All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=722867

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgmoiSuAoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg6pTAygkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg6r33pAkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg6oD3twgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgqs2J6woM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgmqKrsQkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg6o2hmwoM


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

chromium-rel-win7-gpu-ati
linux-release
win-high-dpi

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8979425435148761680

=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: winx64ati_perf_bisect
  Benchmark    : blink_perf.layout
  Metric       : floats_100_100/floats_100_100

Revision             Result                  N
chromium@471247      93.7155 +- 6.12544      21      good
chromium@471290      94.8575 +- 8.54434      21      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests blink_perf.layout

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8979425435148761680

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4834083458252800


| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Speed>Bisection.  Thank you!

Looks like the ref build jumped too. probably a change in the test?

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8979401386486749072

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

Hi maxmorin@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 : Max Morin
  Commit : a50d914e18c636fb8a37bf622ef8bd8728c9d2b7
  Date   : Fri May 12 08:35:02 2017
  Subject: Revert "[reland] Support generation of component targets for Mojom C++ bindings"

Bisect Details
  Configuration: winx64ati_perf_bisect
  Benchmark    : blink_perf.layout
  Metric       : floats_100_100/floats_100_100
  Change       : 3.04% | 91.9491666667 -> 94.7458333333

Revision             Result                   N
chromium@471247      91.9492 +- 0.447257      6      good
chromium@471249      92.5223 +- 1.13489       6      good
chromium@471250      94.9545 +- 0.880077      6      bad       <--
chromium@471253      93.299 +- 1.64826        6      bad
chromium@471258      95.3053 +- 2.20293       6      bad
chromium@471269      93.1587 +- 0.533122      6      bad
chromium@471290      94.7458 +- 2.96992       6      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release_x64 --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests blink_perf.layout

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8979401386486749072

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4834083458252800


| O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq
|  X  | for more information addressing perf regression bugs. For feedback,
| / \ | file a bug with component Speed>Bisection.  Thank you!

That CL (#471250) is a revert of CL #471238, but looking at the graphs it looks like the rise corresponds to a dip at about ~469150, so I don't think my revert is related.