Hi Benoit, On Tue, Dec 07, 2010 at 06:04:18PM +0100, "Benoît Monin" wrote: > Hi, > > I'm experiencing some sigsegv when running some tests with a 3G modem. > I have 2 active connections : > - an ethernet link, configured with dhcp. Typical setup is > ip = 10.23.218.30 > mask = 255.255.240.0 > gw = 10.23.208.1 > - a 3G modem, configured with ofono. Typical setup is > ip = 10.248.160.97 > mask = 255.255.255.255 > gw = 0.0.0.0 > > The sim card used is configured with a dedicated password authenticated > APN connected to our internal network through vpn. So both links ends > up on the same network, with different mask. This is a bogus network > setup, and the values provided by the gsm operator looks strange but > connman should not crash. I agree.
> I'm simulating gsm coverage loss with an antenna attenuator and connman > crashes when the gsm signal is back or when calling ofono > activate_context again. > > The decoded backtrace : > -------- backtrace -------- > [0]: update_order() [connection.c:466] > [1]: __connman_profile_changed() [profile.c:182] > [2]: __connman_service_indicate_state() [service.c:3046] > [3]: set_configuration() [network.c:574] > [4]: set_connected_fixed() [network.c:591] > [5]: connman_network_set_connected() [network.c:812] > [6]: signal_filter() [watch.c:403] > [7]: message_filter() [watch.c:530] > [8]: message_dispatch() [mainloop.c:80] > [9]: main() [main.c:262] > [10]: _start() [start.S:122] > --------------------------- > > I can post the full syslog or run some other tests if needed. The full syslog would be a good start. > I'll try to reproduce the bug differently, it seems to be related > to gateway management. Can you reliably reporduce it ? If so, running connmand through gdb would be useful as well. The gdb backtraces are usually more useful than our custom ones. Cheers, Samuel. -- Intel Open Source Technology Centre http://oss.intel.com/ _______________________________________________ connman mailing list [email protected] http://lists.connman.net/listinfo/connman
