Hi Gary, Yes it was wrong before refactoring, I'll correct this before pushing./Thanks HansN
-----Original Message----- From: Gary Lee [mailto:gary....@dektech.com.au] Sent: den 15 september 2015 05:43 To: Hans Nordebäck; praveen.malv...@oracle.com; nagendr...@oracle.com Cc: opensaf-devel@lists.sourceforge.net Subject: Re: [PATCH 1 of 1] amfd: Range based for loop not correct in map_su_to_node [#1478] Hi Hans Please refer to one comment marked with [GL]. On 14/09/15 22:23, Hans Nordeback wrote: > osaf/services/saf/amf/amfd/ckpt_dec.cc | 4 +++- > osaf/services/saf/amf/amfd/node.cc | 2 +- > osaf/services/saf/amf/amfd/sgproc.cc | 4 +++- > osaf/services/saf/amf/amfd/su.cc | 20 ++++++++++---------- > 4 files changed, 17 insertions(+), 13 deletions(-) > > > diff --git a/osaf/services/saf/amf/amfd/ckpt_dec.cc > b/osaf/services/saf/amf/amfd/ckpt_dec.cc > --- a/osaf/services/saf/amf/amfd/ckpt_dec.cc > +++ b/osaf/services/saf/amf/amfd/ckpt_dec.cc > @@ -3043,9 +3043,11 @@ static uint32_t dec_ng_admin_state(AVD_C > AVD_AVND *node = avd_node_get(*iter); > AVD_SU *su = NULL; > //If this node has any susi on it. > - for (const auto& su : node->list_of_su) > + for (const auto& tmp : node->list_of_su) { > + su = tmp; > if (su->list_of_susi != NULL) > break; > + } > if ((ng->saAmfNGAdminState == SA_AMF_ADMIN_SHUTTING_DOWN) && > (su != NULL)) > /* Still some assignments are not removed on the node. > if ng in SHUTTING_DOWN state and contoller role > changes then > diff --git a/osaf/services/saf/amf/amfd/node.cc > b/osaf/services/saf/amf/amfd/node.cc > --- a/osaf/services/saf/amf/amfd/node.cc > +++ b/osaf/services/saf/amf/amfd/node.cc > @@ -88,7 +88,7 @@ bool AVD_AVND::is_node_lock() { > for (curr_susi = su->list_of_susi; > (curr_susi) && ((SA_AMF_HA_QUIESCING != curr_susi->state) || > ((AVD_SU_SI_STATE_UNASGN == curr_susi->fsm))); > - curr_susi = curr_susi->su_next); \ > + curr_susi = curr_susi->su_next); > if (curr_susi) > return false; > } > diff --git a/osaf/services/saf/amf/amfd/sgproc.cc > b/osaf/services/saf/amf/amfd/sgproc.cc > --- a/osaf/services/saf/amf/amfd/sgproc.cc > +++ b/osaf/services/saf/amf/amfd/sgproc.cc > @@ -712,7 +712,9 @@ void avd_su_oper_state_evh(AVD_CL_CB *cb > */ > avd_node_oper_state_set(node, > SA_AMF_OPERATIONAL_DISABLED); > node->recvr_fail_sw = true; > - for (const auto& su : node->list_of_su) { > + // TODO (uabhano) is the below code correct? > + for (const auto& i_su : node->list_of_su) { > + TRACE("Is this for loop correct with > regard to i_su and su > +below? %s", i_su->name.value); [GL] As you suspect, looks like below should be i_su->set_readiness_state? Looks like it was wrong even before refactoring. > > su->set_readiness_state(SA_AMF_READINESS_OUT_OF_SERVICE); > } > } /* if > (n2d_msg->msg_info.n2d_opr_state.node_oper_state == > SA_AMF_OPERATIONAL_DISABLED) */ > diff --git a/osaf/services/saf/amf/amfd/su.cc > b/osaf/services/saf/amf/amfd/su.cc > --- a/osaf/services/saf/amf/amfd/su.cc > +++ b/osaf/services/saf/amf/amfd/su.cc > @@ -489,6 +489,8 @@ static AVD_AVND *map_su_to_node(AVD_SU * > AVD_SU *su_temp = NULL; > AVD_AVND *node = NULL; > std::set<std::string>::const_iterator node_iter; > + std::vector<AVD_SU*>::const_iterator su_iter; > + std::vector<AVD_SU*> *su_list = nullptr; > > TRACE_ENTER2("'%s'", su->name.value); > > @@ -517,20 +519,18 @@ static AVD_AVND *map_su_to_node(AVD_SU * > osafassert(node); > > if (su->sg_of_su->sg_ncs_spec == true) { > - for (const auto& su_temp : node->list_of_ncs_su) { > - if (su_temp->sg_of_su == su->sg_of_su) > - break; > - } > + su_list = &node->list_of_ncs_su; > + } else { > + su_list = &node->list_of_su; > } > > - if (su->sg_of_su->sg_ncs_spec == false) { > - for (const auto& su_temp : node->list_of_su) { > - if (su_temp->sg_of_su == su->sg_of_su) > - break; > - } > + for (su_iter = su_list->begin(); su_iter != su_list->end(); > ++su_iter) { > + su_temp = *su_iter; > + if (su_temp->sg_of_su == su->sg_of_su) > + break; > } > > - if (su_temp == NULL) > + if (su_iter == su_list->end()) > goto done; > } > ------------------------------------------------------------------------------ _______________________________________________ Opensaf-devel mailing list Opensaf-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/opensaf-devel
