You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Background: We have a unique scenario where our organization hosts a repository that is synced to another repository (identical name/code/etc) in an open-source project. This open-source project uses a physical build machine, as it was donated by some of our community members.
Problem: The problem we're having is that our workflow file specifies the action-zephyr-setup@v1 action, but because we're using a physical build machine, we get an error essentially saying the project has already been initialized (see error below), which is true because the build state persists on this machine.
Error:
Run west init -l application --mf west.yml
west init -l application --mf west.yml
west update -o=--depth=1 -n
shell: bash --noprofile --norc -e -o pipefail {0}
FATAL ERROR: already initialized in /__w/ocre-runtime/ocre-runtime, aborting.
Hint: if you do not want a workspace there,
remove this directory and re-run this command:
/__w/ocre-runtime/ocre-runtime/.west
Proposed Solution: We could add a cleanup step before this action to ensure that the project is not initialized however, it would be nicer/cleaner if this action provided an option to clean the workspace beforehand or ignore if a workspace is already initialized.
The text was updated successfully, but these errors were encountered:
Proposed Solution: We could add a cleanup step before this action to ensure that the project is not initialized however, it would be nicer/cleaner if this action provided an option to clean the workspace beforehand or ignore if a workspace is already initialized.
It's really not up to the action to prepare the workspace, if you have that initialized by a previous run that's honestly not a great design, you are basically not having isolation between runs. Having an option to skip the initialization may make sense if you have a solid use case for it but from what you describe you really should setup the system to start clean or add a clean step before the action in the workflow, maybe just a git clean -fxd or something would be enough.
Hey @fabiobaltieri , yeah I agree. For some reason it looks like these steps are not running in our build container, which is the real issue. Will continue digging on my side to see what the fix is here.
Background: We have a unique scenario where our organization hosts a repository that is synced to another repository (identical name/code/etc) in an open-source project. This open-source project uses a physical build machine, as it was donated by some of our community members.
Problem: The problem we're having is that our workflow file specifies the
action-zephyr-setup@v1
action, but because we're using a physical build machine, we get an error essentially saying the project has already been initialized (see error below), which is true because the build state persists on this machine.Error:
Proposed Solution: We could add a cleanup step before this action to ensure that the project is not initialized however, it would be nicer/cleaner if this action provided an option to clean the workspace beforehand or ignore if a workspace is already initialized.
The text was updated successfully, but these errors were encountered: