* overhaul vector_type, make int8x4_t real vector instead of aliasing from int32_t

* initial implementation for magic number division and DynamicMerge_v2_magic_division that uses it

* turn off DynamicMerge_v2_magic_division that use magic number division by default

* Hybrid direct + implicit GEMM forward convolution NCHWc v5r1. Input tensor bypass LDS. Support fp32/fp16/int8

* use address_space(4) in kernel signature to fix performance issue when passing tensor descriptor from host to kernel by (void) pointers

* remove passing by pointer* option (only use pass by value or void*)