#1823 Amf: Standby Amfd crash during Nodegroup delete

[Nagendra Kumar]

Steps to reproduce

1. Start one controller.
2. In Standby Amfd keep a sleep of 7 sec after confg read and before applier set as below:

+++ b/osaf/services/saf/amf/amfd/role.cc
@@ -219,6 +219,9 @@ uint32_t avd_standby_role_initialization
}
cb->init_state = AVD_CFG_DONE;
LOG_ER("1. Waiting ..........................");
sleep(7);
LOG_ER("2. Waiting ..........................");

if (avd_imm_applier_set() != SA_AIS_OK) {
LOG_ER("avd_imm_applier_set FAILED");

1. Run the below steps:

immcfg -c "SaClmNode" -a saClmNodeLockCallbackTimeout=50000000000 safNode=PL-5,safCluster=myClmCluster
immcfg -c "SaAmfNode" -a saAmfNodeSuFailoverMax=2 -a saAmfNodeSuFailOverProb=1200000000000 -a saAmfNodeClmNode=safNode=PL-5,safCluster=myClmCluster safAmfNode=PL-5,safAmfCluster=myAmfCluster
immcfg -c "SaClmNode" -a saClmNodeLockCallbackTimeout=50000000000 safNode=PL-6,safCluster=myClmCluster
immcfg -c "SaAmfNode" -a saAmfNodeSuFailoverMax=2 -a saAmfNodeSuFailOverProb=1200000000000 -a saAmfNodeClmNode=safNode=PL-6,safCluster=myClmCluster safAmfNode=PL-6,safAmfCluster=myAmfCluster
immcfg -c "SaAmfNodeGroup" -a saAmfNGNodeList=safAmfNode=PL-5,safAmfCluster=myAmfCluster safAmfNodeGroup=nagu,safAmfCluster=myAmfCluster
immcfg -m -a saAmfNGNodeList+=safAmfNode=PL-6,safAmfCluster=myAmfCluster safAmfNodeGroup=nagu,safAmfCluster=myAmfCluster

1. Start SC-2 controller and when the following logs comes issue the command below:
   May 11 13:56:15 PM_SC-2 osafamfd[11589]: ER 1. Waiting .........................
   immcfg -m -a saAmfNGNodeList-=safAmfNode=PL-5,safAmfCluster=myAmfCluster safAmfNodeGroup=nagu,safAmfCluster=myAmfCluster
2. Now, runt the following commands:
   amf-adm lock safAmfNode=PL-5,safAmfCluster=myAmfCluster
   amf-adm lock-in safAmfNode=PL-5,safAmfCluster=myAmfCluster
   immcfg -d safAmfNode=PL-5,safAmfCluster=myAmfCluster
   immcfg -d safAmfNodeGroup=nagu,safAmfCluster=myAmfCluster
   Observed behaviour

---

Standby Amfd will crash.

Expected behaviour

Amfd shouldn't crash.

Error messages

May 11 13:56:32 PM_SC-2 osafamfnd[11599]: WA AMF director unexpectedly crashed
May 11 13:56:32 PM_SC-2 osafamfnd[11599]: Rebooting OpenSAF NodeId = 131599 EE Name = , Reason: local AVD down(Adest) or both AVD down(Vdest) received, OwnNodeId = 131599, SupervisionTime = 60
May 11 13:56:32 PM_SC-2 opensaf_reboot: Rebooting local node; timeout=60

Back trace:

Core was generated by `/usr/local/lib/opensaf/osafamfd --tracemask=0xffffffff'.
Program terminated with signal 11, Segmentation fault.

0 0x00007f9770886812 in __strlen_sse2 () from /lib64/libc.so.6

(gdb) bt
0 0x00007f9770886812 in strlen_sse2 () from /lib64/libc.so.6
1 0x00007f977084e3dd in vfprintf () from /lib64/libc.so.6
2 0x00007f97708f2609 in vsnprintf_chk () from /lib64/libc.so.6
3 0x00007f9772610724 in output () from /usr/local/lib/libopensaf_core.so.0
4 0x00007f9772610b4b in _logtrace_trace () from /usr/local/lib/libopensaf_core.so.0
5 0x0000000000441028 in ng_ccb_completed_cb(CcbUtilOperationData*) () at nodegroup.cc:471
6 0x000000000042ffd1 in ccb_completed_cb(unsigned long long, unsigned long long) () at imm.cc:988
7 0x00007f97721b7efe in imma_process_callback_info () from /usr/local/lib/libSaImmOi.so.0
8 0x00007f97721bacd6 in imma_hdl_callbk_dispatch_all () from /usr/local/lib/libSaImmOi.so.0
9 0x00007f97721aef44 in saImmOiDispatch () from /usr/local/lib/libSaImmOi.so.0
10 0x0000000000407c20 in main () at main.cc:717

[Nagendra Kumar]

changeset: 7618:adc8204ae015
branch: opensaf-4.7.x
parent: 7615:5ca256b58104
user: Nagendra Kumarnagendra.k@oracle.com
date: Thu May 12 16:13:06 2016 +0530
summary: amfd: avoid null pointer access [#1823]

changeset: 7619:6821f86a70ad
branch: opensaf-5.0.x
parent: 7616:48e30c0d5dfd
user: Nagendra Kumarnagendra.k@oracle.com
date: Thu May 12 16:13:46 2016 +0530
summary: amfd: avoid null pointer access [#1823]

changeset: 7620:36ecc9e083d1
tag: tip
parent: 7617:a6f87948dab5
user: Nagendra Kumarnagendra.k@oracle.com
date: Thu May 12 16:13:59 2016 +0530
summary: amfd: avoid null pointer access [#1823]

[staging:adc820]
[staging:6821f8]
[staging:36ecc9]