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

Issue 792050 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner:
Closed: Dec 2017
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

4.4%-7.1% regression in thread_times.key_mobile_sites_smooth at 519889:520079

Project Member Reported by pmeenan@chromium.org, Dec 5 2017

Issue description

See the link to graphs below.
 
All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=792050

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=6257a27fb24bb8f4d708e76a36419c7496f7aafb7c391c7cb48b0ca887e30c4a


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

android-nexus5
android-nexus5X

=== BISECT JOB RESULTS ===
Perf regression found but unable to continue

Bisect was stopped because a commit couldn't be classified as either
good or bad.


Bisect Details
  Configuration: android_nexus5X_perf_bisect
  Benchmark    : thread_times.key_mobile_sites_smooth
  Metric       : thread_renderer_compositor_cpu_time_per_frame/http___www.reddit.com_r_programming_comments_1g96ve


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...www.reddit.com.r.programming.comments.1g96ve thread_times.key_mobile_sites_smooth

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8961033403867193264


For feedback, file a bug with component Speed>Bisection
Cc: dtapu...@chromium.org
Owner: dtapu...@chromium.org
Status: Assigned (was: Untriaged)

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

Hi dtapuska@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 : Dave Tapuska
  Commit : ff7f1749d10b1fce53591cae122639cf17848e53
  Date   : Wed Nov 29 04:13:44 2017
  Subject: Enable mojo input messages on all platforms other than Android webview.

Bisect Details
  Configuration: android_nexus5_perf_bisect
  Benchmark    : thread_times.key_mobile_sites_smooth
  Metric       : thread_renderer_compositor_cpu_time_per_frame/http___www.reddit.com_r_programming_comments_1g96ve

Revision             Result                    N
chromium@519945      2.14118 +- 0.023413       6      good
chromium@519978      2.15999 +- 0.044185       6      good
chromium@519988      2.13639 +- 0.0318712      6      good
chromium@519992      2.14563 +- 0.0240007      6      good
chromium@519995      2.14671 +- 0.0308824      6      good
chromium@519996      2.21356 +- 0.0199775      6      bad       <--
chromium@520013      2.22752 +- 0.026772       6      bad
chromium@520079      2.24098 +- 0.0367058      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 --story-filter=http...www.reddit.com.r.programming.comments.1g96ve thread_times.key_mobile_sites_smooth

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

Debug information about this bisect:
  https://chromeperf.appspot.com/buildbucket_job_status/8961022243350782144


For feedback, file a bug with component Speed>Bisection
 Issue 792049  has been merged into this issue.
Status: WontFix (was: Assigned)
This is mojo input messages getting enabled. The thread times increased on the compositor thread but decreased on the IO thread which is expected.

Only regressions were in thread timings not in latency so this is expected due to the work transitioning threads due to the way mojo works.
 Issue 792004  has been merged into this issue.

Sign in to add a comment