* wip allreduce in op

* wip

* wip

* wip

* wip adding test

* wip for conflict with mp mode

* fix tests test=develop

* fix cpu build test=develop

* fix travis clang format test=develop

* fix cpu build test=develop

* update api.spec test=develop

* delete comment test=develop

* fix cpplint test=develop

* fix test=develop

* follow comment test=develop

* add file test=develop

* fix build test=develop

* update test=develop

* to be compatible with sync_bn, and fix mp mode in develop test=develop

test=develop

* Support Sync Batch Norm.
* Note, do not enable it in one device.

Usage:

build_strategy = fluid.BuildStrategy()
build_strategy.sync_batch_norm = True
binary = fluid.compiler.CompiledProgram(tp).with_data_parallel(
        loss_name=loss_mean.name,
        build_strategy=build_strategy)

test=develop
This reverts commit 358e657f.

test=develop
This reverts commit 064512aa.

* remove workspace_handle in conv2d_cudnn
test=develop

* remove workspace_handle
test=develop

* fix bug
test=develop

* make test_conv2d_op SERIAL
test=develop

* save memory in conv_cudnn
test=develop

* enhance thread safety
test=develop

* enhance temporary allocator
test=develop

* Add excess fraction
test=develop

* follow comments
test=develop

* fix bug and code refine
test=develop

* fix memory size check
test=develop

* rename reuse_tmp_allocation_excess_fraction
test=develop

test=develop

test=develop

test=develop

test=develop

* refine tensor
test=develop

* refine tensor
test=develop

* fix device_context log
test=develop

* Add Temporal Allocator

* add Temporay Allocator to DeviceContext
test=develop

* code refine
test=develop

* fix mean_iou
test=develop

* Add DeviceTemporaryAllocator
test=develop

* fix conv_op bug
test=develop

* small fix
test=develop

* code refine
test=develop

* log refine
test=develop

* fix unit test
test=develop

* move double check

* refine concat_and_split
test=develop

* add limit_of_temporary_allocation
test=develop

* fix name
test=develop

The macro should be defined by compiler rather than by source.

test=develop

* refine cublase
test=develop

* code refine

* refine cublas

* add GEMME_EX

* add enable_cublas_tensor_op_math doc and add cublasCall
test=develop

* fix CublasCall for cuda version
test=develop

* fix error
test=develop

* fix GEMM_EX to be compatible with gcc 4.8
test=develop

* add GEMM_EX
test=develop

* to compatiable with gcc4.8
test=develop

test=develop

test=develop

* feat(platform): lazy initialization of devicecontext in pool

Use std::async(deferer, []{...}) to lazy initialize DeviceContext in Pool

test=develop

* Add future includes

test=develop

test=develop

test=develop

Make all blob info saved in global device context to be thread based.
Meanwhile save thread id in thread local storage in ParallelDo

test=develop

test=develop

This reverts commit 1d91a49d.

* some trivial opt

* remove the fix of lod_tensor and shrink_rnn_memory_op

* refine ShrinkRNNMemoryOp

test=develop

This reverts commit 151e169e.