All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=675672

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgr5n-ugkM


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

chromium-rel-mac10

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8992821290116189968

=== PERF REGRESSION ===


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

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


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Remove SecurityChips experiment and flags
Author  : estark
Commit description:
  
We are launching security chips/verbose security states to stable in
M55, and thus removing the Finch experiments and command-line flags.

The behavior we are launching is show-all and animate-nonsecure-only
(security verbose states are shown for all states but animated for
non-secure only).

BUG=647762

Review-Url: https://codereview.chromium.org/2560203003
Cr-Commit-Position: refs/heads/master@{#438553}
Commit  : b6c0b5412248ac3001905d439cd9ebd4b2d0f7ec
Date    : Wed Dec 14 18:51:26 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@438527  40.6157  1.59194  6  good
chromium@438543  40.773   1.76466  6  good
chromium@438551  46.9247  34.2561  6  good
chromium@438552  47.4602  33.2507  6  good
chromium@438553  81.91    5.40308  6  bad    <--
chromium@438555  83.9515  14.9847  6  bad
chromium@438559  80.5543  5.03372  6  bad
chromium@438591  79.1949  44.5802  9  bad
chromium@438654  74.8572  36.5091  6  bad

Bisect job ran on: mac_10_10_perf_bisect
Bug ID: 675672

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.daum.net. page_cycler_v2.intl_ko_th_vi
Test Metric: timeToFirstContentfulPaint_avg/pcv1-warm/http___www.daum.net_
Relative Change: 84.31%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_10_perf_bisect/builds/2515
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8992821290116189968


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5827629964328960

| 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 Tests>AutoBisect.  Thank you!

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8992804197458381712

===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Remove SecurityChips experiment and flags
Author  : estark
Commit description:
  
We are launching security chips/verbose security states to stable in
M55, and thus removing the Finch experiments and command-line flags.

The behavior we are launching is show-all and animate-nonsecure-only
(security verbose states are shown for all states but animated for
non-secure only).

BUG=647762

Review-Url: https://codereview.chromium.org/2560203003
Cr-Commit-Position: refs/heads/master@{#438553}
Commit  : b6c0b5412248ac3001905d439cd9ebd4b2d0f7ec
Date    : Wed Dec 14 18:51:26 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N   Good?
chromium@438538  45.1286  39.984   14  good
chromium@438546  41.4188  1.39418  9   good
chromium@438550  41.625   2.95451  9   good
chromium@438552  41.1893  1.63121  6   good
chromium@438553  75.3374  54.3701  9   bad    <--
chromium@438568  77.349   82.1652  21  bad
chromium@438597  79.6311  59.9344  14  bad
chromium@438656  81.2674  44.1434  9   bad

Bisect job ran on: mac_10_11_perf_bisect
Bug ID: 675672

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.daum.net. page_cycler_v2.intl_ko_th_vi
Test Metric: timeToFirstContentfulPaint_avg/pcv1-warm/http___www.daum.net_
Relative Change: 74.54%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/mac_10_11_perf_bisect/builds/1135
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8992804197458381712


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5215539008372736

| 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 Tests>AutoBisect.  Thank you!

spqchan are you able to run this Mac benchmark locally? Kind of mystifying...

I've been looking into this and am hoping someone from the perf team might have some insight.

It's a little slow going because I can't reproduce locally, only on the bots, and not reliably there either.

First, the culprit CL enables by default behavior that was already enabled in fieldtrial_testing_config.json. So I'm flummoxed as to how it could have any noticeable effects whatsoever.

But if we put that aside for the moment, the interesting thing about http://wwww.daum.net (the regressed test) is that it has a password field on page load, which will cause the renderer to send an IPC to the browser [1] to animate in the "Not secure" omnibox warning. I don't really understand how that could cause a 94% regression to first contentful paint, but it must be related to that...

[1] https://cs.chromium.org/chromium/src/third_party/WebKit/Source/core/dom/Document.cpp?q=incrementpasswordcount&sq=package:chromium&l=6446

+kouhei, do you see anything in the traces that could help us understand what's going on here?

Didn't get a response from benchmark owners, and this is 8 months old and benchmark has been removed in the meantime; closing.