expand WF obligations when checking method calls
Don't wrap a bunch of signatures in `FnPtr` then check their WF; instead, check the WFness of each input/output separately.
This is useful for the new trait solver, since because we stall on root obligations we end up needing to repeatedly recompute the WFness of possibly very large function signature types if it ends up bottoming out in ambiguity.
This may also give us more chances to hit the WF fast path for certain types like built-ins.
Finally, this just seems conceptually correct to do. There's nothing conceptually that suggests that wrapping the function signature in an fn pointer makes sense at all to do; I'm guessing that it was just convenient so that we didn't have to register WF obligations in a loop, but it doesn't affect the readability of this code at all.
Improve formatting of doc code blocks
We don't currently apply automatic formatting to doc comment code blocks. As a
result, it has built up various idiosyncracies, which make such automatic
formatting difficult. Some of those idiosyncracies also make things harder for
human readers or other tools.
This PR makes a few improvements to doc code formatting, in the hopes of making
future automatic formatting easier, as well as in many cases providing net
readability improvements.
I would suggest reading each commit separately, as each commit contains one
class of changes.
get rid of some false negatives in rustdoc::broken_intra_doc_links
rustdoc will not try to do intra-doc linking if the "path" of a link looks too much like a "real url".
however, only inline links (`[text](url)`) can actually contain a url, other types of links (reference links, shortcut links) contain a *reference* which is later resolved to an actual url.
the "path" in this case cannot be a url, and therefore it should not be skipped due to looking like a url.
fixes https://github.com/rust-lang/rust/issues/54191
to minimize the number of false positives that will be introduced, the following heuristic is used:
If there's no backticks, be lenient revert to old behavior.
This is to prevent churn by linting on stuff that isn't meant to be a link.
only shortcut links have simple enough syntax that they
are likely to be written accidentlly, collapsed and reference links
need 4 metachars, and reference links will not usually use
backticks in the reference name.
therefore, only shortcut syntax gets the lenient behavior.
here's a truth table for how link kinds that cannot be urls are handled:
| | is shortcut link | not shortcut link |
|--------------|--------------------|-------------------|
| has backtick | never ignore | never ignore |
| no backtick | ignore if url-like | never ignore |
Detect more `cfg`d out items in resolution errors
Use a visitor to collect *all* items (including those nested) that were stripped behind a `cfg` condition.
```
error[E0425]: cannot find function `f` in this scope
--> $DIR/nested-cfg-attrs.rs:4:13
|
LL | fn main() { f() }
| ^ not found in this scope
|
note: found an item that was configured out
--> $DIR/nested-cfg-attrs.rs:2:4
|
LL | fn f() {}
| ^
note: the item is gated here
--> $DIR/nested-cfg-attrs.rs:1:35
|
LL | #[cfg_attr(all(), cfg_attr(all(), cfg(FALSE)))]
| ^^^^^^^^^^
```
```
error[E0433]: failed to resolve: could not find `doesnt_exist` in `inner`
--> $DIR/diagnostics-cross-crate.rs:18:23
|
LL | cfged_out::inner::doesnt_exist::hello();
| ^^^^^^^^^^^^ could not find `doesnt_exist` in `inner`
|
note: found an item that was configured out
--> $DIR/auxiliary/cfged_out.rs:6:13
|
LL | #[cfg(false)]
| ----- the item is gated here
LL | pub mod doesnt_exist {
| ^^^^^^^^^^^^
```
Use a visitor to collect *all* items (including those nested) that were stripped behind a `cfg` condition.
```
error[E0425]: cannot find function `f` in this scope
--> $DIR/nested-cfg-attrs.rs:4:13
|
LL | fn main() { f() }
| ^ not found in this scope
|
note: found an item that was configured out
--> $DIR/nested-cfg-attrs.rs:2:4
|
LL | fn f() {}
| ^
note: the item is gated here
--> $DIR/nested-cfg-attrs.rs:1:35
|
LL | #[cfg_attr(all(), cfg_attr(all(), cfg(FALSE)))]
| ^^^^^^^^^^
```
Remove the witness type from coroutine *args* (without actually removing the type)
This does as much of rust-lang/rust#144157 as we can without having to break rust-lang/rust#143545 and/or introduce some better way of handling higher ranked assumptions.
Namely, it:
* Stalls coroutines based off of the *coroutine* type rather than the witness type.
* Reworks the dtorck constraint hack to not rely on the witness type.
* Removes the witness type from the args of the coroutine, eagerly creating the type for nested obligations when needed (auto/clone impls).
I'll experiment with actually removing the witness type in a follow-up.
r? lcnr
add unsupported_calling_conventions to lint list
Seems like you can emit lints without them being on the list, but users cannot control them then... *oops*.
Add tracing to step.rs and friends
Adds tracing calls to functions in `step.rs` (01717ffecfd47eb51f4877da6ad867b329a1ddd5), to friend functions related to evaluation and stepping (cbfa7c4b96b2ea26c1db185da9b59506bf8c8e55), and adds a new trait method `EnteredTraceSpan::or_if_tracing_disabled` (f0d0d1f5ecdf174696c8a74a5bc98967a2751c93).
Adding `EnteredTraceSpan::or_if_tracing_disabled` is optional and is only useful to avoid having both `tracing::info!()` calls (that existed before) and `enter_trace_span!()` calls (that this PR adds) that would be redundant and would slow down the collection of traces. I say it is optional because it adds some cognitive complexity around `EnteredTraceSpan`, which is possibly not worth the reduced redundancy. Let me know if I should revert that commit.
The tracing calls added in this PR are meant to make it easier to understand what was being executing at a particular point when looking at a trace. But they are likely not useful for the purpose of understanding which components are fast/slow, hence why I used `tracing_separate_thread` for them. After opening a trace generated using the code in this PR in https://ui.perfetto.dev, and after executing the following query and then pressing on "Show debug track", you will see something like the following image in the timeline:
```sql
select slices.id, ts, dur, track_id, category, args.string_value as name, depth, stack_id, parent_stack_id, parent_id, slices.arg_set_id, thread_ts, thread_instruction_count, thread_instruction_delta, cat, slice_id from slices inner join args USING (arg_set_id) where args.key = "args." || slices.name and name = "step"
```
<img width="739" height="87" alt="image" src="https://github.com/user-attachments/assets/74ad9619-9a1f-40e5-9ef4-3db31e33d6e1" />
Make tier 3 musl targets link dynamically by default
Since we don't build std for these and don't provide any support for them, these can trivially be changed to link dynamically by default.
Extend `is_case_difference` to handle digit-letter confusables
This PR extends `is_case_difference` to handle digit-letter confusables
Add support for detecting 0/O, 1/l, 5/S, 8/B, 9/g confusables in error suggestions.
r? `@estebank`
Deduplicate `IntTy`/`UintTy`/`FloatTy`.
There are identical definitions in `rustc_type_ir` and `rustc_ast`. This commit removes them and places a single definition in `rustc_ast_ir`. This requires adding `rust_span` as a dependency of `rustc_ast_ir`, but means a bunch of silly conversion functions can be removed.
r? `@fmease`
Consider operator's span when computing binop expr span
When computing the span of a binop consisting of `lhs` and `rhs`, we previously just took the spans of `lhs.span.to(rhs.span)`. In the case that both `lhs` and `rhs` are both arguments to a macro, this can produce a wildly incorrect span.
To fix this, first compute the span between `lhs` and the binary operator, which will cause `lhs` to possibly be adjusted to a relevant macro metavar, and then compute that span extended to `rhs`, which will cause it to also be adjusted to a relevant macro metavar.
This coincidentally fixes a FIXME in `tests/ui/lint/wide_pointer_comparisons.rs` and suppresses a nonsense suggestion.
Uniform `enter_trace_span!` and add documentation
1. The latest changes to `enter_trace_span!` were ported from Miri (see https://github.com/rust-lang/miri/pull/4452#discussion_r2204958019), so now both the `rustc_const_eval` and the Miri macro accept the same syntax. Furthermore, the Miri macro was changed to just call rustc_const_eval`'s, to avoid duplication.
2. I made the `layout_of` (& friends) calls use that new syntax, e.g. `enter_trace_span!(layouting::layout_of, ...)`
3. I made sure the macro specifies all types/traits/macros it refers to using `$crate::`, so the macro works anywhere independently of which types/traits/macros are imported in the context it is used in.
4. I added documentation, examples and tips to the macro's doc. To make the rustdoc compile I had to add some hidden lines (`#`), but now it acts as a compilation test which will avoid reintroducing issue 3. in the future. I will also create a documentation file with everything one needs to know about tracing at a later point, but I figured adding some of that info directly on the tracing macro makes it more discoverable.
5. In `stack.rs` I made it so that the `"frame"` span has a field named "frame" (instead of "message") with the data about the frame. This field used to be called "message" (tracing's default) since it was previously formatted using `"{}", instance`, and now I replaced it with `frame = %instance`.
Rollup of 4 pull requests
Successful merges:
- rust-lang/rust#136840 (Fix linker-plugin-lto only doing thin lto)
- rust-lang/rust#144053 (Remove install Rust script from CI)
- rust-lang/rust#144297 (Make `libtest::ERROR_EXIT_CODE` const public to not redefine it in rustdoc)
- rust-lang/rust#144721 (`std_detect`: Linux 6.16 support for RISC-V)
r? `@ghost`
`@rustbot` modify labels: rollup
Fix linker-plugin-lto only doing thin lto
When rust provides LLVM bitcode files to lld and the bitcode contains
function summaries as used for thin lto, lld defaults to using thin lto.
This prevents some optimizations that are only applied for fat lto.
We solve this by not creating function summaries when fat lto is
enabled. The bitcode for the module is just directly written out.
An alternative solution would be to set the `ThinLTO=0` module flag to
signal lld to do fat lto.
The code in clang that sets this flag is here:
560149b5e3/clang/lib/CodeGen/BackendUtil.cpp (L1150)
The code in LLVM that queries the flag and defaults to thin lto if not
set is here:
e258bca950/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp (L4441-L4446)
try-job: x86_64-gnu-debug
try-job: aarch64-gnu-debug
This commit changes it to store a `Region` instead of a `RegionVid` for the `Var` cases:
- We avoid having to call `Region::new_var` to re-create `Region`s from
`RegionVid`s in a few places, avoiding the interning process, giving a
small perf win. (At the cost of the type allowing some invalid
combinations of values.)
- All the cases now store two `Region`s, so the commit also separates
the `ConstraintKind` (a new type) from the `sub` and `sup` arguments
in `Constraint`.
