Enabling the SSE Wilson Dslash

This native build of Chroma is going to be very slow. On SSE enabled systems you should switch on the SSE assembler optimised dslash. This can be done with the --enable-sse-wilson-dslash to chroma. Depending on the precision of the underlying QDP++ build, chroma will select the right dslash for you (both single and double precision are available). If you were doing a parallel build, that would also be detected and the right dslash would be selected. The previous command becomes:

/home/bj/src/chroma/configure CXXFLAGS="" CFLAGS="" \
  --with-qdp=/home/bj/install/qdp_scalar_single \
  --enable-sse-wilson-dslash \
  --prefix=/home/bj/install/chroma_scalar_single