55865 2011-03-07 02:13:55 -0800 [EFL] Crash on quit when the Web Inspector debugger is paused 2017-03-11 10:35:48 -0800 1 1 1 Unclassified WebKit Web Inspector 528+ (Nightly build) PC Linux RESOLVED WONTFIX InRadar P2 Major --- 0 m.malecki webkit-unassigned dbates inspector-bugzilla-changes l.slachciak mcatanzaro webkit-bug-importer oldest_to_newest 363273 0 m.malecki 2011-03-07 02:13:55 -0800 The crash happens when Web Inspector is on and javascript is paused. In this case, the debugger calls WebCore::ScriptDebugServer::pauseIfNeeded(), and inside it enters another level of event loop. The problem is that once the loop is exited, the state of data, especially *page->group()->pages().begin(), is unstable, and this page is probably during destruction process or it may even be a dangling pointer. This way, when setJavaScriptPaused with this page results in crash. I suspect the system doesn't predict that the event loop may happen to be exited in this function instead of the main event loop in the application that uses webkit. This behavior can be reproduced with WebKit-efl, as well as with Arora [Webkit-Qt] (although I haven't observed it with Rekonq [Webkit-KDE]): 1. Run Arora with any page (preferably under gdb because without it the crash may be unable to be observed) 2. Open Web Inspector 3. Click "Scripts" and pause javascript (make sure that you can see the backtrace) 4. Close the Arora window I know you can think that it's insignificant that a crash happens when you close the program. The problem is, though, that this problem occurs also when you close a widget realized by WRT engine, and in this case the whole WRT engine crashes, not only the widget being closed. 370079 1 dbates 2011-03-18 19:32:29 -0700 *** This bug has been marked as a duplicate of bug 44230 *** 378049 2 timothy 2011-04-01 13:54:24 -0700 I don't think this is a dupe of bug 44230, that bug is about Mac and about closign the window. This is abotu EFL and Qt in Arora. 378052 3 timothy 2011-04-01 13:55:46 -0700 Reopening. 972071 4 bburg 2014-01-25 16:00:13 -0800 Is this still an issue? Migrating to new component. 972072 5 webkit-bug-importer 2014-01-25 16:00:44 -0800 <rdar://problem/15909919> 1286283 6 mcatanzaro 2017-03-11 10:35:48 -0800 Closing this bug because the EFL port has been removed from trunk. If you feel this bug applies to a different upstream WebKit port and was closed in error, please either update the title and reopen the bug, or leave a comment to request this.