Jeff Davis
list-mod_python at empires.org
Fri Sep 13 15:06:03 EST 2002
The backtrace is below. I think that we're getting it taken care of. I narrowed it to an uninitialized variable (see another message I sent earlier with the subject: "more info about the segfault"). Thanks for letting me know how to use gdb! Couldn't have done it without gdb, that's for sure. Regards, Jeff #0 0x402611bb in strlen () from /lib/libc.so.6 No symbol table info available. #1 0x4015d9b4 in apr_pstrcat () from /usr/lib/libapr.so.0 No symbol table info available. #2 0x40350f88 in python_handler () from /usr/lib/apache2/modules/mod_python.so No symbol table info available. #3 0x40352817 in PythonPostReadRequestHandler () from /usr/lib/apache2/modules/mod_python.so No symbol table info available. #4 0x08094f90 in ap_run_post_read_request () No symbol table info available. #5 0x08093a1e in ap_read_request () No symbol table info available. #6 0x0806915c in _start () No symbol table info available. #7 0x0808e7c1 in ap_run_process_connection () No symbol table info available. #8 0x0808ebac in ap_process_connection () No symbol table info available. #9 0x08080468 in ap_graceful_stop_signalled () No symbol table info available. #10 0x0808054d in ap_graceful_stop_signalled () No symbol table info available. ---Type <return> to continue, or q <return> to quit--- #11 0x08080684 in ap_graceful_stop_signalled () No symbol table info available. #12 0x08080aa6 in ap_mpm_run () No symbol table info available. #13 0x08087743 in main () No symbol table info available. #14 0x402070bf in __libc_start_main () from /lib/libc.so.6 No symbol table info available. On Friday 13 September 2002 02:47 am, Martin Pool wrote: > On 13 Sep 2002, Jeff Davis <list-mod_python at empires.org> wrote: > > > Try running apache under gdb? Something like this: > > > > > > # gdb /usr/sbin/apache2 > > > [blah blah] > > > gdb> run -X > > > > > > Then try to access a page. Apache should stop in gdb, and it might > > > tell you where the problem is. > > > > I tried both gdb and valgrind and supplied the -X argument for each. When > > I hit a page, all it said was "Segmentation Fault" (for valgrind, and gdb > > was similar: "Program received signal SIGSEGV, Segmentation fault."). GDB > > also said "0x402611bb in strlen () from /lib/libc.so.6" after the > > SIGSEGV, which might be helpful. > > OK, good. Now once you get to that point, type > > gdb> backtrace full > > to get some info on where the call to strlen() that caused trouble > came from.
|