add wait event for kernel

 add event

 add event

 commit event

update fetch kernel

 update softmax reshape kernel

update fetch kernel

update fetch kernel

 correct conv

 add fetch kernel for CPU

 add fetch kernel for FPGA, GPU_MALI

# Conflicts:
#	src/operators/kernel/fpga/feed-kernel.cpp

update conv kernel

add fetch kernel for CPU

fix depthwise_conv_kernel of opencl

 fix image memory copy bug