All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=763087

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=173146870304e4e57b96cf12ae7c6bc16b78161dbfb0d25aef5d08faf3d88d30


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

android-nexus5X
android-one
chromium-rel-mac12

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8969082248446770848

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

Hi xidachen@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 : Xida Chen
  Commit : ea91b206232bff046594f6f76b547e20c324cc37
  Date   : Wed Sep 06 01:51:20 2017
  Subject: Remove the content feature "ReducedSoftTileMemoryLimitOnLowEndAndroid"

Bisect Details
  Configuration: android_one_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_chrome:gpu:effective_size_avg/browse_tech/browse_tech_discourse_infinite_scroll
  Change       : 211.65% | 461880.0 -> 1439458.66667

Revision             Result                  N
chromium@499765      461880 +- 0.0           6      good
chromium@499822      461880 +- 0.0           6      good
chromium@499836      461880 +- 0.0           6      good
chromium@499843      461880 +- 0.0           9      good
chromium@499844      1379384 +- 1910124      7      bad       <--
chromium@499845      1393948 +- 1533536      9      bad
chromium@499847      1483149 +- 878545       6      bad
chromium@499850      1838136 +- 1071869      6      bad
chromium@499878      1439459 +- 1427133      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=browse.tech.discourse.infinite.scroll system_health.memory_mobile

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969082248446770848


For feedback, file a bug with component Speed>Bisection

tdresser@: it seems something is wrong with the bisect results. A while ago, I added the content feature to reduce the soft tile memory limit on Android devices with system RAM <= 512MB, which was designed for a finch experiment. The suspected CL here removes that feature because we are done with the experiment. I can understand the regression on Android one perf bot, such as this one:
https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg9aKr4ggM

We can see that there is a big perf improvement before 493956, which is likely because I added that content feature.

But I don't understand the regression for nexus5X and mac because the content feature doesn't apply.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8969008926264102704

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8969008845688822624

tdresser@: I kicked off another bisect job on Mac.

 Issue 763331  has been merged into this issue.

=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : Sunny Sachanandani
  Commit : 2eb3c8ac7ffc6840dc07f5e8312576e107eece23
  Date   : Wed Sep 06 02:01:21 2017
  Subject: cc: Flush after every tile task set completes.

Bisect Details
  Configuration: android_nexus5X_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:native_heap:proportional_resident_size_avg/background_news/background_news_nytimes
  Change       : 10.79% | 74465752.4444 -> 82500284.0

Revision             Result                    N
chromium@499616      74465752 +- 12388877      9       good
chromium@499747      74285528 +- 11897562      9       good
chromium@499812      72987096 +- 2262368       9       good
chromium@499829      74697331 +- 14644644      14      good
chromium@499837      73726481 +- 8831158       6       good
chromium@499841      72584892 +- 955467        6       good
chromium@499843      71778321 +- 579652        6       good
chromium@499844      73671697 +- 9455605       6       good
chromium@499845      82446581 +- 2549422       9       bad       <--
chromium@499877      82500284 +- 3154644       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 --story-filter=background.news.nytimes system_health.memory_mobile

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969008926264102704


For feedback, file a bug with component Speed>Bisection

=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : Sunny Sachanandani
  Commit : 2eb3c8ac7ffc6840dc07f5e8312576e107eece23
  Date   : Wed Sep 06 02:01:21 2017
  Subject: cc: Flush after every tile task set completes.

Bisect Details
  Configuration: mac_10_12_perf_bisect
  Benchmark    : system_health.memory_desktop
  Metric       : memory:chrome:all_processes:reported_by_chrome:malloc:effective_size_avg/browse_social/browse_social_twitter_infinite_scroll
  Change       : 9.76% | 107940628.667 -> 118476545.0

Revision             Result                     N
chromium@499748      107940629 +- 2558371       6       good
chromium@499802      108417650 +- 5601602       6       good
chromium@499829      109110038 +- 4169466       6       good
chromium@499842      109549850 +- 5220995       6       good
chromium@499844      111197293 +- 13747826      14      good
chromium@499845      116391515 +- 12035083      14      bad       <--
chromium@499846      117941772 +- 8045397       9       bad
chromium@499849      118071352 +- 8199392       6       bad
chromium@499855      118476545 +- 6258029       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=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=browse.social.twitter.infinite.scroll system_health.memory_desktop

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8969008845688822624


For feedback, file a bug with component Speed>Bisection