From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-98354][MODERATE REGULAR] RDMA/mad: Fix receive buffer leak when PKey enforcement fails [ Upstream Date: Thu, 08 Oct 2026 12:52:09 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-98354 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: 3 X-AL-KERNEL-Commit: 4617c9a856188674dddb0ca66979746d07688579 List-Id: CVE: CVE-2026-98354 Priority: MODERATE REGULAR AL-KERNEL base severity: MODERATE KPANIC flag: NO Patch: RDMA/mad: Fix receive buffer leak when PKey enforcement fails [ Upstream Commit: 4617c9a856188674dddb0ca66979746d07688579 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=4617c9a856188674dddb0ca66979746d07688579 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98354 Analysis date: Thu, 08 Oct 2026 12:52:09 -0400 ActionableScore: 4 ActionableScore lower bound: 3 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-2026-98354 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98354 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-2026-98354 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-98354 CHECK WITH IMPACT FROM ORIG NN IMPORTANT | AEGIS IMPACT MODERATE_KPANIC AEGIS_AS=1 LEGACY=MODERATE RESULT=DIFF (legacy and Aegis reconciliation differ; DECISION: legacy MODERATE vs Aegis MODERATE_KPANIC) Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:A/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H';*CWE-401;*CWE-772;**CWE-400;Other CVSS 'AV:A/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L';BEST CVSS score: '6.5';DESCR 'A receive buffer leak in the RDMA MAD receive path occurs when PKey enforcement rejects a management datagram before recv_buf is linked into rmpp_list. ib_free_recv_mad then finds an empty list and does not release the associated ib_mad_private allocation. A remote node on an InfiniBand or OPA fabric can repeatedly send MADs with an invalid PKey and gradually exhaust kernel memory. For the CVSS the PR:N because no privileges on the victim are required beyond access to the affected fabric. The attack vector is AV:A because the attacker needs adjacent access to the InfiniBand or OPA network and this service is normally not exposed to the public Internet. Impact is denial of service through resource exhaustion. Confidentiality and integrity impact are not indicated because the patch shows a memory leak without a memory corruption primitive.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Borderline / manual review recommended (with actual score 4) MAYBE LEAK ERRORPATH UAF NETWORK LINUS DECREASED_TO_MODERATE70_BASED_ON_ACTIONABLESCORELOWERTHAN5 DECREASED_TO_MODERATEREG_BASED_ON_FALSEPOSCHECKOFKP - - checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=4 ActionableScoreLower=3 PrimitiveClass=NONE RealUAF=NO HostAdminRequired=NO NamespaceScopedPrivilege=NO ImportantCandidate=NO ManualReview=RECOMMENDED AutoDowngradeAllowed=YES 1. ActionableScore * Raw conservative score: 3 * Conservative score: 3 * Raw paranoid score: 4 * Paranoid score: 4 * Final recommended bucket: **Borderline / manual review recommended**:: 2. Signal breakdown * +2 Directly network-triggerable from an adjacent RDMA fabric node. The commit explicitly states that a remote node can repeatedly send MADs with an invalid PKey. * +1 Concrete persistent kernel resource leak. On the PKey-enforcement error path, `ib_free_recv_mad()` receives an empty `rmpp_list`, while the caller has already transferred ownership and sets its receive pointer to NULL. The allocated `ib_mad_private` therefore cannot be recovered by normal cleanup. * +1 paranoid only: sustained resource-exhaustion potential. Each invalid MAD leaks roughly 300 bytes for IB port MADs or roughly 2 KB for OPA, and the attacker can repeat the request sequence. Not awarded: * No UAF, stale post-free use, double release, refcount mutation, OOB access, or arbitrary write is established. * No controlled reclaim, attacker-controlled kernel-memory replacement, callback influence, information disclosure, integrity impact, or privilege escalation is established. * No Reliable kernel crash / strong DoS point is awarded. The evidence establishes gradual memory growth, not a demonstrated immediate panic or deterministic host-wide exhaustion. * No generic niche-subsystem penalty is applied. InfiniBand and OPA are specialized, but the prompt provides no evidence that the attacker cannot reach an exposed fabric port. * No host-admin penalty is applied. The complete trigger described in the commit is a remote invalid MAD from a fabric-connected peer. 3. Reachability analysis A remote peer with access to the affected InfiniBand or OPA fabric can trigger the vulnerable receive path by transmitting management datagrams with an invalid PKey. No local code execution or administrative privilege on the victim is required, so HostAdminRequired is NO. The attack is adjacent-network reachable rather than Internet reachable. The attacker must be connected to, or otherwise have access to, the RDMA fabric that can deliver MAD traffic to the target port. This is appropriately modeled as AV:A, not AV:N. Namespace-scoped privilege is not relevant to the complete trigger. The vulnerable operation is reached from an external RDMA fabric peer, not through an attacker-created namespace capability or local control-plane interface. NamespaceScopedPrivilege is NO. No race, special timing window, device reset, firmware compromise, or prior host-side privileged configuration is established as necessary for the malformed-MAD trigger. The target must operate an affected RDMA MAD receive path on an accessible fabric port. 4. Primitive and severity interpretation The exact primitive is a persistent memory/resource leak caused by incorrect ownership cleanup on the PKey rejection path. `recv_buf` is not linked into `rmpp_list` before `ib_mad_enforce_security()` can fail. The error handler consequently finds no buffer to free, while the caller assumes ownership was transferred and does not retain a cleanup reference. This is neither active structured-object corruption nor fixed stale-storage UAF. There is no object freed before a later access, no second lifetime mutation, no refcount takeover, and no post-free object reuse. The issue is therefore PrimitiveClass NONE rather than CLASS_B or CLASS_C. The demonstrated consequence is progressive kernel memory consumption under repeated invalid MAD traffic. A hostile fabric peer may eventually create a denial of service if enough allocations accumulate, but the supplied patch does not establish memory corruption, confidentiality impact, integrity impact, cross-tenant accounting bypass, or privilege escalation. ImportantCandidate is NO because this is a resource leak without a concrete corruption or security-boundary primitive. AutoDowngradeAllowed is YES because no unresolved UAF, authorization bypass, shared-page ownership violation, or active lifetime corruption remains. Manual review is still recommended because the trigger is attacker-controlled network traffic on an RDMA fabric and the practical exhaustion rate depends on deployment-specific memory limits, MAD rate limiting, fabric isolation, and available kernel memory. 5. One-sentence report phrase A remote InfiniBand or OPA fabric peer can repeatedly send MADs with an invalid PKey to leak receive-buffer allocations in the RDMA MAD error path, potentially causing denial of service through sustained kernel memory exhaustion. 6. Manual review recommendation MANUAL CHECK RECOMMENDED The issue is remotely triggerable from an adjacent RDMA fabric and can accumulate kernel-memory leaks, but no memory-corruption or privilege-escalation primitive is established. 7. Call-site confidence Call-site confidence: high ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: RDMA/mad: Fix receive buffer leak when PKey enforcement fails [ Upstream Commit: 4617c9a856188674dddb0ca66979746d07688579 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=4617c9a856188674dddb0ca66979746d07688579 Changed files: drivers/infiniband/core/mad.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=4617c9a856188674dddb0ca66979746d07688579 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-98354 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-98354 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:N/I:N/A:L The Best / paranoid CVSS vector: AV:A/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H The Best / paranoid CVSS score: 6.5 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: 3 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).