`target add` and `component add` should succeed if target/component is already installed

[brson]

For automation purposes it's nice for this command to 'just work' when these commands are done redundantly. Change these commands to issue an 'info' indicating the thing is already installed and return successfully.

[theindigamer]

For cases when the target is distinct from the toolchain (say target=x86_64-pc-windows-gnu and toolchain=stable-x86_64-unknown-linux-gnu), there is an info issued along the lines of:

 info: component 'rust-std' for target 'x86_64-pc-windows-gnu' is up to date

Call order for issuing this info: bare_add_component -> ??? -> update -> ??? -> fmt (in notifications.rs) which calls ComponentAlreadyInstalled.

The error only happens when the target is a part of the toolchain (say target=x86_64-unknown-linux-gnu and toolchain=stable-x86_64-unknown-linux-gnu). I think this error is coming from here.

1. Should this error be matched and an info be issued in the corresponding functions target_add and component_add in rustup_mode.rs?
2. Should the info be issued using the same notification mechanism or do you just use a simple info! macro in the definitions for target_add/component_add?

[nre-ableton]

I'm also running into this "error" and find it very confusing. I am not sure why rustup would fail in this case when everything is in an "ok" state. Are there any plans to fix this?