All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=794313

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=dcfbf71db998ab0fe5c4d6880b7d12ca74e55312a2f9f859c4af9552f0c57dba


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

android-nexus5
android-nexus5X
android-nexus6
android-one

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8960374633736550544

=== Auto-CCing suspected CL author brianderson@chromium.org ===

Hi brianderson@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : Brian Anderson
  Commit : 242b9b0d6af44d087f29d3a1ed389242efcda5fc
  Date   : Wed Dec 06 21:06:26 2017
  Subject: cc:: Track timing of frame sources in LatencyInfo.

Bisect Details
  Configuration: android_nexus5_perf_bisect
  Benchmark    : thread_times.key_silk_cases
  Metric       : thread_browser_cpu_time_per_frame/silk_finance.html
  Change       : 4.02% | 2.280067313 -> 2.37164237447

Revision             Result                     N
chromium@522187      2.28007 +- 0.0523858       6      good
chromium@522190      2.2882 +- 0.0276666        6      good
chromium@522191      2.36301 +- 0.00903047      6      bad       <--
chromium@522192      2.36187 +- 0.0317181       6      bad
chromium@522196      2.37028 +- 0.0218946       6      bad
chromium@522204      2.37767 +- 0.0446171       6      bad
chromium@522221      2.34964 +- 0.0368948       6      bad
chromium@522254      2.36862 +- 0.0348853       6      bad
chromium@522320      2.37164 +- 0.0204598       6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=silk.finance.html thread_times.key_silk_cases

More information on addressing performance regressions:
  http://g.co/ChromePerformanceRegressions

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8960374633736550544


For feedback, file a bug with component Speed>Bisection

brianderson: Looks like your CL regressed the thread_times benchmark, can you PTAL?

Adding more components to LatencyInfo is expected to take more time, but the magnitude of these regressions makes me think a lighter weight LatencyInfo is more important than I initially thought.

Yup, there's a reason we want to clean it up!

This might actually be a reasonable case study for LatencyInfo design.

Want to try just adding the new components as fields of LatencyInfo instead of components, and seeing how different the performance characteristics are?

Maybe we could start over, with LatencyInfoV2: https://docs.google.com/document/d/1Gtaw7g7PgvSrTH2sbK-POslMrUYtJbcnMbk75JrBO20/edit#heading=h.tkgxrst8sjcb