Fix #6087

* WIP send recv op

* WIP send recv

* put grpc impl in details

* put grpc impl in details

* update wip

* update proto

* update proto

* update proto

* clean cmake

* wip on op implementations

* wip on op implementations

* compile ok adding ut

* wip unitest

* add extern cares for linking

* wip add ut

* working version send recv

* revert optimizer.py

* update test cmake

* add libtool to dockerfile

* update cmake dependency

* update cmake depends

* update cmake grpc depends

* fix cmake dependency

* fix compile error

* fix compile

* follow comments

* update

* update copyfrom

* "make global tensor function independently"

* "replace functor"

* "fix inline template error"

* "fix tensor array with CopyFrom"

* "fix other case use CopyFrom"

* "move the op interface hardly"

* "fix operators"

* "fix typo"

* "delete dynamic recurrent rnn and fix gru_unit in debugmode"

* "fix unique_ptr copy"

* "fix cuda copy"

* "fix namespace error"

* "removed nccl python test"

* "fix include error"

* "fix typo"

* fix copy util test

* is_training to is_test in dropout op

* handle dropout and batch_norm operator when prune pdesc in testing mode

* handle dropout and batch_norm operator when prune pdesc in testing mode

* add get_inference_program method

* fix dropout op

* fix ci

* test data after each batch training

* refine code

* refine test_book3

* fix ci

* follow comments

* Change all `data_type` in Python to `dtype`

* Change `date_type` in C++ to `dtype`

* Refine

* enforce shape of backward target to be {1}

* fix test_regularizer.py

* rm unused code

* fix backward_test

* fix a type bug

* fix test_program

* first commit

* Python API for while op

* Python Unittest for simple while_op forward

* fix out to be list

* Fix UT

* VarType

* Fix several bugs

* Fix bug

* Fix bug

* Fix Bug

* Fix bug

* Fix unittest

* Remove debug log

* Add comments

* add PADDLE_ENFORCE

* while_grad_op first commit

* Add `BlockDescBind::FindRecursiveOrCreateVar()` and fix bugs

* not sure how to setdim of while outputs

* push for test

* add executor vlog

* fix bug of while_op cond

* Several enhancement for code

1. Backward always infer shape & infer var type. Since there are RENAME
variables will be created when creating backward operator, but their
shape & var types are not inferenced.
2. Never use SomePtr-> directly, since every pointer could be nullptr if
it is a function return value. Add `detail::Ref` to cast pointer to
reference safely.
3. Enhance error message for backward.
4. Infer data type of variable in `sum` and `tensor_write`

* Fix bugs of while_op gradient

* Fix several bugs of while_op grad

* fix fill zeros like

* fix 3 >= 3

* fix place holder shouldn't be null

* fail on sum op

* Fix SumOp of TensorList

* clean up

* pass while test

* fix test_array_write_read

* pass sum op

* Support int/int64 for fill_constant_batch_size_like

* Fix compile

* fix gitignore

* refine cmake file

* Conditional Block Forward

* Assign Operator.

Out=X, when type in [LoDTensor/SelectedRows/LoDTensorArray]

* Stash

* Add Scope::Rename

it is useful in gradient phase of an operator with block

* ConditionalBlock Grad Done

* Add comments

* yapf format code

* Assign Operator.

Out=X, when type in [LoDTensor/SelectedRows/LoDTensorArray]

* Follow comments

it is useful in gradient phase of an operator with block

* first commit

* Python API for while op

* Python Unittest for simple while_op forward

* fix out to be list

* Fix UT

* VarType

* Fix several bugs

* Fix bug

* Fix bug

* Fix Bug

* Fix bug

* Fix unittest

* Remove debug log

* Add comments

* add PADDLE_ENFORCE

* while_grad_op first commit

* Add `BlockDescBind::FindRecursiveOrCreateVar()` and fix bugs

* refine code

* fix unittest bug

* increament is default inplace

* Add LoDRankTable

LoD Rank Table stores the `level` of `lod` which is ordered by sequence
length in descending order. It is useful when implement dynamic RNN and
is shared by dynamic RNN memory, dynamic RNN slice input and dynamic
RNN slice output operators.

* Add skeleton for array_to_lod_tensor and lod_tensor_to_array

* Add VarType::LoDTensorArray

* Add PyBind of LoDTensorArray

* Add InferVarType

* Add first unittest

* Add ut

* Add unittest

* Add unittest

* Add unittests

* update

* init

* add infershape for lod_tensor_to_array_op

* compelete array_to_lod_tensor_op

* copy data

* clean code

* clean code

* Fix unittest data

* fix bugs

* fix compile error

* Refine TensorToArrayOp

* refactor array_to_lod_tensor

* Unittest

* fix bugs

* Fix unittest

* Fix unittest

* debug

* Debug

* Fix unittest

* clean code

* refactor

* use ostream

* update test

* fix gpu build error

* make gpu test pass

* Chage `IndicateDataType` to `GetKernelType`. Make it easier to
  understand.
* Change `OpKernelKey` to `OpKernelType`
* Make operator developers can customize which kernel the operator will
  use in runtime.

* Compare Operator

* Follow comments

* Use stable_sort in lod_rank_table

It is easy to debug and test when use `stable_sort`and the time
complexity is not changed.

* Add LoDTensorArray

* Stash

* Better debug message for IsInitialized

* Stash

* Better debug message for IsInitialized

* Complete array read/write op unittests

* Add unittest, Gradient of array read/write

* Follow comments

An important change on lod tensor indexing. A higher level offset will be based on its next level rather than an absolute offset.

* Use stable_sort in lod_rank_table

It is easy to debug and test when use `stable_sort`and the time
complexity is not changed.

* Add LoDTensorArray

* Stash

* Better debug message for IsInitialized

* Stash

* Better debug message for IsInitialized

* Complete array read/write op unittests

* Add GetFineGrainedLoDLength and AppendLoD

* Follow comments and fix bugs

* fix a compile error

* fix a compile bug

* Use stable_sort in lod_rank_table

It is easy to debug and test when use `stable_sort`and the time
complexity is not changed.

* Add LoDTensorArray

It is easy to debug and test when use `stable_sort`and the time
complexity is not changed.

* add acc layer
* memory log level change from 3 to 10
* use gaussian random to init conv parameters
* use initializer
* fix import
* batch_norm use helper to create persistable var
* refine code
* train only 2 batches for test
* use g_program and g_init_program
* use XavierInitializer to init fc parameter

* Add LoDRankTable

LoD Rank Table stores the `level` of `lod` which is ordered by sequence
length in descending order. It is useful when implement dynamic RNN and
is shared by dynamic RNN memory, dynamic RNN slice input and dynamic
RNN slice output operators.

* Add InferVarType

* Init commit

* Make executor use ProgramDescBind

* Change Attribute from BlockDesc to BlockDescBind

* Since we will get the program desc in RNN, just BlockDesc is not
  enough.

* Add DeviceContext to Executor API

* Rewrite RNN

* Pass Python

* AddBiasOp does not care num_flatten_dims

* Stash

* Fix MacOS Compile

* Pass RNN forward

* add python test

* refactor test

* Make compile pass

* add gradopmaker

* First draft done

* Polish code

* add grad op maker and grad infershape

* Polish code

* Fix backward.cc bug

* Fix infershape

* Rename function

* add backward test

* simplify recurrent test

* Update

* Pass unittest

* Add comments & refine test

* Add comments

* refactor test

* Complete Unittest

* fix StepScopes enforce

* Remove unused unittest

* no type error

* Update

* Make RNN Pass unittest

Set gradient's data type based on its forward variable

* Init commit

* Make executor use ProgramDescBind

* Change Attribute from BlockDesc to BlockDescBind

* Since we will get the program desc in RNN, just BlockDesc is not
  enough.

* add shareLod

* fix sequence_conv grad infershape

* add batch_norm_layer

* add img_conv_group layer and test

* add check to Tensor.type()

* forward can run

* with backward

* change label data time from int32 to int64

* refine code

* follow comment