Evolution crashes almost randomly

[ubuntolo]

Binary package hint: evolution

Filling this bug report in order to provide additional informaions, as requested in #579928.
As already reported:

I think this should be considered a major: never experienced such a disaster with Evolution, and I'm using it from two years ago. Evolution just keeps crashing randomly: it can happen while editing a message or simply by itself, not necessarily while sending/receiving messages, but often while (apparently) doing absolutely nothing. The program just closes silently, without any kind of warning, eventually it recovers uncompleted messages at next restart, but it continues with it's strange behavior.
I hope this can be fixed quickly, because e-mailing is a quite fundamental activity.

ProblemType: Crash
DistroRelease: Ubuntu 10.04
Package: evolution 2.28.3-0ubuntu9
ProcVersionSignature: Ubuntu 2.6.32-22.33-generic 2.6.32.11+drm33.2
Uname: Linux 2.6.32-22-generic i686
NonfreeKernelModules: nvidia
Architecture: i386
CrashCounter: 1
Date: Sat May 22 13:17:48 2010
ExecutablePath: /usr/bin/evolution
InstallationMedia: Ubuntu 10.04 LTS "Lucid Lynx" - Release i386 (20100429)
ProcCmdline: evolution
ProcEnviron:
 LANG=it_IT.utf8
 SHELL=/bin/bash
SegvAnalysis:
 Segfault happened at: 0x34bbfee <g_slice_alloc+1070>: mov (%eax),%eax
 PC (0x034bbfee) ok
 source "(%eax)" (0x00000000) not located in a known VMA region (needed readable region)!
 destination "%eax" ok
SegvReason: reading NULL VMA
Signal: 11
SourcePackage: evolution
StacktraceTop:
 g_slice_alloc () from /lib/libglib-2.0.so.0
 g_slice_alloc0 () from /lib/libglib-2.0.so.0
 ?? () from /usr/lib/libgnome-keyring.so.0
 gnome_keyring_is_available ()
 e_passwords_init () from /usr/lib/libedataserverui-1.2.so.8
Title: evolution crashed with SIGSEGV in g_slice_alloc()
UserGroups: adm admin cdrom dialout lpadmin plugdev sambashare

[Pedro Villavicencio (pedro)]

Thank you taking the time to report this bug and helping to make Ubuntu better. However, processing the crash report to get detailed information for the developers failed as the retracer did not generate a useful symbolic stack trace.
Please try to obtain a backtrace manually following the instructions at http://wiki.ubuntu.com/DebuggingProgramCrash and upload the backtrace (as an attachment) to the bug report. This will greatly help us in tracking down your problem.

[Charles Profitt (cprofitt)]

Pedro:

Is that documentation (for 8.04) still valid?

[ubuntolo (ovitaerc)]

That's fine, I'll try to do it, but considering that the process is quite long and the but shows itself randomly it may take me a few days before having the backtrace. Will posto it as soon as I can.

[Rick Vause (rvause)]

I am also experience the same, seemingly random problem. Attaching backtrace.

[gabak (gab77)]

just load evolution with 2 or 3gb of emails and it will crash. I have a customer who has many years of emails and when i import all his emails from thunderbird from windows this problem started to happend.
If you have few emails , evolution works just fine. Even after an update to evolution 2.30

[ubuntolo (ovitaerc)]

Hi gabak, forgive me, but I'm not sure that it's a size related bug. With my Evolution archive, approximately 400 MB of e-mails, the bug still happens, so I think that the size is not important at all.
Forgive me, but still I have no time to produce the backtrace, I hope that you can use the one produced by Rick in the meanwhile.

[ubuntolo (ovitaerc)]

The bug is still marked as Incomplete: isn't the backtrace posted by Rick valid for debugging? Please Pedro, let us know.

[Pedro Villavicencio (pedro)]

Looks like a memory corruption crash, you need to get a valgrind log, Please read following the instructions at https://wiki.ubuntu.com/Valgrind and attach the file to the bug report. This will greatly help us in tracking down your problem. And to all the reporters, if you think you're having the same issue please open a new bug report with apport and let us decide if it's the same issue or not. Thanks.

[Rick Vause (rvause)]

Took all day for a crash but here's a valgrind log.

[ar-jar (arto-jarvinen-letter)]

The 2.30.1.2 exhibits the same behavior, i.e. crashes silently when left alone (and sometimes when you open a mail message too). Thanks guys for working on this! -A

[Nils (nils-fishtown)]

I have the same problem (2.28.3), just crashes randomly, the error now in terminal was:
***MEMORY-ERROR***: evolution[29764]: GSlice: assertion failed: sinfo->n_allocated > 0

[Jean-Peer Lorenz (peer.loz)]

@maintainer: Why is it still marked as 'Incomplete'? Please give us a hint what information is requiered to work on this issue.

This is a major and critical bug (especially in a LTS version).

[ubuntolo (ovitaerc)]

Maybe it's due to the fact that I didn't post my logs? But Rick provided them, so I think it should be considered as a completed bug report now, am I worng? Anyway I will not have the time to setup all the things needed to produce those logs before the next weekend, so I hope that someone can begin working on Rick's files...
I agree in considering this bug as a major one, we're talking about random and frequent crashes in the main e-mail application of the distribution... Seems that other GNOME ditros haven't reported this bug: can it suggest that the bug is in another Evolution-connected component (the new global messaging menu for example?)?

[Pedro Villavicencio (pedro)]

We need a new log with dbgsym, please install the libcamel1.2-14-dbgsym, evolution-dbgsym, evolution-data-server-dbgsym, libglib2.0-0-dbgsym and libgtk2.0-0-dbgsym, We need a backtrace and a valgrind log, Nils, Jean and Ubuntolo if you think you're getting the same crash please provide the logs and let us to determine if that's the same or not, commenting on random bugs because the title is probably the thing you're facing isn't good and is not going to help us to solve the issue, thanks.

[Matthias Braun (matze-braunis)]

I had the same issue. After some updates in the last weeks.
At least for me it seemed like evolution always crashed when it wanted to ask me for a password.
So after checking my gnome-keyring configuration (some stuff in these Session/Start-Applications) settings were disabled for me. And I had an extra entry in /etc/xdg/autostart/gnome-keyring.desktop.dpkg-bak.
Anyway my crashs are gone now so unfortunately I can't create a backtrace or valgrind dump. I'm also not sure which of my steps exactly made it working again.

[LANSA (juan-manuelr)]

I had similar issue since may 2010. Evolution closes with no warning. When I run Evolution with the GDB I got the following error message:
***MEMORY-ERROR***: evolution[7628]: GSlice: assertion failed: sinfo->n_allocated > 0

Program received signal SIGABRT, Aborted.
[Switching to Thread 0xad8f7b70 (LWP 8263)]
0xb7fe2422 in __kernel_vsyscall ()
(gdb)

I made a fresh Ububtu 10.04 installation and i updated the Evolution 2.28.3 to 2.30.1.2 but still have same issue.
I'm using the Evolution with five email accounts, four from google apps and gmail (POP3) and one from my local ISP (POP3) but I discover that a Google Calendar appointment message kills immediately Evolution.
I'm guessing that the bug could be addressed to google accounts.

[ubuntolo (ovitaerc)]

Well, I've worked on this in the weekend. I can attach a backtrace at the moment. After running Evolution for about 60 hours under the valgrind control, however, the bug has not still shown. So I'm afraid I will have to wait more. Evolution running inside valgrind is very slow, maybe the bug will be slow too! :)
Well, here is the backtrace, can you take a look at it, Pedro?

[Sebastien Bacher (seb128)]

valgrind sometime workaround crashes but still list errors

[ubuntolo (ovitaerc)]

Well, and how can I know if the log (my log file was about 90 MB when I left my pc this morning if I can remember and if I wasn't sleeping) already has catched the error condition or not? I'm afraid I can't be sure until I'll se Evolution quitting... Ideas?

[Sebastien Bacher (seb128)]

usually "Invalid read" or "Invalid write" in non python code are errors, you might want to get a log without using --leak-check=full since it's nor required for crashers and would reduce a lot the log

[ubuntolo (ovitaerc)]

For the maintainers: after about 80 hours Evolution refuses to crash, valgrind log has grown 185 MB, and I'm starting tobe afraid that waiting more would be useless. Perhaps valgrind is correcting the problem while controlling the program, so I will not get a crash anymore. However, What to do now? I can try to compress the log file and see if the result will be lighter enough to be posted here, but I doubt it. I can close Evolution, trow the log file and start everything again with the option suggested by Sebastien, hoping that I will get a crash this time. I can even just wait for another week and win the first prize for the larger log in debugging history! :)
Waiting for instructions, I'll do what you ask.

[ubuntolo (ovitaerc)]

Please maintainers, tell me something, I'm staying with Evolution as if my 2 GHz pc was a Z80 since days ago. I would love to contribute to remove the Incomplete status, but Evolutions doesn't crashes while valgrind is checking it, so what should I do?

[bobzr (bobzr)]

This bug is really a disaster, I liked Evolution so much but I'm now considering switching to Thunderbird. Lightning for TB 3 is still beta but yet seems to work better than Evo, which is supposed to be stable.
If it's left idle for some time it simply closes.

[ubuntolo (ovitaerc)]

Still I'm waiting here for some indication from maintainers, still valgrind is running and writing such a huge log, but it seems that it will be of no help. So, is there anyone?

[ubuntolo (ovitaerc)]

Well, after more than a week I had to shut sown Evolution, the valgrind log is 437 MB long, but it just take 21,5 MB compressed. If maintainers think so I can upload it here, or even send it by e-mail, as you prefer. I'm sure that during this week the crash condition must have appeared more than once, but probably valgrind fixed it "on the fly" so Evolution didn't crashed at all. Can I do something more?

[Nils (nils-fishtown)]

I just installed Valgrind and Valkyrie but a have no clue what to put in the Valkyrie options???
Can someone tell me what to put in there?

[Nils (nils-fishtown)]

Tried some things in terminal and got this, dont know if it is correct and also don't know where and if there is a log???

nils@nils-desktop:~$ evolution #valgrind ps -ax
** (evolution:2850): DEBUG: mailto URL command: evolution %s
** (evolution:2850): DEBUG: mailto URL program: evolution

***MEMORY-ERROR***: evolution[2850]: GSlice: assertion failed: sinfo->n_allocated > 0
Afgebroken
nils@nils-desktop:~$

[Sebastien Bacher (seb128)]

you likely want to consider not using "-leak-check=full" and get a small log and stop when you start getting invalid read or write in the log

[Nils (nils-fishtown)]

Then I got this, again don't know if it is wright and also: where can I find the log?
nils@nils-desktop:~$ evolution #valgrind ps -leak-check=full
** (evolution:2435): DEBUG: mailto URL command: evolution %s
** (evolution:2435): DEBUG: mailto URL program: evolution

***MEMORY-ERROR***: evolution[2435]: GSlice: assertion failed: sinfo->n_allocated > 0
Afgebroken

[ubuntolo (ovitaerc)]

Hi, for several reasons I can't do any more testing for the next weeks, so to complete this bug report I'm posting my full and huge valgrind log. It's full of invalid read, so I think it should be helpful. The uncompressed file size is more than 400 MB, check the available space before expanding it.

[Nils (nils-fishtown)]

Could someone tell me where to find the log? Or what to do exactly?

[Eric (evdijken)]

For some unknown reason Evolution is running normal the last few days. Has there been some kind of update wich could have effect on the running of Evolution?

[ubuntolo (ovitaerc)]

@Nils: I was able to collect the logs by following the instructions on comments #2, 9 and 15.

@Eric: It seems you're right, I've noticed an evolution-data-server package update in the last few days, but I can't use my pc very much at the moment so I can't confirm that the problem is fixed.

[Francesco Ruvolo (ruvolof)]

This bug affects me, too.
I don't know if it can be usefull, but this is what I get with gdb.

[ubuntolo (ovitaerc)]

Well, message #18 contains the gdb log, and #31 has the valgrind one, so I don't understand why this bug report is still marked as incomplete.

[Nils (nils-fishtown)]

It looks like the latest update off Ubuntu solved this one for me, still on Evolution 2.28.3 but no crashes anymore?

[Edward Parker (bontano)]

I got so sick of this problem that I installed another version of Evolution that was circulating. It had a variety of issues, but at least it didn't crash randomly. When Ubuntu 10.10 came out, I decided to do a completely fresh install, so I am back to using the standard distribution. Now, Evolution is crashing randomly again. I'm really getting frustrated with Evo, for it seems like the list of bugs I experience is growing, not shrinking. If this keeps up I may have to start using Gmail, though I am not fond of relying on a web interface for mail.

[Launchpad Janitor (janitor)]

[Expired for evolution (Ubuntu) because there has been no activity for 60 days.]

[Olha Honcharova (sleepythomasina)]

I use evolution regularily and have the problem of quite often random crashes, too.
See attachment for one full backtrace.

[Olha Honcharova (sleepythomasina)]

I use evolution regularily and have the problem of quite often random crashes, too.
See attachment for one full backtrace.

[Olha Honcharova (sleepythomasina)]

I have Ubuntu 10.04 and the latest official evolution package.

[Olha Honcharova (sleepythomasina)]

Another backtrace.

[Olha Honcharova (sleepythomasina)]

my ~/.xsession-errors contains:

(evolution:3647): evolution-shell-WARNING **: The name org.freedesktop.NetworkManager was not provided by any .service files

[Olha Honcharova (sleepythomasina)]

I hope someone takes attention to this nasty bug. I like evolution very much, it is nicely designed.

Please find valgrind memcheck in attachment.

[Olha Honcharova (sleepythomasina)]

I hope someone takes attention to this nasty bug. I like evolution very much, it is nicely designed.

Please find valgrind memcheck in attachment.

[shnifti (b-rous)]

same for me. Mostlikely randomly on receiving and when I hit the "send" button. Very nasty bug. Have it since over one year. Maybe some buggy old config because I always copied .evolution to new ubuntu versions. Had the bug at least in 10.04 and 10.10.
Sucks... Otherwise really like evolution.

[baldbrad (baldbrad)]

Beautiful Thai Women

Thai women for dating or marriage
Free To Join ... Lifetime Upgrade
Dating

[Edward Parker (bontano)]

Over a year since my last update, and my Evo is still crashing randomly. It appears that it only happens when I am viewing/moving about in the calendar (crashes at least 2x per day). If there is some information that I can provide (log files or something) that might help troubleshoot, I'd be happy to provide it. Evo just crashes with no apparent error messages.

[agnul (sucrabu)]

Me too!

Evolution worked almost flawlessly for a couple of years on my previous laptop, running 10.10. New laptop, running 64bit oneiric, evolution manually installed and data imported from a backup. Evolution crashes randomly: at least a couple of times every morning, while starting up, while refreshing folders, while doing absolutely nothing. And after a while, with me doing absolutely nothing it starts working. Oh and have I mentioned the sudden crashes if I unplug the network? Or if I put evolution in "online" mode after switching it offline before unplugging the network?

(Sorry for ranting)

[agnul (sucrabu)]

I've been running evolution with full logging over the last couple of weeks and it kept crashing. Most of the time the last lines in the log files (see attachment) seem to be about talking with the IMAP server at the other end (which is zimbra). Hope that helps.

[agnul (sucrabu)]

I've collected a couple of gdb backtraces, hope it helps.

Evolution crashed a couple of times while I left it unattended, opening a mail message every now an then (the evo-unattended... files), and another time just after startup, when I clicked the "send/receive" button. (evo-send-receive.txt).

[Soporte CACyA (34-soporte)]

I use evolution regularily and I need it but I have the same problem of quite often random crashes. I need that this problem resolved. Thanks in advanced.

[Bernie Innocenti (codewiz)]

Not sure if this is the same bug. I get random crashes from evolution_3.6.0-0ubuntu3:

Program received signal SIGABRT, Aborted.
[Switching to Thread 0x7fff41982700 (LWP 30165)]
0x00007ffff3c5f425 in __GI_raise (sig=<optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
64 ../nptl/sysdeps/unix/sysv/linux/raise.c: No such file or directory.
(gdb) bt
#0 0x00007ffff3c5f425 in __GI_raise (sig=<optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
#1 0x00007ffff3c62b8b in __GI_abort () at abort.c:91
#2 0x00007ffff421f290 in mem_error (format=format@entry=0x7ffff429a2d5 "assertion failed: %s") at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:1442
#3 0x00007ffff4266d5b in slab_allocator_free_chunk (chunk_size=chunk_size@entry=16, mem=mem@entry=0x7fff3c00c8a0) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:1324
#4 0x00007ffff426709e in magazine_cache_trim (stamp=<optimized out>, ix=<optimized out>, allocator=<optimized out>) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:672
#5 magazine_cache_push_magazine (ix=<optimized out>, magazine_chunks=<optimized out>, count=<optimized out>) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:703
#6 0x00007ffff421f759 in thread_memory_magazine2_unload (ix=<optimized out>, tmem=<optimized out>) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:802
#7 0x00007ffff4267f2e in g_slice_free1 (mem_size=<optimized out>, mem_block=0x7fff2637a9a0) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:1093
#8 0x00007ffff423bcb0 in g_hash_table_remove_node (hash_table=hash_table@entry=0x555555c8e180, i=i@entry=216688, notify=notify@entry=1) at /build/buildd/glib2.0-2.34.1/./glib/ghash.c:481
#9 0x00007ffff423c3fa in g_hash_table_remove_internal (hash_table=0x555555c8e180, key=0x7fff2637a9a0, notify=1) at /build/buildd/glib2.0-2.34.1/./glib/ghash.c:1274
#10 0x00007ffff6d6708d in camel_pstring_free () from /usr/lib/libcamel-1.2.so.40
#11 0x00007ffff6d0aafd in ?? () from /usr/lib/libcamel-1.2.so.40
#12 0x00007ffff4268cfd in g_slist_foreach (list=<optimized out>, func=0x7ffff6d0abc0 <camel_message_info_free>, user_data=0x0) at /build/buildd/glib2.0-2.34.1/./glib/gslist.c:894
#13 0x00007ffff6d0c6a3 in ?? () from /usr/lib/libcamel-1.2.so.40
#14 0x00007ffff6d50edc in ?? () from /usr/lib/libcamel-1.2.so.40
#15 0x00007ffff47b9e3e in run_in_thread (job=<optimized out>, c=0x55555a53b520, _data=0x555559f43b90) at /build/buildd/glib2.0-2.34.1/./gio/gsimpleasyncresult.c:869
#16 0x00007ffff47a8236 in io_job_thread (data=0x55555a6e37a0, user_data=<optimized out>) at /build/buildd/glib2.0-2.34.1/./gio/gioscheduler.c:162
#17 0x00007ffff4270e62 in g_thread_pool_thread_proxy (data=<optimized out>) at /build/buildd/glib2.0-2.34.1/./glib/gthreadpool.c:309
#18 0x00007ffff4270645 in g_thread_proxy (data=0x55555a6a2940) at /build/buildd/glib2.0-2.34.1/./glib/gthread.c:797
#19 0x00007ffff3fefe9a in start_thread (arg=0x7fff41982700) at pthread_create.c:308
#20 0x00007ffff3d1ccbd in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:112
#21 0x0000000000000000 in ?? ()

[Bernie Innocenti (codewiz)]

Another crash. Now I'm getting them very often, within 5-10 minutes from startup.
This one looks like heap corruption:

***MEMORY-ERROR***: evolution[30408]: GSlice: assertion failed: sinfo->n_allocated > 0

Program received signal SIGABRT, Aborted.
[Switching to Thread 0x7fffbd579700 (LWP 30605)]
0x00007ffff3c5f425 in __GI_raise (sig=<optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
64 ../nptl/sysdeps/unix/sysv/linux/raise.c: No such file or directory.

(gdb) bt
#0 0x00007ffff3c5f425 in __GI_raise (sig=<optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64
#1 0x00007ffff3c62b8b in __GI_abort () at abort.c:91
#2 0x00007ffff421f290 in mem_error (format=format@entry=0x7ffff429a2d5 "assertion failed: %s") at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:1442
#3 0x00007ffff4266d5b in slab_allocator_free_chunk (chunk_size=chunk_size@entry=16, mem=mem@entry=0x55555738f4e0) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:1324
#4 0x00007ffff426709e in magazine_cache_trim (stamp=<optimized out>, ix=<optimized out>, allocator=<optimized out>) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:672
#5 magazine_cache_push_magazine (ix=ix@entry=0, magazine_chunks=<optimized out>, count=<optimized out>) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:703
#6 0x00007ffff42671d2 in private_thread_memory_cleanup (data=0x7fffb80008c0) at /build/buildd/glib2.0-2.34.1/./glib/gslice.c:770
#7 0x00007ffff3fefc83 in __nptl_deallocate_tsd () at pthread_create.c:156
#8 0x00007ffff3fefea8 in start_thread (arg=0x7fffbd579700) at pthread_create.c:315
#9 0x00007ffff3d1ccbd in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:112
#10 0x0000000000000000 in ?? ()

[Bernie Innocenti (codewiz)]

I've upgraded to evolution_3.6.2-0ubuntu2 from raring and I still get the random crashes.

My laptop has a second root partition with Fedora 18 and a shared /home. The evolution binary in Fedora evolution-3.6.2-3.fc18.x86_64 can access the same user profile without crashing.

[Andres Gomez (Tanty) (tanty)]

Taking comment #55, is this the same than bug 1169694 ?

[Charles Willams (chuck-itadmins)]

and I get the following constantly which leads to me starting evo more during the day than just about anything else:

***MEMORY-ERROR***: evolution[9061]: GSlice: assertion failed: sinfo->n_allocated > 0
Aborted (core dumped)

[Emiel Kollof (emiel-kollof)]

Same crashes here, with MAPI plugin connected to an Exchange 2010 server. Evolution version evolution 3.6.4-0ubuntu1

Stacktrace:
 #0 0x00007ff2b7179037 in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56
         resultvar = 0
         pid = 8037
         selftid = 8037
 #1 0x00007ff2b717c698 in __GI_abort () at abort.c:90
         save_stage = 2
         act = {__sigaction_handler = {sa_handler = 0x33, sa_sigaction = 0x33}, sa_mask = {__val = {51, 140680513553696, 140680499170453, 140677358813189, 0, 1, 140680430390504, 140680434305096, 140680513553696, 140680437170165, 1
40680499197733, 10, 18446744073709551615, 140680434309600, 10, 4294967295}}, sa_flags = 10, sa_restorer = 0x7ff2bb479a00}
         sigs = {__val = {32, 0 <repeats 15 times>}}

[Jörg Frings-Fürst (jff-de)]

10.04 is not longer supported
change status to invalid (see gnome-bugs)