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

Issue 713119 link

Starred by 1 user
Status: WontFix
Owner:
jbroman@chromium.org
Closed: Apr 2017
Cc:
jbroman@chromium.org
█ alexclarke@chromium.org
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

5% regression in system_health.memory_mobile at 463382:463479

Project Member Reported by alexclarke@chromium.org, Apr 19 2017 
See the link to graphs below.

Comment 1 by alexclarke@chromium.org, Apr 19 2017 

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=713119

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg_I-5qwkM


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

android-nexus6
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Apr 19 2017

Cc: jbroman@chromium.org
Owner: jbroman@chromium.org

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

Hi jbroman@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 : jbroman
  Commit : 73b6feb516207f9d3578485f6624ab32378364c8
  Date   : Tue Apr 11 00:40:04 2017
  Subject: Hold script source as a simple string in ScheduledAction.

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:native_heap:proportional_resident_size_avg/background_search/background_search_google
  Change       : 2.42% | 44942044.0 -> 46030556.0

Revision             Result                   N
chromium@463381      44942044 +- 790888       6      good
chromium@463430      44944092 +- 481813       6      good
chromium@463455      44858076 +- 617936       6      good
chromium@463467      44891868 +- 318695       6      good
chromium@463468      46431281 +- 901447       6      bad       <--
chromium@463469      46008881 +- 1334542      6      bad
chromium@463470      46178183 +- 397910       6      bad
chromium@463473      45973383 +- 1535643      6      bad
chromium@463479      46030556 +- 876562       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=background.search.google system_health.memory_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981887991015228080

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5005075585433600


| 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!

Comment 4 by jbroman@chromium.org, Apr 19 2017 

Hmm, this bisect is suspicious. It looks solid, but I'm not sure how storing less data in ScheduledAction should increase memory usage.
Project Member

Comment 7 by 42576172...@developer.gserviceaccount.com, Apr 19 2017 


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

Suspected Commit
  Author : jbroman
  Commit : 73b6feb516207f9d3578485f6624ab32378364c8
  Date   : Tue Apr 11 00:40:04 2017
  Subject: Hold script source as a simple string in ScheduledAction.

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:native_heap:proportional_resident_size_avg/background_search/background_search_google
  Change       : 1.80% | 44994552.4444 -> 45930260.8889

Revision             Result                   N
chromium@463300      44994552 +- 884294       9       good
chromium@463400      45125413 +- 1445201      14      good
chromium@463450      44932145 +- 477172       6       good
chromium@463463      45112199 +- 866352       9       good
chromium@463466      44973617 +- 516585       6       good
chromium@463467      44981809 +- 837233       6       good
chromium@463468      46436060 +- 814031       6       bad       <--
chromium@463469      46319153 +- 509587       6       bad
chromium@463475      46391516 +- 1298773      6       bad
chromium@463500      45930261 +- 1400083      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.search.google system_health.memory_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981871837486568768

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6133213371564032


| 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!

Comment 8 by jbroman@chromium.org, Apr 19 2017

Status: Assigned (was: Untriaged)
Hrrrm. That's very mysterious. I hope it isn't something silly like "we're now just below a GC threshold".
Project Member

Comment 9 by 42576172...@developer.gserviceaccount.com, Apr 19 2017 


=== BISECT JOB RESULTS ===
NO Perf regression found, tests failed to produce values

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:native_heap:proportional_resident_size_avg/background_search/background_search_google

Revision             Result                  N
chromium@462500      44694017 +- 796552      21      good

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.search.google system_health.memory_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981872481238843456

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6344323362193408


| 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!
Project Member

Comment 11 by 42576172...@developer.gserviceaccount.com, Apr 19 2017 


=== BISECT JOB RESULTS ===
NO Perf regression found, tests failed to produce values

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:native_heap:proportional_resident_size_avg/background_search/background_search_google

Revision             Result                   N
chromium@462500      44888235 +- 2380274      21      good

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.search.google system_health.memory_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981857881174829248

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6344323362193408


| 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!
Project Member

Comment 13 by 42576172...@developer.gserviceaccount.com, Apr 21 2017 


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

Suspected Commit
  Author : jbroman
  Commit : 73b6feb516207f9d3578485f6624ab32378364c8
  Date   : Tue Apr 11 00:40:04 2017
  Subject: Hold script source as a simple string in ScheduledAction.

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:chrome:all_processes:reported_by_os:system_memory:native_heap:proportional_resident_size_avg/background_search/background_search_google
  Change       : 2.91% | 45024134.6667 -> 46285133.7778

Revision             Result                   N
chromium@463381      45024135 +- 1046007      9       good
chromium@463430      45206293 +- 2567692      9       good
chromium@463455      44872753 +- 679258       6       good
chromium@463467      45081674 +- 946163       14      good
chromium@463468      46079310 +- 1479523      9       bad       <--
chromium@463469      45881491 +- 2479001      14      bad
chromium@463470      45862181 +- 1937646      14      bad
chromium@463473      46212316 +- 592166       9       bad
chromium@463479      46285134 +- 527646       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.search.google system_health.memory_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981755010437296064

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6344323362193408


| 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!

Comment 14 by jbroman@chromium.org, Apr 27 2017

Status: WontFix (was: Assigned)
Closing as WontFix. The graph seems surprising (continual increases for half a dozen patches, then a gradual decrease?), and sending a revert patch to the same bot (25 repeats) shows that this patch's impact on memory for this metric is well within the noise: https://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/3281
►

Sign in to add a comment