Use correct overhead level for benchmarks

tools/perf/benchmarks/blob_storage.py

 # Copyright 2015 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.
 
 from core import perf_benchmark
 
 from telemetry import benchmark
 from telemetry.timeline import chrome_trace_category_filter
 from telemetry.web_perf import timeline_based_measurement
 
 import page_sets
 
 
 BLOB_CATEGORY = 'Blob'
 TIMELINE_REQUIRED_CATEGORY = 'blink.console'
 
 
 # http://crbug.com/499325
 # http://crbug.com/595069
 @benchmark.Disabled('android', 'win')
 class BlobStorage(perf_benchmark.PerfBenchmark):
   """Timeline based measurement benchmark for Blob Storage."""
 
   page_set = page_sets.BlobWorkshopPageSet
 
   def CreateTimelineBasedMeasurementOptions(self):
-    cat_filter = chrome_trace_category_filter.CreateMinimalOverheadFilter()
+    cat_filter = chrome_trace_category_filter.ChromeTraceCategoryFilter()
     cat_filter.AddIncludedCategory(BLOB_CATEGORY)
     cat_filter.AddIncludedCategory(TIMELINE_REQUIRED_CATEGORY)
 
     return timeline_based_measurement.Options(
         overhead_level=cat_filter)
 
   # Necessary until crbug.com/544306 is finished.
   def SetExtraBrowserOptions(self, options):
     options.AppendExtraBrowserArgs(
         ['--enable-experimental-web-platform-features'])
 
   @classmethod
   def Name(cls):
     return 'blob_storage.blob_storage'
 
   @classmethod
   def ValueCanBeAddedPredicate(cls, value, is_first_result):
     if ('blob-writes' not in value.name and
         'blob-reads' not in value.name):
       return False
     return value.values != None