New issue
Advanced search Search tips

Issue 647311 link

Starred by 5 users

Issue metadata

Status: Fixed
Owner:
Closed: Oct 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 2
Type: Bug



Sign in to add a comment

SitePerProcessInteractiveBrowserTest.FullscreenElementInSubframe flaky on win_chromium_x64_rel_ng

Project Member Reported by thakis@chromium.org, Sep 15 2016

Issue description

Cc: alex...@chromium.org
Components: Internals>Sandbox>SiteIsolation
Labels: -Pri-3 Pri-2
Owner: kenrb@chromium.org
Status: Assigned (was: Untriaged)
Indeed, this has started being very flaky on Windows bots in the last day or so:

https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=interactive_ui_tests%20(with%20patch)&tests=SitePerProcessInteractiveBrowserTest.FullscreenElementInSubframe

All the failures I checked hit the following DCHECK:

[5640:4772:0914/151751:FATAL:cross_process_frame_connector.cc(150)] Check failed: false.

	base::debug::StackTrace::StackTrace [0x026D1B17+23]
	logging::LogMessage::~LogMessage [0x026636E1+49]
	content::CrossProcessFrameConnector::TransformPointToLocalCoordSpace [0x0191137E+172]
	content::RenderWidgetHostViewChildFrame::TransformPointToLocalCoordSpace [0x01938E8C+49]
	content::DelegatedFrameHost::TransformPointToCoordSpaceForView [0x01B24725+54]
	content::RenderWidgetHostViewAura::TransformPointToCoordSpaceForView [0x01A4D310+23]
	content::RenderWidgetHostInputEventRouter::SendMouseEnterOrLeaveEvents [0x01A47E64+611]
	content::RenderWidgetHostInputEventRouter::RouteMouseEvent [0x01A473A2+151]
	content::RenderWidgetHostViewAura::OnMouseEvent [0x01A4BAF6+1204]
	ui::EventHandler::OnEvent [0x02972F6D+41]
	ui::EventDispatcher::DispatchEvent [0x0297672B+61]
	ui::EventDispatcher::ProcessEvent [0x02976A56+142]
	ui::EventDispatcherDelegate::DispatchEventToTarget [0x02976935+42]
	ui::EventDispatcherDelegate::DispatchEvent [0x029767D2+135]
	ui::EventProcessor::OnEventFromSource [0x029779A4+304]
	ui::EventSource::DeliverEventToProcessor [0x02977615+80]
	ui::EventSource::SendEventToProcessor [0x029776F6+214]
	views::DesktopWindowTreeHostWin::HandleScrollEvent [0x02638D49+20]
	views::HWNDMessageHandler::HandleMouseEventInternal [0x0264ADF9+752]
	views::HWNDMessageHandler::HandleMouseMessage [0x0264B0D6+45]
	content::LegacyRenderWidgetHostHWND::OnMouseRange [0x01A02EB8+215]
	content::LegacyRenderWidgetHostHWND::_ProcessWindowMessage [0x01A033C6+196]
	content::LegacyRenderWidgetHostHWND::ProcessWindowMessage [0x01A030F8+33]
	ATL::CWindowImplBaseT<ATL::CWindow,ATL::CWinTraits<1073741824,0> >::WindowProc [0x01A03256+72]
	gapfnScSendMessage [0x762A62FA+818]
	GetThreadDesktop [0x762A6D3A+215]
	CharPrevW [0x762A77C4+312]
	DispatchMessageW [0x762A788A+15]
	base::MessagePumpForUI::ProcessMessageHelper [0x026DF017+199]
	base::MessagePumpForUI::ProcessPumpReplacementMessage [0x026DF15D+237]
	base::MessagePumpForUI::ProcessMessageHelper [0x026DEFF5+165]
	base::MessagePumpForUI::DoRunLoop [0x026DE691+65]
	base::MessagePumpWin::Run [0x026DF37A+74]
	base::MessageLoop::RunHandler [0x02671C47+103]
	base::RunLoop::Run [0x026817C9+41]
	content::RunThisRunLoop [0x0219799C+49]
	content::MessageLoopRunner::Run [0x021977A1+105]
	content::ExecuteScriptAndExtractString [0x021982D4+286]
	content::ExecuteScript [0x0219806F+52]
	SitePerProcessInteractiveBrowserTest_FullscreenElementInSubframe_Test::RunTestOnMainThread [0x00E25193+2306]
	InProcessBrowserTest::RunTestOnMainThreadLoop [0x01EE555C+180]
	content::BrowserTestBase::ProxyRunTestOnMainThreadLoop [0x0219BF49+175]
	ChromeBrowserMainParts::PreMainMessageLoopRunImpl [0x0201E91C+3288]
	ChromeBrowserMainParts::PreMainMessageLoopRun [0x0201DC0C+167]
	content::BrowserMainLoop::PreMainMessageLoopRun [0x0188DD68+112]
	content::StartupTaskRunner::RunAllTasksNow [0x01AB91D4+38]
	content::BrowserMainLoop::CreateStartupTasks [0x0188C163+394]
	content::BrowserMainRunnerImpl::Initialize [0x0188EF93+632]
	content::BrowserMain [0x0188AE0E+117]
	content::RunNamedProcessTypeMain [0x01ED6DB3+206]
	content::ContentMainRunnerImpl::Run [0x01ED6CB4+274]
	content::ContentMain [0x01ED6081+35]
	content::BrowserTestBase::SetUp [0x0219C52F+964]
	InProcessBrowserTest::SetUp [0x01EE58CE+268]
	testing::internal::HandleExceptionsInMethodIfSupported<testing::TestCase,void> [0x025E5498+32]
	testing::Test::Run [0x025EBDBE+51]
	testing::TestCase::Run [0x025EBE94+133]
	testing::internal::UnitTestImpl::RunAllTests [0x025EC213+433]
	testing::internal::HandleExceptionsInMethodIfSupported<testing::internal::UnitTestImpl,bool> [0x025E54DC+32]
	testing::UnitTest::Run [0x025EC03D+133]
	base::TestSuite::Run [0x0265B998+95]
	InteractiveUITestSuiteRunner::RunTestSuite [0x00E63E2E+52]

Ken, I see you added that check in https://codereview.chromium.org/2184033003.  Any ideas on what could change in the last day or so to start triggering that DCHECK?  Could it be your https://codereview.chromium.org/2229463004?

I'll disable the test on Windows in the meantime.

Comment 2 by kenrb@chromium.org, Sep 15 2016

It is fairly certain to be that CL. I'll have a look.
Cc: kenrb@chromium.org
 Issue 647339  has been merged into this issue.
Test will be disabled in https://codereview.chromium.org/2341683004/
Project Member

Comment 5 by bugdroid1@chromium.org, Sep 15 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e18b55477c0890c1516d940ac8791dc7aa304275

commit e18b55477c0890c1516d940ac8791dc7aa304275
Author: alexmos <alexmos@chromium.org>
Date: Thu Sep 15 19:48:57 2016

Disable FullscreenElementInSubframe test on Windows due to being flaky.

BUG= 647311 
TBR=kenrb@chromium.org

Review-Url: https://codereview.chromium.org/2341683004
Cr-Commit-Position: refs/heads/master@{#418935}

[modify] https://crrev.com/e18b55477c0890c1516d940ac8791dc7aa304275/chrome/browser/site_per_process_interactive_browsertest.cc

 Issue 647519  has been merged into this issue.
 Issue 647549  has been merged into this issue.
Also disabling FullscreenElementInABAAndExitViaEscapeKey for the same error. If more start to flake as well, I'll probably disable the entire suite.
Project Member

Comment 9 by bugdroid1@chromium.org, Sep 16 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/f36f4339c6a7678bf66d7ed3d522cccd593e3d62

commit f36f4339c6a7678bf66d7ed3d522cccd593e3d62
Author: iclelland <iclelland@chromium.org>
Date: Fri Sep 16 14:30:02 2016

Disable FullscreenElementInABAAndExitViaEscapeKey test on Windows due to flakes

BUG= 647311 
TBR=kenrb@chromium.org

Review-Url: https://codereview.chromium.org/2342853003
Cr-Commit-Position: refs/heads/master@{#419163}

[modify] https://crrev.com/f36f4339c6a7678bf66d7ed3d522cccd593e3d62/chrome/browser/site_per_process_interactive_browsertest.cc

Project Member

Comment 10 by bugdroid1@chromium.org, Oct 14 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/9380a25fe8cc25c6c588218a2af44f34a4d115be

commit 9380a25fe8cc25c6c588218a2af44f34a4d115be
Author: kenrb <kenrb@chromium.org>
Date: Fri Oct 14 14:21:41 2016

Account for failure of coordinate space transformations in browser

In some situations, DCHECKs were triggering due to failed
transformations of points between cc::Surface coordinate spaces. This is
because of raciness that can occur when we cache RenderWidgetHostView
pointers: specifically, if the RWHV has changed its Surface due to
(for example) a resize, and the new SurfaceID has not yet propagated
through an embedding renderer process and back to the browser, then
the browser process does not have enough information to compute the
transformation.

This CL allows the RenderWidgetHostInputEventRouter to handle that
situation gracefully, rather than hitting a DCHECK.

BUG= 647311 
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2417463003
Cr-Commit-Position: refs/heads/master@{#425316}

[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/chrome/browser/site_per_process_interactive_browsertest.cc
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/frame_host/cross_process_frame_connector.cc
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/frame_host/cross_process_frame_connector.h
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/frame_host/render_widget_host_view_child_frame.cc
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/frame_host/render_widget_host_view_child_frame.h
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/delegated_frame_host.cc
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/delegated_frame_host.h
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/render_widget_host_input_event_router.cc
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/render_widget_host_view_aura.cc
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/render_widget_host_view_aura.h
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/render_widget_host_view_base.cc
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/render_widget_host_view_base.h
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/render_widget_host_view_mac.h
[modify] https://crrev.com/9380a25fe8cc25c6c588218a2af44f34a4d115be/content/browser/renderer_host/render_widget_host_view_mac.mm

Comment 11 by kenrb@chromium.org, Oct 14 2016

Status: Fixed (was: Assigned)

Sign in to add a comment