[breaking] Remove the predictor param, allow fallback to prediction using DMatrix. (#9129)

- A `DeviceOrd` struct is implemented to indicate the device. It will eventually replace the `gpu_id` parameter.
- The `predictor` parameter is removed.
- Fallback to `DMatrix` when `inplace_predict` is not available.
- The heuristic for choosing a predictor is only used during training.

doc/tutorials/dask.rst

@@ -148,8 +148,8 @@ Also for inplace prediction:
 
 .. code-block:: python
 
-  booster.set_param({'predictor': 'gpu_predictor'})
-  # where X is a dask DataFrame or dask Array containing cupy or cuDF backed data.
+  # where X is a dask DataFrame or dask Array backed by cupy or cuDF.
+  booster.set_param({"gpu_id": "0"})
   prediction = xgb.dask.inplace_predict(client, booster, X)
 
 When input is ``da.Array`` object, output is always ``da.Array``.  However, if the input

doc/tutorials/saving_model.rst

@@ -173,7 +173,6 @@ Will print out something similar to (not actual output as it's too long for demo
         "gradient_booster": {
           "gbtree_train_param": {
             "num_parallel_tree": "1",
-            "predictor": "gpu_predictor",
             "process_type": "default",
             "tree_method": "gpu_hist",
             "updater": "grow_gpu_hist",