Merging these functions allows merging some loops, which makes the
results (particularly after SIMD optimizations) much faster.
(cherry picked from commit f8bed30d8b)
2.8 KiB
2.8 KiB