CVE-2025-38085: mm/hugetlb: fix huge_pmd_unshare() vs GUP-fast race
In the Linux kernel, the following vulnerability has been resolved:
mm/hugetlb: fix huge_pmd_unshare() vs GUP-fast race
huge_pmd_unshare() drops a reference on a page table that may have
previously been shared across processes, potentially turning it into a
normal page table used in another process in which unrelated VMAs can
afterwards be installed.
If this happens in the middle of a concurrent gup_fast(), gup_fast() could
end up walking the page tables of another process. While I don't see any
way in which that immediately leads to kernel memory corruption, it is
really weird and unexpected.
Fix it with an explicit broadcast IPI through tlb_remove_table_sync_one(),
just like we do in khugepaged when removing page tables for a THP
collapse.
Security readout for executives and security teams
Plain-English summary
CVE-2025-38085 is a Linux kernel memory-management race involving huge pages. In unusual timing, one process could cause kernel page-table walking code to follow another process's page tables. The source text does not claim immediate kernel memory corruption, active exploitation, or a CVSS severity.
Executive priority
Treat as a kernel maintenance priority, not an emergency from the supplied evidence. Schedule patch validation and rollout through normal kernel update channels, with higher priority for shared compute and multi-tenant infrastructure.
Technical view
The bug is in mm/hugetlb: huge_pmd_unshare() can drop a reference to a shared page table while concurrent GUP-fast is walking it. The fix adds explicit synchronization through tlb_remove_table_sync_one(), matching existing THP-collapse handling.
Likely exposure
Exposure is limited to Linux systems running affected kernel builds listed in the CVE record or downstream distributions carrying the vulnerable code. Systems using HugeTLB-related paths are the most relevant validation targets, but the bundle does not quantify prerequisites or reachability.
Exploitation context
The bundle marks KEV as false. It includes upstream stable commits, a Project Zero issue, and Debian LTS advisories, but no cited source states active exploitation, public exploit availability, or confirmed privilege escalation impact.
Researcher notes
Key uncertainty is impact. The upstream description says the behavior is unexpected but does not identify immediate kernel memory corruption. Focus research on race reachability, HugeTLB usage, GUP-fast interaction, and downstream backport status.
Mitigation direction
Apply vendor or distribution kernel updates that include the upstream stable fixes.
Review Debian LTS advisories if running affected Debian LTS kernels.
Check Linux vendor guidance before assuming a specific fixed package version.
Prioritize internet-facing, multi-user, container, and virtualization hosts for kernel update review.
Validation and detection
Inventory deployed Linux kernel versions across servers, containers hosts, and appliances.
Compare kernel builds against the CVE record and vendor security advisories.
Confirm whether the relevant upstream stable fix is backported by your distribution.
Track reboot completion where kernel package updates require it.
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Potential ATT&CK relevance
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
cve · low confidence lookup
CVE-2025-38085 mapping review
Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.