| 16 Oct 2025 |
 matthewcroughan | Ah okay, so ls | 17:39:27 |
| matthewcroughan | * Ah okay, so is that not equivalent to "the perl stdenv" ? | 17:39:37 |
| matthewcroughan | Is (buildPerlPackage {}).stdenv equivalent to the normal stdenv? | 17:40:19 |
 Alyssa Ross | yes | 17:40:22 |
| Alyssa Ross | I'm pretty sure this is it https://github.com/NixOS/nixpkgs/blob/fc4041932ec2f3ad8d59a6d57a8a89f700084fff/pkgs/development/perl-modules/generic/default.nix#L63 | 17:41:50 |
| Alyssa Ross | if it's canExecute cross, it uses the build platform's perl, which in this case has glibc-specifics baked into it | 17:42:26 |
 emily | it's the same either way, just mini vs. not, right? | 17:44:42 |
| emily | but very plausible it'll be invoking the wrong one | 17:44:51 |
| Alyssa Ross | It's splicing | 17:46:09 |
| Alyssa Ross | we don't actually want spliced pkgsBuildHost.perl here | 17:46:17 |
| Alyssa Ross | because that's the glibc one | 17:47:07 |
| Alyssa Ross | since we know canExecute is true, I think we want pkgsHostTarget.perl, which will have the constants right | 17:47:24 |
 dramforever | i'm starting to think fb-re2 is cursed | 17:48:33 |
| dramforever | i'm not entirely sure what's wrong | 17:49:37 |
| dramforever | but i think it literally is just using the wrong type | 17:49:45 |
| matthewcroughan | It hasn't been touched in a while, so that's plausible | 17:50:29 |
| matthewcroughan | I don't think it ever built, I think I figured out how to remove it from the nix deps somehwo | 17:50:37 |
| matthewcroughan | * I don't think it ever built, I think I figured out how to remove it from the nix deps somehow | 17:50:40 |
| matthewcroughan | https://github.com/facebook/pyre2/pull/25 | 17:51:08 |
| matthewcroughan | but yes it hasn't been touched in 6 years | 17:51:13 |
| matthewcroughan | But it is a dependency of mercurial, which Nixpkgs needs for building Nix tests I guess | 17:52:08 |
| matthewcroughan | https://github.com/NixOS/nixpkgs/blob/nixos-unstable/pkgs/by-name/me/mercurial/package.nix#L9 | 17:52:26 |
| matthewcroughan | hmm, but it is optional | 17:52:30 |
| matthewcroughan | All that is left then, is to figure out why pkgsMusl.pkgsStatic.musl doesn't build | 17:56:53 |
| matthewcroughan | rm -f lib/libc.a
aarch64-unknown-linux-musl-ar rc lib/libc.a obj/src/aio/aio.lo obj/src/aio/aio_suspend.lo obj/src/aio/lio_listio.lo obj/src/complex/__cexp.lo obj/src/complex/__cexpf.lo obj/src/complex/cabs.lo obj/src/complex/cabsf.lo obj/src/complex/cabsl.lo obj/src/complex/cacos.lo obj/src/complex/cacosf.lo obj/src/complex/cacosh.lo obj/src/complex/cacoshf.lo obj/src/complex/cacoshl.lo obj/src/c>
aarch64-unknown-linux-musl-gcc -std=c99 -nostdinc -ffreestanding -fexcess-precision=standard -frounding-math -fno-strict-aliasing -Wa,--noexecstack -D_XOPEN_SOURCE=700 -I./arch/aarch64 -I./arch/generic -Iobj/src/internal -I./src/include -I./src/internal -Iobj/include -I./include -g -O2 -fno-align-jumps -fno-align-functions -fno-align-loops -fno-align-labels -fira-region=one -fira>
-Wl,-e,_dlstart -o lib/libc.so obj/src/aio/aio.lo obj/src/aio/aio_suspend.lo obj/src/aio/lio_listio.lo obj/src/complex/__cexp.lo obj/src/complex/__cexpf.lo obj/src/complex/cabs.lo obj/src/complex/cabsf.lo obj/src/complex/cabsl.lo obj/src/complex/cacos.lo obj/src/complex/cacosf.lo obj/src/complex/cacosh.lo obj/src/complex/cacoshf.lo obj/src/complex/cacoshl.lo obj/src/complex/cacosl>
collect2: fatal error: cannot find 'ld'
compilation terminated.
| 17:57:57 |
| Alyssa Ross | chaining like that isn't really expected to work aiui | 17:58:46 |
| matthewcroughan | pkgsMusl.pkgsLLVM is perfect | 17:59:03 |
| matthewcroughan | maybe static isn't | 17:59:14 |
| matthewcroughan | somehow pkgs.nixos {} with the build/hostPlatform set correctly, is invoking the equivalent of pkgsMusl.pkgsStatic.musl | 18:00:22 |
| matthewcroughan | This is just a normal native musl system that maybe wants that because of some bug | 18:00:53 |
