Go: Performance optimization for IndexByte for Power
I added a new implementation for both bytes·IndexByte and strings·IndexByte for Power architecture. The new lazy search-based algorithm gives a speedup of up to 15x over the previous implementation. This