]> git.hungrycats.org Git - linux/commitdiff
mm: exclude HugeTLB pages from THP page_mapped() logic
authorSteve Capper <steve.capper@arm.com>
Thu, 28 Apr 2016 23:18:24 +0000 (16:18 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 4 May 2016 21:49:09 +0000 (14:49 -0700)
commit 66ee95d16a7f1b7b4f1dd74a2d81c6e19dc29a14 upstream.

HugeTLB pages cannot be split, so we use the compound_mapcount to track
rmaps.

Currently page_mapped() will check the compound_mapcount, but will also
go through the constituent pages of a THP compound page and query the
individual _mapcount's too.

Unfortunately, page_mapped() does not distinguish between HugeTLB and
THP compound pages and assumes that a compound page always needs to have
HPAGE_PMD_NR pages querying.

For most cases when dealing with HugeTLB this is just inefficient, but
for scenarios where the HugeTLB page size is less than the pmd block
size (e.g.  when using contiguous bit on ARM) this can lead to crashes.

This patch adjusts the page_mapped function such that we skip the
unnecessary THP reference checks for HugeTLB pages.

Fixes: e1534ae95004 ("mm: differentiate page_mapped() from page_mapcount() for compound pages")
Signed-off-by: Steve Capper <steve.capper@arm.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Ingo Molnar <mingo@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>
include/linux/mm.h

index 516e149443397dcf712e4e62aca6fb3c78e542b6..2332bd805e5767a7335fba9a2b19daa1e46f0d0c 100644 (file)
@@ -1010,6 +1010,8 @@ static inline bool page_mapped(struct page *page)
        page = compound_head(page);
        if (atomic_read(compound_mapcount_ptr(page)) >= 0)
                return true;
+       if (PageHuge(page))
+               return false;
        for (i = 0; i < hpage_nr_pages(page); i++) {
                if (atomic_read(&page[i]._mapcount) >= 0)
                        return true;