refactor(cli/rustup-mode): merge self-update checking logic in rustup update

[rami3l]

No description provided.

[djc]

This is redoing something that I undid only recently. I don't see much value in it -- it just obscures the original meaning and saves a few characters.

[rami3l]

@djc I still think the double negation here in !cfg!("no-*") is a bit confusing, but you can totally have different opinions.

[djc]

If this is the reason, I think you can cfg!() instead of an attribute.

[rami3l]

Sorry for not having explained the full background until now, but when looking at #4565 I realized that the self-update cleanup code is very curiously written in two parts, so I want to merge them together first to make the overall logic clearer.

However, since this change is not directly related to #4565, I planned on making two separate PRs...

[rami3l]

@djc I'm not entirely sure about your proposals, but maybe this could be simplified to just the following?

    if cfg!(feature = "no-self-update") {
        info!("self-update is disabled for this build of rustup");
        info!("any updates to rustup will need to be fetched with your system package manager")
    } else if self_update_mode == SelfUpdateMode::CheckOnly {
        check_rustup_update(&dl_cfg).await?;
    }

... the idea is that this block covers all the cases where self_update() is not executed, and when I add the extra if self_update block in a second PR, it'll be clearer when each case is handled.

Or you might have different opinions regarding the fix?

[rami3l]

I've made #4566 to hopefully make my original intentions clearer.

[rami3l]

Closing in favor of #4566.