OperatorWithKernel

* User can register OpKernel to its Ops. The OpKernelMap saved in
  OperatorWithKernel. Each Op which inherits OperatorWithKernel will
  use `OpKernel::Compute` instead of Run.

Remove Dim::contiguous and Dim::contiguous_strides

add tensor interfaces

Paddle's data block is row-major order, while Dim::contiguous and
Dim::contiguous_strides are based on column-order. So remove them to
prevent misuse.

Refine CUDA Related libraries

Error fix in doc on Write New Layers

Fix new remote updater for go pserver

Add OperatorBase.

issue: https://github.com/PaddlePaddle/Paddle/issues/2790

Paddle design the Operator with Kernel. OperatorBase has no type and device information when create, One operator can have multiple kernels, Operator will choose a kernel to run according to context. The kernel should be bind to Operator before or during Operator running.

1. Add member variable 'DDim dims_' and a getter function 'dims()'.
'dims' is supposed to hold tensor's shape during Op::InferShape.
2. Remove 'mutable_data' which use default Place. User must specify a
explicit Place when call 'mutable_data'.
3. A PlaceHolder may be shared by more than one tensor, and some of them may be the others' slices. So we add a new member variable 'offset_' for Tensor, which is used to show the byte offset between PlaceHolder::ptr_ and where tensor's data really begins.
4. Add functions 'ShareDataFrom' and 'Slice' for Tensor.

TODO: Tensor needs a 'CopyFrom' function.

* Move static variable defined in .cc

We cannot define static variable in .h, because it will be multi-defined
errors.

Also fix some cpp syntax, like:

* Prefer to use algorithm not manually for-loop, to make code more
  readable.
* Remove unused `()`.
* Enforce take a bool. It is no need `xxx==true`.
* Use range-based for-loop iterator from op_desc.attrs

* Fix a protential static variable init order error

* Expose paddle.framework by pybind11

* Export paddle.framework.{Scope, Variable} to paddle.v2.framework.core.
* See python/paddle/v2/framework/tests/test_scope.py for Python usage
* See paddle/pybind/pybind.cc for C++ bind code.

* add copyright

add TaskFail interface

Fix slow parsing a recursive depends topology

* generic.cmake can propogate dependencies through libraries. It is no
  need to specific all dependencies.

* Fix  #2797
* It because trainer_config_helpers' __dfs_travel__ did not record the
  node which travelled, and if the topology has a recursive dependency,
  there are some nodes will be travelled multiple times.
* Add a `travelled` set to record which node is travelled.
* Also add a unittest for this situation.

update docs url

fix dynload link error: undefined reference dlsym

Network implement