From mboxrd@z Thu Jan 1 00:00:00 1970 From: AL-KERNEL To: kernel-cve@kernelcve.org Subject: [CVE-2026-89581][MODERATE 7.0] bpf, x86: Fix per-CPU address resolution into an extended register Date: Fri, 11 Sep 2026 22:59:19 -0400 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-AL-KERNEL-CVE: CVE-2026-89581 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: 5 X-AL-KERNEL-ActionableScore-Lower: unknown X-AL-KERNEL-Commit: 638bc3aada8ecdece184d5c15b100d489c9cccd7 List-Id: CVE: CVE-2026-89581 Priority: MODERATE 7.0 AL-KERNEL base severity: MODERATE KPANIC flag: YES Patch: bpf, x86: Fix per-CPU address resolution into an extended register Commit: 638bc3aada8ecdece184d5c15b100d489c9cccd7 Upstream patch: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=638bc3aada8ecdece184d5c15b100d489c9cccd7 Original CVE announcement: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89581 Analysis date: Fri, 11 Sep 2026 22:59:19 -0400 ActionableScore: 5 ActionableScore lower bound: unknown 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-2026-89581 is available here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89581 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-2026-89581 MODERATE https://www.kernelcve.org/list/?q=CVE-2026-89581 CHECK WITH IMPACT FROM ORIG NN MODERATE Maybe valid. Check manually. Hints by AL-KERNEL: The best (paranoid) CVSS is 'AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H';CWE-682;**CWE-476;CWE-119;Other CVSS 'AV:L/AC:H/PR:L/UI:N/S:U/C:N/I:N/A:H';BEST CVSS score: '7';DESCR 'An x86 BPF JIT code generation bug resolves per CPU addresses into the wrong register when the destination is an extended register. The REX prefix uses REX.B instead of REX.R, so destinations such as R5, R7, R8, and R9 alias lower registers and the intended BPF register can keep an unadjusted per CPU pointer. A local actor who can load and run affected BPF programs can trigger a kernel crash when the JITed program dereferences the wrong pointer. For the CVSS the PR:L value is used because reliable triggering requires local BPF program execution or delegated BPF capability rather than full root in many practical environments. The issue is not network reachable. Impact is at least denial of service via kernel panic. Because this is a verifier versus JIT semantic mismatch with possible frame pointer or argument register corruption, the paranoid impact includes potential confidentiality and integrity impact and should be manually reviewed.';YES REQUIRES MANUAL CHECK; ,and ActionableScore result is Actionable Moderate / Strong Important candidate requiring manual review (with actual score 5) YES OOB BPF HARDWARE LINUS TEST KPANIC INCREASED_TO_MODERATE70_BASED_ON_ACTIONABLESCOREHIGHEREQTHAN5 YES NO checked ====================================================================== ACTIONABLESCORE ANALYSIS ====================================================================== ActionableScore=5 PrimitiveClass=CLASS_A RealUAF=NO HostAdminRequired=UNKNOWN NamespaceScopedPrivilege=UNKNOWN ImportantCandidate=YES ManualReview=REQUIRED AutoDowngradeAllowed=NO 1. ActionableScore * Raw conservative score: 4 * Conservative score: 5 * Applied conservative floor/cap: 4 -> 5 due to Class A semantic reinterpretation in BPF JIT code generation with demonstrated kernel panic and unresolved register-corruption leverage. * Raw paranoid score: 5 * Paranoid score: 5 * Applied paranoid floor/cap: none. BPF privileged-control-plane cap does not reduce the score below 5. * Final recommended bucket: **Actionable Moderate / Strong Important candidate requiring manual review**:: 2. Signal breakdown Positive signals: * +2 Strong attacker-influenced semantic reinterpretation: BPF/JIT semantics diverge from the verifier and internal BPF register model because REX.B is emitted where REX.R is required. The same BPF operation is therefore executed against a different x86 register than intended. * +1 Reliable kernel crash / strong DoS: the supplied trace shows a kernel page fault and panic from JITed BPF execution. * +1 Weak exploitation concern: the patch explicitly states that R7 aliases RBP and can corrupt the frame pointer, while R8 and R9 alias argument registers. This is more than a generic crash, but it does not establish arbitrary read/write or controlled privilege escalation. * +1 Paranoid only, integrity concern from active register corruption: the wrong-register write affects live kernel execution state in JITed code. This supports manual review, but not automatic CLASS_A arbitrary-write scoring. Non-awarded signals: * RealUAF not awarded: no use-after-free, stale lifetime, RCU, or repeated release/refcount primitive is established. * Controlled reclaim not demonstrated. * Arbitrary or directed kernel write not demonstrated. * Confidentiality impact not awarded because no disclosure path is shown. * Strong LPE +2 not awarded because the patch does not establish credential manipulation, callback control, controlled object replacement, or arbitrary R/W. * Local unprivileged trigger not awarded because complete below-host-admin BPF load/JIT/execute reachability is not established. * Remote reachability not awarded. This is local BPF program execution, not network packet processing. * Extra availability not awarded because the same panic is already counted as reliable DoS. * Race/timing penalty not applied because this is deterministic JIT misencoding, not a timing race. * Host-admin penalty not applied because host-admin-equivalent privilege is not positively established from the supplied input. BPF access may be CAP_BPF/CAP_SYS_ADMIN/delegated depending on product policy. 3. Reachability analysis The trigger requires loading and executing an affected BPF program that reaches the internal per-CPU address resolution instruction with an extended destination register. The supplied context says the issue was exposed by BPF_GCC selftests because GCC kept several per-CPU addresses live at once, while clang typically avoided the affected register allocation pattern. For BPF/XDP privilege separation: * BPF load privilege: unresolved from the patch alone. Modern systems often require CAP_BPF, CAP_SYS_ADMIN, CAP_PERFMON, or restrictive sysctls, but the exact evaluated product policy is not supplied. * Attach privilege: unresolved. Some BPF program types and attach points require additional capabilities. * Netdevice privilege: not central to the supplied trigger. * Namespace privilege: not established. The patch does not prove that the complete load, JIT, attach, and trigger sequence works inside attacker-created namespaces. * Root-controlled sysctls: unprivileged BPF may be disabled by default, but product-specific state is not supplied. Therefore HostAdminRequired=UNKNOWN and NamespaceScopedPrivilege=UNKNOWN. The issue is not network reachable. Prior privileged setup may exist depending on BPF policy and JIT availability, but it is not proven as a universal prerequisite from the supplied patch. Race repeatability is not relevant. This is an instruction encoding bug, not a race. 4. Primitive and severity interpretation The exact primitive is a BPF JIT code-generation semantic mismatch on x86. The BPF operation intends to add the per-CPU offset into the destination register, but the generated REX prefix extends the wrong ModRM field. For extended registers, the hardware instruction therefore updates an aliased lower register and leaves the intended destination unchanged. This is not fixed stale storage and not UAF. It is active generated kernel code executing with incorrect register semantics. The demonstrated consequence is a kernel panic when the program dereferences a pointer that was not converted into a valid per-CPU address. The speculative but patch-supported concern is broader register corruption. R7 can corrupt RBP, and R8/R9 can corrupt argument registers. That creates meaningful unresolved exploitation leverage beyond a simple NULL dereference, but the evidence still does not prove arbitrary kernel read/write, controlled memory corruption, or privilege escalation. PrimitiveClass=CLASS_A because the patch directly establishes a strong verifier/JIT semantic reinterpretation of attacker-influenced BPF program behavior into different active machine-register behavior. ImportantCandidate=YES because BPF JIT semantic mismatches with frame-pointer or argument-register corruption should not be automatically treated as ordinary DoS. AutoDowngradeAllowed=NO because automatic downgrade would discard an unresolved active code-generation security primitive. Likely operational severity: actionable Moderate at minimum, with Important review justified. A final Important classification depends on product BPF reachability and whether further analysis can turn wrong-register corruption into a useful memory-access or privilege-escalation primitive. 5. One-sentence report phrase An x86 BPF JIT encoding bug can resolve per-CPU addresses into the wrong extended register, causing kernel panic and leaving unresolved frame-pointer or argument-register corruption risk for local actors able to load and run affected BPF programs. 6. Manual review recommendation MANUAL CHECK REQUIRED Reason: this is not only a crash trace. It is a verifier-versus-JIT semantic mismatch in generated kernel code, and the patch explicitly identifies possible frame-pointer and argument-register corruption. 7. Call-site confidence Call-site confidence: high ====================================================================== UPSTREAM PATCH SUMMARY ====================================================================== Patch: bpf, x86: Fix per-CPU address resolution into an extended register Commit: 638bc3aada8ecdece184d5c15b100d489c9cccd7 Upstream URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=638bc3aada8ecdece184d5c15b100d489c9cccd7 Commit description: The destination of the per-CPU address MOV is encoded in ModRM.reg, which is extended by REX.R, but the REX prefix is built with add_1mod(), which sets REX.B. REX.B extends ModRM.rm and SIB.base, and this instruction addresses memory as disp32 with no base, so the bit has no effect at all and the high register bit is simply lost. Every is_ereg() destination therefore resolves to the wrong register, picking whichever one shares the low three bits: R5 -> RAX R7 -> RBP R8 -> RSI R9 -> RDI With BPF_REG_5, whose reg2hex is 0, the emitted 65 49 03 04 25 add %gs:,%rax adds the per-CPU offset to RAX rather than R8. The destination keeps the unadjusted address and RAX is clobbered, so the program goes on to dereference a pointer that was never made per-CPU: BUG: unable to handle page fault for address: 0000607e386a8894 RIP: bpf_prog_707837aafd2aa9ae_update_percpu_data+0x93/0xc9 Call Trace: __bpf_prog_test_run_raw_tp+0x2dc/0x7d0 __flush_smp_call_function_queue+0x1e9/0xc80 Kernel panic - not syncing: Fatal exception in interrupt R5 is the mildest of the four, aliasing a scratch register and faulting at the store. R7 aliases RBP and would corrupt the frame pointer, R8 and R9 alias the argument registers. Use add_2mod() so the register goes through REX.R, matching how add_2reg() places it in ModRM.reg and how emit_priv_frame_ptr() hardcodes 0x4c for the same instruction with R9. Encodings for the non-extended registers are unchanged. Problem showed up when trying to resurrect BPF_GCC CI (selftests built with BPF_GCC). This has gone unnoticed because clang reloads the address into R1 before each per-CPU access, so the destination is never an extended register. GCC keeps several per-CPU addresses live at once, and test_progs-bpf_gcc panics the kernel in global_percpu_data/init, where the address of a .percpu variable ends up in R5. Fixes: 7bdbf74 ("bpf: add special internal-only MOV instruction to resolve per-CPU addrs") Signed-off-by: Vineet Gupta Reviewed-by: Eduard Zingerman Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260814220254.3797467-2-vineet.gupta@linux.dev Signed-off-by: Eduard Zingerman Signed-off-by: Greg Kroah-Hartman Changed files: arch/x86/net/bpf_jit_comp.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=638bc3aada8ecdece184d5c15b100d489c9cccd7 ====================================================================== DETAILED REPORT METHODOLOGY ====================================================================== The original Linux kernel CVE announcement for CVE-2026-89581 can be found here: https://lore.kernel.org/linux-cve-announce/?q=CVE-2026-89581 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:H The Best / paranoid CVSS vector: AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H The Best / paranoid CVSS score: 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: unknown Paranoid ActionableScore: 5 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).