{ "dataType": "CVE_RECORD", "dataVersion": "5.2", "cveMetadata": { "cveId": "CVE-2025-40169", "assignerOrgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "state": "PUBLISHED", "assignerShortName": "Linux", "dateReserved": "2025-04-16T07:20:57.176Z", "datePublished": "2025-11-12T10:46:51.736Z", "dateUpdated": "2026-08-05T12:08:14.652Z" }, "containers": { "cna": { "providerMetadata": { "orgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67", "shortName": "Linux", "dateUpdated": "2026-08-05T12:08:14.652Z" }, "descriptions": [ { "lang": "en", "value": "In the Linux kernel, the following vulnerability has been resolved:\n\nbpf: Reject negative offsets for ALU ops\n\nWhen verifying BPF programs, the check_alu_op() function validates\ninstructions with ALU operations. The 'offset' field in these\ninstructions is a signed 16-bit integer.\n\nThe existing check 'insn->off > 1' was intended to ensure the offset is\neither 0, or 1 for BPF_MOD/BPF_DIV. However, because 'insn->off' is\nsigned, this check incorrectly accepts all negative values (e.g., -1).\n\nThis commit tightens the validation by changing the condition to\n'(insn->off != 0 && insn->off != 1)'. This ensures that any value\nother than the explicitly permitted 0 and 1 is rejected, hardening the\nverifier against malformed BPF programs." } ], "metrics": [ { "cvssV3_1": { "version": "3.1", "vectorString": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H", "baseScore": 7.8, "baseSeverity": "HIGH" }, "scenarios": [ { "lang": "en", "value": "AV:L - The vulnerable code is reached through the local `bpf(BPF_PROG_LOAD)` syscall, which invokes `bpf_check()` → `do_check()` → `check_alu_op()`. Although the resulting program later executes on packet receive, loading the malformed instruction requires local system access.\nAC:L - Exploitation is fully deterministic: emit a `BPF_DIV`/`BPF_MOD` instruction with `off = -1`, set the dividend to `LLONG_MIN` and the divisor to `-1`, and run it. There is no race, no timing window, and no dependence on memory layout or system state the attacker cannot control.\nPR:L - `sysctl_unprivileged_bpf_disabled` defaults to 0 unless `CONFIG_BPF_UNPRIV_DEFAULT_OFF` is built in (Android and many embedded/distro configs leave it enabled), and `kernel/bpf/syscall.c:2941` explicitly permits `BPF_PROG_TYPE_SOCKET_FILTER` and `BPF_PROG_TYPE_CGROUP_SKB` loads with no `bpf_cap`; the program is then triggered via `SO_ATTACH_BPF` on the attacker's own socket. Where unprivileged BPF is disabled, `bpf_token_capable()` grants `ns_capable()`-level CAP_BPF inside a delegated user namespace, so real root is never required.\nUI:N - The attacker loads and triggers the BPF program entirely on their own, by attaching it to a socket they own and sending themselves a packet. No victim action of any kind is involved.\nS:U - The verifier bypass, the JIT'd signed division, and the resulting fault all occur within the kernel's own security authority. No VM, IOMMU, or hypervisor boundary is crossed.\nC:H - On the s390 JIT the `switch (off)` in the DIV/MOD handlers has no default, so no divide instruction is emitted while the trailing `lgr %dst,%rc` still executes, copying the uninitialized kernel scratch register `%r0`/`%r1` — potentially a kernel pointer — into a BPF register that the program can write to a map and read from userspace, repeatably and at attacker-chosen program points. The x86 divide-error oops additionally dumps full register and stack state, defeating KASLR.\nI:H - The verifier accepts an instruction whose runtime semantics it never models — the JIT executes a signed division while the fixup layer applies only the unsigned guards, and on s390 uncontrolled kernel register contents are injected into verified program state. This is a verifier validation bypass in which JIT-emitted code and the verifier's model of that code diverge, so integrity of the sandbox's guarantees is not preserved.\nA:H - Because `is_sdiv`/`is_smod` in `do_misc_fixups()` test `insn->off == 1`, the `LLONG_MIN sdiv -1` overflow guards are skipped while the x86-64 JIT still emits `cqo; idiv`, producing a #DE with no exception-table entry — `do_trap()` reaches `die(\"divide error\")`, oopsing the kernel. Triggered from a socket-filter/tc/XDP program this fires in softirq context, yielding \"Fatal exception in interrupt\" and a full system panic." } ] } ], "affected": [ { "product": "Linux", "vendor": "Linux", "defaultStatus": "unaffected", "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git", "programFiles": [ "kernel/bpf/verifier.c" ], "versions": [ { "version": "ec0e2da95f72d4a46050a4d994e4fe471474fd80", "lessThan": "3bce44b344040e5eef3d64d38b157c15304c0aab", "status": "affected", "versionType": "git" }, { "version": "ec0e2da95f72d4a46050a4d994e4fe471474fd80", "lessThan": "5017c302ca4b2a45149ad64e058fa2d5623c068f", "status": "affected", "versionType": "git" }, { "version": "ec0e2da95f72d4a46050a4d994e4fe471474fd80", "lessThan": "21167bf70dbe400563e189ac632258d35eda38b5", "status": "affected", "versionType": "git" }, { "version": "ec0e2da95f72d4a46050a4d994e4fe471474fd80", "lessThan": "55c0ced59fe17dee34e9dfd5f7be63cbab207758", "status": "affected", "versionType": "git" } ] }, { "product": "Linux", "vendor": "Linux", "defaultStatus": "affected", "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git", "programFiles": [ "kernel/bpf/verifier.c" ], "versions": [ { "version": "6.6", "status": "affected" }, { "version": "0", "lessThan": "6.6", "status": "unaffected", "versionType": "semver" }, { "version": "6.6.112", "lessThanOrEqual": "6.6.*", "status": "unaffected", "versionType": "semver" }, { "version": "6.12.53", "lessThanOrEqual": "6.12.*", "status": "unaffected", "versionType": "semver" }, { "version": "6.17.3", "lessThanOrEqual": "6.17.*", "status": "unaffected", "versionType": "semver" }, { "version": "6.18", "lessThanOrEqual": "*", "status": "unaffected", "versionType": "original_commit_for_fix" } ] } ], "cpeApplicability": [ { "nodes": [ { "operator": "OR", "negate": false, "cpeMatch": [ { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.6", "versionEndExcluding": "6.6.112" }, { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.6", "versionEndExcluding": "6.12.53" }, { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.6", "versionEndExcluding": "6.17.3" }, { "vulnerable": true, "criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*", "versionStartIncluding": "6.6", "versionEndExcluding": "6.18" } ] } ] } ], "references": [ { "url": "https://git.kernel.org/stable/c/3bce44b344040e5eef3d64d38b157c15304c0aab" }, { "url": "https://git.kernel.org/stable/c/5017c302ca4b2a45149ad64e058fa2d5623c068f" }, { "url": "https://git.kernel.org/stable/c/21167bf70dbe400563e189ac632258d35eda38b5" }, { "url": "https://git.kernel.org/stable/c/55c0ced59fe17dee34e9dfd5f7be63cbab207758" } ], "title": "bpf: Reject negative offsets for ALU ops", "x_generator": { "engine": "bippy-1.2.0" } } } }