| 14 Jul 2025 |
QuadRadical (Ping) | im testing alyssas pr rn | 13:57:56 |
QuadRadical (Ping) | iti seems fairly reasonable to do if it works | 13:58:07 |
dramforever | i think the cross pkg-config can set its own env vars | 13:58:11 |
QuadRadical (Ping) | * it seems fairly reasonable to do if it works and doesnt break other packages | 13:58:15 |
K900 | Oh | 13:58:18 |
K900 | Oh. | 13:58:21 |
K900 | Yeah disregard me | 13:58:33 |
K900 | I am not thinking good today | 13:58:46 |
Artturin | In reply to @emilazy:matrix.org is there any convention that non-Rust packages follow https://github.com/NixOS/nixpkgs/blob/bf15e6ca04705033f4cfef069248ef55d8675cb9/pkgs/build-support/pkg-config-wrapper/setup-hook.sh#L26 | 14:00:08 |
John Ericson | It would be nice to patch that library to look for prefixed ones | 14:02:44 |
Alyssa Ross | You tried patching it before and didn't finish it :P | 14:03:09 |
John Ericson | But that depends on the rust triple not being different, maybe | 14:02:56 |
Alyssa Ross | Yes, it does | 14:03:25 |
Alyssa Ross | You'd need a mapping | 14:03:30 |
Alyssa Ross | We discussed this three years ago and IIRC you came around to my pkg-config hook change | 14:03:46 |
emily | yeah I am fine with this solution | 14:05:16 |
emily | Alyssa Ross: replaceStrings ["-"] ["_"] is not necessary, it checks both | 14:05:27 |
Alyssa Ross | In reply to @qyliss:fairydust.space We discussed this three years ago and IIRC you came around to my pkg-config hook change Sadly I don't seem to have decryption keys beyond 2023 | 14:05:57 |
emily | there's also a very dubious
pkgs/development/compilers/rust/cargo.nix
111: HOST_PKG_CONFIG_PATH = "${pkgsBuildBuild.pkg-config}/bin/pkg-config";
these days
| 14:05:58 |
emily | that is almost certainly wrong | 14:06:02 |
emily | and this is a silly hack that can go away
pkgs/build-support/rust/build-rust-package/default.nix
141: PKG_CONFIG_ALLOW_CROSS = if stdenv.buildPlatform != stdenv.hostPlatform then 1 else 0;
| 14:06:14 |
emily | other than that the PR LGTM | 14:06:30 |
Alyssa Ross | I guess I'll give it a rebase... | 14:06:55 |
emily | though we have a field for this these days right? | 14:07:22 |
emily | the env form | 14:07:24 |