This package was being included as a RRECOMMENDS from libcrypto, but we don't
need it in the initramfs at all. Since our recovery image is already pretty big
compared with the recovery partition size on ccimx6ul platforms with 256 MB of
storage, remove it to make sure the image can fit.
https://jira.digi.com/browse/DEL-7253
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
Add systemd auto-getty.service to run agetty on the console defined in the
kernel command line on platforms with systemd support.
This service replaces serial-getty@.service as the one responsible for
opening a tty port and prompting for a login name when
TRUSTFENCE_CONSOLE_DISABLE is configured.
https://jira.digi.com/browse/DEL-7242
Signed-off-by: Gonzalo Ruiz <Gonzalo.Ruiz@digi.com>
We have specific scripts per platform to export the qt5 environment variables,
because using the default backend-specific scripts ported from the meta-imx
layer results in mismatches in specific cases (for example, 6UL framebuffer
images only support linuxfb, but the qt5-fb.sh script exports eglfs instead).
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
Wrapping unterminated lines doesn't work for scripts (even
if they contain the backslash).
Substitute with full non-wrapped if/elif sentences.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DEL-7252
In the ccimx6sbc, since each SOC has a separate set of device trees, we can use
the soc_family environment variable when generating the name of the final
device tree in the bootscript. However, for the ccimx6qpsbc, there is only one
set of device trees with the "imx6qp" prefix, even though there are SOC
variants such as the imx6dp. Running the bootscript on a ccimx6qpsbc with an
imx6dp SOC assembled will result in a failure, since it will look for a
non-existing device tree starting with a "imx6dp" prefix.
Stop using the soc_family variable and hardcode "imx6qp" instead.
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
We originally ignored the weston-init.bbappend in meta-freescale back in thud
because we were basing our recipe off of NXP's sumo BSP, and NXP's weston-init
customizations were incompatible with the ones in meta-freescale. Remove it
from the BBMASK to be able to incorporate the customizations in meta-freescale.
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
With the latest version of imx-mkimage, the original dtb file is being
removed at the end of the build.
Add a patch to preserve it and be able to run several builds during the
same compile run, like we do for imx-boot signing.
Signed-off-by: Gonzalo Ruiz <Gonzalo.Ruiz@digi.com>
The ccimx6 uses the g2d weston renderer to avoid graphical glitches, but the
default renderer code doesn't handle suspend/resume cycles properly, causing
other issues (notably in the framerate and the desktop not showing 66% of the
time when resuming from suspend). Add a patch to mitigate these issues.
https://jira.digi.com/browse/DEL-7236
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
Without this option, moving any element on the desktop (cursor or windows) will
cause an undesired "trailing" effect, where parts of the moved elements remain
on the desktop.
https://jira.digi.com/browse/DEL-7221
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
With this change, the ccimx6sbc and ccimx6qpsbc default images will now use the
xwayland backend instead of x11.
https://jira.digi.com/browse/DEL-7221
Signed-off-by: Mike Engel <Mike.Engel@digi.com>
This workaround is no longer needed since ccimx6/6qp now use Linux v5.4
https://jira.digi.com/browse/DEL-7221
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
Remove all v4.9 recipes and build Linux using the same branch on all platforms.
The .inc files were only needed because of the existence of multiple Linux
versions, but that's not the case anymore, so remove them.
https://jira.digi.com/browse/DEL-7221
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
The CST package requires byacc to compile, and even though this dependency is
met when building images for the target, said dependency needs to be made
explicit when the package is built for the SDK in order to avoid build errors.
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
With the previous solution we would need to generate
multiple overlays for each soc_type, so if we have a
new soc type (for example the solo), we would need
to generate 3 different overlays.
Signed-off-by: Francisco Gil Martinez <francisco.gilmartinez@digi.com>
This commit fixed the sdcard generation when encryption is enabled.
In the sdcard image always is included the signed image instead of the
encrypted.
https://jira.digi.com/browse/DEL-7200
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
Recipe from poky layer only provides the native support
but to generate the sdk we need also the nativesdk.
https://jira.digi.com/browse/DEL-7200
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
This version supports encryption for devices with Advanced High Assurance Boot
(AHAB) capabilities. This commit also updates and simplifies Digi custom
patches.
https://jira.digi.com/browse/DEL-7175
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
This package needs to be installed in the rootfs for the demos to work.
While at it, mark this change as Digi-specific to avoid removing it in future
recipe updates.
Reported-by: Alfonso Martin Rey <alfonso.martinrey@digi.com>
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
(cherry picked from commit 07f1fa88f2)
Add running 'mmc partconf 0 1 1 1' command to the uuu scripts after
writing the new U-Boot into the internal eMMC.
This is required for blank eMMCs on CC8MN and CC8X SOMs where the
'update' command has never been run to be able to boot from eMMC.
Signed-off-by: Gonzalo Ruiz <Gonzalo.Ruiz@digi.com>
This commit removes the v4l2 example because they are superseded
by gstreamer examples.
Signed-off-by: Mike Engel <Mike.Engel@digi.com>
https://jira.digi.com/browse/DEL-7211
This commit fixes an intermittent issue when we are cloning several
repos with a huge history, throwing an unexpected error.
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
The old fsl_otp driver is deprecated and we need to determine
if the device is closed using a new property of the device tree
called 'digi,tf-closed'.
Assume the device is open if the property is not found.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DEL-7185
(cherry picked from commit fc1a736758)
The encryption of the U-Boot environment uses the HWID as key
modifier. The HWID was being read using the fsl_otp driver sysfs
entries and over two words. The driver is now deprecated and also
the cc8x and cc8m platforms have different number of HWID words.
This patch modifies the function that reads the HWID words by
using new entries on the device tree 'digi,hwid_X' where X is the
index of the HWID word.
It also removes the need to select CONFIG_MD5 which is already
auto-selected on sandbox_defconfig by other config switches.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DEL-7185
(cherry picked from commit 8ced362766)
- OTP section
- Use just one fixed path to the OTP nvmem device descriptor (instead
of a loop).
- Use '-v' for dumping all values with hexdump.
- TrustFence section
- For cc8x, determine if the device is closed by checking a new boolean
property on the DT: digi,tf-open|closed
- For the rest, check the SEC_CONFIG[1] bit using the new nvmem
descriptor.
- Change log from 'Device status' to 'Security status'
- Report UNKNOWN if nvmem device does not exist.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DEL-7185
(cherry picked from commit f6a8de0067)
The old fsl_otp driver is deprecated and we need to determine
if the device is closed using a new property of the device tree
called 'digi,tf-closed'.
Assume the device is open if the property is not found.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DEL-7185
The encryption of the U-Boot environment uses the HWID as key
modifier. The HWID was being read using the fsl_otp driver sysfs
entries and over two words. The driver is now deprecated and also
the cc8x and cc8m platforms have different number of HWID words.
This patch modifies the function that reads the HWID words by
using new entries on the device tree 'digi,hwid_X' where X is the
index of the HWID word.
It also removes the need to select CONFIG_MD5 which is already
auto-selected on sandbox_defconfig by other config switches.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DEL-7185
- OTP section
- Use just one fixed path to the OTP nvmem device descriptor (instead
of a loop).
- Use '-v' for dumping all values with hexdump.
- TrustFence section
- For cc8x, determine if the device is closed by checking a new boolean
property on the DT: digi,tf-open|closed
- For the rest, check the SEC_CONFIG[1] bit using the new nvmem
descriptor.
- Change log from 'Device status' to 'Security status'
- Report UNKNOWN if nvmem device does not exist.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://jira.digi.com/browse/DEL-7185
Gabriel Valcazar
Gonzalo Ruiz
Hector Palacios
Arturo Buzarra
Mike Engel
Francisco Gil Martinez
Mattias Lindblad