--

Commit: seco-ne/kernel/linux-guf@ffcfe481

arm:dts: Moved muxing of EN_VBAT to generic SANTVEND out group

The pin muxing wasn't loaded correctly.

--

Commit: seco-ne/kernel/linux-imx-kuk@047033f8

dts:mc3: Fixed reset polarity of GPIO-Expander

The polarity of the reset is active high.

--

Commit: seco-ne/kernel/linux-guf@7d439e13



panel-dt: Add handling of "sel6_8_polarity" property in device tree

Add support of optional "sel6_8_polarity" property to device tree,
which defines the polarity of the "sel6_8" gpio pin. Because some panels
(such as Kyocera TCG070WVLQ-AN*174) need "sel6_8" signal to be inverted
for proper operation.

Signed-off-by: Dmitry Petrov <dmitry.petrov@rtsoft.de>

--

Commit: seco-ne/tools/gfxml2dto@7ee50f01



dto: panel: Add "sel6_8_polarity" property

Add support of "sel6_8_polarity" node to the panel fragment of device
tree. This property is handled by "panel-dt" driver to set the polarity
of "sel6_8" gpio pin. The naming is inherited from corresponding XML
files used by gfxml2dto.

Signed-off-by: Dmitry Petrov <dmitry.petrov@rtsoft.de>

--

Commit: seco-ne/kernel/linux-guf@8cd96a28



arm:dts:san: fix i2c1 gpio numbers

scl-gpios and sda-gpios were mixed up with each other.

Signed-off-by: Mikhail Vanyulin <mikhail.vanyulin@rtsoft.de>

The revision was wrong because of a failed integration.

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@71ff8334

uboot:dts: add pull-up for SAI1_TXD1 and tidy up other boot pin configs

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@0da7c2ef

uboot:board: propagate board and module info to environment during boot

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@2ff9b17e

uboot:board: improve boot output describing the system

This commit uses the new baseboard detection to make the descriptive
output during boot ("This is a...") more meaningful.

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@703075a3

uboot:cmd: refactor guf command

This commit removes all code which was previously used to detect the
board, as this functionality was moved to
board/keithkoep/common/baseboard*. Only code with the purpose to do the
bootmode detection remains.

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@edeaf423

uboot:board: add generic platform and baseboard revision detection

This commit implements generic platform and baseboard revision detection
for SBC and SOM systems. It currently supports PFID based baseboard
detection.

- baseboard.h: common interface to query baseboard type and revision. It
  defines the board types, detects whether PFIDs are used or some other
  system and forwards the calls to the proper implementation.
- baseboard_pfids.h: type and revision detection using PFIDs
- baseboard_util.h: utility functions for reading GPIO values

Some additional functions were made available in kuk_boards.h to
implement the changes.

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@054cf0b7

uboot:config: enable OF_LIBFDT_OVERLAY config option

This change is required to make the fdt apply u-boot command available
which is used for applying device tree overlays.

--

Commit: seco-ne/kernel/linux-imx-kuk@2d9b37d0

arm64:dts: Re-enable VPU on tanaro and trizeps8plus

In user space, the CMA buffer nodes are now handled correctly, so
video playback works now.

BCS 746-001325

--

Commit: seco-ne/kernel/linux-imx-kuk@678f7c35

ucb1400 touch: fix exception at boot and allow disabling of pressure

This patch fixes the exception/warning regarding imbalanced enable_irq()
and adds a globbal variable to allow disabling of pressure measurements.
Currently pressure measurements are still enabled, though, because in
current state the touch behaves even worse with pressure disabled.

--

Commit: seco-ne/kernel/linux-imx-kuk@cc51fd1a

dts:mc3: Added UCB1400

Added the UCB1400 codec to the SAI of the MC3.

BCS 746-001152

--

Commit: seco-ne/kernel/linux-imx-kuk@7f7e71a3

dts:mc3: Added ADV_PD and ADV_INT GPIO names

BCS 746-001077

--

Commit: seco-ne/kernel/linux-imx-kuk@a79800f1

dts:aarch64:mc3: Enable PD signal of ADV7535 by default

The ADV7535 should have high impedance outputs, when HDMI on the
backplane is not used. Add a gpio-reset for the PD pin and set it to the
initially-in-reset state. Only if the hdmi-backplane-mc3.dtsi file is
included, this property is removed.

--

Commit: seco-ne/tools/gf-emc-test-suite@08842749



bball: fix multiple definition linking error

Build fails with following errors:
ld: git/testcases/bball/tga_reader.h:34: multiple definition of
`TGA_READER_ABGR'; git/testcases/bball/tga_reader.h:34: first defined here
ld: git/testcases/bball/tga_reader.h:33: multiple definition of
`TGA_READER_ARGB'; git/testcases/bball/tga_reader.h:33: first defined here

Add extern to those definitions in header file.

Signed-off-by: Mikhail Vanyulin <mikhail.vanyulin@rtsoft.de>

--

Commit: seco-ne/kernel/linux-imx-kuk@eb328a82

fsl_sai: disable irqs during AC97 register access

If interrupts occur after enabling the SAI RMR but before enabling the SAI TMR during AC97 read register accesses, the FIFO can run full of empty frames and we are not able to receive the expected answer from the codec. Therefore this patch disables IRQs for the current CPU during this critical step.

BCS: 746-001295

--

Commit: seco-ne/yocto/infrastructure/gitlab-ci@23680f48

Fix check for branch protection state

Since the changes of 939ade2b9f98b3e1569332270a7ab16ce7d70b4a we observe
that pipelines are executed on non-protected branches, which is not what
we want.

GitLab's documentation was a bit unclear on the according CI variables
CI_COMMIT_REF_PROTECTED and CI_MERGE_REQUEST_TARGET_BRANCH_PROTECTED, as
stated in this stack overflow answer:
https://stackoverflow.com/a/59023344/3018229

We need to explicitly check for the variable values to be "true" instead
of just checking whether they are set.

--

Commit: seco-ne/tools/gfxml2dto@4c5c76e0



main: add -f argument

If -f argument is provided, fsl flag will be set to true to config.

Signed-off-by: Mikhail Vanyulin <mikhail.vanyulin@rtsoft.de>

--

Commit: seco-ne/tools/gfxml2dto@33884991



dto: select functions based on fsl boolean flag

If fsl flag is true, overlays will be created for nodes from NXP kernel.

Signed-off-by: Mikhail Vanyulin <mikhail.vanyulin@rtsoft.de>

--

Commit: seco-ne/tools/gfxml2dto@a88c7fba



guf_xml_config.h: add fsl boolean flag for guf_xml_data struct

It will be used to select which functions will be used to create overlays
for LVDS related nodes.

Signed-off-by: Mikhail Vanyulin <mikhail.vanyulin@rtsoft.de>

--

Commit: seco-ne/tools/gfxml2dto@e2dddaee



dto: add functions to add lvds related overlays for fsl based imx6 dtb

Functions are based on guf_xml_to_dt.c from linux-imx-4.1.15 from Jethro BSP.

Signed-off-by: Mikhail Vanyulin <mikhail.vanyulin@rtsoft.de>

--

Commit: seco-ne/tools/secure-element-examples@ab59e351

CI: Add gitlab-ci files to the repository

--

Commit: seco-ne/kernel/linux-imx-kuk@f3fffaa5

fix AC97 read timeouts

Previously we see AC97 read register calls timing out every once in a
while. Adding some debug outputs, these mostly seem to be cases, where
the SAI driver doesn't find the expected answer in the received frames.

Three potential causes have been identified:
 * frame "border" occuring exactly between enabling RMR and TMR,
 resulting in receiving an additional empty RX frame (4 instead of 3)
 until the answer arrives
 * kernel preemption between RMR and TMR resulting in potentially lots
 of empty frames accumulating in the RX-FIFO
 * runnig in a timeout, because the code gave up after 63us already (3
 frames), although due to the asynchronous timing we might have to wait
 up to almost 104us (5 frames)

This patch adds fixes for all three cases. During first manual tests no
more timeouts could be produced.

BCS: 746-001295

--

Commit: seco-ne/kernel/linux-imx-kuk@5550ff9a

dual-espresso:arm64:dts: Set SD-card drive strength to DSE_X6

Some Trizeps8plus modules do not detect SD-cards correctly when DSE_X4
is set. Set drivce strength to DSE_X6, as all cards seem to work with
this value. The DSE_X6 value has already passed an emc pre-test.

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@7797a210

config:tr8plus: Added GPIO read cmd to defconfig

The read command was cherry-picked with ca1b358c.

BCS 746-000956

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@ca1b358c



cmd: gpio: Add `gpio read` subcommand

As explained in commit 4af2a33ee5b9 ("cmd: gpio: Make `gpio input`
return pin value again") the `gpio input` is used in scripts to obtain
the value of a pin, despite the fact that CMD_RET_FAILURE is
indistinguishable from a valid pin value.
To be able to detect failures and properly use the value of a GPIO in
scripts we introduce the `gpio read` command that sets the variable
`name` to the value of the pin. Return code of the `gpio read` command
can be used to check for CMD_RET_SUCCESS or CMD_RET_FAILURE.
CONFIG_CMD_GPIO_READ is used to enable the `gpio read` command.

Signed-off-by: Diego Rondini <diego.rondini@kynetics.com>

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@5682c066

dts:mc3: Added status and error LEDs

BCS 746-000956

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@37f22694

defconfig:tr8plus: Added LED commands to config

--

Commit: seco-ne/kernel/linux-imx-kuk@0f55bd22

dts:mc3: Added baseboard DT to Makefile

--

Commit: seco-ne/kernel/linux-imx-kuk@59c191e3

dts:mc3: Added test baseboard DT

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@255421d4

cmd:guf: Fix board detection of MC3

The board detection wasn't stable for the MC3 because it used a pin that
could toggle. Reworked the detection for the MC3 to use a combination of fuses
and processor type.

--

Commit: seco-ne/kernel/linux-imx-kuk@00c08635

dts:mc3: Updated the pixel clock of the LCDIF2

We now have a LVDS receiver that can handle slower pixel clocks.
Reduced the pixel clock further, to stay inside the display specs.

BCS 746-001313

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@ffece0f2

cmd:guf: Clean-up of SECO NE sub-system

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@63310364

spl:ttl: Set TTLOUT1 on GPIO-Expander

BCS 746-001155

--

Commit: seco-ne/kernel/linux-imx-kuk@24cde151

dts:mc3: Fixed OTG mode on DWC3.0

When the dr_mode is set to peripheral, the CC logic driver (PTN5150) does
not start correctly and the power delivery isn't activated on the
USB-C port.

BCS 746-001313

--

Commit: seco-ne/yocto/infrastructure/gitlab-ci@939ade2b

Depend intgration pipeline execution on branch protection state

Previously we had the master branch hard-coded in a variable in order to
determine whether to run the integration pipeline (on non-master branch)
or the merge pipeline (on master branch). This variable had to be
adapted everytime we copy a branch, which can potentially be forgotten
and lead to problems.

Remove the hard-coded branch and use the branch protection state for
this decision instead in order to make this more flexible.

--

Commit: seco-ne/yocto/infrastructure/gitlab-ci@dbc49ad0

Simplify yamllint job rules

The rules for the yamllint jobs have been defined in the .yamllint class
before. They were overridden in the manifest pipeline's yamllint job,
though, which has its own rules, so we don't need them to be inherited
in this case. And in the manifest integration pipeline the rules can be
much easier, so define them there instead of at the class-level.

--

Commit: seco-ne/yocto/infrastructure/gitlab-ci@cd8de4ff

.gitlab-ci: add analyze stage for limiting YAML script blocks

--

Commit: seco-ne/kernel/linux-imx-kuk@1b29c4bb

driver:gpio: Get input status from PI4IO

The driver didn't support the reading of the input status yet.

BCS 746-001155

--

Commit: seco-ne/kernel/linux-imx-kuk@3185c515

dts:mc3: Added old v1r0 DTBs

--

Commit: seco-ne/kernel/linux-guf@4c573ecd

arm:dts:santvend: Remove regulator for EN_VBAT to allow access using gpiod

--

Commit: seco-ne/kernel/linux-imx-kuk@d425df6e

dts:arm64:mx8m: Set cma size of mx8m to 320 MiB

By default, the cma size is set to 920MiB, which cannot be initialized
on some of our some of our machines because the memory is too small.
Previously, the cma value was bypassed by setting a suitable value
in the kernel command line.
To make the devicetree config usable, the cma size is set to 320 MiB.
The value seems to be suitable for EGL surfaces and video playback.
By default libimxdmabuffer (required for vpu video playback) expects
the cma heap node to be available at /dev/dma_heap/linux,cma-uncached.
This node is only created when the devicetree cma declaration is used.
However, it seems to be possible to create a soft-link if the node is
named differently.

BCS 746-001058

--

Commit: seco-ne/kernel/linux-imx-kuk@08a37c0b

dts:arm64:dual-espresso: Fix pfid gpio assignment

All pfid gpios were shifted by one field in the "gpio-line-names" assignment.

BCS 746-001162

--

Commit: seco-ne/kernel/linux-imx-kuk@1ecd3b66

dts:mc3: Updated the LED pins

The LED pins have changed. Updated the pinctrl group and changed the led
function.

--

Commit: seco-ne/kernel/linux-imx-kuk@871e0e9a

dts:mc3: Added PFID GPIOs

--

Commit: seco-ne/kernel/linux-imx-kuk@cbe9a8a2

dts:mc3: Removed old TTL define and fixed error in GPIO expander

--

Commit: seco-ne/kernel/linux-imx-kuk@cf0157dc

dts:mc3: Added GPIO-Expander and moved TTL-Pins

Added GPIO-Expander and connected the TTL-Pins to it.

BCS 746-001155

--

Commit: seco-ne/kernel/linux-imx-kuk@51802610

dts:arm64:dual-espresso: Add gpios for baseboard rev detection

BCS 746-001162

--

Commit: seco-ne/yocto/infrastructure/gitlab-ci@e943a55e

Use project path instead of name in integration branch name

Usually project name and path are identical after project creation. It
is possible, though, to change the name of a project without changing
its path.
When triggering a pipeline on the integration branch, we are using the
CI_PROJECT_NAME variable to reference the project. Unexpectedly, this
variable contains the *path*, not the *name* of the project (see
https://gitlab.com/gitlab-org/gitlab/-/issues/20030).
Hence we have to use the project path in all other places where the
integration branch name is constructed in our Python code.

--

Commit: seco-ne/yocto/infrastructure/gitlab-ci@bd7025b8

Manifest: run automatic pipeline on tags

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@7477fa0a

config:tr8m: Added Edgehog boot mode

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@e69e22ff

sysdata: Moved update of current slot to correct position

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@6264db01

sysdata: Re-added export of slot_partition

The changes were overwritten by the backport in commit 22379fa5.

BCS 746-001150

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@2e96669b

env: Fixed edgehog boot function

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@5cc2ea0b

sysdata: Set fixed part for environment location

Loading the part doesn't work at the moment. Use a fixed part of 0.

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@dbb19b87

sysdata: Updated to current version

Pulled the changes to sysdata from https://git.seco.com/arm/nxp/imx/develop/
yocto/5.x/u-boot-seco-imx/-/commit/c2d100148f4f50dc977891a580fe5980b399ecce

BCS 746-001143

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@23d5a499

sysdata: Disable the sysdata cmd by default

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@30e42b75

sysdata: Write current slot to the environment

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@bebff0ac

submodule:mkimage: Fixed mkimage submodule

The mkimage submodule is used to bundle the U-Boot image with
the firmware for various hardware components (e.g. ram).

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@81950919

sysdata: Added SECO sysdata to U-Boot

The Sysdata Environment is an additional environment for storing
system related data. Currently, it is mainly used for setting the
boot partition and keeping track of boot failures.

It is developed by SECO S.p.A.

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@116cf6da

env: Added function to load and run scripts from ext4

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@2ce4b46c

cmd: Activated the boot mode command for TANARO

The boot mode command (bmode) can change the start-up path (USB,
SD-Card, etc.) from within U-Boot.

--

Commit: seco-ne/3rd-party/kuk/uboot-imx-kuk@300b72aa

format: Added clang-format and tidy configurations

--

Commit: seco-ne/kernel/linux-imx-kuk@23b50d6e

driver:dwmac: Removed failing check for VLAN promisc mode

The commit is a workaround to re-allow the internal mac to be added to a
software bridge.

Also see https://community.nxp.com/t5/i-MX-Processors-Knowledge-Base/
Workaround-for-issue-Bridge-mode-on-EQoS-module-will-not-work/ta-p/1559302

BCS 746-001292

--

Commit: seco-ne/tools/gf-emc-test-suite@e40b72b7

ltp emc-testsuite: tr8mp-pbb testfiles

Add testfiles for immunity and emission file and add comments how to disable the
serial debug port.
Add relay test as script.

BCS 746-001210