gnome-shell (login screen) crashes with SIGSEGV in __GI_getenv("EXPAT_ACCOUNTING_DEBUG")

Bug #2039340 reported by Anadon
10
This bug affects 1 person
Affects Status Importance Assigned to Milestone
gnome-shell (Ubuntu)
New
High
Unassigned

Bug Description

Last time I hacked around a bug like this, it had portions in PAM, apparmor, Seahorse, and Gnome so this time I'd like someone more focused on broad stability to look at things.

The problem is simple. Before the upgrade to 23.10, I could log in using the "Ubuntu", "Ubuntu on X.org". "Gnome", and "Gnome on X.org" WMs. Now I can't. I can still login using XFCE and TTY.

Revision history for this message
Anadon (anadon) wrote :
Revision history for this message
Anadon (anadon) wrote :
Revision history for this message
Steve Langasek (vorlon) wrote :

What is the behavior when you try to log in? Do you get an authentication failure message, or does it appear there is a successful login followed by a screen flicker and a return to the login screen?

Are there failure messages in /var/log/auth.log?

Changed in pam (Ubuntu):
status: New → Incomplete
Revision history for this message
Anadon (anadon) wrote :

There is some funky stuff going on in auth.log.

I go to log in, enter my password, it starts the process, then lands me back at the gdm login screen after a second.

Revision history for this message
Steve Langasek (vorlon) wrote :

2023-10-14T01:44:00.468542-04:00 anadon gdm-password]: pam_unix(gdm-password:session): session opened for user anadon(uid=1000) by (uid=0)
2023-10-14T01:44:00.473683-04:00 anadon systemd-logind[1169]: New session 6 of user anadon.
2023-10-14T01:44:00.522860-04:00 anadon (systemd): pam_unix(systemd-user:session): session opened for user anadon(uid=1000) by (uid=0)
2023-10-14T01:44:00.658692-04:00 anadon gdm-password]: gkr-pam: unlocked login keyring
2023-10-14T01:44:01.673108-04:00 anadon gnome-keyring-daemon[3237]: The PKCS#11 component was already initialized
2023-10-14T01:44:01.673392-04:00 anadon gnome-keyring-daemon[3560]: discover_other_daemon: 1
2023-10-14T01:44:01.674538-04:00 anadon gnome-keyring-daemon[3237]: The Secret Service was already initialized
2023-10-14T01:44:01.674742-04:00 anadon gnome-keyring-daemon[3562]: discover_other_daemon: 1
2023-10-14T01:44:01.677463-04:00 anadon gnome-keyring-daemon[3563]: discover_other_daemon: 1
2023-10-14T01:44:09.242277-04:00 anadon PackageKit: uid 1000 is trying to obtain org.freedesktop.packagekit.system-sources-refresh auth (only_trusted:0)
2023-10-14T01:44:09.243637-04:00 anadon PackageKit: uid 1000 obtained auth for org.freedesktop.packagekit.system-sources-refresh
2023-10-14T01:44:10.632860-04:00 anadon gdm-password]: pam_unix(gdm-password:session): session closed for user anadon
2023-10-14T01:44:10.643754-04:00 anadon systemd-logind[1169]: Session 6 logged out. Waiting for processes to exit.
2023-10-14T01:44:10.644366-04:00 anadon systemd-logind[1169]: Removed session 6.

This appears to show a successful login with gdm. So if launching the graphical session fails afterwards, that must be an issue with gdm or some other downstream component. Reassigning.

affects: pam (Ubuntu) → gdm3 (Ubuntu)
Changed in gdm3 (Ubuntu):
status: Incomplete → New
Revision history for this message
Sebastien Bacher (seb128) wrote :

The journal shows a gnome-shell error

Bail out! Gjs:ERROR:../gi/wrapperutils.h:847:static Prototype* GIWrapperPrototype<Base, Prototype, Instance, Info>::create_prototype(Info*, GType) [with Base = InterfaceBase; Prototype = InterfacePrototype; Instance = InterfaceInstance; Info = _GIBaseInfoStub; GType = long unsigned int]: assertion failed: (gtype != G_TYPE_INVALID)

Do you have local extensions installed? Could you try if a new user has the same issue?

affects: gdm3 (Ubuntu) → gnome-shell (Ubuntu)
Changed in gnome-shell (Ubuntu):
importance: Undecided → High
status: New → Incomplete
Revision history for this message
Anadon (anadon) wrote :

No local extensions are installed. Only the default system extensions <email address hidden> and <email address hidden> are installed following https://askubuntu.com/questions/1133782/command-to-list-installed-and-enabled-gnome-extensions. XFCE4 is now also broken so I'm without a GUI for the moment.

I did try with a blank new user and they have no such issues.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

There's also a lot of this in the journal:

  GNOME Shell crashed with signal 11

so to help us find the cause of the crash please follow these steps:

1. Look in /var/crash for crash files and if found run:
    ubuntu-bug YOURFILE.crash
Then tell us the ID of the newly-created bug.

2. If step 1 failed then look at https://errors.ubuntu.com/user/ID where ID is the content of file /var/lib/whoopsie/whoopsie-id on the machine. Do you find any links to recent problems on that page? If so then please send the links to us.

Please take care to avoid attaching .crash files to bugs as we are unable to process them as file attachments. It would also be a security risk for yourself.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Also just in case you've found a new form of bug 2034619 please try adding this to /etc/environment:

  MUTTER_DEBUG_KMS_THREAD_TYPE=user

and then reboot.

no longer affects: ubuntu
Revision history for this message
Anadon (anadon) wrote :

I am unable to feasibly comply with running anything from apport. Apport uses `#!/usr/bin/python3` instead of `#!/usr/bin/env python3` across its files, which runs into issues with my Guix installation's instance of Glibc even when removing all the environmental variables that allow Guix to hook into the terminal. Apport must be doing something clever but not quite clever enough to isolate itself from potentially broken system components? In it's place, I'll attach the crash file.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Thanks for the crash file. That's certainly a segfault so let's make this bug about that.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

#0 __GI_getenv (name=0x7fc095eefdd8 "EXPAT_ACCOUNTING_DEBUG") at ./stdlib/getenv.c:31
        ep = 0x559dca3a7720
        len = 22
#1 0x00007fc095eee73f in ?? () from /lib/x86_64-linux-gnu/libexpat.so.1
No symbol table info available.
#2 0x00007fc095ed4f88 in ?? () from /lib/x86_64-linux-gnu/libexpat.so.1
No symbol table info available.
#3 0x00007fc095ede760 in ?? () from /lib/x86_64-linux-gnu/libexpat.so.1
No symbol table info available.
#4 0x00007fc096ebf221 in ?? () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#5 0x00007fc096ebfcce in ?? () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#6 0x00007fc096ebfea9 in ?? () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#7 0x00007fc096ec051d in ?? () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#8 0x00007fc095edc7fe in ?? () from /lib/x86_64-linux-gnu/libexpat.so.1
No symbol table info available.
#9 0x00007fc095ed9de4 in ?? () from /lib/x86_64-linux-gnu/libexpat.so.1
No symbol table info available.
#10 0x00007fc095edb9bd in ?? () from /lib/x86_64-linux-gnu/libexpat.so.1
No symbol table info available.
#11 0x00007fc095edfa9a in XML_ParseBuffer () from /lib/x86_64-linux-gnu/libexpat.so.1
No symbol table info available.
#12 0x00007fc096ebf37d in ?? () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#13 0x00007fc096ebfcce in ?? () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#14 0x00007fc096ea7df1 in ?? () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#15 0x00007fc096ea2dfd in ?? () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#16 0x00007fc096ea80dd in FcInit () from /lib/x86_64-linux-gnu/libfontconfig.so.1
No symbol table info available.
#17 0x00007fc096eeb561 in ?? () from /lib/x86_64-linux-gnu/libpangoft2-1.0.so.0
No symbol table info available.
#18 0x00007fc09a170331 in ?? () from /lib/x86_64-linux-gnu/libglib-2.0.so.0
No symbol table info available.
#19 0x00007fc099897ada in start_thread (arg=<optimized out>) at ./nptl/pthread_create.c:444
        ret = <optimized out>
        pd = <optimized out>
        out = <optimized out>
        unwind_buf = {cancel_jmp_buf = {{jmp_buf = {140465186371568, -680333647052097841, -120, 11, 140724282146384, 140463721930752, 652596815861298895, 652963903777715919}, mask_was_saved = 0}}, priv = {pad = {0x0, 0x0, 0x0, 0x0}, data = {prev = 0x0, cleanup = 0x0, canceltype = 0}}}
        not_first_call = <optimized out>
#20 0x00007fc09992847c in clone3 () at ../sysdeps/unix/sysv/linux/x86_64/clone3.S:78
No locals.

summary: - On upgrade from 23.04 to 23.10, cannot login using Gnome or Ubuntu WM
- entries
+ gnome-shell (login screen) crashes with SIGSEGV in
+ __GI_getenv("EXPAT_ACCOUNTING_DEBUG")
no longer affects: apparmor (Ubuntu)
Changed in gnome-shell (Ubuntu):
status: Incomplete → Triaged
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

This probably has the same root cause as bug 2034524

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Thank you for taking the time to report this bug and helping to make Ubuntu better. Please execute the following command only once, as it will automatically gather debugging information, in a terminal:

  apport-collect 2039340

When reporting bugs in the future please use apport by using 'ubuntu-bug' and the name of the package affected. You can learn more about this functionality at https://wiki.ubuntu.com/ReportingBugs.

Changed in gnome-shell (Ubuntu):
status: Triaged → Incomplete
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Looks like it might be another form of https://gitlab.gnome.org/GNOME/mutter/-/issues/2463

Revision history for this message
Anadon (anadon) wrote : Re: [Bug 2039340] Re: gnome-shell (login screen) crashes with SIGSEGV in __GI_getenv("EXPAT_ACCOUNTING_DEBUG")

Doing so is no longer representative of my system. I had to transfer
packages co-installed via Guix then reinstall all installed packages due to
required work which required my system in a working state.

Apport needs to change it's hard coded shell path to Python in order to
have made this not an issue from earlier.

On Mon, Oct 23, 2023, 12:11 AM Daniel van Vugt <email address hidden>
wrote:

> Thank you for taking the time to report this bug and helping to make
> Ubuntu better. Please execute the following command only once, as it
> will automatically gather debugging information, in a terminal:
>
> apport-collect 2039340
>
> When reporting bugs in the future please use apport by using 'ubuntu-
> bug' and the name of the package affected. You can learn more about this
> functionality at https://wiki.ubuntu.com/ReportingBugs.
>
> ** Changed in: gnome-shell (Ubuntu)
> Status: Triaged => Incomplete
>
> ** Bug watch added: gitlab.gnome.org/GNOME/mutter/-/issues #2463
> https://gitlab.gnome.org/GNOME/mutter/-/issues/2463
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/2039340
>
> Title:
> gnome-shell (login screen) crashes with SIGSEGV in
> __GI_getenv("EXPAT_ACCOUNTING_DEBUG")
>
> Status in gnome-shell package in Ubuntu:
> Incomplete
>
> Bug description:
> Last time I hacked around a bug like this, it had portions in PAM,
> apparmor, Seahorse, and Gnome so this time I'd like someone more
> focused on broad stability to look at things.
>
> The problem is simple. Before the upgrade to 23.10, I could log in
> using the "Ubuntu", "Ubuntu on X.org". "Gnome", and "Gnome on X.org"
> WMs. Now I can't. I can still login using XFCE and TTY.
>
> To manage notifications about this bug go to:
>
> https://bugs.launchpad.net/ubuntu/+source/gnome-shell/+bug/2039340/+subscriptions
>
>

Changed in gnome-shell (Ubuntu):
status: Incomplete → New
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

This is probably fixed by the fix for bug 2036651.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.