GHSA-P3PR-3QQ8-W67C
Vulnerability from github – Published: 2026-08-15 06:32 – Updated: 2026-08-19 09:31In the Linux kernel, the following vulnerability has been resolved:
ntfs: skip extent mft records in writeback to prevent deadlock
This patch fixes the ABBA deadlock between extent_lock and extent mrec_lock triggered by xfstests generic/113, that occurs since the commit 6994acf33bae ("ntfs: use base mft_no when looking up base inode for extent record").
Path A (inode writeback): VFS writeback -> ntfs_write_inode() -> __ntfs_write_inode() -> mutex_lock(&ni->extent_lock) -> mutex_lock(&tni->mrec_lock)
Path B (MFT folio writeback): VFS writeback of $MFT dirty folios -> ntfs_mft_writepages() -> ntfs_write_mft_block() -> ntfs_may_write_mft_record() -> holds one extent mrec_lock from a previous iteration -> tries to acquire another base inode extent_lock
By removing all extent_lock and extent mrec_lock acquisition from the MFT folio writeback path, the ABBA lock ordering is eliminated:
Path A: __ntfs_write_inode(): extent_lock -> mrec_lock Path B (removed): ntfs_write_mft_block(): mrec_lock -> extent_lock
Path B is always redundant for extent records because:
-
mark_mft_record_dirty(ext_ni) does NOT dirty the MFT folio. It only sets NInoDirty(ext_ni) and marks the base VFS inode dirty via __mark_inode_dirty(I_DIRTY_DATASYNC), which triggers Path A. Therefore, normal extent modifications never create a situation where the MFT folio is dirty and Path B is not scheduled.
-
The MFT folio only gets dirtied via ntfs_mft_mark_dirty() inside ntfs_mft_record_alloc(). But all identified callers in attrib.c (ntfs_attr_add, ntfs_attr_record_move_away, ntfs_attr_make_non_resident, ntfs_attr_record_resize) follow through with mark_mft_record_dirty(), which triggers Path A to write the complete record.
-
ntfs_evict_big_inode() calls ntfs_commit_inode() before freeing extent inodes, ensuring all dirty extents are flushed via Path A before the base inode leaves the icache.
{
"affected": [],
"aliases": [
"CVE-2026-72203"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-08-15T06:21:38Z",
"severity": "HIGH"
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\nntfs: skip extent mft records in writeback to prevent deadlock\n\nThis patch fixes the ABBA deadlock between extent_lock and extent\nmrec_lock triggered by xfstests generic/113, that occurs since the commit\n6994acf33bae (\"ntfs: use base mft_no when looking up base inode for\n\t\textent record\").\n\nPath A (inode writeback):\n VFS writeback\n -\u003e ntfs_write_inode()\n -\u003e __ntfs_write_inode()\n -\u003e mutex_lock(\u0026ni-\u003eextent_lock)\n -\u003e mutex_lock(\u0026tni-\u003emrec_lock)\n\nPath B (MFT folio writeback):\n VFS writeback of $MFT dirty folios\n -\u003e ntfs_mft_writepages()\n -\u003e ntfs_write_mft_block()\n -\u003e ntfs_may_write_mft_record()\n -\u003e holds one extent mrec_lock from a previous iteration\n -\u003e tries to acquire another base inode extent_lock\n\nBy removing all extent_lock and extent mrec_lock acquisition from the MFT\nfolio writeback path, the ABBA lock ordering is eliminated:\n\nPath A: __ntfs_write_inode(): extent_lock -\u003e mrec_lock\nPath B (removed): ntfs_write_mft_block(): mrec_lock -\u003e extent_lock\n\nPath B is always redundant for extent records because:\n\n1. mark_mft_record_dirty(ext_ni) does NOT dirty the MFT folio.\n It only sets NInoDirty(ext_ni) and marks the base VFS inode dirty\n via __mark_inode_dirty(I_DIRTY_DATASYNC), which triggers Path A.\n Therefore, normal extent modifications never create a situation where\n the MFT folio is dirty and Path B is not scheduled.\n\n2. The MFT folio only gets dirtied via ntfs_mft_mark_dirty() inside\n ntfs_mft_record_alloc(). But all identified callers in attrib.c\n (ntfs_attr_add, ntfs_attr_record_move_away,\n ntfs_attr_make_non_resident, ntfs_attr_record_resize) follow through\n with mark_mft_record_dirty(), which triggers Path A to write the\n complete record.\n\n3. ntfs_evict_big_inode() calls ntfs_commit_inode() before freeing extent\n inodes, ensuring all dirty extents are flushed via Path A before the\n base inode leaves the icache.",
"id": "GHSA-p3pr-3qq8-w67c",
"modified": "2026-08-19T09:31:22Z",
"published": "2026-08-15T06:32:15Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-72203"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/76bc14c7097ff678b2b5dbfd4fa33b46897d87ce"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/7ffa8f3d30236e0ab897c30bdb01224ff1fe1c89"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/f831ab09d521898bf1dd99bf5adfd630ea1428e3"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
"type": "CVSS_V3"
}
]
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.