[RFC] stanger/cinterion branch

[Aleksander Morgado]

Hey Matthew & Dan,

>
> I ran the code a few thousand times for connecting -> data connection ->
> disconnect through our CI suite. That looks good. However I did find during
> a better manual testing that if you remove the SIM while a data connection
> is up the dbus API isn't updated with the disconnect event.
>
> In:
>    swwan_check_status_ready()
>          ....
>
>     response = mm_base_modem_at_command_finish (modem, res, &error);
>
>     if (!response) {
>
>         g_task_return_error (task, error);
>
>         goto out;
>
>     }
>
>
> We fall into this and exit without updating the status. I think it should
> be:
>
>         g_task_return_int (task, (gssize)
> MM_BEARER_CONNECTION_STATUS_DISCONNECTED);
>
>

What's the output of the SWWAN status check when the SIM is gone? is
it saying "ERROR"?

>
>
>> I have no idea how that could happen.
>
>
> Test weird ways, get weird problems lol. I tracked this down to the very
> weird way I tested, which wasn't valid in the end so disregard.
>

Ok.

>> In which step do you get the "unspecified gprs error"?
>
>
> Not in a step but in a setup, using non-OEM SIMs. The guy's at Cinterion
> told us to try and set this value manually based on the ICCID :0, yikes.
> This is only really a problem for USA users using non-OEM SIMs, so it may
> not be a problem that needs to be solved. I.E. we will have to since we have
> custom SIMs for both Ver & AT&T, but the main MM user's may never run into
> this. I guess knowing that do you think it's important to figure out
> something in the MM scope or should it just be handled on it's own by users
> in this corner case?
>

This issue seems to be very specific to a given user setup, don't
know. I'm tempted to say we shouldn't have anything generic in
ModemManager to handle that.

> Other than that this looks solid. Exciting times!

Unless Dan has any objection, I'll get the branch merged into git
master in the following days.

-- 
Aleksander
https://aleksander.es