re2c: add some key reverse dependencies to passthru.tests

[risicle]

Description of changes

Packages that have a large reverse dependency set are hard to test because a full rebuild is infeasible. But on the other hand, it's tricky for reviewers to remember which reverse dependencies of a particular package are particularly important to ensure are working or are known to be particularly sensitive to changes in the depended package.

Therefore it makes sense to add a small selection of such packages to passthru.tests. Done here for re2c. Interesting thematic relationship between chosen packages just a coincidence.

See also #167079, #167092 in the same vein.

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• 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/)
• 22.05 Release Notes (or backporting 21.11 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
  • (Release notes changes) Ran nixos/doc/manual/md-to-db.sh to update generated release notes
• Fits CONTRIBUTING.md.

[trofi]

Another possible heavy user of re2c is php.

[SuperSandro2000]

@ofborg eval

[Artturin]

CryptOpenSSLRandom should be marked as broken on aarch64-darwin since it doesn't build there so the ci won't fail

[SuperSandro2000]

CryptOpenSSLRandom should be marked as broken on aarch64-darwin since it doesn't build there so the ci won't fail

#172387