Merge branch 'develop' of https://github.com/PaddlePaddle/Paddle into add-async-ssa-graph-executor-communicator
test=develop

* fuse optimizer

* Revert "[MKL-DNN] Fix to crash of Transformer when mkldnn is to be used (#16233)"

This reverts commit 13816dd4.
Apart from enabling transformer for MKL-DNN

* Revert "- MKL-DNN pooling updated to set_prim_desc"

This reverts commit c63f6b20.

Conflicts:
	paddle/fluid/operators/mkldnn/concat_mkldnn_op.cc

* Revert "[MKL-DNN] MKL-DNN specific Tensor modification (#15429)"

test=develop

This reverts commit dec9cf53.

* - concat compilation fix

- lint

test=develop

- Lint fixes

test=develop

- Lint fixes

test=develop

- Fix Transpose MKLDNN op

test=develop

Fix grad op makers

* fix threaded executor bug
test=develop

* change the order of class member
test=develop

* Fix Travis CI
test=develop

Clean codes and fix some bugs.

Revert "Fix allocator bug"

* modify the interface of Pass::Allay
test=develop

* Polish code
test=develop

* Fix Travis CI
test=develop

* fix Pass::Apply interface
test=develop

* Fix Travis CI
test=develop

* Make unitest of fsp op faster and more stable.
test=develop

* Skip unitest of fsp op.
test=develop

Add SpectralNormGradOpDescMaker

test=develop

Fix gc bug

test=develop

test=develop

add include headers to fix travis-ci
test=develop

* Refine default MKL-DNN Pass order

test=develop

* Add comment to default MKL-DNN Pass list

test=develop

* C-API quantization core

test=develop
Co-authored-by: NSylwester Fraczek <sylwester.fraczek@intel.com>

* Decouple Quantizer from AnalysisPredictor

test=develop

* fixes after review

test=develop

* renamed mkldnn quantize stuff

test=develop

* remove ifdef from header file

test=develop

* test=develop, fix space_to_depth_doc

* test=develop, refine indent

* test=develop, refine code

* test=develop, add api spec

test=develop

Fix allocator bug

test=develop

test=develop

test=develop

test=develop

* Disable compare for accuracy issue.

test=develop

* Add todo comments to show more information.

test=develop

Enhance gc to support deleting tensor buffer in advance

* test=develop, add install check

* test=develop, add install check scripts

* test=develop, refine language

* test=develop, add api spec

* test=develop, change cdn to bj to pass ci

test=develop