Fix Linux UTF-8 decoding

Author: juj

tools/shared.py

@@ -144,20 +144,20 @@ def get_finished_process():
             # then look again if any process has completed.
             try:
               out, err = processes[0][1].communicate(0.2)
-              return (0, out, err)
+              return (0, out.decode('UTF-8'), err.decode('UTF-8'))
             except subprocess.TimeoutExpired:
               pass
 
         j, out, err = get_finished_process()
         idx, finished_process = processes[j]
         del processes[j]
         if pipe_stdout:
-          std_outs += [(idx, out.decode('UTF-8'))]
+          std_outs += [(idx, out)]
         if check and finished_process.returncode != 0:
           if out:
-            logger.info(out.decode('UTF-8'))
+            logger.info(out)
           if err:
-            logger.error(err.decode('UTF-8'))
+            logger.error(err)
           raise Exception('Subprocess %d/%d failed with return code %d! (cmdline: %s)' % (idx + 1, len(commands), finished_process.returncode, shlex_join(commands[idx])))
         num_completed += 1