mm/ksm.c: fix inconsistent accounting of zero pages
authorClaudio Imbrenda <imbrenda@linux.vnet.ibm.com>
Tue, 10 Apr 2018 23:29:41 +0000 (16:29 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Tue, 24 Apr 2018 07:36:22 +0000 (09:36 +0200)
commit a38c015f3156895b07e71d4e4414289f8a3b2745 upstream.

When using KSM with use_zero_pages, we replace anonymous pages
containing only zeroes with actual zero pages, which are not anonymous.
We need to do proper accounting of the mm counters, otherwise we will
get wrong values in /proc and a BUG message in dmesg when tearing down
the mm.

Link: http://lkml.kernel.org/r/1522931274-15552-1-git-send-email-imbrenda@linux.vnet.ibm.com
Fixes: e86c59b1b1 ("mm/ksm: improve deduplication of zero pages with colouring")
Signed-off-by: Claudio Imbrenda <imbrenda@linux.vnet.ibm.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
mm/ksm.c

index 6cb60f46cce55761b0ff9d3523be69a706523972..5b6be9eeb095adfef11b1e89a0192dbf69df0a42 100644 (file)
--- a/mm/ksm.c
+++ b/mm/ksm.c
@@ -1133,6 +1133,13 @@ static int replace_page(struct vm_area_struct *vma, struct page *page,
        } else {
                newpte = pte_mkspecial(pfn_pte(page_to_pfn(kpage),
                                               vma->vm_page_prot));
+               /*
+                * We're replacing an anonymous page with a zero page, which is
+                * not anonymous. We need to do proper accounting otherwise we
+                * will get wrong values in /proc, and a BUG message in dmesg
+                * when tearing down the mm.
+                */
+               dec_mm_counter(mm, MM_ANONPAGES);
        }
 
        flush_cache_page(vma, addr, pte_pfn(*ptep));