All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=766561

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=28806631c3f736a66d6c8f09a5a927515c0caff04c6c73899f1632cefebfa2b4


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

chromium-rel-mac11-pro
chromium-rel-win10
chromium-rel-win7-gpu-ati

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8968030610777784352

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

Hi clamy@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 : clamy
  Commit : 0dbded05ad0ba1c3d8388fd649130509ab97f07a
  Date   : Fri Sep 15 14:44:08 2017
  Subject: Enable PlzNavigate on trunk

Bisect Details
  Configuration: winx64ati_perf_bisect
  Benchmark    : loading.desktop
  Metric       : cpuTimeToFirstMeaningfulPaint_avg/pcv1-warm/ja.wikipedia
  Change       : 57.48% | 96.8781111111 -> 152.4395

Revision             Result                  N
chromium@502243      96.8781 +- 7.26453      9      good
chromium@502249      96.2232 +- 2.32039      6      good
chromium@502250      95.671 +- 3.3133        6      good
chromium@502251      150.068 +- 4.70145      6      bad       <--
chromium@502252      148.943 +- 5.79555      6      bad
chromium@502254      169.373 +- 105.381      6      bad
chromium@502264      140.538 +- 60.0319      9      bad
chromium@502284      151.892 +- 3.52208      6      bad
chromium@502324      152.44 +- 5.34633       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 --story-filter=ja.wikipedia loading.desktop

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8968030610777784352


For feedback, file a bug with component Speed>Bisection

Yes this is likely caused by enabling PlzNavigate. However, the graphs linked show a regression for a specific page (wikipedia) but an improvement for all pages overall.
https://chromeperf.appspot.com/report?sid=7e360733d829f8d484f938fcff6cfc8385be694a536237a8025040b5889351e7&start_rev=497695&end_rev=502692
https://chromeperf.appspot.com/report?sid=cea0d11048189b54f094789abd2ae0811816cbc5032091c5b1575f92e9a5c987&start_rev=499265&end_rev=502675
https://chromeperf.appspot.com/report?sid=65e9271b3a5fa65b7cebd80ccf6abc596c4d4b824b398121a875e428f1028957&start_rev=497727&end_rev=502698

I suspect the issue on wikipedia is the same as the one behind b/f regression: we shoudl switch to a Mojo stream to commit faster.

The specific page that is regressing is wikipedia, and its TTFCP is ~100ms, making it very quick to load. We've had similar issues with fast loading b/f navigations, which can take longer when PlzNavigate is enabled because teh time to commit a navigation is longer.

 Issue 766560  has been merged into this issue.

 Issue 766567  has been merged into this issue.

 Issue 766563  has been merged into this issue.

 Issue 767725  has been merged into this issue.

WontFix-ing per explanation in #5