Since our recipes handle native builds differently, port the latest version
of the recipes from meta-freescale to our layer while keeping our
customizations. Use the same revision for all platforms.
Adapt patches for lf-6.6.23-2.0.0 release and add the "Upstream-Status" tag to
them to avoid QA errors.
For now, use the imx-boot recipe from meta-freescale, but there's a chance we
might need to port the version of the recipe in meta-imx.
https://onedigi.atlassian.net/browse/DEL-9011https://onedigi.atlassian.net/browse/DEL-9081
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
The recipe for 4.2.0.imx has been upstreamed, so merge both of our .bb files
for 4.0.0.imx and 4.2.0.imx into a single .bbappend file. Adapt our patches to
support the ccimx93 and ccimx91, and while at it, add the "Upstream-Status"
tag to them to avoid QA errors
https://onedigi.atlassian.net/browse/DEL-9011https://onedigi.atlassian.net/browse/DEL-9081
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
Back in Yocto 4.0, we applied some imx recipe changes exclusively for the
ccimx93 from meta-imx's rel_imx_6.1.55_2.2.0 tag, but now these changes have
been upstreamed in the meta-freescale scarthgap branch. Since there currently
isn't an NXP BSP based on scarthgap, make all platforms use the upstreamed
versions of the recipes, which are the same versions or newer as the ones
available in meta-imx
Remove the recipes for:
* firmware-ele-imx
* gstreamer1.0-plugins-bad
* gstreamer1.0-plugins-base
* gstreamer1.0-plugins-good
* gstreamer1.0-plugins-good
* gstreamer1.0-plugins-ugly
* gstreamer1.0
* imx-atf
* imx-boot-firmware-files
* imx-codec
* imx-g2d-samples
* imx-m33-demos
* imx-m4-demos
* imx-m7-demos
* imx-parser
* imx-pxp-g2d
* libdrm
* linux-imx-headers
* optee-test
* wayland
* wayland-protocols
* weston
* xwayland
Apply these changes:
* Remove ccimx93-exclusive changes in gstreamer1.0-meta-base
* Modify imx-gst1.0-plugin bbappend so the upstream version is used in all
platforms
* Remove imx-alsa-plugins bbappend so the upstream version is used in all
platforms
* Remove firmware-nxp-wifi bbappend since the files we need to install seem
to already be included in the meta-freescale recipe
* Remove imx-test and imx-lib bbappends since our custom changes have
already been included in the meta-freescale recipes
https://onedigi.atlassian.net/browse/DEL-9011https://onedigi.atlassian.net/browse/DEL-9081
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
The SCP firmware introduces too much verbosity (with
timestamp included) in the middle of the boot log
between the TF-A and U-Boot.
Reduce the log and remove compiler flags for errors
on unused variables (banner strings).
This removes the following messages from SCP firmware:
[ 0.000000] SCP-firmware v2.12.0-dev
[ 0.000000]
[ 0.000000] [FWK] Module initialization complete!
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
ST_OPTEE_DEBUG_TRACE is set to 0 if ST_DEBUG_TRACE is set to 0.
However, on the optee source code, if ST_OPTEE_DEBUG_TRACE=0 the
log level is automatically set to 3 (INFO) resulting in a very
verbose optee log.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
DEY generates the ccmp25 boot artifacts on subdirectories of the main
deploy folder. The firmware installation script expects to have them on
the deploy directory, so create the proper symlinks.
https://onedigi.atlassian.net/browse/DEL-9120
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
Add support based on STM release openstlinux-6.1-yocto-mickledore-mpu-v24.06.26.
https://onedigi.atlassian.net/browse/DEL-8995
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
Add open-source implementation of the OpenGL API support based on v23.0.3
version from STM release openstlinux-6.1-yocto-mickledore-mpu-v24.06.26.
https://onedigi.atlassian.net/browse/DEL-8995
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
Add GPU support based on v6.4.15 version from STM release
openstlinux-6.1-yocto-mickledore-mpu-v24.06.26.
https://onedigi.atlassian.net/browse/DEL-8995
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
Add support based on v2.8 version from STM release
openstlinux-6.1-yocto-mickledore-mp2-v23.12.06.
https://onedigi.atlassian.net/browse/DEL-8995
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
Add support based on v3.19.0 version from STM release
openstlinux-6.1-yocto-mickledore-mp2-v23.12.06.
https://onedigi.atlassian.net/browse/DEL-8995
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
This commits adds the CCMX91 platform to the DEY
build system. Furthermore, it creates generic ccimx9
support to be used for the CCiMX91 and CCiMX93
platform.
https://onedigi.atlassian.net/browse/DEL-9106
Signed-off-by: Isaac Hermida <isaac.hermida@digi.com>
Signed-off-by: Mike Engel <Mike.Engel@digi.com>
When TrustFence is enabled, use the HUK programmed on the OTP
bits for the ccmp15 platform.
Signed-off-by: Hector Palacios <hector.palacios@digi.com>
https://onedigi.atlassian.net/browse/DEL-9121
By default, the secure storage path in the REE is "/var/lib/tee". It is
part of the rootfs, and thus, it gets lost on a firmware update.
This commit changes that path to a different partition "/mnt/data/tee"
when Trustfence file-based encryption is enabled.
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
Building Optee trusted applications (TA) depends on optee_client and the TA
devkit provided by optee_os. Our toolchain provides those dependencies, but
the SDK script which configures the environment for standalone building,
is not configuring some variables needed to build trusted applications.
This commit extends the SDK environment script to allow building TAs.
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
The patches have been backported from the lf-6.1.36-2.1.0 release of
imx-mkimage.
https://onedigi.atlassian.net/browse/DUB-1081
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
The patches have been backported from the lf-6.1.36-2.1.0 release of
imx-mkimage.
https://onedigi.atlassian.net/browse/DUB-1081
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
Optee-client provides the TEE Client API as defined by the GlobalPlatform TEE standard.
It is required to communicate with a Trusted Application (TA) running in a Trusted OS.
https://onedigi.atlassian.net/browse/DEL-8970
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
Several things were wrong after the latest update to version 4.0: the
tee-supplicant path, some settings in the systemd unit, etc.
This commit fixes the installation so the optee test suite completes again.
https://onedigi.atlassian.net/browse/DEL-8989
Signed-off-by: Javier Viguera <javier.viguera@digi.com>
This commit fixes the set_fip_sign_key() function to match the new keys format
where there is a key_pass file for each key, no longer needing to search with
the key index.
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
This commit fixes the set_fip_sign_key() function to match the new keys format
where there is a key_pass file for each key, no longer needing to search with
the key index.
Signed-off-by: Arturo Buzarra <arturo.buzarra@digi.com>
This is an NXP change that reverts a mainline weston commit form v9.0.0, in
which the mouse cursor only gets activated when there is mouse movement. This
change was only being included in the weston v10.0.X i.MX forks.
For platforms that don't use these weston forks (ccimx93 uses the v11.0.X fork
and ccmp15 uses mainline weston), the mouse cursor doesn't load right away when
booting the system, which causes apps that are automatically launched (such as
the LVGL demo) to not register the mouse, rendering said apps unresponsive to
it.
Port NXP's change to all of the weston versions we currently use to avoid this
problem.
https://onedigi.atlassian.net/browse/DEL-8865
Signed-off-by: Gabriel Valcazar <gabriel.valcazar@digi.com>
Gabriel Valcazar
Javier Viguera
Arturo Buzarra
Francisco Gil
Hector Palacios
Mike Engel
Isaac Hermida