Leo Martinez wrote: Hi, can you run hipd with gdb as follows and report the stack trace: gdb hipd <repeat the crash> backtrace Thanks and sorry for the delay in response.
Hello again everyone,I've been running through the scenario of the last message and found out some interesting things about it:- Whenever testing the Handover scenario of the manual (chapter 12) I always receive a SEGMENTATION FAULT error on the daemon of the crash host immediately after I have deleted the IPv6 address (crash). Reading the daemons log, its right after receiving the order to delete the SA of the old IPv6 address... I tested as well creating the new IPv6 address before deleting the old one as in a make-before-break attempt to see if no error appeared, both had the same outcome./debug(update.c:2460@hip_send_update): Mobility with single SA pair, readdress with no rekeying//debug(update.c:2461@hip_send_update): Reusing old SA/ /debug(update.c:2484@hip_send_update): make_new_sa=0/ /debug(update.c:2510@hip_send_update): not creating a new SA/ /debug(update.c:2526@hip_send_update): Reusing old SPI//debug(update.c:2579@hip_send_update): Netlink event was del, removing SAs for the address for this entry//debug(xfrmapi.c:410@hip_xfrm_state_delete): IPV6 SA deletion/ /debug(xfrmapi.c:415@hip_xfrm_state_delete): sport 0, dport 0//debug(xfrmapi.c:431@hip_xfrm_state_delete): deleting xfrm state with spi 0xb8de9c1b/ /debug(xfrmapi.c:432@hip_xfrm_state_delete): SA peer addr: 0x3ffe0000000000000000000000000004//debug(xfrmapi.c:410@hip_xfrm_state_delete): IPV6 SA deletion/ /debug(xfrmapi.c:415@hip_xfrm_state_delete): sport 0, dport 0//debug(xfrmapi.c:431@hip_xfrm_state_delete): deleting xfrm state with spi 0x6730a71/ /debug(xfrmapi.c:432@hip_xfrm_state_delete): SA peer addr: 0x3ffe0000000000000000000000000002//debug(update.c:2078@hip_update_src_address_list): //info(hadb.c:769@hip_hadb_get_peer_addr): entry def addr: 3ffe:0000:0000:0000:0000:0000:0000:0002//debug(hadb.c:1912@hip_hadb_get_spi_in_list): SPI=0xb8de9c1b/ /debug(update.c:2220@hip_update_src_address_list): I is now 0/ */Segmentation fault/*- Also, I tested having an IPv4 address assigned as well on the same interface and processed the same instructions. Curiously I found that the daemon selects the IPv4 address as the prefered locator and uses it to continue the previous SA. However, there is a problem since it creates a bogus IPv6 address and is never able to send the UPDATE messages. Heres a display of the SA before and after this case. Notice the SRC Addr after the mobility has been made.# hipconf get ha all Sending user message 22 to HIPD on socket 3 Sent 40 bytes Waiting to receive daemon info. 216 bytes received from HIP daemon HA is ESTABLISHED Local HIT: 2001:0010:28bf:1eef:6b9f:7003:ab62:d3d3 Peer HIT: 2001:0013:12af:9aa5:0e68:d8df:e9e4:abb2 Local LSI: 1.0.0.1 Peer LSI: 1.0.0.2 Local IP: 2001:0660:3203:1000:0280:06ff:febe:9116 Local NAT traversal UDP port: 0 Peer IP: 2001:0660:3203:1000:0215:e9ff:fe7c:39f4 Peer NAT traversal UDP port: 0 Peer hostname: hal root@bob:/home/bob# hipconf get ha all Sending user message 22 to HIPD on socket 3 Sent 40 bytes Waiting to receive daemon info. 216 bytes received from HIP daemon HA is ESTABLISHED Local HIT: 2001:0010:28bf:1eef:6b9f:7003:ab62:d3d3 Peer HIT: 2001:0013:12af:9aa5:0e68:d8df:e9e4:abb2 Local LSI: 1.0.0.1 Peer LSI: 1.0.0.2 /* Local IP: c544:0000:7eff:fffe:0000:0000:0000:0000*/ Local NAT traversal UDP port: 0 Peer IP: 2001:0660:3203:1000:0215:e9ff:fe7c:39f4 Peer NAT traversal UDP port: 0 Peer hostname: halAt the moment I have not been able to perform the handover tests described in chapter 12. Any of you guys have some suggestions regarding these observations?Thanks a lot. Best regards, Leo