git master ZTE testing results

[Aleksander Morgado]

On 29/01/14 20:46, Aleksander Morgado wrote:
>> 1) crash when deleting bearer with just a number, I guess we should
>> > detect this and either accept the '0' like we do for -m and -i or print
>> > a warning.  But not critical.
>> >
>> > $ mmcli -m 0 --delete-bearer=0
>> >
>> > (mmcli:6928): GLib-CRITICAL **: g_variant_new_object_path: assertion
>> > 'g_variant_is_object_path (object_path)' failed
>> > Segmentation fault
>> >
>> > 2) NM relies on PPP termination to notify MM that the bearer is
>> > disconnected, which apparnetly doesn't happen here.  NM needs to
>> > explicitly tell MM to terminate the connection after killing PPP.  NM
>> > also has an issue with handling of WWAN network ports that are not
>> > marked FLAG_WWAN.  Obviously NM issues, not MM.
>> >
>> > 3) MF627, MF691 (Rocket2, Icera), K3805-Z (Icera), and AC2726 (CDMA) all
>> > work OK.  I don't have any ZTE QMI devices apparently...
>> >
>> > So if it wasn't clear, the only critical bugs I found were NM git
>> > master bugs, not MM ones.
>> >
> Good! I'll look at the MM issue with the bearer deletion tomorrow, and
> tag the 1.2 release afterwards.

Fixed both --delete-bearer and --messaging-delete-sms to accept both a
bearer/sms index and the full bearer/sms DBus path.

-- 
Aleksander