[PATCH 05/16] P2P: Send M2D event on the global control interface

[Peer, Ilan]

Hi Jouni,

> -----Original Message-----
> From: Jouni Malinen [mailto:j at w1.fi]
> Sent: Friday, June 19, 2015 10:49
> To: Peer, Ilan
> Cc: hostap at lists.shmoo.com
> Subject: Re: [PATCH 05/16] P2P: Send M2D event on the global control
> interface
> 
> On Wed, Jun 17, 2015 at 04:24:51PM +0300, Ilan Peer wrote:
> > As this is not an interface specific event, and the P2P related events
> > should be sent on the global control interface.
> 
> > diff --git a/wpa_supplicant/wps_supplicant.c
> > b/wpa_supplicant/wps_supplicant.c @@ -584,7 +584,8 @@ static void
> wpa_supplicant_wps_event_m2d(struct wpa_supplicant *wpa_s,
> >  	wpas_notify_wps_event_m2d(wpa_s, m2d);  #ifdef CONFIG_P2P
> >  	if (wpa_s->parent && wpa_s->parent != wpa_s) {
> > -		wpa_msg(wpa_s->parent, MSG_INFO, WPS_EVENT_M2D
> > +		wpa_msg_global(wpa_s->global->p2p_init_wpa_s,
> > +			       MSG_INFO, WPS_EVENT_M2D
> >  			"dev_password_id=%d config_error=%d",
> >  			m2d->dev_password_id, m2d->config_error);
> 
> Couldn't this break some existing uses? As an example, wlan0 could be used
> to manage P2P operations and when a separate group interface is used, the
> WPS-M2D event used to go out on that group interface's ctrl_iface and wlan0
> interface's ctrl_iface. With this change, that latter would not happen and if an
> upper layer application was only registered for events on wlan0, it would not
> see this new global event.
> 

Yes, this would break this use case your described. However, in the case of a dedicated P2P Device interface the only alternative is to use the global control interface (and with higher layer applications that use the global control interface for P2P operations should really also listen on it for the events).

Would you consider a solution that fires the event on both the parent interface and the global interface as a valid one? 

Regards,

Ilan.