* Modify MarkTrtEngineOutputs API

* update

* fix bug

* fix bug

* fix bug

* fix bug

* rerun ci

* turn off shared_phi

* update

* update

* update

* fix contiguous

* polish code of pass and executor

* update ut

* [NewIR]Refine CINN Dilact directory

* fix conflict

* fix deps

* fix unittest deps

* link C++ tests to libpaddle.so except windows

* fix compile kill-9 bug

* fix compile kill-9 bug

* fix compile kill-9 bug

* fix compile kill-9 bug

* Fix attr type error like concat axis

* Fix None input error

* Fix intermediate output

* support vjp code gen

---------
Co-authored-by: N0x45f <wangzhen45@baidu.com>

* support index_select op

* index_sample in cpu

* support index_sample in gpu

* change data_transform

* fix api gen and use skip_transform in yaml

* [NewIR]Part-2.1 Refactor NewIRCompiler to support Group Ops

* fix gflags link error

* fix include ir_printer.h

* fix unittest

* fix conflict

* fix flags

* fix comment

[clang-tidy] No.34,36 enable performance-noexcept-move-constructor,modernize-use-transparent-functors (#56261)

* fix

* fix

* CI

* fix

* fix

* fix

* fix

* fix

* fix

* fix

* fix

* fix

* fix

* CI

* fix

* CI

* Generate pd_op.parsed.yaml from pd_op.yaml

* Generate pd_op.parsed.yaml from pd_op.yaml

* fix bug

* bug fix

* bug fix

* bug fix

* 向pd_ops.yaml中新增算子 & 修改pd_ops.parsed.yaml存放路径

* 修复路径依赖bug & 添加 .gitignore文件

* fix bug - compat input args in save_combine op

* fix compat file

* fix set_value_with_tensor yaml

* split backward op in original yaml file

* add send_v2 & recv_v2

* fix bug

* fix bug

* fix install check bug

* fix bug

* Add elementwise_add support into NHWC IR

* update

* fix batch norm grad args def

* fix bug

* fix combine slice bug

* fix slice bug

* update builtin split

* disable using kernel resigter dtype

* polish code

* disable some test

* move matmul spmd rules into phi

* add basic infer spmd utils

* addspmd factory

* fix compile error

* add unittest

* refine infer spmd test and utils

* debug infer spmd test

* adapt python test

* poish details

* change to vector attr arg

* revert needless change

* update matmul spmd rule test

* remove original rule

* polish details

* fix marco error

* add comment

* pass backward test

* fix compile error

* add cmake rule for spmd_rules_test

* add dist meta tensor

* update pybind impl

* add marco for rules

* update

* repalce gflags header

* replace DEFINE_<type> with PD_DEFINE_<type>

* fix bug

* fix bug

* fix bug

* update cmake

* add :: before some paddle namespace

* fix link error

* fix CI-Py3

* allow commandline parse

* fix SetFlagsFromEnv

* fix bug

* fix bug

* fix CI-CINN

* fix CI-Coverage-build

* fix CI-Windows-build

* fix CI-Inference

* fix bug

* fix bug

* fix CI-CINN

* fix inference api test

* fix infer_ut test

* revert infer_ut gflags usage

* update

* fix inference

* remove flags export macro

* revert inference demo_ci gflags usage

* update

* update

* update

* update

* update

* update

* update

* update

* fix bug when turn on WITH_GFLAGS

* turn on WITH_GFLAGS

* fix bug when turn on WITH_GFLAGS

* fix bug when turn on WITH_GFLAGS

* update

* update and add unittest

* add unittest

* fix conflict

* rerun ci

* update

* resolve conflict

* add_arg_mapping_for_fetch

* fix

* fix

* for verify

fluid operator support new comm library

* u

* u

* u

* compatiable new comm library upgrade for c_allgather, c_reduce, c_reduce_scatter and c_scatter.

* Remove useless comments in process_group.py

* Polish code style.

* Fix some problems.

* Remove use fluid api in phi comm_context_manager.

* Add PPADDLE_WITH_CUDA and PADDLE_WITH_NCCL micro judgement.

* Fix bug of HIP architecture.

* Fix some problems.
1. remove useless loggings.
2. Fix conditional compilation for HIP.
3. Fix problems of test_pass_generation_pipeline.py. It calls paddle.distributed.init_parallel_env() at first,
then auto.Engine calls _init_comm(), which will calls process_group.instantiate(). However, init_parallel_env() will call
paddle.distributed.barrier(), it will call CreateNCCLEnvCache and create corresponding NCCLCommContext. But dev_id is not
set, as a result, NCCLCommContext's dev_ctx is not initialized.

* Fix some problems.

* Polish code.

* Polish code.

* Revert compatiable upgrade for communication operators. Their upgrades
will be submitted in another PR.

* Remove StaticTCPStore.

* Remove useless modification.

* Remove useless set_cuda_device_id.

* Polish code.

* Remove fluid header files in phi files.

* Remove useless comments.

* Fix problems of hip arch.

* Fix some problems.

* Polish code.

* Polish code style.

---------
Co-authored-by: hitywt <yuwentao126@126.com>

* Register LegacyKernelDialect & Rigister LegacyKernelOp

* fix code style

* delete LegacyKernelDialect ,register LegacyKernelOp into PaddleKernelDialect

* fix bug

* change as reviewed comments

* bug fix

* bug fix

* try to restart coverage CI

* pass legacy op to kernel pass

* fix code style

* fix code style

* fix code style

* [clang-tidy] enable `hicpp-exception-baseclass` and fix existing errors

* config

* update error format to pass the ci check (at least 20 chars)

* [AutoParallel][NewIR] support calc_sync/comm_sync/send_v2/recv_v2

* pre-commit

* rm unittest

* tiny fix

* api_gen support send_v2's output is empty

* fix format

* python_c_gen support send_v2

When using paddle2cinn, CompilationContext.with_instantiate_variables should be set to false, otherwise CINN will instant and manage variables memory, this leads to double the memory usage, which eventually leads to out of memory error.
This PR will set CompilationContext.with_instantiate_variables to false before context pass to constructing the graph compiler.

* add vjp code gen for SplitOp

* change vjp manual file name