Add skeletons of `mul`, `rowwise_add`, `sigmoid`, `softmax` ops

Fix some link errors about NNPACK.

refine tensor's conditional compilation, remove Tensor::numel_ and add DDim::size()

Enable tensor test

* Also fix unit test

Python Generate OpCreation Methods by OpProto

* add inputs

* add ut for multiple inputs

* fix AddToLayer

* op_desc -> op_proto

* CreateArgumentOffsetMap -> CreateInOutOffsetMap

* move CreateInOutOffsetMap from OperatorBase to op registry

* arg_idxs_ -> in_out_idxs_

* Implement InferShape and register them, give a stub Kernel method
  by LOG(INFO)

CMake `op_library` function

Update the dockerfile to build cross-compiling image for Android.

Refactor DDim's product() and add slice_ddim()

* add NDEBUG switch to PADDLE_ENFORCE

Fix fetch record from master

Change to meet cpplint style

"fix dense vector shape in mq2007"

* using etcd as fault tolerant training

* update

* workable version, ft not tested

* small fix

* update

* remove TODO

"fix unrar in docker"