| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
add new MCCMNC of EE UK(23430,23431,23432,23433,23434,23476,23501,23502,23577)
add new MCCMNC of ATT(310650,310980,90118)
add new MCCMNC of Verizon(312770)
add new MCCMNC of Vodafone(46601,46603)
add new MCCMNC of Orange(21409,23105,26005)
|
|
add new device(0x413c:0x81e4 0x413c:0x81e6)
|
|
This URC is emitted when the SIM phonebook is initialized. We don't care
about that but don't want this URC to mess with the initialization
sequence either.
Signed-off-by: Sven Schwermer <sven.schwermer@disruptive-technologies.com>
|
|
Signed-off-by: Sven Schwermer <sven.schwermer@disruptive-technologies.com>
|
|
This is the factory-default setting. Make sure to support it as well as
the ECM/RNDIS settings. The L610 in GTUSBMODE=31 only enumerates with
serial USB interfaces using the option driver, so we need to allow that
driver too.
Signed-off-by: Sven Schwermer <sven.schwermer@disruptive-technologies.com>
|
|
Ignore the error SIM_NOT_INSERTED when ModemManager disable modem after SIM card removed
Fixes https://gitlab.freedesktop.org/mobile-broadband/ModemManager/-/issues/578
|
|
|
|
|
|
If any of the control ports ends up timing out 10 consecutive times,
the modem will be flagged as invalid and it will be reprobed from
scratch.
This allows us to detect modems that end up irresponsive in QMI or
MBIM while they're still exposed in e.g. the USB bus.
|
|
|
|
This fixes a bug introduced in a2a0e2d754. Since the state pointer is
always non-NULL, the function will always fail if GTRNDIS is inactive,
i.e. the modem will return state=0 and no cid field.
Signed-off-by: Sven Schwermer <sven.schwermer@disruptive-technologies.com>
|
|
The ECM dialling guide requires to check whether RNDIS is already active
before attempting to establish an ECM/RNDIS connection.
If it is active (regardless of its settings), we will disconnect it
first, before attempting the new connection with the user-provided
settings.
Signed-off-by: Sven Schwermer <sven.schwermer@disruptive-technologies.com>
|
|
Fix segfault that can occur during message text decoding if there are no
elements in SMS text
|
|
The required libqmi version was updated in configure.ac to 1.31.7 in
commit ac06052b97cc02e60ae1ed7a551727b3c14d6a22, but we forgot to do
the same in the meson setup.
|
|
software_package_version
Currently, MBIM broadband Telit modem does not store firmware revision
into 'software_package_version', so it won't be possible to specialize
behavior based on modem model (using mm_telit_model_from_revision).
This change overloads load-revision in MBIM broadband Telit in order to
call parent's loading and store the returned 'caps_firmware_info' into
'software_package_version' via mm_shared_telit_store_revision.
|
|
|
|
The new QCDM probing mechanism does not probe and grab QCDM ports
if the plugin is not allowed, so QCDM ports currently ignored in
compositions can be safely marked as ID_MM_PORT_TYPE_QCDM.
|
|
The new QCDM port probing and grabbing mechanism implements a new
property for detecting the plugins which require the QCDM port to
be probed and grabbed.
Add the property to the affected plugins.
|
|
The QCDM/DIAG port is usually nowadays exclusively used by applications
gathering traces from the modem, so avoid port probing and grabbing
when plugins set the property MM_PLUGIN_ALLOWED_QCDM.
A new property MM_PLUGIN_REQUIRED_QCDM is created for those plugins
requiring the QCDM port for properly using a modem.
|
|
Signed-off-by: Sven Schwermer <sven.schwermer@disruptive-technologies.com>
|
|
|
|
In mm_telit_model_from_revision there is nothing that prevents to use
the input revision string if it is NULL, which currently happens if
the modem is pure MBIM (i.e. LE910-V2 pid:0x0032).
Fix the above error returning a default safe value and not using the
NULL variable.
|
|
ModemManager handles suspend and resume signals sent from powerd
Fixes https://gitlab.freedesktop.org/mobile-broadband/ModemManager/-/issues/547
|
|
Tested w/ Lenovo firmware T99W175.F0.1.0.0.9.DT.003.055 (old unlock method) and
T99W175.F0.1.0.0.9.DT.003.079 (new unlock method)
|
|
We don't have a clear way to report "unknown" signal quality, so for
now just skip printing it altogether in the mmcli human output if the
modem is not yet enabled.
|
|
E.g. if we reach the failed state before the bearer list is created.
|
|
Instead of not creating a modem object, create it in failed state with
the "unknown capabilities" failed state reason.
|
|
|
|
A modem using an eSIM without profiles should not be allowed to get
enabled, it should be really treated as a modem without a physical
SIM.
|
|
|
|
These are current settings in the modem, they don't really depend on
whether there is a SIM or not.
|
|
|
|
|
|
Avoid calling the MMBaseSim class method directly, and use the helper
mm_base_sim_load_sim_identifier() instead.
Also, rework a bit the logic to have a single completion method once
we know the current ICCID.
|
|
|
|
If the reported eSIM doesn't have profiles, there is no point in
trying to load properties like ICCID, IMSI or operator name/id.
If the reported SIM is a physical SIM, there is no point in trying
to load properties like EID or eSIM status.
|
|
|
|
|
|
If we created the SIM objects during the SIM slots processing, we
should not attempt to reload the settings that were already set during
the re-initialization, as that may end up clearing the initial
properties.
|
|
All properties loaded during initialization are meant to be loaded
only once, no point in saying that for each property.
|
|
|
|
|
|
E.g. if the error reported by the MbimDevice is a timeout error, it
means the connection attempt is still ongoing in the device, we need
to ask it to stop.
|
|
|
|
|
|
|
|
Some PCI modems prefer customizing the subsytem vendor ID,
instead of the vendor ID.
Add a filter for the couple vendor/subsystem vendor IDs.
|
|
Subsystem vendor ID can be used for identifying PCI modems,
so expose the property.
|
