[PATCH 4.9 05/69] thp: fix MADV_DONTNEED vs. MADV_FREE race

From: Greg Kroah-Hartman
Date: Wed Apr 19 2017 - 10:56:39 EST


4.9-stable review patch. If anyone has any objections, please let me know.

------------------

From: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>

commit 58ceeb6bec86d9140f9d91d71a710e963523d063 upstream.

Both MADV_DONTNEED and MADV_FREE handled with down_read(mmap_sem).

It's critical to not clear pmd intermittently while handling MADV_FREE
to avoid race with MADV_DONTNEED:

CPU0: CPU1:
madvise_free_huge_pmd()
pmdp_huge_get_and_clear_full()
madvise_dontneed()
zap_pmd_range()
pmd_trans_huge(*pmd) == 0 (without ptl)
// skip the pmd
set_pmd_at();
// pmd is re-established

It results in MADV_DONTNEED skipping the pmd, leaving it not cleared.
It violates MADV_DONTNEED interface and can result is userspace
misbehaviour.

Basically it's the same race as with numa balancing in
change_huge_pmd(), but a bit simpler to mitigate: we don't need to
preserve dirty/young flags here due to MADV_FREE functionality.

[kirill.shutemov@xxxxxxxxxxxxxxx: Urgh... Power is special again]
Link: http://lkml.kernel.org/r/20170303102636.bhd2zhtpds4mt62a@xxxxxxxxxxxxxxxxxx
Link: http://lkml.kernel.org/r/20170302151034.27829-4-kirill.shutemov@xxxxxxxxxxxxxxx
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
Acked-by: Minchan Kim <minchan@xxxxxxxxxx>
Cc: Minchan Kim <minchan@xxxxxxxxxx>
Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
Cc: Hillf Danton <hillf.zj@xxxxxxxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>

---
mm/huge_memory.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)

--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1380,8 +1380,7 @@ bool madvise_free_huge_pmd(struct mmu_ga
deactivate_page(page);

if (pmd_young(orig_pmd) || pmd_dirty(orig_pmd)) {
- orig_pmd = pmdp_huge_get_and_clear_full(tlb->mm, addr, pmd,
- tlb->fullmm);
+ pmdp_invalidate(vma, addr, pmd);
orig_pmd = pmd_mkold(orig_pmd);
orig_pmd = pmd_mkclean(orig_pmd);