From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Phil Bouchard Sent: Sunday, July 06, 2008 3:46 PM
I get the following backtrace, but I haven't investiguated thta much yet. It looks to be an STL problem but I'll check this out tomorrow:
#0 0x0804fc7a in std::_List_base<std::pair<char*, char*>, boost::fast_pool_allocator<std::pair<char*, char*>, boost::default_user_allocator_new_delete, boost::details::pool::pthread_mutex, 32u> >::_M_clear ( this=0x8dc2008) at /usr/include/c++/4.3/bits/list.tcc:77 #1 0x0804fcdb in ~_List_base (this=0x8dc2008) at /usr/include/c++/4.3/bits/stl_list.h:358 #2 0x0804fcef in ~list (this=0x8dc2008) at /usr/include/c++/4.3/bits/stl_list.h:417 #3 0x0804fd4f in ~auto_ptr (this=0x80685b8) at /usr/include/c++/4.3/backward/auto_ptr.h:173 #4 0x0804fe80 in ~pool (this=0x80685a0) at ../../../boost/detail/sh_owned_base_nt.hpp:71 #5 0x006bc527 in exit () from /lib/tls/libc.so.6 #6 0x006a6ded in __libc_start_main () from /lib/tls/libc.so.6 #7 0x08048df1 in _start ()
Just for grins, try printing a message out from the destructor of owned_base::pool (or setting a breakpoint there) and see how many times you wind up there. My hunch is that this is an ODR issue... -Chris