Re: GN_OP_OnSMS in nk6510

[Pawel Kot]

Hi,

On 7/4/07, Bastien Nocera <address@hidden> wrote:
> Because I don't use horrible horrible globals, and I want to provide my
> own callback data. As in:

Globals are usually horrible but not always :)

> /* Set up the new SMS callback */
>                 gn_data_clear(&l->phone_state->data);
>                 l->phone_state->data.on_sms = phonemgr_listener_new_sms_cb;
>                 l->phone_state->data.user_data = l;
>                 if (gn_sm_functions (GN_OP_OnSMS, &l->phone_state->data, 
> &l->phone_state->state) == GN_ERR_NONE) {
>                         l->supports_sms_notif = TRUE;
>                 }
>
> And the callback would look like:
> static gn_error
> phonemgr_listener_new_sms_cb (gn_sms *message, struct gn_statemachine *state, 
> void *user_data)
> {
>         PhonemgrListener *l = (PhonemgrListener *) user_data;
>         gn_sms *msg;
>
>         /* The message is allocated on the stack in the driver, so copy it */
>         msg = g_memdup (message, sizeof (gn_sms));
>         g_async_queue_push (l->queue, msg);
>         g_idle_add ((GSourceFunc) phonemgr_listener_push, l);
>
>         return GN_ERR_NONE;
> }
>
> PhonemgrListener is a GObject, and this sort of construct is usual in
> signal/callback driven applications.

Fair enough. Just one more question, why do you need this:
+       void *call_notification_user_data;
} at_driver_instance;
?

take care,
pkot
--
Pawel Kot