| 21 Jan 2025 |
 Tristan Ross | Something kinda like this | 04:41:03 |
| Tristan Ross | Just trying to read stdenv.cc.is* is able to cause infinite recursion in certain cases | 04:41:22 |
| Tristan Ross | It seems when the stdenv shifts to the dependencies of a derivation that's needed for the stdenv, the stdenv being used for those dependencies is borked in some way which is able to cause infinite recursion. | 04:46:00 |
| Tristan Ross | How can I have a setup hook add a step to the fixup phase? | 05:26:11 |
| Tristan Ross | I want a function to execute that does stuff during the fixup phase | 05:26:27 |
| Tristan Ross | But I don't want to override the fixup phase | 05:26:35 |
| Tristan Ross | Oh, preFixupPhases | 05:27:45 |
 John Ericson | In reply to @emilazy:matrix.org
which thing? https://github.com/philiptaron/nixpkgs/tree/gcc-ng-14 | 07:10:02 |
 Randy Eckenrode | Splitting up GCC like LLVM is? | 14:02:13 |
| Randy Eckenrode | @rosscomputerguy:matrix.org: I left some questions/comments re: Darwin on your toolchains PR. | 14:02:49 |
 Philip Taron (UTC-8) | Hmm. nix-build -A tests.stdenv.outputs-no-out is broken on master. | 14:22:26 |
| Philip Taron (UTC-8) | > /nix/store/shkw4qm9qcw5sc5n1k5jznc83ny02r39-default-builder.sh: line 1: pop_var_context: head of shell_variables not a function context
| 14:22:42 |
| Philip Taron (UTC-8) | * Hmm. nix-build -A tests.stdenv.outputs-no-out is broken on master for me. | 14:22:52 |
 FliegendeWurst | most likely not the real error? https://github.com/NixOS/nixpkgs/issues/214822 | 15:48:00 |
| Philip Taron (UTC-8) | Fixed quite a while ago, so unless it regressed... | 15:48:29 |
| Philip Taron (UTC-8) | (There's also no error otherwise.) | 15:48:38 |
| Philip Taron (UTC-8) | I'm bisecting. | 15:48:47 |
| Philip Taron (UTC-8) | It works on 24.11, so it's not too bad a bisect. | 15:49:32 |
| Philip Taron (UTC-8) | Found it: https://github.com/NixOS/nixpkgs/pull/357053#issuecomment-2605142607 | 16:06:42 |
| Tristan Ross | In reply to @reckenrode:matrix.org
@rosscomputerguy:matrix.org: I left some questions/comments re: Darwin on your toolchains PR. Thank you. I'm hoping we can get this merged soon. It'll be great once this PR is fully working. | 16:51:08 |
| Tristan Ross | https://github.com/NixOS/nixpkgs/pull/375435 this is going to be cool too | 16:51:45 |
| Tristan Ross | I just need to figure out why the hook isn't executing. | 16:52:01 |
| Tristan Ross | I was using the makeSetupHook but switched to the setupHook/setupHooks attribute and that causes it to not apply anymore. | 16:52:44 |
 emily | lmk when the PR is ready for another review | 16:53:03 |
| Tristan Ross | But GNU hello does work with bolt if you compiled it with LLVM | 16:53:21 |
| Tristan Ross | In reply to @emilazy:matrix.org
lmk when the PR is ready for another review It could use one, I need to figure out why Darwin has mass rebuilds | 16:53:55 |
| emily | lemme know when you figure that out. though a mass-rebuild may be correct if useLLVM was conditioning on something that actually applies to Darwin (which is not useLLVM) | 16:55:02 |
| Randy Eckenrode | In reply to @rosscomputerguy:matrix.org
Thank you. I'm hoping we can get this merged soon. It'll be great once this PR is fully working. The only other question I really still have is regarding libunwind. It’s linked implicitly by libSystem on Darwin. | 16:55:08 |
| Randy Eckenrode | In reply to @rosscomputerguy:matrix.org
It could use one, I need to figure out why Darwin has mass rebuilds Could be because some of the libunwind cases are different for Darwin. | 16:56:49 |
| Tristan Ross | In reply to @reckenrode:matrix.org
The only other question I really still have is regarding libunwind. It’s linked implicitly by libSystem on Darwin. Yeah which is why I switched to libunwind-system | 16:57:17 |
