{libGL,libGLU,libglut}: don’t use deprecated stubs on Darwin

[emilazy]

cc @reckenrode who had some reservations about the choice of apple-sdk over null here.

I have commits to make channel eval pass with null, but they’re a little ugly (e.g. changing packages that already don’t eval on Darwin, fixing stuff in KDE and Perl that doesn’t expect null build inputs) and since it broke a good handful of things in Nixpkgs, even ones that aren’t doing incorrect string interpolations, I figure it’d probably break external users too, and apple-sdk is strictly more correct in a sense (although sadly it becomes less strictly more correct in the presence of multiple SDK versions).

I’ve left in the truly benign/net‐positive clean‐ups I did on the road to making null work, though. I’m happy to change this to use null and restore the additional fix‐up commits if that’s preferred.

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandboxing enabled in nix.conf? (See Nix manual)
  • sandbox = relaxed
  • sandbox = true
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 25.05 Release Notes (or backporting 24.11 and 25.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
• Fits CONTRIBUTING.md.

---

Add a 👍 reaction to pull requests you find important.

[emilazy]

Switched this to null instead, and added back all the changes I had to make for eval to pass when I was trying that out originally one by one, as if it’d be any different this time just because the things that break it are so silly. Oh well. I’m not sure if null is actually the ideal thing here, but apple-sdk is sort of wrong/weird in the multi‐version case, and libintl is precedent for doing it this way, so… here it is.

[reckenrode]

Does this work? When the cc-wrapper hook processes the flags, it resolves environment variables?

[emilazy]

I think that they get expanded inside env.*? Maybe with structured attributes you have to do it in a Bash phase instead? I could just recommend that instead.

[reckenrode]

I think it has to be done in Bash in one of the phases. Otherwise, I think you’ll get something like:

clang: warning: argument unused during compilation: '-L$SDKROOT/System/Library/Frameworks/OpenGL.framework/Versions/Current/Libraries' [-Wunused-command-line-argument]

[emilazy]

Amended appropriately. Though since only one package in the tree even does this, probably it wasn’t even worth documenting.

[reckenrode]

Otherwise LGTM.

[K900]

LGTM

[emilazy]

Merging to unblock Darwin deprecations for the 25.11 deadline, but keeping the door open to changing null to something more compatible if necessary for #400739 concerns.