Created attachment 168547 [details]
testcase

Attached is a small testcase demonstrating the issue. This is on Fedora 13 x86_64 with glib2 2.24.1 and goffice 0.8.8

[jonn@jon-workstation massifg]$ gcc -g goffice-corrupt-mem.c `pkg-config --libs --cflags libgoffice-0.8 glib-2.0`
[jonn@jon-workstation massifg]$ ./a.out 
** Message: Testing glib functions
** Message: Malloc works
*** glibc detected *** ./a.out: malloc(): memory corruption: 0x0000000001abee50 ***

Expected behaviour would be for this to work in the same way as when using the standard mem_table. 

GDB backtrace:
[jonn@jon-workstation massifg]$ gdb ./a.out 
GNU gdb (GDB) Fedora (7.1-32.fc13)
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-redhat-linux-gnu".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /home/jonn/work/c/massifg/a.out...done.
(gdb) run
Starting program: /home/jonn/work/c/massifg/a.out 
[Thread debugging using libthread_db enabled]
** Message: Testing glib functions
** Message: Malloc works
*** glibc detected *** /home/jonn/work/c/massifg/a.out: malloc(): memory corruption: 0x0000000000655e50 ***
^C
Program received signal SIGINT, Interrupt.
__lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:97
97	2:	movl	%edx, %eax
(gdb) bt

+ Trace 223355