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

Issue 790877 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: May 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug

Blocking:
issue 703263



Sign in to add a comment

LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame is flaky

Project Member Reported by xing...@intel.com, Dec 1 2017

Issue description

Ubuntu 16.04 (Didn't test this on other platform)
LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunMultiThread_DelegatingRenderer

LOG:
[----------] 1 test from LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame
[ RUN      ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunMultiThread_DelegatingRenderer
../../cc/trees/layer_tree_host_unittest.cc:5924: Failure
Value of: TestEnded()
  Actual: true
Expected: false
../../cc/trees/layer_tree_host_unittest.cc:5924: Failure
Value of: TestEnded()
  Actual: true
Expected: false
../../cc/trees/layer_tree_host_unittest.cc:5924: Failure
Value of: TestEnded()
  Actual: true
Expected: false
../../cc/trees/layer_tree_host_unittest.cc:5944: Failure
      Expected: will_begin_impl_frame_count_
      Which is: 10
To be equal to: did_finish_impl_frame_count_
      Which is: 9
[  FAILED  ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunMultiThread_DelegatingRenderer (80 ms)
[----------] 1 test from LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame (80 ms total)

[----------] Global test environment tear-down
[==========] 1 test from 1 test case ran. (80 ms total)
[  PASSED  ] 0 tests.
[  FAILED  ] 1 test, listed below:
[  FAILED  ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunMultiThread_DelegatingRenderer

 1 FAILED TEST
[1477:1478:1201/100954.400942:956752809732:ERROR:kill_posix.cc(84)] Unable to terminate process group 1479: No such process (3)
[1/1] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunMultiThread_DelegatingRenderer (80 ms)
1 test failed:
    LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunMultiThread_DelegatingRenderer (../../cc/trees/layer_tree_host_unittest.cc:5961)
Tests took 0 seconds.
 
Components: Internals>Compositing
Status: Available (was: Untriaged)
-untriaged +available.

How flaky is this test? Can it be reliably reproduced within a few retries?
FindIt is currently complaining that the SingleThread version of this test is also flaky:

https://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer&testType=cc_unittests

The failing output seems to consistently be:

[ RUN      ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer
../../cc/trees/layer_tree_host_unittest.cc(5960): error: Expected equality of these values:
  will_begin_impl_frame_count_
    Which is: 11
  kExpectedNumImplFrames
    Which is: 10
../../cc/trees/layer_tree_host_unittest.cc(5961): error: Expected equality of these values:
  did_finish_impl_frame_count_
    Which is: 11
  kExpectedNumImplFrames
    Which is: 10
[  FAILED  ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer (69 ms)
[ RUN      ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer
../../cc/trees/layer_tree_host_unittest.cc(5960): error: Expected equality of these values:
  will_begin_impl_frame_count_
    Which is: 11
  kExpectedNumImplFrames
    Which is: 10
../../cc/trees/layer_tree_host_unittest.cc(5961): error: Expected equality of these values:
  did_finish_impl_frame_count_
    Which is: 11
  kExpectedNumImplFrames
    Which is: 10
[  FAILED  ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer (66 ms)
[ RUN      ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer
../../cc/trees/layer_tree_host_unittest.cc(5960): error: Expected equality of these values:
  will_begin_impl_frame_count_
    Which is: 11
  kExpectedNumImplFrames
    Which is: 10
../../cc/trees/layer_tree_host_unittest.cc(5961): error: Expected equality of these values:
  did_finish_impl_frame_count_
    Which is: 11
  kExpectedNumImplFrames
    Which is: 10
[  FAILED  ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer (68 ms)
[ RUN      ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer
../../cc/trees/layer_tree_host_unittest.cc(5960): error: Expected equality of these values:
  will_begin_impl_frame_count_
    Which is: 11
  kExpectedNumImplFrames
    Which is: 10
../../cc/trees/layer_tree_host_unittest.cc(5961): error: Expected equality of these values:
  did_finish_impl_frame_count_
    Which is: 11
  kExpectedNumImplFrames
    Which is: 10
[  FAILED  ] LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame.RunSingleThread_DelegatingRenderer (67 ms)

Comment 4 by danakj@chromium.org, Jan 17 2018

Cc: sunn...@chromium.org briander...@chromium.org

Comment 5 by danakj@chromium.org, Jan 17 2018

Summary: LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame is flaky (was: LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame i sflaky)
Labels: -Pri-3 Pri-2
sheriff: This flaked again in the same way: <https://uberchromegw.corp.google.com/i/chromium.win/builders/Win7%20Tests%20%28dbg%29%281%29/builds/65952>. Bumping priority.
Project Member

Comment 7 by bugdroid1@chromium.org, May 10 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/2fceb01d7d68414f6580d2b87198312d4e468124

commit 2fceb01d7d68414f6580d2b87198312d4e468124
Author: Weiliang Chen <weiliangc@chromium.org>
Date: Thu May 10 15:33:53 2018

cc: Fix Flaky LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame

LayerTreeHostTestWillBeginImplFrameHasDidFinishImplFrame was testing
that each |WillBeginImplFrame| triggers a corresponding
|DidFinishImplFrame|. Since the test is only focused on compositor
thread, stop using commit on main thread to drive the test. Also call
EndTest on thread as soon as we reach expected number of frames, and
stop using whether we expect a |BeginMainFrame| as signal for end of
test.

R=danakj

Bug:  790877 
Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel;master.tryserver.blink:linux_trusty_blink_rel
Change-Id: Ie8847e2ddec49a1191e87a0a4608cb2bce4eb952
Reviewed-on: https://chromium-review.googlesource.com/1052926
Reviewed-by: danakj <danakj@chromium.org>
Commit-Queue: weiliangc <weiliangc@chromium.org>
Cr-Commit-Position: refs/heads/master@{#557524}
[modify] https://crrev.com/2fceb01d7d68414f6580d2b87198312d4e468124/cc/trees/layer_tree_host_unittest.cc

Owner: weiliangc@chromium.org
Status: Started (was: Available)
Status: Fixed (was: Started)

Sign in to add a comment