* add some ops to train ssd on kunlun

* update test_fill_any_like_op_xpu.py

* merge conflict of paddle_gtest_main.cc

* modify FLAGS_npu_precision_mode and default not to call aclSetCompileopt

* Support npu kernel for fill_any_like op

* modify the description of exception

* remove useless template element

* remove useless decorator

* fix the code format error