All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=768048

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=1e1451ac560a48f75eb060731b49fcf88a627d4078a2a062d76a96c4b5d5298e


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

chromium-rel-mac-retina
chromium-rel-mac12
chromium-rel-mac12-mini-8gb
chromium-rel-win7-x64-dual

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8967713329278604800

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

Bisect Details
  Configuration: mac_10_12_perf_bisect
  Benchmark    : loading.desktop
  Metric       : timeToFirstMeaningfulPaint_avg/warm/Free.fr

Revision             Result                  N
chromium@503183      319.932 +- 479.547      21      good
chromium@503261      326.071 +- 561.596      21      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=Free.fr loading.desktop

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8967713329278604800


For feedback, file a bug with component Speed>Bisection

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8966722389905735504

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12ac38b6780000

Kicked two more bisects because I think the graph show a clear regression despite previous result.

P.S: YAY , my first pinpoint job \o/, let's see how it will do :P

📍 Pinpoint job completed.
https://pinpoint-dot-chromeperf.appspot.com/job/12ac38b6780000

Couldn't reproduce a difference.

=== BISECT JOB RESULTS ===
Perf regression found but unable to narrow commit range

Build failures prevented the bisect from narrowing the range further.


Bisect Details
  Configuration: mac_retina_perf_bisect
  Benchmark    : loading.desktop
  Metric       : timeToFirstMeaningfulPaint_avg/cold/Free.fr
  Change       : 20.11% | 337.012357143 -> 354.77475

Suspected Commit Range
  10 commits in range
  https://chromium.googlesource.com/chromium/src/+log/c887f0ac8126dcda89ec57dfec130faa030ae144..953c88223dbf2b2a0fec6f0cbf229492f9bade2b


Revision             Result                  N
chromium@503109      337.012 +- 281.227      14       good
chromium@503118      334.619 +- 181.297      21       good
chromium@503119      ---                     ---      build failure
chromium@503120      ---                     ---      build failure
chromium@503121      ---                     ---      build failure
chromium@503122      ---                     ---      build failure
chromium@503123      ---                     ---      build failure
chromium@503124      ---                     ---      build failure
chromium@503125      ---                     ---      build failure
chromium@503126      ---                     ---      build failure
chromium@503127      ---                     ---      build failure
chromium@503128      369.567 +- 199.623      14       bad
chromium@503147      338.812 +- 127.975      9        bad
chromium@503185      354.775 +- 124.831      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=Free.fr loading.desktop

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8966722389905735504


For feedback, file a bug with component Speed>Bisection

I manually scanned the range. THe only plausible CL in that 10 CL range seems 

https://chromium.googlesource.com/chromium/src/+/b2ca65cb5a61eb6f69a573a336cbd1f9ea794fcd

All the others are tests or iOS stuff.

My CL only affects pages with ES6 module scripts (which is a recent new feature), so I feel it's unlikely to affect tests existing since before (I haven't checked whether the test really uses ES6 modules or not though).

Also, according to Comment #8, the fluctuation is large and the performance change looks within fluctuation.

Bisect couldn't reproduce and this test fluctuates a lot.