This currently the #2 most verbose warning during testing. > 15430 [NNNNN] WARNING: No inner window available!: file dom/base/nsGlobalWindow.cpp, line 9767 This warning [1], from |nsGlobalWindow::GetKeyboardIndicators|, shows up in the following test suites: > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-e10s-1-bm54-tests1-linux64-build1.txt:1047 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-browser-chrome-7-bm54-tests1-linux64-build0.txt:969 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-e10s-3-bm121-tests1-linux64-build5.txt:938 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-e10s-5-bm68-tests1-linux64-build4.txt:924 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-reftests-e10s-bm68-tests1-linux64-build1.txt:904 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-e10s-6-bm53-tests1-linux64-build2.txt:851 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-browser-chrome-5-bm131-tests1-linux64-build5.txt:757 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-browser-chrome-4-bm122-tests1-linux64-build12.txt:717 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-e10s-2-bm114-tests1-linux64-build1.txt:670 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-e10s-8-bm54-tests1-linux64-build5.txt:620 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-browser-chrome-6-bm53-tests1-linux64-build5.txt:610 > mozilla-central_ubuntu64_vm-debug_test-mochitest-other-bm131-tests1-linux64-build4.txt:545 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-browser-chrome-3-bm115-tests1-linux64-build13.txt:505 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-browser-chrome-2-bm131-tests1-linux64-build4.txt:497 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-reftests-bm53-tests1-linux64-build5.txt:453 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-e10s-4-bm68-tests1-linux64-build5.txt:393 > mozilla-central_ubuntu64_vm-debug_test-mochitest-browser-chrome-4-bm115-tests1-linux64-build2.txt:392 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-browser-chrome-1-bm116-tests1-linux64-build2.txt:328 > mozilla-central_ubuntu64_vm-debug_test-mochitest-browser-chrome-5-bm122-tests1-linux64-build1.txt:258 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-1-bm131-tests1-linux64-build5.txt:250 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-e10s-7-bm113-tests1-linux64-build4.txt:216 > mozilla-central_ubuntu64_vm-debug_test-mochitest-jetpack-bm115-tests1-linux64-build7.txt:215 > mozilla-central_ubuntu64_vm-debug_test-mochitest-2-bm131-tests1-linux64-build7.txt:200 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-8-bm131-tests1-linux64-build0.txt:179 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-8-bm118-tests1-linux64-build3.txt:179 > mozilla-central_ubuntu64_vm-debug_test-mochitest-8-bm54-tests1-linux64-build0.txt:153 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-2-bm125-tests1-linux64-build2.txt:150 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-5-bm68-tests1-linux64-build4.txt:140 > mozilla-central_ubuntu64_vm-debug_test-mochitest-media-e10s-bm54-tests1-linux64-build0.txt:138 > mozilla-central_ubuntu64_vm-debug_test-mochitest-browser-chrome-2-bm131-tests1-linux64-build8.txt:134 > mozilla-central_ubuntu64_vm-debug_test-mochitest-media-bm114-tests1-linux64-build0.txt:128 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-3-bm68-tests1-linux64-build4.txt:124 > mozilla-central_ubuntu64_vm-debug_test-mochitest-browser-chrome-7-bm122-tests1-linux64-build7.txt:115 > mozilla-central_ubuntu64_vm-debug_test-mochitest-browser-chrome-1-bm131-tests1-linux64-build4.txt:113 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-7-bm123-tests1-linux64-build6.txt:111 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-6-bm115-tests1-linux64-build5.txt:104 > mozilla-central_ubuntu64_vm-debug_test-mochitest-1-bm130-tests1-linux64-build6.txt:102 > mozilla-central_ubuntu64_vm-debug_test-mochitest-3-bm118-tests1-linux64-build7.txt:101 > mozilla-central_ubuntu64_vm-debug_test-mochitest-browser-chrome-3-bm118-tests1-linux64-build2.txt:94 > mozilla-central_ubuntu64_vm-debug_test-mochitest-e10s-4-bm131-tests1-linux64-build3.txt:87 > mozilla-central_ubuntu64_vm-debug_test-mochitest-5-bm122-tests1-linux64-build8.txt:80 > mozilla-central_ubuntu64_vm-debug_test-mochitest-devtools-chrome-3-bm117-tests1-linux64-build8.txt:75 > mozilla-central_ubuntu64_vm-debug_test-mochitest-7-bm115-tests1-linux64-build4.txt:73 > mozilla-central_ubuntu64_vm-debug_test-mochitest-browser-chrome-6-bm118-tests1-linux64-build4.txt:64 > mozilla-central_ubuntu64_vm-debug_test-mochitest-6-bm117-tests1-linux64-build3.txt:53 > mozilla-central_ubuntu64_vm-debug_test-mochitest-devtools-chrome-6-bm54-tests1-linux64-build4.txt:51 > mozilla-central_ubuntu64_vm-debug_test-mochitest-4-bm122-tests1-linux64-build3.txt:50 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-5-bm54-tests1-linux64-build2.txt:30 > mozilla-central_ubuntu64_vm-debug_test-mochitest-devtools-chrome-8-bm122-tests1-linux64-build1.txt:27 > mozilla-central_ubuntu64_vm-debug_test-mochitest-devtools-chrome-5-bm54-tests1-linux64-build4.txt:26 > mozilla-central_ubuntu64_vm-debug_test-mochitest-devtools-chrome-4-bm131-tests1-linux64-build5.txt:22 > mozilla-central_ubuntu64_vm-debug_test-mochitest-devtools-chrome-2-bm113-tests1-linux64-build6.txt:18 > mozilla-central_ubuntu64_vm-debug_test-mochitest-devtools-chrome-1-bm53-tests1-linux64-build5.txt:17 > mozilla-central_ubuntu64_vm-debug_test-mochitest-devtools-chrome-7-bm68-tests1-linux64-build6.txt:16 > mozilla-central_ubuntu64_vm-debug_test-crashtest-e10s-bm53-tests1-linux64-build4.txt:12 > mozilla-central_ubuntu64_vm-debug_test-mochitest-gl-e10s-bm120-tests1-linux64-build6.txt:9 > mozilla-central_ubuntu64_vm-debug_test-crashtest-bm67-tests1-linux64-build6.txt:8 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-4-bm123-tests1-linux64-build8.txt:6 > mozilla-central_ubuntu64_vm-debug_test-mochitest-gl-bm53-tests1-linux64-build4.txt:6 > mozilla-central_ubuntu64_vm-debug_test-reftest-e10s-2-bm118-tests1-linux64-build10.txt:5 > mozilla-central_ubuntu64_vm-debug_test-reftest-e10s-1-bm120-tests1-linux64-build10.txt:5 > mozilla-central_ubuntu64_vm-debug_test-jsreftest-e10s-bm131-tests1-linux64-build9.txt:5 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-3-bm116-tests1-linux64-build8.txt:4 > mozilla-central_ubuntu64_vm-debug_test-reftest-4-bm52-tests1-linux64-build7.txt:4 > mozilla-central_ubuntu64_vm-debug_test-reftest-3-bm131-tests1-linux64-build5.txt:4 > mozilla-central_ubuntu64_vm-debug_test-reftest-2-bm118-tests1-linux64-build2.txt:4 > mozilla-central_ubuntu64_vm-debug_test-reftest-1-bm131-tests1-linux64-build5.txt:4 > mozilla-central_ubuntu64_vm-debug_test-jsreftest-bm131-tests1-linux64-build9.txt:4 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-8-bm54-tests1-linux64-build0.txt:2 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-7-bm116-tests1-linux64-build0.txt:2 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-6-bm54-tests1-linux64-build1.txt:2 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-2-bm122-tests1-linux64-build5.txt:2 > mozilla-central_ubuntu64_vm-debug_test-web-platform-tests-1-bm118-tests1-linux64-build1.txt:2 It shows up in 8205 tests. A few of the most prevalent: > 1561 - Shutdown > 99 - jetpack-package/addon-sdk/source/test/test-simple-prefs.js.testUnloadOfDynamicPrefGeneration > 51 - dom/html/test/test_fullscreen-api.html > 42 - dom/plugins/test/mochitest/test_pluginstream_err.html > 42 - browser/components/sessionstore/test/browser_formdata_cc.js > 40 - browser/components/privatebrowsing/test/browser/browser_privatebrowsing_windowtitle.js > 40 - browser/base/content/test/general/browser_overflowScroll.js > 37 - dom/manifest/test/browser_ManifestObtainer_obtain.js > 37 - browser/base/content/test/general/browser_sanitizeDialog.js > 36 - toolkit/components/passwordmgr/test/browser/browser_passwordmgrdlg.js [1] https://dxr.mozilla.org/mozilla-central/rev/68c0b7d6f16ce5bb023e08050102b5f2fe4aacd8/dom/base/nsGlobalWindow.cpp#9767

Looks like the only caller is nsBaseWidget::NotifyUIStateChanged().

I'm guessing this is a regression from bug 1241764.

We see an awful lot of these in the Thunderbird test suite.

Wayne, have you confirmed this is a regression? Kyle, WDYT?

A stack would be useful.

(In reply to Andrew Overholt [:overholt] from comment #4) > Wayne, have you confirmed this is a regression? No, I have not. I have not tested anything. A presumption in part based on comment 2

Eric, can you provide a stack for khuey?

This is the log with stack dumping enabled while running: > ./mach test dom/html/test/test_fullscreen-api.html Is this enough info for you Kyle?

How is that calling GetKeyboardIndicators?

(In reply to Kyle Huey [:khuey] (khuey@mozilla.com) from comment #9) > How is that calling GetKeyboardIndicators? Looks like maybe mach did a bad job at symbolizing the log. Results from gdb make more sense: > Breakpoint 1, nsGlobalWindow::GetKeyboardIndicators (this=<optimized out>, aShowAccelerators=<optimized out>, > aShowFocusRings=<optimized out>) at /home/erahm/dev/mozilla-central/dom/base/nsGlobalWindow.cpp:9779 > 9779 NS_WARNING("No inner window available!"); > (gdb) bt > #0 0x00007fffea365be0 in nsGlobalWindow::GetKeyboardIndicators(bool*, bool*) (this=<optimized out>, aShowAccelerators=<optimized out>, aShowFocusRings=<optimized out>) at /home/erahm/dev/mozilla-central/dom/base/nsGlobalWindow.cpp:9779 > #1 0x00007fffea344075 in nsGlobalWindow::InitializeShowFocusRings() (this=0x7fffd4634800) > at /home/erahm/dev/mozilla-central/dom/base/nsGlobalWindow.cpp:9714 > #2 0x00007fffea343d82 in nsGlobalWindow::nsGlobalWindow(nsGlobalWindow*) (this=0x7fffd4634800, aOuterWindow=0x7fffd4633800) > at /home/erahm/dev/mozilla-central/dom/base/nsGlobalWindow.cpp:1228 > #3 0x00007fffea37797e in nsGlobalChromeWindow::nsGlobalChromeWindow(nsGlobalWindow*) (this=0x7fffd4634800, aOuterWindow=0x7fffffffb8a7) > at /home/erahm/dev/mozilla-central/dom/base/nsGlobalWindow.h:1932 > #4 0x00007fffea34a458 in nsGlobalChromeWindow::Create(nsGlobalWindow*) (aOuterWindow=0x7fffffffb8a7) > at /home/erahm/dev/mozilla-central/dom/base/nsGlobalWindow.cpp:13157 > #5 0x00007fffea348a30 in nsGlobalWindow::SetNewDocument(nsIDocument*, nsISupports*, bool) (this=0x7fffd4633800, aDocument= > 0x7fffd46d2000, aState=0x0, aForceReuseInnerWindow=<optimized out>) > at /home/erahm/dev/mozilla-central/dom/base/nsGlobalWindow.cpp:2550 > #6 0x00007fffebadbf8f in nsDocumentViewer::InitInternal(nsIWidget*, nsISupports*, mozilla::gfx::IntRectTyped<mozilla::gfx::UnknownUnits> const&, bool, bool, bool) (this=0x7fffd593ad80, aParentWidget=<optimized out>, aState=0x0, aBounds=..., aDoCreation=true, aNeedMakeCX=<optimized out>, aForceSetNewDocument=<optimized out>) at /home/erahm/dev/mozilla-central/layout/base/nsDocumentViewer.cpp:874 > #7 0x00007fffebadb9e2 in nsDocumentViewer::Init(nsIWidget*, mozilla::gfx::IntRectTyped<mozilla::gfx::UnknownUnits> const&) (this=0x7fffd4633800, aParentWidget=0x7fffffffb8a7, aBounds=...) at /home/erahm/dev/mozilla-central/layout/base/nsDocumentViewer.cpp:618 > #8 0x00007fffebed7394 in nsDocShell::SetupNewViewer(nsIContentViewer*) (this=0x7fffd4633000, aNewViewer=<optimized out>) > at /home/erahm/dev/mozilla-central/docshell/base/nsDocShell.cpp:9251 > #9 0x00007fffebed6d44 in nsDocShell::Embed(nsIContentViewer*, char const*, nsISupports*) (this=0x7fffd4633000, aContentViewer=0x7fffd593ad80, aCommand=0x7fffffffb8a6 "", aExtraInfo=0x7fffeea55be0 <vtable for nsGlobalChromeWindow+1128>) > at /home/erahm/dev/mozilla-central/docshell/base/nsDocShell.cpp:7141 > #10 0x00007fffebedb401 in nsDocShell::CreateAboutBlankContentViewer(nsIPrincipal*, nsIURI*, bool) (this=0x7fffd4633000, aPrincipal= > 0x7fff80004005, aBaseURI=0x0, aTryToSaveOldPresentation=<optimized out>) > at /home/erahm/dev/mozilla-central/docshell/base/nsDocShell.cpp:7971 > #11 0x00007fffebf3bc71 in nsWebShellWindow::Initialize(nsIXULWindow*, nsIXULWindow*, nsIURI*, int, int, bool, nsITabParent*, nsWidgetInitData&) (this=<optimized out>, aParent=0x6469752e6c746e69, aOpener=<optimized out>, aUrl=0x7fffd587aa00, aInitialWidth=<optimized out>, aInitialHeight=<optimized out>, aIsHiddenWindow=<optimized out>, aOpeningTab=<optimized out>, widgetInitData=...) > at /home/erahm/dev/mozilla-central/xpfe/appshell/nsWebShellWindow.cpp:218 > #12 0x00007fffebf3a029 in nsAppShellService::JustCreateTopWindow(nsIXULWindow*, nsIURI*, unsigned int, int, int, bool, nsITabParent*, nsWebShellWindow**) (this=<optimized out>, aParent=0x0, aUrl=0x7fffd587aa00, aChromeMask=4094, aInitialWidth=1, aInitialHeight=1, aIsHiddenWindow=<optimized out>, aOpeningTab=0x0, aResult=<optimized out>) at /home/erahm/dev/mozilla-central/xpfe/appshell/nsAppShellService.cpp:742 > #13 0x00007fffebf39c55 in nsAppShellService::CreateHiddenWindowHelper(bool) (this=0x7fffd46a3340, aIsPrivate=<optimized out>) > at /home/erahm/dev/mozilla-central/xpfe/appshell/nsAppShellService.cpp:140 > #14 0x00007fffec1a4752 in nsAppStartup::CreateHiddenWindow() (this=<optimized out>) > at /home/erahm/dev/mozilla-central/toolkit/components/startup/nsAppStartup.cpp:245 > #15 0x00007fffec200fc4 in XREMain::XRE_mainRun() (this=0x7fffffffc680) > at /home/erahm/dev/mozilla-central/toolkit/xre/nsAppRunner.cpp:4273 > #16 0x00007fffec201720 in XREMain::XRE_main(int, char**, nsXREAppData const*) (this=0x7fffffffc680, argc=<optimized out>, argv=<optimized out>, aAppData=<optimized out>) at /home/erahm/dev/mozilla-central/toolkit/xre/nsAppRunner.cpp:4441 > #17 0x00007fffec201c48 in XRE_main(int, char**, nsXREAppData const*, uint32_t) (argc=-18265, argv=0x7fffffffb8a6, aAppData=0x7fffeea55be0 <vtable for nsGlobalChromeWindow+1128>, aFlags=<optimized out>) at /home/erahm/dev/mozilla-central/toolkit/xre/nsAppRunner.cpp:4547 > #18 0x000000000040513e in main(int, char**, char**) (argc=5, argv=<optimized out>, envp=0x7fffffffda68, xreDirectory=<optimized out>) > at /home/erahm/dev/mozilla-central/browser/app/nsBrowserApp.cpp:220 > #19 0x000000000040513e in main(int, char**, char**) (argc=5, argv=<optimized out>, envp=0x7fffffffda68) > at /home/erahm/dev/mozilla-central/browser/app/nsBrowserApp.cpp:360

Ok. Moving the keyboard indicator/etc handling onto the outer window probably makes sense here. I'm not going to have time to do it in the short term though.

Since we're aiming for a "ship no regressions" policy, can you help mentor someone on this, Kyle?

Yes.

I can work on this bug.

Bug 1174789 removed method |nsGlobalWindow::GetKeyboardIndicator| but similar logs as below still show on the latest m-c. I'll refer to bug 1174798 for the fix and keep checking other places printing similar logs. === [Child 13117] WARNING: No inner window available!: file /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp, line 10026 [Parent 13039] WARNING: No inner window available!: file /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp, line 9694

(In reply to Ben Tian [:btian] from comment #15) > [Child 13117] WARNING: No inner window available!: file > /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp, > line 10026 Callstack of above log while testing dom/html/test/test_fullscreen-api.html Breakpoint 1, nsGlobalWindow::ShouldShowFocusRing (this=0x7fffccd34000) at /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp:10026 10026 FORWARD_TO_INNER(ShouldShowFocusRing, (), false); (gdb) bt #0 nsGlobalWindow::ShouldShowFocusRing (this=0x7fffccd34000) at /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp:10026 #1 0x00007fffefb8324b in nsGlobalWindow::SetKeyboardIndicators (this=0x7fffccd34000, aShowAccelerators=UIStateChangeType_Set, aShowFocusRings=UIStateChangeType_Set) at /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp:10049 #2 0x00007ffff1815f8e in mozilla::dom::TabChild::Init (this=0x7fffd297dc00) at /home/bentian/WORKSPACE/projects/gecko/central/dom/ipc/TabChild.cpp:859 #3 0x00007ffff18146a6 in mozilla::dom::TabChild::Create (aManager=0x7fffe397e9d8, aTabId=..., aContext=..., aChromeFlags=2148798462) at /home/bentian/WORKSPACE/projects/gecko/central/dom/ipc/TabChild.cpp:579 #4 0x00007ffff182ddd8 in mozilla::dom::nsIContentChild::AllocPBrowserChild (this=0x7fffe397e9d8, aTabId=..., aContext=..., aChromeFlags=@0x7fffffffc940: 2148798462, aCpID=..., aIsForApp=@0x7fffffffc8b0: false, aIsForBrowser=@0x7fffffffc920: false) at /home/bentian/WORKSPACE/projects/gecko/central/dom/ipc/nsIContentChild.cpp:61 #5 0x00007ffff17c9425 in mozilla::dom::ContentChild::AllocPBrowserChild (this=0x7fffe397e020, aTabId=..., aContext=..., aChromeFlags=@0x7fffffffc940: 2148798462, aCpID=..., aIsForApp=@0x7fffffffc8b0: false, aIsForBrowser=@0x7fffffffc920: false) at /home/bentian/WORKSPACE/projects/gecko/central/dom/ipc/ContentChild.cpp:1527 #6 0x00007fffef05af5d in mozilla::dom::PContentChild::OnMessageReceived (this=0x7fffe397e020, msg__=...) at /home/bentian/WORKSPACE/projects/gecko/central/obj-x86_64-pc-linux-gnu/ipc/ipdl/PContentChild.cpp:7424 #7 0x00007fffee9dc315 in mozilla::ipc::MessageChannel::DispatchAsyncMessage (this=0x7fffe397e088, aMsg=...) at /home/bentian/WORKSPACE/projects/gecko/central/ipc/glue/MessageChannel.cpp:1658 #8 0x00007fffee9dbe00 in mozilla::ipc::MessageChannel::DispatchMessage(IPC::Message&&) (this=0x7fffe397e088, aMsg=<unknown type in /home/bentian/WORKSPACE/projects/gecko/central/obj-x86_64-pc-linux-gnu/dist/bin/libxul.so, CU 0x20e9db9, DIE 0x2186531>) at /home/bentian/WORKSPACE/projects/gecko/central/ipc/glue/MessageChannel.cpp:1596 #9 0x00007fffee9dbb26 in mozilla::ipc::MessageChannel::OnMaybeDequeueOne (this=0x7fffe397e088) at /home/bentian/WORKSPACE/projects/gecko/central/ipc/glue/MessageChannel.cpp:1563 #10 0x00007fffee9f93c6 in mozilla::detail::RunnableMethodArguments<>::applyImpl<mozilla::ipc::MessageChannel, bool (mozilla::ipc::MessageChannel::*)()>(mozilla::ipc::MessageChannel*, bool (mozilla::ipc::MessageChannel::*)(), mozilla::Tuple<>&, mozilla::IndexSequence<>) (o=0x7fffe397e088, m=(bool (mozilla::ipc::MessageChannel::*)(mozilla::ipc::MessageChannel * const)) 0x7fffee9db9b0 <mozilla::ipc::MessageChannel::OnMaybeDequeueOne()>, args=...) at /home/bentian/WORKSPACE/projects/gecko/central/obj-x86_64-pc-linux-gnu/dist/include/nsThreadUtils.h:729 #11 0x00007fffee9f908a in mozilla::detail::RunnableMethodArguments<>::apply<mozilla::ipc::MessageChannel, bool (mozilla::ipc::MessageChannel::*)()>(mozilla::ipc::MessageChannel*, bool (mozilla::ipc::MessageChannel::*)()) ( this=0x7fffe396d6b8, o=0x7fffe397e088, m=(bool (mozilla::ipc::MessageChannel::*)(mozilla::ipc::MessageChannel * const)) 0x7fffee9db9b0 <mozilla::ipc::MessageChannel::OnMaybeDequeueOne()>) at /home/bentian/WORKSPACE/projects/gecko/central/obj-x86_64-pc-linux-gnu/dist/include/nsThreadUtils.h:736 #12 0x00007fffee9f8b5d in mozilla::detail::RunnableMethodImpl<bool (mozilla::ipc::MessageChannel::*)(), false, true>::Run (this=0x7fffe396d680) at /home/bentian/WORKSPACE/projects/gecko/central/obj-x86_64-pc-linux-gnu/dist/include/nsThreadUtils.h:764 #13 0x00007fffee9e7789 in mozilla::ipc::MessageChannel::RefCountedTask::Run (this=0x7fffe394b810) at /home/bentian/WORKSPACE/projects/gecko/central/obj-x86_64-pc-linux-gnu/dist/include/mozilla/ipc/MessageChannel.h:476 #14 0x00007fffee9e7998 in mozilla::ipc::MessageChannel::DequeueTask::Run (this=0x7fffe39b9bb0) at /home/bentian/WORKSPACE/projects/gecko/central/obj-x86_64-pc-linux-gnu/dist/include/mozilla/ipc/MessageChannel.h:495 #15 0x00007fffee21d9e5 in nsThread::ProcessNextEvent (this=0x7fffe228ab00, aMayWait=false, aResult=0x7fffffffce8f) at /home/bentian/WORKSPACE/projects/gecko/central/xpcom/threads/nsThread.cpp:1068 #16 0x00007fffee286a67 in NS_ProcessNextEvent (aThread=0x7fffe228ab00, aMayWait=false) at /home/bentian/WORKSPACE/projects/gecko/central/xpcom/glue/nsThreadUtils.cpp:290 #17 0x00007fffee9dff83 in mozilla::ipc::MessagePump::Run (this=0x7fffe394cd80, aDelegate=0x7fffffffd220) at /home/bentian/WORKSPACE/projects/gecko/central/ipc/glue/MessagePump.cpp:100 #18 0x00007fffee9e0957 in mozilla::ipc::MessagePumpForChildProcess::Run (this=0x7fffe394cd80, aDelegate=0x7fffffffd220) at /home/bentian/WORKSPACE/projects/gecko/central/ipc/glue/MessagePump.cpp:317 #19 0x00007fffee95048d in MessageLoop::RunInternal (this=0x7fffffffd220) at /home/bentian/WORKSPACE/projects/gecko/central/ipc/chromium/src/base/message_loop.cc:235 #20 0x00007fffee950422 in MessageLoop::RunHandler (this=0x7fffffffd220) at /home/bentian/WORKSPACE/projects/gecko/central/ipc/chromium/src/base/message_loop.cc:228

(In reply to Ben Tian [:btian] from comment #15) > [Child 13117] WARNING: No inner window available!: file > /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp, > line 10026 This log only shows during TabChild::Init as comment 16. I'll confirm whether |mInnerWindow| should be nullptr then. If it's normal I'll attach a patch to remove the warning. > [Parent 13039] WARNING: No inner window available!: file > /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp, > line 9694 This log shows during GC as following. === Breakpoint 1, nsGlobalWindow::GetExistingListenerManager (this=0x7fffd2421800) at /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp:9694 9694 FORWARD_TO_INNER(GetExistingListenerManager, (), nullptr); (gdb) bt #0 nsGlobalWindow::GetExistingListenerManager (this=0x7fffd2421800) at /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp:9694 #1 0x00007fffe6facd85 in mozilla::dom::TraceBlackJS (aTrc=0x7fffdcacf5e0, aGCNumber=203, aIsShutdownGC=false) at /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsCCUncollectableMarker.cpp:496 #2 0x00007fffe652dfc3 in XPCJSRuntime::TraceNativeBlackRoots (this=0x7fffe0753800, trc=0x7fffdcacf5e0) at /home/bentian/WORKSPACE/projects/gecko/central/js/xpconnect/src/XPCJSRuntime.cpp:619 #3 0x00007fffe545d910 in mozilla::CycleCollectedJSRuntime::TraceBlackJS (aTracer=0x7fffdcacf5e0, aData=0x7fffe0753800) at /home/bentian/WORKSPACE/projects/gecko/central/xpcom/base/CycleCollectedJSRuntime.cpp:788 #4 0x00007fffeb7731ef in js::gc::GCRuntime::markRuntime (this=0x7fffdcacd600, trc=0x7fffdcacf5e0, traceOrMark=js::gc::GCRuntime::MarkRuntime, lock=...) at /home/bentian/WORKSPACE/projects/gecko/central/js/src/gc/RootMarking.cpp:343 #5 0x00007fffeb2e1b75 in js::gc::GCRuntime::beginMarkPhase (this=0x7fffdcacd600, reason=JS::gcreason::FULL_GC_TIMER, lock=...) at /home/bentian/WORKSPACE/projects/gecko/central/js/src/jsgc.cpp:3910 #6 0x00007fffeb2e8d91 in js::gc::GCRuntime::incrementalCollectSlice (this=0x7fffdcacd600, budget=..., reason=JS::gcreason::FULL_GC_TIMER, lock=...) at /home/bentian/WORKSPACE/projects/gecko/central/js/src/jsgc.cpp:5893 #7 0x00007fffeb2e9ba8 in js::gc::GCRuntime::gcCycle (this=0x7fffdcacd600, nonincrementalByAPI=false, budget=..., reason=JS::gcreason::FULL_GC_TIMER) at /home/bentian/WORKSPACE/projects/gecko/central/js/src/jsgc.cpp:6181 #8 0x00007fffeb2ea108 in js::gc::GCRuntime::collect (this=0x7fffdcacd600, nonincrementalByAPI=false, budget=..., reason=JS::gcreason::FULL_GC_TIMER) at /home/bentian/WORKSPACE/projects/gecko/central/js/src/jsgc.cpp:6291 #9 0x00007fffeb2ea55c in js::gc::GCRuntime::startGC (this=0x7fffdcacd600, gckind=GC_NORMAL, reason=JS::gcreason::FULL_GC_TIMER, millis=0) at /home/bentian/WORKSPACE/projects/gecko/central/js/src/jsgc.cpp:6364 #10 0x00007fffeb2ed90c in JS::StartIncrementalGC (cx=0x7fffdcacd000, gckind=GC_NORMAL, reason=JS::gcreason::FULL_GC_TIMER, millis=0) at /home/bentian/WORKSPACE/projects/gecko/central/js/src/jsgc.cpp:7245 #11 0x00007fffe70a7181 in nsJSContext::GarbageCollectNow (aReason=JS::gcreason::FULL_GC_TIMER, aIncremental=nsJSContext::IncrementalGC, aShrinking=nsJSContext::NonShrinkingGC, aSliceMillis=0) at /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsJSEnvironment.cpp:1223 #12 0x00007fffe70a6fec in FullGCTimerFired (aTimer=0x7fffc99fd660, aClosure=0x0) at /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsJSEnvironment.cpp:1175 #13 0x00007fffe554ab6a in nsTimerImpl::Fire (this=0x7fffc99fd660) at /home/bentian/WORKSPACE/projects/gecko/central/xpcom/threads/nsTimerImpl.cpp:524 #14 0x00007fffe55237ab in nsTimerEvent::Run (this=0x7fffd33600c8) at /home/bentian/WORKSPACE/projects/gecko/central/xpcom/threads/TimerThread.cpp:286 #15 0x00007fffe55296eb in nsThread::ProcessNextEvent (this=0x7ffff6b7b500, aMayWait=false, aResult=0x7fffffffc08f) at /home/bentian/WORKSPACE/projects/gecko/central/xpcom/threads/nsThread.cpp:1068 #16 0x00007fffe559276d in NS_ProcessNextEvent (aThread=0x7ffff6b7b500, aMayWait=false) at /home/bentian/WORKSPACE/projects/gecko/central/xpcom/glue/nsThreadUtils.cpp:290 #17 0x00007fffe5cebc89 in mozilla::ipc::MessagePump::Run (this=0x7fffe218dc00, aDelegate=0x7ffff6b426e0)

(In reply to Ben Tian [:btian] from comment #17) > This log only shows during TabChild::Init as comment 16. I'll confirm > whether |mInnerWindow| should be nullptr then. If it's normal I'll attach a > patch to remove the warning. *remove the warning when TabChild::Init calls into the function.

(In reply to Ben Tian [:btian] from comment #17) > This log only shows during TabChild::Init as comment 16. I'll confirm > whether |mInnerWindow| should be nullptr then. If it's normal I'll attach a > patch to remove the warning. |mInnerWindow| becomes non-nullptr once it's assigned in [1] after TabChild::Init. [1] https://dxr.mozilla.org/mozilla-central/source/dom/base/nsGlobalWindow.cpp#2826 Attached WIP patch removes warning below by skipping ShouldShowFocusRing check. > [Child 13117] WARNING: No inner window available!: file > /home/bentian/WORKSPACE/projects/gecko/central/dom/base/nsGlobalWindow.cpp, > line 10026

Change: - rename |InitKeyboardIndicators| to |SetInitialKeyboardIndicators|.

(In reply to Ben Tian [:btian] from comment #20) > WIP Patch (v2): Remove "no inner window" warning during TabChild::Init() by > skipping ShouldShowFocusRing check The following shows remaining and reduced # of "no inner window" warning, with attached patch, on the most prevalent tests in comment 1: > 2 (-42) - dom/html/test/test_fullscreen-api.html > 0 (-30) - dom/plugins/test/mochitest/test_pluginstream_err.html > 0 (-84) - browser/components/sessionstore/test/browser_formdata_cc.js > 2 (-34) - browser/components/privatebrowsing/test/browser/browser_privatebrowsing_windowtitle.js > 0 (-80) - browser/base/content/test/general/browser_overflowScroll.js > 0 (-76) - dom/manifest/test/browser_ManifestObtainer_obtain.js > 0 (-4) - browser/base/content/test/general/browser_sanitizeDialog.js > 0 (-4) - toolkit/components/passwordmgr/test/browser/browser_passwordmgrdlg.js

Comment on attachment 8771329 [details] [diff] [review] Patch 1 (v2): Remove "no inner window" warning during TabChild::Init() by skipping ShouldShowFocusRing check Kyle, Can you review my patch that removes "no inner window" warning during TabChild::Init by skipping |ShouldShowFocusRing| check? Bug 1174798 change removed comment 1 warning but other ones remain as comment 15. This patch removes the more frequent one with comment 16 call stack, and comment 21 lists reduced # of warning on comment 1 tests. Try result is in https://treeherder.mozilla.org/#/jobs?repo=try&revision=2c253a9322ec

Why is skipping the check ok?

Comment on attachment 8771329 [details] [diff] [review] Patch 1 (v2): Remove "no inner window" warning during TabChild::Init() by skipping ShouldShowFocusRing check Review of attachment 8771329 [details] [diff] [review]: ----------------------------------------------------------------- ::: dom/base/nsGlobalWindow.cpp @@ +3695,5 @@ > +void > +nsPIDOMWindowOuter::SetInitialKeyboardIndicators( > + UIStateChangeType aShowAccelerators, UIStateChangeType aShowFocusRings) > +{ > + MOZ_ASSERT(IsOuterWindow()); Please add an assertion that we do *not* have an inner window here.

As :btian explained to me in person, if we trigger the warning, then the function returned false.

Revise per Kyle's suggestion.

Pushed by ryanvm@gmail.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/ca723ddedfe5 Remove "no inner window" warning during TabChild::Init() by skipping ShouldShowFocusRing check. r=khuey

This seems to have caused android crashes, backed out in https://hg.mozilla.org/integration/mozilla-inbound/rev/b4d52dc84179 https://treeherder.mozilla.org/logviewer.html#?job_id=32237465&repo=mozilla-inbound#L2343

So just to sum up previous comments, it looks like this bug has morphed into |nsGlobalWindow::ShouldShowFocusRing| spewing warnings -- |nsGlobalWindow::GetKeyboardIndicator| seems okay (in the not in my top 10 warnings sense) -- and is by far the most verbose warning during testing. Here's the full report: > 23806 WARNING: No inner window available!: file dom/base/nsGlobalWindow.cpp, line 10026 This warning [1] shows up in the following test suites: > 1604 - desktop-test-linux64/debug-mochitest-browser-chrome-e10s-7 bc7 > 1546 - desktop-test-linux64/debug-web-platform-tests-e10s-7 7 > 1506 - desktop-test-linux64/debug-web-platform-tests-e10s-1 1 > 1416 - desktop-test-linux64/debug-mochitest-browser-chrome-e10s-6 bc6 > 1406 - desktop-test-linux64/debug-mochitest-browser-chrome-e10s-3 bc3 > 1394 - desktop-test-linux64/debug-web-platform-tests-e10s-8 8 > 1292 - desktop-test-linux64/debug-mochitest-browser-chrome-e10s-4 bc4 > 1266 - desktop-test-linux64/debug-web-platform-tests-e10s-4 4 > 1260 - desktop-test-linux64/debug-web-platform-tests-e10s-5 5 > 1206 - desktop-test-linux64/debug-web-platform-tests-e10s-9 9 > 1134 - desktop-test-linux64/debug-web-platform-tests-e10s-3 3 > 1030 - desktop-test-linux64/debug-mochitest-browser-chrome-e10s-2 bc2 > 1028 - desktop-test-linux64/debug-web-platform-tests-e10s-2 2 > 1022 - desktop-test-linux64/debug-mochitest-browser-chrome-e10s-1 bc1 > 932 - desktop-test-linux64/debug-web-platform-tests-reftests-e10s Wr > 882 - desktop-test-linux64/debug-mochitest-browser-chrome-e10s-5 bc5 > 504 - desktop-test-linux64/debug-web-platform-tests-e10s-11 11 > 498 - desktop-test-linux64/debug-web-platform-tests-e10s-12 12 > 346 - desktop-test-linux64/debug-web-platform-tests-e10s-6 6 > 268 - desktop-test-linux64/debug-mochitest-clipboard-e10s cl > 268 - desktop-test-linux64/debug-mochitest-e10s-1 1 > 238 - desktop-test-linux64/debug-mochitest-e10s-5 5 > 222 - desktop-test-linux64/debug-web-platform-tests-e10s-10 10 > 190 - desktop-test-linux64/debug-mochitest-e10s-3 3 > 146 - desktop-test-linux64/debug-mochitest-e10s-9 9 > 128 - desktop-test-linux64/debug-mochitest-2 2 > 126 - desktop-test-linux64/debug-mochitest-e10s-8 8 > 126 - desktop-test-linux64/debug-mochitest-e10s-10 10 > 108 - desktop-test-linux64/debug-mochitest-e10s-7 7 > 74 - desktop-test-linux64/debug-mochitest-browser-chrome-5 bc5 > 64 - desktop-test-linux64/debug-firefox-ui-functional-local-e10s en-US > 62 - desktop-test-linux64/debug-mochitest-e10s-4 4 > 58 - desktop-test-linux64/debug-mochitest-e10s-2 2 > 56 - desktop-test-linux64/debug-mochitest-e10s-6 6 > 52 - desktop-test-linux64/debug-mochitest-chrome-1 c1 > 50 - desktop-test-linux64/debug-firefox-ui-functional-remote-e10s en-US > 32 - desktop-test-linux64/debug-mochitest-gpu-e10s gpu > 32 - desktop-test-linux64/debug-mochitest-4 4 > 28 - desktop-test-linux64/debug-mochitest-browser-chrome-2 bc2 > 26 - desktop-test-linux64/debug-mochitest-chrome-3 c3 > 24 - desktop-test-linux64/debug-mochitest-media-e10s mda > 16 - desktop-test-linux64/debug-mochitest-5 5 > 10 - desktop-test-linux64/debug-mochitest-webgl-e10s gl > 8 - desktop-test-linux64/debug-mochitest-8 8 > 8 - desktop-test-linux64/debug-crashtest-e10s C > 8 - desktop-test-linux64/debug-mochitest-clipboard cl > 8 - desktop-test-linux64/debug-mochitest-browser-chrome-6 bc6 > 6 - desktop-test-linux64/debug-mochitest-chrome-2 c2 > 6 - desktop-test-linux64/debug-mochitest-browser-chrome-3 bc3 > 6 - desktop-test-linux64/debug-mochitest-3 3 > 6 - desktop-test-linux64/debug-mochitest-10 10 > 4 - desktop-test-linux64/debug-jsreftest-e10s-1 J1 > 4 - desktop-test-linux64/debug-reftest-e10s-5 R5 > 4 - desktop-test-linux64/debug-reftest-no-accel-e10s-1 Ru1 > 4 - desktop-test-linux64/debug-reftest-no-accel-e10s-8 Ru8 > 4 - desktop-test-linux64/debug-reftest-e10s-4 R4 > 4 - desktop-test-linux64/debug-reftest-e10s-6 R6 > 4 - desktop-test-linux64/debug-reftest-no-accel-e10s-3 Ru3 > 4 - desktop-test-linux64/debug-reftest-e10s-7 R7 > 4 - desktop-test-linux64/debug-reftest-e10s-2 R2 > 4 - desktop-test-linux64/debug-reftest-e10s-3 R3 > 4 - desktop-test-linux64/debug-reftest-no-accel-e10s-7 Ru7 > 4 - desktop-test-linux64/debug-reftest-no-accel-e10s-5 Ru5 > 4 - desktop-test-linux64/debug-jsreftest-e10s-2 J2 > 4 - desktop-test-linux64/debug-reftest-e10s-8 R8 > 4 - desktop-test-linux64/debug-reftest-e10s-1 R1 > 4 - desktop-test-linux64/debug-reftest-no-accel-e10s-4 Ru4 > 4 - desktop-test-linux64/debug-reftest-no-accel-e10s-6 Ru6 > 4 - desktop-test-linux64/debug-reftest-no-accel-e10s-2 Ru2 > 2 - desktop-test-linux64/debug-mochitest-1 1 It shows up in 8073 tests. A few of the most prevalent: > 694 - [e10s] Shutdown > 96 - [e10s] dom/tests/mochitest/bugs/test_bug260264.html > 84 - [e10s] browser/components/sessionstore/test/browser_formdata_cc.js > 78 - [e10s] browser/base/content/test/general/browser_overflowScroll.js > 74 - [e10s] dom/manifest/test/browser_ManifestObtainer_obtain.js > 56 - [e10s] browser/components/sessionstore/test/browser_615394-SSWindowState_events.js > 54 - [e10s] toolkit/mozapps/extensions/test/browser/browser_discovery.js > 54 - [e10s] browser/base/content/test/referrer/browser_referrer_middle_click_in_container.js > 52 - [e10s] browser/components/sessionstore/test/browser_522545.js > 50 - [e10s] browser/base/content/test/referrer/browser_referrer_middle_click.js [1] https://hg.mozilla.org/mozilla-central/annotate/f44bb9de08ad/dom/base/nsGlobalWindow.cpp#l10026

(In reply to Wes Kocher (:KWierso) from comment #28) > This seems to have caused android crashes, backed out in > https://hg.mozilla.org/integration/mozilla-inbound/rev/b4d52dc84179 > https://treeherder.mozilla.org/logviewer.html#?job_id=32237465&repo=mozilla- > inbound#L2343 I'm investigating the android crash. Still figuring how to run mochitest on Android emulator with gdb for call stack.

(In reply to Ben Tian [:btian] from comment #30) > (In reply to Wes Kocher (:KWierso) from comment #28) > > This seems to have caused android crashes, backed out in > > https://hg.mozilla.org/integration/mozilla-inbound/rev/b4d52dc84179 > > https://treeherder.mozilla.org/logviewer.html#?job_id=32237465&repo=mozilla- > > inbound#L2343 The crash seems to result from the MOZ_ASSERT of no inner window added in comment 26, since [1] with the assertion has no crash. The comment 22 try result [2] before the change also matches the finding. [1] https://treeherder.mozilla.org/#/jobs?repo=try&revision=be77d82ffa78&selectedJob=24591353 [2] https://treeherder.mozilla.org/#/jobs?repo=try&revision=2c253a9322ec&selectedJob=24022859 Next I'll check whether method |GetCurrentInnerWindow| is suitable here or |mInnerWindow| is non-nullpter then. If the latter the patch may require further revision.

(In reply to Ben Tian [:btian] from comment #31) > > > https://treeherder.mozilla.org/logviewer.html#?job_id=32237465&repo=mozilla- > > > inbound#L2343 > Next I'll check whether method |GetCurrentInnerWindow| is suitable here or > |mInnerWindow| is non-nullpter then. If the latter the patch may require > further revision. |mInnerWindow| is non-nullpter during TabChild:;Init on Fennec, as the warning msg shows only on linux64 debug testing. A try with quick build flag hack is running to verify. https://treeherder.mozilla.org/#/jobs?repo=try&revision=55e45cac991b

Change: - skip ShouldShowFocusRing check only if |mInnerWindow| is nullptr

Comment on attachment 8775841 [details] [diff] [review] [final] Patch 1: Skip ShouldShowFocusRing check during TabChild::Init() if no inner window is available, r=khuey Kyle, Can you review my updated patch to solve comment 28 crash on Fennec debug build? The patch skips ShouldShowFocusRing check during TabChild::Init() ONLY if |mInnerWindow| is nullptr. So Fennec doesn't skip the check but linux64 does per comment 31 and comment 32. Try result is in https://treeherder.mozilla.org/#/jobs?repo=try&revision=a0290fbf2e33

Pushed by cbook@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/2b634a956b26 Skip ShouldShowFocusRing check during TabChild::Init() if no inner window is available. r=khuey

Thank you very much Ben :) In case you didn't know that, we are deprecating btpp-* whiteborad, and have a plan to map btpp-* into priorities. https://wiki.mozilla.org/Platform/DOM_Bug_Triage

Hi Ben, Since this bug is a regression and also affects 49/50, do you want to uplift this for 49/50 if this patch is not too risky?

I suggest no need to uplift, as the linux64-debug-only log spam has no user impact.

Since there is no user impact, it's just a log spam. Let's let it ride the train on 51. Mark 48/49/50 as won't fix.