"Generic Scan"

Single block generic parallel scan (CUDA)

Compilation

# run from repo dir
nvcc -o out/generic-scan generic-scan.cu

TODOs

• push the integer version (done by Nadhir)
• fix running prefix sum direction (right to left currently)
• write a generic kernel (using C++ templates and functors)
• write a generic prescan (exclusive scan)
• cite Hillis & Steele's paper