better (maybe) freeing of RA process when an UE is removed from MAC
The RA process may be in the state: RA_active == FALSE and wait_ack_Msg4 == 1 Testing only the rnti should be enough, if everything else works properly. Also, don't break, just in case. (Not sure it makes a difference.)
This commit is contained in:
@@ -365,8 +365,7 @@ int rrc_mac_remove_ue(module_id_t mod_idP,rnti_t rntiP)
|
|||||||
RA_TEMPLATE *RA_template;
|
RA_TEMPLATE *RA_template;
|
||||||
for (i=0;i<NB_RA_PROC_MAX;i++) {
|
for (i=0;i<NB_RA_PROC_MAX;i++) {
|
||||||
RA_template = (RA_TEMPLATE *)&eNB_mac_inst[mod_idP].common_channels[pCC_id].RA_template[i];
|
RA_template = (RA_TEMPLATE *)&eNB_mac_inst[mod_idP].common_channels[pCC_id].RA_template[i];
|
||||||
if ((RA_template->RA_active == TRUE) &&
|
if (RA_template->rnti == rntiP){
|
||||||
(RA_template->rnti == rntiP)){
|
|
||||||
RA_template->RA_active=FALSE;
|
RA_template->RA_active=FALSE;
|
||||||
RA_template->generate_rar=0;
|
RA_template->generate_rar=0;
|
||||||
RA_template->generate_Msg4=0;
|
RA_template->generate_Msg4=0;
|
||||||
@@ -374,7 +373,7 @@ int rrc_mac_remove_ue(module_id_t mod_idP,rnti_t rntiP)
|
|||||||
RA_template->timing_offset=0;
|
RA_template->timing_offset=0;
|
||||||
RA_template->RRC_timer=20;
|
RA_template->RRC_timer=20;
|
||||||
RA_template->rnti = 0;
|
RA_template->rnti = 0;
|
||||||
break;
|
//break;
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
if (ret == 0) {
|
if (ret == 0) {
|
||||||
|
|||||||
Reference in New Issue
Block a user