Entreprise française de cybersécurité depuis 2004
☎ 03 60 47 09 81 - info@securiteinfo.com
☎ 03 60 47 09 81 - info@securiteinfo.com
In the Linux kernel, the following vulnerability has been resolved:mm/hugetlb: unshare page tables during VMA split, not beforeCurrently, __split_vma() triggers hugetlb page table unsharing throughvm_ops->may_split(). This happens before the VMA lock and rmap locks aretaken - which is too early, it allows racing VMA-locked page faults in ourprocess and racing rmap walks from other processes to cause page tables tobe shared again before we actually perform the split.Fix it by explicitly calling into the hugetlb unshare logic from__split_vma() in the same place where THP splitting also happens. At thatpoint, both the VMA and the rmap(s) are write-locked.An annoying detail is that we can now call into the helperhugetlb_unshare_pmds() from two different locking contexts:1. from hugetlb_split(), holding: - mmap lock (exclusively) - VMA lock - file rmap lock (exclusively)2. hugetlb_unshare_all_pmds(), which I think is designed to be able to call us with only the mmap lock held (in shared mode), but currently only runs while holding mmap lock (exclusively) and VMA lockBackporting note:This commit fixes a racy protection that was introduced in commitb30c14cd6102 ("hugetlb: unshare some PMDs when splitting VMAs"); thatcommit claimed to fix an issue introduced in 5.13, but it should actuallyalso go all the way back.[jannh@google.com: v2]
No PoCs from references.
- https://github.com/ARPSyndicate/cve-scores
- https://github.com/w4zu/Debian_security