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

Issue 635965 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Last visit > 30 days ago
Closed: Aug 2016
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression

Blocked on:
issue 636310



Sign in to add a comment

33.2%-46.2% regression in v8.top_25_smooth at 410359:410398

Project Member Reported by rmcilroy@chromium.org, Aug 9 2016

Issue description

See the link to graphs below.
 

===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@410376  50.6611  2.21316  8  good
chromium@410382  50.8965  2.14248  8  good
chromium@410383  57.369   3.28462  5  bad
chromium@410384  56.6733  2.93591  8  bad
chromium@410385  59.1867  1.48646  5  bad
chromium@410387  56.83    3.15041  8  bad

Bisect job ran on: android_s5_perf_bisect
Bug ID: 635965

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests v8.key_mobile_sites_smooth
Test Metric: v8_gc_total_outside_idle/v8_gc_total_outside_idle
Relative Change: 11.96%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_s5_perf_bisect/builds/904
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004789132565261312


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5834817137016832

| 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!

===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@410376  50.6611  2.21316  8  good
chromium@410382  50.8965  2.14248  8  good
chromium@410383  57.369   3.28462  5  bad
chromium@410384  56.6733  2.93591  8  bad
chromium@410385  59.1867  1.48646  5  bad
chromium@410387  56.83    3.15041  8  bad

Bisect job ran on: android_s5_perf_bisect
Bug ID: 635965

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests v8.key_mobile_sites_smooth
Test Metric: v8_gc_total_outside_idle/v8_gc_total_outside_idle
Relative Change: 11.96%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_s5_perf_bisect/builds/904
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004789132565261312


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5834817137016832

| 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!
Blockedon: 636310
Cc: hpayer@chromium.org
Owner: hpayer@chromium.org
Another situation where the bisect isn't descending into the V8 roll.

Looks likely to be "[heap] Use smaller minimum allocation limit growing step when optimizing for memory usage." though. Assigning to Hannes for investigation.
Project Member

Comment 8 by 42576172...@developer.gserviceaccount.com, Aug 11 2016


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

Hi hpayer@chromium.org, 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 : [heap] Use smaller minimum allocation limit growing step when optimizing for memory usage.
Author  : hpayer
Commit description:
  
BUG= chromium:634900 

Review-Url: https://codereview.chromium.org/2223493002
Cr-Commit-Position: refs/heads/master@{#38435}
Commit  : caf5c5a194c67b6dae05f8fac8217eab25184bc4
Date    : Mon Aug 08 11:32:01 2016


===== TESTED REVISIONS =====
Revision                       Mean     Std Dev  N  Good?
chromium@410376                350.576  20.3859  5  good
chromium@410382                333.772  6.54493  5  good
chromium@410382,v8@79d9e18cbe  333.651  11.4342  5  good
chromium@410382,v8@caf5c5a194  415.702  13.7047  5  bad    <--
chromium@410382,v8@2ed9e6e634  406.256  17.2774  5  bad
chromium@410383                423.014  15.4521  5  bad
chromium@410384                415.378  14.6255  5  bad
chromium@410385                424.287  18.9582  5  bad
chromium@410387                424.733  22.7199  5  bad

Bisect job ran on: android_one_perf_bisect
Bug ID: 635965

Test Command: src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --also-run-disabled-tests v8.top_25_smooth
Test Metric: v8_gc_total_outside_idle/v8_gc_total_outside_idle
Relative Change: 21.15%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_one_perf_bisect/builds/1516
Job details: https://chromeperf.appspot.com/buildbucket_job_status/9004789132565261312


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5834817137016832

| 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!
Project Member

Comment 9 by bugdroid1@chromium.org, Aug 12 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/v8/v8.git/+/b992a8e1899091e3c180a050c1a63fa1e29886d5

commit b992a8e1899091e3c180a050c1a63fa1e29886d5
Author: hpayer <hpayer@chromium.org>
Date: Fri Aug 12 08:13:35 2016

Fix wrong state of IsolateInBackgroundNotification and IsolateInForegroundNotification.

BUG= chromium:636368 , chromium:635965 , chromium:634900 

Review-Url: https://codereview.chromium.org/2245483004
Cr-Commit-Position: refs/heads/master@{#38598}

[modify] https://crrev.com/b992a8e1899091e3c180a050c1a63fa1e29886d5/src/isolate.cc

With the CL in #9, all graphs should recover except Android One. There the regression is intentional.
Labels: TradeOff
Status: Fixed (was: Assigned)

Sign in to add a comment