Normalize types when deducing closure signature from supertraits
Elaborated supertraits should be normalized, since there's no guarantee they don't contain projections 😅Fixes#104025
r? types
Add a codegen test for rust-lang/rust#96152
This is a regression test for https://github.com/rust-lang/rust/issues/96152, it is intended to check that our codegen for a particular strict provenance pattern is always as good as the ptr2int2ptr/provenance-ignoring style.
r? `@nikic`
rustdoc: various cross-crate reexport fixes
Fixes for various smaller cross-crate reexport issues.
The PR is split into several commits for easier review. Will be squashed after approval.
Most notable changes:
* We finally render late-bound lifetimes in the generic parameter list of cross-crate functions & methods.
Previously, we would display the re-export of `pub fn f<'s>(x: &'s str) {}` as `pub fn f(x: &'s str)`
* We now render unnamed parameters of cross-crate functions and function pointers as underscores
since that's exactly what we do for local definitions, too. Mentioned as a bug in #44306.
* From now on, the rendering of cross-crate trait-object types is more correct:
* `for<>` parameter lists (for higher-ranked lifetimes) are now shown
* the return type of `Fn{,Mut,Once}` trait bounds is now displayed
Regarding the last list item, here is a diff for visualization (before vs. after):
```patch
- dyn FnOnce(&'any str) + 'static
+ dyn for<'any> FnOnce(&'any str) -> bool + 'static
```
The redundant `+ 'static` will be removed in a follow-up PR that will hide trait-object lifetime-bounds if they coincide with [their default](https://doc.rust-lang.org/reference/lifetime-elision.html#default-trait-object-lifetimes) (see [Zulip discussion](https://rust-lang.zulipchat.com/#narrow/stream/266220-rustdoc/topic/clean_middle_ty.3A.20I.20need.20to.20add.20a.20parameter/near/307143097)). `FIXME(fmease)`s were added.
``@rustbot`` label A-cross-crate-reexports
r? ``@GuillaumeGomez``
Implement `std::marker::Tuple`, use it in `extern "rust-call"` and `Fn`-family traits
Implements rust-lang/compiler-team#537
I made a few opinionated decisions in this implementation, specifically:
1. Enforcing `extern "rust-call"` on fn items during wfcheck,
2. Enforcing this for all functions (not just ones that have bodies),
3. Gating this `Tuple` marker trait behind its own feature, instead of grouping it into (e.g.) `unboxed_closures`.
Still needing to be done:
1. Enforce that `extern "rust-call"` `fn`-ptrs are well-formed only if they have 1/2 args and the second one implements `Tuple`. (Doing this would fix ICE in #66696.)
2. Deny all explicit/user `impl`s of the `Tuple` trait, kinda like `Sized`.
3. Fixing `Tuple` trait built-in impl for chalk, so that chalkification tests are un-broken.
Open questions:
1. Does this need t-lang or t-libs signoff?
Fixes#99820
Rollup of 7 pull requests
Successful merges:
- #103012 (Suggest use .. to fill in the rest of the fields of Struct)
- #103851 (Fix json flag in bootstrap doc)
- #103990 (rustdoc: clean up `.logo-container` layout CSS)
- #104002 (fix a comment in UnsafeCell::new)
- #104014 (Migrate test-arrow to CSS variables)
- #104016 (Add internal descriptions to a few queries)
- #104035 (Add 'closure match' test to weird-exprs.rs.)
Failed merges:
r? `@ghost`
`@rustbot` modify labels: rollup
Migrate test-arrow to CSS variables
There should be no UI changes. I kept both `color` and `background-color` properties even though only the ayu theme is actually completely making use of them on hover.
r? ``@notriddle``
rustdoc: clean up `.logo-container` layout CSS
This commit should result in no appearance changes.
To make the logo container exactly the desired height, you want to get rid of the part of the box used for typographic descenders (you know, the part of g, y, and j that descends below the baseline). After all, it contains no text, but the space is still left open in the layout by default, because `<img>` is `display:inline`. The CSS used to employ three different tricks to accomplish this:
* By making `.sidebar .logo-container` a flex container, the image becomes a flex item and is [blockified], without synthesizing any inline boxes. No inline boxes means no descenders.
* By giving `.mobile-topbar .logo-container` a max-height exactly the same as the height of the image plus the padding, the descender area gets cut off.
* By setting `.sub-logo-container { line-height: 0 }`, we ensure that the only box that contributes to the height of the line box is the image itself, and not any zero-content text boxes that neighbor it. See the [logical height algorithm].
This commit gets rid of the first two hacks, leaving only the third, since it requires only one line of code to accomplish and doesn't require setting the value based on math.
[blockified]: https://drafts.csswg.org/css-flexbox-1/#flex-items
[logical height algorithm]: https://www.w3.org/TR/css-inline-3/#inline-height
rework applying closure requirements in borrowck
Previously the promoted closure constraints were registered under the category `ConstraintCategory::ClosureBounds` in `type_check::prove_closure_bounds()` and then mapped back their original category in `regions_infer::best_blame_constraint` using the complicated map `closure_bounds_mapping`.
Now we're registering promoted constraints under their original category and span earlier in `type_check::prove_closure_bounds`.
See commit messages.
Fixes#99245
