GHSA-84RP-G9WQ-249M
Vulnerability from github – Published: 2026-09-04 18:31 – Updated: 2026-09-04 18:31In the Linux kernel, the following vulnerability has been resolved:
mm/hugetlb: fix swap entry corruption when clearing uffd-wp at fork()
copy_hugetlb_page_range() clears the uffd-wp bit of migration and hwpoison entries with huge_pte_clear_uffd_wp(), which operates on the present-PTE bit position. Swap entries keep the uffd-wp state elsewhere -- the migration branch reads and sets it with pte_swp_uffd_wp() and pte_swp_mkuffd_wp() -- and the present-PTE position falls into the swap payload. On x86-64 it lands in the inverted swap offset, where a naturally-aligned hugetlb PFN always has the affected bit set, so the clear advances the encoded PFN by two pages.
No userfaultfd needs to be involved: the clear is guarded only by the child VMA not being uffd-wp registered, so a plain fork() with an in-flight hugetlb migration entry (or a poisoned hugetlb page) corrupts the entry copied into the child. Instrumenting the clear and forking after MADV_HWPOISON on a 2MB anon hugetlb page shows:
offset before=120e00 offset after =120e02
The fallout is mostly latent: rmap walks match migration entries by folio range and remove_migration_pte() rebuilds the PTE from the folio, so a within-folio PFN skew heals once migration completes. But any path that re-encodes the corrupted offset -- e.g. hugetlb_change_protection() rewriting a writable migration entry via make_readable_migration_entry(swp_offset(entry)) -- propagates it.
Migration entries legitimately carry uffd-wp, so clear it with pte_swp_clear_uffd_wp(), matching copy_nonpresent_pte() and move_huge_pte().
A hwpoison entry, on the other hand, never carries the uffd-wp bit: it is installed fresh by make_hwpoison_entry() (try_to_unmap_one() does not preserve uffd-wp on the hwpoison path) and hugetlb_change_protection() leaves hwpoison entries untouched. There was nothing to clear there, only the corruption, so drop the clear entirely.
{
"affected": [],
"aliases": [
"CVE-2026-80893"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-09-04T18:17:57Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\nmm/hugetlb: fix swap entry corruption when clearing uffd-wp at fork()\n\ncopy_hugetlb_page_range() clears the uffd-wp bit of migration and hwpoison\nentries with huge_pte_clear_uffd_wp(), which operates on the present-PTE\nbit position. Swap entries keep the uffd-wp state elsewhere -- the\nmigration branch reads and sets it with pte_swp_uffd_wp() and\npte_swp_mkuffd_wp() -- and the present-PTE position falls into the swap\npayload. On x86-64 it lands in the inverted swap offset, where a\nnaturally-aligned hugetlb PFN always has the affected bit set, so the\nclear advances the encoded PFN by two pages.\n\nNo userfaultfd needs to be involved: the clear is guarded only by the\nchild VMA not being uffd-wp registered, so a plain fork() with an\nin-flight hugetlb migration entry (or a poisoned hugetlb page) corrupts\nthe entry copied into the child. Instrumenting the clear and forking\nafter MADV_HWPOISON on a 2MB anon hugetlb page shows:\n\n offset before=120e00\n offset after =120e02\n\nThe fallout is mostly latent: rmap walks match migration entries by folio\nrange and remove_migration_pte() rebuilds the PTE from the folio, so a\nwithin-folio PFN skew heals once migration completes. But any path that\nre-encodes the corrupted offset -- e.g. hugetlb_change_protection()\nrewriting a writable migration entry via\nmake_readable_migration_entry(swp_offset(entry)) -- propagates it.\n\nMigration entries legitimately carry uffd-wp, so clear it with\npte_swp_clear_uffd_wp(), matching copy_nonpresent_pte() and\nmove_huge_pte().\n\nA hwpoison entry, on the other hand, never carries the uffd-wp bit: it is\ninstalled fresh by make_hwpoison_entry() (try_to_unmap_one() does not\npreserve uffd-wp on the hwpoison path) and hugetlb_change_protection()\nleaves hwpoison entries untouched. There was nothing to clear there, only\nthe corruption, so drop the clear entirely.",
"id": "GHSA-84rp-g9wq-249m",
"modified": "2026-09-04T18:31:33Z",
"published": "2026-09-04T18:31:33Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-80893"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/2b9a07002c2f296aa6a9c591213933d3492e3089"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/2fa11c60c9c06bafc19cf4d9efdaa36a38079e87"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/69cb5825d9988c7944bc9f1dc08cb233655405a7"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/83abe2fd5b3aeb3123b5408a5a91709c5538fb23"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/8b0de7005b148738d79d6c45594d566489948a68"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/f1b1311c0352873137768bac5a126e491271a747"
}
],
"schema_version": "1.4.0",
"severity": []
}
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.