flake: update nixpkgs: 22.11 -> 23.05#8569
Conversation
|
|
|
After the nixpkgs fix, |
|
What's the status on this? Is the stuff we need in Nixpkgs already? |
|
This pull request has been mentioned on NixOS Discourse. There might be relevant details there: |
|
Once CI passes, we should just merge. There's nothing blocking it. |
|
I'm not sure why macos is failing on: |
|
Yes, it's in there and I had updated the lockfile and the commands above passed for me. I thought CI also passed but I'm not sure. |
|
It seems that linux passes, but macos fails. I don't see the history, but did the same error happen before on linux? |
|
All tests including this one seem to pass on nix master, but I have no idea why this PR could make a difference here. |
|
I can't reproduce it on aarch64-darwin and I don't have access to an intel machine. Maybe you can pass |
|
I have access to a native x86_64-darwin, if you tell me what to run/inspect. |
|
|
|
Uh, I don't know :-/ (repeatedly) |
|
I tried to run the same with newer nix than was on that machine and it all passed. |
Judging from Probably the hard link optimization wasn't triggered in an older version of nix. As that runs async, it could be quite non-deterministic. It would be great to have a slightly more functional default nix store when nix runs in the sandbox. We'll need that for effective use of RFC 92 dynamic derivations anyway. For a test like this, we run nix just for instantiation, which is basically the same requirement that we have for 92, and we've established that a very restrictive recursive nix socket that basically only allows |
|
I haven't mentioned that the machine I tried is configured close to defaults, i.e. without sandboxing. |
|
I set up a jobset for this PR: https://hydra.nixos.org/eval/1797165
There are two other problems:
|
|
The crashing So we may need to decide whether we care more about coverage analysis or property testing... |
|
Yikes. These lines in the log seem the describe the cause:
|
|
We have Hydra results! https://hydra.nixos.org/eval/1798739?compare=1798723&full=0#tabs-now-fail
|
Interface has changed upstream. It *should* be fine to test 23.05's other Nix versions as those *should* succeed, but that's not the case and it's obfuscating our terrible CI setup's log.
f5f1637 to
d3588a6
Compare
The lowdown input can't be updated; `nix build` would fail to find it. Co-authored-by: Robert Hensing <[email protected]>
d3588a6 to
539cc5e
Compare
|
Meeting notes 2023-09-01
|
|
This pull request has been mentioned on NixOS Discourse. There might be relevant details there: |
|
@roberth I updated this PR to use the latest 23.05, but it's still failing in the aws-sdk-cpp build: https://hydra.nixos.org/build/233688522 |
|
https://hydra.nixos.org/build/233688522 and https://hydra.nixos.org/build/232704350 look like the same derivation? I think we need to manually restart the failing builds in Hydra, since the 23.05 bump did not include a mass rebuild. |
|
That failing build has been retried several times already; you can see some attempts e.g. on https://hydra.nixos.org/build/232704299#tabs-buildsteps (You point to two different builds of nix but both depend on the same aws-sdk-cpp.) |
|
I'm having network difficulties with my aarch64-darwin node, so I can't troubleshoot that build right now. |
This is broken because of a change in systemd in NixOS 23.05. It fails with Failed to mount proc (type proc) on /proc (MS_NOSUID|MS_NODEV|MS_NOEXEC ""): Operation not permitted
|
I've disabled the systemd-nspawn test. That looks like a regression/change in systemd in NixOS 23.05 and I have no idea how to fix it. |
|
Thanks @edolstra for taking a look at this. |
https://hydra.nixos.org/build/235888160 This is needed because Nixpkgs now contains dangling symlinks (pkgs/test/nixpkgs-check-by-name/tests/symlink-invalid/pkgs/by-name/fo/foo/foo.nix).
| #inputs.nixpkgs.url = "github:NixOS/nixpkgs/nixos-23.05-small"; | ||
| inputs.nixpkgs.url = "github:edolstra/nixpkgs/fix-aws-sdk-cpp"; |
There was a problem hiding this comment.
https://nixpk.gs/pr-tracker.html?pr=256102 once this hits small, we can go back.
|
Merging this since the jobset is now green: https://hydra.nixos.org/eval/1799383 |
| # Run systemd-nspawn in a Nix build. | ||
| host.succeed("nix build -v --auto-allocate-uids --sandbox -L --offline --impure --file ${./systemd-nspawn.nix} --argstr nixpkgs ${nixpkgs}") | ||
| host.succeed("[[ $(cat ./result/msg) = 'Hello World' ]]") | ||
| #host.succeed("nix build -v --auto-allocate-uids --sandbox -L --offline --impure --file ${./systemd-nspawn.nix} --argstr nixpkgs ${nixpkgs}") |
These statements are taken from the `lib/test/release.nix` tests, which previously also worked on darwin. Unblocks NixOS/nix#8569 when backported
Motivation
Pinned nixpkgs commit over half a year old.
Context
#1254 (comment)
Checklist for maintainers
Maintainers: tick if completed or explain if not relevant
tests/**.shsrc/*/teststests/nixos/*Priorities
Add 👍 to pull requests you find important.