From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6) Gecko/20040312 Description of problem: I upgraded a working fedora core 1 system to fedora core 2 test2. Right after starting X from the command line, gnome-panel crashes and I get a window that says 'The Application "gnome-panel" has quit unexpectedly'. Subsequent attempts to run the panel fail. I get the same results regardless of kernel (fedora 2.6.3 and 2.6.4 or stock 2.6.5). I also upgraded to all the latest developement packages, with no change. This machine is an athlon and the video card is an nvidia, the panel crashes with both "nv" and "nvidia" drivers. Nothing shows up in any logs. I also upgraded a pentium-M laptop with radeon video from fedora core 1 to test2, and the panel works fine. So I'm stumped. Version-Release number of selected component (if applicable): 2.5.92-1 and 2.6.0-1 How reproducible: Always Steps to Reproduce: 1. upgrade fedora core 1 to fedora core 2 test2 on an athlon 2. startx 3. curse at screen Actual Results: Big stop-sign dialog box with the above quoted message. Expected Results: working gnome-panel Additional info:
Is this still happening for you? If it is could you install the appropriate gnome-panel debuginfo package and obtain a stack trace using the dialog get pops up when the panel crashes? Thanks :)
Yes, it is still happening as of gnome-panel-2.6.0-4. I installed the debuginfo package, but I'm not familiar with debuginfo and I have no idea what to do with it. The obvious things have failed. Google has failed. Please advise.
I've got similar problem. I updated a RH9 (dual 1ghz intel processors) directly to FC2T2. On RH9, I had previously upgraded Evoluation to 1.4.6 which had installed some gnome files. During the FC2T2 upgrade, there was a message about some inconsistent gnome file(s) that might cause problems. I performed the same FC2T2 upgrade on two other machines that had fresh FC2 installs and am not seeing the problem. I have latest updates from upd2date as of 4/10. I was wondering if it was some left-over configuration files in home directory from RH9. I defined a brand new user with no previous files of any kind, and it has the same failure
Dan, you need to enable core dumps by adding ulimit -c <max core size in KB> to .bashrc (and relogin). Save the crash dump via bug buddy. If you see undefined symbols in the crash dump that seem relevant you can add debuginfo rpms for the corresponding packages and repeat the crash.
Dan: as Leonard says ... we really need a stack trace to figure out why this is happening for you ...
Created attachment 99419 [details] stack trace from gdb Ok, here is the gdb "bt" output. Hope this is what you need. If there is a way to use the debuginfo package to get better information, I need to be told explicitly how.
thanks, this is a problem with the gnome-vfs menu method
Still a problem with gnome-panel-2.6.0-7 gnome-vfs2-2.6.0-5 sean
It happens for me if I have a .desktop file in /etc/X11/applnk which is not world-readable. Trivially reproducable as follows: sudo chmod og-r /etc/X11/applnk/Graphics/gv.desktop Sean, get it working again you should just be able to find the offending file and make it readable again. Perhaps run gnome-panel under 'strace' and see what the last file it touches before the SEGV is.
*** Bug 121330 has been marked as a duplicate of this bug. ***
Well, close. It dies because a .desktop file doesn't *exist* , or, more accurately, is a broken symlink. Thanks sean open("/usr/share/applications/gimp-1.3.desktop", O_RDONLY|O_LARGEFILE) = -1 ENOENT (No such file or directory) --- SIGSEGV (Segmentation fault) @ 0 (0) --- write(3, "\33\0\2\0\0\0\0\0", 8) = 8 write(3, " \0\2\0\0\0\0\0", 8) = 8 write(3, "+\0\1\0", 4) = 4 read(3, 0xbffff1b4, 32) = -1 EAGAIN (Resource temporarily unavailable) select(4, [3], NULL, NULL, NULL) = 1 (in [3]) read(3, "\1\2I\0\0\0\0\0\23\0\340\1\0\0\0\0\320M\206\10`\335\265"..., 32) = 32 clone(child_stack=0, flags=CLONE_CHILD_CLEARTID|CLONE_CHILD_SETTID|SIGCHLD, child_tidptr=0x4002b048) = 28835 waitpid(28835, (gnome_segv:28835): Gtk-WARNING **: Ignoring the separator setting [WIFEXITED(s) && WEXITSTATUS(s) == 0], 0) = 28835
After doing my own strace, I removed a bad symbolic link at /usr/share/gnome/apps/Games/gnomame.desktop, and now the panel works.
* Thu Apr 22 2004 Mark McLoughlin <markmc> - 2.6.0-6 - Fix crash when .desktop files aren't readable (bug #120014) (already fixed upstream in desktop-file-utils 0.5)
Created attachment 99660 [details] User built rpm I made a rpm, my first time, to upgrade desktop-file-utils because 0.4 is still the current one in development, you might be able to use this to resolve gnome-panel crashes. I have no idea if I did this right but I am tired of not being able to use GNOME.
Sorry for the confusion - the bug is fixed in gnome-vfs 2.6.0-6. If you upgrade to that it should fix the problem. The desktop-file-utils package isn't really affected even though the fix is in desktop-file-utils. gnome-vfs2 has a copy of some of the stuff in desktop-file-utils (Attaching binary RPMs to bugzilla prolly won't make the bugzilla admins happy btw :-)
My gnome-panel still crashes every time with SIGSEGV. I upgraded RH9 to Fedora Core 2 by booting their FC2 install CD and doing the upgrade. I have following (relevant) rpms installed: gnome-panel-2.6.0-9 gnome-vfs-1.0.5-18 gnome-vfs2-2.6.0-8 and now also, desktop-file-utils-0.5-1 but still crashes. With gdb: (gdb) add-symbol-file /usr/lib/debug/usr/lib/gnome-vfs-2.0/modules/libmenu.so.debug ....and so on...couple of symbol-files loaded... Source directories searched: /usr/src/debug/gnome-vfs-2.7.91:/usr/src/debug/gnome-panel-2.7.91.1:$cdir:$cwd (gdb) r Starting program: /usr/bin/gnome-panel [Thread debugging using libthread_db enabled] [New Thread -151128512 (LWP 13205)] Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -151128512 (LWP 13205)] 0x00de6483 in directory_monitor_callback () from /usr/lib/gnome-vfs-2.0/modules/libmenu.so (gdb) generate-core-file Saved corefile core.13205 (gdb) bt #0 0x00de6483 in directory_monitor_callback () from /usr/lib/gnome-vfs-2.0/modules/libmenu.so #1 0x00de68fb in g_canonicalize_file_name () from /usr/lib/gnome-vfs-2.0/modules/libmenu.so #2 0x00dec7b0 in menu_cache_get_menu_for_file () from /usr/lib/gnome-vfs-2.0/modules/libmenu.so #3 0x00def79a in desktop_entry_tree_diradd_callback_set () .... 2nd try: (gdb) r Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -151128512 (LWP 14588)] 0x00707016 in malloc () from /lib/tls/libc.so.6 (gdb) bt #0 0x00707016 in malloc () from /lib/tls/libc.so.6 #1 0x0093d527 in g_malloc () from /usr/lib/libglib-2.0.so.0 #2 0x00de6897 in directory_monitor_callback () from /usr/lib/gnome-vfs-2.0/modules/libmenu.so (gdb) list directory_monitor_callback 2036 void directory_monitor_callback (GnomeVFSMonitorHandle *handle, 2037 const gchar *monitor_uri, 2038 const gchar *info_uri, 2039 GnomeVFSMonitorEventType event_type, 2040 gpointer user_data) 2041 { 2042 MenuMethod * method; 2043 DesktopEntryTree * tree = user_data; 2044 2045 method = method_checkout (); (gdb) 2046 2047 if (tree) 2048 desktop_entry_tree_invalidate( tree, 2049 desktop_entry_tree_get_menu_file_name (tree), 2050 gnome_vfs_get_local_path_from_uri (info_uri)); 2051 2052 method_return (method); 2053 } Even created a new user and logged in through gdm with that user, but gnome-panel does not work.
I got gnome-panel somehow working by renaming /etc/gnome-vfs-2.0/modules/default-modules.conf to /etc/gnome-vfs-2.0/modules/default-modules.conf.rpmorig and copying /etc/gnome-vfs-2.0/modules/default-modules.conf.with-menu-editing to /etc/gnome-vfs-2.0/modules/default-modules.conf After that gnome-panel didn't crash anymore, but now it gives these error messages to stderr: ** (gnome-panel:5954): WARNING **: No toplevel on which to load object '2t1071354047ut938781u500p1249r167139504k3221215788' ** (gnome-panel:5954): WARNING **: No toplevel on which to load object '1t1052118433ut397747u500p1160r1622858785k3221217532' ** (gnome-panel:5954): WARNING **: No toplevel on which to load object '0t1052083621ut909683u500p1160r1347554718k3221215548'
On my FC3 completely updated, when I login, gnome-panel crashes. I've tried to follow what you said before and I've made the strace (I hope it is right and that it could be helpfull). Any idea how to solve the problem? Thanks, Teo (gdb) r Starting program: /usr/bin/gnome-panel [Thread debugging using libthread_db enabled] [New Thread -151140128 (LWP 3683)] ** (gnome-panel:3683): WARNING **: Error in parse: Errore alla linea 4: Il carattere '%' non è valdo come inizio del nome di una entità : il carattere &inizia una entità ; se questo simbolo non vuole essere l'inizio di una entità , utilizzate & ** (gnome-panel:3683): WARNING **: Errore alla linea 4: Il carattere 'on è valdo come inizio del nome di una entità : il carattere &inizia una entità ; se questo simbolo non vuole essere l'inizio di una entità , utilizzate & Program received signal SIGSEGV, Segmentation fault. [Switching to Thread -151140128 (LWP 3683)] 0x00647118 in g_error_free () from /usr/lib/libglib-2.0.so.0 (gdb) bt #0 0x00647118 in g_error_free () from /usr/lib/libglib-2.0.so.0 #1 0x080a5268 in egg_recent_model_read (model=0x23, file=0xfee8c6ac) at egg-recent-model.c:739 #2 0x080a5afb in egg_recent_model_get_list (model=0x9b7ed40) at egg-recent-model.c:1429 #3 0x080a619e in egg_recent_model_changed (model=0x9b7ed40) at egg-recent-model.c:1671 #4 0x0808f994 in panel_recent_append_documents_menu (top_menu=0x23) at panel-recent.c:287 #5 0x0808c62e in panel_menu_bar_append_actions_menu (menubar=0x9b6a9c8) at panel-menu-bar.c:148 #6 0x0808ca34 in panel_menu_bar_instance_init (menubar=0x9b6a9c8, klass=0x9b6a000) at panel-menu-bar.c:230 #7 0x006d75cc in g_type_create_instance () from /usr/lib/libgobject-2.0.so.0 #8 0x006be5a1 in g_object_new () from /usr/lib/libgobject-2.0.so.0 #9 0x006bd83f in g_object_newv () from /usr/lib/libgobject-2.0.so.0 #10 0x006be459 in g_object_new_valist () from /usr/lib/libgobject-2.0.so.0 #11 0x006be578 in g_object_new () from /usr/lib/libgobject-2.0.so.0 #12 0x0808cb84 in panel_menu_bar_load_from_gconf (panel=0x9b1b2a8, locked=1, position=0, exactpos=1, id=0x9b4d8c8 "menu_bar") at panel-menu-bar.c:311 #13 0x08070d4c in panel_applet_load_idle_handler (dummy=0x0) at applet.c:776 #14 0x00655b5a in g_child_watch_add () from /usr/lib/libglib-2.0.so.0 #15 0x006527bb in g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0 ---Type <return> to continue, or q <return> to quit--- #16 0x00654242 in g_main_context_acquire () from /usr/lib/libglib-2.0.so.0 #17 0x006544ef in g_main_loop_run () from /usr/lib/libglib-2.0.so.0 #18 0x07630fbe in gtk_main () from /usr/lib/libgtk-x11-2.0.so.0 #19 0x0806342e in main (argc=2, argv=0xfee8eeb4) at main.c:99 (gdb)
I had the same problem Matteo reported. Both gnome-panel (when clicking on the redhat logo) and file-roller (on start-up) crashed and quit unexpectedly. I found out that the problem was inside the .recently-used file. Here it is my line 4: <URI>file:///home/andrea/wip/Libri/Informatica/Programmazione/Linguaggi/Visual%20Basic/Visual%20Basic%20&%20DB.doc</URI> As you can see, the file name contains & and not & therefore the fatal error. Simply deleting the file solve this issue. However, I don't know the offending program who written that line incorrectly. Andrea.
Same problem here when running desktop from exceed on a pc and rolling over one of the applications menus gnome quits... latest fedora core 3...
Well, this bug never got fixed. I never got gnome-menu working. Now I am going to upgrade from fc2 to fc3, but because this bug I won't make an update, although I would otherwise and would be so much easier. Have to do it hard way or I am pretty sure gnome-panel wouldn't work.
I got it fixed. It's a profile issue, either you have to create a new home directory or backup your home directory & delete the following files/folders. .config .eggcups .gconfd .gnome .gnome2 .gnome2_private .gstreamer-0.8 .gtkrc-1.2-gnome2 .ICEauthority .kde .mcop .mcoprc .metacity .nautilus .qt .recently-used .rhn-applet.conf restart & login back, you should be good.