* "add c++ side kernel selection"

* "add multiple kernel op test"

* "kernel selection only support cudnn"

* "better formatter"

* "small fix with UseCPU"

* "depends on change interface Get(Place, Library)"

* "fix CI"

* "fix python cudnn test"

* "leave the register cudnn op to another PR"

* "fix CI"

* "use all kernel by default"

* "fix CI"

* Add squared error layers doc

* Add doc for gru_unit

* Remove cdot which isn't supported

* Update layers.rst

* Update layers.rst (minor)

x_num_col_dims => num_flatten_dims

* "remove cudnndevicecontext"

* "remove unused init code"

* "fix hash functions"

* Addign document for fluid split_lod_tensor

* Adding document for fluid merge_lod_tensor