triaged - leave as need-triage until more info available.

Some kind of backtrace needed.

See: http://wiki.merproject.org/wiki/Debugging

Note: new connman available: http://git.kernel.org/?p=network/connman/connman.git;a=blobdiff;f=ChangeLog;h=f7a6cacb50e7c6b53e59b685818072bf5fa77b81;hp=2b35dd140107fca99d5d45b71cef21753fb6a89e;hb=442b1fe603e005814f592a3dbcf0d0bfb13f961c;hpb=04cac51a36804f9d1327351144014501e528f824

Created attachment 3 [details]
Core

(sb2-obs-rpm-install-gdb) bt
#0  print_backtrace (offset=2) at src/log.c:189
#1  0x0002d4f0 in signal_handler (signo=<optimized out>) at src/log.c:219
#2  <signal handler called>
#3  table_cleanup (table=0x0) at src/iptables.c:1155
#4  0x0005839c in iptables_init (table_name=0x97888 "nat") at src/iptables.c:1244
#5  0x00058a38 in iptables_command (argv=0x940b8, argc=5) at src/iptables.c:1576
#6  __connman_iptables_command (format=0x69b58 "-t nat -F POSTROUTING") at src/iptables.c:1804
#7  0x00054e80 in disable_nat (interface=<optimized out>) at src/tethering.c:330
#8  0x000557e0 in __connman_tethering_update_interface (interface=0x0) at src/tethering.c:409
#9  0x0003fd68 in __connman_notifier_default_changed (service=0x0) at src/notifier.c:410
#10 0x0003af9c in service_indicate_state (service=0x95350) at src/service.c:4145
#11 0x0003b3a8 in __connman_service_ipconfig_indicate_state (service=0x95350, new_state=CONNMAN_SERVICE_STATE_DISCONNECT, 
    type=CONNMAN_IPCONFIG_TYPE_IPV6) at src/service.c:4422
#12 0x000311e0 in set_connected (user_data=0x92538) at src/network.c:1084
#13 0x000316a0 in connman_network_set_connected (network=0x92538, connected=0) at src/network.c:1145
#14 0x0002f72c in __connman_device_disable (device=0x93ae8) at src/device.c:310
#15 0x0002f810 in remove_device (device=0x93ae8) at src/device.c:358
#16 0x0002f8cc in remove_driver (driver=0x76bfc) at src/device.c:378
#17 connman_device_driver_unregister (driver=0x76bfc) at src/device.c:431
#18 0x000231ec in callback_system_killed () at gsupplicant/supplicant.c:331
#19 g_supplicant_unregister (callbacks=<optimized out>) at gsupplicant/supplicant.c:3461
#20 0x0001c168 in wifi_exit () at plugins/wifi.c:1443
#21 0x0002e064 in __connman_plugin_cleanup () at src/plugin.c:212
#22 0x000123d4 in main (argc=74708, argv=0xaed797a4) at src/main.c:380

triaged

looks like a bug - iptables_init() shouldn't fail

possibly needs HA CONFIG_ checks to ensure kernel has NAT

possible depends on 228

(In reply to comment #4)
> triaged
> 
> looks like a bug - iptables_init() shouldn't fail
> 
> possibly needs HA CONFIG_ checks to ensure kernel has NAT
> 
> possible depends on 228

No, it's not dependent on 228.

I have the same bug in Nemo Mobile on my N900, but in my case connman dies by itself some time after connecting to my access point. I presume, this is caused by some IPv6 routing advertisements in my network.

The problem is that Linux kernel in Nemo Mobile doesn't have iptables modules at all. As a result, iptables_init() function just returns NULL when trying to insmod ip_tables module, which leads to segfault in table_cleanup() function invoked with NULL argument.

There's no need to fix this bug, it was fixed long ago in upstream:
http://git.kernel.org/?p=network/connman/connman.git;a=commitdiff;h=a4ae1cf0ac362fdda02522eb8bed94059d3ed3e8

Just include the newest version of connman in Mer build, and the problem will go away.

On the other hand, building the Linux kernel with iptables will be beneficial as well.

There'll be a new Nemo image today, can one of you verify if this is fixed there?

Confirm fixed in 20120702 (although the settings app has a new bug... Nemo#203)