修复了fused_transformer_encoder_layer fine-tune过程发现的一些问题：

    fused_attention_op添加attn_mask=None的支持：PR
    pre_layer_norm处理问题：PR
    参数处理，计算错误的问题：PR
    add_bias计算错误问题：PR
    添加pure fp16的支持：PR

* add mlperf optimization PRs

* update

* Fix bug when pre_layer_norm is false.

功能：本PR的目标是提高attention模块的计算性能。
为了减少框架层对op的调度开销，本PR通过在C++层手动实现attention模块，对外提供attention 大op；
为了减少防存开销，本PR采取了两种优化方法：
（1）在q,k,v计算时通过共享输入X，将该处的gemm，transpose和bias add从三次调用减少为一次；
（2）使用kernel融合优化技术，在不同cuda kernel之间通过寄存器传输数据；

* add op: fused_feedforward(backward) (#35611)

这个PR是fused_feedforward反向的代码

相关kernel实现：fused_dropout_act_bias, fused_residual_dropout_bias, fused_layernorm_residual_dropout_bias

fused_feedforward是一个融合算子，该算子对transformer模型的feed forward层的算子进行融合和封装，使得前端只呈现一个接口，通过融合减少部分访存和kernel launch的时间，以此提升性能。

* Move fused_attention and fused_feedforward functional api path to incubate (#36704)

将 #35905 和 #35843 PR中新增的的python api接口移到incubate目录下。

This is a fusion operator to compute feed forward layer in transformer model architecture.

功能：本PR的目标是提高attention模块的计算性能。
为了减少框架层对op的调度开销，本PR通过在C++层手动实现attention模块，对外提供attention 大op；
为了减少防存开销，本PR采取了两种优化方法：
（1）在q,k,v计算时通过共享输入X，将该处的gemm，transpose和bias add从三次调用减少为一次；
（2）使用kernel融合优化技术，在不同cuda kernel之间通过寄存器传输数据；

功能：本PR的目标是提高attention模块的计算性能。
为了减少框架层对op的调度开销，本PR通过在C++层手动实现attention模块，对外提供attention 大op；
为了减少防存开销，本PR采取了两种优化方法：
（1）在q,k,v计算时通过共享输入X，将该处的gemm，transpose和bias add从三次调用减少为一次；
（2）使用kernel融合优化技术，在不同cuda kernel之间通过寄存器传输数据；

In fused_attention op and fused_ffn op, the fused bias_add+dropout+residual+layernorm kernel or bias_add+dropout+residual kernel is used. To ease the use of this kernel, we provide a wrapper in this PR.
1.To reuse the increment computing code, we exact the corresponding code to "GetSeedDataAndIncrement" routine in dropout_impl_util.h.
2.The fused_dropout_helper.h provides the fused dropout kernel wrapper.

Note: the test of this warper will be provided in the following fused_attention_op and fused_ffn PRs.

* Fix a bug in ReadData, ReadDataBc and ReadDataReduce when NX != 1
* Update the implement of reduceAnyKernel according to kernel primitive api

* broadcast qkv_op

* use PADDLE_ENFORCE_GT to replace assert

Fused elementwise_add, dropout, elementwise_add and layer_norm into one operator, only support Forward. 
No Python API changed.

Implement FunctionTraits to support two kinds of elementwise functor and remove some old codes for broadcast. (#35688)

Revert "Implement FunctionTraits to support two kinds of elementwise functor and remove some old codes for broadcast. (#35487)" (#35686)

Implement FunctionTraits to support two kinds of elementwise functor and remove some old codes for broadcast. (#35487)

* Add fusion_lstm INT8 PTQ

* Correct mkldnn_cache_capacity and enable fc_lstm_fuse_pass only for this test

* Change mkldnn_cache_capacity

Describe

Add feed_forward for fused attention op.
(1) Encapsulate matmul impl (forward and backward) used in attention op.
(2) Implement bias_add (forward and backward) used in attention op.

Refactor the organization of layer_norm cuda impl so that it can be reused in fused attention op.

    Extract the layer_norm cuda impl form layer_norm_op.cu to layer_norm_kernel.cu.h.
    Define fused/attention_layer_norm.h, which can be used in fused attention op in next PR.

This PR adds fused transformer related files defining c interface including class, function etc..

* Small changes related to BF16 fusion_gru and fusion_lstm

* Correct to pass arg by value

* Add conditions to rnn op

* Correct the spelling mistake

* Improving the test with checking activation

* Trigger CI

* fix two error message

* fix two error message

* fix error

* fix error

* fix error

* fix error

* fix some error message

* fix some error

* fix error

* fix some error

* fix some error

* fix some error

* fix one error

* fix some error

* fix seven error message

* fix error

* fix error

* fix error

* fix error

* fix some error message

* fix error

* fix some error

* fix some error

* fix four error message

* fix error

* fix error

* fix test_unpool_op

* fix test_inplace_addto_strategy

* fix test_conv2d_fusion_op

* fix test_imperative_lod_tensor_to_selected_rows, test_imperative_selected_rows_to_lod_tensor

* fix test_dot_op

* fix test_correlation_op

* fix tracer

* fix test_memcpy_op

* delete include framework.pb.h

* fix error

* [ROCM] update fluid operators for rocm (part3), test=develop

* fix clang format error, test=develop

* Modify relu native implementation

* fix GPU performance

* 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

* changed stream handling

* minor change

* added datatype to GetExpectedKernelType()

* added reading stream from TLS

This reverts commit d834f4e6.

* 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

* More precise mkldnn kernel choice in GetExpectedKernelType

* Fixes after review

* Refresh develop for CI

* CI experiment

* get back from CI exper