* simply C++ error stack once again, test=develop

* refactor code remove string pointer and recursive, test=develop

* fix multhead matmul's instable
test=develop

* fix multihead matmul bug
test=develop

* fix converage problem
test=develop
Co-authored-by: NZhaolong Xing <nhzlx.dragon@gmail.com>

* Fix dygraph grad bugs (#25781)

* fix double grad visitid unit; test=develop

* change name hash_pair to HashPair; test=develop

* follow comment; test=develop

* remove manual seed; test=develop

* change create_graph from True to False; test=develop

* Fix the double grad bug for the star gan. (#25655)

* update the retain_graph parameter doc. test=develop

* Add some error messages for the op without double grads.

* fix the test_imperative_double_grad UT.

* fix conflict

* fix conflict

* fix code

cherry pick encryption api
Co-authored-by: NYanghello <915769235@qq.com>
Co-authored-by: NYanghello <yangqingyou@baidu.com>

Fix the condition of concat dimension judgment.

* ignore warnings of external libraries, test=develop (#24193)

* fix repeat definitions in liengine.cc, test=develop (#25020)

* remove paddle_use_kernel and paddle_use_op. test=develop (#25189)

* fix compile for lite subgraph. test=develop (#25285)

* [CI] [Lite-Subgraph] CI add lite subgraph check. (#25346)

* supports xpu runtime, test=develop (#25554)

* fix cmake of lite, test=develop (#25680)

* change commit files, test=release/1.8
Co-authored-by: NWilber <jiweibo@baidu.com>

* fix softmax_with_cross_entropy cuda kernel overflow bug, test=develop

* replace old macro & for condition, test=develop

* polish details, test=develop

* Add default cudnn lib path (#25175)

* add default cudnn lib path, test=develop

* change default path in func, test=develop

* move to linux branch, test=develop

* fix var error in other plat, test=develop

* Refactor dynamic dso search functions (#25214)

* refactor dynamic dso search func, test=develop

* polish details, test=develop

* polish detail based review comments, test=develop

* revert string type change, test=develop

* Polish install error hint message (#25531)

* polish install error hint msg, test=develop

* fix variable error, test=develop

* polish hint messgae again

* Append error op hint for GradOpMaker (#24750)

* append error op hint for grad op maker, test=develop

* add unittests for coverage, test=develop

* append try-catch to opbase run, test=develop (#24870)

* add scale_x_y for yolo_box, test=develop

* add scale_x_y for yolo_box, test=develop

* refine eps in iou_similarity, test=develop

* fix softmax with cross entropy out of bound; test=develop (#25549)

* fix index error; test=develop

* fix concat shape error
test=develop

test=release/1.8

* test=release/1.8, cherry fix fluid.embedding

test=release/1.8

* fix bug of prelu when rank not equal 4, test=develop

* fix prelu inference, test=develop

* fix api, test=develop

* fix shape when mode is chennel, test=develop

* remove debug code, test=develop

* add unittest, test=develop

* fix beam_search infershape, test=develop

* fix beam search op unittest, test=develop

* cherry-pick #24977

* raise index error when slice out of range; test=develop

* add uni test; test=develop

* fix format error; test=develop

* add comment for py::index_error; test=develop

* polish error message; test=develop

* polish error message; test=develop

PR types: Bug fixes
PR changes:Ops
Describe:
Fix the broadcast bug in elementwise ops, when elementwise ops do not have fast gradient calculate way，use the default way to calculate.