[Performance] Make IdHashMap parallel (#4881)

* make IdHashMap parallel

* fix

* Update array_utils.h

src/array/cpu/array_utils.h

@@ -7,6 +7,7 @@
 #define DGL_ARRAY_CPU_ARRAY_UTILS_H_
 
 #include <dgl/aten/types.h>
+#include <dgl/runtime/parallel_for.h>
 #include <parallel_hashmap/phmap.h>
 
 #include <unordered_map>
@@ -81,9 +82,13 @@ class IdHashMap {
     const IdType* ids_data = static_cast<IdType*>(ids->data);
     const int64_t len = ids->shape[0];
     IdArray values = NewIdArray(len, ids->ctx, ids->dtype.bits);
-    IdType* values_data = static_cast<IdType*>(values->data);
-    for (int64_t i = 0; i < len; ++i)
-      values_data[i] = Map(ids_data[i], default_val);
+    IdType* values_data = values.Ptr<IdType>();
+    runtime::parallel_for(
+        0, len, 1000, [=] (size_t begin, size_t end) {
+          for (size_t i = begin; i < end; ++i) {
+            values_data[i] = Map(ids_data[i], default_val);
+          }
+        });
     return values;
   }