All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=692946

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg2K3KowoM


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

android-one

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8987514784755399632

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

Hi verwaest@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 : verwaest
  Commit : d891b500535b1833d03f25e878bffede56ad3abc
  Date   : Mon Feb 13 18:49:23 2017
  Subject: [IC] Also use LoadNonexistentDH for keyed loads

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_chrome:v8:effective_size_avg/load_search/load_search_yahoo
  Change       : 8.86% | 5006144.0 -> 5449890.66667

Revision                           Result                  N
chromium@450107                    5006144 +- 236398       6      good
chromium@450173                    5014021 +- 344186       6      good
chromium@450206                    5006031 +- 357983       6      good
chromium@450222                    5040145 +- 301388       6      good
chromium@450226                    5053798 +- 293001       6      good
chromium@450227                    5021269 +- 346747       6      good
chromium@450227,v8@d891b50053      5488662 +- 64788.7      6      bad       <--
chromium@450227,v8@2dab40ccda      5460663 +- 86113.0      6      bad
chromium@450227,v8@673bbcbb9d      5458973 +- 75444.6      6      bad
chromium@450228                    5444203 +- 106601       6      bad
chromium@450230                    5457530 +- 118861       6      bad
chromium@450238                    5449891 +- 69656.4      6      bad

Please refer to the following doc on diagnosing memory regressions:
  https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

To Run This Test
  src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=load.search.yahoo system_health.memory_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8987514784755399632

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6079475122438144


| 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!

The Cl pointed out by the regression increases use of data handlers for something that's likely a dictionary mode object. Since we don't yet share complex data handlers, this causes an increase in memory usage.

We should look into reducing the overhead of complex data handlers, possibly by sharing them. At least the nonexistent case.

Issue 699462 has been merged into this issue.

 Issue 701769  has been merged into this issue.