See upcoming testcase. Current trunk Mozilla builds crash when clicking on the button. Doesn't crash with 2005-09-21 build, crashes with 2005-09-22 build, so my guess this happens because of the fix for bug 1156.

Most of the time, Firefox just exists without showing up of the Talkback Agent, but once I got a Talkback ID: TB10387141M (which doesn't look useful to me)

probably a duplicate of bug 309706

###!!! ASSERTION: Plugin performed illegal operation: 'PR_FALSE', file d:/moz_sr c/mozilla/modules/plugin/base/src/nsPluginHostImpl.cpp, line 6197 I see the assertion but the dialog does not show up under windows, which makes me believe that we shutdown somewhere else with a exit(0) or something. The assert sits inside HandleBadPlugin http://lxr.mozilla.org/seamonkey/source/modules/plugin/base/src/nsPluginHostImpl.cpp#6189

hmmm setting display:table on a <embed> sounds like (Bor)IsSpecialHTMLContent,

forget comment 4, its like Biesi said NTDLL! 7c911230() nsDebugImpl::Assertion(nsDebugImpl * const 0x018cbe18, const char * 0x03e1fc08 `string', const char * 0x03e1dd2c `string', const char * 0x03e1df4c `string', int 0x00001835) line 266 NSGlue_Assertion(const char * 0x03e1fc08 `string', const char * 0x03e1dd2c `string', const char * 0x03e1df4c `string', int 0x00001835) line 107 nsPluginHostImpl::HandleBadPlugin(nsPluginHostImpl * const 0x03279d94, PRLibrary * 0x00000000, nsIPluginInstance * 0x0318beb8) line 6197 + 33 bytes PluginWndProc(HWND__ * 0x000f03ea, unsigned int 0x00000210, unsigned int 0xc1380001, long 0x00040362) line 340 + 241 bytes USER32! 77d18734() USER32! 77d18816() USER32! 77d1c63f() USER32! 77d1e905() NPDSPLAY! 5f749f36() NPDSPLAY! 5f749604() NPDSPLAY! 5f74a3e4() NPDSPLAY! 5f749547() NPDSPLAY! 5f74974f() USER32! 77d18734() USER32! 77d18816() USER32! 77d1c63f() USER32! 77d1e905() PluginWndProc(HWND__ * 0x000f03ea, unsigned int 0x00000210, unsigned int 0xc1380001, long 0x00040362) line 340 + 86 bytes USER32! 77d18734() USER32! 77d18816() USER32! 77d1c63f() USER32! 77d1e905() NPDSPLAY! 5f749f36() NPDSPLAY! 5f749604() NPDSPLAY! 5f74a3e4() NPDSPLAY! 5f749547() NPDSPLAY! 5f74974f() USER32! 77d18734() USER32! 77d18816() USER32! 77d1c63f() USER32! 77d1e905() PluginWndProc(HWND__ * 0x000f03ea, unsigned int 0x00000210, unsigned int 0xc1380001, long 0x00040362) line 340 + 86 bytes and so on....

msg = 0x00000210 WM_PARENTNOTIFY wparam = 0xc1380001 lparam = 0x00040362 http://msdn.microsoft.com/library/default.asp?url=/library/en-us/winui/winui/windowsuserinterface/windowing/windows/windowreference/windowmessages/wm_parentnotify.asp The low word of wparam is 0x0001 WM_CREATE c138 is the identifier of the child window 0x00040362 is the handle of the child window

Is this still a problem, now that bug 309706 is fixed?

NO longer crahses for me.

Gecko/20051031 Firefox/1.6a1 exits on the testcase as before

bernd: well, which hour on that day? I checked that patch in at 10:29 PST.

yep thats fixed with 20051101 *** This bug has been marked as a duplicate of 309706 ***