{"id":"CVE-2026-74618","title":"In the Linux kernel, the following vulnerability has been resolved:\n\nbinfmt_misc: don't warn when the mount is completed from another user namespace\n\nfsopen() records the caller's user namespace in fc->user_ns and hands\nback an ordinary …","summary":"In the Linux kernel, the following vulnerability has been resolved:\n\nbinfmt_misc: don't warn when the mount is completed from another user namespace\n\nfsopen() records the caller's user namespace in fc->user_ns and hands\nback an ordinary …","severity":"none","published":"2026-08-22","updated":"2026-10-14","sourceUpdated":"2026-10-14T13:10:00.187","source":"NVD","sourceUrl":"https://nvd.nist.gov/vuln/detail/CVE-2026-74618","references":[{"url":"https://git.kernel.org/stable/c/047f927f54c6c17593e93aafe82dcb7acdda2a71","label":"416baaa9-dc9f-4396-8d5f-8c081fb06d67"},{"url":"https://git.kernel.org/stable/c/24e95a24f151ce40d5fc1b3a6cefbcda8ded736c","label":"416baaa9-dc9f-4396-8d5f-8c081fb06d67"},{"url":"https://git.kernel.org/stable/c/37cf5cf1320a84a17225a1690547b8a0812ca94e","label":"416baaa9-dc9f-4396-8d5f-8c081fb06d67"},{"url":"https://git.kernel.org/stable/c/79fdf39f1a31f88cb3833b6f8091fbf6acdca2c6","label":"416baaa9-dc9f-4396-8d5f-8c081fb06d67"}],"tags":["nvd"],"epss":0.00168,"epssPercentile":0.05618,"ingestedAt":"2026-10-14T13:30:58.517Z","slug":"CVE-2026-74618","body":"## Overview\n\nIn the Linux kernel, the following vulnerability has been resolved:\n\nbinfmt_misc: don't warn when the mount is completed from another user namespace\n\nfsopen() records the caller's user namespace in fc->user_ns and hands\nback an ordinary file descriptor. Nothing ties the task that calls\nfsconfig(FSCONFIG_CMD_CREATE) to the task that created the context. The\nfd is inherited across fork() and exec() and it can be passed over a\nunix socket.\n\nCompleting a context from another user namespace is allowed on purpose.\nvfs_cmd_create() authorizes the create with mount_capable(), which for\nFS_USERNS_MOUNT checks ns_capable(fc->user_ns, CAP_SYS_ADMIN), and that\nsucceeds for a task holding CAP_SYS_ADMIN in an ancestor of fc->user_ns.\nSo an unprivileged task can reach the WARN_ON() in bm_fill_super():\ncreate a user and a mount namespace in a child, call\nfsopen(\"binfmt_misc\") there, send the fscontext fd to the parent and let\nthe parent issue FSCONFIG_CMD_CREATE. Both namespaces come from a plain\nunshare(1) and no capability is needed anywhere:\n\n  WARNING: fs/binfmt_misc.c:938 at bm_fill_super+0xa2/0xc0 [binfmt_misc]\n  CPU: 15 UID: 1000 PID: 3243382 Comm: fswarn\n  Call Trace:\n   get_tree_keyed+0x7d/0xb0\n   bm_get_tree+0x34/0x90 [binfmt_misc]\n   vfs_get_tree+0x2a/0x100\n   vfs_cmd_create+0x60/0xf0\n   __do_sys_fsconfig+0x4b2/0x500\n\nThe child needs the mount namespace because fsopen() itself gates on\nmay_mount(), which asks for CAP_SYS_ADMIN in the user namespace owning\nthe caller's mount namespace. fsconfig() doesn't repeat that check.\n\nIt is a WARN_ON() and not a WARN_ON_ONCE(), so the condition can be\nraised in a loop to taint the kernel and flood the log, and it panics a\nkernel booted with panic_on_warn.\n\nKeep refusing the mount and stop warning about it. Nothing in\nbm_fill_super() depends on the two namespaces matching, it derives\neverything from sb->s_user_ns.\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":[]}