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

Issue 631039 link

Starred by 1 user
Status: Duplicate
Merged: issue 495650
Owner: ----
Closed: Jul 2016
Cc:
danakj@chromium.org
ajuma@chromium.org
msramek@chromium.org
mpear...@chromium.org
█ ymalik@chromium.org
mark@chromium.org
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 2
Type: Bug



Sign in to add a comment

virtual/threaded/fast/scroll-behavior/ tests very flaky on Win 7 (dbg)

Project Member Reported by msramek@chromium.org, Jul 25 2016 
Win 7 (dbg):
https://build.chromium.org/p/chromium.webkit/builders/WebKit%20Win7%20%28dbg%29

Most of the 20 past builds were red. Almost all flaky tests are from the virtual/threaded/fast/scroll-behavior/ directory, e.g.:

virtual/threaded/fast/scroll-behavior/subframe-scrollBy.html
virtual/threaded/fast/scroll-behavior/main-frame-element-scroll.html
virtual/threaded/fast/scroll-behavior/main-frame-element-scrollBy.html
virtual/threaded/fast/scroll-behavior/main-frame-element-scrollTo.html
virtual/threaded/fast/scroll-behavior/main-frame-scroll.html
virtual/threaded/fast/scroll-behavior/overflow-scroll-scroll.html

failing in various subsets at the same time.
Project Member

Comment 1 by bugdroid1@chromium.org, Jul 25 2016 

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/d65439f03bfed76eb84c1f86bd3277879b6cdc2d

commit d65439f03bfed76eb84c1f86bd3277879b6cdc2d
Author: msramek <msramek@chromium.org>
Date: Mon Jul 25 14:58:27 2016

Disable the flaky virtual/threaded/fast/scroll-behavior/ tests on Win7

NOTRY=True
TBR=ymalik@chromium.org
BUG= 631039 

Review-Url: https://codereview.chromium.org/2175333002
Cr-Commit-Position: refs/heads/master@{#407476}

[modify] https://crrev.com/d65439f03bfed76eb84c1f86bd3277879b6cdc2d/third_party/WebKit/LayoutTests/TestExpectations

Comment 2 by msramek@chromium.org, Jul 25 2016

Cc: ymalik@chromium.org
Going 200 builds backwards on the bot, this seems to have always been flaky.

ymalik@, can you help triage / investigate?

Comment 3 by ymalik@chromium.org, Jul 25 2016

Cc: ajuma@chromium.org
@ajuma, looks like the failure is caused by this check failing: https://cs.chromium.org/chromium/src/cc/scheduler/scheduler.cc?rcl=0&l=160

Do you know what could be going on?

Comment 4 by mpear...@chromium.org, Jul 25 2016

Cc: danakj@chromium.org
This message is the same one discussed in  bug 495650  ("Browser test flakiness in SchedulerStateMachine::DidSwapBuffersComplete").  Perhaps this is a dup of that one?

Comment 5 by ajuma@chromium.org, Jul 25 2016 

Yes, it looks similar to me. Looking at the stack (https://storage.googleapis.com/chromium-layout-test-archives/WebKit_Win7__dbg_/6615/layout-test-results/virtual/threaded/fast/scroll-behavior/main-frame-element-scroll-stderr.txt), we're inside a call to DisplayScheduler::DrawAndSwap, we haven't yet called DidSwapBuffers for the current frame, then we run a callback (TestDelegatingOutputSurface::DrawCallback) that calls DidSwapBuffersComplete but there aren't any pending frames.

Comment 6 by danakj@chromium.org, Jul 25 2016

Mergedinto: 495650
Status: Duplicate (was: Available)

Comment 7 Deleted

►

Sign in to add a comment