Additionally the two follow up commits are reverted too:
Revert "lpa: Remove dependencies to NRSW"
Revert "lpa: Use latest version with patched toby driver"
This reverts commit
8f2fca9c5f,
c1187b5553 and
c18ced1d01.
id:379399
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
For most of the overrides the order is not important since all overrides
are used.
But when using a file specific to a machine, only one override is used
and in this case the order is important: the last override matching is
used. So we must keep more specific overrides at the end.
id:366284
This override was useful in the past when we had a lot of features
defined directly using the machine name. But now it is not needed
anymore.
It is actually the source of problems when using specific files for both
hw24 and 25: Since the HW24 overrides was at the end, the files for it
were always used and the files for HW25 were ignored.
id:366284
The default APN configuration is only required for TOBY-L2. For LARA, it
is properly handled through the QMI interface.
For LARA the only supported profile is RMNET, so this is not
configurable.
tpID: #354963
LARA-L6 is using a QMI interface so we need to enable QMI and the
"generic" plugin.
Also we need to backport libqmi to have the version matching
ModemManager 1.14.
BugzID: 81947
Now all scenarios with a user module are covered:
- static setup with a user module will create a bridge with static IP
- dhcp setup with a user module will put a static IP on the UM and DHCP
on the main interface
BugzID: 81969
This is a setup required for the factory testing image: the user module
must be reachable from the CPU. And since the subnet of the fct ethernet
port and the subnet of the user module are the same, we must bridge them
together.
BugzID: 81969
This is the new VCU pro based on HW26 but which includes the user
module.
Because of this change, we cannot reuse HW21 nor HW26 machines.
BugzID: 81969
After discussion with the maintainers, the problem actually lies in the
kernel drivers where a dash is forbidden. But the da9063-hwmon driver
was using a deprecated function which was not checking the name.
BugzID: 81093
These two hardware have the ability to read out the wwan modem supply to
know when it is completely turned off.
This is abstracted to wwan-config through lmsensors rules.
BugzID: 80542
It is suspected that having this variable containing a local path is the
reason why when building from sstate, the kernel sometimes need to be
rebuild.
BugzID: 80858
Deploying directly the DEPLOY_DIR_IMAGE bypasses the shared state
mechanism implemented in the deploy class, leading to the image not
being properly deployed when building entirely from sstate.
BugzID: 80331
As explained in the comment, the start of this service may include some
power cycles of the modem and also the startup of ModemManager. All of
this may take longer than the default 90 seconds in some special cases.
This should anyway not be a problem since this service is self resilient
and can handle errors during startup.
BugzID: 80178
As for the kernel, we don't need to build them for every MACHINE and
will save some build time when building them only once per architecture.
BugzID: 80015
As explained in the comments, in yocto the kernel is expected to be
built for each MACHINE and not for each architecture as we do.
Therefor all tools related to kernel must also be adapted.
BugzID: 80015
These dependencies lead to a partial rebuild of the kernel when
switching machine, even where they are actually using the exact same
kernel sources and config.
BugzID: 80015
Instead of having one kernel recipe with two hashes, it makes more sense
and is easier to maintain two recipes with the same name but different
versions. The shared part is in an include file.
Each machine then chooses which version it wants to use with
PREFERRED_VERSION variable.
BugzID: 79987
It happens that a soft reset is failing and blocking the chip in a state
where it always times out. The only wait to get out of this is to hard
reset the bluetooth chip.
BugzID: 77171
Detecting if the mac address must be inverted means doing two resets of
the chip in a short period when the mac must be inverted.
These two resets is sometimes letting the BT chip in a state where it
does not answer for more than a minute.
Also fixes getopts string: options -w and -b don't have arguments.
BugzID: 77171
settin mac address of bt and wlan is now triggered in two separated
services.
BugzID: 77171
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
added several enhancements to the mac-address-set script:
- splittet wlan and bt address setting (use -w / -b options)
- fixed upper and lower case problem with vcu1 addresses
- fixed hexadecimal support with vcu2 addresses
- added testify support (incl. log_test function)
BugzID: 77171
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
adding a unit test script which uses testify framework to test the
mac-address-set script for vuc1 and vcu2.
BugzID: 77171
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
if a device is never getting ready on startup sequence, the wait_on()
function runs forever. with this patch it aborts after 10 seconds.
BugzID: 77171
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
If NetworkManager starts before, it may change the mac address for
scanning and store the chip mac address. When finaly using a static mac
address (for an established connection or AP) it will restore the chip
address instead of our address.
BugzID: 77171
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
After an overall refactoring this script now does:
* Wait on devices being ready before changing the address
* Handle errors when setting wlan or bluetooth mac address
* Handle different MAC address schemes
* Handle different behaviors when setting wlan mac:
On HW21, the address must be set while the interface is down, but on
HW23 it must be set while the interface is up and if it is down, no
error is thrown, so the modified address is checked.
* Handle different behaviors while setting BT mac:
On HW21, the address must be given with the bytes in reverse order.
We therefore first try with the right order and check the address and
retry with the reversed address if it failed.
BugzID: 77171
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
hw23: Force wpa-supplicant and hostapd versions
meta-netmodule-wlan provides modified versions of these tools, but they
seem to not work properly on HW23. Since this HW is anyway not supported
in meta-netmodule-wlan, we force the usage of latest stable version.
BugzID: 77173
Co-authored-by: Patrick Zysset <patrick.zysset@netmodule.com>
Reviewed-on: https://git.netmodule.intranet/yoctoproject/meta-netmodule-bsp/pulls/111
Co-Authored-By: Alexandre Bard <alexandre.bard@netmodule.com>
Co-Committed-By: Alexandre Bard <alexandre.bard@netmodule.com>
those files reside now in the repository itself, thus the recipe
needed to be adapted
BugzID: 75350
Signed-off-by: Marc Mattmueller <marc.mattmueller@netmodule.com>
With this change we are building the same kernel config for all
hardwares belonging to the same platform. The configs only differ from
one or two drivers which are enabled or not and they are anyway not
loaded as long as the devicetree is not matching.
But having only one config for several hardwares is reducing maintenance
(like enabling or disabling kernel security features) and also reducing
build time when building for different machines.
BugzID: 75223
The two "ifs" are covering cases where the fitImage is not used and we
don't support this case.
And as the Module.symvers comment says, this has been fixed upstream in
kernel.bbclass.
BugzID: 75220
Our u-boot-nm.inc was just a copy of an outdated u-boot.inc.
I now changed the u-boot-nm.inc file to be a common configuration file
for our u-boot recipes.
BugzID: 75187
Signed-off-by: Lucien Mueller <lucien.mueller@netmodule.com>
ATM the system state framework only works properly if we have the da9063
with working ignition.
BugzID: 74987
Signed-off-by: Lucien Mueller <lucien.mueller@netmodule.com>
These values are out of date with the real kernel versions and may be
misleading. Since we have only one version of the kernel, we can remove
them and "git" will be used as default value.
BugzID: 74959
These symlinks were required for compatibility with NRSW u-boot.
Since we are now using fitimage, the name of the devicetree is not
relevant since it is directly part of the fitimage.
BugzID: 74958
Instead of rewriting the full fitimage_assemble, we can just append to
the part generating the kernel entry, therefore reducing the duplicated
code.
BugzID: 74530
With this approach, we just need to change a machine.conf file to add
advanced-gnss to machine_feature in machine.conf to enable gnss-mgr,
force gpsd to start with 115200 bitrate and to disable gpsd service
because it will be started by gnss-mgr.
BugzID: 74458
And also reduce the number of created artifacts and simplify do_deploy.
The rm commands are not required. bitbake is smart enough to remove the
old deployed artifacts of a recipe.
BugzID: 74357
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
This will bring pciutils automatically to all machines:
- all armada machines
- all am33x machines
- imx8 hw23
BugzID: 73589
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
and remove the machine feature on hw26 since it is derived from hw21
machine config.
BugzID: 73593
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
In order to pull in some chip specific firmware / drivers or tool there
are also one specific wifi feature per wifi chip:
- ath-wifi for atheros chips
- ti-wifi for TI chips
- ublox-wifi for u-blox chips
BugzID: 75593
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
Because we don't have the kernel class in the netmodule-fitimage recipe,
we have to add this line there.
Signed-off-by: Lucien Mueller <lucien.mueller@netmodule.com>
This recipe lets us create a fitimage outside of the kernel recipe.
For this to work we need to create just the zImage from the standard
kernel recipe.
BugzID: 73616
Signed-off-by: Lucien Mueller <lucien.mueller@netmodule.com>
Reason for reverting to originial state:
- adaption of the v2x-fw-load script
- re-introduction of the v2x recipe in all of our images
- no difference when wainting on USB disconnect event or directly
releaseing the gnss module from its reset
BTW: The original timeout of 50s can be explained as the SCU held a
delay of 40s until powering the v2x module. This delay shall be
removed in the next couple of days.
BugzID: 72787
Signed-off-by: Marc Mattmueller <marc.mattmueller@netmodule.com>
we power cylce now the power of the v2x module hence it is no longer
needed.
BugzID: 72787
Signed-off-by: Marc Mattmueller <marc.mattmueller@netmodule.com>
the v2x-fw-load script was refactored and ready now a config file. The v2x
recipe installs a v2x config file depending on a distro variable
V2X_ENABLE_FW_LOAD. Inthere it is possible to declare a parameter line like
'is-placeholder-firmware=true' to skip the firmware load in the v2x-fw-load
script. The skip includes also the detection of the DFU mode of the module.
These changes enables us to introduce the v2x package in all our images so
that the service doesn't fail anymore when we are using our placeholder
firmware.
BugzID: 72787
Signed-off-by: Marc Mattmueller <marc.mattmueller@netmodule.com>
start-up changes:
- usb-hub is just released from reset
- gnss-init service calls a script twice:
- as pre-step: checking the hub state depending on the start
reason:
- at reboot --> just go on, at power-up
- at power-up --> wait for a USB disconnect
of a device for a certain
time
- as start step: releasing gnss module from reset
shutdown changes:
- no resets were triggered
BugzID: 72787
Signed-off-by: Marc Mattmueller <marc.mattmueller@netmodule.com>
The bootloader might already release the usb-hub, the v2x module and
the gnss modem from reset. Thus the usb-hub and the gnss modem are put
into reset first so that a proper enumeration would be possible.
BugzID: 72787
Signed-off-by: Marc Mattmueller <marc.mattmueller@netmodule.com>
The v2x firmware load script contained also the usb hub reset which
affects the GNSS modem. This means that the GNSS modem worked only
when the v2x recipe was enabled. But without firmware the v2x-service
fails at start-up which affects all other images except vcu.
Therefore the usb-hub reset part was extracted from the v2x firmware
load script so that we decouple those two functions. ZF/OM need a
failing v2x service when no firmware is loaded, thus the v2x service
was moved back to the vcu image (see meta-netmodule-om).
The systemd services v2x and gnss-init depend now on usb-hub-reset
whereas gnss-init additionally depends on v2x (if available).
BugzID: 72787
Signed-off-by: Marc Mattmueller <marc.mattmueller@netmodule.com>
The descriptor.py file was not properly packaged because the variable
PYTHON_SITEPACKAGES_DIR was not defined. It is defined in the
python3-dir class.
BugzID: 72580
This updated firmware is allowing the kernel core to access the I2C bus
of the SCU core.
It also properly forwards the interrupts from the SCU core to linux.
BugzID: 70694
There is no guarantee that a fix delay is actually enough, because with
the BT_EN gpio, we only control activation of bluetooth on the chip. But
we don't know the status of the whole chip.
Also hciattach has anyway a timeout of 5 seconds, so 0.1 seconds does
probably not make a big difference in this regard.
So in order to make it the safest possible, we can only rely a failure
handling and retrying. The number 5 for retries has been defined since
repeated restart of the service with only 3 retries leads to failure
pretty quickly. At 4 there is no failure anymore. So with 5 we are sure
the risk of failure is really low.
BugzID: 70195
This package contains gpioset, gpiofind, etc
This tools are required to access the gpios from shell scripts.
Also fixes the RDEPENDS ${PN}
BugzID: 65420
Name the BT_EN gpio in the device tree when you want to use this service.
It has to be found with the command: "gpiofind BT_EN".
Signed-off-by: Lucien Mueller <lucien.mueller@netmodule.com>
The default behavior is to try to build u-boot-initial-env
Which is not implemented in our u-boot repo.
Part of yocto update to dunfell release.
BugzID: 69636
Afer upgrading meta-freescale to dunfell, the new default tune is
a35-crypto instead of aarch64.
Part of yocto update to dunfell release.
BugzID: 69636
system-state-framework: added recipes for the broker, extmod and
sysstate parts.
move include files to ${STAGING_KERNEL_BUILDDIR}/include/broker/
Signed-off-by: Lucien Mueller <lucien.mueller@netmodule.com>
Co-authored-by: Ramon Moesching <ramon.moesching@netmodule.com>
Reviewed-on: https://git.netmodule.intranet/yoctoproject/meta-netmodule-bsp/pulls/70
The name to use is actually the name defined in the layer.conf of the
target layer and not the name of the directory itself (which can
actually change).
BugzID: 69469
It was already the case for HW18: eth2 was the default.
Now with HW25, we can't rely either on eth0 name since it is assigned by
default to the wrong port.
We are therefor using lan0 for HW25 and make the nm-conf recipe more
generic.
BugzID: 69468
nmhw21 and hw26 are using common u-boot code base.
Thus we are creating symlink from nmhw21 to hw26.
With this fix jenkins packaging stage is working again.
BugzID: 69325
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
This patch is removing wlan config to use the backports instead. But the
backports are not compatible with other kernel than virtual/kernel.
We must therefor rely in the default wlan for the moment
BugzID: 69147
Changing KERNEL_PACKAGE_NAME allows this variant of the kernel to live
its own life in tmp/work/ instead of being shared in tmp/work-shared.
It is also not providing virtual/kernel.
This way, the "official" kernel (virtual/kernel) is never modified and
rebuilt, leading to strange behavior regarding deployed licenses.
BugzID: 69147
Since imx-boot is different than a classic u-boot we need to handle it
separatly. bootloader-config does not support it so we use this variable
to remove it at build time.
BugzID: 69098
In order to make the service start faster we make it less nice.
It will therefor have more cpu resources to setup the modem faster. The
overall goal being to have WWAN connectivity earlier during boot
process.
BugzID: 67742
Due typo and wrong usage bbappends were never applied to
meta-ublox-module layer.
BugzID 68069
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
After changing the distro config to allow reproducible builds, a side
effect was that the date printed during uboot startup was 1 Jan 1970.
In order to get the real build date in uboot, we must unset the
SOURCE_DATE_EPOCH variable.
BugzID: 67738
The wifi and bluetooth chip ti wl18xx firmware files are handled by seperate
meta-netmodule-wlan layer.
BugzID: 67576
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
The DLM board does not have V2X module. Therefor we must remove related
software components. These components will there be enabled by looking
at v2x in MACHINE_FEATURES.
BugzID: 66088
Instead of checking if we are building for an IMX target to include or
not the jody bbappends, we are now including them only when the ublox
meta-layer is present, which offer more scalability for futur HW design.
BugzID: 65947
DLM board requires a different SCU firmware than HW23.
This firmware includes changes in the branch "dlm" from imx8-scfw repo:
https://git.netmodule.intranet/hancock/imx8-scfw/src/branch/dlm
(latest revision : 0356a979932ae24ef1b9d6f0c3b33c82abd42948)
BugzID: 65899
gnss-mgr is only present on HW21 and therefor gpsd will not be started
on other HWs if the service is disabled.
Also the bitrate is configured by gnss-mgr so we must keep gpsd with
default bitrate detection mode.
The tool was previously just printing log messages. Using this log
library, the timestamps in journald will be more precise and it will be
possible to filter by log level.
BugzID: 65586
The default config file has been emptied to not set dummy values
A example file has been added so the user knows which fields can be set.
BugzID: 65564
Our used ubxlib was forked and diverged too much from the original
one. Now there is a naming conflict when installing our image in
the car drive tests.
Therefore our ubxlib is renamed to nmubxlib
BugzID: 65478
Signed-off-by: Marc Mattmueller <marc.mattmueller@netmodule.com>
It can happen that hciattach fails because of timeout.
Reexecuting the service in this case makes sure that the bluetooth
device is handled by bluetoothctl.
BugzID: 65391
It may happen that the service is failing, because the modem is in an
unpredictable state. To avoid this kind of cases and increase the
resilience of the system, systemd will now restart the service if it
fails, leading to a full power cycle of the modem.
BugzID: 63740
scons was called manually in do_install step with different parameters
than in do_compile step, leading to a recompilation in do_install (but
without -j4), leading it to take very long.
BugzID: 62046
Get rid of issue: Fix overlapping ranges for Switzerland and
Liechtenstein
Fixed with upstream commit: 5dd4dbf709d38aab21cba0183cf26510054449f0
BugzID: 61219
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
Making it clear what is related to netmodule version and what can be
appended to any gpsd version.
Additional small changes:
1. Removed udev rules createing /dev/gps0
Not needed since we have proper udev rules in their own repo now.
2. Removed patch 0001-gps_shm_close-Free-privdata.patch
It was not used referenced anywhere, probably a remans of an old
version.
3. Use gnss0 as default TTY device on all targets.
It avoids rebuilding everything when changing MACHINE
4. Update checksums for mainline gpsd archive.
This recipe alone is still not building but at least the checksums are
now correct.
BugzID: 62684
These definitions were here since the beginning of the project and
nobody really knows what they are used for.
Now it is the source of a build error because it is overwritting
definitions from upstream yocto.
Removing is fixing error and seems to fix other warnings related to
licenses too.
BugzID: 62473
neom8tool is part of gps-utils, the gnss0.ubx config file
is now part of the newly created gps-utils-congf package.
BugzID: 62085
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
Issue: after a reboot with no power cycle the gnss reveiver
was in the wrong state and did not send any NMEA data for gpsd to use.
Solution: Added a controlled GNSS start on boot to make sure that
the receiver is in the correct state.
BugzID: 62382
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
This tool is parsing first sector of MMC to find bootloaders, and then
reports the data of these bootloaders to /run/bootloader.
It also creates the file /dev/bootloaderX for each bootloader, allowing
to do a safe and redundant update.
BugzID: 62317
Adoption of recent changes in udev rules and gpsd build recipe
to gnss-config script.
BugzID: 62295
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
The pythonpath is now fixed in 3.20/netmodule branch on netmodule
gpsd gitlab repository. therfore the patch is no longer needed in the recipe.
BugzID: 61939
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
This tool is not used anymore and has been replaced by the more
generic nmhw-fwupdate tool.
BugzID: 62102
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
gnss-save-on-shutdown is a service, makes gnss receiver state persistent
before shut down and check if state got loaded again upon boot
on shutdown:
contents of receiver ram gets dumped to receiver flash
on boot:
checks if ram dump got loaded from receiver flash back into receiver ram
BugzID: 60669
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
The gnss0.config file still used the old name for gnss-config.service.
Changed it to the current name.
BugzID: 61707
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
wwan-config is a rework of ublox-gsm-config.
It is more generic and allows better extension and customization.
BugzID: 61873
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
we are only using hardware specific rules from now. Generic rules don't make sense.
Some rules can be common to different hardwares but none can
be completely generic.
BugzID: 61729
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
Add packageconfig to keep only the ublox tool.
Missing files if this config is applied are:
- firmware itself (UBX_M8_301_ADR_421_NEO_M8L.bin)
- map of the flash (ubx-fw-flash.xml)
- wrapper to flash the modem easily (gnss-fw-update)
BugzID: 60600
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
Vendor, model, firmware, ubx-protocol version, supported satelite
systems and supported augmentation services ar now dumped into a config
file at /run/gnss
Renamed gnss.config to gnss0.config to be in line with wwan rework.
Add grep as depency because -P option is required and is not available
in busybox grep.
BugzID: 61707
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
Added patch that reverts patch from gpsd mainline,
because it messes with the buildflags.
BugzID: 60987
Signed-off-by: Lucien Mueller <lucien.mueller@netmodule.com>
Changes to baud rate config script:
+ more appropriate name
+ only one service
+ baud rate detector is faster in special case
+ shellcheck conform
- no longer creates default config file if file is missing, instead
throws error
BugzID: 60698
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
The SERIAL_CONSOLES defaults to ttyS0 and
therefore a getty on ttyS0 is opened by systemd.
set USE_VT to 0 because we don't have virtual ttys.
BugzID: 61141
Signed-off-by: Lucien Mueller <lucien.mueller@netmodule.com>
M4 CPU is not used, leave all the pads and peripherals to the main CPU.
Reference: f010eb32247e0b06cfd63294ad8099f1889e3430
Remote: git.netmodule.intranet:hancock/imx8-scfw.git
BugzID: 61160
added patch file to ubxtool. Since ubxtool is now patched, assigning
the path of python3 env to PYTHONPATH and then exporting it in the
receiver config script is no longer needed.
BugzID: 61081
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
Added recipes aswell as scripts to automatically configure the baud
rate of ttyS3 and the GNSS receiver (NEO-M8L) with the baud rate
specified in a new config file called /etc/gnss.conf
tty interface also configured if receiver config gets launched manually.
If config-gnss-receiver was launched manually it now executes
config-gnss-tty to configure the tty interface before starting gpsd.
BugzID: 60698
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
With this whitelist, we make sure that only toby-l2 devices are
handled by ModemManager and that it won't mess up with other serial
or net devices.
BugzID: 59772
BugzID: 59771
The wowlan feature is broken in the driver provided by cypress,
and although we don't need this feature for the moment, networkmanager
is calling some related functions when stopping which leads a kernel
crash.
BugzID: 58213
serial gnss receiver baudrate configuration with gpsctl tool is too
shaky and not reliable. gps CLI tools calls are reconfigure
baudrate back to 9600 instead of forced 115200 baud.
Configuration is then blocking whole gpsd action and no gnss positon
data are delivered.
BugzID: 60550
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
This new recipe can be configured using PACKAGECONFIG to enable
different config option like DHCP or unmanaged devices.
BugzID: 60752
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
When running the script through RPM installer, the cat process
does not stop when receiving the TERM signal. So the KILL signal
has to be used to make sure that the script is not blocking.
BugzID: 60344
ubxtool is required for gnss receiver configuration
fork gpsd recipe from openembedded:
ubxtool was introduced in version 3.18
BugzID: 60545
Signed-off-by: Tobias Jäggi <tobias.jaeggi@netmodule.com>
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
configure baudrate of serial gnss receiver with gpsctl.
Todo: generic solutions for all hardware
future-proof: support for more that one gnss module
BugzID: 60330
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
This feature had been disabled to prevent use of random mac address.
But apparently the random mac is only used for scanning and, when
actualy connecting to an AP, the real mac address is used.
Furthermore this config is generating errors in some cases when
using access point mode.
BugzID: 60405
Make wl12xx and wl18xx main package RDEPENDS on minimal package.
Because of package split content of minimal package is not out of the
box also in main package.
FILES_${PN} package defintions doesn't allow to assaign same file to
different packages. During parsing process files are assaign only first
(from left to right). Futher infos FILES in yocto manual.
BugzID: 60410
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
Get rid of rootfs postinstall scripts for removing manually unused
TI wl12xx and 18xx firmware binaries. So let's define new packages
with only necessary files and call it *-minimal.
BugzID: 60410
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
After a disconnection from network side, ModemManager can reconnect
itself if the modem requests it. This commit makes sure that a
reconnection of the modem makes the GSM connection active in NM.
BugzID: 60029
Allowing configuration of default bearer through modemamanger (0001)
Replacing the old dirty hack for default context by a better handling
specific to ublox devices. (0002 and 0003).
Handling reconnect requests from the modem (0004).
BugzID: 59580
dnsmasq being now used for dns caching we add the google dns server
as fallback if other DNS are not reachable.
In case no DNS at all is configured in NM, dnsmasq will not be
started and in this case the google dns cannot be used either.
BugzID: 59522
This new release fixes the issue of an IP address change on the
modem that is not noticed by ModemManager for u-blox modems.
BugzID: 59189
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
Avoid using machine branches which can lead into failling bitbake
parsing process when branch does not exists.
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
wlconf-bin is fetching from git repository with branch name which is
derivied from MACHINE variable.
wlconf-bin is used for ti wl18xx bt/wifi chip which use mainly used on
nmhw21 and similar boards. On imx8-nmhw23 we are using u-blox jody-w1.
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
These variables may change in updates either in poky repo or
meta-openembedded, using directly the variable instead of the
static string name make it more failsafe to future upgrades.
Having external patch in Yocto means we need to maintain it and check
that it applies without conflicts after every device tree change.
Let's move this temporary change to kernel git instead.
This reverts commit 2bfe8d31c7.
BugzID: 58086
assign ublox neo gnss always name gps0 (/dev/gps0)
typ fix in url removement
BugzID: 58211
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
initialize jody with patchram file and bring up hci0 interface
use systemd for enabling and disabling:
(requests BT_EN pin)
systemctl start jody-w1-bt-init (default after boot)
systemctl stop jody-w1-bt-init
BugzID: 58086
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
Temporary workaround until issues are found.
rts and cts blocking serial communications, without rts and cts
serial communication works fine
BugzID: 58086
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
To avoid problems when meta-freescale is not present, recipes-imx is masked
when building for other HWs.
meta-freescale us now dynamically added to bblayers when building for HW23.
BugzID: 57911
This line needs to be enable to connect the physical SIM first.
Because it is only possible to detect if a physical SIM is present
or not. The eSIM will always be shown as present, even if not
configured.
BugzID: 57829
With current src revision we are checking out kernel 4.14.78
Clearing cache and sstate won't fix issue
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
To apply: include meta-netmodule-bsp/recipes-bsp/imx8 to local bblayer
Requires: offical meta-freescale layer
linux kernel: use 4.14 kernel exclusive for imx8
BugzID: 57544
Signed-off-by: Ramon Moesching <ramon.moesching@netmodule.com>
on nmhw21 the following mac address scheme is used:
- Ethernet: xx:xx:xx:50:xx:xx
- User Module: xx:xx:xx:60:xx:xx
- Wifi: xx:xx:xx:70:xx:xx
- BT: xx:xx:xx:80:xx:xx
Note: this scheme is only valid for nmhw21.
BugzID: 54514
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
zImage is not required anymore since we are now packing everything
in fitImages.
Plus, KERNEL_IMAGETYPES should not be modified manually, it is
generated using KERNEL_IMAGETYPE and KERNEL_ALT_IMAGETYPE
Note that the file kernel-fitimage-fpga.bbclass contains
overwrite of functions present in meta/kernel-fitimage.bbclass
and may not always be consistent with it
This script sets mac address of wifi and bluetooth chip
according to serial number
Also add NetworkManager config to not use random mac address on wifi
BugzID: 54514
u-boot deployed files of am335x based platforms (nrhw16/20) should be renamed according former nmhw21 refactoring.
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
since u-boot is backward compatible and today mainly used on v2 boards,
this naming is misleading.
BugzId: 55650
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
with latest nmhw21 boards 0xffff have been introduced as a new end tag
(with or without intention). Tag 65535 is now handled the same way as
Tag 0 (aka end-tag).
BugzId: 55411
Signed-off-by: Patrick Zysset <patrick.zysset@netmodule.com>
saveenv # save the environment in eeprom (ignore message during bootup)
- boot to Linux:
boot
Do not overwrite the original partition, u-boot will search a valid FPGA there. If you want to overwrite the other partition, make sure you copy the /logic folder from the original partition to the new one!
To learn how to build an image and to find the build dependencies visit:
testify is a lightweight unit testing framework for bash
# Usage
clone this repository `git clone https://github.com/zombieleet/testify.git` here.
run the mac-address-set-test.sh file and check the console output.
# Commands
all subcommands to the assert functions requres 4 arguments, the first argument is the actual value to test for, while the second argument
is the expected value, the thrid argument is a description of the test , while the fourth argument is a short description of what the test output should be
**expect** Compares two values
`assert expect "$(Name 'Jane' 'Doe')" "John Doe" "Test for Name Function" "should fail"`
To test the output of a function you have to use command substitution
You can also test single values
`assert expect "victory" "favour" "Test for Name comparison" "This should fail"`
testing for mathematical expressions
`assert expect "$((2+2))" "4" "Test for Simple Math Operation" "It should succeed"`
**regex** Does a regular expression match. The second argument to this subcommand should be a regular expression
`assert regex "What is the difference between 6 and half a dozen" "[[:digit:]]" "Match Number Regular Expression" "It should succeed"`
**status** Test for any status code. The second argument should be the expected status code. The first argument to this subcommand should be a command name, and it should not be passed as a command substitution but it should be passed as just a string wrapped in double quotes.
The arguments to the function should also be in the double quotes. Arguments with spaced should be wrapped in single quotes
`assert status "ls ." "0" "List in current dir" "it should return 0"`
**done** This should be last subcommand to call, it does not require any argument
# LICENSE
GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.