{"id":"CVE-2026-98989","title":"namespace: check a recursive bind mount for mount namespace loops","summary":"In the Linux kernel, the following vulnerability has been resolved:\n\nnamespace: check a recursive bind mount for mount namespace loops\n\ndo_loopback() refuses to bind mount the file of a mount namespace that\nis as old as the caller's or o…","severity":"none","vendor":"Linux","product":"Linux","affected":["Linux >= e149ed2b805fefdccf7ccdfc19eca22fdd4514ac < a0ba48256f5dff6f75a45f8ce39fe57f21df1022","Linux >= e149ed2b805fefdccf7ccdfc19eca22fdd4514ac < 37a82f400efa4d263542a5fa41755b21057f1e5d","Linux >= e149ed2b805fefdccf7ccdfc19eca22fdd4514ac < c640a01a78e42302879006ed7b7d5a9a943ba815","Linux 3.19"],"published":"2026-10-23","updated":"2026-10-23","sourceUpdated":"2026-10-23T12:20:08.366Z","source":"CVEORG","sourceUrl":"https://www.cve.org/CVERecord?id=CVE-2026-98989","references":[{"url":"https://git.kernel.org/stable/c/a0ba48256f5dff6f75a45f8ce39fe57f21df1022"},{"url":"https://git.kernel.org/stable/c/37a82f400efa4d263542a5fa41755b21057f1e5d"},{"url":"https://git.kernel.org/stable/c/c640a01a78e42302879006ed7b7d5a9a943ba815"}],"tags":["cve.org"],"ingestedAt":"2026-10-23T13:24:55.546Z","slug":"CVE-2026-98989","body":"## Overview\n\nIn the Linux kernel, the following vulnerability has been resolved:\n\nnamespace: check a recursive bind mount for mount namespace loops\n\ndo_loopback() refuses to bind mount the file of a mount namespace that\nis as old as the caller's or older because a mount namespace that holds\na mount of its own file, or of an ancestor's, would cause a cycle. But\nit only checks the dentry the bind mount starts from. With MS_REC\neverything below it is copied and do_loopback() passes\nCL_COPY_MNT_NS_FILE so mount namespace files below the source are\ncopied.\n\nThat's fine for a source in the caller's own mount namespace. Every\nmount namespace file in there passed the same check when it was\nbind-mounted. It isn't fine for a source in another mount namespace.\nmay_copy_tree() accepts a bind mount of any nsfs or pidfs file no matter\nwhat mount namespace it lives in so that /proc/<pid>/ns/<ns> and pidfds\ncan be bind mounted. If we create a bind-mount stack of mount namespaces\nfile descriptors the kernel will copy them irrespective of their\nancestoral relationship to the mount namespace in question:\n\n  151 146 0:7 net:[4026531833] /tmp/nrc/y rw - nsfs nsfs rw\n  152 151 0:7 mnt:[4026532293] /tmp/nrc/y rw - nsfs nsfs rw\n\nMount 152 is a mount of the mount namespace file of mount namespace\n4026532293 inside mount namespace 4026532293. That namespace and every\nmount in it are leaked... All it takes is a process in an older mount\nnamespace that stacks the file and repeating it leaks without limit:\n\n  after control: Shmem:               380 kB\n  child: mount(/proc/self/fd/6, MS_BIND|MS_REC): ok\n  after cycle: Shmem:             65916 kB\n\ndo_move_mount() runs check_for_nsfs_mounts() over a detached tree for\nexactly that reason. Do the same for the copy before it is grafted.\n\n## Affected\n\n- `Linux >= e149ed2b805fefdccf7ccdfc19eca22fdd4514ac < a0ba48256f5dff6f75a45f8ce39fe57f21df1022`\n- `Linux >= e149ed2b805fefdccf7ccdfc19eca22fdd4514ac < 37a82f400efa4d263542a5fa41755b21057f1e5d`\n- `Linux >= e149ed2b805fefdccf7ccdfc19eca22fdd4514ac < c640a01a78e42302879006ed7b7d5a9a943ba815`\n- `Linux 3.19`\n\n## Remediation\n\nRefer to the linked advisories for vendor-supplied fixes and affected version ranges.","depth":"sunlit","depthScore":3,"depthScoreParts":{"impact":2.8,"likelihood":0,"exploitation":0,"ransomware":0},"changes":[]}