Not sure if I understand the logic properly here. Seems we're only tracking the ofono attached state, to make that work better from the software level, but isn't you "lying" when you're returning NOT_REGISTERED here?
Shouldn't we also set the modem state to reflect the returned value here? Guess this goes back to the airplane mode support.
Is ofono calling set_attached with 0 when you disable the connection from the network indicator? If so, we might probably want to change the modem state from there instead.
73 - gd->status = status; REGISTRATION_ STATUS_ REGISTERED || REGISTRATION_ STATUS_ ROAMING) ; false; return !REGISTERED"); REGISTRATION_ STATUS_ NOT_REGISTERED;
74 - gd->tech = tech;
75 + gd->rild_status = status;
76 +
77 + attached = (status == NETWORK_
78 + status == NETWORK_
79 +
80 + if (attached && gd->ofono_attached == FALSE) {
81 + DBG("attached=true; ofono_attached=
82 + status = NETWORK_
83 + }
84
Not sure if I understand the logic properly here. Seems we're only tracking the ofono attached state, to make that work better from the software level, but isn't you "lying" when you're returning NOT_REGISTERED here?
Shouldn't we also set the modem state to reflect the returned value here? Guess this goes back to the airplane mode support.
Is ofono calling set_attached with 0 when you disable the connection from the network indicator? If so, we might probably want to change the modem state from there instead.