resort .cu headers, set clang-format not sort include block and consider .cu as main source file (#43633)

* [Phi] Migrate mode_op and mode_grad_op into Phi

* fix omp

* add ifdef

* migrate infershape

* modify according reviewer

* first commit

* cpu kernel

* first version

* fix compile error

* fix compile error

* delete v2

* fix

* fix

* add alias

* fix

* fix

* fix

* fix error

* fix

* fix

* fix

* fix format