We have assumed until now that all QCDM ports are based on TTY
drivers, e.g. exposed via USB.
This may no longer be true, so allow creating QCDM ports with
an explicit subsystem instead of defaulting always to TTY.
Split in its own method the per-subsystem port creation mechanism, and
apply all common AT port settings (e.g. response parser, flags) in a
single place.
This check is pointless given that we're not implementing API, if
anything it should be an assert. Anyway, just get rid of it, so that
we don't need to update it on every new subsystem we add as supported.
Back in Linux < 3.6 days, the cdc-wdm ports exposed by the QMI driver
were flagged as owned by the 'usb' subsystem. That changed in 3.6 when
the subsystem was renamed to 'usbmisc':
https://mail.gnome.org/archives/networkmanager-list/2012-June/msg00125.html
So, rename the port subsystem type enumn to 'usbmisc'.
It is no longer true that all QMI ports are exposed by the qmi_wwan
driver and that all MBIM ports are exposed by the cdc_mbim driver.
There are other generic setups that allow exposing these types of
ports using different drivers, and usually we can also know the type
of port in advance via other means. Therefore, allow adding udev port
type hints for QMI and MBIM ports as well.
The rules to force ignoring certain network ports because the modem is
using some specific drivers should definitely only be applied for the
very known qmi_wwan and cdc_mbim drivers.
If network ports for QMI or MBIM modems are exposed using different
network drivers, don't ignore them.
There should be no need to do an early check to filter out ports of
the wrong subsystem.
For the user of these methods it is irrelevant if the FALSE is
returned because the port is of the wrong subsystem, or because the
test wasn't added to be probed, or because the test actually failed.
In other words, the ports where the test succeeded will only have
succeeded if they are of the correct subsystem and if the test was
actually executed successfully.
The ModemManager-names.h header file is generated on git builds, and
shipped within the release tarball, so treat it as a maintainer
managed file that should only be removed on 'maintainer-clean'.
This allows us to avoid requiring xsltproc to build release tarballs.
This is not just a rename of the rule, we also now avoid doing an
explicit check on the port name as well, and we rely on subsystem
checks only; i.e. the same logic applied for net ports.
The port candidate rules already do a 'cdc-wdm*' device name check
so it shouldn't be a big deal.
If multiple kernel device types inherit from MMKernerDevice, and those
are compared against each other, the current logic returns TRUE if the
G_OBJECT_TYPE of one of them is smaller than the other. This function is
checking for equality, so returning FALSE is enough.
The numbers associated to each port mode given by the AT^GETPORTMODE
response are not USB interface numbers, they are 'port numbers'.
Moreover, these numbers may start either at 0 or at 1, depending on
the firmware.
The only reasonable way to parse this response is to just gather the
order of all the port modes reported, and apply the modes to each
serial port found in the system in the same order.
Fixes https://gitlab.freedesktop.org/mobile-broadband/ModemManager/-/issues/239
We will use one single method to apply port type hints, not a mix of
them:
* If AT^GETPORTMODE is supported, prefer its hints over any other
method.
* Otherwise, try to guess hints from USB interface descriptions.
* And if none of the plugin-specific hints are supported, we'll
default to applying generic port type hints from udev tags.
Once the hints have been applied by one of the methods above, the
fallback hint sequences are run:
* Flag the first cdc-wdm port as primary if no other port has been
flagged as primary.
* Flag the USB interface 0 as PPP if no other port type hint has
been set in any other port.
The logic applying all these procedures has been refactored so that we
have separate functions for each, which is much easier to read and
follow, even if it requires multiple iterations over the port probe
list.
Fixes https://gitlab.freedesktop.org/mobile-broadband/ModemManager/-/issues/238
The mm_base_modem_get_port_*() returns a full reference, we should use
the mm_base_modem_peek_port_*() methods instead. Also, refactor a bit
the logic because both ports are really configured in the same way, so
just apply the same setup to both.
The class object definition should always be last in the file,
following the interface definitions. The actual method implementations
should be given before any other type system method (i.e. before even
the _new() method).
There is no need to reload it on every settings update attempt; just
load it once when the 3GPP interface is initialized, and re-use the
loaded value on every new update attempt.
Attempting to change the initial EPS bearer settings while in full
functionality mode shouldn't happen, so make sure we don't attempt to
do that if going into low power mode fails.
Multiple changes that shouldn't affect behavior:
* Avoid reusing the same context and state machine for the set and
the load operations, because they truly have different behaviors.
* Setup the common load operation in a separate async method, and
reuse the common operation for both the runtime state loading and
the settings configuration loading.
* Avoid having a "generic step ready" method, and instead provide
proper ready methods for each step, so that we can give
comprehensive warning logs when things fail.
* Use the common CFUN? response parser instead of a custom
implementation.
So that the list of ports shown in the Ports DBus property is also
alphabetically sorted by port name, instead of having a mess like
this:
-----------------------------
System | device: qcom-soc
| drivers: bam-dmux
| plugin: qcom-soc
| primary port: rpmsg0
| ports: rmnet5 (net), rmnet_usb0 (unknown), rmnet4 (net),
| rpmsg1 (at), rmnet3 (net), rpmsg0 (qmi), rmnet2 (net), rmnet1 (net),
| rmnet7 (net), rmnet0 (net), rmnet6 (net)
At the moment, ignored ports show up as (unknown) in the ports list
in mmcli. This makes it look like something went wrong while probing.
Actually ModemManager already tracks unknown and ignored ports separately
(MM_PORT_TYPE_UNKNOWN vs MM_PORT_TYPE_IGNORED) but the API always exposes
them as MM_MODEM_PORT_TYPE_UNKNOWN.
Add MM_MODEM_PORT_TYPE_IGNORED and use this for ignored ports so they
show up as (ignored) instead in mmcli.
If the modem goes away (ports removed) during the initialization
phase (e.g. while QMI clients are being allocated), the MMPortQmi
object will be closed and it will lose its internal QmiDevice.
We should therefore consider the lack of QmiDevice a valid usecase in
track_qmi_device_removed() and return a GError when that happens.
#0 0x00007fb544618cc9 in raise () from /lib/libc.so.6
#1 0x00007fb54461bd68 in abort () from /lib/libc.so.6
#2 0x00007fb544e2213d in g_assertion_message () from /usr/lib/libglib-2.0.so.0
#3 0x00007fb544e221ba in g_assertion_message_expr () from /usr/lib/libglib-2.0.so.0
#4 0x00000000004be584 in track_qmi_device_removed ()
#5 0x00000000004be5e3 in allocate_next_client ()
#6 0x00000000004be7b1 in qmi_port_allocate_client_ready ()
#7 0x00007fb5453690a3 in g_task_return_now () from /usr/lib/libgio-2.0.so.0
#8 0x00007fb54536967e in g_task_return () from /usr/lib/libgio-2.0.so.0
#9 0x00000000004dd8f8 in allocate_client_ready ()
#10 0x00007fb5453690a3 in g_task_return_now () from /usr/lib/libgio-2.0.so.0
#11 0x00007fb54536967e in g_task_return () from /usr/lib/libgio-2.0.so.0
#12 0x00007fb54591d4de in allocate_cid_ready () from /usr/lib/libqmi-glib.so.5
...
SIMs can be created with an EID fetched during load_sim_slots
while initializing the modem, if present.
Since load_eid would be implemented with the same mechanism
we avoid using it here (if Get Slot Status fails once, it
probably doesn't make a lot of sense to try it again).
All BCD-encoded strings used by MM currently have the low nybble
of each byte come before the high nybble, but some strings (such
as the EID string returned by QMI Get Slot Status) are meant to
be read in order with the high nybble before the low one. As such,
extend mm_bcd_to_string to decode both.
