Move Copy out from memory.h into memcpy.h

Slight simplification of eigen.h in hope to make it more readable.

Add CPU/GPU Memcpy in memory folder

Simplify Tensor implementation

enable Intel MKLDNN library and Intel MKL small package

Mobilenet gpu implementation

Implement some basic op kernel

1. Change PODDeleter's template parameter 'PlaceType' to 'Place'.

2. Limit PODDeleter and Tensor::mutable_data()'s `T` to POD type.

* add cross-entropy-op

* add infershape and compute

* implement Infershape and compute of onehotcrossentropy op

fix client discover pserver context cancelled

It's already fixed by Wuyi's PR, but his PR may take some time to
merge, but I want to get this change in ASAP.

Update Memory ReadMe Doc

Copyright

fix recurrent_group parsing bug.

fix flowers dataset bug