add source code for reduction benchmarks

src/calc-res-gsl.h

src/calc-res.h

src/fused-atomic-gsl.h

src/fused-atomic.h

src/fused-cub-atomic.h

src/fused-cub-gsl-atomic.h

src/fused-shared-atomic.h

src/fused-shared-gsl-atomic.h

src/fused-shfl-atomic.h

src/fused-shfl-gsl-atomic.h

src/fused-shfl2-atomic.h

src/fused-shfl2-gsl-atomic.h

src/main.cpp

src/red-atomic-gsl.h

src/red-atomic.h

src/red-base.h

src/red-cub-atomic.h

src/red-cub-gsl-atomic.h

src/red-cub-gsl-cascade.h

src/red-shared-gsl-atomic.h