include/linux/page_ref.h: ensure page_ref_unfreeze is ordered against prior accesses
authorWill Deacon <will.deacon@arm.com>
Mon, 10 Jul 2017 22:48:28 +0000 (15:48 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Mon, 10 Jul 2017 23:32:31 +0000 (16:32 -0700)
commit108a7ac448caff8e35e8c3f92f65faad893e5aca
tree5f89dd3053a152be5dedb660a232584ec35b507f
parentbaabda261424517110ea98c6651f632ebf2561e3
include/linux/page_ref.h: ensure page_ref_unfreeze is ordered against prior accesses

page_ref_freeze and page_ref_unfreeze are designed to be used as a pair,
wrapping a critical section where struct pages can be modified without
having to worry about consistency for a concurrent fast-GUP.

Whilst page_ref_freeze has full barrier semantics due to its use of
atomic_cmpxchg, page_ref_unfreeze is implemented using atomic_set, which
doesn't provide any barrier semantics and allows the operation to be
reordered with respect to page modifications in the critical section.

This patch ensures that page_ref_unfreeze is ordered after any critical
section updates, by invoking smp_mb() prior to the atomic_set.

Link: http://lkml.kernel.org/r/1497349722-6731-3-git-send-email-will.deacon@arm.com
Signed-off-by: Will Deacon <will.deacon@arm.com>
Acked-by: Steve Capper <steve.capper@arm.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
include/linux/page_ref.h