All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=759487

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=e81f0d8d927333eceea017672ae9b691a016037f96efbad99844079295ffb8c5


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

chromium-rel-mac11
chromium-rel-mac11-pro

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8970027616188684720

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

Hi wangxianzhu@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 : Xianzhu Wang
  Commit : 27d732d9364e1f139dafdbacc813318b5a9a852f
  Date   : Fri Aug 18 23:18:12 2017
  Subject: Reland Devirtualize LayoutBoxModelObject logical orientation border/margin/padding accessors

Bisect Details
  Configuration: mac_10_11_perf_bisect
  Benchmark    : speedometer2
  Metric       : Inferno-TodoMVC/Speedometer2
  Change       : 2.56% | 983.828583333 -> 1009.03875

Revision             Result                  N
chromium@495668      983.829 +- 14.7053      6      good
chromium@495727      977.422 +- 5.7638       6      good
chromium@495735      979.403 +- 14.5894      6      good
chromium@495736      1003.48 +- 10.7163      6      bad       <--
chromium@495737      1019.89 +- 10.4716      6      bad
chromium@495739      1001.9 +- 11.3286       6      bad
chromium@495742      999.44 +- 9.78019       6      bad
chromium@495756      999.187 +- 13.9442      6      bad
chromium@495785      1009.04 +- 17.9863      6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=Speedometer2 speedometer2

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8970027616188684720


For feedback, file a bug with component Speed>Bisection

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14df783e780000

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8966647452416210256

📍 Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/14df783e780000

Reland Devirtualize LayoutBoxModelObject logical orientation border/margin/padding accessors
By wangxianzhu@chromium.org · Fri Aug 18 23:18:12 2017
chromium @ 27d732d9364e1f139dafdbacc813318b5a9a852f

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

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

Hi wangxianzhu@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 : Xianzhu Wang
  Commit : 27d732d9364e1f139dafdbacc813318b5a9a852f
  Date   : Fri Aug 18 23:18:12 2017
  Subject: Reland Devirtualize LayoutBoxModelObject logical orientation border/margin/padding accessors

Bisect Details
  Configuration: mac_10_11_perf_bisect
  Benchmark    : speedometer2
  Metric       : Inferno-TodoMVC/Speedometer2

Revision             Result                  N
chromium@495668      983.22 +- 6.93299       6      good
chromium@495727      979.609 +- 16.7987      6      good
chromium@495735      981.659 +- 7.03415      6      good
chromium@495736      1006.88 +- 11.4014      6      bad       <--
chromium@495737      1018.94 +- 11.239       6      bad
chromium@495739      1006.46 +- 13.4816      6      bad
chromium@495742      1006.39 +- 9.81364      6      bad
chromium@495756      998.547 +- 8.01985      6      bad
chromium@495785      998.784 +- 9.26209      6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=Speedometer2 speedometer2

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8966647452416210256


For feedback, file a bug with component Speed>Bisection

I think this is related to compiler optimization. The performance of the benchmarks has improved after this slight regression.