[SR-Users] Proposal for dialog: DLG_STATE_EARLY

[Daniel-Constantin Mierla]

Hello,

I have some comments related to the patches, as I couldn't dig much into 
sources due to traveling constraints. See them inline.

On 6/14/13 2:10 PM, Halina Nowak wrote:
> These modifications were implementated for dialogs having PRACK and UPDATE
>
> --- a/modules/dialog/dlg_handlers.c    Fri Jun 14 13:45:41 2013 +0200
> +++ b/modules/dialog/dlg_handlers.c    Fri Jun 14 13:55:24 2013 +0200
> @@ -1249,16 +1249,18 @@
>      }
>
>      if ( (event==DLG_EVENT_REQ || event==DLG_EVENT_REQACK)
> -    && new_state==DLG_STATE_CONFIRMED) {
> +         && (new_state==DLG_STATE_CONFIRMED || 
> new_state==DLG_STATE_EARLY)) {

This above is to catch PRACK, right? UPDATE should be sent after 200ok, 
or is allowed also for early dialogs?

>
>          timeout = get_dlg_timeout(req);
>          if (timeout!=default_timeout) {
>              dlg->lifetime = timeout;
>          }
> -        if (update_dlg_timer( &dlg->tl, dlg->lifetime )==-1) {
> +        if (new_state!=DLG_STATE_EARLY) {
> +          if (update_dlg_timer( &dlg->tl, dlg->lifetime )==-1) {
>              LM_ERR("failed to update dialog lifetime\n");
> +          }
>          }
> -        if (update_cseqs(dlg, req, dir)!=0) {
> +        if ((event != DLG_EVENT_REQACK) && (update_cseqs(dlg, req, 
> dir)!=0)) {
>              LM_ERR("cseqs update failed\n");
>          } else {
>              dlg->dflags |= DLG_FLAG_CHANGED;
>
> --- a/modules/dialog/dlg_hash.c    Fri Jun 14 13:45:41 2013 +0200
> +++ b/modules/dialog/dlg_hash.c    Fri Jun 14 13:55:24 2013 +0200
> @@ -883,6 +883,7 @@
>              break;
>          case DLG_EVENT_REQACK:
>              switch (dlg->state) {
> +                         case DLG_STATE_EARLY:
>                  case DLG_STATE_CONFIRMED_NA:
>                      dlg->state = DLG_STATE_CONFIRMED;
Here it seems to go to state DLG_STATE_CONFIRMED due to an ACK even 
there was no 200ok reply. Why is needed like that or have I 
misunderstood something?

Cheers,
Daniel

-- 
Daniel-Constantin Mierla - http://www.asipto.com
http://twitter.com/#!/miconda - http://www.linkedin.com/in/miconda

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sip-router.org/pipermail/sr-users/attachments/20130621/5cd29709/attachment-0001.html>