produce "_Unwind_Resume" call seeing under gbd backtracing

Here is the problem which I'm trying to address: We've got a core dump while processing data. The result of backtracing is:

#0 0x00a99402 in __kernel_vsyscall ()
#1 0x00306df0 in raise () from /lib/libc.so.6
#2 0x00308701 in abort () from /lib/libc.so.6
#3 0x001c4530 in _gnu_cxx::_verbose_terminate_handler() () from /usr/lib/libstdc++.so.6
#4 0x001c1f35 in ?? () from /usr/lib/libstdc++.so.6
#5 0x001c12ca in ?? () from /usr/lib/libst开发者_JAVA百科dc++.so.6
#6 0x001c1d99 in __gxx_personality_v0 () from /usr/lib/libstdc++.so.6
#7 0x00d1c7e6 in ?? () from /lib/libgcc_s.so.1
#8 0x00d1cb62 in _Unwind_Resume () from /lib/libgcc_s.so.1
........

I've looked through the code base of our application and it is not clear that the problem is due to uncaught exception or whatever (but i know it somehow connected with exceptions because of _Unwind_Resume call is there). So I'm trying to write simple program which also fails with core dump and its gdb backtracing contains the lines above.

os: CentOS, compiler: gnu gcc 4.1.2, language: c/c++

Any suggestions about the problem/code would be much appreciated