1.9.8: Segfault when entering transaction for 2006/07/04

Derek Atkins warlord at MIT.EDU
Wed Jul 5 17:53:20 EDT 2006

Quoting Thomas Klausner <tk at giga.or.at>:

> On Wed, Jul 05, 2006 at 05:40:36PM -0400, Derek Atkins wrote:
>> Huh.  I've got GCC 4.1.1 on my FC5 box and it works just fine.  Did you
>> rebuild all of ports from scratch after you upgraded your system?
> Yes, I did.

Huh.  I'm really at a loss here.

> What is happening at this point?
> Is it external, e.g. guile?

Nope.  What's going on is that the keypress event is passed from X to gdk
to gtk to our code and then it's trying to redraw the table.  But when it
tries to redraw the table it's SEGV'ing instead.  This is the line where
it's crashing:

        result = gnc_table_direct_update (table, virt_loc,
                                          &new_start, &new_end,

I've no idea why it's crashing here.  This code hasn't changed in a very long
time.  Indeed, the annotation shows that this code hasn't changed since 2002:
  6626  (peticolas 2002-01-21):                 result = 
(table, virt_loc,
  6626  (peticolas 2002-01-21):
  6626  (peticolas 2002-01-21):
  6626  (peticolas 2002-01-21):
  &new_start, &new_end,
  6626  (peticolas 2002-01-21):

So...  I suspect it's a build environment problem, or a compiler bug, or
a bug in your gtk, gdk, or x libraries.   What versions of those do you have?
I'll note that Nobody else can reproduce this.  I cannot.  David Hampton
cannot.  Nobody else on IRC can reproduce it, either, so.. Right now I
think this is you.


> Thomas

       Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
       Member, MIT Student Information Processing Board  (SIPB)
       URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
       warlord at MIT.EDU                        PGP key available

More information about the gnucash-devel mailing list