* Optimize topk's performance when k is small and input_width is large

* 修改blockdim设置逻辑

* Update top_k_function_cuda.h

* first commit

* clarify the quotes

* change code style format

* rerun for ci

* add fp16 support

* update

* update half

* code format

* fix unittest

* fix rocm compile error

* code format

* code format

* fix rocm compile error

* fix rocm compile error

This reverts commit e8de9dfd.

* use cinn in the paddle inference

* fix some cmake errors

* Avoid division by zero in the arange_kernel.

* Avoid dynamic ops.

* Remove some useless codes.

* Use OpTransInfo to encapsulate some codes used in the build_cinn_pass.

* sort out index

* for add_bias

* migrate sigmoid with cross entropy, and tile xpu kernels to phi, test=kunlun

* migrate add_n kernep to phi, test=kunlun

[PHI] Migrate concat+grad, expand+grad, fill_constant, nearest_interp and bilinear_interp oneDNN kernels (#45863)

* Migrate concat+grad, expand+grad, fill_constant, nearest_interp_v2 and bilinear_interp_v2 oneDNN kernels to PHI

* Remove old namespace variable

* Fix invalid out dims error

* Add mutable_data method to concat output

* Add check for -1 dim before computing out_dims

* Capitalize oneDNNGetDataType function name

* Change fill_constant kernel to correct PHI kernel

* Attempt to fix dims error

* Fix fill_constant (full) kernel

* optimize adaptive_pooling_op (forward)

* fix bug of AdaptiveKernelMaxPool2dWithIdx

* fix bug of AdaptiveKernelPool2D

* init

* remove softmaxop

* merge dev

* correct dir

* style

* Convert split, pad and pad3d kernels

* Convert slice+grad oneDNN fluid kernels to PHI

* change out->mutable_data to dev_ctx.Alloc

* First approach

* Shape kernel corrected

* Compilation error fixed

* Resize corrected

* Registered types added

* Mistake corrected & types added

* sum kernel deleted

* move_merge_selected_rows

* update code

* fix linspace error in amp

* fix log

* fix amp error

* Revert "Simplify size op impl (#45808)"

This reverts commit c252b1de.

* fix_seg

* fix detection
Co-authored-by: NChen Weihang <sunny_cwh@163.com>

* fix linspace error in amp

* fix log

* fix amp error

* fix ocr error which caused by amp

* add more check

* rename dtype ns

* polish code comments

* polish data_device_transform.cc

* sparse infer_meta

* fix wrong eigen header include

* fix complie bug

* fix nan_inf_utils_detail

* fix resource_manager

* fix conv_miopen_helper

* move sum

* fix ci bugs

* fix ci bugs

* fix set_lod bugs

* fix infershape bugs

* fix ci bugs

* fix ci unittest bug

* fix ci bugs

* perfect code

* update code according comment

* add unittest

* fix ci bugs

This reverts commit c252b1de.

* fix return order error and duplicate results with specific inputs

* first commit

* refine code with template argument

* refine code with template argument

* add ternary broadcast test file

* add ternary broadcast test file

* fix accoriding to ci

* fix op-benchmark ci error

* perfect softmax functor

* fix compile bugs

* fix ci bugs

* Fix bug of reduce_sum op. When input.numel() > INT32_MAX, its result
is wrong.

* Fix some problems.
1. Change fluid head files to phi files.
2. Delete useless code.
3. Fix code style problems.

* Fix some code style problems.

* Fix some code style problems.

* support int64 non-broadcast

* support broadcast case for int64 index

* fix bug

* support more Arity

* remove some codes

* upgrade patchelf to v0.15.0 to pass CI build

* fix bug

* fix patchelf installation

* add debug flags

* remove useless codes

* fix viterbi_decode and set_value op uts

* remove always enable int64

* - mul & matmul changes

- fix

- bs16 correction of strides

* - cosmetic fixes

* - lint

* - fix

* - fix

* - format -> mem_desc

* - fix

* - fix

* - fix

* - fix

* - fix

Support 0 shapes input Tensor for MKL slice kernel

* first commit

* fix some bugs in code

* fix bugs

* to optimize merge one dimension feature

* support bmm and bmm_grad in xpu

* add error removal

* test=kunlun

* refactor code for better structure

* test=kunlun

* add fp16 kernel for bmm

* test=kunlun