installCheckPhase fails when building lix with a forgejo action #310
Labels
No labels
Area/build-packaging
Area/cli
Area/evaluator
Area/fetching
Area/flakes
Area/language
Area/profiles
Area/protocol
Area/releng
Area/remote-builds
Area/repl
Area/store
bug
crash 💥
Cross Compilation
devx
docs
Downstream Dependents
E/easy
E/hard
E/help wanted
E/reproducible
E/requires rearchitecture
imported
Needs Langver
OS/Linux
OS/macOS
performance
regression
release-blocker
RFD
stability
Status
blocked
Status
invalid
Status
postponed
Status
wontfix
testing
testing/flakey
ux
No milestone
No project
No assignees
4 participants
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: lix-project/lix#310
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Describe the bug
When running the CI for my infra on podman, building lix fails at the
installCheckPhase
, with 12 failed checks.Some of the failures are due to #157, but other are not:
I added the logs of these tests.
Steps To Reproduce
I agree that these error may not be easy to reproduce, given the complicated build process, it uses https://git.hubrecht.ovh/hubrecht/nix-modules/src/branch/main/services/forgejo-nix-runners/default.nix to run nix builds inside podman containers through forgejo actions...
Expected behavior
The tests to not fail
nix --version
outputAdditional context
Add any other context about the problem here.
I managed to capture the full logs of the installChecks failure
So when it ssh's into localhost it then can't find busybox? wat
worth noting that ssh to localhost does not invoke ssh, instead it invokes
bash -c
. perhaps that interacts somehow.Reproducer split out from the nixos stuff:
virtualisation.podman.enable = true;
and get xonsh available.Dump this in
sandbox.xsh
:Then in xonsh:
Inside the container:
This will explode in the same way as described above, I think. FWIW: your container setup is pretty busted and has bad perms on /tmp among other things.
There are complaints that
/nix/store/l85ia9bwz4v0k09v716sx2f2cjar3jvp-busybox
doesn't exist. This is copied from/nix/store/svsnsvyz3h7vxkcssalms7c5y4pq5n5h-busybox-static-x86_64-unknown-linux-musl-1.36.1/bin/busybox
via the$busybox
variable in the tests, which becomes--arg busybox /nix/store/svsnsvyz3h7vxkcssalms7c5y4pq5n5h-busybox-static-x86_64-unknown-linux-musl-1.36.1/bin/busybox
, which is a path literal that causes that copy.I don't think that that copy is busted, so maybe it got deleted????? idk!!!
I don't know why the copy doesn't appear to actually happen.
I've attached test logs from this container. You can view the junit one with https://github.com/lukejpreston/xunit-viewer/ which is slightly nicer.
At this point I don't have many more spoons to fight this issue.