New issue
Advanced search Search tips

Issue 899006 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: Oct 26
Cc:
EstimatedDays: ----
NextAction: ----
OS: Linux , Android , Windows , Mac
Pri: 2
Type: Bug

Blocking:
issue chromedriver:2629



Sign in to add a comment

ChromeDriver tests in FrameSwitchingTest are failing on Waterfall

Project Member Reported by khachatryan@chromium.org, Oct 25

Issue description

Due to the commit r602299 ChromeDriver tests for classes FrameSwitchingTest and GetLogsTest are failing on Waterfall which breaks ChromeDriver builds on Waterfall.

https://chromium.googlesource.com/chromium/src/+/1316c14aaac985a39d3beee0721d68f6063891a5

To reproduce the issue please run the following tests from 'chromium' project:

1. From 'chromium' root directory go to ../src/chrome/test/chromedriver/test
2. run the following command:

python run_java_tests.py --also-run-disabled-tests --chrome=/<path_to_chrome_binaries>/chrome --chromedriver=/<path_to_chromedriver_binaries>/chromedriver --filter=FrameSwitchingTest.*

This will run all tests from FrameSwitchingTest class.

I would suggest to create a shell script and run the above command from inside the shell with a loop count. Please have count of 10 or 20, because the issue is flaky.

Tests are failing on following error:
...
================================================================================
========== org.openqa.selenium.FrameSwitchingTest.testShouldThrowFrameNotFoundExceptionLookingUpSubFramesWithSuperFrameNames (0.654s)
org.openqa.selenium.WebDriverException: unknown error: session deleted because of page crash
from tab crashed
  (Session info: chrome=72.0.3591.0)
  (Driver info: chromedriver=2.43.600419 (5804c136290b50e8d9e4fc81ed7873787c076292),platform=Linux 4.17.0-3rodete2-amd64 x86_64) (WARNING: The server did not provide any stacktrace information)
Command duration or timeout: 0 millisecond
...
 
Blocking: chromedriver:2629
Labels: -Pri-3 Pri-2
Description: Show this description
I believe this is same as Issue 898772.  The CL has been reverted few hours ago, the bots should be okay soon.

Tests are passing with the Chrome built from revision 602298 and failing on revision 602299, so I don't see if this might be related to the Issue 898772

To clarify, revision 602299 has been reverted to fix Issue 898772.

Do you a link to failed bots?  Or at least a Chrome stack trace?  Just to double check my assumption.

https://ci.chromium.org/p/chromium/g/chromium.chromedriver/console

Click on green bars for Linux, Mac and Win and you'll see the build history for each platform.


Now they are showing green because we are disabled those tests.

Cc: mustaq@chromium.org
Owner: khachatryan@chromium.org
I couldn't find any relevant failure in Linux Builder:
https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Linux%20Builder/107143

If you think revision 602299 is the cause, this has been reverted.
Otherwise, this is some other problem.

khachatryan@chromium.org: Could you please check if the revert of r602299 resolved the problem?

For the record, here is the stack trace from a failure:

Received signal 11 SEGV_MAPERR 000000000020
#0 0x7fcfaebfa98f base::debug::StackTrace::StackTrace()
#1 0x7fcfaebfa501 base::debug::(anonymous namespace)::StackDumpSignalHandler()
#2 0x7fcfa9be9330 <unknown>
#3 0x7fcfaca1ad20 <unknown>
#4 0x7fcfb1d7b5cd blink::FrameTree::IsDescendantOf()
#5 0x7fcfb18e9d72 blink::LocalFrame::HasTransientUserActivation()
#6 0x7fcfb1d2e916 blink::FrameLoader::CommitNavigation()
#7 0x7fcfb193bc42 blink::WebLocalFrameImpl::CommitNavigation()
#8 0x7fcfb21b2199 content::RenderFrameImpl::CommitNavigation()
#9 0x7fcfad0c7c67 content::mojom::FrameNavigationControlStubDispatch::AcceptWithResponder()
#10 0x7fcfb21ca946 content::mojom::FrameNavigationControlStub<>::AcceptWithResponder()
#11 0x7fcfaec4e438 mojo::InterfaceEndpointClient::HandleValidatedMessage()
#12 0x7fcfaf19a295 IPC::(anonymous namespace)::ChannelAssociatedGroupController::AcceptOnProxyThread()
#13 0x7fcfaf198269 _ZN4base8internal7InvokerINS0_9BindStateIMN3IPC12_GLOBAL__N_132ChannelAssociatedGroupControllerEFvN4mojo7MessageEEJ13scoped_refptrIS5_ENS0_13PassedWrapperIS7_EEEEEFvvEE3RunEPNS0_13BindStateBaseE
#14 0x7fcfaeb78f91 base::debug::TaskAnnotator::RunTask()
#15 0x7fcfaebb826c base::sequence_manager::internal::ThreadControllerImpl::DoWork()
#16 0x7fcfaeb78f91 base::debug::TaskAnnotator::RunTask()
#17 0x7fcfaeb781bf base::MessageLoop::RunTask()
#18 0x7fcfaeb78542 base::MessageLoop::DoWork()
#19 0x7fcfaeb7ac9a base::MessagePumpDefault::Run()
#20 0x7fcfaeb96df5 base::RunLoop::Run()
#21 0x7fcfb2a6473f content::RendererMain()
#22 0x7fcfae7d3db7 content::RunZygote()
#23 0x7fcfae7d4f4d content::ContentMainRunnerImpl::Run()
#24 0x7fcfae8066ab service_manager::Main()
#25 0x7fcfae7d33c1 content::ContentMain()
#26 0x7fcfac9351b3 ChromeMain
#27 0x7fcfa3d31f45 __libc_start_main
#28 0x7fcfac93502a _start
Status: Fixed (was: Untriaged)
Reverting r602299 resolved the issue.

I tested with Chromium	72.0.3592.0 (Developer Build) (64-bit)
Revision	6fd4ae8a480a8d88820a7bc933763443cf8b68cf-refs/heads/master@{#602962}

Sign in to add a comment