For QCDM devices we want most of what MMSerialPort does, but not
the AT command handling stuff since the commands and responses
aren't AT commands nor are they even strings. So convert everything
that MMSerialPort does into a GByteArray, and let MMAtSerialPort
handle the conversion to strings when necessary.
If the modem echoed commands by default (since we may not have
initialized the modem yet), the echoed command would confuse
the PIN check reply parser.
Normally this would get done by the prober, but if the device
has a PIN enabled it'll reject almost all commands so the +CPMS?
in the prober will fail. Thus we have to do it after we've unlocked
the device.
Even just walking sysfs for driver and parent devices takes
time for ports we know we'll never use, so take a short-cut
and save some startup time. This reduces the startup
overhead to some 15%.
Some devices won't get to the initialization stage where we send
CMEE=1 (for numeric error codes) before they return some errors,
so handle the string representation of CME error codes too.
Clients can check the property to determine lock/unlock status and thus
unlock the modem before trying to connect if required.
Bits of the patch by dcbw (see the bug).
Some cards (Novatel S720 for example) can take a long time to start
a data call if the device isn't activated on the network or the
signal strength is low.
Some modems key the AT+CSS? response off their 1X state, so if the
modem has EVDO service but no 1X service, AT+CSS? will provide incorrect
registration state information and the registration checking will
end too early. Allow modems that can handle more specific registration
checking to skip the AT+CSS? part.
There are more places to handle, but these are the most critical. If
the modem is removed while a command is in-progress, the mm-callback-info
code will set info->modem to NULL. Make sure we check for that in
callbacks and return a reasonable error. Previous code would just
blindly forge ahead and die on a null dereference.
