[Bisect] Account for tir_label-level summaries in parse_metric

scripts/slave/recipe_modules/auto_bisect/config_validation.py

 # Copyright 2016 The Chromium Authors. All rights reserved.
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
 import re
 
 # Note: this module is tested by a unit test config_validation_test.py,
 # rather than recipe simulation tests.
 
 _BISECT_CONFIG_SCHEMA = {
@@ skipping 88 matching lines @@
   except ValueError:
     return  # The revisions could be sha1 hashes.
   if earlier >= later:
     raise ValidationFail('Order of good_revision (%d) and bad_revision(%d) '
                          'is reversed.' % (earlier, later))
 
 
 def _validate_metric(bisect_mode, metric):  # pragma: no cover
   if bisect_mode not in ('mean', 'std_dev'):
     return
-  if not (isinstance(metric, basestring) and metric.count('/') == 1):
+  if not (isinstance(metric, basestring) and ((metric.count('/') == 1) or
+                                              (metric.count('/') == 2))):
     raise ValidationFail('Invalid value for "metric": %s' % metric)