!VRULIdgoKmKPzJZzjj:nixos.org

Nix Package Manager development

707 Members
For people hacking on Nix: https://github.com/NixOS/nix Nix maintainers can be reached here.155 Servers

Load older messages


SenderMessageTime
19 Sep 2024
@joerg:thalheim.ioMic92I can't find the flake in question either09:06:18
@aleksana:mozilla.orgaleksana (force me to bed after 18:00 UTC)
In reply to @jade_:matrix.org
i suspect my config is not good enough, so i defer to someone with one that will actually do this
The nixosTests.installer.simpleUefiSystemdBoot just contains a dead simple configuration.nix, but that is triggering the problem
09:13:53
@aleksana:mozilla.orgaleksana (force me to bed after 18:00 UTC) It's in nixpkgs/nixos/tests/installer.nix 09:13:53
@joerg:thalheim.ioMic92Ok. I will try locally09:20:18
@aleksana:mozilla.orgaleksana (force me to bed after 18:00 UTC)Maybe we should just open fsantizers and attach valgrind, and fix as many potential errors around libexpr as possible?09:22:36
@jade_:matrix.orgjade_wont help if the fuckup is gc induced09:32:07
@jade_:matrix.orgjade_gc means no sanitizers09:32:14
@aleksana:mozilla.orgaleksana (force me to bed after 18:00 UTC)Boehm gc 💀09:39:14
@joerg:thalheim.ioMic92I haven't seen this bug in nix-eval-jobs which doesn't enable GC. And I eval a fair bit of NixOS tests.10:06:42
@joerg:thalheim.ioMic92* I haven't seen this bug in nix-eval-jobs which doesn't enable GC. And I eval a fair bit of NixOS tests on a daily base.10:07:34
@joerg:thalheim.ioMic92This is the nixos configuration from simple-uefi-system-boot: https://gist.github.com/Mic92/0243df2a0b47bbc44ae0384de5f49aa911:04:22
@joerg:thalheim.ioMic92Copied out of the test VM11:04:32
@aleksana:mozilla.orgaleksana (force me to bed after 18:00 UTC)Reproducibly irreproducible11:05:58
@aleksana:mozilla.orgaleksana (force me to bed after 18:00 UTC)I thought running the actual checkout of that hydra job might be better, but still no luck11:06:45
@joerg:thalheim.ioMic92https://i.imgur.com/UYk2V3c.jpeg Let's try some more cores11:32:44
@joerg:thalheim.ioMic92I scheduled 6.4k evaluations11:45:08
@k900:0upti.meK900https://hydra.nixos.org/build/273036630/nixlog/45/tail13:14:48
@k900:0upti.meK900Same error on new Hydra eval on aarch6413:14:53
@k900:0upti.meK900I'm going to revert in a few hours unless there's strong objections13:15:10
@joerg:thalheim.ioMic92I cannot reproduce this bug after thousands of evaluations13:40:35
@k900:0upti.meK900I'm just going to make an executive decision and revert then13:41:06
@k900:0upti.meK900Because this looks like it's going to take a while to investigate13:41:59
@joerg:thalheim.ioMic92But how userfacing is the issue if no one has a proper reproducer?13:42:29
@lily:lily.flowersLily Foster
In reply to @joerg:thalheim.io
But how userfacing is the issue if no one has a proper reproducer?
If it's breaking hydra channel blockers, then well...
13:42:48
@k900:0upti.meK900It's at least hydra-facing13:42:50
@emilazy:matrix.orgemilyit's user-facing in that it's blocking the channels13:42:50
@k900:0upti.meK900Given we have a lot more users running evals than Hydras13:42:58
@k900:0upti.meK900I'd expect it to be hit13:43:04
@joerg:thalheim.ioMic92It's not broken in hydra just now: https://hydra.nixos.org/job/nixos/unstable-small/nixos.tests.installer.simpleUefiSystemdBoot.aarch64-linux13:43:29
@joerg:thalheim.ioMic92Interestingly it was the same eval that broke on x86_64 and aarch6413:44:05

Show newer messages


Back to Room ListRoom Version: 6