From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-97409][MODERATE REGULAR] nvme-fc: Do not cancel requests in io target before it is initialized [ Upstream Date: Thu, 24 Sep 2026 14:27:53 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-97409 X-AL-KERNEL-Priority: MODERATE REGULAR X-AL-KERNEL-Severity: MODERATE REGULAR X-AL-KERNEL-Base-Severity: MODERATE X-AL-KERNEL-KPANIC: NO X-AL-KERNEL-ActionableScore: 4 X-AL-KERNEL-ActionableScore-Lower: unknown X-AL-KERNEL-Commit: da930308b4bc24c335574e85ca84ff70a988cf05 List-Id: CVE: CVE-2026-97409 Priority: MODERATE REGULAR AL-KERNEL base severity: MODERATE KPANIC flag: NO Patch: nvme-fc: Do not cancel requests in io target before it is initialized [ Upstream Commit: da930308b4bc24c335574e85ca84ff70a988cf05 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=da930308b4bc24c335574e85ca84ff70a988cf05 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97409 Analysis date: Thu, 24 Sep 2026 14:27:53 -0400 ActionableScore: 4 ActionableScore lower bound: unknown Actionable bucket: Borderline / manual review recommended Manual review required: YES Summary: No one-sentence report phrase was found. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-97409 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97409 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 REGULAR 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-2026-97409 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-97409 CHECK WITH IMPACT FROM ORIG NN LOW Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H';CWE-362;CWE-908;CWE-667;Other CVSS 'AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:L';BEST CVSS score: '5.3';DESCR 'An NVMe FC controller in CONNECTING state can schedule ioerr_work after an admin request timeout before the IO tagset has been initialized. The worker can then attempt to iterate IO requests based on queue_count even though the IO tagset does not exist. A related race can occur when ioerr_work runs while the IO tagset is being removed after a failed queue setup. This was observed as a lockdep warning and may cause a local availability impact through a stalled controller recovery path or failed storage connectivity. For the CVSS the PR:N because the triggering condition can originate from a remote NVMe FC endpoint and does not require privileges on the affected host. The attack surface is adjacent network access to the Fibre Channel fabric rather than public Internet access. The paranoid score retains C:N and I:N because the patch does not demonstrate a UAF dereference or another memory corruption primitive. Impact is at least denial of service through controller or storage path disruption. YES REQUIRES MANUAL CHECK.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4) YES LOCK WARNONLY NVME LINUS INCREASED_TO_MODERATE7_FROM_LOW_BASED_ON_GUESSCVSS_AND_ACTIONABLESCOREHIGHER4 DECREASED_TO_MODERATEREG_BASED_ON_FALSEPOSCHECKOFKP NO NO checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=4 PrimitiveClass=CLASS_B RealUAF=YES HostAdminRequired=UNKNOWN NamespaceScopedPrivilege=UNKNOWN ImportantCandidate=NO ManualReview=RECOMMENDED AutoDowngradeAllowed=YES 1. ActionableScore * Raw conservative score: 2 * Conservative score: 4 * Raw paranoid score: 2 * Paranoid score: 4 * Applied conservative floor: Real UAF floor 4, 2 -> 4. * Applied paranoid floor: Real UAF floor 4, 2 -> 4. * Final recommended bucket: **Borderline / manual review recommended**:: 2. Signal breakdown * +2 Real UAF or active lifetime violation. `ioerr_work` can execute `blk_mq_tagset_busy_iter()` while the IO tagset is absent or being removed. The added `cancel_work_sync()` before `nvme_remove_io_tag_set()` directly establishes that workqueue execution must be synchronized with IO tagset teardown. * +0 Reliable kernel crash or strong DoS. The supplied evidence shows a lockdep warning caused by using an uninitialized tagset. It does not demonstrate an Oops, panic, or host-wide crash. * +0 Remote reachable. An NVMe FC peer or fabric event may contribute to the timeout and recovery path, but the supplied material does not establish that an attacker can independently create or impersonate the required Fibre Channel target connection. * +0 Local unprivileged trigger. No evidence establishes that an ordinary local user can configure or drive the complete NVMe FC controller connection and failure path. * +0 Strong exploitation leverage. Controlled reclaim, replacement of freed tagset internals, callback control, arbitrary read, arbitrary write, and refcount takeover are not demonstrated. * +0 Structured stale-object exploitation leverage. `blk_mq_tagset_busy_iter()` consumes active block-layer state, but the patch does not establish attacker control over post-free tagset contents or a meaningful reused-object interpretation. * +0 Concrete repeated release or refcount mutation. The patch prevents stale iteration during teardown. It does not show a second put, free, release, or refcount mutation through the same dangling reference. * +0 Confidentiality and integrity impact. No disclosure, write primitive, policy bypass, or protected-object modification is established. * +0 Extra availability impact. The observed warning and any possible stale access are the same underlying availability concern. No distinct persistent availability consequence is shown. * No race penalty. The failure sequence involves asynchronous work and teardown, but the supplied evidence does not prove an unusually narrow, one-shot, externally uncontrollable timing window. 3. Reachability analysis The immediate trigger is an NVMe FC controller entering `CONNECTING`, followed by an administrative request timeout or connection failure that schedules `ctrl->ioerr_work`. The vulnerable conditions arise when the IO tagset has not yet been initialized, or when queue setup fails and the IO tagset is removed while error work remains runnable. A remote NVMe FC target or an attacker with access to the Fibre Channel fabric could potentially influence connection failures or timeouts. However, the evidence does not establish that an unprivileged attacker can create the required controller, attach to the relevant FC fabric, impersonate the target, or perform the complete trigger sequence. Therefore, below-host-admin reachability is not positively established. No namespace-scoped privilege path is shown. NVMe FC controller creation, FC transport setup, host bus adapter configuration, and access to the relevant storage fabric may require prior administrative provisioning, but the exact host privilege boundary is not visible. `HostAdminRequired=UNKNOWN` is therefore appropriate rather than assuming either host-admin-only or ordinary-user reachability. The issue depends on NVMe over Fibre Channel support, an FC HBA, and an existing or attempted FC controller connection. These are deployment prerequisites, but they are not scored as a negative niche penalty because the supplied evidence does not establish that the attacker cannot reach an already deployed NVMe FC environment. 4. Primitive and severity interpretation The patch establishes an asynchronous lifetime synchronization defect. `ctrl->ioerr_work` can enter `__nvme_fc_abort_outstanding_ios()` and iterate the IO tagset based on `queue_count`, even though `ctrl->tagset` has not been initialized. Separately, queued or executing error work can overlap IO tagset removal after a failed queue setup. This is an active structured block-layer tagset lifetime issue rather than a simple value-validation defect. The potentially stale operation is `blk_mq_tagset_busy_iter()` over tagset internals after the tagset lifetime is no longer guaranteed. The added `cancel_work_sync()` before tagset removal is direct evidence that worker completion is required before destroying the IO tagset resources. The available evidence supports `CLASS_B` and `RealUAF=YES`, because asynchronous worker access can outlive the relevant tagset resources. It does not establish controlled reclaim, attacker-controlled replacement contents, a directed memory access, a second lifetime-mutating operation, or a concrete guest, container, or remote privilege boundary bypass. The demonstrated result is a lockdep warning for an uninitialized tagset. A controller recovery failure, stalled storage path, or kernel fault remains plausible, but an actual crash is not shown. This is therefore a constrained asynchronous teardown UAF rather than an Important-preserving corruption primitive. ImportantCandidate is `NO`, and automatic downgrade is allowed after normal subsystem verification because meaningful exploitation leverage beyond crash or storage disruption is not established. 5. One-sentence report phrase An NVMe FC error-recovery worker can iterate an uninitialized or concurrently removed IO tagset during controller connection failure, causing a lockdep warning and potential storage-path denial of service. 6. Manual review recommendation MANUAL CHECK RECOMMENDED The patch directly fixes asynchronous worker versus tagset teardown ordering, so the exact behavior of `nvme_remove_io_tag_set()` and `blk_mq_tagset_busy_iter()` in the affected kernel version should be checked for a reachable post-free dereference or only a warning-level invalid-state path. 7. Call-site confidence Call-site confidence: medium ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: nvme-fc: Do not cancel requests in io target before it is initialized [ Upstream Commit: da930308b4bc24c335574e85ca84ff70a988cf05 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=da930308b4bc24c335574e85ca84ff70a988cf05 Changed files: drivers/nvme/host/fc.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=da930308b4bc24c335574e85ca84ff70a988cf05 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-97409 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-97409 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:H/PR:N/UI:N/S:U/C:N/I:N/A:L The Best / paranoid CVSS vector: AV:A/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:H The Best / paranoid CVSS score: 5.3 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: unknown Paranoid ActionableScore: 4 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: NO Published priority for this report (same as in Subject): MODERATE REGULAR 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).