The odds are something failed and we hit a case where a null pointer wasn't expected. We died in: #0 0x400254ee in nsWebShell::OnConnectionsComplete (this=0x8175bb0) at nsWebShell.cpp:2126 2126 NS_RELEASE(presContext); (gdb) p presContext $1 = (suspicious *) 0x0 (gdb) l 2116 if (nsnull != mContentViewer && 2117 NS_OK == mContentViewer->QueryInterface(kIDocumentViewerIID, (void**)&docViewer)) { 2118 nsIPresContext *presContext; 2119 if (NS_OK == docViewer->GetPresContext(presContext)) { 2120 nsEventStatus status = nsEventStatus_eIgnore; 2121 nsMouseEvent event; 2122 event.eventStructType = NS_EVENT; 2123 event.message = NS_PAGE_LOAD; 2124 rv = mScriptGlobal->HandleDOMEvent(*presContext, &event, nsnull, NS_EVENT_FLAG_INIT, status); 2125 2126 NS_RELEASE(presContext); 2127 } 2128 NS_RELEASE(docViewer); 2129 } 2130 } 2131 2132 /* 2133 *Fire the EndLoadURL(...) notification... 2134 */ 2135 if (((nsnull != mURLListener) || (nsnull != mContainer)) && (nsnull != mContentViewer)) { Here is the stack trace: #0 0x400254ee in nsWebShell::OnConnectionsComplete (this=0x8175bb0) at nsWebShell.cpp:2126 #1 0x4001d0e7 in nsDocLoaderImpl::AreAllConnectionsComplete (this=0x8106de8) at nsDocLoader.cpp:1481 #2 0x4001d105 in nsDocLoaderImpl::AreAllConnectionsComplete (this=0x92045e0) at nsDocLoader.cpp:1490 #3 0x4001d029 in nsDocLoaderImpl::LoadURLComplete (this=0x92045e0, aURL=0x9208a70, aBindInfo=0x914dac8, aStatus=-2142568447) at nsDocLoader.cpp:1451 #4 0x4001e219 in nsDocumentBindInfo::OnStopBinding (this=0x914dac8, aURL=0x9208a70, aStatus=2152398849, aMsg=0xbfffe888) at nsDocLoader.cpp:2000 #5 0x40616aef in bam_exit_routine (URL_s=0x8c3aa68, status=-201, window_id=0x8d95920) at nsNetService.cpp:1403 #6 0x4063a961 in net_CallExitRoutine ( exit_routine=0x406169e4 <bam_exit_routine(URL_Struct_ *, int, MWContext_ *) at nsNetService.cpp:1371>, URL_s=0x8c3aa68, status=-201, format_out=102, window_id=0x8d95920) at mkgeturl.c:1142 #7 0x4063e3c1 in net_InterruptActiveStream (entry=0x94f95e8, show_warning=1 '\001') at mkgeturl.c:3514 #8 0x4063e582 in NET_InterruptStream (nurl=0x8c3aa68) at mkgeturl.c:3593 #9 0x40615526 in nsNetlibService::InterruptStream (this=0x806ee60, aURL=0x9208a70) at nsNetService.cpp:636 #10 0x4001dbe0 in nsDocumentBindInfo::Stop (this=0x914dac8) at nsDocLoader.cpp:1767 #11 0x4001d236 in nsDocLoaderImpl::StopBindInfoEnumerator (aElement=0x914dac8, aData=0x0) at nsDocLoader.cpp:1515 #12 0x4082d19a in nsSupportsArray::EnumerateForwards (this=0x8c77d80, aFunc=0x4001d1e8 <nsDocLoaderImpl::StopBindInfoEnumerator(nsISupports *, void *) at nsDocLoader.cpp:1509>, aData=0x0) at nsSupportsArray.cpp:318 #13 0x4001c51f in nsDocLoaderImpl::Stop (this=0x92045e0) at nsDocLoader.cpp:1024 #14 0x40023f90 in nsWebShell::Stop (this=0x9615330) at nsWebShell.cpp:1465 #15 0x4002263f in nsWebShell::Destroy (this=0x9615330) at nsWebShell.cpp:828 #16 0x40391ea0 in nsHTMLFrameInnerFrame::~nsHTMLFrameInnerFrame (this=0x94acd18, __in_chrg=3) at nsFrameFrame.cpp:421 #17 0x402d1a54 in nsFrame::DeleteFrame (this=0x94acd18, aPresContext=@0x9689410) at nsFrame.cpp:379 #18 0x4029a4f3 in nsFrameList::DeleteFrames (this=0x943464c, aPresContext=@0x9689410) at nsFrameList.cpp:28 #19 0x402cfcfd in nsContainerFrame::DeleteFrame (this=0x9434610, aPresContext=@0x9689410) at nsContainerFrame.cpp:78 #20 0x4029a4f3 in nsFrameList::DeleteFrames (this=0x8f55144, aPresContext=@0x9689410) at nsFrameList.cpp:28 #21 0x402cfcfd in nsContainerFrame::DeleteFrame (this=0x8f55108, aPresContext=@0x9689410) at nsContainerFrame.cpp:78 #22 0x402e732f in nsLineBox::DeleteLineList (aPresContext=@0x9689410, aLine=0x8c50c70) at nsLineBox.cpp:233 #23 0x402c476f in nsBlockFrame::DeleteFrame (this=0x8f688a8, aPresContext=@0x9689410) at nsBlockFrame.cpp:488 #24 0x402c22fa in nsAreaFrame::DeleteFrame (this=0x8f688a8, aPresContext=@0x9689410) at nsAreaFrame.cpp:101 #25 0x4029a4f3 in nsFrameList::DeleteFrames (this=0x96c4a34, aPresContext=@0x9689410) at nsFrameList.cpp:28 #26 0x402cfcfd in nsContainerFrame::DeleteFrame (this=0x96c49f8, aPresContext=@0x9689410) at nsContainerFrame.cpp:78 #27 0x4029a4f3 in nsFrameList::DeleteFrames (this=0x94a91a4, aPresContext=@0x9689410) at nsFrameList.cpp:28 #28 0x402cfcfd in nsContainerFrame::DeleteFrame (this=0x94a9168, aPresContext=@0x9689410) at nsContainerFrame.cpp:78 #29 0x4030076e in ViewportFrame::DeleteFrame (this=0x94a9168, aPresContext=@0x9689410) at nsViewportFrame.cpp:111 #30 0x402effaf in PresShell::~PresShell (this=0x8b60ed0, __in_chrg=3) at nsPresShell.cpp:529 #31 0x402efd50 in PresShell::Release (this=0x8b60ed0) at nsPresShell.cpp:471 #32 0x4001986d in DocumentViewerImpl::~DocumentViewerImpl (this=0x95c6cf0, __in_chrg=3) at nsDocumentViewer.cpp:216 #33 0x40019595 in DocumentViewerImpl::Release (this=0x95c6cf0) at nsDocumentViewer.cpp:153 #34 0x4002213f in nsWebShell::Embed (this=0x8175bb0, aContentViewer=0x952b280, aCommand=0x95e2190 "view", aExtraInfo=0x0) at nsWebShell.cpp:674 #35 0x4001df69 in nsDocumentBindInfo::OnStartBinding (this=0x8acd918, aURL=0x94d6ac0, aContentType=0x9537318 "text/html") at nsDocLoader.cpp:1892 #36 0x4061d59b in NET_NGLayoutConverter (format_out=38, converter_obj=0x0, URL_s=0x96ab508, context=0x962d690) at nsStubContext.cpp:763 #37 0x40646ebd in NET_StreamBuilder (format_out=38, URL_s=0x96ab508, context=0x962d690) at mkstream.c:237 #38 0x405f08eb in NET_CacheConverter (format_out=38, converter_obj=0x0, URL_s=0x96ab508, window_id=0x962d690) at mkcache.c:1689 #39 0x40646ebd in NET_StreamBuilder (format_out=102, URL_s=0x96ab508, context=0x962d690) at mkstream.c:237 #40 0x405d31a4 in net_setup_http_stream (ce=0x964bf58) at mkhttp.c:2804 #41 0x405d4291 in net_ProcessHTTP (ce=0x964bf58) at mkhttp.c:3477 #42 0x4063de53 in NET_ProcessNet (ready_fd=0x833ce98, fd_type=2) at mkgeturl.c:3312 #43 0x40645d31 in NET_PollSockets () at mkselect.c:298 #44 0x406168c2 in nsNetlibService::NetPollSocketsCallback (aTimer=0x82192e0, aClosure=0x806ee60) at nsNetService.cpp:1220 #45 0x404bedf9 in TimerImpl::FireTimeout (this=0x82192e0) at nsTimer.cpp:73 #46 0x404bf2e2 in nsTimerExpired (aCallData=0x82192e0) at nsTimer.cpp:189 #47 0x409ee990 in g_timeout_dispatch (source_data=0x94fa9e8, current_time=0xbffff6e0, user_data=0x82192e0) at gmain.c:1147 #48 0x409edc83 in g_main_dispatch (current_time=0xbffff6e0) at gmain.c:647 #49 0x409ee20f in g_main_iterate (block=1, dispatch=1) at gmain.c:854 #50 0x409ee391 in g_main_run (loop=0x81d2cb8) at gmain.c:912 #51 0x4091b44b in gtk_main () at gtkmain.c:475 #52 0x40049b94 in nsAppShell::Run (this=0x809c2d8) at nsAppShell.cpp:207 #53 0x8053882 in nsNativeViewerApp::Run (this=0x80953f0) at nsGTKMain.cpp:42 #54 0x8053a96 in main (argc=3, argv=0xbffff788) at nsGTKMain.cpp:97

Setting target milestone to M4 for now...

I tried www.yahoo.com in both apprunner and viewer and they did not crash. I tried running the debug robot in viewer and it shows a dialog on which only the widgets repaint and no urls get loaded in viewer. Am moving this bug to M5 for further analysis, given that www.yahoo.com loads fine if we type the url in the url bar.

nisheeth, was this Linux specific?

I see the behavior described in my earlier post on Linux and Win32. I haven't looked at the Mac, yet, but because www.yahoo.com is part of the smoke tests for the daily build, I am assuming that the page does not crash the Mac builds.

could i get some details about how are you reproducing this bug? -thanx

I just tried this again on the Linux build of viewer and it seems like the debug robot is broken. The steps I took are: 1) Choose Debug..Debug Robot from the menu 2) Hit Start Nothing happens. Hitting cancel does not get rid of the Debug Robot dialog. As I said in my earlier post going to www.yahoo.com by typing the url directly into the url bar does not crash viewer. So am setting the milestone to M6. I'll look into why the debug robot is not doing its thing for M6...

Putting on [PP] radar.

Given other higher priority bugs, this needs to move to M8...

Updating summary...

Kevin recently checked in a bunch of fixes to the debug robot. Please retest this on all platforms. Thanks.

Moving out to M10...

Claudius or kipp, can you test and see if this is still broken on Linux? I'd do this myself but I'm working from home right now and will leave for vacation tomorrow. If this is fixed, please close the bug. If not, I'll take a look at this when I get back. Thanks.

I can't get Viewer to run on Linux right now and my Mac viewer (19990830) wouldn't run the robot at all. So I think I'll be stuck on this one for a while.

Setting milestone to M15...

Does anybody care about viewer's debug robot anymore? Troy and buster, what do you think about latering this bug? I think this is a relatively low priority bug because we have other ways in which we can traverse web pages automatically (chofmann's browser buster tests, for example).

The robot is the automated web crawler, right? I think this is an important tool that ought to work, and should be fixed. I feel the same way about viewer -f <file of urls>, which is also broken (I hope for the same reason) which I use constantly when it works. Without the web crawler, we run the risk of always testing the same n pages and not getting sufficient breadth of testing.

Moving bugs out by one milestone...

non-essential for m16

Marking M18...

Marking nsbeta3-

Updating QA Contact

Chris, this bug has been lying on my plate for a long time. I haven't had the cycles to fix it and probably won't in the near future. Passing it on to you to see if you or somebody else in the layout team has cycles to work on it. Also marking it helpwanted. In case you don't find an owner, please pass it back to me and I'll future it. Thanks!

Nisheeth, sorry but I have to give it back.

This bug has been marked "future" because the original netscape engineer working on this is over-burdened. If you feel this is an error, that you or another known resource will be working on this bug,or if it blocks your work in some way -- please attach your concern to the bug for reconsideration.

Correction: Changing QA contact for the Embed API bugs to David Epstein.

Cleaning up Status Whiteboard.

Comment #15 From Nisheeth Ranjan 2000-03-14 17:03 states that the robot may not be needed > bug because we have other ways in which we can traverse web pages automatically > (chofmann's browser buster tests, for example). Comment #16 From buster@formerly-netscape.com.tld 2000-03-14 17:13 thinks > this is an important tool that ought to work, and should be fixed. following comments are administrative only, last one 2001-10-16 10:43 PST Cleaning up Status Whiteboard. I suggest resolving this bug as WONTFIX, as there has been no action for nearly 5 years, and little (visible) action before.

marking wontfix as old Netcape bug (i don't think we use the robot tool anymore)