as in subject, I've just installed freetype 2 from their cvs (http://savannah.nongnu.org/cvs/?group=freetype) and I can't start xfce4 4.2 anymore. This is my .xsession-errors: /etc/X11/gdm/PreSession/Default: Registering your session with wtmp and utmp /etc/X11/gdm/PreSession/Default: running: /usr/bin/X11/sessreg -a -w /var/log/wtmp -u /var/run/utmp -x "/var/gdm/:0.Xservers" -h "" -l ":0" "ghepeu" /etc/X11/gdm/Xsession: Beginning session setup... /etc/X11/gdm/Xsession: Setup done, will execute: /usr/bin/ssh-agent -- startxfce4 /usr/bin/startxfce4: X server already running on display :0 xrdb: "Xft.autohint" on line 6 overrides entry on line 2 /home/ghepeu/.config/xfce4//xinitrc: line 52: xscreensaver: command not found Agent pid 22697 /home/ghepeu/.config/xfce4//xinitrc: line 63: 22699 Segmentation fault $xfcesm Agent pid 22697 killed line 62-63 of xinitrc are xfcesm=`which xfce4-session` case "x$xfcesm" in Gnome 2.10 works flawlessy.
First of all: Is there any special reason why you use an unstable version of freetype? CVS versions are usually open to breakage. It doesn't tell anything that Gnome works. Now on with the potential bug: I need more information to fix this bug. Which splash engine do you use? Does the crash also occur if you disable the splash screen? Can you provide a backtrace?
well, there was a bug with the bytecode interpreter in release 2.1.9 so I was using 20041231 CVS until a few days ago, then I checked a few CVS snapshot without doing a system-wide installation. As for the instability, feature freeze and new stable release are scheduled soon: http://lists.gnu.org/archive/html/freetype-devel/2005-03/msg00050.html. The problem seems to be related to splash screen, if I disable it xfce starts, and I can trigger the segfault just by previewing them in settings manager. Can you provide me instructions for a useful backtrace?
Which splash screen, or does it happen with all of them?
mice works, balou and simple don't
Created attachment 188 Fix for the MCS manager crash This patch fixes the MCS manager crash, which was caused by an initialized value.
For the startup crash, I need a backtrace of the crash. To get a usable backtrace: 1) Build xfce4-session with --enable-debug=full 2) Set ulimit -c unlimited in your shell (add it to xinitrc before the xfce4-session line) 3) start session, xfce4-session should leave a core or core.<pid> file in your home directory (depends on the configuration of your linux systems) 4) run: gdb /path/to/xfce4-session core 5) type bt
I followed your instructions but I couldn't find any core file, so I tried running xfce4-session from gdb but what I get doesn't seems to me very useful. Maybe I made a mistake... GNU gdb 6.0 Copyright 2003 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "i686-pc-linux-gnu"...Using host libthread_db librar$ (gdb) run --display :1 Starting program: /usr/bin/xfce4-session --display :1 Program received signal SIGSEGV, Segmentation fault. 0x00000000 in ?? () (gdb) bt #0 0x00000000 in ?? () (gdb) bt full #0 0x00000000 in ?? () No symbol table info available. (gdb) Btw, I've just discovered that with these freetype matlab 7.0 crashes and Xorg with it, so I think I'll make a bug report to their bugzilla or in their mailing list too.
I'm pretty sure there's nothing xfce4-session can do in order to fix/work-around this issue. xfce4-session does not use freetype directly. Instead xfce4-session uses high-level functions from Gdk and Pango, which in turn use Xft, which in turn uses freetype. So they must have made a mistake or changed the semantics without fixing the libraries that use freetype. Anyway, I'm curious to know what the freetype people say about this.
(In reply to comment #5) > Created an attachment (id=188) [edit] > Fix for the MCS manager crash > > This patch fixes the MCS manager crash, which was caused by an initialized > value. Just to check is this engine->start which isn't initialised and therefore is this the same as the following Debian bug? http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=303010 Thanks.
My XFCE 4.2.0 installation crashes after I updated freetype to version 2.1.10 (release). pango was compiled against FT 2.1.9. Also, I do not use xfce4-session.
I'm resolving this one for now. We don't use freetype directly, so it's either a freetype bug or a problem with the pango freetype backend.