unity-greeter crashed with SIGSEGV in FT_Set_Transform()

StacktraceTop:
 FT_Set_Transform (face=0x7ff134ead9db, matrix=0x7fff8f9ad520, delta=0x0) at /build/buildd/freetype-2.4.4/freetype-2.4.4/src/base/ftobjs.c:469
 _cairo_ft_unscaled_font_set_scale (unscaled=0x19d5e90, scale=<optimized out>) at /build/buildd/cairo-1.10.2/src/cairo-ft-font.c:780
 _cairo_ft_font_face_scaled_font_create (abstract_font_face=0x1a1c740, font_matrix=0x7ff134e6a1a0, ctm=0x7ff134e6a228, options=0x7fff8f9ad860, font_out=0x7fff8f9ad628) at /build/buildd/cairo-1.10.2/src/cairo-ft-font.c:1845
 INT_cairo_scaled_font_create (font_face=0x1a1c740, font_matrix=<optimized out>, ctm=<optimized out>, options=0x7fff8f9ad860) at /build/buildd/cairo-1.10.2/src/cairo-scaled-font.c:1053
 _cairo_gstate_ensure_scaled_font (gstate=0x7ff134e6a138) at /build/buildd/cairo-1.10.2/src/cairo-gstate.c:1803

This particular code path isn't there anymore, but I can't say it won't crash on other similar lines in the code. It *seems* like something broken in cairo though, as at first blush, unity-greeter seems to be doing reasonable cairo API things.

(do you still get this on up to date oneiric?)

On Fri, 2011-08-26 at 17:33 +0000, Michael Terry wrote:
> (do you still get this on up to date oneiric?)

I got it this morning on an up-to-date as of then. Aug 26th. I believe
my stack is one of the dups, might be a slightly different code path.

Yes, newest crash report I have submitted as bug 835290.

apport information

Tried to attach the crash report which I have originally submitted as bug 835290 with "apport-collect 833619", but it uploaded only Dependencies.txt and stopped without any error messages (bug in apport?).

I believe I am seeing this as well. I installed yesterday (08.30.11) and get a blank screen on boot, but I can ssh into the machine. Starting lightdm results in a blank screen and a syslog entry like this:

Aug 31 11:23:43 mother kernel: [ 1230.216768] unity-greeter[6557]: segfault at 0 ip 003abceb sp bfe350f0 error 4 in libgio-2.0.so.0.2916.0[30d000+14a000]

Can I do anything to help?

The issue got no recent duplicate or comment, is it still happening? Doesn't seem an Oneiric stopper bug...

No recent duplicates and no comments since then, so let's assume that this is fixed. If it still happens, can you please file a new apport crash report? Then we'll get an up to date stack trace. Thanks!

Just upgraded to release version of Oneiric, and I get this error. Intel video hardware.

great job assuming that this is fixed. typical of ubuntu to launch with critical bugs. obviously, yes, I too am seeing this issue. maybe you'll get this one right in time for 2012...

I'm not sure if it's the same fault but I 've just upgraded to 11.10 on a Dell Inspiron 6400 laptop - roughly 6 years old, with an Intel chipset graphics card. I get a plain brown screen when I try to log in and find this in dmesg:

compiz[5413]: segfault at 0 ip 012fd746 sp bfbf2020 error 4 in libunityshell.so[1156000+20b000]
[ 3633.533213] compiz[5571]: segfault at 0 ip 01385746 sp bfa810e0 error 4 in libunityshell.so[11de000+20b000]

I can log in using Unity2, but even that appears to be erratic.

I have the same problem -- but I can't launch any display manager -- xdm, gdm, lightdm, etc. all report a segfault in the libgio2.0.so file. This is most definitely
not fixed! I have tried reinstalling glib2.0-0, the package that contains this file, but it doesn't fix the problem. In fact, the old .so file and the new one are identical.
Perhaps then this is a problem with glib and not a specific greeter. One thing I think that could be improved is the error message: I only get the segfault error message in the /var/log/syslog file,
but then there is a wealth of specific information in the ~/.xsession-errors file. Could this be improved, so that x-session-manager tells you to check that file? I didn't even know it existed.

In my case, the problem seems to have something to do with DBus: I get many .xsession-errors entries that report that the greeter program and session manager programs
cannot connect to DBus, like this:

** (gnome-settings-daemon:3109): CRITICAL **: DBus AddMatch() for dconf path '/org/gnome/[some-path]': The connection is closed
g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream returned 0 bytes on an async read (g-io-error-quark, 0). Exiting.

I don't get why this causes a segfault, though. This is all very confusing, and I have no idea which package is causing the problem!

Sorry about that, I posted that in the wrong tab. Terminal browsers are quite confusing...

I have been able to trigger this bug as part of purging out various GNOME data/similar packages. I suspect the greeter is trying to open a file that isn't there now, or similar.