More cfg alias targets + disabled related clippy lint

[Jan561]

Adds the remaining supported platforms as cfg alias.

This PR also disables this lint.

Checklist:

• I have read CONTRIBUTING.md
• I have written necessary tests and rustdoc comments
• A change log has been added if this PR modifies nix's API

[SteveLauC]

Thanks!

[asomers]

Why is there a need for those aliases?

[SteveLauC]

Why is there a need for those aliases?

Yeah, they are currently not in use, should I revert this PR?

I merged this PR because it is trivial and maybe there will be a use case for it in the future, when we don't want those target_os = "xxx" stuff and fully switch to aliases

[asomers]

Yes, I think we should revert it , just so we can reenable that lint. I've seen that lint fire correctly before.

[Jan561]

1. Either we use aliases for all supported targets, or we don't use them at all. Something in between is the worst option. I don't want to double check every time if there is a alias configured for the target.

2. Converts AddressFamily to struct with associated consts #2210 already uses these aliases, and having to configure these in a PR when needing them feels wrong.

3. This lint is necessary if we want to use #[cfg(linux)] instead of #[cfg(target_os = "linux")]

[asomers]

I don't think we should use them at all, if they only cover single targets. Aliases like apple_family are useful because they combine multiple targets. But the linux alias only saves a few characters. And I want to restore that lint precisely because I've seen people accidentally do #[cfg(linux)] where they meant #[cfg(target_os = "linux")]

[SteveLauC]

#2210 already uses these aliases, and having to configure these in a PR when needing them feels wrong.

This is my fault, I am sorry about it

[Jan561]

#2210 already uses these aliases, and having to configure these in a PR when needing them feels wrong.

This is my fault, I am sorry about it

All good, just need to tweak my codegen script a bit.