mm: thp: fix MADV_DONTNEED vs migrate_misplaced_transhuge_page race condition

Patch series "migrate_misplaced_transhuge_page race conditions".

Aaron found a new instance of the THP MADV_DONTNEED race against
pmdp_clear_flush* variants, that was apparently left unfixed.

While looking into the race found by Aaron, I may have found two more
issues in migrate_misplaced_transhuge_page.

These race conditions would not cause kernel instability, but they'd
corrupt userland data or leave data non zero after MADV_DONTNEED.

I did only minor testing, and I don't expect to be able to reproduce this
(especially the lack of ->invalidate_range before migrate_page_copy,
requires the latest iommu hardware or infiniband to reproduce).  The last
patch is noop for x86 and it needs further review from maintainers of
archs that implement flush_cache_range() (not in CC yet).

To avoid confusion, it's not the first patch that introduces the bug fixed
in the second patch, even before removing the
pmdp_huge_clear_flush_notify, that _notify suffix was called after
migrate_page_copy already run.

This patch (of 3):

This is a corollary of ced10803 ("thp: fix MADV_DONTNEED vs.  numa
balancing race"), 58ceeb6b ("thp: fix MADV_DONTNEED vs.  MADV_FREE
race") and 5b7abeae ("thp: fix MADV_DONTNEED vs clear soft dirty
race).

When the above three fixes where posted Dave asked
https://lkml.kernel.org/r/929b3844-aec2-0111-fef7-8002f9d4e2b9@intel.com
but apparently this was missed.

The pmdp_clear_flush* in migrate_misplaced_transhuge_page() was introduced
in a54a407f ("mm: Close races between THP migration and PMD numa
clearing").

The important part of such commit is only the part where the page lock is
not released until the first do_huge_pmd_numa_page() finished disarming
the pagenuma/protnone.

The addition of pmdp_clear_flush() wasn't beneficial to such commit and
there's no commentary about such an addition either.

I guess the pmdp_clear_flush() in such commit was added just in case for
safety, but it ended up introducing the MADV_DONTNEED race condition found
by Aaron.

At that point in time nobody thought of such kind of MADV_DONTNEED race
conditions yet (they were fixed later) so the code may have looked more
robust by adding the pmdp_clear_flush().

This specific race condition won't destabilize the kernel, but it can
confuse userland because after MADV_DONTNEED the memory won't be zeroed
out.

This also optimizes the code and removes a superfluous TLB flush.

[akpm@linux-foundation.org: reflow comment to 80 cols, fix grammar and typo (beacuse)]
Link: http://lkml.kernel.org/r/20181013002430.698-2-aarcange@redhat.comSigned-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reported-by: Aaron Tomlin <atomlin@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

mm/migrate.c

@@ -2031,15 +2031,26 @@ int migrate_misplaced_transhuge_page(struct mm_struct *mm,
 	entry = maybe_pmd_mkwrite(pmd_mkdirty(entry), vma);
 
 	/*
-	 * Clear the old entry under pagetable lock and establish the new PTE.
-	 * Any parallel GUP will either observe the old page blocking on the
-	 * page lock, block on the page table lock or observe the new page.
-	 * The SetPageUptodate on the new page and page_add_new_anon_rmap
-	 * guarantee the copy is visible before the pagetable update.
+	 * Overwrite the old entry under pagetable lock and establish
+	 * the new PTE. Any parallel GUP will either observe the old
+	 * page blocking on the page lock, block on the page table
+	 * lock or observe the new page. The SetPageUptodate on the
+	 * new page and page_add_new_anon_rmap guarantee the copy is
+	 * visible before the pagetable update.
 	 */
 	flush_cache_range(vma, mmun_start, mmun_end);
 	page_add_anon_rmap(new_page, vma, mmun_start, true);
-	pmdp_huge_clear_flush_notify(vma, mmun_start, pmd);
+	/*
+	 * At this point the pmd is numa/protnone (i.e. non present) and the TLB
+	 * has already been flushed globally.  So no TLB can be currently
+	 * caching this non present pmd mapping.  There's no need to clear the
+	 * pmd before doing set_pmd_at(), nor to flush the TLB after
+	 * set_pmd_at().  Clearing the pmd here would introduce a race
+	 * condition against MADV_DONTNEED, because MADV_DONTNEED only holds the
+	 * mmap_sem for reading.  If the pmd is set to NULL at any given time,
+	 * MADV_DONTNEED won't wait on the pmd lock and it'll skip clearing this
+	 * pmd.
+	 */
 	set_pmd_at(mm, mmun_start, pmd, entry);
 	update_mmu_cache_pmd(vma, address, &entry);
 
@@ -2053,7 +2064,7 @@ int migrate_misplaced_transhuge_page(struct mm_struct *mm,
 	 * No need to double call mmu_notifier->invalidate_range() callback as
 	 * the above pmdp_huge_clear_flush_notify() did already call it.
 	 */
-	mmu_notifier_invalidate_range_only_end(mm, mmun_start, mmun_end);
+	mmu_notifier_invalidate_range_end(mm, mmun_start, mmun_end);
 
 	/* Take an "isolate" reference and put new page on the LRU. */
 	get_page(new_page);