Protect the invocation of the _scan_done() callback function with the
global lock to avoid that it is called while the module is being unloaded
and the data structures have been freed.
Additionally, the commit also adds a patch to reduce the log level of the
driver that is logging some annoying messages.
https://jira.digi.com/browse/DEL-3607https://jira.digi.com/browse/DEL-3393
Signed-off-by: Pedro Perez de Heredia <pedro.perez@digi.com>
This parameter allows to set the time interval for scheduled scans in
seconds. This is usually managed by the wpa-supplicant dynamically, but
for example with DFS channels and non broadcasting SSIDs re-associations
take too long if the periodic scans time interval is not reduced.
https://jira.digi.com/browse/CC6UL-594
Signed-off-by: Alex Gonzalez <alex.gonzalez@digi.com>
This is the manufacturing test harness used for the CC6UL module during
manufacturing RF test.
Signed-off-by: Sebastian Pastor <sebastian.pastor@digi.com>
Refresh the patches with GIT so they apply cleanly using "git am".
Otherwise they fail with:
Applying: openssl_helper: use /dev/urandom as seed source
error: corrupt patch at line 16
Patch failed at 0003 openssl_helper: use /dev/urandom as seed source
Applying: hab4_pki_tree.sh: usa a random password for the default PKI generation
warning: keys/hab4_pki_tree.sh has type 100755, expected 100644
Notice that they were not failing in Yocto, as it does not use "git am"
to apply patches, but it's better to have the patches correctly done.
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
* Device Cloud has been renamed to Remote Manager.
* Removed "Data points" known issue (fixed in CloudConnector)
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
Ship libgles3-mesa-dev package along with libgles2-mesa-dev because
otherwise GLES3 headers are missing from the toolchain and we have build
failures for example compiling QT applications.
Exactly the same workaround is done for the vivante implementation in
meta-fsl-arm layer.
Original patch (not merged into oe-core):
https://patchwork.openembedded.org/patch/134244/
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
https://jira.digi.com/browse/DEL-3088
According to NXP i.MX Linux Reference Manual:
The full CAAM function is exclusive with the Mega/Fast mix off feature
in DSM. If CAAM is enabled, the Mega/Fast mix off feature needs to be
disabled, and the user should
"echo enabled > /sys/bus/platform/devices/2100000.aips-bus/2100000.caam/2101000.jr0/power/wakeup"
after the kernel boots up, and then Mega/Fast mix will keep the power
on in DSM.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DEL-3636
It's not enabled by default. It's controlled by DEY_IMAGE_INSTALLER
variable in the project's local.conf.
https://jira.digi.com/browse/DEL-3591
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
On a CC6, the 'env default -a' command deletes the 'soc_family' variable
that is used on the U-Boot boot scripts to configure the name of the
device tree file to load.
Then if we try to boot into recovery directly (without a reset) because
the variable is not available, it fails to compose the correct name for
the fdt, and gives the error:
reading uImage--ccimx6sbc-id129.dtb
** Unable to read file uImage--ccimx6sbc-id129.dtb **
Force a reset, so the 'soc_family' variable is available again and the
correct fdt_file is loaded.
While on it, add some blank lines to the 'recovery reboot' message, so
it's easier to read.
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
The caam encryption check was only done from fw_env_write and fw_env_read
functions, which are not called when using the functions exported as a library.
Move the check_caam_encryption() call to fw_env_open(), which is called from
all code paths. A similar check for AES encryption cannot be moved because it
requires the AES key as an argument.
https://jira.digi.com/browse/DEL-3616
Signed-off-by: Diaz de Grenu, Jose <Jose.DiazdeGrenu@digi.com>
Add critical section in suspend script to avoid that more than
one instance can be executed concurrently. This happens, for
instance, when the power button key is pressed while the system
is resumming, causing issues on the wifi interface.
https://jira.digi.com/browse/DEL-3431
Signed-off-by: Pedro Perez de Heredia <pedro.perez@digi.com>
The profiles created for the sgtl5000 sound card didn't have any priorities
assigned, making pulseaudio map a different profile than before and causing the
sink to remain unloaded.
This commit fixes this by assigning a priority to each profile. The values
were taken directly from the default profile to assure that pulseaudio's
behavior is the same using one profile or the other.
https://jira.digi.com/browse/DEL-3602
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
NAND partitions may have pre-existing UBI volume information
and that will tell U-Boot to use 'ubiwrite' to keep UBI wear
leveling information instead of erasing the NAND partition
beforehand. Given that the UBI support in U-Boot is not
particularly complete and stable, and specially if the NAND
partition table has been changed, this could result into
problems during the UBI volume initialization or during the
update process. To prevent such problems it is better that
the NAND partitions are erased.
This loses the UBI wear leveling information but provides
a more reliable deployment install script.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DUB-718
We have seen corner cases (e.g. when the partition to format has already
some corrupted UBI metadata) where ubiformat stops and ask the user for
confirmation to do the format. We cannot have such interactivity in the
recovery ramdisk so use the '-y' argument to assume yes for all
questions.
Also use the '-q' (quiet) parameter instead of the '/dev/null'
redirections to remove command's output.
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
Modify patch to clean compiler warning:
warning: passing argument 1 of 'sysfs_mmcboot_set_protection' discards
'const' qualifier from pointer target type
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
Remove the commit that removes the p2p0 interface at boot. This interface
is the one used to establish p2p connections. Configuring the wlan0 command
for p2p does not work.
https://jira.digi.com/browse/DEL-3390
Signed-off-by: Alex Gonzalez <alex.gonzalez@digi.com>
The qualcomm wlan driver creates two interfaces at load time, wlan0 and
p2p0. Both of them should load the module if not already loaded.
Signed-off-by: Alex Gonzalez <alex.gonzalez@digi.com>
This commit adds a wpa supplicant P2P configuration file and configures
the p2p0 interface automatically at pre-up.
Signed-off-by: Alex Gonzalez <alex.gonzalez@digi.com>
Pedro Perez de Heredia
Alex Gonzalez
Sebastian Pastor
Javier Viguera
Hector Palacios
David Escalona
Alejandro Navarro
Diaz de Grenu, Jose
Gabriel Valcazar