Build fails in single-user mode because install tests fail #365
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
3 participants
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: lix-project/lix#365
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 trying to build recent versions of Lix using the Nix packaging in single-user mode, there are a lot of test failures in the installCheckPhase.
Steps To Reproduce
ff08d95420
(latest as of time of writing).nix-build -A packages.x86_64-linux.nix --store [some directory]
.error: home directory '/homeless-shelter' exists; please remove it to assure purity of builds without sandboxing
.Expected behavior
The build succeeds.
nix --version
outputnix (Lix, like Nix) 2.90.0-pre0381aacff0be
Additional context
On
b6799ab037
(2.90-beta.1), the build succeeds. I'm currently running another bisect because the first result was garbage.Thanks for debugging this! Single user mode on Linux is a kind of bad configuration that I regret that we support but we definitely shouldn't regress it accidentally; honestly it ought to be made to perform identically to multi user because that should be possible. Interested to find the results of your debugging.
According to the bisect,
0565f97e78
is the first bad commit (which still seems a bit weird). FWIW, the first result gave677cf75473
, which is the commit right before that, so there may be some nondeterminism involved.Reverting
0565f97e78
does in fact "fix" the build.This issue was mentioned on Gerrit on the following CLs: