New issue
Advanced search Search tips

Issue 795780 link

Starred by 1 user

Issue metadata

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



Sign in to add a comment

59.8% regression in system_health.common_desktop at 524359:524389

Project Member Reported by chiniforooshan@chromium.org, Dec 18 2017

Issue description

See the link to graphs below.
 
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=795780

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=589cd38a281acfa40d4ebf3a5edf5f16f3e7bdb931376c536ab20429c7dd0b70


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

linux-release
Project Member

Comment 2 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

๐Ÿ“ Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/168a4dc6040000
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

๐Ÿ“ Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/168a4dc6040000
Project Member

Comment 4 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

๐Ÿ“ Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/16f441a6040000
Project Member

Comment 5 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

Cc: thomasanderson@chromium.org dnicoara@chromium.org dpranke@chromium.org
Owner: thomasanderson@chromium.org
Status: Assigned (was: Untriaged)
๐Ÿ“ Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/16f441a6040000

Statically link fontconfig on Linux
By thomasanderson@chromium.org ยท Fri Dec 15 15:53:31 2017
chromium @ 4f1a60552748d150783655ee6f7b6f7084e3af95

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions
Cc: sullivan@chromium.org
+sullivan@ addressing_performance_regressions.md says that the bisect bot should have posted a message on this bug with instructions on how to reproduce the test, but it's missing in this case.  Which test do I need to run?
I think the documentation is stale and should be updated.

Meanwhile, the pinpoint job (https://pinpoint-dot-chromeperf.appspot.com/job/16f441a6040000) has this information about the failure:

benchmark: system_health.common_desktop
story: browse.search.google.india
configuration: linux-release

To run the test locally, use:

src/tools/perf/run_benchmark -v --browser=release --output-format=html --pageset-repeat=1 --also-run-disabled-tests --story-filter=browse.search.google.india system_health.common_desktop --output-dir=<SOME DIR>

Project Member

Comment 8 by 42576172...@developer.gserviceaccount.com, Dec 18 2017

๐Ÿ“ Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/15df5476040000
Thanks chiniforooshan@.  I ran the tests locally and got the attached results.  Runs 2 and 3 are without the CL, and 4 and 5 are with the CL applied.  I don't see a significant difference in the timings though.
results.tar.gz
8.1 MB Download
Project Member

Comment 10 by 42576172...@developer.gserviceaccount.com, Dec 19 2017

๐Ÿ“ Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/15df5476040000

Statically link fontconfig on Linux
By thomasanderson@chromium.org ยท Fri Dec 15 15:53:31 2017
chromium @ 4f1a60552748d150783655ee6f7b6f7084e3af95

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions
Oh, sorry. If you want to try with local changes, instead of --browser=release please use --browser=exact --browser-executable=<path to your locally built chrome binary>.
Project Member

Comment 12 by 42576172...@developer.gserviceaccount.com, Dec 22 2017

๐Ÿ“ Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/14fc9d31040000
Project Member

Comment 13 by 42576172...@developer.gserviceaccount.com, Dec 22 2017

Cc: drott@chromium.org kojii@chromium.org e...@chromium.org
Owner: drott@chromium.org
๐Ÿ“ Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/14fc9d31040000

Roll HarfBuzz to 1.7.3 plus hb_set fix
By drott@chromium.org ยท Wed Dec 20 13:26:18 2017
chromium @ a14e00ae9bc6a9263bbfb249ec51b4ed6a7635c1

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions
Status: Fixed (was: Assigned)
Looks like things went back to normal after the harfbuzz roll (also the CL was reverted before the hb roll, but pinpoint didn't identify that as the source of the speedup)

Sign in to add a comment