CVE-2024-52319

What it actually does

Untrusted fault address passed to clear_gigantic_page() can be unaligned to huge-page granularity, causing the function to write outside the intended gigantic-page region and produce memory corruption or information disclosure.

Why it matters to the business

Systems running affected kernels risk silent data corruption, crashes, or leakage of sensitive kernel or user data; unpatched hosts face elevated operational and compliance risk until upgraded.

How an attack unfolds

An attacker triggers a hugetlb page fault at a misaligned virtual address; the kernel passes that address through folio_zero_user() into clear_gigantic_page(), which performs writes based on the unaligned pointer, corrupting adjacent memory or exposing prior contents.

How to fix it

Upgrade the Linux kernel to a version containing commit b79b6fe0737f233f0be1465052b7f0e75f324735 or later; apply the corresponding stable or distribution update that aligns the address passed to clear_gigantic_page() and renames the parameter to addr_hint.

How to confirm the fix worked

Confirm the running kernel contains the fix commit or the patched clear_gigantic_page() implementation; re-test hugetlb fault paths with deliberately misaligned addresses to verify alignment enforcement.

Typical attack path

1. Map or fault a gigantic huge page via hugetlbfs
2. Supply a fault address not aligned to the huge-page size
3. Trigger the unaligned path in hugetlb_no_page()
4. Observe corruption, crash, or leaked data in adjacent memory

How to detect it

Examine kernel logs for oops or warnings referencing clear_gigantic_page or hugetlb_no_page; look for unexpected memory corruption symptoms or information leaks correlated with hugetlb page faults.

What your security team should do

• Prioritize kernel upgrades on systems using hugetlbfs or gigantic pages
• Monitor vendor advisories for the exact fixed kernel package versions
• Restrict hugetlb usage to trusted workloads until patches are deployed
• Review logs for unexpected hugetlb-related oops or corruption after upgrade