From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-97991][IMPORTANT] vdpa_sim_blk: reject out-of-range sector starts [ Upstream Date: Fri, 25 Sep 2026 07:42:46 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-97991 X-AL-KERNEL-Priority: IMPORTANT X-AL-KERNEL-Severity: IMPORTANT X-AL-KERNEL-Base-Severity: IMPORTANT X-AL-KERNEL-KPANIC: YES X-AL-KERNEL-ActionableScore: 7 X-AL-KERNEL-ActionableScore-Lower: 6 X-AL-KERNEL-Commit: 14acb0d44b7d90a077beb7b92b444b90bafda9d9 List-Id: CVE: CVE-2026-97991 Priority: IMPORTANT AL-KERNEL base severity: IMPORTANT KPANIC flag: YES Patch: vdpa_sim_blk: reject out-of-range sector starts [ Upstream Commit: 14acb0d44b7d90a077beb7b92b444b90bafda9d9 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=14acb0d44b7d90a077beb7b92b444b90bafda9d9 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97991 Analysis date: Fri, 25 Sep 2026 07:42:46 -0400 ActionableScore: 7 ActionableScore lower bound: 6 Actionable bucket: Strong Important candidate Manual review required: YES Summary: No one-sentence report phrase was found. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-97991 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97991 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: IMPORTANT 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-2026-97991 IMPORTANT https://www.kernelcve.org/list/?q=CVE-2026-97991 CHECK WITH IMPACT FROM ORIG NN IMPORTANT Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H';*CWE-787;CWE-125;CWE-191;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:C/C:L/I:L/A:H';BEST CVSS score: '8.8';DESCR 'A memory corruption issue in the vDPA simulated block device occurs because vdpasim_blk_check_range can continue after an invalid start sector and an unsigned capacity subtraction can underflow. A guest can submit an out of range virtio block request that reaches blk->buffer plus an invalid offset. VIRTIO_BLK_T_OUT and VIRTIO_BLK_T_WRITE_ZEROES can cause out of bounds writes in host kernel memory. VIRTIO_BLK_T_IN can copy data beyond the host buffer to the guest and may disclose host memory. For the CVSS the PR:L reflects that reliable triggering normally requires control over raw virtio block requests from a guest or equivalent control over the exposed vDPA block device. The issue is not directly network reachable. It is a guest to host attack surface in deployments that expose the vdpa_sim_blk device to an untrusted guest. Impact is at least a host denial of service and may include confidentiality integrity impact or privilege escalation due to explicit host kernel out of bounds read and write primitives.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate (with actual score 7) YES WRITE OOB LINUS KPANIC NO NO checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=7 ActionableScoreLower=6 PrimitiveClass=CLASS_A RealUAF=NO HostAdminRequired=NO NamespaceScopedPrivilege=NO ImportantCandidate=YES ManualReview=REQUIRED AutoDowngradeAllowed=NO 1. ActionableScore * Raw conservative score: 6 * Conservative score: 6 * Raw paranoid score: 7 * Paranoid score: 7 * Final recommended bucket: **Strong Important candidate**:: Conservative arithmetic: * +2 attacker-influenced host-kernel out-of-bounds write. A guest-controlled sector offset reaches `blk->buffer + offset` and is used by `memcpy`, `_copy_from_iter`, or `memset`. * +1 confidentiality impact. `VIRTIO_BLK_T_IN` copies data from an invalid host buffer offset back to the guest through `_copy_to_iter`. * +1 integrity impact. `VIRTIO_BLK_T_OUT` and `VIRTIO_BLK_T_WRITE_ZEROES` write past the simulated block buffer into host kernel memory. * +1 guest-to-host security-boundary relevance. The vulnerable requests originate from guest-controlled virtio block operations and affect host-kernel memory. * +1 practical chain. The supplied patch description directly establishes guest request input, invalid offset creation through unsigned underflow, and host out-of-bounds read/write sinks. Paranoid addition: * +1 realistic host denial-of-service potential. The established host-kernel OOB write and OOB read paths can plausibly fault or corrupt host kernel state, even though the supplied input does not include a demonstrated panic trace. 2. Signal breakdown Positive signals: * +2 Strong corruption primitive. The patch explicitly identifies attacker-influenced out-of-bounds writes through `VIRTIO_BLK_T_OUT` and `VIRTIO_BLK_T_WRITE_ZEROES`. This is not merely invalid-state validation or a warning-only issue. * +1 Confidentiality plausible. `VIRTIO_BLK_T_IN` can copy bytes beyond `blk->buffer` into the guest-visible response path. This is a concrete read-back/disclosure path, not only an internal invalid read. * +1 Integrity plausible. Guest-controlled data or zeroes can be written beyond the intended host buffer. * +1 Cross-boundary relevance. A guest request can reach host-kernel memory operations after the simulated vDPA block device has been exposed to that guest. * +1 Practical exploit chain. The patch establishes the complete primitive chain: invalid sector start, unsigned underflow, invalid offset, and a host copy or memset sink. * +1 Paranoid host DoS potential. Host-kernel OOB accesses can realistically cause a host fault or corruption-induced failure. Important non-awarded signals: * Controlled arbitrary kernel write is not demonstrated. The destination is attacker-selected relative to `blk->buffer`, but exact arbitrary host address control is not established. * Controlled reclaim, callback corruption, function-pointer control, credential overwrite, and demonstrated privilege escalation are not established. * Remote network reachability is not established. The trigger is delivered through a guest virtio block interface, not directly through a network listener. * No UAF, stale reference, repeated release, refcount mutation, or lifetime race is established. * No extra Availability point is awarded in the conservative score because a host crash is plausible but not explicitly demonstrated by the supplied material. 3. Reachability analysis The evaluated attacker is a malicious guest, or an actor controlling raw block requests issued by a guest that has been given a `vdpa_sim_blk` backed virtio block device. The guest does not need host administrator privilege to submit malformed block requests after the device is exposed. Therefore, `HostAdminRequired=NO` describes the trigger-side attacker model. A host administrator normally must configure and expose the simulated vDPA block device before a guest can use it, but this is prior deployment setup rather than a privilege required by the attacker to issue the malformed request. The supplied evidence does not depend on Linux user namespaces or network namespaces. The issue is not a namespace-scoped `CAP_NET_ADMIN` problem, therefore `NamespaceScopedPrivilege=NO`. No special physical hardware is required because this affects the in-kernel `vdpa_sim_blk` simulated block backend. However, exposure requires a deployment that uses the vDPA simulator and attaches the resulting virtio block device to a guest. This is a configuration prerequisite, but it does not negate the guest-to-host attack surface once deployed. No timing race is involved. The vulnerability is an ordered input-validation failure caused by accepting an invalid starting sector and then performing unsigned arithmetic. 4. Primitive and severity interpretation The concrete primitive is an attacker-influenced out-of-bounds host-kernel read and write relative to `blk->buffer`. `vdpasim_blk_check_range()` previously logged an invalid `start_sector` but continued validation. The later unsigned capacity subtraction can underflow, allowing an invalid byte offset to reach data-path operations: * `VIRTIO_BLK_T_OUT` copies guest-controlled data into host memory beyond the simulated block buffer. * `VIRTIO_BLK_T_WRITE_ZEROES` zeroes host memory beyond the simulated block buffer. * `VIRTIO_BLK_T_IN` copies host memory beyond the simulated block buffer back to the guest. This is not a fixed stale-buffer UAF, cleanup-only dereference, or crash-only invalid pointer. It is a direct memory-corruption and disclosure primitive crossing from a guest-controlled request path into host-kernel memory handling. The primitive is classified as `CLASS_A` because the patch directly establishes guest-influenced OOB writes and a guest-visible OOB read path. The patch does not establish arbitrary-address write, controlled host-object replacement, callback control, or a working guest-to-host escape. Those remain speculative consequences and are not scored as established exploitation results. `ImportantCandidate=YES` and `ManualReview=REQUIRED` are appropriate because the host-kernel OOB read/write primitive is concrete and crosses a virtualization boundary. `AutoDowngradeAllowed=NO` is appropriate because automatically treating this as an ordinary simulated-device validation bug would discard an established host-memory corruption and disclosure primitive. Likely operational severity is Important-class when `vdpa_sim_blk` is deployed for untrusted guests. In deployments where the simulator is not enabled or no guest can access its virtio block device, the vulnerable path is not reachable. 5. One-sentence report phrase A malicious guest can submit an out-of-range virtio block request to the vDPA simulated block device, causing an unsigned range-validation underflow that leads to host-kernel out-of-bounds reads and writes, with potential host memory disclosure, corruption, and denial of service. 6. Manual review recommendation MANUAL CHECK REQUIRED The patch establishes guest-controlled host-kernel OOB read and write sinks, but further analysis is needed to determine the practical degree of target control, disclosure reliability, host crash behavior, and feasibility of guest-to-host privilege escalation. 7. Call-site confidence Call-site confidence: high ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: vdpa_sim_blk: reject out-of-range sector starts [ Upstream Commit: 14acb0d44b7d90a077beb7b92b444b90bafda9d9 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=14acb0d44b7d90a077beb7b92b444b90bafda9d9 Changed files: drivers/vdpa/vdpa_sim/vdpa_sim_blk.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=14acb0d44b7d90a077beb7b92b444b90bafda9d9 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-97991 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97991 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:L/PR:L/UI:N/S:C/C:L/I:L/A:H The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H The Best / paranoid CVSS score: 8.8 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: 6 Paranoid ActionableScore: 7 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: IMPORTANT KPANIC detected for this report: YES Published priority for this report (same as in Subject): IMPORTANT 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).