GHSA-9FQF-4MVH-6FX9
Vulnerability from github – Published: 2026-07-19 18:31 – Updated: 2026-07-20 15:31In the Linux kernel, the following vulnerability has been resolved:
mm/rmap: initialize nr_pages to 1 at loop start in try_to_unmap_one
Initialize nr_pages to 1 at the start of each loop iteration, like folio_referenced_one() does.
Without this, nr_pages computed by a previous folio_unmap_pte_batch() call can be reused on a later iteration that does not run folio_unmap_pte_batch() again.
mmap a 64K large folio with MAP_ANONYMOUS | MAP_DROPPABLE, then call madvise(MADV_FREE), then make the last page device-exclusive via HMM_DMIRROR_EXCLUSIVE.
Trigger node reclaim through sysfs. Now, in try_to_unmap_one(), we will first clear the first 15 out of 16 entries mapping the lazyfree folio. This will set nr_pages to 15. In the next pvmw walk, this nr_pages gets reused on a device-exclusive pte, thus potentially corrupting folio refcount/mapcount.
At the moment, I have a userspace program which can make the kernel spit out a trace, but the blow up is in folio_referenced_one(), because there are existing bugs in the interaction between device-private and rmap (which too I am investigating). I did a one liner kernel change to avoid going into folio_referenced_one(), and the kernel blows up at folio_remove_rmap_ptes in try_to_unmap_one which is what I wanted.
Note that the bug is there not since file folio batching but lazyfree folio batching, since device-exclusive only works for anonymous folios.
Userspace visible effect is simply kernel crashing somewhere due to refcount/mapcount corruption.
{
"affected": [],
"aliases": [
"CVE-2026-63950"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-07-19T16:17:13Z",
"severity": "HIGH"
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\nmm/rmap: initialize nr_pages to 1 at loop start in try_to_unmap_one\n\nInitialize nr_pages to 1 at the start of each loop iteration, like\nfolio_referenced_one() does.\n\nWithout this, nr_pages computed by a previous folio_unmap_pte_batch() call\ncan be reused on a later iteration that does not run\nfolio_unmap_pte_batch() again.\n\nmmap a 64K large folio with MAP_ANONYMOUS | MAP_DROPPABLE, then call\nmadvise(MADV_FREE), then make the last page device-exclusive via\nHMM_DMIRROR_EXCLUSIVE.\n\nTrigger node reclaim through sysfs. Now, in try_to_unmap_one(), we will\nfirst clear the first 15 out of 16 entries mapping the lazyfree folio. \nThis will set nr_pages to 15. In the next pvmw walk, this nr_pages gets\nreused on a device-exclusive pte, thus potentially corrupting folio\nrefcount/mapcount.\n\nAt the moment, I have a userspace program which can make the kernel spit\nout a trace, but the blow up is in folio_referenced_one(), because there\nare existing bugs in the interaction between device-private and rmap\n(which too I am investigating). I did a one liner kernel change to avoid\ngoing into folio_referenced_one(), and the kernel blows up at\nfolio_remove_rmap_ptes in try_to_unmap_one which is what I wanted.\n\nNote that the bug is there not since file folio batching but lazyfree\nfolio batching, since device-exclusive only works for anonymous folios.\n\nUserspace visible effect is simply kernel crashing somewhere due to\nrefcount/mapcount corruption.",
"id": "GHSA-9fqf-4mvh-6fx9",
"modified": "2026-07-20T15:31:53Z",
"published": "2026-07-19T18:31:47Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-63950"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/0fcc34d0d8fefca4fea349e45c10e3a3d90350eb"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/3f8968e9cbf95d5d87d32218906cab0b9b9eddbe"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/f611db9b771b2b6775357555d2517af044fca4f0"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/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.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.