Rewrite sparse dmatrix using callbacks. (#7092)

- Reduce dependency on dmlc parsers and provide an interface for users to load data by themselves.
- Remove use of threaded iterator and IO queue.
- Remove `page_size`.
- Make sure the number of pages in memory is bounded.
- Make sure the cache can not be violated.
- Provide an interface for internal algorithms to process data asynchronously.

tests/cpp/gbm/test_gbtree.cc

@@ -155,7 +155,8 @@ TEST(GBTree, ChoosePredictor) {
   ASSERT_TRUE(data.HostCanWrite());
 
   // pull data into device.
-  data = HostDeviceVector<Entry>(data.HostVector(), 0);
+  data.HostVector();
+  data.SetDevice(0);
   data.DeviceSpan();
   ASSERT_FALSE(data.HostCanWrite());