projects / oota-llvm.git / history
? search:
summary | shortlog | log | commit | commitdiff | tree
first ⋅ prev ⋅ next
Allow to fold vector load if there is more than one bitcast, so in the case:
[oota-llvm.git] / test / CodeGen / X86 / avx2-vperm.ll
2012-04-15 Elena DemikhovskyAdded VPERM optimization for AVX2 shuffles blob | commitdiff | raw
C/C++ LLVM-based compilers that forbids OOTA behaviors