mash_browser_tests taking over one hour to run on the bots |
|||
Issue descriptionThis build: https://luci-milo.appspot.com/buildbot/tryserver.chromium.linux/linux_chromium_chromeos_ozone_rel_ng/292360 I see this in the logs: [1220/070510.170155:FATAL:env.cc(93)] Check failed: env. Env::CreateInstance must be called before getting the instance of Env. #0 0x000003309a4e base::debug::StackTrace::StackTrace() #1 0x000003320eca logging::LogMessage::~LogMessage() #2 0x000004d5f047 aura::Env::GetInstance() #3 0x000004d75b70 aura::WindowEventDispatcher::SynthesizeMouseMoveAfterChangeToWindow() #4 0x000004d6eec4 aura::Window::~Window() #5 0x000004d6f6a9 aura::Window::~Window() #6 0x000004d64f28 aura::WindowTreeClient::~WindowTreeClient() #7 0x000004d655a9 aura::WindowTreeClient::~WindowTreeClient() #8 0x000005f8e521 views::MusClient::~MusClient() #9 0x000005f8e729 views::MusClient::~MusClient() #10 0x000005f8daf6 views::AuraInit::~AuraInit() #11 0x0000032bc2a1 mash::quick_launch::QuickLaunch::~QuickLaunch() #12 0x0000032bc329 mash::quick_launch::QuickLaunch::~QuickLaunch() #13 0x0000048a7840 service_manager::ServiceContext::~ServiceContext() #14 0x0000048a7879 service_manager::ServiceContext::~ServiceContext() #15 0x0000032b93ae mash::MashPackagedService::~MashPackagedService() #16 0x0000048a7840 service_manager::ServiceContext::~ServiceContext() #17 0x0000032fc1af (anonymous namespace)::StartChildApp() #18 0x0000017c9694 _ZN4base8internal7InvokerINS0_9BindStateIPFvN4mojo16InterfaceRequestIN4mash5mojom10LaunchableEEEEJEEES9_E3RunEPNS0_13BindStateBaseEOS8_ #19 0x000001e2f962 service_manager::RunStandaloneService() #20 0x0000032fbee8 RunMashBrowserTests() #21 0x0000032fbdb3 main #22 0x7f2d9e753f45 __libc_start_main #23 0x0000006266c1 <unknown> It's currently been running for 75 minutes. Is something up here? Ken do you know who might own this?
,
Jan 10 2017
,
Jan 10 2017
mash_browser_tests takes < 10 seconds in the most recent builds I checked. I only see one failure in the last 200 builds. I'll open a separate bug for flake.
,
Jan 10 2017
Actually, I only see one flaky failure in mash_browser_tests in 500 runs, so I think we're OK on flake for now. |
|||
►
Sign in to add a comment |
|||
Comment 1 by kbr@chromium.org
, Dec 21 2016Components: -Internals>GPU>Testing UI>Shell>WindowManager
Labels: -Pri-2 Proj-Ozone Pri-1