* Merge conflicts with 1.4

* Resolve conflicts

* Fix some comments

test=release/1.4

* Fix py3 error

test=release/1.4

* Security issue (#16774)

* disable memory_optimize and inpalce strategy by default, test=develop

* fix security issue
http://newicafe.baidu.com:80/issue/PaddleSec-3/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-8/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-12/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-32/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-35/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-37/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-40/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-43/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-44/show?from=page
http://newicafe.baidu.com:80/issue/PaddleSec-45/show?from=page

test=develop

* revert piece.cc, test=develop

* adjust api.cc,test=develop

* fix overflow by int32 mul test=develop (#16794)

* fix overflow by int32 mul test=develop

* fix reference nullptr

* fix codestyle test=develop

* modify to point in ContextProjectFunctor test=develop

* modify to point in ContextProjectFunctor test=develop

* modify . to -> test=develop

* test=release/1.4 cherry-pick (#16783) (#16794) (#16774)
fix security issue

test=develop

test=develop

test=develop

test=develop

Make parent_idx a dispensable output for beam_search op to support models saved by older paddle version. (#16106)

test=develop

Make parent_idx a dispensable output for beam_search op to support models saved by older paddle version. (#16106)

test=develop

test=develop

test=develop

test=develop

test=develop

test=develop

* Revert "Optimze Gelu with MKL Erf function (#15770)"

This reverts commit 676995c8.

* test=develop

* Optimize for gelu operator

* Set up the low accuracy mode of MKL ERF function.

test=develop

* Only enable MKLML ERF when OS is linux

* Use the speical mklml version included vmsErf function to verify gelu mkl kernel.

test=develop

* Add the CUDA macro to avoid NVCC's compile issue.

test=develop

* Add the TODO comments for mklml library modification.

test=develop

* Clean Code

test=develop

* Add the comment of marco for NVCC compiler.

test=develop

test=develop

test=develop

test=develop

test=develop

* Refine beam_search_op to output an extra parent_idx tensor.
test=develop

* Fix the unittest test_beam_search_op.
test=develop

* Fix the merging mistake.
test=develop

test=develop

test=develop

test=develop

* Refine the beam_search op and test.

* A basic CUDA implementation of beam_search for small batch_size.

* Implement CUDA kernel for beam_search_op.

* Use multiple CUDA threads in the same block to select the top beam.

* Update the python api of beam_search op.

* Enable extend function in CPU kernel of beam_search op.

* Unify the CUDA codes.
test=develop

* Unify the CPU kernel of beam_search op.

* Ensure the seletced items of beam_search_op's CPU kernel sorted by scores.

* Update the description of beam_search in API.spec.

* Enable the use of CUDA kernel in beam_search op.

* Exclude the beam_search's CUDA unittest when there is no CUDA gpu, and delete some debuging statements.
test=develop

* Follow comments.
test=develop

* Call the CPU kernel for beam_search op when batch_size > 4.
test=develop

* Remove the except of is_empty op in PrepareData.
test=develop

* nce add check sample lables, test=develop

* mem opt

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* test=develop

* refine code  test=develop

* refine code  test=develop

* refine code  test=develop

* refine code  test=develop

* refine with cub test=develop

* fix mkldnn test && remove comments && test=develop

* polish code && test=develop

* add only_forward test && test=develop

* refactor tree2col operator with new memory mechanism test=develop

* test=develop

* test=develop

* Modified API according to panyx0718 test=develop

* fix API change according to heavengate test=develop

* Modify API comment test=develop

test=develop

* clip softmax for fp16

* updates

* fuse xent support fp16 test=develop

* wip

* wip

* add simple row reduce

* wip fp16 accurate softmax

* add accurate softmax kernel for fp16 test=develop

* update test=develop

* fix cpu build test=develop

* update api.spec test=develop

* follow comments test=develop

* fix build test=develop

* fix trt build test=develop

* fix inference build test=develop

* fix merge test=develop

* update test=develop

* try fix build test=develop

* fix build test=develop

* rename real_exp test=develop

* fortest

* remove hacky kernels test=develop

* clean up test=develop