On 24/01/2010 07:42, Marco Atzeri wrote:
> My question however is about the gdb backtrace, is the
> "Cannot access memory at address 0x1a" at # 9 normal ?
I wouldn't read too much into it. Could indicate that there's some kind of
corruption at the top end of the stack, or it could just be gdb getting
confused about a value that was in a register at the start of execution but
has long since been lost.
> Program received signal SIGSEGV, Segmentation fault.
> 0x20f414d0 in ?? ()
> (gdb) backtrace
> #0 0x20f414d0 in ?? ()
> #1 0x6112e117 in __call_exitprocs () from /usr/bin/cygwin1.dll
> #2 0x6110b6da in exit () from /usr/bin/cygwin1.dll
> #3 0x61004e5a in cygwin_exit () from /usr/bin/cygwin1.dll
> #4 0x610b6838 in _sigfe () from /usr/bin/cygwin1.dll
> #5 0x0022bba8 in ?? ()
> #6 0x6c4e401e in
> cygstdc++-6!_ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc ()
> from /usr/bin/cygstdc++-6.dll
> #7 0x6e5c8c65 in _fu4930___ZSt4cerr () at
> ../../octave_clone/src/toplev.cc:625
> #8 0x6e58fd9c in _fu4441___ZN8file_ops14static_members8instanceE ()
> at ../../octave_clone/src/octave.cc:876
> #9 0x00401167 in main (argc=Cannot access memory at address 0x1a
> ) at ../../octave_clone/src/main.c:35
Looks like maybe an exception was thrown in ostream::operator<<(const
char*). Looks also like somewhere up the stack is
file_ops::static_members::instance(), and it's quite possibly trying to print
something to std::cerr. Possibly a NULL pointer?
Also, you want to set breakpoints on __call_exitprocs and on atexit, and see
what functions are getting registered, whether there's a corrupt entry
pointing at 0x20f414d0 or whether one of the exitprocs is going astray and
jumping into space.
cheers,
DaveK
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
