Bug#345156: mysql-admin: Segfaults when User Administration is clicked.

[Adam Majer]

package mysql-admin
forwarded 345156 http://bugs.mysql.com/16119
tag 345156 + confirmed
thanks


Hi,

I tracked the problem down to what look like uninitialized variables.
The patch that "fixed" the segfault is just,

--- MAUserAdministrationPanel.cc        2006-01-01 18:08:53.000000000 -0600
+++ MAUserAdministrationPanel.cc.new    2006-01-01 18:08:28.000000000 -0600
@@ -73,6 +73,8 @@
       _first_show(true), _dirty(false),
       _ignore_entry_changes(false), _ignore_limit_changes(false)
 {
+  _apply_changes_on_user_switch= false;
+  _discard_changes_on_user_switch= false;
   _instance= data->get_instance();
  
  
_instance->signal_disconnect().connect(sigc::mem_fun(*this,&MAUserAdministrationPanel::disconnected_cb));



The segfault occurs in validate() function where _current_user is NULL
but that is not checked. Adding the following small patch fixed that
problem but the real cause appeared to be from uninitialized vvariables
(see above).


@@ -1686,6 +1688,9 @@
   Glib::ustring pw1, pw2;
   Glib::ustring msg;
 
+//  if( _current_user == 0 )
+//    return false;
+
   do
   {
     if (_current_user->hosts_num == 0)


But I will not add any of these patches to Debian package because they
don't work. They just stop mysql-admin from crashing. What happens is
mysql-admin just stops responding as you've see on your MP machine. I'm
guessing there are a number of race conditions as well.

I'll forward this upstream.

- Adam

PS. This is not the same bug that is on Amd64. That one is caused by
casting pointer <-> int which are not the same sizes there.


Matthijs Mohlmann wrote:

>Package: mysql-admin
>Version: 1.1.5-1
>Followup-For: Bug #345156
>
>Hi,
>
>This happens for me too with two different behaviours:
>
>- Dual MP 1800+ it just hangs
>(strace output)
>gettimeofday({1136128055, 770188}, NULL) = 0
>poll([{fd=3, events=POLLIN}, {fd=4, events=POLLIN}], 2, 0) = 0
>gettimeofday({1136128055, 771678}, NULL) = 0
>futex(0xbfc572a8, FUTEX_WAIT, 2, NULL
>
>- Celeron 2.66Ghz
>(gdb output)
>Program received signal SIGSEGV, Segmentation fault.
>[Switching to Thread -1225316128 (LWP 2449)]
>0x0808b75f in Glib::ListHandle<Gtk::CellRenderer*,
>Glib::Container_Helpers::TypeTraits<Gtk::CellRenderer*> >::~ListHandle()
>
>Hope this will help.
>  
>


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]