Re: [Gtk-osx-users] Fwd: Bluefish crash
- From: John Ralls <jralls ceridwen fremont ca us>
- To: GTK+-2 OSX Users <gtk-osx-users lists sourceforge net>
- Subject: Re: [Gtk-osx-users] Fwd: Bluefish crash
- Date: Wed, 2 Mar 2011 16:56:41 -0800
On Mar 2, 2011, at 5:46 AM, Olivier Sessink wrote:
> Hi all,
>
> I received a bugreport with backtrace, and looking at the backtrace
> below I get the feeling that the crash might be in the GtkOSXApplication
> integration code. The NS_ and CFX_ functions seem to be
> Carbon/Quartz/OSX functions that finally call a gtk_ function again.
>
> Anybody an idea where to start debugging this?
>
> Olivier
> -------- Original Message --------
> Subject: Bluefish crash
> Date: Tue, 1 Mar 2011 16:28:12 -0500
>
> Olivier,
>
> I had a project open - all files were saved.
> I was exiting Bluefish normally - I got a wait cursor - and then the
> following occurred:
>
> (running MacOSX 10.6.6 Bluefish 2.0.3-rc2-3.dmg)
>
> Process: bluefish [159]
> Path: /Applications/Bluefish.app/Contents/MacOS/bluefish
> Identifier: nl.openoffice.bluefish
> Version: 2.0.1 (1)
> Code Type: X86 (Native)
> Parent Process: launchd [101]
>
> Date/Time: 2011-03-01 16:20:18.387 -0500
> OS Version: Mac OS X 10.6.6 (10J567)
> Report Version: 6
>
> Interval Since Last Report: 114981 sec
> Crashes Since Last Report: 1
> Per-App Interval Since Last Report: 62140 sec
> Per-App Crashes Since Last Report: 1
> Anonymous UUID: 08700A3A-670E-49C5-8182-DC2ED2CC37F4
>
> Exception Type: EXC_BAD_ACCESS (SIGBUS)
> Exception Codes: KERN_PROTECTION_FAILURE at 0x0000000000000006
> Crashed Thread: 0 Dispatch queue: com.apple.main-thread
>
> Thread 0 Crashed: Dispatch queue: com.apple.main-thread
> 0 libgtk-quartz-2.0.0.dylib 0x00505d63 gtk_target_list_find + 65
> 1 libgtk-quartz-2.0.0.dylib 0x0050afef -[GtkDragSourceOwner
> pasteboard:provideDataForType:] + 183
> 2 com.apple.AppKit 0x9456e10f __NSPasteboardProvideData
> + 176
No, it's a bug in the gtk-quartz drag'n drop code. What's happened is that the user has
done a drag and then closed the window that was the source. At this point, Gtk *should* recognize
that the source is no longer available and either load the OSX clipboard with the data types that haven't
already been provided or tell OSX to clear the clipboard entirely. It does neither (I've made a couple of passes at getting it to clear the clipboard, but haven't figured out the right magic). Later, when the App shuts down, OSX sends an event to retrieve the rest of the data types, but the owner object has been deleted, causing an access violation and the crash.
Regards,
John Ralls
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]