[PATCH] D18201: Switch over targets to use AtomicExpandPass, and clean up target atomics code.

[Tim Northover via llvm-commits]

t.p.northover added a comment.

On the plus side, GCC itself seems to be as buggy as clang (i.e. it loads with a cmpxchg16b with no loop to ensure success) so maybe real processors are more forgiving than the documentation suggests.

Tim.


http://reviews.llvm.org/D18201