* onednn: rename macro to PADDLE_WITH_DNNL

* onednn: rename macro to CINN_WITH_DNNL

* [NewIR]Add cinn RuntimeDialect and JitKernelOp

* remove PointerAttribute register

* fix comment

* refactor codegen for cinn

* add to_string to some type which can't be += with string

* fix multi-thread bug caused by static var

* delete dead code and comment

Update Save/Load Interface to 2.0

1. `Parallel Compiler`：
    - 合并`FLAGS_cinn_parallel_compile_size`和`FLAGS_cinn_parallel_compile_thread`，通过`FLAGS_cinn_parallel_compile_thread`即可指定编译时使用的线程数，所有的`fusion_groups`将会平均分配到可用的线程上
    - 增强编译完成后返回的信息，除`instruction`外，将`lowered_function`、`source_code`、`source_ptx`返回，供上层进一步使用
2. Debug信息：
    - 新增`FLAGS_ cinn_dump_group_lowered_func`、`FLAGS_cinn_dump_group_source_code`、`FLAGS_ cinn_dump_group_ptx`、`FLAGS_ cinn_dump_group_instruction`，可分别按`fusion_groups`储存编译的每个阶段中的中间代码
    - 重新整理`graph_visualization`，所有的可视化图、单测代码均能正确分组储存
3. Bug修复：
    - 修复`MakeDirectory`不能正确创建文件夹的问题
4. 其他：
    - 清除了一些无用代码

* add ir core

* add test

* modify name

* merge

* add test for __eq__

* shield  test for __eq__

* --amend

* Update new_ir_compiler.cc

* [CINN Frontend] Optimize user interface, polish fuse_pass classes"

* Fix some compile error

* [NewIR]Split NewIRCompiler with .h/.cc and decoupling compilatiom with cinncore

* fix cmake

* fix CINN_ONLY

Remove the remaining old schedules.

test_dot_merger和test_dot_merger_pass在CI和本地环境中会随机出现segmentation fault错误，导致单测随机挂，经排查后发现是ginac库相关的问题，现避免使用MathIsZero绕过

* [NewIR]Add ConvertIRType and fix some TODO for IR+CINN

* modify into GPUPlace

* [0D-Tensor] CINN supports fill_constant, fix infershape and pass

* fix infershape of fill_constant

* add back fill_constant to zero_tensor_trick_pass

* Refactor the error message system

* fix header

* fix compile

* Just set threshold to 6 in op mapper relu6

* Remove attrs in op mapper relu6

In preparation for the improvement of the graph compiler, the deprecated old graph compiler was cleaned up.

* Clarify cinn/ir dirs [Part1]

* addition of cinn/ir/op dir

* change header inludsion of ir/ir_operator.h to ir/op/ir_operator.h

* merge with develop changes

* relocate libschedule_desc_proto.a

* remove extra ir_schedule_error.cc

* addition for schedule/ir_schedule_error files

* [0D-Tensor] CINN supports argmax, fix infershape

* [0D-Tensor] CINN supports argmax, fix infershape

Remove old schedules.

* [CINN] Schedule error message optimization

* format code style

* add test

* fix format

* using CINN_THROW and using flags

* optimize error msg

* do not use abtract class of error hanlder

* fix header

* new group fuse pass api

* fix header

* update

* change logic of get master node to fix bug

* revert update for ReduceFuseReduce

* modify according review

* modify by review

* refine

* update

* fix code-format

* [CINN] comb the op lowering code

* [CINN] format code of OpLower

* squash pick the poc code
* fix build after rebase
* fix int8 conv and fc uts
* Fix and clean-up Get_SRC_Scale_Memory
* fix floating point fc uts
* fix test_analyzer_int8_googlenet
* test_analyzer_int8_mobilenetv1
* fix int8 mobilenet v2 and v3
* fix build error after rebase
* [oneDNN] rename library version
* fix conv bias datatype
* try to fix import error
* fix rebase error
* [oneDNN] pack library into python wheel
* add MKLDNN_SHARED_LIB_3 to env_dict
* fix test_analyzer_bert
* fix fill_constant op kernel
* fix ernie and matmul op ut
* fix softplus ut
* fix conv+relu6 fusion ut
* fix hardswish fusion
* fix quant+transpose fusion ut
* fixsgd ut
* fix int8 matmul with flatten
* fix fc+scale fusion
* fix conv/matmul+gelu fusion uts
* fix rebase error
* Revert "fix conv/matmul+gelu fusion uts"
This reverts commit 47eb5e49972bd8f7271a233def9bfb3e98ce78e1.
* upgrade to onednn v3.1
* remove older version onednn
* use densetensor::data() for achieving mean and var in layernorm impl
* comments for atol of integer tests
* fix clang-format
* Revert "remove older version onednn"
This reverts commit 783e57ddfd4401254596eae7d47adb9b03590c09.
* improve binary handle
* fix expand kernel
* Revert "use densetensor::data() for achieving mean and var in layernorm impl"
* always use forward_inference for conv
* remove activation scales
* rollback changes to mkldnn.cmake
* address comments
* port changes to dequantize kernel
* fix merge error
* fix fused_elementwise_kernel
* upgrade onednn version to v3.1.1
* fix some approval error
* fix error msg format
* remove old onednn libs
* try to fix symbolic link issue
* fix cinn test case segfault
* do not explicit link test with onednn
* remove unnecessary changes
* integrate CINN with onednn v3
* link with mkldnn project
* fix cinn build file

---------
Co-authored-by: NTomasz Socha <tomasz.socha@intel.com>
Co-authored-by: NChen, Xinyu1 <xinyu1.chen@intel.com>
Co-authored-by: Ntianshuo78520a <707759223@qq.com>

This PR delete middle ir_visitor class and thus we can avoid middle virtual function call and codes look more clean
pcard-72718

* remove some pybind interface in cinn to fix compile problem

* modify cmake

* fix cmake

* add log for build cinn whl

* fix ninja for cinn

* fix conflict

Pcard-72384
paddle with cinn的情况下支持cinn的c++ ut。 统一了cinn和paddle的third_party的c++版本，cinn third party不使用C++17