The page migration code employs try_to_unmap() to try and unmap the source page. This is accomplished by using rmap_walk to find all vmas where the page is mapped. This search stops when page mapcount is zero. For shared PMD huge pages, the page map count is always 1 no matter the number of mappings. Shared mappings are tracked via the reference count of the PMD page. Therefore, try_to_unmap stops prematurely and does not completely unmap all mappings of the source page.

This problem can result is data corruption as writes to the original source page can happen after contents of the page are copied to the target page. Hence, data is lost.

This problem was originally seen as DB corruption of shared global areas after a huge page was soft offlined due to ECC memory errors. DB developers noticed they could reproduce the issue by (hotplug) offlining memory used to back huge pages. A simple testcase can reproduce the problem by creating a shared PMD mapping (note that this must be at least PUD_SIZE in size and PUD_SIZE aligned (1GB on x86)), and using migrate_pages() to migrate process pages between nodes while continually writing to the huge pages being migrated.

To fix, have the try_to_unmap_one routine check for huge PMD sharing by calling huge_pmd_unshare for hugetlbfs huge pages. If it is a shared mapping it will be 'unshared' which removes the page table entry and drops the reference on the PMD page. After this, flush caches and TLB.

Fixes: 39dde65c9940 ("shared page table for hugetlb page") Signed-off-by: Mike Kravetz mike.kravetz@oracle.com --- v2: Fixed build issue for !CONFIG_HUGETLB_PAGE and typos in comment

include/linux/hugetlb.h | 6 ++++++ mm/rmap.c | 21 +++++++++++++++++++++ 2 files changed, 27 insertions(+)

diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 36fa6a2a82e3..7524663028ec 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -170,6 +170,12 @@ static inline unsigned long hugetlb_total_pages(void) return 0; }

+static inline int huge_pmd_unshare(struct mm_struct *mm, unsigned long *addr, + pte_t *ptep) +{ + return 0; +} + #define follow_hugetlb_page(m,v,p,vs,a,b,i,w,n) ({ BUG(); 0; }) #define follow_huge_addr(mm, addr, write) ERR_PTR(-EINVAL) #define copy_hugetlb_page_range(src, dst, vma) ({ BUG(); 0; }) diff --git a/mm/rmap.c b/mm/rmap.c index 09a799c9aebd..cf2340adad10 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -1409,6 +1409,27 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma, subpage = page - page_to_pfn(page) + pte_pfn(*pvmw.pte); address = pvmw.address;

+ /* + * PMDs for hugetlbfs pages could be shared. In this case, + * pages with shared PMDs will have a mapcount of 1 no matter + * how many times they are actually mapped. Map counting for + * PMD sharing is mostly done via the reference count on the + * PMD page itself. If the page we are trying to unmap is a + * hugetlbfs page, attempt to 'unshare' at the PMD level. + * huge_pmd_unshare clears the PUD and adjusts reference + * counting on the PMD page which effectively unmaps the page. + * Take care of flushing cache and TLB for page in this + * specific mapping here. + */ + if (PageHuge(page) && + huge_pmd_unshare(mm, &address, pvmw.pte)) { + unsigned long end_add = address + vma_mmu_pagesize(vma); + + flush_cache_range(vma, address, end_add); + flush_tlb_range(vma, address, end_add); + mmu_notifier_invalidate_range(mm, address, end_add); + continue; + }

if (IS_ENABLED(CONFIG_MIGRATION) && (flags & TTU_MIGRATION) &&