Files
vuln-cve/data/2025/38xxx/CVE-2025-38306.json
T

159 lines
9.0 KiB
JSON

{
"dataType": "CVE_RECORD",
"dataVersion": "5.2",
"cveMetadata": {
"cveId": "CVE-2025-38306",
"assignerOrgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"state": "PUBLISHED",
"assignerShortName": "Linux",
"dateReserved": "2025-04-16T04:51:24.002Z",
"datePublished": "2025-07-10T07:42:16.806Z",
"dateUpdated": "2026-08-05T12:01:05.449Z"
},
"containers": {
"cna": {
"providerMetadata": {
"orgId": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
"shortName": "Linux",
"dateUpdated": "2026-08-05T12:01:05.449Z"
},
"descriptions": [
{
"lang": "en",
"value": "In the Linux kernel, the following vulnerability has been resolved:\n\nfs/fhandle.c: fix a race in call of has_locked_children()\n\nmay_decode_fh() is calling has_locked_children() while holding no locks.\nThat's an oopsable race...\n\nThe rest of the callers are safe since they are holding namespace_sem and\nare guaranteed a positive refcount on the mount in question.\n\nRename the current has_locked_children() to __has_locked_children(), make\nit static and switch the fs/namespace.c users to it.\n\nMake has_locked_children() a wrapper for __has_locked_children(), calling\nthe latter under read_seqlock_excl(&mount_lock)."
}
],
"metrics": [
{
"cvssV3_1": {
"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"
},
"scenarios": [
{
"lang": "en",
"value": "AV:L - The vulnerability is reached through local syscalls — `open_by_handle_at(2)` drives the unlocked `has_locked_children()` traversal while `mount(2)`/`umount(2)` mutate the `mnt_mounts` list. No network or physical access is involved.\nAC:L - The attacker controls both sides of the race entirely, spinning one thread on `open_by_handle_at(..., O_DIRECTORY)` and another on bind-mount/umount in their own mount namespace; the check runs before the file handle is even copied in or validated, so the window can be hammered without needing a valid handle or any condition outside attacker control.\nPR:L - An unprivileged local user reaches the vulnerable call by running `unshare -Urm`, since the gate only needs `CAP_SYS_ADMIN` over the mount namespace's owning user namespace — which is the attacker's own newly created user namespace. It is likewise reachable by any process that is root inside an ordinary container.\nUI:N - The attacker triggers everything from their own threads; no victim action, mount, or file access by another user is required.\nS:U - The corrupted state and the resulting impact are both within the kernel's own security authority, with no crossing of a VM, hypervisor, or IOMMU boundary.\nC:H - Traversing the concurrently modified list reads freed `struct mount` memory and feeds a stale `mnt_mountpoint` into `is_subdir()`/`d_ancestor()`, chasing attacker-groomable pointers; additionally, a raced result defeats the locked-mount gate, granting `HANDLE_CHECK_SUBTREE` and letting the attacker open directory handles into subtrees deliberately hidden by locked mounts.\nI:H - The use-after-free is on `mnt_cache` objects the attacker can spray at will via `mount()`, the standard basis for heap-corruption and control-flow primitives; the same bypass also yields a writable directory fd anchor beneath mounts that locking was meant to protect.\nA:H - `list_del_init()` on a mount being unmounted leaves a self-referential entry, trapping the unsynchronized traverser in an unkillable in-kernel infinite loop (soft/hard lockup), and the freed-mount dereference produces an oops — Al Viro describes it as \"an oopsable race.\""
}
]
}
],
"affected": [
{
"product": "Linux",
"vendor": "Linux",
"defaultStatus": "unaffected",
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"programFiles": [
"fs/namespace.c"
],
"versions": [
{
"version": "620c266f394932e5decc4b34683a75dfc59dc2f4",
"lessThan": "6482c3dccbfb8d20e2856ce67c75856859930b3f",
"status": "affected",
"versionType": "git"
},
{
"version": "620c266f394932e5decc4b34683a75dfc59dc2f4",
"lessThan": "287c7d34eedd37af1272dfb3b6e8656f4f026424",
"status": "affected",
"versionType": "git"
},
{
"version": "620c266f394932e5decc4b34683a75dfc59dc2f4",
"lessThan": "1f282cdc1d219c4a557f7009e81bc792820d9d9a",
"status": "affected",
"versionType": "git"
}
]
},
{
"product": "Linux",
"vendor": "Linux",
"defaultStatus": "affected",
"repo": "https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git",
"programFiles": [
"fs/namespace.c"
],
"versions": [
{
"version": "6.11",
"status": "affected"
},
{
"version": "0",
"lessThan": "6.11",
"status": "unaffected",
"versionType": "semver"
},
{
"version": "6.12.46",
"lessThanOrEqual": "6.12.*",
"status": "unaffected",
"versionType": "semver"
},
{
"version": "6.15.3",
"lessThanOrEqual": "6.15.*",
"status": "unaffected",
"versionType": "semver"
},
{
"version": "6.16",
"lessThanOrEqual": "*",
"status": "unaffected",
"versionType": "original_commit_for_fix"
}
]
}
],
"cpeApplicability": [
{
"nodes": [
{
"operator": "OR",
"negate": false,
"cpeMatch": [
{
"vulnerable": true,
"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*",
"versionStartIncluding": "6.11",
"versionEndExcluding": "6.12.46"
},
{
"vulnerable": true,
"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*",
"versionStartIncluding": "6.11",
"versionEndExcluding": "6.15.3"
},
{
"vulnerable": true,
"criteria": "cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*",
"versionStartIncluding": "6.11",
"versionEndExcluding": "6.16"
}
]
}
]
}
],
"references": [
{
"url": "https://git.kernel.org/stable/c/6482c3dccbfb8d20e2856ce67c75856859930b3f"
},
{
"url": "https://git.kernel.org/stable/c/287c7d34eedd37af1272dfb3b6e8656f4f026424"
},
{
"url": "https://git.kernel.org/stable/c/1f282cdc1d219c4a557f7009e81bc792820d9d9a"
}
],
"title": "fs/fhandle.c: fix a race in call of has_locked_children()",
"x_generator": {
"engine": "bippy-1.2.0"
}
}
}
}