powerpc: remove superfluous PageTail checks on the pte gup_fast
authorAndrea Arcangeli <aarcange@redhat.com>
Wed, 2 Nov 2011 20:37:03 +0000 (13:37 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Wed, 2 Nov 2011 23:06:57 +0000 (16:06 -0700)
This part of gup_fast doesn't seem capable of handling hugetlbfs ptes,
those should be handled by gup_hugepd only, so these checks are
superfluous.

Plus if this wasn't a noop, it would have oopsed because, the insistence
of using the speculative refcounting would trigger a VM_BUG_ON if a tail
page was encountered in the page_cache_get_speculative().

Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Acked-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
arch/powerpc/mm/gup.c

index b9e1c7ff5f6d33d8a111c10ab9b3551f6df6f335..d7efdbf640c7d5e39cd8c3c55f79aa0b8ca64f10 100644 (file)
 
 #ifdef __HAVE_ARCH_PTE_SPECIAL
 
-static inline void get_huge_page_tail(struct page *page)
-{
-       /*
-        * __split_huge_page_refcount() cannot run
-        * from under us.
-        */
-       VM_BUG_ON(page_mapcount(page) < 0);
-       VM_BUG_ON(atomic_read(&page->_count) != 0);
-       atomic_inc(&page->_mapcount);
-}
-
 /*
  * The performance critical leaf functions are made noinline otherwise gcc
  * inlines everything into a single function which results in too much
@@ -58,8 +47,6 @@ static noinline int gup_pte_range(pmd_t pmd, unsigned long addr,
                        put_page(page);
                        return 0;
                }
-               if (PageTail(page))
-                       get_huge_page_tail(page);
                pages[*nr] = page;
                (*nr)++;