Hi Marcel,
and thanks for the valid comments.
> Hi Petteri Tikander,
>
> > ---
> > drivers/atmodem/sim.c | 23 +++++++++++++++++++++
> > include/sim.h | 3 ++
> > src/sim.c | 52
> > +++++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 78
> > insertions(+), 0 deletions(-)
>
> these need to be split into two separate patches no matter what. First
> one is adding a feature to the core. The other is implementing it in the
> atom driver.
OK.
>
> I know that you guys talked with Denis about this, so I might be just
> missing something here.
>
You did not necessarily missed anything:) So logically presim-state in SIM-
enabled FDN has sense (or even switching to OFF_LINE-state). It can be my
misunderstanding. We had some discussion with Denis of this issue, and I
understood it so, that oFono-api has to remove rest of the atoms, if FDN is
enabled (so also SIM- and voicecall atoms). At least it was easy way to filter
any calls.
Well, the removal of the SIM-atom (what my patch-code did) wasn't a good idea
from me. Seems that SIM-initialization routine will be done every time when
waking up from the POWER-OFF state to the emergency-mode :(
Anyway, I ensured the situation (staying in the presim-state) by next kind of
test (by making some test-code, where FDN-enabled SIM causes keeping in the
PRE_SIM-state):
1) enable FDN in SIM
2) enable modem:
- create SIM-atom/voice call atom/.../read some EFs/.../read
IMSI-code/read
FDN enabled-status
3) dial to some number in the FDN-list
4) call passed!
Well, why the call passed to B-subscriber: it shouldn't because modem should
be offline (AT+CFUN=1 not given yet). But my test-modem doesn't understand
'AT+CFUN=4'- command, so it can pass calls anyway. OK, most modems will
not start dialing. But anyway, if modem is in the offline, it still receives
'ATD####'-command through atom/driver-chain, handles it, and probably returns
some error.
So the basic question is now, should the voicecall-atom also be removed in
the call-restrictions (FDN/BDN-situations etc), causing calls to be filtered in
very early phase. Or voicecall-atom not even created until switching to
ON_LINE (emergency-mode logic switches). Or is it OK to accept situation
mentioned below (let trust the modem to filter calls), and keep the voicecall-
logic visible all the time? Or, probably to add some condition in voicecall
atoms's methods for canceling calls when not being in the MODEM_ON_LINE-state.
> The requirement for minimum FDN support is really to power down the
> modem and not leave it in pre_sim state for emergency calls?
>
Thanks.
Br, Petteri
_______________________________________________
ofono mailing list
ofono@ofono.org
http://lists.ofono.org/listinfo/ofono
