python3Packages.torch: explicitly include libopenmp on Darwin x86_64

[sarahec]

PR #424873 fixed multiple builds hanging by disabling a "duplicate"* copy of the OpenMP dylib on Darwin. While this fixed, aarch64, this broke compilation on x86_64 (see #430472).

This PR omits the "duplicate" copy only on aarch64 and restores it on x86_64.

* "Duplicate": two copies of the dylib visible while running. Normally this is a non-issue on Darwin, but FAISS crashes on ARM with multiple copies of openMP visible. (See pytorch/pytorch#149201 (comment))

Fixes #430472

Things done

• Built on platform:
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• Tested, as applicable:
  • NixOS tests in nixos/tests.
  • Package tests at passthru.tests.
  • Tests in lib/tests or pkgs/test for functions and "core" functionality.
• Ran nixpkgs-review on this PR. See nixpkgs-review usage.
• Tested basic functionality of all binary files, usually in ./result/bin/.
• Nixpkgs Release Notes
  • Package update: when the change is major or breaking.
• NixOS Release Notes
  • Module addition: when adding a new NixOS module.
  • Module update: when the change is significant.
• Fits CONTRIBUTING.md, pkgs/README.md, maintainers/README.md and other READMEs.

---

Add a 👍 reaction to pull requests you find important.

[sarahec]

nixpkgs-review result

aarch64-Darwin

Generated using nixpkgs-review.

Command: nixpkgs-review pr 430808
Commit: c387b7908bd35d57b3e1df6e587dd4dd547c88bd

[sarahec]

nixpkgs-review result

x86_64-Darwin

Generated using nixpkgs-review.

Command: nixpkgs-review pr 430808

[sarahec]

Normally an empty nixpkgs-review would worry me, but I confirmed building on both aarch64 and x86-64. This PR maintains the status quo for aarch64 builds and uses the non-broken x86_64 builds from the prior PR. So I guess it makes sense.

[GaetanLepage]

Thanks!