All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=732315

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg5tXgqwoM


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

android-one

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8977001689284578352

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

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:java_heap:proportional_resident_size_avg/background/after_https_mobile_twitter_com_justinbieber_skip_interstitial_true

Revision             Result                 N
chromium@476608      5602109 +- 238351      21      good
chromium@476640      5594990 +- 280248      21      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 memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8977001689284578352

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6694552821301248


| 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!

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976981851972255184

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976981688121749008

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

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:java_heap:proportional_resident_size_avg/background/after_https_mobile_twitter_com_justinbieber_skip_interstitial_true

Revision             Result                 N
chromium@476608      5579337 +- 251836      21      good
chromium@476640      5573876 +- 264664      21      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 memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976981688121749008

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6694552821301248


| 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!

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

Hi estade@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 : Evan Stade
  Commit : 90cf9d727da221f76e5cb7d34aa09c8e711c1e68
  Date   : Fri Jun 02 00:37:26 2017
  Subject: Clean up views::Background.

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:java_heap:proportional_resident_size_avg/background/after_https_mobile_twitter_com_justinbieber_skip_interstitial_true
  Change       : 1.29% | 5563392.0 -> 5636681.14286

Revision             Result                  N
chromium@476500      5563392 +- 140523       14      good
chromium@476501      5681493 +- 72039.9      6       bad       <--
chromium@476502      5679787 +- 32564.7      6       bad
chromium@476504      5641216 +- 112975       6       bad
chromium@476507      5636608 +- 156151       6       bad
chromium@476513      5656235 +- 163254       6       bad
chromium@476525      5655552 +- 106693       6       bad
chromium@476550      5624718 +- 220460       9       bad
chromium@476601      5632683 +- 227641       9       bad
chromium@476700      5636681 +- 143438       14      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 memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976981851972255184

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5687106636087296


| 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!

I feel very skeptical that my CL could cause this. It should have had zero impact on android.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976896989357221504

=== BISECT JOB RESULTS ===
Bisect failed for unknown reasons

Please contact the team (see below) and report the error.


Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:java_heap:proportional_resident_size_avg/background/after_https_mobile_twitter_com_justinbieber_skip_interstitial_true


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 memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976896989357221504

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5800070315769856


| 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!

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976879737827416192

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

Hi nainar@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 : nainar
  Commit : e6cbdad3d4c498263f2ab30a2e193e3a634afb02
  Date   : Fri Jun 02 08:38:01 2017
  Subject: Remove superfluous diff code in ComputedStyle::UpdatePropertySpecificDifferences

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:java_heap:proportional_resident_size_avg/background/after_https_mobile_twitter_com_justinbieber_skip_interstitial_true
  Change       : 1.07% | 5576704.0 -> 5640850.28571

Revision             Result                  N
chromium@476500      5576704 +- 127462       9       good
chromium@476600      5570706 +- 74700.9      14      good
chromium@476602      5680981 +- 53726.7      6       good
chromium@476603      5820245 +- 166987       6       bad       <--
chromium@476604      5896021 +- 94933.3      6       bad
chromium@476607      5692416 +- 93032.0      6       bad
chromium@476613      5664597 +- 68177.4      6       bad
chromium@476625      5706069 +- 29772.4      6       bad
chromium@476651      5684322 +- 311750       21      bad
chromium@476700      5640850 +- 140453       14      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 memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976879737827416192

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5800070315769856


| 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!

Sorry don't think its me either. I removed dead code. 

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8976828357235679312

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

Hi chengx@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 : chengx
  Commit : 61e622b8b398999d4884231999375245c07dad01
  Date   : Wed May 31 20:59:27 2017
  Subject: Cancel the JumpList update if top 5 most visited URLs are unchanged

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:java_heap:proportional_resident_size_avg/background/after_https_mobile_twitter_com_justinbieber_skip_interstitial_true
  Change       : 1.05% | 5555565.71429 -> 5620280.88889

Revision             Result                  N
chromium@476000      5555566 +- 167796       14      good
chromium@476016      5632512 +- 67081.8      6       good
chromium@476017      5834240 +- 133039       6       bad       <--
chromium@476018      5748736 +- 67163.8      6       bad
chromium@476020      5728597 +- 58335.1      6       bad
chromium@476024      5744299 +- 74557.7      6       bad
chromium@476032      5743275 +- 79129.0      6       bad
chromium@476063      5647189 +- 71481.0      6       bad
chromium@476125      5648725 +- 174080       9       bad
chromium@476250      5658917 +- 334925       21      bad
chromium@476500      5629806 +- 259954       14      bad
chromium@477000      5620281 +- 154343       9       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 memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976828357235679312

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5315162266927104


| 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!

My CL is Windows specific, so I don't think it can cause any regression on Android. 

 Issue 729521  has been merged into this issue.

I think this is issue 724032 again, as the bisect is on java_heap and the number an all three bisect results above look like going up over time.

+primiano, +hjd FYI

I'm going to increase the priority of that bug, we'll need to find a solution, otherwise our bisects on that metric will all be hit or miss.

As to the cause of the regression, I'm suspecting  issue 729521 .

ulan, hablich: can you help triage while hpayer is ooo?

Java heap should be completely independent form the JavaScript heap. I'd be surprised if Hannes' CL affects Java heap.

1) Noise/regression ratio seems high since bisect bot is finding different culprits.
2) I checked couple of graphs and they recovered at TOT.

Should we won't fix this based on 1 and 2?

Thanks, Ulan. WontFix-ing sgtm.