mm: hugetlb: return immediately for hugetlb page in __delete_from_page_cache()
authorNaoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Mon, 10 Jul 2017 22:47:35 +0000 (15:47 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Mon, 10 Jul 2017 23:32:30 +0000 (16:32 -0700)
We avoid calling __mod_node_page_state(NR_FILE_PAGES) for hugetlb page
now, but it's not enough because later code doesn't handle hugetlb
properly.  Actually in our testing, WARN_ON_ONCE(PageDirty(page)) at the
end of this function fires for hugetlb, which makes no sense.  So we
should return immediately for hugetlb pages.

Link: http://lkml.kernel.org/r/1496305019-5493-3-git-send-email-n-horiguchi@ah.jp.nec.com
Signed-off-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm/filemap.c

index 3247b4208034723c82ad05f0ca4cc880cbdc91d9..a49702445ce05beeb8d80b46f0ee57c116986be2 100644 (file)
@@ -239,14 +239,16 @@ void __delete_from_page_cache(struct page *page, void *shadow)
        /* Leave page->index set: truncation lookup relies upon it */
 
        /* hugetlb pages do not participate in page cache accounting. */
-       if (!PageHuge(page))
-               __mod_node_page_state(page_pgdat(page), NR_FILE_PAGES, -nr);
+       if (PageHuge(page))
+               return;
+
+       __mod_node_page_state(page_pgdat(page), NR_FILE_PAGES, -nr);
        if (PageSwapBacked(page)) {
                __mod_node_page_state(page_pgdat(page), NR_SHMEM, -nr);
                if (PageTransHuge(page))
                        __dec_node_page_state(page, NR_SHMEM_THPS);
        } else {
-               VM_BUG_ON_PAGE(PageTransHuge(page) && !PageHuge(page), page);
+               VM_BUG_ON_PAGE(PageTransHuge(page), page);
        }
 
        /*