Fix ordering of operands on lowering of atomicrmw min/max nodes on ARM.
[oota-llvm.git] / test / CodeGen / X86 / neg_cmp.ll
2012-04-27 Chad RosierAdd x86-specific DAG combine to simplify: