[Debugger] Sorting op-time breakdown for quicker analysis.

python/tvm/contrib/debugger/debug_result.py

@@ -228,6 +228,11 @@ def display_debug_result(self):
                 node_data = [name, op, time_us, time_percent, shape, inputs, outputs]
                 data.append(node_data)
                 eid += 1
+
+        # Sort on the basis of execution time. Prints the most expensive ops in the start.
+        data = sorted(data, key=lambda x: x[2], reverse=True)
+        # Insert a row for total time at the end.
+        data.append(["Total_time", "-", round(total_time * 1000000, 3), "-", "-", "-", "-", "-"])
         fmt = ""
         for i, _ in enumerate(header):
             max_len = len(header[i])