Re: [Boost-users] [interprocess] Is there a bug in spinlock atomic implementations 1.39 and 1.40?