seahorse crashes immediately

[Sean McNeil]

On Fri, 2004-09-24 at 22:14, Joe Marcus Clarke wrote:
> On Thu, 2004-09-23 at 19:25, Sean McNeil wrote:
> > On Thu, 2004-09-23 at 16:10, Joe Marcus Clarke wrote:
> > > On Wed, 2004-09-22 at 02:31, Sean McNeil wrote:
> > > 
> > > > This is what I got:
> > > 
> > > Can you send the output of "bt full".  Specifically, I want to know what
> > > node is.  Also, do you have malloc scrubbing enabled?  If so, does it
> > > still crash if you disable it?
> > 
> > I've tried this with an /etc/malloc.conf of "aj" and without an
> > /etc/malloc.conf.
> > 
> > Here is the bt full:
> 
> Thanks, but where is frame #0?

Was caught up in some other things and I guess I got sloppy on
cut/paste.  Here is the first few frames again:

(gdb) r
Starting program: /usr/X11R6/bin/seahorse

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1 (LWP 100205)]
0x0000000202f55f8f in g_type_check_is_value_type (type=8595649152)
    at gtype.c:3249
3249      if (node && node->mutatable_check_cache)
(gdb) bt full
#0  0x0000000202f55f8f in g_type_check_is_value_type (type=8595649152)
    at gtype.c:3249
No locals.
#1  0x0000000202f43faa in g_signal_newv (signal_name=0x0, itype=5714304,
    signal_flags=G_SIGNAL_RUN_LAST, class_closure=0x56f980,
accumulator=0,
    accu_data=0x0, c_marshaller=0x1, return_type=4, n_params=1,
    param_types=0x569a00) at gsignal.c:1267
        name = (gchar *) 0x569a10 "add"
        signal_id = 0
        i = 0
        node = (SignalNode *) 0x0
        __PRETTY_FUNCTION__ = "g_signal_newv"
#2  0x0000000202f44b98 in g_signal_new_valist (signal_name=0x41adec
"add",
    itype=5714304, signal_flags=G_SIGNAL_RUN_LAST,
class_closure=0x56f980,
    accumulator=0, accu_data=0x0, c_marshaller=0x1, return_type=1,
n_params=1,
    args=0x7fffffffe2e0) at gsignal.c:1370
        param_types = (GType *) 0x569a00
        i = 5378304
        signal_id = 4294959840

I've looked over the code and I think I know what is going on (maybe). 
GType is a gulong.  This is 64 bits on amd64.  There is a define

#define SEAHORSE_TYPE_KEY		(seahorse_key_get_type ())

that is passing what is suppose to be a GType as a vararg to
g_signal_new.  I did a break in seahorse_key_get_type and looked at what
the key is suppose to be and I looked at the param_type[0] which is
passed in:

(gdb) p key_type
$4 = 5714560

(gdb) p param_types[0]
$5 = 8595649152

Cheers,
Sean