From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-93827][MODERATE 7.0] virtio-fs: avoid double-free on failed queue setup [ Upstream Date: Thu, 24 Sep 2026 21:07:15 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-93827 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: 3fbc5ee776fefbc6ea34d2518f1f5f90aa2e1dad List-Id: CVE: CVE-2026-93827 Priority: MODERATE 7.0 AL-KERNEL base severity: MODERATE KPANIC flag: YES Patch: virtio-fs: avoid double-free on failed queue setup [ Upstream Commit: 3fbc5ee776fefbc6ea34d2518f1f5f90aa2e1dad Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=3fbc5ee776fefbc6ea34d2518f1f5f90aa2e1dad Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93827 Analysis date: Thu, 24 Sep 2026 21:07:15 -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-2026-93827 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93827 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-2026-93827 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-93827 CHECK WITH IMPACT FROM ORIG NN MODERATE Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H';CWE-415;*CWE-416;CWE-763;Other CVSS 'AV:A/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:H';BEST CVSS score: '8.8';DESCR 'A double free in the virtio fs probe failure path occurs because virtio_fs_setup_vqs frees fs vqs and fs mq_map after virtio_find_vqs fails but leaves the pointers non NULL. Later probe cleanup drops the final kobject reference and virtio_fs_ktype_release frees the same allocations again, creating a kernel memory corruption condition. A malicious or malformed virtio fs device can trigger the failure by advertising more request queues than the transport provides. The attack vector is adjacent AV:A because the attacker must control a virtio device or virtual machine device configuration and it is not a network protocol attack. For the CVSS the PR:N because no guest user account or local code execution is required when an attacker controls the presented virtio device. Impact is at least guest kernel denial of service and may include confidentiality and integrity impact through double free based memory corruption. Manual review is required to determine exploitability for the affected allocator and virtualization configuration.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 6) YES DANGER DISK VIRT ERRORPATH UAF HARDWARE LINUS KPANIC KPANIC INCREASED_TO_HIGH_BASED_ON_GUESSCVSS DECREASED_TO_MODERATE7_BASED_ON_FALSEPOSCHECKOFDH NO NO 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 * Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**:: No score floor or cap changed the arithmetic. The Real UAF FLOOR 5 is independently satisfied because the first kfree can end allocation lifetime, the exact pointers remain reachable, and the later kobject release performs a concrete second free. 2. Signal breakdown Positive signals: * +2 Real UAF or lifetime violation. `fs->vqs` and `fs->mq_map` are freed in `virtio_fs_setup_vqs()` after `virtio_find_vqs()` failure, remain stored as non NULL pointers, and are later freed again by `virtio_fs_ktype_release()`. * +1 Concrete repeated release or lifetime mutation. The same exact stale references reach a later `kfree()`. `kfree()` is a real allocation lifetime operation and can act on allocator metadata for storage whose original allocation lifetime may have ended. * +1 Reliable kernel crash or strong DoS. A kernel double-free can trigger allocator detection, corruption handling, BUG, Oops, or panic during virtio-fs probe cleanup. * +1 Firmware or device mediated external influence. A malformed or malicious virtio-fs backend can advertise more request queues than the transport provides, causing `virtio_find_vqs()` to fail and driving the vulnerable cleanup sequence. * +1 Paranoid only. The second `kfree()` is a concrete write-like allocator lifetime mutation through stale storage. It may affect allocator state before a diagnostic failure, although controlled reuse or allocator manipulation is not demonstrated. Negative signals: * No host-admin penalty. Control of the virtio device backend or VM device configuration is required for the malicious-device scenario, but the supplied evidence does not establish whether the attacker must be a host administrator, a lower-privileged VMM operator, a compromised backend, or another device-control principal. * No hard-race penalty. This is an ordered probe failure and cleanup sequence, not a narrow timing race. * No niche subsystem penalty. virtio-fs is specialized, but the supplied evidence establishes a concrete device-controlled failure path. Specialization alone is not a valid negative signal. Important non-awarded signals: * Controlled reclaim is not demonstrated. * Arbitrary kernel write is not demonstrated. * Controlled replacement of either freed allocation is not demonstrated. * Callback or function-pointer influence is not demonstrated. * Confidentiality and integrity consequences are not awarded merely because a double-free can theoretically corrupt reused memory. * Direct remote network reachability is not established. The relevant input is delivered through a virtio device or backend, not a network protocol. * Structured stale-object exploitation leverage is not awarded. The stale objects are fixed heap allocations used in probe cleanup, and no later consumption of attacker-controlled object fields is shown. * Weak LPE concern is not separately awarded because the same relevant fact is already counted as the concrete second lifetime mutation. * Extra Availability impact is not awarded because it would duplicate the kernel-crash DoS signal. Repeated-release evaluation: * First release can end lifetime: Yes. `kfree(fs->vqs)` and `kfree(fs->mq_map)` release the allocations. * Exact same references survive: Yes. The patch sets both fields to NULL immediately after the first frees, proving they otherwise remain reachable through `struct virtio_fs`. * Second operation is a real lifetime mutation: Yes. `virtio_fs_ktype_release()` later invokes `kfree()` on the same stale pointers. * Concrete repeated release/refcount signal: Awarded, +1. * Paranoid repeated-lifetime leverage: Awarded, +1, because the second `kfree()` can operate on freed allocator-managed storage. This does not assume controlled reclaim or arbitrary write. * FLOOR 5: Applies, but does not alter the conservative raw score because it already equals 5. 3. Reachability analysis The evaluated attacker is not an ordinary guest user. The triggering condition is a malformed or malicious virtio-fs device that advertises more request queues than the underlying transport actually provides. A guest-only attacker normally cannot directly alter the virtio device queue configuration presented by the hypervisor or backend. A local guest process mounting or accessing virtio-fs is also not shown to control the relevant queue count or force the failed probe sequence after device initialization. The likely attacker roles are a malicious or compromised virtio-fs backend, a malicious VMM process, or an actor able to configure the virtual hardware exposed to a guest. The supplied patch does not establish whether this ability necessarily requires host-administrator-equivalent privilege. Therefore `HostAdminRequired=UNKNOWN` is appropriate rather than assuming either YES or NO. No namespace-scoped capability path is shown. Network namespaces, user namespaces, containers, and guest-local capabilities do not establish control over virtio queue advertisement. Therefore `NamespaceScopedPrivilege=UNKNOWN`. The complete trigger requires prior exposure of a virtio-fs device to the guest. This is normally a host or VMM provisioning action, but the supplied evidence does not resolve whether the attacker is the provisioning actor, controls an existing backend, or exploits a compromised device implementation. No unusual physical hardware is required. The relevant condition is virtual device or transport behavior. The sequence is deterministic once the malformed queue configuration is present and does not depend on a narrow race window. 4. Primitive and severity interpretation The exact primitive is a concrete double-free in the virtio-fs probe failure path: `virtio_fs_setup_vqs()` frees `fs->vqs` and `fs->mq_map` then leaves both pointers non NULL then `virtio_fs_probe()` drops the final kobject reference then `virtio_fs_ktype_release()` frees the same allocations again. This is a real lifetime violation and directly implies stale post-lifetime references. The patch is not merely defensive cleanup because assigning NULL after the first `kfree()` specifically prevents the second `kfree()` during kobject release. The stale storage consists of fixed heap allocations associated with `struct virtio_fs`, and the demonstrated stale use is in error and destructor cleanup. This limits the currently proven exploitation surface compared with an attacker-controlled active-object UAF. However, this is stronger than an ordinary stale cleanup dereference because the later operation is a second concrete allocation lifetime mutation through the exact potentially dangling references. The first operation may free the allocations. The second operation actively invokes `kfree()` on the same freed allocations. This can result in allocator corruption or allocator-detected double-free termination. No controlled reuse, arbitrary write, active object substitution, callback control, information disclosure, privilege escalation, or guest-to-host escape is established by the supplied evidence. The appropriate class is `CLASS_B`, not `CLASS_A`. A meaningful kernel lifetime-corruption primitive exists and requires manual exploitability review, but strong exploitation leverage is not proven. `ImportantCandidate=YES` is appropriate because the concrete second free may mutate allocator state through freed storage, and the actual behavior depends on allocator hardening, allocation cache behavior, and whether the attacker can influence reuse through the malicious backend or surrounding guest activity. `AutoDowngradeAllowed=NO` is appropriate because automatically treating this as an ordinary probe-failure crash would discard a concrete repeated lifetime mutation through stale heap pointers. The lack of a demonstrated arbitrary write does not make the primitive safe to automatically close. Operationally, the likely immediate consequence is guest kernel DoS during virtio-fs device probe failure. The higher-impact confidentiality, integrity, or privilege-escalation outcomes remain unproven and should not be treated as established solely from the double-free class. 5. One-sentence report phrase A malformed virtio-fs device can trigger a double-free during failed virtqueue setup because freed queue arrays remain referenced and are freed again during kobject release, causing guest kernel memory corruption and potential denial of service. 6. Manual review recommendation MANUAL CHECK REQUIRED A concrete double-free performs a second lifetime mutation on the same potentially freed heap allocations. Review allocator behavior, virtio-fs backend trust boundaries, VMM attacker roles, and possible allocation-reuse influence before downgrading. 7. Call-site confidence Call-site confidence: high ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: virtio-fs: avoid double-free on failed queue setup [ Upstream Commit: 3fbc5ee776fefbc6ea34d2518f1f5f90aa2e1dad Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=3fbc5ee776fefbc6ea34d2518f1f5f90aa2e1dad Changed files: fs/fuse/virtio_fs.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=3fbc5ee776fefbc6ea34d2518f1f5f90aa2e1dad ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-93827 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-93827 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:A/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:H The Best / paranoid CVSS vector: AV:A/AC:L/PR:N/UI:N/S:U/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: 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).