amd64 back end: generate improved SIMD64 code.

amd64 back end: generate improved SIMD64 code.

For most SIMD operations that happen on 64-bit values (as would arise from MMX
instructions, for example, such as Add16x4, CmpEQ32x2, etc), generate code
that performs the operation using SSE/SSE2 instructions on values in the low
halves of XMM registers.  This is much more efficient than the previous scheme
of calling out to helper functions written in C.  There are still a few SIMD64
operations done via helpers, though.