The description for the runner in the UI is by default sthg like
"npm_nixos_d0544ed48909" i.e., the name of the attribute.
I wanted to have a more user-friendly description and added a
description to the service.
Seems like gitlab-runner doesn't like having both fields set:
"Cannot use two forms of the same flag: description name"
so I used one or the other.
Add missing packaging module, fixes the following backtrace while trying
to use labgrid-client:
Traceback (most recent call last):
File "/nix/store/izj1dw63bqh47bq10v1b2d7qqg9al7f7-python3.9-labgrid-0.4.1/bin/.labgrid-client-wrapped", line 6, in <module>
from labgrid.remote.client import main
File "/nix/store/izj1dw63bqh47bq10v1b2d7qqg9al7f7-python3.9-labgrid-0.4.1/lib/python3.9/site-packages/labgrid/__init__.py", line 1, in <module>
from .target import Target
File "/nix/store/izj1dw63bqh47bq10v1b2d7qqg9al7f7-python3.9-labgrid-0.4.1/lib/python3.9/site-packages/labgrid/target.py", line 9, in <module>
from .driver import Driver
File "/nix/store/izj1dw63bqh47bq10v1b2d7qqg9al7f7-python3.9-labgrid-0.4.1/lib/python3.9/site-packages/labgrid/driver/__init__.py", line 4, in <module>
from .serialdriver import SerialDriver
File "/nix/store/izj1dw63bqh47bq10v1b2d7qqg9al7f7-python3.9-labgrid-0.4.1/lib/python3.9/site-packages/labgrid/driver/serialdriver.py", line 5, in <module>
from packaging import version
ModuleNotFoundError: No module named 'packaging'
- Drop pkgs/tools/admin/ansible and move everything into top-level and
throws into central aliases.nix
- Drop the Ansible 2.8 throw
- Remove Ansible 2.9/2.10, both will be EOL before the 22.05 release
- Remove Ansible 2.11, it will go EOL during the 22.05 release
- Expose the collections as `python3Packages.ansible`
Closes: #157591
Commit 6df37c9aab bumped the version of
evtest but failed to update the hash. As a result, hosts which
already have evtest-1.33 source present will build the old version but
label it as evtest-1.34. Hosts which lack the older source code will
fail their builds. This commit corrects the issue.
We should think about a way to get Hydra to catch issues like this.
Maybe require that if two FODs have different hashes then they must
have different `${pname}-${version}`s? Only for FODs, of course.