Send each character through the state machine one at a time. Some
modems (various Quectel ones for example) can process multiple at
the same time, and while that should be easily implemented using
the new changes to send_dtmf() it's not done in any plugin yet.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Split into groups of non-pause characters and single pause characters
by themselves and return as an array of strings.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Generalize the timeout/stop logic that QMI requires into the
base call. We'll use this to implement the pause character (,)
generically.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Instead of hard-coding a 500ms DTMF tone duration, let applications
use a specific duration at call setup time.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Split out the things we actually use from MMBaseModem into
separate properties that can be independently implemented
(or just left NULL) for unit tests.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Add an interface and implementation for a port scheduler that round-
robins between ports the scheduler is attached to, serializing command
execution among one or more MMPortSerial instances.
Theory of operation:
Sources (e.g. MMPort subclasses) register themselves with the scheduler.
Each source notifies the scheduler whenever its command queue depth changes,
for example when new commands are submitted, when commands are completed,
or when commands are canceled.
The scheduler will round-robin between all sources with pending commands,
sleeping when there are no pending commands from any source.
For each source with a pending command the scheduler will emit the
'send-command' signal with that source's ID. The given source should
send the next command in its queue to the modem.
When that command is finished (either successfully or with an error/timeout)
the source must call mm_port_scheduler_notify_command_done() to notify the
scheduler that it may advance to the next source with a pending command, if
any. If the 'send-command' signal and the notify_command_done() call are not
balanced the scheduler may stall.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
If the object has an owner ID already but doesn't yet have enough
information to build its log ID, don't cache a less-than-useful
log ID string for it.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Some older MTK-based phones (Tecno T528 NEW) reply to +CPIN with
an error but are (obviously) unlocked, otherwise they wouldn't
be able to boot up to the point of talking to a computer. Since
they are actually unlocked they will happily give out the IMSI
even if +CPIN errors.
To work around the +CPIN issue treat the device as unlocked if we
can successfully read the IMSI.
Fixes: https://gitlab.freedesktop.org/mobile-broadband/ModemManager/-/issues/980
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Some Sierra devices omit the leading <CR> from call start/end URCs
like NO CARRIER and CONNECT, which caused the echo-removal code to
remove from the response buffer because the leading <CR><LF> did
not exist. That can break call control and hangup handling on
those devices.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Mainly because we need somewhere to stash the call end regex, and
it's silly to have a 3rd instance of that in MMPortSerialAt when
we already have one in the serial parsers that MMPortSerialAt
relies on pretty heavily.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Add some test responses from the Quectel EG915Q and some other Quectel devices.
Then try to parse those responses.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
MMSmsList has logic to allow multiple unstored messages as long as
they use SMS_PART_INVALID_INDEX. These messages aren't stored so
they don't have an index.
But the MBIM and QMI modems used index 0 for unstored messages,
meaning there could only ever be one and also that multipart
messages would fail to be combined.
Let's fix that by just using SMS_PART_INVALID_INDEX for every
scenario where an unstored message comes in.
Fixes: https://gitlab.freedesktop.org/mobile-broadband/ModemManager/-/issues/979
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Instead of going through the messaging interface, which just asks
the MMSmsList anyway, just go straight to the list.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Move error paths after getting a multipart reference to before getting
one, to reduce the possiblity that we get a reference from the MMSmsList
that ultimately doesn't end up getting used.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Instead of getting it from the MMIfaceModemMessaging when the
SMS is stored. This flattens dependencies between MMBaseSms and
MMIfaceModemMessaging.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
This commit moves creation of the MMBaseSms objects out of MMSmsList and up
into MMIfaceModemMessaging (which is already a MMBroadbandModem) and
the MMBroadbandModem subclasses themselves.
This flattens the creation of MMBaseSms objects by passing them down
from the object that creates the SMS parts, rather than having a
convoluted callback scheme relying on MMSmsList and MMBaseSms having
direct knowledge of their owning modem.
The goal is to eventually remove usage of MMBaseModem from MMBaseSms
and MMSmsList so that we can test them more easily.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Simplify MMBaseSms (making it easier to use from testcases) by
splitting the AT-specific code into MMSmsAt rather than keeping
it in the base class.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
CBM isn't as complicated as SMS, and the CBM class itself doesn't
need to be overridden by QMI/MBIM/etc. Now that the log parent
and connection binding no longer require MMBaseModem we can flatten
the CBM object creation and remove usage of MMBaseModem.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Continues removing usage of MMBaseModem in a bunch of files
by splitting out bits of its usage to separate interfaces.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Rather than make all of them rely on MMBaseModem for it. This
lets us disentangle dependencies for easier unit testing.
For interfaces, rather than casting directly to MMBaseModem
use intermediate interfaces (MMIfaceAuth and MMIfaceOpLock)
that tests can fake out.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
AT docs for many (most?) Quectel devices state that QPOWD=1 receives
the OK response very quickly but the caller must wait for the
POWERED DOWN URC to avoid data loss. Do that.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
When matching profiles ignore disabled ones. When activating a
specific profile index, return an error if that profile is marked
as disabled.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Previously, a disabled profile would cause Profile Manager interface
enabling to fail because get_profile_settings_ready() returned an
error for disabled profiles.
But 3GPP profile objects already have support for the enable/disabled
setting, so just hook that up and stop returning an error for disabled
ones.
This allows successful initialization of the 3GPP Profile Manager
on devices that contain disabled profiles.
While we're at it, allow creating disabled profiles, and changing
profiles back and forth from enabled<->disabled.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
We want ports to be closed as early as possible; so make sure they
are closed in dispose() which is run explicitly by MMDevice when
cleaning up and removing the modem. Otherwise they could stay
open longer than we want, causing problems for suspend/resume.
This caused problems with suspend because cancelling the
MMBaseModem's cancellable triggered an idle handler that held
a reference to the modem, preventing finalize() from running
until after the suspend code was finished, thus preventing
ports from being closed and cleaned up. Closing them in
dispose() handles this.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
base_modem_invalid_idle() needs to keep a reference to the
MMBaseModem until its done. But if the modem is cancelled during
removal right before we're trying to forget about it, keeping
that reference can prevent everything getting cleaned up by
MMBaseModem's dispose() until after the mainloop has run again
and all our removal operations are supposedly complete.
MMDevice uses an explicit dispose() trigger that mitigates
most of this (and requires it for other reasons, like breaking
reference cycles between the modem and it's various child
classes) but it's somewhat error-prone to keep the modem
around if we don't really need to.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Every use of mm_device_remove_modem() was already trying to cancel the
modem's canellable, so let's just put it there and simplify the callers.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
This refreshes operator information after manual registration. QMI and
MBIM implementations return operator info along with registration info
so they handle this easily, while with AT-based devices we must run
additional commands to retrieve that after the manual registration has
finished.
Suggested by: Kirill Buksha <kirill.buksha@axians.rs>
Fixes: https://gitlab.freedesktop.org/mobile-broadband/ModemManager/-/issues/971
Signed-off-by: Dan Williams <dan@ioncontrol.co>
When an outgoing call is created and added to the call list its state
is UNKNOWN. The call list emits the ADDED signal which MMIfaceModemVoice
listens for to check whether to begin polling call state.
But since UNKNOWN is not an in-progress call state polling isn't started.
And nothing was rechecking whether to start polling when the call state
changed to something other than UNKNOWN.
Do that so that we can track call progress as it dials out and is
answered.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Missing ! for strncmp() meant the test was backwards and would
erroneously report 1xRTT and EVDOr0 when those access technologies
were not in use.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Dell Snapdragon X62 before FCC Unlock reports no radio interfaces
from QMI and no device capabilities from MBIM. Instead of asserting
in mm_filter_supported_modes() when called from
mm_supported_modes_from_qmi_supported_modes_context() gracefully
handle the lack of supported modes.
Signed-off-by: Dan Williams <dan@ioncontrol.co>
Dan Williams