treewide: Remove use of apple_sdk framework stub paths

[booxter]

This patch covers all packages that use stub paths in some way in their
build instructions.

(Except those listed in this comment.)

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.

[booxter]

I've manually checked these packages on aarch64-darwin. Results are as follows:

vengi-tools: broken since 24.05: https://hydra.nixos.org/build/260818375
cardpeek: broken since 24.11: https://hydra.nixos.org/build/278953995
gildas: builds
nextpnr: broken in staging-next: https://hydra.nixos.org/build/294373417
odin: builds
opensc: builds
bullet: builds
bullet-roboschool: broken in 24.05: https://hydra.nixos.org/build/260816205
libsamplerate: builds
netcdffortran: builds
jaxlib: not utilized since the switch to -bin
pyscard: builds
bazel-test-cpp: passes
xorg.xorgserver: broken in staging-next: https://hydra.nixos.org/build/294644953 (dependency issue)

I've also checked for those already broken packages that failures are similar to what can be seen on hydra. So I hope this doesn't break anything that wasn't broken before.

OpenSC can be cleaned up because it detects the right path for PSCS if configure option is not passed:

opensc> PC/SC default provider:  /System/Library/Frameworks/PCSC.framework/PCSC
opensc> PKCS11 default provider: /nix/store/2mph7dm80lksszb10gf4mank6yp73zfp-opensc-0.26.1/lib/opensc-pkcs11.so
opensc> PKCS11 onepin provider:  /nix/store/2mph7dm80lksszb10gf4mank6yp73zfp-opensc-0.26.1/lib/onepin-opensc-pkcs11.so

bullet can be cleaned up more, need to remove patchPhase sections that tried to work around old SDK.

[booxter]

Not sure if this commit / PR should have treewide as its title tag. Please advise.

[NickCao]

The PR title should be treewide, but the changes should be in individual commits for individual packages.

[booxter]

I separated each package into a separate commit. I also split out more elaborate pieces out into individual PRs:

• nextpnr: 0.7.0 -> 0.8.0 #399630
• vengi-tools: fix build on darwin #399609
• cardpeek: fix darwin build #399602
• libAppleWM: fix build failure due to missing xorg-macros #399600 (to fix build of a dependency for xorg.xorgserver)

With these merged, we should have no packages using old apple_sdk stub paths in-tree.

[Astavie]

How come this is being removed? I remember the odin package requiring these inputs to compile on macos, has there been a recent change so that is not required anymore?

[yzx9]

How come this is being removed? I remember the odin package requiring these inputs to compile on macos, has there been a recent change so that is not required anymore?

Yes, the Darwin SDK has been fully reworked in #346043. The old-style SDKs were just no-op placeholders now.

[emilazy]

Looks great on the whole, thanks! Some nits.

[smaret]

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 399040 --package gildas

---

aarch64-darwin

✅ 1 package built:

• gildas

[smaret]

The change in gi/gildas/package.nix looks good to me. Thanks!

[emilazy]

Looks great, thanks! I plan to do a nixpkgs-review and then merge within a day or two. Some nits if you feel like addressing them before then, but I won’t block on them.

[emilazy]

Okay, when I said I’d run nixpkgs-review I hadn’t looked at the rebuilds total. But I built every directly touched package on x86_64-darwin except for xorg.xorgserver (didn’t want to build on staging and too lazy to test it on top of the libAppleWM change, plus it was broken before anyway), bullet-roboschool (broken on x86_64-darwin), and gildas (just takes too long). Thank you!