Bug 732018 – Crash due to passing set GError to glib-networking

After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.

Bug 732018 - Crash due to passing set GError to glib-networking


Summary:	Crash due to passing set GError to glib-networking


Status:	RESOLVED FIXED

Product:	evolution-data-server
Classification:	Platform
Component:	Mailer
Version:	3.12.x (obsolete)
Hardware:	Other Linux

Importance:	Normal critical
Target Milestone:	---
Assigned To:	evolution-mail-maintainers
QA Contact:	Evolution QA team

URL:
Whiteboard:

Duplicates:	733519 748448 753659 (view as bug list)
Depends on:
Blocks:

Reported:	2014-06-21 22:01 UTC by Eternal Sorrow
Modified:	2015-08-19 15:05 UTC

See Also:
GNOME target:	---
GNOME version:	---

Description Eternal Sorrow 2014-06-21 22:01:39 UTC

Evolution crashes randomly.

Error message and gdb backtrace:


GLib-Net:ERROR:/tmp/portage/net-libs/glib-networking-2.40.1-r1/work/glib-networking-2.40.1/tls/gnutls/gtlsconnection-gnutls.c:688:end_gnutls_io: assertion failed: (!error || !*error)

Program received signal SIGABRT, Aborted.

+ Trace 233716

Thread 140735730640640 (LWP 2259)

#0 raise
from /lib64/libc.so.6
#1 abort
from /lib64/libc.so.6
#2 g_assertion_message
#3 g_assertion_message_expr
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/glib/gtestutils.c line 2306
#4 end_gnutls_io
#5 g_tls_connection_gnutls_write
at /tmp/portage/net-libs/glib-networking-2.40.1-r1/work/glib-networking-2.40.1/tls/gnutls/gtlsconnection-gnutls.c line 1528
#6 g_tls_output_stream_gnutls_write
at /tmp/portage/net-libs/glib-networking-2.40.1-r1/work/glib-networking-2.40.1/tls/gnutls/gtlsoutputstream-gnutls.c line 60
#7 g_output_stream_write
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gio/goutputstream.c line 219
#8 stream_write
at camel-stream.c line 162
#9 camel_stream_write
#10 stream_write
at camel-pop3-stream.c line 184
#11 camel_stream_write
at camel-stream.c line 539
#12 engine_command_queue
at camel-pop3-engine.c line 286
#13 camel_pop3_engine_command_new
at camel-pop3-engine.c line 450
#14 pop3_store_authenticate_sync
at camel-pop3-store.c line 687
#15 camel_service_authenticate_sync
at camel-service.c line 2117
#16 mail_authenticator_try_password_sync
at e-mail-authenticator.c line 164
#17 e_source_authenticator_try_password_sync
at e-source-authenticator.c line 449
#18 source_registry_authenticate_authenticate_cb
at e-source-registry.c line 1927
#19 ffi_call_unix64
from /usr/lib64/libffi.so.6
#20 ffi_call
from /usr/lib64/libffi.so.6
#21 g_cclosure_marshal_generic
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gclosure.c line 1445
#22 g_closure_invoke
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gclosure.c line 768
#23 signal_emit_unlocked_R
#24 g_signal_emitv
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gsignal.c line 3045
#25 e_dbus_authenticator_proxy_g_signal
at e-dbus-authenticator.c line 1248
#26 ffi_call_unix64
from /usr/lib64/libffi.so.6
#27 ffi_call
from /usr/lib64/libffi.so.6
#28 g_cclosure_marshal_generic
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gclosure.c line 144
#29 g_type_class_meta_marshal
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gclosure.c line 961
#30 g_closure_invoke
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gclosure.c line 768
#31 signal_emit_unlocked_R
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gsignal.c line 3589
#32 g_signal_emit_valist
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gsignal.c line 3307
#33 g_signal_emit
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gobject/gsignal.c line 3363
#34 on_signal_received
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gio/gdbusproxy.c line 917
#35 emit_signal_instance_in_idle_cb
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gio/gdbusconnection.c line 3750
#36 g_idle_dispatch
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/glib/gmain.c line 5319
#37 g_main_dispatch
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/glib/gmain.c line 3064
#38 g_main_context_dispatch
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/glib/gmain.c line 3663
#39 g_main_context_iterate
#40 g_main_loop_run
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/glib/gmain.c line 3928
#41 e_source_registry_authenticate_sync
at e-source-registry.c line 2190
#42 mail_session_authenticate_sync
at e-mail-session.c line 1439
#43 camel_session_authenticate_sync
at camel-session.c line 1537
#44 pop3_store_connect_sync
at camel-pop3-store.c line 558
#45 service_shared_connect_thread
at camel-service.c line 558
#46 g_task_thread_pool_thread
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/gio/gtask.c line 1213
#47 g_thread_pool_thread_proxy
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/glib/gthreadpool.c line 307
#48 g_thread_proxy
at /tmp/portage/dev-libs/glib-2.40.0/work/glib-2.40.0/glib/gthread.c line 764
#49 start_thread
from /lib64/libpthread.so.0
#50 clone
from /lib64/libc.so.6

Comment 1 André Klapper 2014-06-21 22:37:57 UTC

Which exact version is this about?

Comment 2 Eternal Sorrow 2014-06-22 00:38:32 UTC

3.12.3

Comment 3 Milan Crha 2014-11-20 13:59:31 UTC

Thanks for a bug report. This is similar to bug #737903, at least in a sense that it is crashing due to gnutls asserting an 'error' parameter. Could I ask what authentication type your POP3 account uses, please?

I'm not sure whether end_gnutls_io() was about to use it, but it was ready to propagate an error "Error writing data to TLS socket: %s", from which I suppose there was some issue connecting to the server.

The next time, please check for any private information in the backtrace before submitting it, it can contain various private information, like passwords, server addresses, email addresses and any other such kind of information which one usually doesn't want to share in public. I usually search for "pass" at least (quotes for clarity only).

Comment 4 Milan Crha 2014-11-21 10:22:41 UTC

*** Bug 733519 has been marked as a duplicate of this bug. ***

Comment 5 Milan Crha 2014-11-21 10:28:12 UTC

The above duplicate has the same crash, but from IMAP. So we have currently IMAP and POP3, while there was also one for SMTP (bug #737903). I'm only unsure whether it's related to a certain authentication method or a general error reported, but not handled properly.

Comment 6 Milan Crha 2014-11-21 11:40:50 UTC

I fixed the POP3 crash, but I'm not sure with the IMAP+ one, because it was missing line numbers and such (debugging info for evolution-data-server). Nonetheless, there were plenty of places which could be related to the crash, thus I fixed at least those.

Created commit 8eef965 in eds master (3.13.8+) [1]
Created commit f4742e1 in eds evolution-data-server-3-12 (3.12.9+)

[1] https://git.gnome.org/browse/evolution-data-server/commit/?id=8eef965

Comment 7 Eternal Sorrow 2014-11-23 02:32:48 UTC

(In reply to comment #3)
Sorry for the long reply, I was far away from my Linux laptop.
> what authentication type your POP3 account uses, please?
STARTTLS
> The next time, please check for any private information in the backtrace before
> submitting it, it can contain various private information, like passwords,
> server addresses, email addresses and any other such kind of information which
> one usually doesn't want to share in public.
Oh, my bad :) But I already changed all these passwords, so it's OK.

Comment 8 Milan Crha 2015-02-23 14:21:44 UTC

It turned out that the crash can still happen in some cases. I added a workaround and claim an error on console, instead of let glib-networking crash the application with the below change.

Created commit a17c38c in eds master (3.15.91+)

Comment 9 Milan Crha 2015-04-28 10:51:41 UTC

*** Bug 748448 has been marked as a duplicate of this bug. ***

Comment 10 Milan Crha 2015-08-19 15:05:10 UTC

*** Bug 753659 has been marked as a duplicate of this bug. ***