New SX editor sometimes crashes gnucash

[David Reiser]

On 22 Jan 2007, at 12:33:54 PM, Josh Sled wrote:

> On Sun, January 21, 2007 8:15 pm, David Reiser wrote:
>> On 21 Jan 2007, at 3:04:07 PM, Josh Sled wrote:
>>> On Sun, 2007-01-21 at 14:19 -0500, David Reiser wrote:
>>>> [...]
>>>     (gdb) backtrace
>>>     [...goodness...]
>>>
>>
>> Here's the goodness:
>>
>> (gdb) bt
>> #0  0x9004796c in kill ()
>> #1  0x9012dc14 in abort ()
>> #2  0x02bc2340 in g_logv ()
>> #3  0x02bc2400 in g_log ()
>> #4  0x02bc0ea8 in g_malloc0 ()
>> #5  0x01047914 in gdc_add_tag_markings (cal=0x732a168, tag=1) at gnc-
>> dense-cal.c:1572
>> #6  0x01047bf0 in gdc_model_update_cb (model=0x67945d0, update_tag=1,
>> user_data=0x732a168) at gnc-dense-cal.c:1626
>
> Bingo.
>
>> I'll work on the breakpoints tomorrow. Gotta go be responsible and
>> write a lecture for tomorrow in a class I haven't taught before...
>
> Heh. :)
>
> Well, nevermind the breakpoints;  if you can just cause it again  
> and get
> to the point where you cause that backtrace, just walk up the stack  
> (with
> "up" at the (gdb) prompt), and you should be able to print the  
> value of
> 'num_marks' ... at the point, we can assume it's not valid, but  
> it'd be
> good to know exactly what value it has.

Is this the right snippet?

(gdb) up
#5  0x01047914 in gdc_add_tag_markings (cal=0x732b170, tag=1) at gnc- 
dense-cal.c:1572
1572         dates = g_new0(GDate*, num_marks);


>
> Thanks...
> ...jsled
>
>

--
David Reiser
dbreiser at earthlink.net