Bi-Weekly Scheduled transactions -- bad news...

Robert Heller heller at deepsoft.com
Sun Feb 15 12:30:37 CST 2004 

  Derek Atkins <warlord at MIT.EDU>,
  In a message on Sun, 15 Feb 2004 13:01:19 -0500, wrote :

DA> Well, a segv in free() tends to imply double-free'ing data, or some
DA> other memory corruption problem.  How are your C debugging skills? ;)

Pretty good.  But I know next to nothing about the GNome or Gtk
libraries.  I also did not build GnuCash from sources -- I downloaded a
batch of binary RPMs (for RH 7.3).  I am not totally sure I have all of
the devel RPMs needed to build a debug-able version of GnuCash.

DA> 
DA> -derek
DA> 
DA> Robert Heller <heller at deepsoft.com> writes:
DA> 
DA> >   Derek Atkins <warlord at MIT.EDU>,
DA> >   In a message on Sun, 15 Feb 2004 12:38:00 -0500, wrote :
DA> >
DA> > DA> Did you get a gdb backtrace from the segfault?
DA> >
DA> > Yes.  I clicked on 'File Bug Report' and bug-buddy created one.  I had
DA> > to futz with the bug-buddy Bug Report and I believe I cc'ed the report
DA> > to you.
DA> >
DA> > Here is an 'interesting' piece of it:
DA> >
DA> > #0  0x401734d9 in wait4 () from /lib/libc.so.6
DA> > #1  0x401eb30c in __DTOR_END__ () from /lib/libc.so.6
DA> > #2  0x4044eff8 in gnome_segv_handle () from /usr/lib/libgnomeui.so.32
DA> > #3  <signal handler called>
DA> > #4  0x4013d5d9 in chunk_free () from /lib/libc.so.6
DA> > #5  0x4013d548 in free () from /lib/libc.so.6
DA> > #6  0x402c49f2 in g_free () from /usr/lib/libglib-1.2.so.0
DA> > #7  0x402bbda9 in g_date_free () from /usr/lib/libglib-1.2.so.0
DA> > #8  0x403001ab in _gnc_sxd_free_dates () from /usr/lib/libgncgnome.so.0
DA> > #9  0x402c2caa in g_list_foreach () from /usr/lib/libglib-1.2.so.0
DA> > #10 0x403003ec in putSchedXactionInDialog () from /usr/lib/libgncgnome.so.0
DA> > #11 0x402c2caa in g_list_foreach () from /usr/lib/libglib-1.2.so.0
DA> > #12 0x402fead5 in schedXact_populate () from /usr/lib/libgncgnome.so.0
DA> > #13 0x402fe9d9 in gnc_ui_scheduled_xaction_dialog_create ()
DA> >    from /usr/lib/libgncgnome.so.0
DA> > #14 0x402f4a26 in sxftd_advanced_clicked () from /usr/lib/libgncgnome.so.0
DA> > #15 0x405acd91 in gtk_marshal_NONE__NONE () from /usr/lib/libgtk-1.2.so.0
DA> > ...
DA> >
DA> > It seems that something in the scheduled transaction editor is in
DA> > trouble.  The problem only appeared when I put in Bi-Weekly scheduled
DA> > transactions and seems to have gone away when I deleted the Bi-Weekly
DA> > scheduled transactions.
DA> >
DA> > DA> 
DA> > DA> -derek
DA> > DA> 
DA> > DA> Robert Heller <heller at deepsoft.com> writes:
DA> > DA> 
DA> > DA> > OK, with some poking around I am pretty sure that there is something
DA> > DA> > seriously fishy with Bi-Weekly Scheduled transactions.  I added a few of
DA> > DA> > these -- a set of payments synchronized with my Bi-Weekly paycheck.  I
DA> > DA> > was able to add other Scheduled transactions, but I could not use the
DA> > DA> > Scheduled transactions editor while *adding* a new transaction --
DA> > DA> > GnuCash (/usr/bin/guile) would crash with a SegFault.
DA> > DA> >
DA> > DA> > Eventually, I was able to get the Scheduled transactions editor and by
DA> > DA> > simply deleting the Bi-Weekly transactions (sight unseen), I was able to
DA> > DA> > get things back to functioning sanely (no more crashing).
DA> > DA> >
DA> > DA> > Someone needs to have a seriously close look at the Scheduled
DA> > DA> > transactions code, particularly looking at the Bi-Weekly code.
DA> > DA> >
DA> > DA> > It seems somehow related to the code in gdate.c -- somehow it is
DA> > DA> > mis-creating date strings or something like that.
DA> > DA> >
DA> > DA> >
DA> > DA> >                                      \/
DA> > DA> > Robert Heller                        ||InterNet:   heller at cs.umass.edu
DA> > DA> > http://vis-www.cs.umass.edu/~heller  ||            heller at deepsoft.com
DA> > DA> > http://www.deepsoft.com              /\FidoNet:    1:321/153
DA> > DA> >
DA> > DA> >
DA> > DA> >
DA> > DA> >
DA> > DA> >
DA> > DA> >
DA> > DA> >                                                                                                                     
DA> > DA> > _______________________________________________
DA> > DA> > gnucash-user mailing list
DA> > DA> > gnucash-user at lists.gnucash.org
DA> > DA> > https://lists.gnucash.org/mailman/listinfo/gnucash-user
DA> > DA> >
DA> > DA> >
DA> > DA> 
DA> > DA> -- 
DA> > DA>        Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
DA> > DA>        Member, MIT Student Information Processing Board  (SIPB)
DA> > DA>        URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
DA> > DA>        warlord at MIT.EDU                        PGP key available
DA> > DA> 
DA> > DA>                          
DA> >
DA> >                                      \/
DA> > Robert Heller                        ||InterNet:   heller at cs.umass.edu
DA> > http://vis-www.cs.umass.edu/~heller  ||            heller at deepsoft.com
DA> > http://www.deepsoft.com              /\FidoNet:    1:321/153
DA> >
DA> >
DA> >
DA> >
DA> >
DA> >
DA> >                                                                                                    
DA> > _______________________________________________
DA> > gnucash-user mailing list
DA> > gnucash-user at lists.gnucash.org
DA> > https://lists.gnucash.org/mailman/listinfo/gnucash-user
DA> >
DA> >
DA> 
DA> -- 
DA>        Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
DA>        Member, MIT Student Information Processing Board  (SIPB)
DA>        URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
DA>        warlord at MIT.EDU                        PGP key available
DA> 
DA>                                                                                  

                                     \/
Robert Heller                        ||InterNet:   heller at cs.umass.edu
http://vis-www.cs.umass.edu/~heller  ||            heller at deepsoft.com
http://www.deepsoft.com              /\FidoNet:    1:321/153

More information about the gnucash-user mailing list