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

Issue 788381 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner:
Closed: Jan 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

4.9% regression in system_health.common_desktop at 517943:518061

Project Member Reported by hjd@google.com, Nov 24 2017

Issue description

See the link to graphs below.
 
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, Nov 24 2017

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=788381

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=24e0f6d35122d321b5ea0f67bc79059b37663184a55d006972d1b8a48f6e46dc


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

chromium-rel-mac11-pro
Project Member

Comment 2 by 42576172...@developer.gserviceaccount.com, Nov 24 2017

๐Ÿ“ Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/1061ef07f80000
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Nov 24 2017

๐Ÿ“ Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/1061ef07f80000
Project Member

Comment 4 by 42576172...@developer.gserviceaccount.com, Jan 18 2018

๐Ÿ“ Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/129d37d0840000
Trying a wider bisect.
Project Member

Comment 6 by 42576172...@developer.gserviceaccount.com, Jan 19 2018

Cc: mlippautz@chromium.org u...@chromium.org
Owner: mlippautz@chromium.org
Status: Assigned (was: Untriaged)
๐Ÿ“ Found a significant difference after 1 commit.
https://pinpoint-dot-chromeperf.appspot.com/job/129d37d0840000

[heap] Re-enable parallel marking
By mlippautz@chromium.org ยท Mon Nov 20 13:03:03 2017
v8 @ a9cab08e6c38a11c370350fdc844e32c361179d9

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions
Cc: -mlippautz@chromium.org
Components: Blink>JavaScript>GC
Status: WontFix (was: Assigned)
Thanks for triaging.

This is expected when we have a benchmark that triggers a full garbage collection that is not preceded by incremental or concurrent marking. In this case the parallel marker will take over and and we see a regression in CPU time. Since this blocks the main thread that is fine.

(I observed this behavior locally on a few websites; unfortunately we don't have the V8 metrics in the trace file)

Sign in to add a comment