| 15 Nov 2024 |
 emily | I don't think -nostdlibinc is the fix for that | 08:58:56 |
| emily | ideally we'd get the unwrapped compilers to not inject those GCC directories at all | 08:59:06 |
| emily | rather than using such a blunt tool to get there | 08:59:14 |
| emily | I forget the exact issue I found though | 08:59:29 |
 p14 |
that it's hard to use the libcxxStdenv compiler because it injects libstdc++ directories, sure
This statement reads as a surprise to me - what does 'libcxx' have to do with 'libstdc++'?
| 08:59:30 |
| emily | I think it's something like the GCC toolchain causes it to… something something | 08:59:42 |
| emily | I'll dig up the issue. | 08:59:44 |
| p14 | https://github.com/NixOS/nixpkgs/issues/151879 ? | 09:00:25 |
| emily | that might be the one for the wrapped compiler | 09:00:49 |
| emily | but maybe :) | 09:00:50 |
| emily | there's an Alyssa commit with a reproducer for the wrapped one | 09:00:57 |
| emily | hm, maybe all the reports were just for the wrapped compiler? | 09:01:08 |
| emily | but in that case I have no idea why someone felt the need to bake it into a patch | 09:01:15 |
| emily | let me do some proper digging | 09:01:22 |
| p14 | https://github.com/NixOS/nixpkgs/pull/153963 was the original patch that added it, but I did not understand the justification yet | 09:01:41 |
| emily | wow, backporting a change that incompatible to the previous release | 09:02:16 |
| p14 | No wait, looking at the patch it just changed the implementation | 09:02:19 |
| emily | the past is a foreign country | 09:02:27 |
| emily | In reply to @p14:matrix.org
No wait, looking at the patch it just changed the implementation yeah. but somehow fixed that bug? apparently? | 09:02:32 |
| emily | oh | 09:02:39 |
| emily | I guess the sed just rotted | 09:02:40 |
| emily | 1e26d33371a4a7238a644c49ddae49a4009c927f is the patch with the reproducer btw | 09:02:53 |
| emily | I say the criteria for replacing the patch is to not regress that example and unwrapped compilers can behave however | 09:03:09 |
| emily | anyway, I chased back the -nostdlibinc through Git blame before, let me do it again now that I'm not just venting to the macOS room 😅 | 09:03:34 |
| emily | it broke things for us because libclang is essentially unwrapped | 09:03:46 |
| emily | which is its own problem | 09:03:50 |
| emily | and a good reason we need to move off wrappers in time | 09:04:00 |
| p14 | There are old patches which mess with this. https://github.com/NixOS/nixpkgs/commit/006d8dcdc1127808d6c68f3b72f8c3d0d8e79014 | 09:04:49 |
| p14 | This was the first match for DriverArgs.hasArg via git log -G DriverArgs.hasArg ./pkgs/development/compilers/llvm/. | 09:05:15 |
| emily | uh, wow | 09:05:27 |
