Instead of deciding in advance which data port to use, we let the dialling
operation gather it. For the generic dialling logic, ATD-based, always an
'AT' port will be used as data port, even if we grabbed a 'net' port. Those
plugins that can work with 'net' ports will grab the specific 'net' port
themselves.
If the primary port is gone (e.g. when going to sleep) and we are just in the
middle of a connection attempt, we won't be able to receive any unsolicited
message regarding the status of the attempt. So, if we detect that the port is
forced to get closed, we'll just treat it as a connection failure.
http://code.google.com/p/chromium-os/issues/detail?id=35391
The 'hso' driver tags Icera-based modems' proprietary protocol port
as a DIAG port, but since they aren't Qualcomm-based, it's not
a DIAG port. Just turn off AT probing and let QCDM probing
fail the port instead of assuming it's a DIAG port.
The logic gets completely stuck when this happens:
Stack trace below:
#0 0x77661424 in __kernel_vsyscall ()
#1 0x77337c3c in pthread_cond_wait ()
#2 0x773cebaa in g_cond_wait () from /usr/lib/libglib-2.0.so.0
#3 0x774c03cc in g_cancellable_disconnect () from /usr/lib/libgio-2.0.so.0
#4 0x76955d36 in connect_cancelled_cb (cancellable=0x78e055a0, self=0x78e0b590)
#5 0x77460982 in g_cclosure_marshal_VOID__VOIDv () from /usr/lib/libgobject-2.0.so.0
#6 0x7745ed8a in ?? () from /usr/lib/libgobject-2.0.so.0
#7 0x77478435 in g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0
#8 0x77478eb3 in g_signal_emit () from /usr/lib/libgobject-2.0.so.0
#9 0x774c01eb in g_cancellable_cancel () from /usr/lib/libgio-2.0.so.0
#10 0x776a0eab in mm_bearer_disconnect (self=0x78e0b590, callback=0x776c5980 <disconnect_ready>,
#11 0x776c57de in disconnect_next_bearer (ctx=0x78e12870) at mm-iface-modem-simple.c:898
#12 0x776c58d2 in disconnect_auth_ready (self=0x78df3048, res=0x78e06210, ctx=0x78e12870)
#13 0x774fed25 in g_simple_async_result_complete () from /usr/lib/libgio-2.0.so.0
#14 0x776a8c4e in authorize_ready (authp=0x78db68d0, res=0x76801638, simple=0x78e06210)
#15 0x774fed25 in g_simple_async_result_complete () from /usr/lib/libgio-2.0.so.0
#16 0x774fee3e in ?? () from /usr/lib/libgio-2.0.so.0
#17 0x7738a7a2 in ?? () from /usr/lib/libglib-2.0.so.0
#18 0x7738ce83 in g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0
#19 0x7738d248 in ?? () from /usr/lib/libglib-2.0.so.0
#20 0x7738d6eb in g_main_loop_run () from /usr/lib/libglib-2.0.so.0
#21 0x77696a7d in main (argc=2, argv=0x7fbb1f04) at main.c:158
http://code.google.com/p/chromium-os/issues/detail?id=36448
Instead of getting the port type hints while grabbing each probed port, we now
run a custom init operation in the probing which already gives us the port type,
and therefore allows us to completely skip port probings.
Also now added hints for 'Diag' (QCDM) ports and AT/'Modem' ports (I guess it's
the one for PPP if we don't have a net port, which is unlikely anyway in HSO).
This makes the probing of an Option/HSO modem almost instant.
We will not report 'CS' as a supported mode every time '2G' is supported. This
actually was forcing all plugins to handle a 'CS' fallback when they didn't have
CS-specific mode setup. So, to simplify things, we will only report 'CS' as
supported for those plugins which actually allow to select 'CS' mode (e.g. the
'wavecom' plugin).
If we are requested to cancel the connection, we first need to wait for the
connection attempt to finish before issuing the disconnect command, as otherwise
the modem just returns an error saying that it cannot perform the operation and
at the end we end up with the modem connected but ModemManager thinking that it
isn't.
Both the ModemManager daemon and the mmcli will now include `libmm-glib.h' only.
We also handle two new special `_LIBMM_INSIDE_MM' and `LIBMM_INSIDE_MMCLI'
symbols, which if included before the `libmm-glib.h' library allow us to:
* Don't include the libmm-glib high level API in the ModemManager daemon, as
the object names would clash with those in the core.
* Define some of the methods of helper objects to be included only if compiling
ModemManager daemon or the mmcli.
Different ports of the same modem may get handled by different drivers. We
therefore need to provide a list of drivers (new `Modem.Drivers' property with
signature 'as') instead of just one (removed `Modem.Driver' property with
signature 's').
$ sudo mmcli -m 0 | grep drivers
| drivers: 'qcserial, qmi_wwan'
In the connection sequence, we make sure that the Bearer object is valid by
keeping a reference in the Dial3gppContext (actually, another one kept by the
GSimpleAsyncResult as well). But we are considering here the case where the
connection context is completed and freed by an unsolicited message handler
before we get the reply to the AT command, so to properly ensure that the bearer
object is still valid when we try to check if the context is in the priv struct,
we need to keep an extra ref around.
In the connection sequence, we keep the context in the private info of the
Bearer object, so that we can complete and free it when we receive OWANCALL
unsolicited messages. Now, the reply of the OWANCALL itself may get processed
as an unsolicited message, so effectively we're processing and finishing the
connection/disconnection context *before* we process the reply of the AT
command.
This patch ensures that we do not try to re-use the connection context after it
has been processed by the unsolicited messages handler.
There's no real point in maintaining a separate `MMPlugin' interface, as all the
plugins will inherit from `MMPluginBase', so just merge them and simplify
everything.
Before this, we only exported the modem to DBus when all ports were organized,
in order to make sure that we select as primary port the one we really want and
not the first AT port grabbed. Given that to get all the ports organized we also
needed to wait to get all the ports grabbed, we can now also defer the creation
of the modem object until all the ports get grabbed. This allows us to create
different types of objects based on the ports available (e.g. we can now create
QMI-supported modem objects if we see a QMI port around).
The base MMBearer object will receive the MMBearerProperties configuration,
which contains every generic and plugin-specific set up. Not every configuration
parameter will be used by every implementation (e.g. not every bearer needs
user/password).
The Bearer object will expose every configuration parameter received in its
`Properties' property in the interface, even if it's not really used.
The new `mm_bearer_report_disconnection()' allows subclasses to notify about
being disconnected, so that every layer of inheritance can chain its own code
to reset the connection status.
This commit partially disables the logic included in commit 981222. Now
subclasses (actually, no one) are not allowed to g_object_set() the
MM_BEARER_STATUS property.
Whenever 3GPP location source gets enabled, we'll launch new registration status
checks (to get updated LAC/CI) and new operator code/name checks (to get updated
MCC/MNC).
Additional changes were needed in the HSO plugin, so that the specific location
gathering enabling implemented by the HSO modem chains up parent's one first.
The `physdev' stored in the probe is the sysfs path of the parent device, that
owning all ports. We need to build the setup file path using the port's sysfs
path instead.
