* move operators/math/math_function_* to pten/kernels/func
* namespace from `paddle::operators::math` to `pten::funcs`

* improve backward performance

* support different dtypes for elementwise ops

* Merged LoDTensor with Tensor,test=allcases

* Patched python level LoDTensor

* Patched python level LoDTensor

* Merge Tensor into DenseTensor

* Fixed namespace issues,test=allcases

* Fixed merge issues

* Fixed inference issues

* Fixed NPU test issues

* Fixed merge issues

* Add the transformop parameter in TensorReduceFunctorImpl

* add zero norm, inf norm support for p_norm op

* fix the invalid argument check, fix the dtype problem in test case.

fix invalid read of pnorm gradient function and delete the unused code

Fix the compile problem in norm op in the windows env, add delete some invalid argument 

* Add new norm api, support frobenius norm and p-order vector norm. test==develop
* combine test files, add more attr checks. test=develop

* Implement norm_op by CUDA instead of Eigen.
* Remove the commented code.

* Add normalization operator.
1. Refine the raw norm_op and let it more general to support to normalize Tensor along any axis.
2. There is a bug in l2_normalize API, which lacks sqrt after `reduce_sum`.
3. Use norm_op to refine the l2_normalize API.
4. Fix bug in test_normalization_wrapper.py.

There are mainly following fixes:

- take `DeviceContext` as the template parameter of math functors and OpKernel instead of `Place`
- remove `eigen_device` interface in base class  `DeviceContext`
- remove `GetEigenDevice` interface in `ExecutionContext` and base class `DeviceContext`
- remove unused `platform::EigenDeviceConverter`
- rename `REGISTER_OP_GPU_KERNEL` to `REGISTER_OP_CUDA_KERNEL`
- rename `USE_GPU_ONLY_OP` to `USE_CUDA_ONLY_OP`