{
  "affected": [
    {
      "ranges": [
        {
          "events": [
            {
              "introduced": "5b08bd408534bfb3a7cf5778da5b27d4e4fffe12"
            },
            {
              "fixed": "cf24ce48d60225fdbee68ed2058763a1fd172ff3"
            },
            {
              "fixed": "e89318aa8629969e86548b05e7c19283af3f0efb"
            },
            {
              "fixed": "9688a46802939da28f00cb40e8129615d5d4af39"
            }
          ],
          "repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
          "type": "GIT"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "Linux",
        "name": "Kernel"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "6.11.0"
            },
            {
              "fixed": "6.18.51"
            }
          ],
          "type": "ECOSYSTEM"
        },
        {
          "events": [
            {
              "introduced": "6.19.0"
            },
            {
              "fixed": "7.2.4"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "cna_assigner": "Linux",
    "osv_generated_from": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/89xxx/CVE-2026-89766.json"
  },
  "details": "In the Linux kernel, the following vulnerability has been resolved:\n\npidfd: hold exec_update_lock around namespace ioctl\n\nThe PIDFD_GET_*_NAMESPACE ioctls in pidfd_ioctl() perform a filesystem\ncredentials ptrace access check before handing out a namespace file\ndescriptor.  The accompanying comment states that the code \"mirrors nsfs\nbehavior\", but, unlike the corresponding procfs paths, it does so without\nholding the target task's exec_update_lock.\n\nproc_ns_get_link() and proc_ns_readlink() both take exec_update_lock for\nreading around the ptrace check and the namespace lookup, so that the\ncredentials used for the access decision match those of the task when its\nnamespace is read.  Without it, a caller can pass the check against the\ntarget's old credentials and then read the namespace after the target has\nexecve()'d a setuid binary and committed new credentials -- accessing\nnamespace information it should have been denied.\n\nHold exec_update_lock for reading around the ptrace check and the\nnamespace lookup so that pidfd truly mirrors nsfs behavior, as the comment\nalready claims.  open_namespace() itself runs outside the lock: once a\nnamespace reference is obtained it carries its own refcount and is opened\nwith the caller's own credentials, so a concurrent execve() on the target\ncan no longer affect the outcome.",
  "id": "CVE-2026-89766",
  "modified": "2026-09-13T03:30:18.331585421Z",
  "published": "2026-09-11T19:47:06.597Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://git.kernel.org/stable/c/9688a46802939da28f00cb40e8129615d5d4af39"
    },
    {
      "type": "WEB",
      "url": "https://git.kernel.org/stable/c/cf24ce48d60225fdbee68ed2058763a1fd172ff3"
    },
    {
      "type": "WEB",
      "url": "https://git.kernel.org/stable/c/e89318aa8629969e86548b05e7c19283af3f0efb"
    },
    {
      "type": "ADVISORY",
      "url": "https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/89xxx/CVE-2026-89766.json"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-89766"
    },
    {
      "type": "PACKAGE",
      "url": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git"
    }
  ],
  "schema_version": "1.9.0",
  "summary": "pidfd: hold exec_update_lock around namespace ioctl"
}