On Fri, Apr 22, 2016 at 11:41:49AM +0100, Will Deacon wrote: > For compound atomics performing both a load and a store operation, make > it clear that _acquire and _release variants refer only to the load and > store portions of compound atomic. For example, xchg_acquire is an xchg > operation where the load takes on ACQUIRE semantics. > > Cc: Paul E. McKenney <paul...@linux.vnet.ibm.com>
Thanks! Acked-by: Peter Zijlstra (Intel) <pet...@infradead.org> > Signed-off-by: Will Deacon <will.dea...@arm.com> > --- > Documentation/memory-barriers.txt | 5 +++++ > 1 file changed, 5 insertions(+) > > diff --git a/Documentation/memory-barriers.txt > b/Documentation/memory-barriers.txt > index 3729cbe60e41..05f8011011be 100644 > --- a/Documentation/memory-barriers.txt > +++ b/Documentation/memory-barriers.txt > @@ -464,6 +464,11 @@ And a couple of implicit varieties: > This means that ACQUIRE acts as a minimal "acquire" operation and > RELEASE acts as a minimal "release" operation. > > +A subset of the atomic operations described in atomic_ops.txt have ACQUIRE > +and RELEASE variants in addition to fully-ordered and relaxed (no barrier > +semantics) definitions. For compound atomics performing both a load and a > +store, ACQUIRE semantics apply only to the load and RELEASE semantics apply > +only to the store portion of the operation. > > Memory barriers are only required where there's a possibility of interaction > between two CPUs or between a CPU and a device. If it can be guaranteed that > -- > 2.1.4 >