Split the Add, Sub, and Mul instruction opcodes into separate
[oota-llvm.git] / test / CodeGen / X86 / vec_shuffle-26.ll
2008-11-16 Mon P WangImproved shuffle normalization to avoid using extract...