Add `$message_type` field to distinguish json diagnostic outputs
Currently the json-formatted outputs have no way to unambiguously determine which kind of message is being output. A consumer can look for specific fields in the json object (eg "message"), but there's no guarantee that in future some other kind of output will have a field of the same name.
This PR adds a `"type"` field to add json outputs which can be used to unambiguously determine which kind of output it is. The mapping is:
`diagnostic`: regular compiler diagnostics
`artifact`: artifact notifications
`future_incompat`: Future incompatibility report
`unused_extern`: Unused crate warnings/errors
This matches the "internally tagged" representation for serde enums.
Add place.ty() and Ty build from a kind to smir
Add a method to retrieve the type of a place and a few utility functions needed to build the projection type. I decided to return a result to avoid panicking if the user passes invalid inputs, such as wrong list of locals.
r? `@spastorino`
Rollup of 8 pull requests
Successful merges:
- #117327 (Add documentation for some queries)
- #117835 (Note about object lifetime defaults in does not live long enough error)
- #117851 (Uplift `InferConst` to `rustc_type_ir`)
- #117973 (test: Add test for async-move in 2015 Rust proc macro)
- #117992 (Don't require intercrate mode for negative coherence)
- #118010 (Typeck break expr even if break is illegal)
- #118026 (Don't consider regions in `deref_into_dyn_supertrait` lint)
- #118089 (intercrate_ambiguity_causes: handle self ty infer + reservation impls)
r? `@ghost`
`@rustbot` modify labels: rollup
Don't consider regions in `deref_into_dyn_supertrait` lint
I actually wonder if we should just warn on *any* deref impl with a target type that matches a supertrait by *def-id*.
cc #89460
r? types
Typeck break expr even if break is illegal
Fixes#117821
We were returning immediately when encountering an illegal break. However, this caused problems later when the expr that the break was returning was evaluated during writeback. So now we don't return and instead simply set tainted by error. This lets typeck of break expr to occur even though we've encountered an illegal break.
Don't require intercrate mode for negative coherence
Negative coherence needs to be *sound*, but does not need to be *complete*, since it's looking for the *existence* of a negative goal, not the non-existence of a positive goal.
This removes some trivial and annoying ambiguities when a negative impl has region constraints.
r? lcnr idk if this needs an fcp but if it does, pls kick it off
Uplift `InferConst` to `rustc_type_ir`
We need this in `rustc_type_ir` because the canonicalizer must understand the difference between a const vid and an effect vid. In that way, it's not an implementation detail of the representation of an infer const, but just part of the type ir.
If we find out later on that it's better to leave the representation up to the consumer of `rustc_type_ir`, we could abstract `InferConst` (and probably `InferTy` as well) with some traits, but I don't see the benefit of that indirection currently.
Note about object lifetime defaults in does not live long enough error
This is a aspect of Rust that frequently trips up people who are not aware of it yet. This diagnostic attempts to explain what's happening and why the lifetime constraint, that was never mentioned in the source, arose.
The implementation feels a bit questionable, I'm not sure whether there are better ways to do this. There probably are.
fixes#117835
r? types
Rollup of 8 pull requests
Successful merges:
- #117828 (Avoid iterating over hashmaps in astconv)
- #117832 (interpret: simplify handling of shifts by no longer trying to handle signed and unsigned shift amounts in the same branch)
- #117891 (Recover `dyn` and `impl` after `for<...>`)
- #117957 (if available use a Child's pidfd for kill/wait)
- #117988 (Handle attempts to have multiple `cfg`d tail expressions)
- #117994 (Ignore but do not assume region obligations from unifying headers in negative coherence)
- #118000 (Make regionck care about placeholders in outlives components)
- #118068 (subtree update cg_gcc 2023/11/17)
r? `@ghost`
`@rustbot` modify labels: rollup
We were earlier returning immediately when encountering an illegal break. However, this caused problems later
when the expr that the break was returning was evaluated during writeback. So now we don't return and instead
simply set tainted by error. This lets typeck of break expr to occur even though we've encountered an illegal break.
Fix insertion of statements to be executed along return edge in inlining
Inlining creates additional statements to be executed along the return
edge: an assignment to the destination, storage end for temporaries.
Previously those statements where inserted directly into a call target,
but this is incorrect when the target has other predecessors.
Avoid the issue by creating a new dedicated block for those statements.
When the block happens to be redundant it will be removed by CFG
simplification that follows inlining.
Fixes#117355
Inlining creates additional statements to be executed along the return
edge: an assignment to the destination, storage end for temporaries.
Previously those statements where inserted directly into a call target,
but this is incorrect when the target has other predecessors.
Avoid the issue by creating a new dedicated block for those statements.
When the block happens to be redundant it will be removed by CFG
simplification that follows inlining.
Fixes#117355
This lint is not triggered if any of the following conditions are met:
- The user explicitly annotates the binding with the `()` type.
- The binding is from a macro expansion.
- The user explicitly wrote `let () = init;`
- The user explicitly wrote `let pat = ();`. This is allowed for local
lifetimes.
Make regionck care about placeholders in outlives components
Currently, we don't consider a placeholder type `!T` to be a type component when it comes to processing type-outlives obligations. This means that they are essentially treated like unit values with no sub-components, and always outlive any region. This is problematic for `non_lifetime_binders`, and even more problematic for `with_negative_coherence`, since negative coherence uses placeholders as universals.
This PR adds `Component::Placeholder` which acts much like `Component::Param`. This currently causes a regression in some non-lifetime-binders tests because `for<T> T: 'static` doesn't imply itself when processing outlives obligations, so code like this will fail:
```
fn foo() where for<T> T: 'static {
foo() //~ fails
}
```
Since the where clause doesn't imply itself. This requires making the `MatchAgainstHigherRankedOutlives` relation smarter when it comes to binders.
r? types
Ignore but do not assume region obligations from unifying headers in negative coherence
Partly addresses a FIXME that was added in #112875. Just as we can throw away the nested trait/projection obligations from unifying two impl headers, we can also just throw away the region obligations too.
I removed part of the FIXME that was incorrect, namely:
> Given that the only region constraints we get are involving inference regions in the root, it shouldn't matter, but still sus.
This is not true when unifying `fn(A)` and `for<'b> fn(&'b B)` which ends up with placeholder region outlives from non-root universes. I'm pretty sure this is okay, though it would be nice if we were to use them as assumptions. See the `explicit` revision of the test I committed, which still fails.
Fixes#117986
r? lcnr, feel free to reassign tho.
Recover `dyn` and `impl` after `for<...>`
Recover `dyn` and `impl` after `for<...>` in types. Reuses the logic for parsing bare trait objects, so it doesn't fix cases like `for<'a> dyn Trait + dyn Trait` or anything, but that seems somewhat of a different issue.
Parsing recovery logic is a bit involved, but I couldn't find a way to simplify it.
Fixes#117882
interpret: simplify handling of shifts by no longer trying to handle signed and unsigned shift amounts in the same branch
While we're at it, also update comments in codegen and MIR building related to shifts, and fix the overflow error printed by Miri on negative shift amounts.
Add arm64e-apple-ios & arm64e-apple-darwin targets
This introduces
* `arm64e-apple-ios`
* `arm64e-apple-darwin`
Rust targets for support `arm64e` architecture on `iOS` and `Darwin`.
So, this is a first approach for integrating to the Rust compiler.
## Tier 3 Target Policy
> * A tier 3 target must have a designated developer or developers (the "target
maintainers") on record to be CCed when issues arise regarding the target.
(The mechanism to track and CC such developers may evolve over time.)
I will be the target maintainer.
> * Targets must use naming consistent with any existing targets; for instance, a
target for the same CPU or OS as an existing Rust target should use the same
name for that CPU or OS. Targets should normally use the same names and
naming conventions as used elsewhere in the broader ecosystem beyond Rust
(such as in other toolchains), unless they have a very good reason to
diverge. Changing the name of a target can be highly disruptive, especially
once the target reaches a higher tier, so getting the name right is important
even for a tier 3 target.
Target names should not introduce undue confusion or ambiguity unless
absolutely necessary to maintain ecosystem compatibility. For example, if
the name of the target makes people extremely likely to form incorrect
beliefs about what it targets, the name should be changed or augmented to
disambiguate it.
If possible, use only letters, numbers, dashes and underscores for the name.
Periods (.) are known to cause issues in Cargo.
The target names `arm64e-apple-ios`, `arm64e-apple-darwin` were derived from `aarch64-apple-ios`, `aarch64-apple-darwin`.
In this [ticket,](#73628) people discussed the best suitable names for these targets.
> In some cases, the arm64e arch might be "different". For example:
> * `thread_set_state` might fail with (os/kern) protection failure if we try to call it from arm64 process to arm64e process.
> * The returning value of dlsym is PAC signed on arm64e, while left untouched on arm64
> * Some function like pthread_create_from_mach_thread requires a PAC signed function pointer on arm64e, which is not required on arm64.
So, I have chosen them because there are similar triplets in LLVM. I think there are no more suitable names for these targets.
> * Tier 3 targets may have unusual requirements to build or use, but must not
create legal issues or impose onerous legal terms for the Rust project or for
Rust developers or users.
The target must not introduce license incompatibilities.
Anything added to the Rust repository must be under the standard Rust
license (MIT OR Apache-2.0).
The target must not cause the Rust tools or libraries built for any other
host (even when supporting cross-compilation to the target) to depend
on any new dependency less permissive than the Rust licensing policy. This
applies whether the dependency is a Rust crate that would require adding
new license exceptions (as specified by the tidy tool in the
rust-lang/rust repository), or whether the dependency is a native library
or binary. In other words, the introduction of the target must not cause a
user installing or running a version of Rust or the Rust tools to be
subject to any new license requirements.
Compiling, linking, and emitting functional binaries, libraries, or other
code for the target (whether hosted on the target itself or cross-compiling
from another target) must not depend on proprietary (non-FOSS) libraries.
Host tools built for the target itself may depend on the ordinary runtime
libraries supplied by the platform and commonly used by other applications
built for the target, but those libraries must not be required for code
generation for the target; cross-compilation to the target must not require
such libraries at all. For instance, rustc built for the target may
depend on a common proprietary C runtime library or console output library,
but must not depend on a proprietary code generation library or code
optimization library. Rust's license permits such combinations, but the
Rust project has no interest in maintaining such combinations within the
scope of Rust itself, even at tier 3.
"onerous" here is an intentionally subjective term. At a minimum, "onerous"
legal/licensing terms include but are not limited to: non-disclosure
requirements, non-compete requirements, contributor license agreements
(CLAs) or equivalent, "non-commercial"/"research-only"/etc terms,
requirements conditional on the employer or employment of any particular
Rust developers, revocable terms, any requirements that create liability
for the Rust project or its developers or users, or any requirements that
adversely affect the livelihood or prospects of the Rust project or its
developers or users.
No dependencies were added to Rust.
> * Neither this policy nor any decisions made regarding targets shall create any
binding agreement or estoppel by any party. If any member of an approving
Rust team serves as one of the maintainers of a target, or has any legal or
employment requirement (explicit or implicit) that might affect their
decisions regarding a target, they must recuse themselves from any approval
decisions regarding the target's tier status, though they may otherwise
participate in discussions.
> * This requirement does not prevent part or all of this policy from being
cited in an explicit contract or work agreement (e.g. to implement or
maintain support for a target). This requirement exists to ensure that a
developer or team responsible for reviewing and approving a target does not
face any legal threats or obligations that would prevent them from freely
exercising their judgment in such approval, even if such judgment involves
subjective matters or goes beyond the letter of these requirements.
Understood.
I am not a member of a Rust team.
> * Tier 3 targets should attempt to implement as much of the standard libraries
as possible and appropriate (core for most targets, alloc for targets
that can support dynamic memory allocation, std for targets with an
operating system or equivalent layer of system-provided functionality), but
may leave some code unimplemented (either unavailable or stubbed out as
appropriate), whether because the target makes it impossible to implement or
challenging to implement. The authors of pull requests are not obligated to
avoid calling any portions of the standard library on the basis of a tier 3
target not implementing those portions.
Understood.
`std` is supported.
> * The target must provide documentation for the Rust community explaining how
to build for the target, using cross-compilation if possible. If the target
supports running binaries, or running tests (even if they do not pass), the
documentation must explain how to run such binaries or tests for the target,
using emulation if possible or dedicated hardware if necessary.
Building is described in the derived target doc.
> * Tier 3 targets must not impose burden on the authors of pull requests, or
other developers in the community, to maintain the target. In particular,
do not post comments (automated or manual) on a PR that derail or suggest a
block on the PR based on a tier 3 target. Do not send automated messages or
notifications (via any medium, including via `@)` to a PR author or others
involved with a PR regarding a tier 3 target, unless they have opted into
such messages.
> * Backlinks such as those generated by the issue/PR tracker when linking to
an issue or PR are not considered a violation of this policy, within
reason. However, such messages (even on a separate repository) must not
generate notifications to anyone involved with a PR who has not requested
such notifications.
Understood.
> * Patches adding or updating tier 3 targets must not break any existing tier 2
or tier 1 target, and must not knowingly break another tier 3 target without
approval of either the compiler team or the maintainers of the other tier 3
target.
> * In particular, this may come up when working on closely related targets,
such as variations of the same architecture with different features. Avoid
introducing unconditional uses of features that another variation of the
target may not have; use conditional compilation or runtime detection, as
appropriate, to let each target run code supported by that target.
These targets are not fully ABI compatible with arm64e code.
#73628
Begin to abstract `rustc_type_ir` for rust-analyzer
This adds the "nightly" feature which is used by the compiler, and falls back to more simple implementations when that is not active.
r? `@lcnr` or `@jackh726`
When encountering struct fn call literal with private fields, suggest all builders
When encountering code like `Box(42)`, suggest `Box::new(42)` and *all* other associated functions that return `-> Box<T>`.
Add a way to give pre-sorted suggestions.
Ensure sanity of all computed ABIs
This moves the ABI sanity assertions from the codegen backend to the ABI computation logic. Sadly, due to past mistakes, we [have to](https://github.com/rust-lang/rust/pull/117351#issuecomment-1788495503) be able to compute a sane ABI for nonsensical function types like `extern "C" fn(str) -> str`. So to make the sanity check pass we first need to make all ABI adjustment deal with unsized types... and we have no shared infrastructure for those adjustments, so that's a bunch of copy-paste. At least we have assertions failing loudly when one accidentally sets a different mode for an unsized argument.
To achieve this, this re-lands the parts of https://github.com/rust-lang/rust/pull/80594 that got reverted in https://github.com/rust-lang/rust/pull/81388. To avoid breaking wasm ABI again, that ABI now explicitly opts-in to the (wrong, broken) ABI that we currently keep for backwards compatibility. That's still better than having *every* ABI use the wrong broken default!
Cc `@bjorn3`
Fixes https://github.com/rust-lang/rust/issues/115845
Lint pinned `#[must_use]` pointers (in particular, `Box<T>` where `T` is `#[must_use]`) in `unused_must_use`.
Fixes: #111458
This is motivated by a common async/await pattern:
```rs
fn foo() -> Pin<Box<dyn Future<Output = i32>>> {
Box::pin(async { 42 })
}
// call `foo`, but forget to await the result
foo();
```
Unlike with `async fn` or return position `impl Future`, this does not currently warn the user that the `Future` is unused.
To fix this, I've extended the `unused_must_use` lint to catch `Pin<P>`, where `P` must be used. In particular, this applies to `Pin<Box<T>>`, where `T` must be used. I'm not sure if there are other pointers where this applies, but I can't think of any situation the user wouldn't want to be warned.
Suggest field typo through derefs
Take into account implicit dereferences when suggesting fields.
```
error[E0609]: no field `longname` on type `Arc<S>`
--> $DIR/suggest-field-through-deref.rs:10:15
|
LL | let _ = x.longname;
| ^^^^^^^^ help: a field with a similar name exists: `long_name`
```
CC https://github.com/rust-lang/rust/issues/78374#issuecomment-719564114
Remove option_payload_ptr; redundant to offset_of
The `option_payload_ptr` intrinsic is no longer required as `offset_of` supports traversing enums (#114208). This PR removes it in order to dogfood offset_of (as suggested at https://github.com/rust-lang/rust/issues/106655#issuecomment-1790907626). However, it will not build until those changes reach beta (which I think is within the next 8 days?) so I've opened it as a draft.
When a local binding shadows a fn, point at fn def in call failure
When a local binding shadows a function that is then called, this local binding will cause an E0618 error. We now point not only at the binding definition, but also at the locally defined function of the same name.
```
error[E0618]: expected function, found `&str`
--> $DIR/issue-22468.rs:3:13
|
LL | let foo = "bar";
| --- `foo` has type `&str`
LL | let x = foo("baz");
| ^^^-------
| |
| call expression requires function
...
LL | fn foo(file: &str) -> bool {
| -------------------------- this function of the same name is available here, but it shadowed by the local binding of the same name
```
Fix#53841
Add some additional warnings for duplicated diagnostic items
This commit adds warnings if a user supplies several diagnostic options where we can only apply one of them. We explicitly warn about ignored options here. In addition a small test for these warnings is added.
r? `@compiler-errors`
For now that's the last PR to improve the warnings generated by misused `#[diagnostic::on_unimplemented]` attributes. I'm not sure what needs to be done next to move this closer to stabilization.
When a local binding shadows a function that is then called, this local
binding will cause an E0618 error. We now point not only at the binding
definition, but also at the locally defined function of the same name.
```
error[E0618]: expected function, found `&str`
--> $DIR/issue-22468.rs:3:13
|
LL | let foo = "bar";
| --- `foo` has type `&str`
LL | let x = foo("baz");
| ^^^-------
| |
| call expression requires function
...
LL | fn foo(file: &str) -> bool {
| -------------------------- this function of the same name is avalable here, but it shadowed by the local binding of the same name
```
Fix#53841
When using existing fn as module, don't claim it doesn't exist
Tweak wording of module not found in resolve, when the name exists but belongs to a non-`mod` item.
Fix#81232.
Emit smir
This adds ability to `-Zunpretty=smir` and get smir output of a Rust file, this is obliviously pretty basic compared to `mir` output but I think we could iteratively improve it, and even at this state this is useful for us.
r? ``@celinval``
Remove asmjs
Fulfills [MCP 668](https://github.com/rust-lang/compiler-team/issues/668).
`asmjs-unknown-emscripten` does not work as-specified, and lacks essential upstream support for generating asm.js, so it should not exist at all.
More detail when expecting expression but encountering bad macro argument
On nested macro invocations where the same macro fragment changes fragment type from one to the next, point at the chain of invocations and at the macro fragment definition place, explaining that the change has occurred.
Fix#71039.
```
error: expected expression, found pattern `1 + 1`
--> $DIR/trace_faulty_macros.rs:49:37
|
LL | (let $p:pat = $e:expr) => {test!(($p,$e))};
| ------- -- this is interpreted as expression, but it is expected to be pattern
| |
| this macro fragment matcher is expression
...
LL | (($p:pat, $e:pat)) => {let $p = $e;};
| ------ ^^ expected expression
| |
| this macro fragment matcher is pattern
...
LL | test!(let x = 1+1);
| ------------------
| | |
| | this is expected to be expression
| in this macro invocation
|
= note: when forwarding a matched fragment to another macro-by-example, matchers in the second macro will see an opaque AST of the fragment type, not the underlying tokens
= note: this error originates in the macro `test` (in Nightly builds, run with -Z macro-backtrace for more info)
```
I find `Compilation::and_then` hard to read. This commit removes it,
simplifying the control flow in `run_compiler`, and reducing the number
of lines of code.
In particular, `list_metadata` and `process_try_link` (renamed `rlink`)
are now only called if the relevant condition is true, rather than that
condition being checked within the function.
Currently we have an inconsistency between the "input" and "no input"
cases:
- no input: `rustc --print=sysroot -Whelp` prints the lint help.
- input: `rustc --print=sysroot -Whelp a.rs` prints the sysroot.
It makes sense to print the lint help in both cases, because that's what
happens with `--help`/`-Zhelp`/`-Chelp`.
In fact, the `describe_lints` in the "input" case happens amazingly
late, after *parsing*. This is because, with plugins, lints used to be
registered much later, when the global context was created. But #117649
moved lint registration much earlier, during session construction.
So this commit moves the `describe_lints` call to a single spot for both
for both the "input" and "no input" cases, as early as possible. This is
still not as early as `--help`/`-Zhelp`/`-Chelp`, because `-Whelp` must
wait until the session is constructed.
`rustc_driver_impl::run_compiler` currently has two
`interface::run_compiler` calls: one for the "no input" case, and one
for the normal case.
This commit merges the former into the latter, which makes the control
flow easier to read and avoids some duplication.
It also makes it clearer that the "no input" case will describe lints
before printing crate info, while the normal case does it in the reverse
order. Possibly a bug?
Yes, its type is `EarlyErrorHandler`, but there is another value of that
type later on in the function called `handler` that is initialized with
`sopts.error_format`. So `default_handler` is a better name because it
clarifies that it is initialized with `ErrorOutputType::default()`.
ignore implied bounds with placeholders
given the following code:
```rust
trait Trait {
type Ty<'a> where Self: 'a;
}
impl<T> Trait for T {
type Ty<'a> = () where Self: 'a;
}
struct Foo<T: Trait>(T)
where
for<'x> T::Ty<'x>: Sized;
```
when computing the implied bounds from `Foo<X>` we incorrectly get the bound `X: !x` from the normalization of ` for<'x> <X as Trait>::Ty::<'x>: Sized`. This is a a known bug! we shouldn't use the constraints that arise from normalization as implied bounds. See #109628.
Ignore these bounds for now. This should prevent later ICEs.
Fixes#112250Fixes#107409
new solver normalization improvements
cool beans
At the core of this PR is a `try_normalize_ty` which stops for rigid aliases by using `commit_if_ok`.
Reworks alias-relate to fully normalize both the lhs and rhs and then equate the resulting rigid (or inference) types. This fixes https://github.com/rust-lang/trait-system-refactor-initiative/issues/68 by avoiding the exponential blowup. Also supersedes #116369 by only defining opaque types if the hidden type is rigid.
I removed the stability check in `EvalCtxt::evaluate_goal` due to https://github.com/rust-lang/trait-system-refactor-initiative/issues/75. While I personally have opinions on how to fix it, that still requires further t-types/`@nikomatsakis` buy-in, so I removed that for now. Once we've decided on our approach there, we can revert this commit.
r? `@compiler-errors`
On resolve error of `[rest..]`, suggest `[rest @ ..]`
When writing a pattern to collect multiple entries of a slice in a single binding, it is easy to misremember or typo the appropriate syntax to do so, instead writing the experimental `X..` pattern syntax. When we encounter a resolve error because `X` isn't available, we suggest `X @ ..` as an alternative.
```
error[E0425]: cannot find value `rest` in this scope
--> $DIR/range-pattern-meant-to-be-slice-rest-pattern.rs:3:13
|
LL | [1, rest..] => println!("{rest:?}"),
| ^^^^ not found in this scope
|
help: if you meant to collect the rest of the slice in `rest`, use the at operator
|
LL | [1, rest @ ..] => println!("{rest:?}"),
| +
```
Fix#88404.
Misc changes to StableMIR required to Kani use case.
First, I wanted to say that I can split this review into multiple if it makes reviewing easier. I bundled them up, since I've been testing them together (See https://github.com/rust-lang/project-stable-mir/pull/51 for the set of more thorough checks).
So far, this review includes 3 commits:
1. Add more APIs and fix `Instance::body`
- Add more APIs to retrieve information about types.
- Add a few more instance resolution options. For the drop shim, we return None if the drop body is empty. Not sure it will be enough.
- Make `Instance::body()` return an Option<Body>, since not every instance might have an available body. For example, foreign instances, virtual instances, dependencies.
2. Fix a bug on MIRVisitor
- We were not iterating over all local variables due to a typo.
3. Add more SMIR internal impl and callback return value
- In cases like Kani, we will invoke the rustc_internal run command directly for now. It would be handly to be able to have a callback that can return a value.
- We also need extra methods to convert stable constructs into internal ones, so we can break down the transition into finer grain commits.
- For the internal implementation of Region, we're always returning `ReErased` for now.
It matches the type, and a noun makes more sense than a verb.
The `output_filenames` function still uses a profiling label named
`prepare_outputs`, but I think that makes sense as a verb and can be
left unchanged.
This commit adds warnings if a user supplies several diagnostic options
where we can only apply one of them. We explicitly warn about ignored
options here. In addition a small test for these warnings is added.
Subtree sync for rustc_codegen_cranelift
The main highlights this time are support for AES and SHA256 crypto intrinsics on x86_64 by lowering to inline asm.
r? `@ghost`
`@rustbot` label +A-codegen +A-cranelift +T-compiler
When writing a pattern to collect multiple entries of a slice in a
single binding, it is easy to misremember or typo the appropriate syntax
to do so, instead writing the experimental `X..` pattern syntax. When we
encounter a resolve error because `X` isn't available, we suggest
`X @ ..` as an alternative.
```
error[E0425]: cannot find value `rest` in this scope
--> $DIR/range-pattern-meant-to-be-slice-rest-pattern.rs:3:13
|
LL | [1, rest..] => println!("{rest:?}"),
| ^^^^ not found in this scope
|
help: if you meant to collect the rest of the slice in `rest`, use the at operator
|
LL | [1, rest @ ..] => println!("{rest:?}"),
| +
```
Fix#88404.
Better handle type errors involving `Self` literals
When encountering a type error involving a `Self` literal, point at the self type of the enclosing `impl` and suggest using the actual type name instead.
```
error[E0308]: mismatched types
--> $DIR/struct-path-self-type-mismatch.rs:13:9
|
LL | impl<T> Foo<T> {
| - ------ this is the type of the `Self` literal
| |
| found type parameter
LL | fn new<U>(u: U) -> Foo<U> {
| - ------ expected `Foo<U>` because of return type
| |
| expected type parameter
LL | / Self {
LL | |
LL | | inner: u
LL | |
LL | | }
| |_________^ expected `Foo<U>`, found `Foo<T>`
|
= note: expected struct `Foo<U>`
found struct `Foo<T>`
= note: a type parameter was expected, but a different one was found; you might be missing a type parameter or trait bound
= note: for more information, visit https://doc.rust-lang.org/book/ch10-02-traits.html#traits-as-parameters
help: use the type name directly
|
LL | Foo::<U> {
| ~~~~~~~~
```
Fix#76086.
This was made possible by the removal of plugin support, which
simplified lint store creation.
This simplifies the places in rustc and rustdoc that call
`describe_lints`, which are early on. The lint store is now built before
those places, so they don't have to create their own lint store for
temporary use, they can just use the main one.
Fixed the `has_body()` function operator. Before that, this function was
returning false for all shims.
Change resolve_drop_in_place() to also return an instance for empty
shims, since they may still be required for vtable construction.
Add more APIs to retrieve information about types, and add more instance
resolution options.
Make `Instance::body()` return an Option<Body>, since not every instance
might have an available body. For example, foreign instances, virtual
instances, dependencies.
In cases like Kani, we will invoke the rustc_internal run command
directly for now. It would be handly to be able to have a callback
that can return a value.
We also need extra methods to convert stable constructs into internal
ones, so we can break down the transition into finer grain commits.
Take into account implicit dereferences when suggesting fields.
```
error[E0609]: no field `longname` on type `Arc<S>`
--> $DIR/suggest-field-through-deref.rs:10:15
|
LL | let _ = x.longname;
| ^^^^^^^^ help: a field with a similar name exists: `long_name`
```
CC https://github.com/rust-lang/rust/issues/78374#issuecomment-719564114
When encountering a type error caused by the use of `Self`, suggest
using the actual type name instead.
```
error[E0308]: mismatched types
--> $DIR/struct-path-self-type-mismatch.rs:13:9
|
LL | impl<T> Foo<T> {
| - ------ this is the type of the `Self` literal
| |
| found type parameter
LL | fn new<U>(u: U) -> Foo<U> {
| - ------ expected `Foo<U>` because of return type
| |
| expected type parameter
LL | / Self {
LL | |
LL | | inner: u
LL | |
LL | | }
| |_________^ expected `Foo<U>`, found `Foo<T>`
|
= note: expected struct `Foo<U>`
found struct `Foo<T>`
= note: a type parameter was expected, but a different one was found; you might be missing a type parameter or trait bound
= note: for more information, visit https://doc.rust-lang.org/book/ch10-02-traits.html#traits-as-parameters
help: use the type name directly
|
LL | Foo::<U> {
| ~~~~~~~~
```
Fix#76086.
Ensure strings created with `const_str` get the `unnamed_addr` attribute
This function (`const_str`) is only used when we need to invent a string during codegen -- for example, for a panic message to pass when codegening some of the assert/panic/etc terminators (for stuff like divide by zero).
AFAICT all other consts, such as the user-defined ones from const eval, should already be getting this attribute (things that come from a ConstAllocation do, for example). Which means that the "unnamed" part is even more true than usual here, these aren't strings that even exist as far as the user can tell.
~~Setting this attribute allows LLVM to merge these constants, leading to significant binary size savings (much more than I would expect). On x86_64-unknown-linux-gnu, t takes a build of ripgrep (release without debug info) from 9.7MiB to 6.0MiB (a savings of over 30%!?), and a build of rustc_driver's shared object from 123MiB to 112MiB (less drastic, but still over 10% reduced).~~
~~The effect on ripgrep is substantially reduced on macOS for reasons beyond me (I may have fucked up the test), only saving around 0.2MiB, although rustc_driver is still around 10MB or smaller than it had been previously.~~
~~This raises some questions, such as "does that mean 1/3 of ripgrep was made of division by zero complaints?" I'm not sure, that may be the case. The output of `strings path/to/rg` is \~2MB smaller, so it seems like a lot of it was. Allowing these to be merged presumably also allow functions that contain them to be merged (if the addresses had semantic meaning, then it stands).~~
~~I intend to do some more analysis here, but I got this up as soon as I realized that this attribute was only missing for internal const strings, and all other ones already get it.~~
Edit: The wins are much more marginal, but there's some argument to do this for the sake of consistency.
llvm-wrapper: Pass newly added param to DIBuilder::createStaticMemberType()
This was added in https://github.com/llvm/llvm-project/pull/72234.
DW_TAG_member was the implicit default before.
The LLVM change is quite sinister since due to weakly typed ints and default params, this was still successfully compiling against LLVM but was passing the wrong parameters.
Fix depth check in ProofTreeVisitor.
The hack to cutoff overflows and cycles in the new trait solver was incorrect. We want to inspect everything with depth [0..10].
This fix exposed a previously unseen bug, which caused the compiler to ICE when invoking `trait_ref` on a non-assoc type projection. I simply added the guard in the `AmbiguityCausesVisitor`, and updated the expected output for the `auto-trait-coherence` test which now includes the extra note:
```text
|
= note: upstream crates may add a new impl of trait `std::marker::Send` for type `OpaqueType` in future versions
```
r? `@lcnr`
Add -Z llvm_module_flag
Allow adding values to the `!llvm.module.flags` metadata for a generated module. The syntax is
`-Z llvm_module_flag=<name>:<type>:<value>:<behavior>`
Currently only u32 values are supported but the type is required to be specified for forward compatibility. The `behavior` element must match one of the named LLVM metadata behaviors.viors.
This flag is expected to be perma-unstable.
finish `RegionKind` renaming
second step of https://github.com/rust-lang/types-team/issues/95
continues the work from #117876. While working on this and I encountered a bunch of further cleanup which I'll either open a tracking issue for or will do in a separate PR:
- rewrite the `RegionKind` docs, they still talk about `ReEmpty` and are generally out of date
- rename `DescriptionCtx` to `DescriptionCtxt`
- what is `CheckRegions::Bound`?
- `collect_late_bound_regions` et al
- `erase_late_bound_regions` -> `instantiate_bound_regions_with_erased`?
- `EraseEarlyRegions` visitor should be removed, feels duplicate
r? `@BoxyUwU`
Don't expect a rcvr in `print_disambiguation_help`
We don't necessarily have a receiver when we are both accidentally using the `.` operator *AND* we have more than one ambiguous method candidate.
Fixes#117728
rustc_log: provide a way to init logging based on the values, not names, of the env vars
Miri wants to affect how rustc does logging. So far this required setting environment variables before calling `rustc_driver::init_rustc_env_logger`. However, `set_var` is a function one should really [avoid calling](https://github.com/rust-lang/rust/issues/90308), so this adds the necessary APIs to rustc such that Miri can just pass it the *values* of all the log-relevant environment variables, rather than having to change the global environment.
Add richer structure for Stable MIR Projections
Resolves https://github.com/rust-lang/project-stable-mir/issues/49.
Projections in Stable MIR are currently just strings. This PR replaces that representation with a richer structure, namely projections become vectors of `ProjectionElem`s, just as in MIR. The `ProjectionElem` enum is heavily based off of the MIR `ProjectionElem`.
This PR is a draft since there are several outstanding issues to resolve, including:
- How should `UserTypeProjection`s be represented in Stable MIR? In MIR, the projections are just a vector of `ProjectionElem<(),()>`, meaning `ProjectionElem`s that don't have Local or Type arguments (for `Index`, `Field`, etc. objects). Should `UserTypeProjection`s be represented this way in Stable MIR as well? Or is there a more user-friendly representation that wouldn't drag along all the `ProjectionElem` variants that presumably can't appear?
- What is the expected behavior of a `Place`'s `ty` function? Should it resolve down the chain of projections so that something like `*_1.f` would return the type referenced by field `f`?
- Tests should be added for `UserTypeProjection`
Also shifts comments explaining why Stable MIR drops an optional variant
name field, for `Downcast` projection elements, to the `Place::stable`
function.
target_feature: make it more clear what that 'Option' means
Reading those tables in compiler/rustc_codegen_ssa/src/target_features.rs, it's not immediately clear that `None` means "stable". Let's use an enum with appropriately named variants to make that more obvious.
Build pre-coroutine-transform coroutine body on error
I was accidentally building the post-transform coroutine body, rather than the pre-transform coroutine body. There's no pinning expected here yet, and the return type isn't yet transformed into `CoroutineState`.
Fixes#117670
make `LayoutError::Cycle` carry `ErrorGuaranteed`
Addresses a FIXME, and also I think it's wise for error variants to carry their `ErrorGuaranteed` -- makes it easier to use that `ErrorGuaranteed` for creating, e.g. `TyKind::Error` and other error kinds. Splitting out from #117703.
Add CoroutineWitness to covered types in smir
Previously we accepted `CouroutineWitness` as `unreachable!` but https://github.com/rust-lang/project-stable-mir/issues/50 shows it is indeed reachable, this pr fixes that and covers `CouroutineWitness`
Custom MIR: Support cleanup blocks
Cleanup blocks are declared with `bb (cleanup) = { ... }`.
`Call` and `Drop` terminators take an additional argument describing the unwind action, which is one of the following:
* `UnwindContinue()`
* `UnwindUnreachable()`
* `UnwindTerminate(reason)`, where reason is `ReasonAbi` or `ReasonInCleanup`
* `UnwindCleanup(block)`
Also support unwind resume and unwind terminate terminators:
* `UnwindResume()`
* `UnwindTerminate(reason)`
Cleanup blocks are declared with `bb (cleanup) = { ... }`.
`Call` and `Drop` terminators take an additional argument describing the
unwind action, which is one of the following:
* `UnwindContinue()`
* `UnwindUnreachable()`
* `UnwindTerminate(reason)`, where reason is `ReasonAbi` or `ReasonInCleanup`
* `UnwindCleanup(block)`
Also support unwind resume and unwind terminate terminators:
* `UnwindResume()`
* `UnwindTerminate(reason)`
Always point at index span on index obligation failure
Use more targetted span for index obligation failures by rewriting the obligation cause span.
CC #66023
Remove incorrect transformation from RemoveZsts
Partial removal of storage statements for a local is incorrect, so a decision to optimize cannot be make independently for each statement.
Avoid the issue by performing the transformation completely or not at all.
Remove `-Zperf-stats`.
The included measurements have varied over the years. At one point there were quite a few more, but #49558 deleted a lot that were no longer used. Today there's just four, and it's a motley collection that doesn't seem particularly valuable.
I think it has been well and truly subsumed by self-profiling, which collects way more data.
r? `@wesleywiser`
Rollup of 4 pull requests
Successful merges:
- #114224 (rustc_llvm: Link to libkstat on Solaris/SPARC)
- #117695 (Reorder checks to make sure potential missing expect on Option/Result…)
- #117870 (`fn args_ref_X` to `fn args_X`)
- #117879 (tests: update check for inferred nneg on zext)
r? `@ghost`
`@rustbot` modify labels: rollup
rustc_llvm: Link to libkstat on Solaris/SPARC
getHostCPUName calls into libkstat but as of
LLVM 16.0.6 libLLVMTargetParser is not explicitly
linked against libkstat causing builds to fail
due to undefined symbols.
See also: llvm/llvm-project#64186
Compute layout with spans for better cycle errors in coroutines
Split out from #117703, this PR at least gives us a nicer span to point at when we hit a cycle error in coroutine layout cycles.
Remove `-Zkeep-hygiene-data`.
It was added way back in #28585 under the name `-Zkeep-mtwt-tables`. The justification was:
> This is so that the resolution results can be used after analysis,
> potentially for tool support.
There are no uses of significance in the code base, and various Google searches for both option names (and variants) found nothing of interest. I think this can safely be removed.
r? `@davidtwco`
`ReLateBound` -> `ReBound`
first step of https://github.com/rust-lang/types-team/issues/95
already fairly large xx
there's some future work here I intentionally did not contribute as part of this PR, from my notes:
- `DescriptionCtx` to `DescriptionCtxt`
- what is `CheckRegions::Bound`?
- `collect_late_bound_regions` et al
- `erase_late_bound_regions` -> `instantiate_bound_regions_with_erased`?
- `EraseEarlyRegions` should be removed, feels duplicate
r? `@BoxyUwU`
coverage: Avoid creating malformed macro name spans
This is a workaround for #117788. It detects a particular scenario where we would create malformed coverage spans that might cause `llvm-cov` to immediately exit with an error, preventing the user from processing coverage reports.
The patch has been kept as simple as possible so that it's trivial to backport to beta (or stable) if desired.
---
The `maybe_push_macro_name_span` method is trying to detect macro invocations, so that it can split a span into two parts just after the `!` of the invocation.
Under some circumstances (probably involving nested macros), it gets confused and produces a span that is larger than the original span, and possibly extends outside its enclosing function and even into an adjacent file.
In extreme cases, that can result in malformed coverage mappings that cause `llvm-cov` to fail. For now, we at least want to detect these egregious cases and avoid them, so that coverage reports can still be produced.
