Issue metadata
Sign in to add a comment
|
network_service_browser_tests failing on multiple builders |
||||||||||||||||||||||||||||
Issue descriptionFiled by sheriff-o-matic@appspot.gserviceaccount.com on behalf of fdoray@google.com network_service_browser_tests failing on multiple builders Builders failed on: - Win7 Tests (dbg)(1): https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win7%20Tests%20%28dbg%29%281%29 Failing test: WebVrBrowserTestOpenVrDisabled.TestWebVrNoDevicesWithoutOpenVr ../../chrome/browser/vr/test/xr_browser_test.cc(95): error: Value of: PollJavaScriptBoolean( "isInitializationComplete()", kPollTimeoutShort, browser()->tab_strip_model()->GetActiveWebContents()) Actual: false Expected: true Timed out waiting for JavaScript test initialization. Stack trace: Backtrace: StackTraceGetter::CurrentStackTrace [0x02DAAA9A+90] testing::internal::UnitTestImpl::CurrentOsStackTraceExceptTop [0x02DC194A+74] testing::internal::AssertHelper::operator= [0x02DC145B+75] vr::XrBrowserTestBase::LoadUrlAndAwaitInitialization [0x0299B1AC+364] vr::WebVrBrowserTestOpenVrDisabled_TestWebVrNoDevicesWithoutOpenVr_Test::RunTestOnMainThread [0x029A6024+132] content::BrowserTestBase::ProxyRunTestOnMainThreadLoop [0x05FC8AAA+874] ??$Invoke@P8BrowserTestBase@content@@AEXXZPAV12@$$V@?$FunctorTraits@P8BrowserTestBase@content@@AEXXZX@internal@base@@SAXP8BrowserTestBase@content@@AEXXZ$$QAPAV34@@Z [0x05FCB17C+28] base::internal::InvokeHelper<0,void>::MakeItSo<void (__thiscall content::BrowserTestBase::*const &)(void),content::BrowserTestBase *> [0x05FCB0BF+79] base::internal::Invoker<base::internal::BindState<void (__thiscall content::BrowserTestBase::*)(void),base::internal::UnretainedWrapper<content::BrowserTestBase> >,void __cdecl(void)>::RunImpl<void (__thiscall content::BrowserTestBase::*const &)(void),std [0x05FCB055+85] base::internal::Invoker<base::internal::BindState<void (__thiscall content::BrowserTestBase::*)(void),base::internal::UnretainedWrapper<content::BrowserTestBase> >,void __cdecl(void)>::Run [0x05FCAF3F+63] base::RepeatingCallback<void __cdecl(void)>::Run [0x00B0EB92+50] ChromeBrowserMainParts::PreMainMessageLoopRunImpl [0x07A22447+7447] ChromeBrowserMainParts::PreMainMessageLoopRun [0x07A206AE+398] content::BrowserMainLoop::PreMainMessageLoopRun [0x5D65AE36+342] ??$Invoke@P8BrowserMainLoop@content@@AEHXZPAV12@$$V@?$FunctorTraits@P8BrowserMainLoop@content@@AEHXZX@internal@base@@SAHP8BrowserMainLoop@content@@AEHXZ$$QAPAV34@@Z [0x5D665D8C+28] base::internal::InvokeHelper<0,int>::MakeItSo<int (__thiscall content::BrowserMainLoop::*const &)(void),content::BrowserMainLoop *> [0x5D665CBF+79] base::internal::Invoker<base::internal::BindState<int (__thiscall content::BrowserMainLoop::*)(void),base::internal::UnretainedWrapper<content::BrowserMainLoop> >,int __cdecl(void)>::RunImpl<int (__thiscall content::BrowserMainLoop::*const &)(void),std::t [0x5D665C55+85] base::internal::Invoker<base::internal::BindState<int (__thiscall content::BrowserMainLoop::*)(void),base::internal::UnretainedWrapper<content::BrowserMainLoop> >,int __cdecl(void)>::Run [0x5D665AFF+63] base::RepeatingCallback<int __cdecl(void)>::Run [0x5E05DAF2+50] content::StartupTaskRunner::RunAllTasksNow [0x5E9EE989+137] content::BrowserMainLoop::CreateStartupTasks [0x5D658A75+1493] content::BrowserMainRunnerImpl::Initialize [0x5D66C514+1908] content::BrowserMain [0x5D6522FD+173] content::RunBrowserProcessMain [0x600B803D+141] content::ContentMainRunnerImpl::Run [0x600B9227+903] content::ContentServiceManagerMainDelegate::RunEmbedderProcess [0x600B59C2+50] service_manager::Main [0x594A3EF3+1811] content::ContentMain [0x600B7F2C+92] content::BrowserTestBase::SetUp [0x05FC8429+3097] InProcessBrowserTest::SetUp [0x0561AE39+1961] vr::XrBrowserTestBase::SetUp [0x0299AE40+2448]
,
Sep 28
,
Sep 28
CL to disable test is in the CQ https://chromium-review.googlesource.com/c/chromium/src/+/1252083 Long-term solution might be to increase the value of kPollTimeoutShort.
,
Sep 28
,
Sep 28
My guess is that the root cause is that the tests are being run when they shouldn't be. The VR tests should never be running in anything except xr_browser_tests, as we have specific hardware requirements. I'll take a look into how they got enabled on network_service_browser_tests.
,
Sep 28
Oh wait, this test is one of the few that doesn't have hardware requirements and can run anywhere. Ignore comment #5
,
Sep 28
My best guess given the logs is that promise that getVRDisplays returns took more than 1 second to resolve. This is possible, since we are now spawning an extra process while we weren't before, so potentially on slow machines it could take longer. Increasing the timeout would be reasonable, but it'd be nice to understand how much longer it is taking, and whether the machine is under heavy load at the time (spawning a process is generally quick).
,
Sep 28
The issue does seem to just be the combination of short timeout and lots of tests running simultaneously. One second is plenty when we run the full set of VR tests, as we run everything serially. However, network_service_browser_tests runs a bunch of tests in parallel, so it sometimes takes more than one second to spawn the process and return a device. Since the disable patch is almost through the CQ, I'll wait for that to go in then increase the timeout and reenable.
,
Sep 28
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/6b21dd3be51917d78d998318220b6780780d8f94 commit 6b21dd3be51917d78d998318220b6780780d8f94 Author: Francois Doray <fdoray@chromium.org> Date: Fri Sep 28 19:00:38 2018 Disable WebVrBrowserTestOpenVrDisabled.TestWebVrNoDevicesWithoutOpenVr. TBR=acondor@chromium.org Bug: 890380 Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:linux_vr;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel Change-Id: Ibce268155e198a855e6815bced44578001b8279e Reviewed-on: https://chromium-review.googlesource.com/1252083 Reviewed-by: François Doray <fdoray@chromium.org> Commit-Queue: François Doray <fdoray@chromium.org> Cr-Commit-Position: refs/heads/master@{#595170} [modify] https://crrev.com/6b21dd3be51917d78d998318220b6780780d8f94/chrome/browser/vr/webxr_vr_transition_browser_test.cc
,
Sep 29
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/591949795043a818e50aba8a539094c321a4220c commit 591949795043a818e50aba8a539094c321a4220c Author: bsheedy <bsheedy@chromium.org> Date: Sat Sep 29 00:18:27 2018 Re-enable disabled VR browser test Re-enables WebVrBrowserTestOpenVrDisabled.TestWebVrNoDevicesWithoutOpenVr and increases the timeout that was flakily being hit when many tests were run in parallel. Bug: 890380 Cq-Include-Trybots: luci.chromium.try:android_optional_gpu_tests_rel;luci.chromium.try:linux_optional_gpu_tests_rel;luci.chromium.try:linux_vr;luci.chromium.try:mac_optional_gpu_tests_rel;luci.chromium.try:win_optional_gpu_tests_rel Change-Id: Iebb10f25848ac0ab3a2913fff4e35dad919fd4bc Reviewed-on: https://chromium-review.googlesource.com/1252642 Reviewed-by: Bill Orr <billorr@chromium.org> Commit-Queue: Brian Sheedy <bsheedy@chromium.org> Cr-Commit-Position: refs/heads/master@{#595266} [modify] https://crrev.com/591949795043a818e50aba8a539094c321a4220c/chrome/browser/vr/test/xr_browser_test.cc [modify] https://crrev.com/591949795043a818e50aba8a539094c321a4220c/chrome/browser/vr/test/xr_browser_test.h [modify] https://crrev.com/591949795043a818e50aba8a539094c321a4220c/chrome/browser/vr/webxr_vr_transition_browser_test.cc
,
Oct 2
transferring ownership to bsheedy to decide if further action is needed.
,
Oct 2
Appears to be passing fine since being re-enabled, removing from sheriff queue.
,
Oct 2
|
|||||||||||||||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||||||||||||||
Comment 1 by fdoray@chromium.org
, Sep 28