Using xrandr to change the screen size (or orientation) leads to a crash in xfdesktop. The bug is not 100% reproducible, so it might be some race somewhere in the icon code (the problem doesn't occur when disabling desktop icons). [New Thread 1082132816 (LWP 31584)] Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 47568197576576 (LWP 31577)] 0x00000000004175b6 in xfdesktop_icon_get_extents (icon=0xffff101010101010, extents=0x7fff58f5d040) at xfdesktop-icon.c:200 200 g_return_val_if_fail(XFDESKTOP_IS_ICON(icon) && extents, FALSE); (gdb) (gdb) (gdb) (gdb) (gdb) bt #0 0x00000000004175b6 in xfdesktop_icon_get_extents (icon=0xffff101010101010, extents=0x7fff58f5d040) at xfdesktop-icon.c:200 #1 0x000000000041c2f8 in xfdesktop_icon_view_motion_notify (widget=<value optimized out>, evt=0x866aa0, user_data=0x589170) at xfdesktop-icon-view.c:905 #2 0x00002b4352795b5d in _gtk_marshal_BOOLEAN__BOXED () from /usr/lib/libgtk-x11-2.0.so.0 #3 0x00002b4353e5048a in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0 #4 0x00002b4353e5fa08 in g_signal_chain_from_overridden () from /usr/lib/libgobject-2.0.so.0 #5 0x00002b4353e60c17 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0 #6 0x00002b4353e61013 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0 #7 0x00002b43528929ae in gtk_widget_get_default_style () from /usr/lib/libgtk-x11-2.0.so.0 #8 0x00002b435278f24d in gtk_propagate_event () from /usr/lib/libgtk-x11-2.0.so.0 #9 0x00002b4352790261 in gtk_main_do_event () from /usr/lib/libgtk-x11-2.0.so.0 #10 0x00002b4352b34b0c in _gdk_events_init () from /usr/lib/libgdk-x11-2.0.so.0 #11 0x00002b435440fc84 in g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0 #12 0x00002b4354412acd in g_main_context_check () from /usr/lib/libglib-2.0.so.0 #13 0x00002b4354412dda in g_main_loop_run () from /usr/lib/libglib-2.0.so.0 #14 0x00002b43527905f3 in gtk_main () from /usr/lib/libgtk-x11-2.0.so.0 #15 0x000000000041163c in main (argc=1, argv=0x7fff58f5da18) at main.c:394
Ok, I can reproduce this. It's weird, cuz sometimes it doesn't crash until many minutes later. But I'm getting the same backtrace you're getting. I think the right move here is just to reload all icons on screen size change rather than trying to determine which have fallen off and then moving them (which appears to be buggy anyway, since some icons are disappearing). Not the most performant solution, but it's the safest.
Ok, I'm pretty sure this should be fixed now. I can't reproduce it anymore, anyway.