By shell command

```bash
sed -i 's#ifdef PADDLE_ONLY_CPU#ifndef PADDLE_WITH_GPU#g' `find ./paddle/ -name '*.h' -o -name '*.cc' -o -name '*.cpp' -o -name '*.c' -o -name '*.cu'`
sed -i 's#ifndef PADDLE_ONLY_CPU#ifdef PADDLE_WITH_GPU#g' `find ./paddle/ -name '*.h' -o -name '*.cc' -o -name '*.cpp' -o -name '*.c' -o -name '*.cu'`
```

* Since GPU is an async device by default. We should sync computation
  when Python invoke `run`. So Python can get the correct computation
  result

* Also use `const DeviceContext&` all the time, to prevent `const_cast`

Fix #4169
Fix #3468
Fix #3475

* User can register OpKernel to its Ops. The OpKernelMap saved in
  OperatorWithKernel. Each Op which inherits OperatorWithKernel will
  use `OpKernel::Compute` instead of Run.

* add device_context

* add unittest for device_context

* transfer to use function paddle::platform::throw_on_error

* fix cuda build error

* using dynload functions

* follow comments