New issue
Advanced search Search tips
Note: Color blocks (like █ or █) mean that a user may not be available. Tooltip shows the reason.

Issue 685645 link

Starred by 2 users
Status: Fixed
Owner:
ericrk@chromium.org
Closed: Jan 2017
Cc:
█ pras...@chromium.org
ericrk@chromium.org
tdres...@chromium.org
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

1.5%-9345.1% regression in memory.top_10_mobile at 446136:446174

Project Member Reported by tdres...@chromium.org, Jan 26 2017 
See the link to graphs below.

Comment 1 by tdres...@chromium.org, Jan 26 2017 

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=685645

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsKLe6wkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsKLFpgoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0M_9oAkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsI7Z-wgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0NuN4AsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0MOg9wgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsOXuugkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsKLe6woM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0M_9oAoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0MXD8wsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsIq9uQsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg8OijpgoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0P2M8QoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsI6uqgkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsKGcqgsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0L_UlggM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsNaY5QkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsMHDvQsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0Nv_2AgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsJP8swoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsN3ImwgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0K-B5QsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsOud6wgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsIWW_QoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsKLFpggM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsLLOpQkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsMyT8QoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsIG89ggM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsKGIswsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0JuKggoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsNaY5QoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0OGi6gsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsNnBlQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0K-B5QkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0OGi6gkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0L2pjggM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsNL1vwkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg0M-E-wgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsOOPvQkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsPuNpQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsNipsAsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsMb25wsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg8MyBrAsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsL-QqQsM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsJOYogoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsJHvpwoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsL-QqQkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgsLTdjwoM


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

android-nexus5X
android-one

Comment 4 by tdres...@chromium.org, Jan 26 2017

Cc: ericrk@chromium.org
Nothing in the range the bisect found (https://chromium.googlesource.com/chromium/src/+log/6030ae0eb891f76ca0646bdca08ea6b04898f516%5E..4c759d5e287eda60b6ce8708ca40a22be6436800?pretty=fuller) looks at all plausible.

Eric, there's no way this is your revert, is there?
Project Member

Comment 6 by 42576172...@developer.gserviceaccount.com, Jan 26 2017

Owner: ericrk@chromium.org

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

Hi ericrk@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 : ericrk
  Commit : 4c759d5e287eda60b6ce8708ca40a22be6436800
  Date   : Wed Jan 25 23:20:21 2017
  Subject: Temporarily disable context idle cleanup.

Bisect Details
  Configuration: android_nexus5X_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_chrome:skia:effective_size_avg/foreground/https_mobile_twitter_com_justinbieber_skip_interstitial_true
  Change       : 9345.15% | 212864.0 -> 20105318.0

Revision             Result                   N
chromium@446135      212864 +- 0.0            6      good
chromium@446154      212864 +- 0.0            6      good
chromium@446159      212864 +- 0.0            6      good
chromium@446160      212864 +- 0.0            6      good
chromium@446161      19142246 +- 5274963      6      bad       <--
chromium@446162      20105318 +- 0.0          6      bad
chromium@446164      19142246 +- 5274963      6      bad
chromium@446173      20105318 +- 0.0          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=https.mobile.twitter.com.justinbieber.skip.interstitial.true memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8989390359643471792

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4994994789679104


| 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!
Project Member

Comment 7 by 42576172...@developer.gserviceaccount.com, Jan 26 2017 


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

Suspected Commit
  Author : ericrk
  Commit : 4c759d5e287eda60b6ce8708ca40a22be6436800
  Date   : Wed Jan 25 23:20:21 2017
  Subject: Temporarily disable context idle cleanup.

Bisect Details
  Configuration: android_nexus5X_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_chrome:skia:effective_size_avg/foreground/https_m_facebook_com_rihanna
  Change       : 2285.79% | 201720.0 -> 4812619.33333

Revision             Result                  N
chromium@446135      201720 +- 0.0           6      good
chromium@446154      201720 +- 0.0           6      good
chromium@446159      201720 +- 0.0           6      good
chromium@446160      201720 +- 0.0           6      good
chromium@446161      4813038 +- 0.0          6      bad       <--
chromium@446162      4813038 +- 0.0          6      bad
chromium@446164      4813038 +- 0.0          6      bad
chromium@446173      4812619 +- 2293.13      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=https.m.facebook.com.rihanna memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8989390290255282208

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5343947125161984


| 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!

Comment 8 by ericrk@chromium.org, Jan 26 2017 

This is my change - sorry, should have added a note to the commit. This is temporary to investigate an UMA regression, will revert by Friday.
Project Member

Comment 9 by 42576172...@developer.gserviceaccount.com, Jan 26 2017 


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

Suspected Commit
  Author : ericrk
  Commit : 4c759d5e287eda60b6ce8708ca40a22be6436800
  Date   : Wed Jan 25 23:20:21 2017
  Subject: Temporarily disable context idle cleanup.

Bisect Details
  Configuration: android_nexus5X_perf_bisect
  Benchmark    : memory.top_10_mobile
  Metric       : memory:chrome:all_processes:reported_by_chrome:skia:effective_size_avg/foreground/http_yandex_ru_touchsearch_text_science
  Change       : 3219.49% | 272216.0 -> 9036184.0

Revision             Result              N
chromium@446135      272216 +- 0.0       6      good
chromium@446154      272216 +- 0.0       6      good
chromium@446159      272216 +- 0.0       6      good
chromium@446160      272216 +- 0.0       6      good
chromium@446161      9036184 +- 0.0      6      bad       <--
chromium@446162      9036184 +- 0.0      6      bad
chromium@446164      9036184 +- 0.0      6      bad
chromium@446173      9036184 +- 0.0      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=http.yandex.ru.touchsearch.text.science memory.top_10_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8989384859278091056

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5329596297248768


| 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!

Comment 10 by pras...@chromium.org, Jan 27 2017

Cc: pras...@chromium.org
 Issue 686146  has been merged into this issue.

Comment 11 by pras...@chromium.org, Jan 30 2017

Status: Assigned (was: Untriaged)

Comment 12 by ericrk@chromium.org, Jan 31 2017

Status: Fixed (was: Assigned)
The change which caused this has been reverted.

Comment 13 by benhenry@chromium.org, Mar 11 2017

Labels: Performance-Memory
►

Sign in to add a comment