Bug 1812503 - [abrt] cinnamon-session: match(): cinnamon-session killed by SIGSEGV
Summary: [abrt] cinnamon-session: match(): cinnamon-session killed by SIGSEGV
Keywords:
Status: CLOSED EOL
Alias: None
Product: Fedora
Classification: Fedora
Component: cinnamon-session
Version: 31
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: leigh scott
QA Contact: Fedora Extras Quality Assurance
URL: https://retrace.fedoraproject.org/faf...
Whiteboard: abrt_hash:5a0a08aba6509259924b596bb28...
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2020-03-11 13:01 UTC by James E. LaBarre
Modified: 2020-11-24 17:36 UTC (History)
6 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2020-11-24 17:36:03 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
File: backtrace (104.20 KB, text/plain)
2020-03-11 13:01 UTC, James E. LaBarre
no flags Details
File: core_backtrace (77.86 KB, text/plain)
2020-03-11 13:01 UTC, James E. LaBarre
no flags Details
File: cpuinfo (2.47 KB, text/plain)
2020-03-11 13:01 UTC, James E. LaBarre
no flags Details
File: dso_list (7.64 KB, text/plain)
2020-03-11 13:01 UTC, James E. LaBarre
no flags Details
File: environ (9.85 KB, text/plain)
2020-03-11 13:01 UTC, James E. LaBarre
no flags Details
File: exploitable (82 bytes, text/plain)
2020-03-11 13:01 UTC, James E. LaBarre
no flags Details
File: limits (1.29 KB, text/plain)
2020-03-11 13:02 UTC, James E. LaBarre
no flags Details
File: maps (48.85 KB, text/plain)
2020-03-11 13:02 UTC, James E. LaBarre
no flags Details
File: mountinfo (2.41 KB, text/plain)
2020-03-11 13:02 UTC, James E. LaBarre
no flags Details
File: open_fds (953 bytes, text/plain)
2020-03-11 13:02 UTC, James E. LaBarre
no flags Details
File: proc_pid_status (1.33 KB, text/plain)
2020-03-11 13:02 UTC, James E. LaBarre
no flags Details
.xsession-errors when starting Cinnamon Desktop (2.03 KB, text/plain)
2020-03-12 12:46 UTC, James E. LaBarre
no flags Details
.xsession-errors.old when starting Cinnamon Desktop (30 bytes, text/plain)
2020-03-12 12:47 UTC, James E. LaBarre
no flags Details
tar of /var/log/lightdm/ after Cinnamon Desktop (50.00 KB, application/x-tar)
2020-03-12 12:48 UTC, James E. LaBarre
no flags Details

Description James E. LaBarre 2020-03-11 13:01:34 UTC
Description of problem:
This has happened in both RHEL 7.6 as well as Fedora 31.  Started happening with Cinnamon Desktop (4.4), but also tested with Gnome3 (3.34) and MATE (1.22).  Starting any of these desktop environments from the DM (RHEL 7.6) or LightDM display managers will start loading he desktop environment, then drop back to the display manager/login screen.

Logging in as OpenBox (on RHEL 7.6) LXQt (0.14) or KDE Plasma (5.17) will load those desktops/shells with no problem.

Have installed Fedora 31 on the system that used to have RHEL 7.6, so that particular test environment is not available anymore. 

Also, to verify this isn't an issue with my particular hardware, I copied the home directory from the primary machine (Lenovo ThinkPad P50) to another machine (Lenovo ThinkPad T430).  Previously this method of copying the home directory had worked (testing a mogration from RHEL to Fedora), so it should be a comparable environment.

As this happens in three desktops, I will be pasting this note to those reports as well.

Version-Release number of selected component:
cinnamon-session-4.4.1-1.fc31

Additional info:
reporter:       libreport-2.12.0
backtrace_rating: 4
cgroup:         0::/user.slice/user-1000.slice/session-7.scope
cmdline:        cinnamon-session --session cinnamon
crash_function: match
executable:     /usr/bin/cinnamon-session
journald_cursor: s=9f560bcd5b704e24adefe158709e8d7b;i=7856;b=81f58ca7c29144508286886819caf526;m=709bf484d;t=5a09267f6ea29;x=cc783b71d94b4d5d
kernel:         5.5.8-200.fc31.x86_64
rootdir:        /
runlevel:       N 5
type:           CCpp
uid:            1000
xsession_errors: 

Truncated backtrace:
Thread no. 0 (1 frames)
 #0 match at pcre_exec.c:522

Comment 1 James E. LaBarre 2020-03-11 13:01:46 UTC
Created attachment 1669287 [details]
File: backtrace

Comment 2 James E. LaBarre 2020-03-11 13:01:48 UTC
Created attachment 1669288 [details]
File: core_backtrace

Comment 3 James E. LaBarre 2020-03-11 13:01:49 UTC
Created attachment 1669289 [details]
File: cpuinfo

Comment 4 James E. LaBarre 2020-03-11 13:01:54 UTC
Created attachment 1669290 [details]
File: dso_list

Comment 5 James E. LaBarre 2020-03-11 13:01:56 UTC
Created attachment 1669291 [details]
File: environ

Comment 6 James E. LaBarre 2020-03-11 13:01:57 UTC
Created attachment 1669292 [details]
File: exploitable

Comment 7 James E. LaBarre 2020-03-11 13:02:01 UTC
Created attachment 1669293 [details]
File: limits

Comment 8 James E. LaBarre 2020-03-11 13:02:06 UTC
Created attachment 1669294 [details]
File: maps

Comment 9 James E. LaBarre 2020-03-11 13:02:11 UTC
Created attachment 1669295 [details]
File: mountinfo

Comment 10 James E. LaBarre 2020-03-11 13:02:16 UTC
Created attachment 1669296 [details]
File: open_fds

Comment 11 James E. LaBarre 2020-03-11 13:02:18 UTC
Created attachment 1669297 [details]
File: proc_pid_status

Comment 12 leigh scott 2020-03-11 13:29:58 UTC
There is nothing in the backtrace to suggest it's a cinnamon or lightdm issue

Comment 13 Petr Pisar 2020-03-11 14:26:02 UTC
The logs do not provide any clue on how the PCRE was used. Especially what was the regular expression and options passed to pcre_compile() and what options were passed to pcre_exec(). cinamon-session maintainer, if you believe the is an issue in pcre, please provides these data. Otherwise I have no reproducer and I cannot help you. A Non-truncated stack trace should show you where the pcre_exec() was called from.

The backtrace shows an enormous number of recursive match() calls. It's probably caused by an ill-formed regular expression or a subject string that's prone to a stack exhaustion. Stack exhaustion is not considered as a bug by PCRE. A countermeasure is rephrasing the regular expression or setting lower PCRE recursion limit, or increasing a stack size for the running process.

Comment 14 leigh scott 2020-03-11 15:12:39 UTC
(In reply to Petr Pisar from comment #13)
> The logs do not provide any clue on how the PCRE was used. Especially what
> was the regular expression and options passed to pcre_compile() and what
> options were passed to pcre_exec(). cinamon-session maintainer, if you
> believe the is an issue in pcre, please provides these data. Otherwise I
> have no reproducer and I cannot help you. A Non-truncated stack trace should
> show you where the pcre_exec() was called from.
> 

I can't reproduce the issue so can't provide additional info.
As far as I'm aware cinnamon doesn't use pcre.

Comment 15 Petr Pisar 2020-03-11 15:40:28 UTC
The memory mapping shows /usr/bin/cinnamon-session as the program. E.g. /usr/bin/cinnamon-session loads /usr/lib64/libglib-2.0.so.0 and that library loads /usr/lib64/libpcre.so.1. I did not check if there other transitive dependencies on PCRE.

The stack trace hints that the being matched text contains ";5;7:*.plt=38;5;7:*.pm=38;5;7:" string. That looks like a LS_COLORS environment variable content. However, my LS_COLORS variable does not contain e.g. ".plt=". Perhaps it comes from the reporter's shell profile.

Comment 16 leigh scott 2020-03-11 16:26:04 UTC
(In reply to James E. LaBarre from comment #0)
> Description of problem:
> This has happened in both RHEL 7.6 as well as Fedora 31.  Started happening
> with Cinnamon Desktop (4.4), but also tested with Gnome3 (3.34) and MATE
> (1.22).  Starting any of these desktop environments from the DM (RHEL 7.6)
> or LightDM display managers will start loading he desktop environment, then
> drop back to the display manager/login screen.
> 
> Logging in as OpenBox (on RHEL 7.6) LXQt (0.14) or KDE Plasma (5.17) will
> load those desktops/shells with no problem.
> 
> Have installed Fedora 31 on the system that used to have RHEL 7.6, so that
> particular test environment is not available anymore. 
> 
> Also, to verify this isn't an issue with my particular hardware, I copied
> the home directory from the primary machine (Lenovo ThinkPad P50) to another
> machine (Lenovo ThinkPad T430).  Previously this method of copying the home
> directory had worked (testing a mogration from RHEL to Fedora), so it should
> be a comparable environment.
> 
> As this happens in three desktops, I will be pasting this note to those
> reports as well.
> 
> Version-Release number of selected component:
> cinnamon-session-4.4.1-1.fc31
> 
> Additional info:
> reporter:       libreport-2.12.0
> backtrace_rating: 4
> cgroup:         0::/user.slice/user-1000.slice/session-7.scope
> cmdline:        cinnamon-session --session cinnamon
> crash_function: match
> executable:     /usr/bin/cinnamon-session
> journald_cursor:
> s=9f560bcd5b704e24adefe158709e8d7b;i=7856;b=81f58ca7c29144508286886819caf526;
> m=709bf484d;t=5a09267f6ea29;x=cc783b71d94b4d5d
> kernel:         5.5.8-200.fc31.x86_64
> rootdir:        /
> runlevel:       N 5
> type:           CCpp
> uid:            1000
> xsession_errors: 
> 
> Truncated backtrace:
> Thread no. 0 (1 frames)
>  #0 match at pcre_exec.c:522

Can you log into the working session and run.


gsettings set org.cinnamon.SessionManager debug true


then try to log into cinnamon, then post the files as attachments.

~/.xsession-errors

and

~/.xsession-errors.old


also post the contents of /var/log/lightdm/ in .tar format as an attachment

Comment 17 James E. LaBarre 2020-03-11 17:35:10 UTC
Will run that later today, after a couple of upcoming calls.

Also, as mentioned above, similar behavior was seen in https://bugzilla.redhat.com/show_bug.cgi?id=1812503 and https://bugzilla.redhat.com/show_bug.cgi?id=1794684, where the common factor is that they are all GTK-based environments.  Whether that is relevant I couldn't say.

Comment 18 James E. LaBarre 2020-03-12 12:46:03 UTC
Created attachment 1669644 [details]
.xsession-errors when starting Cinnamon Desktop

Comment 19 James E. LaBarre 2020-03-12 12:47:26 UTC
Created attachment 1669645 [details]
.xsession-errors.old when starting Cinnamon Desktop

Comment 20 James E. LaBarre 2020-03-12 12:48:52 UTC
Created attachment 1669646 [details]
tar of /var/log/lightdm/ after Cinnamon Desktop

Comment 21 James E. LaBarre 2020-03-12 12:52:39 UTC
Also noting; I don't have VMWare installed on this system, especially since I'm working on a different VM env, so I don't know why it would refer to "vmware-user" in the .xsession-error file (fresh install on this system, profile has never run vmware at all).

Comment 22 James E. LaBarre 2020-03-13 19:11:57 UTC
So I went through the steps I had done to migrate my profile to a fresh install, and came across the setting for customizing the colors for the "ls" command (which had been in my old .bashrc file but not the default one created by the install.  Removed this and the login failure went away.  

https://github.com/trapd00r/LS_COLORS

Hadn't changed any of it's default settings yet, so it would be whatever values the sample file has.  Can't understand why a settings file for LS would mess up a login, unless it's because the variable would be large.

Comment 23 Fedora Admin user for bugzilla script actions 2020-06-12 02:46:16 UTC
This package has changed maintainer in the Fedora.
Reassigning to the new maintainer of this component.

Comment 24 Ben Cotton 2020-11-03 16:27:57 UTC
This message is a reminder that Fedora 31 is nearing its end of life.
Fedora will stop maintaining and issuing updates for Fedora 31 on 2020-11-24.
It is Fedora's policy to close all bug reports from releases that are no longer
maintained. At that time this bug will be closed as EOL if it remains open with a
Fedora 'version' of '31'.

Package Maintainer: If you wish for this bug to remain open because you
plan to fix it in a currently maintained version, simply change the 'version' 
to a later Fedora version.

Thank you for reporting this issue and we are sorry that we were not 
able to fix it before Fedora 31 is end of life. If you would still like 
to see this bug fixed and are able to reproduce it against a later version 
of Fedora, you are encouraged  change the 'version' to a later Fedora 
version prior this bug is closed as described in the policy above.

Although we aim to fix as many bugs as possible during every release's 
lifetime, sometimes those efforts are overtaken by events. Often a 
more recent Fedora release includes newer upstream software that fixes 
bugs or makes them obsolete.

Comment 25 Ben Cotton 2020-11-24 17:36:03 UTC
Fedora 31 changed to end-of-life (EOL) status on 2020-11-24. Fedora 31 is
no longer maintained, which means that it will not receive any further
security or bug fix updates. As a result we are closing this bug.

If you can reproduce this bug against a currently maintained version of
Fedora please feel free to reopen this bug against that version. If you
are unable to reopen this bug, please file a new report against the
current release. If you experience problems, please add a comment to this
bug.

Thank you for reporting this bug and we are sorry it could not be fixed.


Note You need to log in before you can comment on or make changes to this bug.