All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=700085

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDglL_0-AgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDglN217woM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg1LLX7wgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg1Jy8owkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg1MSYvwoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDglMXj4QoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg1MiejgsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg1JKg-wkM


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

chromium-rel-mac-retina
chromium-rel-mac11
chromium-rel-mac11-pro
chromium-rel-mac12
chromium-rel-mac12-mini-8gb
chromium-rel-win10
chromium-rel-win7-x64-dual
chromium-rel-win8-dual

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8985574422891611520

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

Hi mlippautz@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 : mlippautz
  Commit : d8e518533a6a987997cdcf438ddc21573eb75de0
  Date   : Mon Mar 06 11:03:10 2017
  Subject: MutationObserver: Keep related nodes and callback alive through tracing

Bisect Details
  Configuration: win_x64_perf_bisect
  Benchmark    : blink_perf.shadow_dom
  Metric       : LargeDistributionWithoutLayout/LargeDistributionWithoutLayout
  Change       : 11.55% | 304.637043837 -> 269.463782451

Revision             Result                  N
chromium@454740      304.637 +- 3.58932      6      good
chromium@454833      307.577 +- 3.25561      6      good
chromium@454841      306.858 +- 2.3398       6      good
chromium@454845      307.337 +- 6.70194      6      good
chromium@454846      305.333 +- 5.92473      6      good
chromium@454847      268.505 +- 5.46367      6      bad       <--
chromium@454849      269.323 +- 1.91602      6      bad
chromium@454865      268.857 +- 2.49665      6      bad
chromium@454897      267.308 +- 4.2155       6      bad
chromium@454960      269.464 +- 4.34157      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.shadow_dom

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8985574422891611520

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5016390446088192


| 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!

Ack. This likely has to do with the swap() performance of HeapVector<TraceWrapperMember>. 

There's nothing we can do here for now. If we would change to looking up object starts through a bitmap then this one will likely disappear.

This also caused a significant (16-18%) regression in many of the querySelector benchmarks:
https://chromeperf.appspot.com/report?sid=35841db9cb36e48d7b35a4d6d1e13c1744995d492b4486361cf72c98a9a63389

This is lots of benchmarks that regressed not just this one.

Took a while but this is fixed now. cd39e5b0ea3c0da423e54066bd97571f82dffaad allowed us to use the fast swap() for HeapVectors again.

I just checked a few of the regressions and they all recovered again:
-  ChromiumPerf/linux-release/blink_perf.parser / query-selector-all-class-first
-  ChromiumPerf/linux-release/blink_perf.parser / query-selector-all-class
-  ChromiumPerf/chromium-rel-mac12-mini-8gb/blink_perf.parser / query-selector-all-class