Basically, MADV_FREE relys on the dirty bit in page table entry to decide whether VM allows to discard the page or not. IOW, if page table entry includes marked dirty bit, VM shouldn't discard the page.
However, if swap-in by read fault happens, page table entry point out the page doesn't have marked dirty bit so MADV_FREE might discard the page wrongly. To fix the problem, this patch marks page table entry of page swapping-in as dirty so VM shouldn't discard the page suddenly under us. With MADV_FREE point of view, marking dirty unconditionally is no problem because we dropped swapped page in MADV_FREE sycall context(ie, Look at madvise_free_pte_range) so every swapping-in pages are no MADV_FREE hinted pages. Cc: Hugh Dickins <[email protected]> Cc: Cyrill Gorcunov <[email protected]> Cc: Pavel Emelyanov <[email protected]> Reported-by: Yalin Wang <[email protected]> Signed-off-by: Minchan Kim <[email protected]> --- mm/memory.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/mm/memory.c b/mm/memory.c index 8a2fc9945b46..d1709f763152 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -2557,9 +2557,11 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, inc_mm_counter_fast(mm, MM_ANONPAGES); dec_mm_counter_fast(mm, MM_SWAPENTS); - pte = mk_pte(page, vma->vm_page_prot); + + /* Mark dirty bit of page table because MADV_FREE relies on it */ + pte = pte_mkdirty(mk_pte(page, vma->vm_page_prot)); if ((flags & FAULT_FLAG_WRITE) && reuse_swap_page(page)) { - pte = maybe_mkwrite(pte_mkdirty(pte), vma); + pte = maybe_mkwrite(pte, vma); flags &= ~FAULT_FLAG_WRITE; ret |= VM_FAULT_WRITE; exclusive = 1; -- 1.9.1 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
