From: Huang Ying <ying.huang@xxxxxxxxx> Subject: mm, hugetlbfs: pass fault address to cow handler This is to take better advantage of the general huge page copying optimization. Where, the target subpage will be copied last to avoid the cache lines of target subpage to be evicted when copying other subpages. This works better if the address of the target subpage is available when copying huge page. So hugetlbfs page fault handlers are changed to pass that information to hugetlb_cow(). This will benefit workloads which don't access the begin of the hugetlbfs huge page after the page fault under heavy cache contention. Link: http://lkml.kernel.org/r/20180524005851.4079-5-ying.huang@xxxxxxxxx Signed-off-by: "Huang, Ying" <ying.huang@xxxxxxxxx> Reviewed-by: Mike Kravetz <mike.kravetz@xxxxxxxxxx> Cc: Michal Hocko <mhocko@xxxxxxxx> Cc: David Rientjes <rientjes@xxxxxxxxxx> Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx> Cc: "Kirill A. Shutemov" <kirill.shutemov@xxxxxxxxxxxxxxx> Cc: Andi Kleen <andi.kleen@xxxxxxxxx> Cc: Jan Kara <jack@xxxxxxx> Cc: Matthew Wilcox <willy@xxxxxxxxxxxxx> Cc: Hugh Dickins <hughd@xxxxxxxxxx> Cc: Minchan Kim <minchan@xxxxxxxxxx> Cc: Shaohua Li <shli@xxxxxx> Cc: Christopher Lameter <cl@xxxxxxxxx> Cc: "Aneesh Kumar K.V" <aneesh.kumar@xxxxxxxxxxxxxxxxxx> Cc: Punit Agrawal <punit.agrawal@xxxxxxx> Cc: Anshuman Khandual <khandual@xxxxxxxxxxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- mm/hugetlb.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) --- a/mm/hugetlb.c~mm-hugetlbfs-pass-fault-address-to-cow-handler +++ a/mm/hugetlb.c @@ -3509,7 +3509,7 @@ static void unmap_ref_private(struct mm_ * Keep the pte_same checks anyway to make transition from the mutex easier. */ static int hugetlb_cow(struct mm_struct *mm, struct vm_area_struct *vma, - unsigned long haddr, pte_t *ptep, + unsigned long address, pte_t *ptep, struct page *pagecache_page, spinlock_t *ptl) { pte_t pte; @@ -3518,6 +3518,7 @@ static int hugetlb_cow(struct mm_struct int ret = 0, outside_reserve = 0; unsigned long mmun_start; /* For mmu_notifiers */ unsigned long mmun_end; /* For mmu_notifiers */ + unsigned long haddr = address & huge_page_mask(h); pte = huge_ptep_get(ptep); old_page = pte_page(pte); @@ -3592,7 +3593,7 @@ retry_avoidcopy: goto out_release_all; } - copy_user_huge_page(new_page, old_page, haddr, vma, + copy_user_huge_page(new_page, old_page, address, vma, pages_per_huge_page(h)); __SetPageUptodate(new_page); set_page_huge_active(new_page); @@ -3826,7 +3827,7 @@ retry: hugetlb_count_add(pages_per_huge_page(h), mm); if ((flags & FAULT_FLAG_WRITE) && !(vma->vm_flags & VM_SHARED)) { /* Optimization, do the COW without a second fault */ - ret = hugetlb_cow(mm, vma, haddr, ptep, page, ptl); + ret = hugetlb_cow(mm, vma, address, ptep, page, ptl); } spin_unlock(ptl); @@ -3980,7 +3981,7 @@ int hugetlb_fault(struct mm_struct *mm, if (flags & FAULT_FLAG_WRITE) { if (!huge_pte_write(entry)) { - ret = hugetlb_cow(mm, vma, haddr, ptep, + ret = hugetlb_cow(mm, vma, address, ptep, pagecache_page, ptl); goto out_put_page; } _
