[cpp-threads] N2052 sequencing revisions &
examples(moderatelylong)
Paul E. McKenney
paulmck at linux.vnet.ibm.com
Sat Mar 10 16:52:22 GMT 2007
On Fri, Mar 09, 2007 at 06:55:26PM -0000, Boehm, Hans wrote:
> In this model, any use of acquire or release can be changed to relaxed
> if there is no matching release or acquire in another thread. This
> allows aggressive compiler synchronization removal, but means you cannot
> rely on the fence-based implementations some people expect.
Certainly things like aliasing can make it extremely difficult to
determine whether or not there is a matching release or acquire in
another thread.
Thanx, Paul
More information about the cpp-threads
mailing list