New issue
Advanced search Search tips

Issue 847942 link

Starred by 2 users

Issue metadata

Status: Started
Owner:
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

1.2%-1.4% regression in memory.top_10_mobile at 558835:560086

Project Member Reported by sullivan@chromium.org, May 30 2018

Issue description

See the link to graphs below.
 
Project Member

Comment 1 by 42576172...@developer.gserviceaccount.com, May 30 2018

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=847942

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=a6481806dfee04517b7d71b835d5671407e1c7fb3fa28b9d92d72c05082bb363


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

android-one
android-webview-nexus6
😿 Pinpoint job stopped with an error.
https://pinpoint-dot-chromeperf.appspot.com/job/169f247a240000

The swarming task 3de3dc491b246b10 expired. The bots are likely overloaded or dead, or may be misconfigured.
Project Member

Comment 5 by 42576172...@developer.gserviceaccount.com, Jun 13 2018

Cc: hajimehoshi@chromium.org foolip@chromium.org dmu...@chromium.org
Owner: dmu...@chromium.org
Status: Assigned (was: Untriaged)
📍 Found significant differences after each of 3 commits.
https://pinpoint-dot-chromeperf.appspot.com/job/149a3ba5240000

Roll in new version of WPT tools by foolip@chromium.org
https://chromium.googlesource.com/chromium/src/+/36670a32fb24c76cbba10511911f17339204d2c9
2.483e+07 → 2.478e+07 (-4.301e+04)

Make MainThreadSchedulerImpl's *TaskRunner() return TaskQueueWithTaskType by hajimehoshi@chromium.org
https://chromium.googlesource.com/chromium/src/+/c70cdb314e37ae58980dd993433748d9c09e64dc
2.478e+07 → 2.504e+07 (+2.593e+05)

[SessionStorageS13N] SessionStorage mojo context, and supporting fixes by dmurph@chromium.org
https://chromium.googlesource.com/chromium/src/+/f36bec395cfd4b4be3efb451810abd4fbbbad887
2.5e+07 → 2.509e+07 (+9.009e+04)

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

Comment 6 by sullivan@google.com, Jun 13 2018

Owner: hajimehoshi@chromium.org
hajimehoshi: looking at the pinpoint graph, this pretty clearly looks like "Make MainThreadSchedulerImpl's *TaskRunner() return TaskQueueWithTaskType"
Cc: altimin@chromium.org
Actually my CL added a very slight overhead, but I think we can ignore the overhead for the benefit of task type information (e.g. we can now know the ratio of task duration for such task runners).

altimin, wdyt?
Status: Started (was: Assigned)
Project Member

Comment 10 by bugdroid1@chromium.org, Jun 26 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/79a25b83e65b5d556e4393baec5487717909dfa4

commit 79a25b83e65b5d556e4393baec5487717909dfa4
Author: Hajime Hoshi <hajimehoshi@chromium.org>
Date: Tue Jun 26 10:59:34 2018

Avoid creation of TaskQueueWithTaskType for each call of MainThreadSchedulerImpl's functions

After https://chromium-review.googlesource.com/c/chromium/src/+/1063670,
memory usages on some devices were slightly increased since
MainThreadSchedulerImpl::DefaultTaskRunner() or other functions created
a TaskQueueWithTaskType object for each call unnecessarily.

This CL fixes this problem by creating necessary objects at
MainThreadSchedulerImpl constructor.

Bug: 847942
Change-Id: I738fc0ac91c49131ef8ecdb68910c46afef09d8b
Reviewed-on: https://chromium-review.googlesource.com/1114405
Commit-Queue: Hajime Hoshi <hajimehoshi@chromium.org>
Reviewed-by: Alexander Timin <altimin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#570370}
[modify] https://crrev.com/79a25b83e65b5d556e4393baec5487717909dfa4/third_party/blink/renderer/platform/scheduler/main_thread/main_thread_scheduler_impl.cc
[modify] https://crrev.com/79a25b83e65b5d556e4393baec5487717909dfa4/third_party/blink/renderer/platform/scheduler/main_thread/main_thread_scheduler_impl.h

Sign in to add a comment