341551 - document.write() without document.close() in popup: ###!!! ASSERTION: EndLoad called early: 'mWriteState == eNotWriting || mWriteState == ePendingClose || mWriteState == eDocumentClosed', file c:/mozilla.org/baseline/mozilla/content

Status: RESOLVED DUPLICATE of bug 332896

(Core :: DOM: Core & HTML, defect)

Description

[Alex Vincent [:WeirdAl]]

Steps to reproduce:
(1) SeaMonkey trunk, load https://bugzilla.mozilla.org/attachment.cgi?id=133934 (from bug 223393)
(2) Close the popup window.

I claim this assertion is bogus; although document.close() dates back to DOM Level 0, web page authors don't always use it.  There are four possible values for mWriteState, and this assertion covers three of them; we fired for the fourth.

 	ntdll.dll!7c901230() 	
 	[Frames below may be incorrect and/or missing, no symbols loaded for ntdll.dll]	
>	xpcom_core.dll!Break(const char * aMsg=0x0012e848)  Line 471	C++
 	xpcom_core.dll!NS_DebugBreak_P(unsigned int aSeverity=1, const char * aStr=0x01e248c0, const char * aExpr=0x01e24860, const char * aFile=0x01e24810, int aLine=1072)  Line 350 + 0xc bytes	C++
 	gklayout.dll!nsHTMLDocument::EndLoad()  Line 1072 + 0x40 bytes	C++
 	gklayout.dll!HTMLContentSink::DidBuildModel()  Line 2181	C++
 	gkparser.dll!CNavDTD::DidBuildModel(unsigned int anErrorCode=2152596471, int aNotifySink=1, nsIParser * aParser=0x04e00488, nsIContentSink * aSink=0x04e5eb74)  Line 458	C++
 	gkparser.dll!nsParser::DidBuildModel(unsigned int anErrorCode=2152596471)  Line 948 + 0x35 bytes	C++
 	gkparser.dll!nsParser::Terminate()  Line 1069 + 0xc bytes	C++
 	gklayout.dll!nsDocument::StopDocumentLoad()  Line 1184	C++
 	gklayout.dll!DocumentViewerImpl::Stop()  Line 1634	C++
 	docshell.dll!nsDocShell::Stop(unsigned int aStopFlags=3)  Line 3297	C++
 	docshell.dll!nsDocShell::Stop(unsigned int aStopFlags=3)  Line 3318	C++
 	docshell.dll!nsDocShell::Destroy()  Line 3584	C++
 	appshell.dll!nsXULWindow::Destroy()  Line 507	C++
 	appshell.dll!nsWebShellWindow::Destroy()  Line 834 + 0x9 bytes	C++
 	appshell.dll!nsWebShellWindow::HandleEvent(nsGUIEvent * aEvent=0x0012f0ac)  Line 393	C++
 	gkwidget.dll!nsWindow::DispatchEvent(nsGUIEvent * event=0x0012f0ac, nsEventStatus & aStatus=nsEventStatus_eIgnore)  Line 1102 + 0xc bytes	C++
 	gkwidget.dll!nsWindow::DispatchWindowEvent(nsGUIEvent * event=0x0012f0ac)  Line 1123	C++
 	gkwidget.dll!nsWindow::DispatchStandardEvent(unsigned int aMsg=101)  Line 1142 + 0x11 bytes	C++
 	gkwidget.dll!nsWindow::ProcessMessage(unsigned int msg=16, unsigned int wParam=0, long lParam=0, long * aRetValue=0x0012f498)  Line 4226	C++
 	gkwidget.dll!nsWindow::WindowProc(HWND__ * hWnd=0x001b04b0, unsigned int msg=16, unsigned int wParam=0, long lParam=0)  Line 1291 + 0x1d bytes	C++
 	user32.dll!77d48734() 	
 	user32.dll!77d48816() 	
 	xpcom_core.dll!nsReadingIterator<char>::advance(int n=0)  Line 174 + 0x18 bytes	C++
 	user32.dll!77d4b4c0() 	
 	user32.dll!77d4b50c() 	
 	ntdll.dll!7c90eae3() 	
 	user32.dll!77d494be() 	
 	user32.dll!77d4b42d() 	
 	user32.dll!77d4baa4() 	
 	user32.dll!77d4b96b() 	
 	user32.dll!77d4b3f9() 	
 	uxtheme.dll!5ad73c20() 	
 	uxtheme.dll!5ad8e300() 	
 	uxtheme.dll!5ad71ac7() 	
 	uxtheme.dll!5ad71b3d() 	
 	uxtheme.dll!5ad8e2d5() 	
 	user32.dll!77d4bb15() 	
 	nspr4.dll!_MD_CURRENT_THREAD()  Line 296 + 0xc bytes	C
 	gkwidget.dll!nsWindow::DefaultWindowProc(HWND__ * hWnd=0x001b04b0, unsigned int msg=274, unsigned int wParam=61536, long lParam=1442394)  Line 1312	C++
 	user32.dll!77d48734() 	
 	user32.dll!77d48816() 	
 	gkwidget.dll!nsBaseWidget::QueryInterface(const nsID & aIID={...}, void * * aInstancePtr=0x001b04b0)  Line 64 + 0x84 bytes	C++
 	user32.dll!77d4c63f() 	
 	user32.dll!77d4c665() 	
 	gkwidget.dll!nsWindow::WindowProc(HWND__ * hWnd=0x0646d3d4, unsigned int msg=1243400, unsigned int wParam=38545384, long lParam=0)  Line 1298 + 0x1f bytes	C++
 	nspr4.dll!PR_GetCurrentThread()  Line 175	C
 	gkwidget.dll!nsWindow::`vftable'()  + 0x23 bytes	C++
 	gkwidget.dll!02469aaf() 	
 	4d890cec()

Comment 1

[Blake Kaplan (:mrbkap) (inactive)]

Except that we should only get into EndLoad when we're in that particular state, and the HTML document needs to do more work to make sure we're in it.

*** This bug has been marked as a duplicate of 332896 ***