r? @Eh2406

(rust_highfive has picked a reviewer for you, use r? to override)

The cache_messages::rustdoc error seems unrelated to this PR, maybe it is caused by the new beta released yesterday?

The cache_messages::rustdoc error seems unrelated to this PR, maybe it is caused by the new beta released yesterday?

Thanks, posted #8653 with a fix.

Thanks! Rebased to rerun CI with your fix.

The Cargo team discussed this at our meeting, and we think that something of this nature should probably go through the RFC process. There are a number of issues with how Cargo discovers the config files to load, and we don't want to incrementally add small changes that would accumulate to something complex and difficult to work with. The comment at #7894 (comment) contains links to related issues, which would be a good place to start collecting all the different use cases and problems (and the following comments have more detail).

Perhaps you and @jsgf could maybe collaborate on that?

Thanks for your comment!

I took a look at the linked issues/proposals and tried to categorize and summarize them: https://internals.rust-lang.org/t/problems-of-cargo-config-files-and-possible-solutions/12987

Closing this in favor of #9030 and the proposal in https://internals.rust-lang.org/t/proposal-move-some-cargo-config-settings-to-cargo-toml/13336. An --ignore-config or --ignore-local-config flag might be still useful when compiling crates that you have no control over (e.g. when running a service like docs.rs), but for most people the new package-specific Cargo.toml keys should be the better solution.