From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-90415][LOW] RDMA/cxgb4: free STAG index when TPT entry write fails [ Upstream Date: Fri, 18 Sep 2026 07:18:03 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-90415 X-AL-KERNEL-Priority: LOW X-AL-KERNEL-Severity: LOW X-AL-KERNEL-Base-Severity: LOW X-AL-KERNEL-KPANIC: NO X-AL-KERNEL-ActionableScore: 2 X-AL-KERNEL-ActionableScore-Lower: 1 X-AL-KERNEL-Commit: 5fe4731bfbd8d83c4a14b4af3a27329969ce8a99 List-Id: CVE: CVE-2026-90415 Priority: LOW AL-KERNEL base severity: LOW KPANIC flag: NO Patch: RDMA/cxgb4: free STAG index when TPT entry write fails [ Upstream Commit: 5fe4731bfbd8d83c4a14b4af3a27329969ce8a99 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=5fe4731bfbd8d83c4a14b4af3a27329969ce8a99 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90415 Analysis date: Fri, 18 Sep 2026 07:18:03 -0400 ActionableScore: 2 ActionableScore lower bound: 1 Actionable bucket: Ordinary Moderate Manual review required: NO Summary: No one-sentence report phrase was found. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-90415 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90415 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: LOW Manual review required: NO A detailed explanation of the methodology and priority rules is included at the end of this message. ====================================================================== AL-KERNEL CLASSIFICATION RESULT ====================================================================== CVE-2026-90415 LOW https://www.kernelcve.org/list/?q=CVE-2026-90415 CHECK WITH IMPACT FROM ORIG NN LOW Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:H';*CWE-772;*CWE-401;**CWE-400;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:L';BEST CVSS score: '4.7';DESCR 'RDMA cxgb4 can leak STAG indexes when write_tpt_entry() allocates a new STAG and write_adapter_mem() later fails before the MR is inserted into driver tracking structures. Repeated failures can exhaust the TPT/STAG resource pool and make memory registration or RDMA operation setup fail, causing local denial of service for the adapter or workloads using it. For the CVSS the PR:L is used because an attacker needs local access to the RDMA device or verbs interface, but does not need full administrator privileges in deployments where RDMA device nodes are delegated to users or containers. The issue is not directly network reachable because the vulnerable allocation and rollback path is reached by local memory registration/control operations, not by remote RDMA packets. Impact is resource exhaustion and availability loss only, with no evidence of UAF, OOB access, information disclosure, or privilege escalation.';NO MANUAL CHECK; ,and ActionableScore result is Ordinary Moderate (with actual score 2) YES SIMPLEFIX LEAK NETWORK HARDWARE LINUS DECREASED_TO_MODERATE_REGULAR_FROM_MODERATE7_BASED_ON_GUESSCVSS_AND_CIANNH DECREASED_TO_LOW_FROM_MODERATE_BASED_ON_ACTIONABLESCORELESSTHAN3 NO NO checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=2 ActionableScoreLower=1 PrimitiveClass=NONE RealUAF=NO HostAdminRequired=UNKNOWN NamespaceScopedPrivilege=NO ImportantCandidate=NO ManualReview=NO AutoDowngradeAllowed=YES 1. ActionableScore * Raw conservative score: 1 * Conservative score: 1 * Raw paranoid score: 2 * Paranoid score: 2 * Final recommended bucket: **Ordinary Moderate**:: No floor or cap changed the score. 2. Signal breakdown Conservative signals: * +1 Concrete persistent kernel resource/reference leak. `write_tpt_entry()` allocates a STAG index with `c4iw_get_resource()`, but on `write_adapter_mem()` failure the index is not returned and no MR is inserted into `rhp->mrs`, so normal deregistration cannot reclaim it. * +0 Reliable kernel crash not awarded. The patch supports RDMA adapter resource leakage and possible later allocation failures, not an immediate host-wide crash. * +0 Confidentiality and integrity not awarded. There is no UAF, OOB access, arbitrary write, disclosure path, ownership violation, or security-boundary bypass. * +0 Remote reachability not awarded. The vulnerable path is local RDMA memory-registration/control-path behavior, not direct processing of remote RDMA packets. * +0 Local unprivileged trigger not awarded. Local access to RDMA userspace interfaces may be delegated, but the supplied evidence does not prove complete ordinary-user reachability. * +0 Strong resource exhaustion not awarded conservatively. The leak is concrete, but practical exhaustion depends on repeated failures of `write_adapter_mem()`. Paranoid additional signal: * +1 Sustained finite-resource exhaustion. If an attacker or workload can repeatedly hit the failing write path, leaked STAG/TPT indexes can accumulate until RDMA memory registration or related operations fail for the device. Negative signals: * No host-admin penalty applied. Host-admin-equivalent privilege is not positively established. * No race/timing penalty applied. This is an ordered error-path leak, not a timing race. * No niche-subsystem penalty applied. RDMA/cxgb4 hardware is hardware-specific, but the prompt requires concrete rare/physical-only evidence before subtracting. Repeated release/refcount checklist: * First release can end lifetime: not applicable. * Exact same reference survives: not applicable. * Second lifetime/refcount mutation: not established. * +1 Concrete repeated release/refcount: not awarded. * Paranoid repeated-refcount leverage: not awarded. * FLOOR 5: not applied. 3. Reachability analysis The likely trigger is local activity that causes `write_tpt_entry()` to allocate a new STAG index and then encounter a `write_adapter_mem()` failure before the MR is inserted into normal driver tracking. This is associated with RDMA memory registration or related cxgb4 driver control operations. Ordinary unprivileged reachability is not fully established from the supplied patch. Some RDMA deployments expose verbs device nodes to non-root users or containers, but the complete ability to force the failing adapter-memory write is unresolved. Therefore `HostAdminRequired=UNKNOWN`, not YES and not NO. The issue is not directly network reachable. Remote RDMA traffic is not shown to reach the allocation-and-rollback bug by itself. A remote peer might influence workload behavior, but the concrete vulnerable operation is local driver resource programming. Namespace-scoped privilege is not established. No evidence shows that the complete RDMA/cxgb4 trigger chain is reachable with attacker-created namespace capabilities. Prior setup exists in the practical sense that the host must have cxgb4 RDMA hardware and the driver active. This is a deployment prerequisite, but not scored as a rare-hardware penalty without stronger evidence. 4. Primitive and severity interpretation The exact primitive is a persistent STAG/TPT resource leak on an error path. It is not memory corruption, not UAF, not double free, and not a repeated release/refcount mutation. The missing rollback leaves an allocated index unreachable by normal deregistration because no MR was inserted into `rhp->mrs`. Demonstrated consequence is local RDMA resource exhaustion over repeated failures, causing denial of service for memory registration or RDMA workloads on the affected adapter. The stronger paranoid interpretation is adapter-level availability loss if the finite STAG/TPT pool can be exhausted. There is no meaningful evidence of exploitation leverage beyond resource exhaustion. No controlled reclaim, arbitrary read/write, callback control, information disclosure, privilege escalation, or authorization bypass is supported. PrimitiveClass is `NONE` because the established issue is a resource leak without a corruption, lifetime, ownership, or security-boundary primitive. ImportantCandidate is NO because this is not an Important-class kernel primitive. AutoDowngradeAllowed is YES because the issue is suitable for ordinary Moderate/Low-like handling unless product-specific evidence shows unprivileged reliable exhaustion at scale. Likely operational severity: Low to ordinary Moderate, mainly affecting RDMA availability on systems using the cxgb4 driver. 5. One-sentence report phrase RDMA/cxgb4 can leak STAG indexes when TPT entry programming fails after allocation, allowing repeated failures to exhaust adapter RDMA resources and cause local denial of service without evidence of memory corruption or privilege escalation. 6. Manual review recommendation NO MANUAL CHECK NEEDED Reason: this is a concrete resource leak only, with no UAF, no OOB access, no authorization bypass, no remote packet trigger, and no demonstrated path to confidentiality or integrity impact. 7. Call-site confidence Call-site confidence: high ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: RDMA/cxgb4: free STAG index when TPT entry write fails [ Upstream Commit: 5fe4731bfbd8d83c4a14b4af3a27329969ce8a99 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=5fe4731bfbd8d83c4a14b4af3a27329969ce8a99 Changed files: drivers/infiniband/hw/cxgb4/mem.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=5fe4731bfbd8d83c4a14b4af3a27329969ce8a99 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-90415 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-90415 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:L The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:H The Best / paranoid CVSS score: 4.7 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: 1 Paranoid ActionableScore: 2 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: LOW KPANIC detected for this report: NO Published priority for this report (same as in Subject): LOW 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).