This fixes a bug with the action's Docker container apparently not being
able to run Node correctly anymore...
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
Coverity issues need to be resolved and closed by owners, not by the
bot. Other wise we will continue scanning them and reporting them over
and over.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Seems like the PDF build hit the "detected dubious ownership in
repository" issue that has already been worked around in other workflow.
Add that step here as well.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Skip the rebase step if it's not a pull request. Schedule and push runs
have no base ref to rebase against anyway, currently the step is failing
and being skipped.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Hardwaremap testing requires installing requirements-run-tests.txt.
This change adds that to the relevant workflow.
Signed-off-by: Lukasz Mrugala <lukaszx.mrugala@intel.com>
Also run some split BT stack tests on the nrf5340 platform
to test this configuration.
Move the selection of which BT tests are built and run
in each platform to files in the tests/bsim/ folder.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
The compliance and coding guidelines workflows are using the same pip
cache key, but they install a different set of packages. This is
probably a copy paste error from the documentation build workflow, which
has been changed to use a hash, and is causing one of the two steps to
restore from a incorrect cache.
Fix this by using a hash of the workflow file itself as a key, as that's
where the python package list is defined.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Add some west update flags to do a shallow fetch of the modules and
skip the tags. That data is not needed anyway, should make the
compliance check initialization a bit faster.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
The license check workflow quietly stopped working after commit
8f66f854c3. Upgrading the checkout action
changed default behavior to only fetch one commit instead of all history
for all branches, which caused an uncaught fatal error in the scancode
action:
fatal: ambiguous argument 'origin/main..': unknown revision or path not
in the working tree.
The scancode action then completed successfully having not actually
checked anything.
Fix this by setting the checkout action fetch depth to 0, similar to
other workflows.
Signed-off-by: Maureen Helm <maureen.helm@analog.com>
As documentation is now built using one of our self-hosted runners,
have the GitHub action only run for the upstream repo, not for potential
forks.
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
Upgrade to v1.2.0, which comes with a bugfix and support for processing
PRs that have removed modifications to the manifest.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
looks like our docker image is way too big for the GH runners, so move
to own runners until we have a better solution.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
PR #63195 has managed to change code under test without
triggering those tests.
This change should remedy that by creating a new workflow,
based on twister_tests.yaml, that runs tests in
scripts/tests/build_helpers if build_helpers are modified.
Such a workflow should be easily extendable if we ever have more
non-Twister things to test in scripts/pylib.
Signed-off-by: Lukasz Mrugala <lukaszx.mrugala@intel.com>
Run all bsim bluetooth controller tests also
on a nrf5340bsim_nrf5340_cpunet, as the controller
runs in a quite different mode.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Update the text so it points to useful parts of the documentation
regarding CI and contribution guidelines.
Also invite users to join Discord to ask for help.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Add a rebase step after checkout for the documentation build CI run.
This is in line with what pretty much all other workfloww already do,
and means that if a PR is opened on a broken tree and the breakage get
fixed, the PR run just have to be retried for the test to pass and we
don't have to ask the authors to rebase and lose any pending approvals.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
Include CI runs to push and pull request against collab- branches so
that thay can be used for detecting CI breakages before trying to open
PRs against main.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
A recent fix in zephyrproject-rtos/action-first-interaction/pull/5 is
now making the bot's comments on PR be actual comments, not PR
"review comments".
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
Let's also run the BabbleSim tests when either
the native simulator or the the include headers
are changed.
The include path was forgotten when creating the filter.
The nrf52_bsim was changed to use the native simulator
in 3a4bebacb1
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Remove the Hardware Support request template. Hardware support requests
generally end up being stale and no one picks them up. This avoids giving
users false hope that simply opening a Hardware Support request will result
in someone implementing the requested support.
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
In addition to the white-box testing and unit tests,
we would like to have some black-box testing for the
console commands of twister.
This serves as a Proof of Concept for future expansion.
Special dummy tests are made for blackbox testing.
Signed-off-by: Lukasz Mrugala <lukaszx.mrugala@intel.com>
Add a scheduled entry to run_assignee to run the module assignment
function periodically, as well as a target to run the issue code when
issue labels are changed.
Signed-off-by: Fabio Baltieri <fabiobaltieri@google.com>
The compile script was modified to fully parallelize
all bsim test images builds.
This speed up the build time, but as we are
building ~80 images, the average load on the system
will skyrocket to over 200.
For CI systems we are seeing in quite many cases,
the CI server loosing connection with the CI runner,
while this part of the job is run, and suspect it is
due to this load.
Nice the job down (by the default +10),
as this should alleviate this problem without slowing
down the job.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
Zephyr scripts do not require documentation dependencies, so let's
move them from scripts/ to doc/.
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Add a new CMake option to append options to SPHINXOPTS. This allows us
to easily extend default SPHINXOPTS. This patch also restores the "-j
auto" option in CI (now that we use a custom runner).
Signed-off-by: Gerard Marull-Paretas <gerard.marull@nordicsemi.no>
Install the natsort package before running the west commands test
suites. The natsort package is required by twister, which is imported in
the runners, since c37deeb0c4.
Signed-off-by: Henrik Brix Andersen <henrik@brixandersen.dk>
So that we see all the potential errors in CI and not just the first
and can assess how much work there is to do.
This was introduced for debuggin but gmarull has suggested keeping it
in.
Signed-off-by: Michael Jones <m.pricejones@gmail.com>
When PDF build fails, we want to see what cmaused the failure, so upload
the log for later analysis.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
As the bsim repo is disabled by default, west list bsim -f {sha}
fails.
Instead get the revision field which works.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
prefix with numbers to get sorting based on usage. Most common files
issue types should come first, with bugs being at the top.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
