fix  #469 add  centra arm func folder

optimize gemm performance

optimize gemm code

fix #458 add kernel init method

fix gemm bug

fix #453 add a Dockerfile for easy setup dev environment

fix #448 fix loop predict bus error

add vector matrix multiplication in Gemm

fix #443 fix dep analysis

fix #442 multithread executor

fix #437 add conv_add_relu fusion op

add 'pld' instruction in Gemm

Create CONTRIBUTING.md