The PLS63 is a drop-in replacement for PLS62 modem, however despite that
fact it has a different USB SoC. As a consequence enumarates
differently on the USB bus.
The modem-manager interface of snapd contains the udev rules for adding
type hints to USB devices. One of the recent updates includes the rules
for PLS63 modem.
This commit makes sure that the Modem Manager is using these type hints
so that it chosses the correct port for data.
Author: Alfonso Sanchez-Beato <email address hidden>
bearer: workaround serial not responding in PLS62-W
Do not try to deactivate the PDP context for the Gemalto PLS62-W modem,
let NM do the job for us. The reason is that due to a modem firmware
bug, if the timing between pppd and us deactivating is unfortunate, the
port ends up in a state in which it is unusable. Note also that this LTE
modem cannot really disconnect while attached, so the AT command always
returns an error anyway. Fixes LP: #1829293.
cinterion: pls62: workaround serial not responding
Do not try to deactivate the PDP context for the Gemalto PLS62-W modem,
let NM do the job for us. The reason is that due to a modem firmware
bug, if the timing between pppd and us deactivating is unfortunate, the
port ends up in a state in which it is unusable. Note also that this LTE
modem cannot really disconnect while attached, so the AT command always
returns an error anyway.
To perform this, the disconnect 3gpp functions have been implemented
in the PLS62 bearer. Fixes LP: #1829293.
02ab1ea...
by
System Enablement CI Bot <email address hidden>
Author: Alfonso Sanchez-Beato <email address hidden>
Ensure we do not get duplicated traces on journal
snapd redirectes all process output to the journal, while MM writes its
ouput to both syslog and to stderr. This provokes duplicated traces that
makes debugging harder than needed. We remove the openlog LOG_PERROR so
only the traces written directly by MM appear. Fixes LP: #1833524.
snapd redirectes all process output to the journal, while MM writes its
ouput to both syslog and to stderr. This provokes duplicated traces that
makes debugging harder than needed. We remove the openlog LOG_PERROR so
only the traces written directly by MM appear. Fixes LP: #1833524.
76d0080...
by
System Enablement CI Bot <email address hidden>
Author: Alfonso Sanchez-Beato <email address hidden>
port-serial: reopen instead of force close on port hangup
Do not force close the port when receiving a hangup from the serial port. It is
possible to receive it in some cases, like when pppd has not properly restored
the port settings and we are not in CLOCAL mode. If this happens, do a port
re-open instead to make sure we have a port with good settings.
port-serial: reopen instead of force close on port hangup
Do not force close the port when receiving a hangup from the serial port. It is
possible to receive it in some cases, like when pppd has not properly restored
the port settings and we are not in CLOCAL mode. If this happens, do a port
re-open instead to make sure we have a port with good settings.
As this modem does not support SWWAN and uses PPP instead of an eth-type
connection, the call to SGAUTH is not supported either. The user and
password need to be used by pppd instead.
Ignore PDP type when searching for a suitable APN in the contexts
created by the modem. These contexts modify the PDP type to reflect
the PDP types really supported by the network, which might be more
restricted than the one the user asks for.