pages(), leaving a device-private PMD. Userspace then calls mbind(), migrate_pages() or set_mempolicy_home_node() on that range. Link: https://lore.kernel.org/20260710105557.1987433-1-usama.arif@linux.dev Link: https://lore.kernel.org/20260710105557.1987433-2-usama.arif@linux.dev Fixes: 368076f ("mm/huge_memory: add device-private THP support to PMD operations") Signed-off-by: Usama Arif Reported-by: sashiko-bot Link: https://sashiko.dev/#/patchset/20260703173903.3789516-1-usama.arif%40linux.dev?part=6 Reviewed-by: Joshua Hahn Reviewed-by: Zi Yan Acked-by: David Hildenbrand (Arm) Reviewed-by: Balbir Singh Reviewed-by: Gregory Price Cc: Alistair Popple Cc: Baolin Wang Cc: Barry Song Cc: Byungchul Park Cc: Dev Jain Cc: "Huang, Ying" Cc: Jann Horn Cc: Johannes Weiner Cc: Lance Yang Cc: Liam R. Howlett Cc: Lorenzo Stoakes Cc: Matthew Brost Cc: Nico Pache Cc: Rakie Kim Cc: Ryan Roberts Cc: Shakeel Butt Cc: Vlastimil Babka Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman Changed files: mm/mempolicy.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=3d927093b2add7ced588a926c3dc5d7a83f18596 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-89758 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89758 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: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).