cling: tidy/simplify + add experimental libc++ support

[thomasjm]

Description of changes

This PR tidies and improves the Cling derivation as part of my work to package xeus-cling.

In particular, this removes the weird auto-detection of flags for the C++ standard library, and replaces it with the actual correct ones.

I also added experimental support for LLVM libc++, for users who want to use this entirely without GCC. This is gated behind a flag.

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/)
• 23.11 Release Notes (or backporting 23.05 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.

[thomasjm]

No pressure to review @RaitoBezarius but I figure you're familiar with this now :)

[thomasjm]

Friendly ping @veprbl , I'm hoping to land this soon to unblock further work like #244777.

[veprbl]

There is some eval issue on ofborg.

[veprbl]

The compiler used to compile Cling may affect the runtime include and lib directories it expects to be run with. Cling builds against (a fork of) Clang, so we prefer to use Clang as the compiler as well for consistency.

Is there a specific problem that you saw, or this is done "just in case"?

[thomasjm]

There wasn't a problem with GCC as far as I'm aware, but it was confusing to mix two compiler toolchains. Especially when trying to debug linking errors and wondering where a given symbol is from.

The main motivation though is to pin down the makeWrapper flags here so we can get rid of the weird autodetection logic we had before. That logic was weird and actually defeated the purpose of exposing the flags, since it was separate from them but also needed at runtime. This issue came up when packaging xeus-cling.