{"api_version":"1","generated_at":"2026-10-27T04:27:50+00:00","cve":"CVE-2026-98982","urls":{"html":"https://cve.report/CVE-2026-98982","api":"https://cve.report/api/cve/CVE-2026-98982.json","docs":"https://cve.report/api","cve_org":"https://www.cve.org/CVERecord?id=CVE-2026-98982","nvd":"https://nvd.nist.gov/vuln/detail/CVE-2026-98982"},"summary":{"title":"namespace: don't let a pseudo dentry become the root of a mount","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nnamespace: don't let a pseudo dentry become the root of a mount\n\nd_alloc_pseudo() hands out dentries for pipes, sockets and other files\nthat are never anyone's child or parent. They carry DCACHE_NORCU and\ndentry_free() frees them right away because no lockless path walk can\never reach them. That holds as long as such a dentry isn't the root of\na mount. But bind mounting /proc/self/fd/<fd> of such a file does\nexactly that.\n\nMost callers of alloc_file_pseudo() put their files on kernel-internal\nmounts and may_copy_tree() refuses those. bpf_token_create() doesn't. It\nplaces the token file on the bpffs mount the caller handed it and that\nmount is in the caller's mount namespace so the clone goes through:\n\n  mount --bind /proc/self/fd/<token> <file>\n  open_tree(tokfd, \"\", AT_EMPTY_PATH | OPEN_TREE_CLONE) + move_mount()\n\n__follow_mount_rcu() then loads ->mnt_root of that mount, reads d_seq\nand d_flags of the dentry and only then checks mount_lock. The dentry\ncan be gone by then. The final mntput() of an unmounted parent unhooks\na child that stayed attached to it under mount_lock alone. When the\nchild's holder does the final put right after that the root is dput()\nand freed immediately while a walker that found the mount hashed still\nlooks at it.\n\nRefuse to clone a mount with a DCACHE_NORCU dentry as its root. Nothing\nsensible can be done with a bind mount of a bpf token anyway.","state":"PUBLISHED","assigner":"Linux","published_at":"2026-10-23 13:17:20","updated_at":"2026-10-24 14:16:50"},"problem_types":[],"metrics":[{"version":"3.1","source":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","type":"Secondary","score":"7.8","severity":"HIGH","vector":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","data":{"version":"3.1","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","baseScore":7.8,"baseSeverity":"HIGH","attackVector":"LOCAL","attackComplexity":"LOW","privilegesRequired":"LOW","userInteraction":"NONE","scope":"UNCHANGED","confidentialityImpact":"HIGH","integrityImpact":"HIGH","availabilityImpact":"HIGH"}},{"version":"3.1","source":"CNA","type":"DECLARED","score":"7.8","severity":"HIGH","vector":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","data":{"baseScore":7.8,"baseSeverity":"HIGH","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","version":"3.1"}}],"references":[{"url":"https://git.kernel.org/stable/c/7e1c466011ba9e64d43075396badc03191bcee44","name":"https://git.kernel.org/stable/c/7e1c466011ba9e64d43075396badc03191bcee44","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/a256e31e0c54c4c37d2523aa52faabdcdd2ebc07","name":"https://git.kernel.org/stable/c/a256e31e0c54c4c37d2523aa52faabdcdd2ebc07","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/b138e16fb1e81376cea89f727dda7bfccd140de4","name":"https://git.kernel.org/stable/c/b138e16fb1e81376cea89f727dda7bfccd140de4","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/00287afa7cd989efdbf5822b7efac33cce7d489a","name":"https://git.kernel.org/stable/c/00287afa7cd989efdbf5822b7efac33cce7d489a","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://www.cve.org/CVERecord?id=CVE-2026-98982","name":"CVE Program record","refsource":"CVE.ORG","tags":["canonical"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-98982","name":"NVD vulnerability detail","refsource":"NVD","tags":["canonical","analysis"]}],"affected":[{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 35f96de04127d332a5c5e8a155d31f452f88c76d a256e31e0c54c4c37d2523aa52faabdcdd2ebc07 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 35f96de04127d332a5c5e8a155d31f452f88c76d 7e1c466011ba9e64d43075396badc03191bcee44 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 35f96de04127d332a5c5e8a155d31f452f88c76d 00287afa7cd989efdbf5822b7efac33cce7d489a git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 35f96de04127d332a5c5e8a155d31f452f88c76d b138e16fb1e81376cea89f727dda7bfccd140de4 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 6.9","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.9 semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.12.113 6.12.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.18.57 6.18.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.2.11 7.2.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.3 * original_commit_for_fix","platforms":[]}],"timeline":[],"solutions":[],"workarounds":[],"exploits":[],"credits":[],"nvd_cpes":[],"vendor_comments":[],"enrichments":{"kev":null,"epss":null,"legacy_qids":[]},"source_records":{"cve_program":{"containers":{"cna":{"affected":[{"defaultStatus":"unaffected","product":"Linux","programFiles":["fs/namespace.c"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"lessThan":"a256e31e0c54c4c37d2523aa52faabdcdd2ebc07","status":"affected","version":"35f96de04127d332a5c5e8a155d31f452f88c76d","versionType":"git"},{"lessThan":"7e1c466011ba9e64d43075396badc03191bcee44","status":"affected","version":"35f96de04127d332a5c5e8a155d31f452f88c76d","versionType":"git"},{"lessThan":"00287afa7cd989efdbf5822b7efac33cce7d489a","status":"affected","version":"35f96de04127d332a5c5e8a155d31f452f88c76d","versionType":"git"},{"lessThan":"b138e16fb1e81376cea89f727dda7bfccd140de4","status":"affected","version":"35f96de04127d332a5c5e8a155d31f452f88c76d","versionType":"git"}]},{"defaultStatus":"affected","product":"Linux","programFiles":["fs/namespace.c"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"status":"affected","version":"6.9"},{"lessThan":"6.9","status":"unaffected","version":"0","versionType":"semver"},{"lessThanOrEqual":"6.12.*","status":"unaffected","version":"6.12.113","versionType":"semver"},{"lessThanOrEqual":"6.18.*","status":"unaffected","version":"6.18.57","versionType":"semver"},{"lessThanOrEqual":"7.2.*","status":"unaffected","version":"7.2.11","versionType":"semver"},{"lessThanOrEqual":"*","status":"unaffected","version":"7.3","versionType":"original_commit_for_fix"}]}],"cpeApplicability":[{"nodes":[{"cpeMatch":[{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"6.12.113","versionStartIncluding":"6.9","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"6.18.57","versionStartIncluding":"6.9","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.2.11","versionStartIncluding":"6.9","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.3","versionStartIncluding":"6.9","vulnerable":true}],"negate":false,"operator":"OR"}]}],"descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\nnamespace: don't let a pseudo dentry become the root of a mount\n\nd_alloc_pseudo() hands out dentries for pipes, sockets and other files\nthat are never anyone's child or parent. They carry DCACHE_NORCU and\ndentry_free() frees them right away because no lockless path walk can\never reach them. That holds as long as such a dentry isn't the root of\na mount. But bind mounting /proc/self/fd/<fd> of such a file does\nexactly that.\n\nMost callers of alloc_file_pseudo() put their files on kernel-internal\nmounts and may_copy_tree() refuses those. bpf_token_create() doesn't. It\nplaces the token file on the bpffs mount the caller handed it and that\nmount is in the caller's mount namespace so the clone goes through:\n\n  mount --bind /proc/self/fd/<token> <file>\n  open_tree(tokfd, \"\", AT_EMPTY_PATH | OPEN_TREE_CLONE) + move_mount()\n\n__follow_mount_rcu() then loads ->mnt_root of that mount, reads d_seq\nand d_flags of the dentry and only then checks mount_lock. The dentry\ncan be gone by then. The final mntput() of an unmounted parent unhooks\na child that stayed attached to it under mount_lock alone. When the\nchild's holder does the final put right after that the root is dput()\nand freed immediately while a walker that found the mount hashed still\nlooks at it.\n\nRefuse to clone a mount with a DCACHE_NORCU dentry as its root. Nothing\nsensible can be done with a bind mount of a bpf token anyway."}],"metrics":[{"cvssV3_1":{"baseScore":7.8,"baseSeverity":"HIGH","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","version":"3.1"},"scenarios":[{"lang":"en","value":"AV:L - Local bpf(BPF_TOKEN_CREATE) reaches bpf_token_create()/alloc_file_pseudo(), after which mount(MS_BIND) or open_tree(OPEN_TREE_CLONE) reaches the patched __do_loopback() on /proc/self/fd/<token>; no network protocol carries the triggering data.\nAC:L - The attacker controls creation and teardown of the token-root mount and can race cleanup_mnt()/dput() against openat() reaching __follow_mount_rcu() using concurrent threads; neither side of the race requires external timing or victim-controlled state.\nPR:L - bpf_token_create() requires CAP_BPF in the bpffs user namespace, while open_tree(OPEN_TREE_CLONE) and move_mount() require CAP_SYS_ADMIN over the mount namespace; because bpf_fs_type has FS_USERNS_MOUNT, an ordinary user can obtain these capabilities through unprivileged user and mount namespaces.\nUI:N - The attacking process performs every required action itself: mounting delegated bpffs, creating the token, cloning and tearing down mounts, and racing openat(); no victim must mount, open, or otherwise process attacker-supplied data.\nS:U - The freed dentry is consumed by __follow_mount_rcu() inside the same host kernel, yielding ordinary local kernel compromise or denial of service rather than crossing a VM, IOMMU, or separate security-authority boundary.\nC:H - cleanup_mnt() can dput() and immediately free the DCACHE_NORCU mount root while __follow_mount_rcu() still reads its d_seq, d_flags, and potentially d_op; attacker-driven dentry-cache reuse makes this kernel UAF capable of high-impact memory disclosure.\nI:H - The freed struct dentry can be reclaimed through attacker-driven dentry allocation, letting heap spraying influence fields and the d_op->d_manage dispatch consumed by __follow_mount_rcu(); this is an exploitable kernel UAF path to arbitrary modification or control-flow hijacking.\nA:H - __follow_mount_rcu() dereferences the mount root after dentry_free() has returned it immediately to dentry_cache, so poisoned or reclaimed contents can cause a kernel oops or panic, and attacker-controlled threads can repeat the race."}]}],"providerMetadata":{"dateUpdated":"2026-10-24T14:09:50.745Z","orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","shortName":"Linux"},"references":[{"url":"https://git.kernel.org/stable/c/a256e31e0c54c4c37d2523aa52faabdcdd2ebc07"},{"url":"https://git.kernel.org/stable/c/7e1c466011ba9e64d43075396badc03191bcee44"},{"url":"https://git.kernel.org/stable/c/00287afa7cd989efdbf5822b7efac33cce7d489a"},{"url":"https://git.kernel.org/stable/c/b138e16fb1e81376cea89f727dda7bfccd140de4"}],"title":"namespace: don't let a pseudo dentry become the root of a mount","x_generator":{"engine":"bippy-1.2.0"}}},"cveMetadata":{"assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","assignerShortName":"Linux","cveId":"CVE-2026-98982","datePublished":"2026-10-23T12:19:59.784Z","dateReserved":"2026-09-25T10:25:14.384Z","dateUpdated":"2026-10-24T14:09:50.745Z","state":"PUBLISHED"},"dataType":"CVE_RECORD","dataVersion":"5.2"},"nvd":{"publishedDate":"2026-10-23 13:17:20","lastModifiedDate":"2026-10-24 14:16:50","problem_types":[],"metrics":{"cvssMetricV31":[{"source":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","type":"Secondary","cvssData":{"version":"3.1","vectorString":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","baseScore":7.8,"baseSeverity":"HIGH","attackVector":"LOCAL","attackComplexity":"LOW","privilegesRequired":"LOW","userInteraction":"NONE","scope":"UNCHANGED","confidentialityImpact":"HIGH","integrityImpact":"HIGH","availabilityImpact":"HIGH"},"exploitabilityScore":1.8,"impactScore":5.9}]},"configurations":[]},"legacy_mitre":{"record":{"CveYear":"2026","CveId":"98982","Ordinal":"1","Title":"namespace: don't let a pseudo dentry become the root of a mount","CVE":"CVE-2026-98982","Year":"2026"},"notes":[{"CveYear":"2026","CveId":"98982","Ordinal":"1","NoteData":"In the Linux kernel, the following vulnerability has been resolved:\n\nnamespace: don't let a pseudo dentry become the root of a mount\n\nd_alloc_pseudo() hands out dentries for pipes, sockets and other files\nthat are never anyone's child or parent. They carry DCACHE_NORCU and\ndentry_free() frees them right away because no lockless path walk can\never reach them. That holds as long as such a dentry isn't the root of\na mount. But bind mounting /proc/self/fd/<fd> of such a file does\nexactly that.\n\nMost callers of alloc_file_pseudo() put their files on kernel-internal\nmounts and may_copy_tree() refuses those. bpf_token_create() doesn't. It\nplaces the token file on the bpffs mount the caller handed it and that\nmount is in the caller's mount namespace so the clone goes through:\n\n  mount --bind /proc/self/fd/<token> <file>\n  open_tree(tokfd, \"\", AT_EMPTY_PATH | OPEN_TREE_CLONE) + move_mount()\n\n__follow_mount_rcu() then loads ->mnt_root of that mount, reads d_seq\nand d_flags of the dentry and only then checks mount_lock. The dentry\ncan be gone by then. The final mntput() of an unmounted parent unhooks\na child that stayed attached to it under mount_lock alone. When the\nchild's holder does the final put right after that the root is dput()\nand freed immediately while a walker that found the mount hashed still\nlooks at it.\n\nRefuse to clone a mount with a DCACHE_NORCU dentry as its root. Nothing\nsensible can be done with a bind mount of a bpf token anyway.","Type":"Description","Title":"namespace: don't let a pseudo dentry become the root of a mount"}]}}}