{
  "affected": [
    {
      "ranges": [
        {
          "events": [
            {
              "introduced": "8eb76cb03f0f6c2bd7b15cf45dcffcd6bd07a360"
            },
            {
              "fixed": "c16c62c91e5d6737a9fbb0a541b8b34610470a2b"
            },
            {
              "fixed": "75b0a6db4300e4c2c9e97a0848deaa7acfb42fb7"
            }
          ],
          "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
          "type": "GIT"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "Linux",
        "name": "Kernel"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "7.0.0"
            },
            {
              "fixed": "7.2.7"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "cna_assigner": "Linux",
    "osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/98xxx/CVE-2026-98149.json"
  },
  "details": "In the Linux kernel, the following vulnerability has been resolved:\n\nbpf: Fix percpu map update indexing with sparse CPU IDs\n\nPer-CPU array, hash, and cgroup storage map updates without BPF_F_CPU\nor BPF_F_ALL_CPUS use a value buffer whose per-CPU slots are packed in\npossible-CPU order. The buffer is sized as:\n\n  round_up(value_size, 8) * num_possible_cpus()\n\nThe update paths iterate over possible CPUs, but use the logical CPU ID\nto calculate the source offset:\n\n  value + size * cpu\n\nThis only works when possible CPU IDs are contiguous starting at zero.\n\nFor example, with a possible CPU mask of 0,2-3, the buffer contains\nthree slots corresponding to CPUs 0, 2, and 3. CPU2 is therefore\nexpected to use slot 1 and CPU3 slot 2. Instead, the current code uses\nslots 2 and 3 respectively, causing incorrect per-CPU values and an\nout-of-bounds read from the update buffer for CPU3.\n\nThe corresponding lookup paths already use a dense offset while\niterating over possible CPUs. Do the same for the array, hash, and\ncgroup storage update paths, advancing the source offset once for each\npossible CPU. BPF_F_ALL_CPUS continues to use the same value for every\nCPU.",
  "id": "CVE-2026-98149",
  "modified": "2026-09-27T03:30:57.386892019Z",
  "published": "2026-09-25T10:36:22.182Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://git.kernel.org/stable/c/75b0a6db4300e4c2c9e97a0848deaa7acfb42fb7"
    },
    {
      "type": "WEB",
      "url": "https://git.kernel.org/stable/c/c16c62c91e5d6737a9fbb0a541b8b34610470a2b"
    },
    {
      "type": "ADVISORY",
      "url": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/98xxx/CVE-2026-98149.json"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-98149"
    },
    {
      "type": "PACKAGE",
      "url": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git"
    }
  ],
  "schema_version": "1.9.0",
  "summary": "bpf: Fix percpu map update indexing with sparse CPU IDs"
}