tl;dr for people familiar with the C/C++11 MM: It's very similar. `relaxed` -> `Opaque`, `seq_cst` -> `volatile`. `acquire` and `release` map more or less the s

tl;dr for people familiar with the C/C++11 MM: It's very similar. `relaxed` -> `Opaque`, `seq_cst` -> `volatile`. `acquire` and `release` map more or less the same. Interestingly, there's no equivalent to C++ `acq_rel` on the RMW operations -- you have to either choose the stronger `volatile` ordering, or do a release fence followed by an acquire RMW (or the reverse). Regular loads/stores may be mixed with atomic ones, but don't give any coherence or forward progress guarantees, and may see word tearing for longs or doubles. The java `fullFence()` is stronger than a C++ `seq_cst` one; inserting one between every pair of Opaque accesses gives them sequential consistency (though, I understand that C++ is probably going to strengthen the semantics of `seq_cst` fences eventually).