All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=907645

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=f55b10a3d040695c52b3121f1a577cc58fcb14a05230ff50892bf86c75f4ff99


Bot(s) for this bug's original alert(s):

linux-perf
mac-10_12_laptop_low_end-perf

system_health.common_desktop - Benchmark documentation link:
  https://bit.ly/system-health-benchmarks

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/1446a497e40000

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/1446a497e40000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/1608abf4140000

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/1608abf4140000

Added URLLoaderThrottleSafeBrowsingChecks to testing config by carlosil@chromium.org
https://chromium.googlesource.com/chromium/src/+/c3639baabf0af95d5eb4eff138c9cff56c8e9120
cpu_time_percentage: 0.9902 → 1.023 (+0.03239)

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Benchmark documentation link:
  https://bit.ly/system-health-benchmarks

From the pinpoint (and the initial graph) it looks like the regression is for about 0.03%, I'm not familiar with perf regressions, so I'm not sure if this is within the noise or something that needs to be reverted/addressed. FWIW this changes a check that was done via a ResourceThrottle, so it now uses a URLLoaderThrottle, so I don't think this should increase CPU usage unless URLLoaderThrottles use more CPU, but that seems external to this CL.