* add code

* fix bug

* fix bug

* support ir fill constant

* add inverse kernel for xpu

* add more kernels

* add index_put kernel for xpu

* add index_put kernel for xpu

* remove unused headers

* refine test

* wait to avoid memory bugs for xpu

* refine inverse

* polish dist teensor design

* adjust constructor

* polish details

* polish details design

* fix compile error

* refactor init tensor impl

* fix reshard test

* polish details

* add unittest for coverage

* Modify kernels to support quantized_matmul

---------
Co-authored-by: Nsuperxf <1208713646@qq.com>

* add complex numel

* change test && add doc

* Add c_embedding forward compat op.

* Fix some bugs.

* Polish code style.

* fix strided slice compute bug

[Inference] Make share_external_data supports bf16 and bool; fix while_op cache_inference_while_scope when using fleet_executor. (#56055)

* 1. make share_external_data supports bf16 and bool; 2. don't drop_kids when cache_inference_while_scope

* fix FLAGS_cache_inference_while_scope

* add unitest

* add unitest

* skip unitest when cudnn_version < 8100

* skip test share_external_data_bf16 when CUDA_ARCH < 80

* fix op translator reshape type

* update

* new ir support vector type place transfer

* add test case

* update

* revert code

* add test assert new ir test

* update

* update

This reverts commit fd765f61.

* add lu_unpack data check

* add error input api test

* add error type info

* [NewIR] support c_broadcast

* add legacyOpList

* refine static op return val

* update

* update

runtime compiler api will only build special target if it is bind.

'--include-path' is not supported by hipcc and "-I/include/folder"
is better choice

fix ut:
        * device_code_test
        * test_code_generator
        * test_fusion_group_pass
        * test_fusion_group_op
Signed-off-by: Njiajuku <jiajuku12@163.com>

* [Fluid] Migrate c_embedding to PHI

* fix

* add python_api

* fix ut

* migrate xpu kernel

* fix windows compile error