Incident ID: 18563830 Stack Signature nsEventStateManager::HandleAccessKey c6b95f66 Product ID FirefoxTrunk Build ID 2006050405 Trigger Time 2006-05-11 06:46:48.0 Platform Win32 Operating System Windows NT 5.1 build 2600 Module firefox.exe + (00282bf1) URL visited <a href='http://landfill.mozilla.org/ryl/random.html'>http://landfill.mozilla.org/ryl/random.html</a> User Comments trying to crash firefox. i have to do something while i wait for trees to burn Since Last Crash 97041 sec Total Uptime 282921 sec Trigger Reason Access violation Source File, Line No. c:\builds\tinderbox\fx-trunk-cairo\winnt_5.2_depend\mozilla\content\events\src\nseventstatemanager.cpp, line 1077 Stack Trace nsEventStateManager::HandleAccessKey nsDocLoader::AddRef The url is a guess at what could possibly be going wrong. note that i was probably actively loading new windows so it's possible my window changed or something.
Created attachment 227557 [details] [diff] [review] proposed patch Changing assertion to if (). (I think) this crash may happen if window is closed or presshell otherwise destroyed before keypress. And anyway, I see no reason for the assertions in this case. Simple if () is enough. Ofc, I can't reproduce the crash atm, but afaik it is still possible to crash. And this patch should not cause any harm :)
This is also quite a 'popular' crash on the 1.8.1 branch (ranked 3rd, as I write this): http://talkback-public.mozilla.org/reports/firefox/FF2x/index.html I can easily reproduce the crash on latest build Firefox2 branch with this url: http://www.mountyhall.com/mountyhall/Forum and then quickly going back and forth in history, using alt+arrow keys left and right. Olli, you want to ask approval1.8.1? on the patch? I hope/expect this patch would fix it.
I'll test first. But probably this could go into 1.8 and 1.8.0
Created attachment 230277 [details] [diff] [review] for 1.8 This fixes the crash also on 1.8
Comment on attachment 227557 [details] [diff] [review] proposed patch And this patch applies almost cleanly to 1.8.0 branch.
Comment on attachment 230277 [details] [diff] [review] for 1.8 a=drivers. Please go ahead and land this on the branch.
This was checked in to 1.8.1, so remowing blocking1.8.1? flag
Comment on attachment 227557 [details] [diff] [review] proposed patch approved for 1.8.0 branch, a=dveditz for drivers