New issue
Advanced search Search tips

Issue 733744 link

Starred by 1 user

Issue metadata

Status: Duplicate
Merged: issue 733705
Owner: ----
Closed: Jun 2017
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

2.5%-32% regression in thread_times.key_silk_cases at 479147:479269

Project Member Reported by majidvp@chromium.org, Jun 15 2017

Issue description

See the link to graphs below.
 

Comment 4 by ajuma@chromium.org, Jun 15 2017

Cc: sunn...@chromium.org
The metric that regressed on Nexus 5 is thread_GPU_cpu_time_per_frame.
One of the changes in the regression range (http://test-results.appspot.com/revision_range?start=479147&end=479249) is enabling the GPU scheduler (r479175).
Project Member

Comment 5 by 42576172...@developer.gserviceaccount.com, Jun 16 2017

Mergedinto: 733705
Status: Duplicate (was: Untriaged)

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

Suspected Commit
  Author : Sunny Sachanandani
  Commit : f9a0da96d26ec1fe2121a8eef62f7bd7e9330555
  Date   : Tue Jun 13 22:16:40 2017
  Subject: Enable gpu scheduler on waterfall and developer builds.

Bisect Details
  Configuration: android_nexus5_perf_bisect
  Benchmark    : thread_times.key_silk_cases
  Metric       : thread_GPU_cpu_time_per_frame/thread_GPU_cpu_time_per_frame
  Change       : 9.47% | 5.60204354697 -> 6.13249837416

Revision             Result                    N
chromium@479146      5.60204 +- 0.079595       6      good
chromium@479172      5.59573 +- 0.118739       6      good
chromium@479174      5.57233 +- 0.0744293      6      good
chromium@479175      6.05088 +- 0.0937384      6      bad       <--
chromium@479176      6.04032 +- 0.0977873      6      bad
chromium@479179      6.01229 +- 0.0966027      6      bad
chromium@479185      6.04477 +- 0.0923399      6      bad
chromium@479198      6.06424 +- 0.150443       6      bad
chromium@479249      6.1325 +- 0.11361         6      bad

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 thread_times.key_silk_cases

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976696547811590128

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5651558601064448


| 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 6 by 42576172...@developer.gserviceaccount.com, Jun 16 2017


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

Suspected Commit
  Author : Sunny Sachanandani
  Commit : f9a0da96d26ec1fe2121a8eef62f7bd7e9330555
  Date   : Tue Jun 13 22:16:40 2017
  Subject: Enable gpu scheduler on waterfall and developer builds.

Bisect Details
  Configuration: android_webview_nexus6_aosp_perf_bisect
  Benchmark    : system_health.memory_mobile
  Metric       : memory:webview:all_processes:reported_by_chrome:gpu:effective_size_avg/load_news/load_news_irctc
  Change       : 32.00% | 3276800.0 -> 4325376.0

Revision             Result                 N
chromium@479146      3276800 +- 0.0         6      good
chromium@479172      3276800 +- 0.0         6      good
chromium@479174      3276800 +- 0.0         6      good
chromium@479175      4500139 +- 957215      6      bad       <--
chromium@479176      4325376 +- 0.0         6      bad
chromium@479179      4325376 +- 0.0         6      bad
chromium@479185      4325376 +- 0.0         6      bad
chromium@479197      4325376 +- 0.0         6      bad
chromium@479247      4325376 +- 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-webview --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=load.news.irctc system_health.memory_mobile

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8976697103221547776

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6229677103906816


| 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!

Sign in to add a comment