move "temp = numel / (dim[0] * dim[1]);"  out to improve prelu performance 

dd preluop support genet && add unit test  close #799

update register list of resnet closed#797

Add debug info for FPGA track close #795

#793:fix tensor print bug on FPGA

fix:#791

Correct FPGA kernels. close  #789

fix:#787

fix:#785

add quantilization func to fusion_fc and fc_relu kernel closed #783

change FPGA kernel pointer type close #781

addeded fpga softmax kernel:#779

Correct feed_op for FPGA track close #777