* added external reorder to profiler

* resolved conflict

* added enable_static

* initial version of lstm, not working yet

* added lstm to operators.cmake

* added vanilla lstm mkldnn op

* added peephole weights integration

* minor changes

* added formatting

* added fusion_lstm_mkldnn to static_whitelist

* added formatting

* removed comment

* moved use_peepholes attribute inside is_cached block

* reverted wrong changes

* minor formatting change

* minor changes

* Add mkldnn nearest_interp and bilinear_interp op
* don't run mkldnn interpolate in default
* add interpolate_mkldnn_pass

* add double grad for conv_transpose

* remove addcmul

* remove unittest and other related code of addcmul

* fix bug

* fix merge conflict

* Add bf16 pool2d and unify bf16 unit tests

* Add change default ops test

* Add multi_gru_fuse_pass and tests

* fix date

* cleaned up headers

* Add multi_gru_seq_fuse_pass and tests

* fix date

* removed unused functions

* Add bf16 matmul, fc, elementwise add and mul

* Correct unit test

* Add multi_gru op and tests

* removed redundant disable_dygraph()

* add mkldnn inplace op version test

* update mkldnn_inplace fuse pass

* update the inplace test

* Add bfloat16 softmax and gelu

* Add pass attr bfloat16_enabled_op_types

* Changes from review

* - Added sum bf16 oneDNN

test=develop

* - Fix to UT of sum bf16

test=develop

* add generate_proposals_v2 op

* add static_mode_white_list

* add Mac CI static list

* add Win CI white_list

* add Coverage and Py3 CI white_list, add test_unittest