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

Issue 610254 link

Starred by 0 users

Issue metadata

Status: WontFix
Owner:
Closed: May 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: All
Pri: 2
Type: Bug-Regression



Sign in to add a comment

17.9%-27.4% regression in blink_perf.paint at 391831:391913

Project Member Reported by alexclarke@chromium.org, May 9 2016

Issue description

See the link to graphs below.
 
All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=610254

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgtLrSvQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg1IvxrwoM


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

android-nexus5X
android-nexus9
Cc: rob.b...@samsung.com
Owner: rob.b...@samsung.com

=== Auto-CCing suspected CL author rob.buis@samsung.com ===

Hi rob.buis@samsung.com, the bisect results pointed to your CL below as possibly
causing a regression. Please have a look at this info and see whether
your CL be related.


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Store separator info of counters as CSSStringValue
Author  : rob.buis
Commit description:
  
According to [1] the separator part of counters() is a <string>, so
prefer storing it as a CSSStringValue.

[1] https://drafts.csswg.org/css-lists-3/#funcdef-counter

Review-Url: https://codereview.chromium.org/1933713002
Cr-Commit-Position: refs/heads/master@{#391869}
Commit  : c58569e4a11e639da68e6442ce8e40bdc835ba00
Date    : Thu May 05 19:34:18 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@391840  110.167  5.40909  8  good
chromium@391859  116.517  5.0356   8  good
chromium@391868  113.689  2.94346  5  good
chromium@391869  99.3195  3.48388  5  bad    <--
chromium@391870  99.4714  2.95839  5  bad
chromium@391871  97.9807  1.84605  5  bad
chromium@391873  98.8925  3.06772  5  bad
chromium@391877  100.106  2.84679  5  bad

Bisect job ran on: android_nexus5X_perf_bisect
Bug ID: 610254

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests blink_perf.paint
Test Metric: large-table-background-change-with-invisible-collapsed-borders/large-table-background-change-with-invisible-collapsed-borders
Relative Change: 9.42%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5X_perf_bisect/builds/162
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9013151101539174384


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5234883148906496

| 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!
Owner: wangxianzhu@chromium.org
My change should not affect painting at all.
Because of the mentioned test I suspect https://codereview.chromium.org/1549693002.
@wangxianzhu can you have a look?
Components: Blink>Paint
Labels: OS-All
The CL improved performance of blink_perf.paint/large-table-background-change-with-visible-collapsed-borders, but regressed blink_perf.paint/large-table-background-change-with-invisible-collapsed-borders. Will profile.


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Store separator info of counters as CSSStringValue
Author  : rob.buis
Commit description:
  
According to [1] the separator part of counters() is a <string>, so
prefer storing it as a CSSStringValue.

[1] https://drafts.csswg.org/css-lists-3/#funcdef-counter

Review-Url: https://codereview.chromium.org/1933713002
Cr-Commit-Position: refs/heads/master@{#391869}
Commit  : c58569e4a11e639da68e6442ce8e40bdc835ba00
Date    : Thu May 05 19:34:18 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev   N  Good?
chromium@391830  104.933  2.45959   5  good
chromium@391854  107.934  6.5883    5  good
chromium@391866  108.185  7.98915   5  good
chromium@391868  110.634  1.73098   5  good
chromium@391869  96.6462  2.25769   5  bad    <--
chromium@391872  95.1682  0.799445  5  bad
chromium@391877  95.9096  1.13639   5  bad

Bisect job ran on: android_nexus5X_perf_bisect
Bug ID: 610254

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests blink_perf.paint
Test Metric: large-table-background-change-with-invisible-collapsed-borders/large-table-background-change-with-invisible-collapsed-borders
Relative Change: 8.60%
Score: 95.0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus5X_perf_bisect/builds/164
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9013115889753140576


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5801062246645760

| 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!
Status: WontFix (was: Assigned)
The performance of the "bad" versions is actually better than the "good" versions (the numbers are the lower the better).

Sign in to add a comment