* add huber_loss for kunlun

* update xpu.cmake

* update unitests

* update unitests

* update elementwise_add

* update elementwise_add

* update elementwise_add

* [XPU]add split op for kunlun2,*test=kunlun

* [XPU]add split op for kunlun2,*test=kunlun

* [XPU]add split op for kunlun,*test=kunlun
Co-authored-by: NQingshuChen <chenqingshu@baidu.com>

* 6 dims fix

* removed limitations of max dims

* unify data layout

* fix test_transfer_layout error

LGTM

* add sigmoid cross entropy with logits to kl1. test=kunlun

* add sigmoid cross entropy with logits to kl1. test=kunlun

Add exp, abs_grad, reciprocal, reciprocal_grad operator for XPU and update xpu2_op_list.h,test=kunlun (#38570)

This PR adds runtime flags run_kp_kernel, which choose which op to run for xpu2. There are two: dynamic linked and built from kp.

* add cpu kernel of lstsq

* update

* modify code style

* modify unittest

* remove support for complex

    将cuSparse的handle与DeviceContext进行绑定，避免op中进行创建和销毁
    添加对cuSparse中dense和sparse转换的API进行封装
    添加对封装的API的单测

* Fix the bug of batch_norm and batch_norm_grad op. Add the "roi_align" and "roi_align_grad" op in xpu2 op list.

* Fix the bug of batch_norm and batch_norm_grad op. Add the "roi_align" and "roi_align_grad" op in xpu2 op list. test=kunlun
Co-authored-by: NZibin <guozibin@baidu.com>

* add align for WorkQueue

* add spinlock

* merge develop

* merge

* Add EventsWaiter

* Revert "Add EventsWaiter"

This reverts commit e206173aa9be7401b83a53581627bfaf557c8fb2.

* update OS info

* split host_event_recorder

* split host_event_recorder

* update

* update

* update

* update

* update

* update

* update
Co-authored-by: Nliutiexing <liutiexing@google.com>

* add argsort/scatter for kunlun

* update test_scatter

* update xpu.cmake

* update xpu.cmake

* fix scatter

* add reduce_prod_xpu. fix reduce_mean_xpu bug.

* iadd reduce_prod_xpu. fix reduce_mean_xpu bug. test=kunlun

* add device-agnostic stream class

* add stream.h

* fix ut

* fix cpu compile

* First set of fixes

* - Make more likely to GetBlob find a blobs

* - Lint

* support external stream.

* update

* update

* update

* Get GPU BasePtr from CUDA allocation

* Fix compile error for ROCm

* Add BasePtr function for IPUPlace in naive_best_fit_allocator.cc

* Add alignment for BuddyAllocator

* Set address alignment of BuddyAllocator to 32 bytes

* Fix CI error

* Remove code for naive_best_fit strategy

* trainer_device fix and checknan tool for psgpu;test=develop

* disable show_one_table;test=develop

* add align for WorkQueue

* add spinlock

* merge develop

* merge

* Add EventsWaiter

* Revert "Add EventsWaiter"

This reverts commit e206173aa9be7401b83a53581627bfaf557c8fb2.

* add os_info

* update

* update

* update

* update

* update

* update for bugfix

* update

* update

* update
Co-authored-by: Nliutiexing <liutiexing@google.com>

* fix cuda atomicAdd for FP16

* try to fix ci

* add align for WorkQueue

* add spinlock

* merge develop

* merge

* Add EventsWaiter

* Revert "Add EventsWaiter"

This reverts commit e206173aa9be7401b83a53581627bfaf557c8fb2.

* Fix RecordEvent
Co-authored-by: Nliutiexing <liutiexing@google.com>

* fix CUDA Graph H2D bug again

* fix no return bug

* format xpu op list

* format xpu op list

* update xpu1 op list