{"id":"CVE-2026-99174","aliases":[],"title":"ext4: free the sbi after a grace period","titleSource":"Linux","descriptionSource":"Linux","products":[{"vendor":"Linux","name":"Linux","platforms":[],"versions":[{"version":"6493792d3299b3e33f887ef8a150099d271faf9c","status":"affected","lessThan":"80d5db1686b05fad6f384be017714297d828a714","versionType":"git"},{"version":"6493792d3299b3e33f887ef8a150099d271faf9c","status":"affected","lessThan":"87d29b8b849738d52fe820d495fa0aa2d5a1ec82","versionType":"git"},{"version":"6493792d3299b3e33f887ef8a150099d271faf9c","status":"affected","lessThan":"c17b44266a1f9fcb49b894d4611d1254bf1aa313","versionType":"git"}],"defaultStatus":"unaffected","source":"Linux","url":"https://www.cve.org/CVERecord?id=CVE-2026-99174"},{"vendor":"Linux","name":"Linux","platforms":[],"versions":[{"version":"5.19","status":"affected"},{"version":"0","status":"unaffected","lessThan":"5.19","versionType":"semver"},{"version":"6.18.59","status":"unaffected","lessThanOrEqual":"6.18.*","versionType":"semver"},{"version":"7.2.13","status":"unaffected","lessThanOrEqual":"7.2.*","versionType":"semver"},{"version":"7.3","status":"unaffected","lessThanOrEqual":"*","versionType":"original_commit_for_fix"}],"defaultStatus":"affected","source":"Linux","url":"https://www.cve.org/CVERecord?id=CVE-2026-99174"}],"remediation":[],"description":"In the Linux kernel, the following vulnerability has been resolved:\n\next4: free the sbi after a grace period\n\next4_get_link() follows a symlink that keeps its target in a data block\nin rcu pathwalk mode with ext4_getblk(EXT4_GET_BLOCKS_CACHED_NOWAIT).\nThat reads s_mount_state in the ASSERT() of ext4_getblk() and again in\next4_es_lookup_extent(), and the lookup bumps the extent status hit or\nmiss counter, all through sb->s_fs_info.\n\next4_put_super() destroys those counters, clears sb->s_fs_info and frees\nthe sbi without waiting for a grace period. A pathwalk that is still in\next4_get_link() when the last reference to the mount goes away then\ndereferences a NULL s_fs_info, the freed sbi or the destroyed percpu\ncounters.\n\nWith a pathwalk held up in ext4_get_link() while the last reference\ngoes away:\n\n  KASAN: null-ptr-deref in range [0x00000000000000a8-0x00000000000000af]\n  RIP: 0010:ext4_getblk+0xf0/0x640\n  Call Trace:\n   <TASK>\n   ext4_get_link+0x26c/0x410\n   pick_link+0xf96/0x1780\n   step_into_slowpath+0x9b8/0x1250\n   path_lookupat+0x484/0xb00\n   filename_lookup+0x1f1/0x5b0\n   vfs_statx+0xd7/0x3b0\n   vfs_fstatat+0x4d/0x90\n   __do_sys_newfstatat+0xa0/0x100\n   do_syscall_64+0x10e/0x5e0\n   entry_SYSCALL_64_after_hwframe+0x77/0x7f\n\nSo far that has been hard to hit. umount(2) waits for a grace period in\nnamespace_unlock() before it drops the mount, so only a mount whose last\nreference goes away elsewhere, like the close of a file or a chdir out of\na lazily unmounted directory, can race a walker.\n\nKeep sb->s_fs_info and free the sbi together with the extent status\ncounters after a grace period. Split the counters out of\next4_es_unregister_shrinker() for that. The shrinker still gets wasted\nin ext4_put_super(), the counters go with the sbi.","severity":"unknown","score":null,"published":"2026-10-28T10:35:28.758Z","modified":"2026-10-28T10:35:28.758Z","withdrawn":false,"knownExploited":false,"affected":[],"references":["https://git.kernel.org/stable/c/80d5db1686b05fad6f384be017714297d828a714","https://git.kernel.org/stable/c/87d29b8b849738d52fe820d495fa0aa2d5a1ec82","https://git.kernel.org/stable/c/c17b44266a1f9fcb49b894d4611d1254bf1aa313"],"sources":[{"name":"CVE","id":"CVE-2026-99174","modified":"2026-10-28T10:35:28.758Z","url":"https://www.cve.org/CVERecord?id=CVE-2026-99174"}],"assessments":[],"sourceAffected":[],"kev":{"status":"not-listed","checkedAt":"2026-10-29T02:20:33.012Z"},"coverage":[{"source":"NVD","status":"not-checked"},{"source":"GHSA","status":"not-checked"},{"source":"EUVD","status":"not-checked"},{"source":"EPSS","status":"not-checked"}],"evidence":[]}