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 581548 - segment fault when saving a file
segment fault when saving a file
Status: RESOLVED FIXED
Product: anjuta
Classification: Applications
Component: plugins: symbol-db
2.27.x
Other All
: Normal critical
: ---
Assigned To: Massimo Cora'
Anjuta maintainers
Depends on:
Blocks:
 
 
Reported: 2009-05-06 03:53 UTC by Minjue
Modified: 2009-06-22 21:28 UTC
See Also:
GNOME target: ---
GNOME version: 2.25/2.26



Description Minjue 2009-05-06 03:53:16 UTC
Steps to reproduce:
1. open a gtk+ or gtkmm project
2. make some change in a .c or .cc file, then press save
3. the user interface of anjuta disappear from the desktop


Stack trace:
here is the gdb info:
minjue:/var/backups/anjuta/anjuta-2.27.1.0# gdb anjuta
GNU gdb (GDB) 6.8.50.20090106-cvs-debian
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
(gdb) run
Starting program: /usr/local/bin/anjuta 
[Thread debugging using libthread_db enabled]

(anjuta:10568): libanjuta-WARNING **: Invalid gconf type for key: anjuta.toolbar.visible

(anjuta:10568): libanjuta-WARNING **: Invalid gconf type for key: anjuta.toolbar.visible

(anjuta:10568): libanjuta-WARNING **: Invalid gconf type for key: anjuta.toolbar.visible

(anjuta:10568): libanjuta-WARNING **: Invalid gconf type for key: view.whitespace
[New Thread 0x7fffe7f72950 (LWP 10573)]
[New Thread 0x7fffe6dad950 (LWP 10574)]
[Thread 0x7fffe6dad950 (LWP 10574) exited]
[New Thread 0x7fffe6dad950 (LWP 10580)]
[New Thread 0x7fffdffff950 (LWP 10581)]
[Thread 0x7fffe7f72950 (LWP 10573) exited]
[New Thread 0x7fffe7f72950 (LWP 10584)]
[New Thread 0x7fffdf7fe950 (LWP 10585)]

(anjuta:10568): libanjuta-symbol-db-CRITICAL **: on_scan_files_end_1: assertion `user_data != NULL' failed

(anjuta:10568): libanjuta-symbol-db-CRITICAL **: on_scan_files_end_1: assertion `user_data != NULL' failed
[Thread 0x7fffe7f72950 (LWP 10584) exited]
[Thread 0x7fffdf7fe950 (LWP 10585) exited]
[Thread 0x7fffdffff950 (LWP 10581) exited]
[New Thread 0x7fffdffff950 (LWP 10588)]

(anjuta:10568): libanjuta-WARNING **: Invalid gconf type for key: view.whitespace
[Thread 0x7fffdffff950 (LWP 10588) exited]
[New Thread 0x7fffdffff950 (LWP 10591)]

(anjuta:10568): GLib-CRITICAL **: g_io_channel_write_chars: assertion `channel != NULL' failed

(anjuta:10568): GLib-CRITICAL **: g_io_channel_flush: assertion `channel != NULL' failed

Program received signal SIGSEGV, Segmentation fault.
anjuta_launcher_send_ptyin (launcher=0xbac660, input_str=
    0xffffc7360a72 <Address 0xffffc7360a72 out of bounds>)
    at anjuta-launcher.c:446
446		while (*input_str);
(gdb) 


Other information:
my platform is debian-sid-amd64
Comment 1 Minjue 2009-05-06 03:58:57 UTC
(gdb) bt
  • #0 anjuta_launcher_send_ptyin
    at anjuta-launcher.c line 446
  • #1 sdb_engine_scan_files_2
    at symbol-db-engine-core.c line 1708
  • #2 ??
    from /usr/lib/libgio-2.0.so.0
  • #3 g_main_context_dispatch
    from /usr/lib/libglib-2.0.so.0
  • #4 ??
    from /usr/lib/libglib-2.0.so.0
  • #5 g_main_loop_run
    from /usr/lib/libglib-2.0.so.0
  • #6 gtk_main
    from /usr/lib/libgtk-x11-2.0.so.0
  • #7 main
    at main.c line 232

Comment 2 Massimo Cora' 2009-05-06 16:35:02 UTC
have you got libgda installed?
do you have /dev/shm?
what about ctags? Is it http://ctags.sourceforge.net/ ?
Comment 3 Minjue 2009-05-07 01:12:32 UTC
I installed libgda-4.1.0 from source code, by the default ./configure,
make, make install steps, and I also installed anjuta in the same
default steps from source.
I have /dev/shm/, it's empty.
I installed ctags by "apt-get install exuberant-ctags" in debian, its
version is Version: 1:5.7-4, from http://ctags.sourceforge.net/.
Comment 4 Massimo Cora' 2009-05-07 17:04:43 UTC
I don't get it here.
It seems it's a problem of ctags because it cannot create the g_io_channels.

Can you please try these steps:

$ sqlite3 /path/your_projectproject/.anjuta_sym_db.db

and paste the output of command ".tables" and "select * from symbol".

I'm on a debian sid and I get

$ ctags --version
Exuberant Ctags 5.7, Copyright (C) 1996-2007 Darren Hiebert
  Compiled: Jun  9 2008, 09:07:41
  Addresses: <dhiebert@users.sourceforge.net>, http://ctags.sourceforge.net
  Optional compiled features: +wildcards, +regex

I'm on a 32bit proc but I have never heard problems with 64.
@Johannes: do you see something in your box?

Comment 5 Massimo Cora' 2009-05-07 17:27:20 UTC
Upon further investigation it seems that 

(anjuta:10568): libanjuta-symbol-db-CRITICAL **: on_scan_files_end_1: assertion
`user_data != NULL' failed

this function is called. This one is called when ctags ends. And in this particular case it seems that it ends prematurely. Is the binary placed under /usr/bin/ctags ?

If you give "ctags some_foo.c" file does it create a "tags" file populated with something?

thanks
Comment 6 Minjue 2009-05-08 04:39:52 UTC
thank you all, I have solved this issue by "aptitude purge emacs23",
now I can use anjuta normally. it's caused by emacs23 replaces ctags to use emacs ctags. please help close this bug.
Comment 7 Massimo Cora' 2009-05-08 06:55:40 UTC
ok thanks.
I'll take this bug opened because I'll implement a check to see if the correct ctags has been installed (maybe parsing the output of ctags --version).

Comment 8 Johannes Schmid 2009-05-08 08:21:51 UTC
Hi!

> thank you all, I have solved this issue by "aptitude purge emacs23",
> now I can use anjuta normally. it's caused by emacs23 replaces ctags to use
> emacs ctags. please help close this bug.

Could you check if the debian package for anjuta solves this correctly and otherwise file a bug against debian that they know that these two packages conflict?

Thanks!

Comment 9 Massimo Cora' 2009-06-22 21:28:26 UTC
I'm closing this one because I've added a anjuta-tags parser program internal to Anjuta.
It seems also that next version 2.28 will be shipped with this one instead of the official 'ctags'.