* First basic implementation

* Add infer example for image_classification

* Address review comments: round 1

Add RunAndGetException in threadpool

Refine buffer receive

Change the behaviour of thread pool. Thread pool will ignore the
thrown exception implicitly. It is hard to debug.
Corrently, ThreadPool::Run will invoke `LOG(FATAL)` if an exception
thrown.

Add type `Reader` for `VarDesc`

Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into dev_make_VarDesc_supporting_multiple_tensor

Fix warnings in multiclass_nms_op.cc.

* Update csp.md

* Update csp.md

* Update csp.md

helper functions fetch_var and get_var

* Add proposed fluid syntax for select statement in Fluid's implementation of CSP

* Fix Typo

* Adding send and recv in design doc

* fix typo

* fixed code

* Adding threading

Refine prior_box_op

add independent sphinx tree for api

Add a new type `Reader` for `VarDesc`, which can holds more than one
LoDTensor.

Add `make clean` in docker/build.sh

remove duplicated mobile index

Fix the error when sorted_key is none in profiler