SIM client

hi, i committed the SIM client process for layer23. it turns reading and writing jobs (from and to SIM files) into APDUs. because the SIM reader is not yet finished, it cannot be tested yet. after attaching the SIM card, a process in mobile/subscriber.c will request all required files form that card before starting the network selection process. (untested, as noted above) the mobility management is now cleaned up: TODO comments in the source code are replaced by SIM update commands to the SIM client. in order to make mobile application work with public networks, we need to perform the following steps: - finish the SIM reader and interface in layer 1 - add BTSAP interface to layer23 - add measurement reports to both TCH and SDCCH4/8 (*) - send updates for measurement report from layer23 via l23 api of layer 1 - add crpyto-key request to layer 1 and l1-l2 interface (dieter ?) * tests showed that the measurement report on SDCCH4/8 is requred. if not sent, the network releases a call (cleanly) after the mobile stays for a some seconds on an SDCCH4/8 (probably waiting for measurements, before completing the call). sending some dummy measurements completed the call successfully. beside that, the following issues need to be solved soon, in order to make the layer 1 firmware usable and stable: - additional structures for the libosmocore, see http://home.eversberg.eu/osmocore.patch <http://home.eversberg.eu/osmocore.patch> - freeze of layer 1 after about 40-60 data messages, see http://home.eversberg.eu/data.patch <http://home.eversberg.eu/data.patch> for testing. after a few location updates firmware freezes. (the location update will fail, because the patch disturbs location updating process, but the bug show up.) - syncing problem - clock drifts away after sync - subchannels 4..8 of SDCCH/8 and maybe - tch mode l1ctl message to select signalling only / speech codec of TCH/F / TCH/H. (including bearer capability processing in layer23) regards, andreas