Re: 2.3.10 barely made it out of the gate :-(

[Willem Riede <wriede riede org>]

On 01/31/2006 05:15:10 AM, Peter Bloomfield wrote:
> On 01/30/2006 05:28:08 PM Mon, Willem Riede wrote:
> > On 01/30/2006 07:34:51 AM, Peter Bloomfield wrote:
> >> On 01/30/2006 06:46:46 AM Mon, Pawel Salek wrote:
> >> > On 01/30/2006 12:02:17 AM, Willem Riede wrote:
> >> >> Crashed within minutes on my dual opteron FC5T2 x86_64:
> >> >>>> *** glibc detected *** /usr/bin/balsa: double free or >>  
> >> corruption
> >> >> (fasttop): 0x0000000000cdbc70 ***
> >> >> I am running it now on a FC5t2 machine, Sempron CPU  
> >> without > problems. What's your system configuration? Balsa  
> >> config? Does > it happen on startup?
> >> 
> >> I started getting segfaults in the glibc allocation routines  
> >> called by g_string_sized_new (FC4 Athlon), but they've gone  
> >> away since I upgraded from glib-2.8.4/gtk-2.8.8 to  
> >> 2.9.5/2.8.11.
> > 
> 
> My bad--those were glib2 versions.
> 
> > As to the error I reported - it was not right at startup, but  
> > after looking at a couple of messages. And it is not  
> > reproducable - happened only once so far.
> > 
> > I could upgrade to gtk2-2.8.11 if that's recommended (it is in  
> > rawhide).
> 
> I don't know that it's recommended--just that it fixed some  
> problems for me!

Hmmm, that is certainly not the extend of my problem, since I rebuilt 
balsa 2.3.10 against:

glib2.x86_64                             2.9.5-1                installed
glib2.i386                               2.9.5-1                installed
glib2-devel.x86_64                       2.9.5-1                installed
gtk2.i386                                2.8.11-1               installed
gtk2.x86_64                              2.8.11-1               installed
gtk2-devel.x86_64                        2.8.11-1               installed

I again was hit by:

IMAP mailbox: imap://wriede athena riede org/RedHat

Gtk-CRITICAL **: _gtk_rbtree_reorder: assertion `tree->root->count == length' failed
aborting...

Program received signal SIGTRAP, Trace/breakpoint trap.
[Switching to Thread 1115699552 (LWP 3540)]
0x00002aaaab4de797 in g_logv () from /usr/lib64/libglib-2.0.so.0
(gdb) bt
#0  0x00002aaaab4de797 in g_logv () from /usr/lib64/libglib-2.0.so.0
#1  0x00002aaaab4dea0a in g_log () from /usr/lib64/libglib-2.0.so.0
#2  0x00002aaaaad0b903 in gtk_range_get_adjustment () from /usr/lib64/libgtk-x11-2.0.so.0
#3  0x00002aaaaad8b976 in gtk_tree_view_get_path_at_pos () from /usr/lib64/libgtk-x11-2.0.so.0
#4  0x00002aaaab279e68 in g_closure_invoke () from /usr/lib64/libgobject-2.0.so.0
#5  0x00002aaaab28956a in g_signal_override_class_closure () from /usr/lib64/libgobject-2.0.so.0
#6  0x00002aaaab28a765 in g_signal_emit_valist () from /usr/lib64/libgobject-2.0.so.0
#7  0x00002aaaab28aa1d in g_signal_emit () from /usr/lib64/libgobject-2.0.so.0
#8  0x000000000047d803 in lbm_sort (mbox=0x8d7b80, parent=0x2aaab3926ec0) at mailbox.c:3136
#9  0x000000000047e32b in lbm_set_threading (mailbox=0x8d7b80, thread_type=Variable "thread_type" is not available.
) at mailbox.c:1893
#10 0x000000000047f799 in libbalsa_mailbox_check (mailbox=0x8d7b80) at mailbox.c:599
#11 0x0000000000449c8d in balsa_window_mailbox_check (mailbox=0x8d7b80) at main-window.c:2967
#12 0x00002aaaab4ebd0e in g_slist_foreach () from /usr/lib64/libglib-2.0.so.0
#13 0x0000000000449e06 in check_messages_thread (list=0x1391480) at main-window.c:2985
#14 0x00000039c310615a in start_thread () from /lib64/libpthread.so.0
#15 0x00000039c21c92bd in clone () from /lib64/libc.so.6
#16 0x0000000000000000 in ?? ()
(gdb)

Anything I can do to facilitate fault isolation?

Thanks, Willem Riede.