Reduce pub exposure in rustc_mir_build#125959
Conversation
Exhaustiveness and usefulness checking are now in `rustc_pattern_analysis`.
A lot of errors don't need to be visible outside the crate, and some other things as well.
|
Some changes occurred in match lowering cc @Nadrieril |
There was a problem hiding this comment.
Many errors.rs/lints.rs files have this kind of structure: public diagnostic struct with public fields.
Do you think it would make sense to deny the unreachable_pub lint in those files?
There was a problem hiding this comment.
This is the first I've heard of this lint! If it's reliable, I think we should use it just about everywhere :)
There was a problem hiding this comment.
As far as I know it is reliable and the only reason it isn't warn-by-default yet is because of pub-fields in non-pub-adt, cf. #120144 (comment).
|
Would like to see another PR changing (almost?) all diagnostic structs to be This is good for now @bors r+ rollup |
…mpiler-errors Rollup of 8 pull requests Successful merges: - rust-lang#125667 (Silence follow-up errors directly based on error types and regions) - rust-lang#125717 (Refactor `#[diagnostic::do_not_recommend]` support) - rust-lang#125795 (Improve renaming suggestion for names with leading underscores) - rust-lang#125865 (Fix ICE caused by ignoring EffectVars in type inference) - rust-lang#125953 (Streamline `nested` calls.) - rust-lang#125959 (Reduce `pub` exposure in `rustc_mir_build`) - rust-lang#125967 (Split smir `Const` into `TyConst` and `MirConst`) - rust-lang#125968 (Store the types of `ty::Expr` arguments in the `ty::Expr`) r? `@ghost` `@rustbot` modify labels: rollup
Rollup merge of rust-lang#125959 - nnethercote:rustc_mir_build-cleanups, r=compiler-errors Reduce `pub` exposure in `rustc_mir_build` r? compiler
|
I have started adding |
|
@nnethercote on the one hand, I'm glad you're doing that for the sake of the codebase, on the other I have a project that uses the rustc rlibs to build its own rustc_driver and there might be "unused pub items" that I might be using. At least two other tools might be silently affected by that too: miri and klint. That doesn't mean we shouldn't enable that lint, just that you should be aware that people might mildly grumble after the clean up. |
|
@estebank: I am familiar with that general problem, and there are now comments on several seemingly unused things in But |
r? compiler