From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-52908][IMPORTANT] RDMA: During rereg_mr ensure that REREG_ACCESS is compatible [ Upstream Date: Fri, 19 Jun 2026 10:58:03 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-52908 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: 8 X-AL-KERNEL-ActionableScore-Lower: 6 X-AL-KERNEL-Commit: 09dc18894148381d3bfc550083b1236043870dce List-Id: CVE: CVE-2026-52908 Priority: IMPORTANT AL-KERNEL base severity: IMPORTANT KPANIC flag: YES Patch: RDMA: During rereg_mr ensure that REREG_ACCESS is compatible [ Upstream Commit: 09dc18894148381d3bfc550083b1236043870dce Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=09dc18894148381d3bfc550083b1236043870dce Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-52908 Analysis date: Fri, 19 Jun 2026 10:58:03 -0400 ActionableScore: 8 ActionableScore lower bound: 6 Actionable bucket: Strong Important candidate / Actionable Moderate at minimum Manual review required: YES Summary: RDMA rereg_mr could change a read-only pinned user memory region to writable RDMA access without revalidating writable pinning, potentially violating GUP and COW protections and allowing unauthorized memory modification from a local RDMA-capable process. ====================================================================== ABOUT THIS REPORT ====================================================================== The original Linux kernel CVE announcement for CVE-2026-52908 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-52908 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-2026-52908 IMPORTANT CHECK Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H';CWE-863;CWE-284;CWE-266;Other CVSS 'AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:H/A:H';BEST CVSS score: '7.8';DESCR 'RDMA rereg_mr could allow an existing user memory region to change from read only to writable access without rechecking that the underlying umem was pinned as writable. This can violate GUP and COW permission assumptions because the same pinned pages may later be exposed to writable RDMA access even though they were originally pinned read only. For the CVSS the PR:L is used because a local user or process with access to the RDMA uverbs device can request MR reregistration without full administrator privileges in many deployments. The issue is not directly network reachable because the vulnerable operation is a local RDMA control path operation. A remote RDMA peer may only become relevant after a local process exposes the MR and rkey. Impact is at least denial of service or memory corruption through unintended writes and may allow privilege escalation depending on the mapped memory and device configuration.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Strong Important candidate / Actionable Moderate at minimum (with actual score 8) YES RACE NETWORK KPANIC NO NO checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=8 ActionableScoreLower=6 ## 1. ActionableScore * Conservative score: 6 * Paranoid score: 8 * Final recommended bucket: **Strong Important candidate / Actionable Moderate at minimum**:: ## 2. Signal breakdown Conservative signals: * Local unprivileged trigger: +1 The vulnerable operation is exposed through RDMA uverbs rereg_mr paths. In many deployments access to RDMA devices is granted to non-root users or service accounts. * Dirty-Pipe-like ownership/COW violation: +2 The bug allows an MR originally backed by read-only pinned umem to be changed to writable access without revalidating writable pinning. This can violate GUP/COW assumptions and expose pages to writes that were not authorized as writable. * Integrity impact plausible: +1 Writable RDMA access to pages not pinned writable can corrupt memory contents that should have remained read-only or COW-protected. * Availability impact realistic: +1 Incorrect RDMA writes into improperly pinned memory can plausibly cause process or kernel instability, data corruption, or crash-like failure modes depending on the mapping and device behavior. * Privileged device-management / DMA-adjacent permission path: +1 The affected code controls RDMA memory registration and DMA-visible page permissions, which is a high-value kernel/device boundary. * Rare/configuration-dependent exposure: -1 Practical exposure requires RDMA stack availability and access to an RDMA-capable device or rxe-like setup. This is not a universal default desktop/server exposure. Paranoid additional signal: * Page-cache or file-backed/COW corruption concern: +2 If the original read-only umem can reference file-backed or COW-shared pages, later writable RDMA access may become a page ownership or page-cache corruption primitive. This is technically defensible from the RO-to-RW without writable pinning bug class, but depends on mapping details and should be manually verified. ## 3. Reachability analysis The bug is not directly network-triggerable. The vulnerable operation is a local RDMA control path through MR reregistration. A remote RDMA peer may become relevant only after a local process exposes the MR and rkey, but the unsafe permission transition is initiated locally. Typical triggering requires access to RDMA uverbs. This is often not full root, because RDMA device access may be delegated to users, HPC jobs, containers, or service accounts. Therefore treating it as PR:L is more realistic than PR:H for triage. Containers or namespaces can matter if RDMA devices are passed through or if rxe is configured, but the exposure is still deployment-dependent. ## 4. Severity interpretation This is stronger than an ordinary Moderate because it is not just a crash or validation cleanup. The patch prevents a transition from read-only to writable RDMA memory access when the underlying umem was not pinned writable, which is a meaningful memory permission and COW boundary issue. Realistic exploitation evidence is not equivalent to a demonstrated arbitrary kernel write or reliable LPE chain. However, the primitive is security-relevant because RDMA writes can affect memory outside normal CPU page permission checks. Conservative handling is Actionable Moderate. Paranoid handling is a Strong Important candidate due to possible page ownership or COW bypass impact. ## 5. One-sentence report phrase RDMA rereg_mr could change a read-only pinned user memory region to writable RDMA access without revalidating writable pinning, potentially violating GUP and COW protections and allowing unauthorized memory modification from a local RDMA-capable process. ## 6. Manual review recommendation MANUAL CHECK REQUIRED This needs manual review because the patch affects RDMA memory registration permissions and may create a writable access path to pages that were only pinned read-only. The key question is whether practical mappings can turn this into file-backed page corruption, COW bypass, or privilege escalation rather than only process-local memory corruption. ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: RDMA: During rereg_mr ensure that REREG_ACCESS is compatible [ Upstream Commit: 09dc18894148381d3bfc550083b1236043870dce Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=09dc18894148381d3bfc550083b1236043870dce Changed files: drivers/infiniband/core/umem.c drivers/infiniband/hw/hns/hns_roce_mr.c drivers/infiniband/hw/irdma/verbs.c drivers/infiniband/hw/mlx4/mr.c drivers/infiniband/hw/mlx5/mr.c drivers/infiniband/sw/rxe/rxe_verbs.c include/rdma/ib_umem.h 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=09dc18894148381d3bfc550083b1236043870dce ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-52908 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-52908 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:U/C:N/I:H/A:H The Best / paranoid CVSS vector: AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H The Best / paranoid CVSS score: 7.8 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: 8 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).