Bug 57343 - Segmenzation fault caused by the "Add User.LoginTime property" commit
Segmenzation fault caused by the "Add User.LoginTime property" commit
Status: RESOLVED FIXED
Product: accountsservice
Classification: Unclassified
Component: general
unspecified
Other All
: medium critical
Assigned To: Matthias Clasen
:
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2012-11-20 21:21 UTC by György Balló
Modified: 2012-11-28 02:20 UTC (History)
4 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description György Balló 2012-11-20 21:21:51 UTC
Since version 0.6.26, I get a segmentation fault when I try to run accounts-daemon, and this causes that GDM fails to run. After bisecting, I found that it caused by the following commit:

accountsservice: Add User.LoginTime property
http://cgit.freedesktop.org/accountsservice/commit/?id=eb1c98bd9785cea849028b8be60d8d16c9abcab0


The output of the accounts-daemon command:

(accounts-daemon:366): GLib-GObject-WARNING **: g_object_set_valist: object class `User' has no property named `ballogy'

(accounts-daemon:366): GLib-GObject-WARNING **: g_object_set_valist: object class `User' has no property named `login-previous_login'
Segmentation fault


The backtrace:

#0  0xb7be37b6 in __strchr_sse2_bsf () from /usr/lib/libc.so.6
#1  0xb7e0ecaf in g_param_spec_pool_lookup () from /usr/lib/libgobject-2.0.so.0
#2  0xb7e09ac9 in g_object_set_valist () from /usr/lib/libgobject-2.0.so.0
#3  0xb7e0a438 in g_object_set () from /usr/lib/libgobject-2.0.so.0
#4  0x0805599a in entry_generator_wtmp (users=
    Python Exception <class 'gdb.error'> There is no member named keys.: 
0x8077d90, state=0xbffff92c) at daemon.c:344
#5  0x08056683 in load_entries (daemon=daemon@entry=0x8074048, 
    users=users@entry=
    Python Exception <class 'gdb.error'> There is no member named keys.: 
0x8077d90, entry_generator=entry_generator@entry=
    0x8055686 <entry_generator_wtmp>) at daemon.c:480
#6  0x0805686b in reload_users (daemon=0x8074048) at daemon.c:551
#7  reload_users_timeout (daemon=0x8074048) at daemon.c:593
#8  0xb7d3d9e0 in ?? () from /usr/lib/libglib-2.0.so.0
#9  0xb7d40793 in g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0
#10 0xb7d40b30 in ?? () from /usr/lib/libglib-2.0.so.0
#11 0xb7d40f8b in g_main_loop_run () from /usr/lib/libglib-2.0.so.0
#12 0x0805baec in main (argc=1, argv=0xbffffba4) at main.c:182
Comment 1 György Balló 2012-11-22 18:14:13 UTC
With version 0.6.29 I still receive a segmentation fault. The warning about login-previous_login was gone, but all others are unchanged.
Comment 2 Matthias Clasen 2012-11-26 13:54:16 UTC
What are the others ?
Comment 3 György Balló 2012-11-26 21:05:00 UTC
I still get the same segfault with version 0.6.29.

Backtrace:
#0  0xb7be37b6 in __strchr_sse2_bsf () from /usr/lib/libc.so.6
#1  0xb7e0ecaf in g_param_spec_pool_lookup () from /usr/lib/libgobject-2.0.so.0
#2  0xb7e09ac9 in g_object_set_valist () from /usr/lib/libgobject-2.0.so.0
#3  0xb7e0a438 in g_object_set () from /usr/lib/libgobject-2.0.so.0
#4  0x0805599a in entry_generator_wtmp (users=
    Python Exception <class 'gdb.error'> There is no member named keys.: 
0x8077d90, state=0xbffffabc) at daemon.c:344
#5  0x08056683 in load_entries (daemon=daemon@entry=0x8074048, 
    users=users@entry=
    Python Exception <class 'gdb.error'> There is no member named keys.: 
0x8077d90, entry_generator=entry_generator@entry=
    0x8055686 <entry_generator_wtmp>) at daemon.c:480
#6  0x0805686b in reload_users (daemon=0x8074048) at daemon.c:551
#7  reload_users_timeout (daemon=0x8074048) at daemon.c:593
#8  0xb7d3d9e0 in ?? () from /usr/lib/libglib-2.0.so.0
#9  0xb7d40793 in g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0
#10 0xb7d40b30 in ?? () from /usr/lib/libglib-2.0.so.0
#11 0xb7d40f8b in g_main_loop_run () from /usr/lib/libglib-2.0.so.0
#12 0x0805baec in main (argc=1, argv=0xbffffd34) at main.c:182
Comment 4 Matthew Monaco 2012-11-27 23:07:36 UTC
György, are you on a 32-bit system? I have a bunch of 64-bits and haven't seen this. Then I started noticing it on my single 32-bit system.
Comment 5 Matthias Clasen 2012-11-27 23:34:02 UTC
That is a good hint.  Please try the fix that I've just pushed.
Comment 6 György Balló 2012-11-28 02:20:55 UTC
Yes, it happened on my 32 bit systems only, and the fix solves the problem. Thanks!