Add `lp64e` RISC-V ABI
This PR adds support for the `lp64e` RISC-V ABI, which is the 64-bit equivalent of the `ilp32e` ABI that is already supported.
For reference, this ABI was originally added to LLVM in [this PR](https://reviews.llvm.org/D70401).
Remove region from adjustments
It's not necessary to store this region, because it's only used in THIR and MemCat/ExprUse, both of which already basically only deal with erased regions anyways.
Rollup of 4 pull requests
Successful merges:
- #132347 (Remove `ValueAnalysis` and `ValueAnalysisWrapper`.)
- #132365 (pass `RUSTC_HOST_FLAGS` at once without the for loop)
- #132366 (Do not enforce `~const` constness effects in typeck if `rustc_do_not_const_check`)
- #132376 (Annotate `input` reference tests)
r? `@ghost`
`@rustbot` modify labels: rollup
Do not enforce `~const` constness effects in typeck if `rustc_do_not_const_check`
Fixes a slight inconsistency between HIR and MIR enforcement of `~const` :D
r? `@rust-lang/project-const-traits`
Remove `ValueAnalysis` and `ValueAnalysisWrapper`.
They represent a lot of abstraction and indirection, but they're only used for `ConstAnalysis`, and apparently won't be used for any other analyses in the future. This commit inlines and removes them, which makes `ConstAnalysis` easier to read and understand.
r? `@cjgillot`
Try to point out when edition 2024 lifetime capture rules cause borrowck issues
Lifetime capture rules in 2024 are modified to capture more lifetimes, which sometimes lead to some non-local borrowck errors. This PR attempts to link these back together with a useful note pointing out the capture rule changes.
This is not a blocking concern, but I'd appreciate feedback (though, again, I'd like to stress that I don't want to block this PR on this): I'm worried about this note drowning in the sea of other diagnostics that borrowck emits. I was tempted to change the level of the note to `.span_warn` just so it would show up in a different color. Thoughts?
Fixes#130545
Opening as a draft first since it's stacked on #131183.
r? `@ghost`
They represent a lot of abstraction and indirection, but they're only
used for `ConstAnalysis`, and apparently won't be used for any other
analyses in the future. This commit inlines and removes them, which
makes `ConstAnalysis` easier to read and understand.
Make sure `type_param_predicates` resolves correctly for RPITIT
After #132194, we end up lowering the item bounds for an RPITIT in an `ItemCtxt` whose def id is the *synthetic GAT*, not the opaque type from the HIR.
This means that when we're resolving a shorthand projection like `T::Assoc`, we call the `type_param_predicates` function with the `item_def_id` of the *GAT* and not the opaque. That function operates on the HIR, and is not designed to work with the `Node::Synthetic` that gets fed for items synthesized by the compiler...
This PR reuses the trick we use elsewhere in lowering, where we intercept whether an item comes from RPITIT lowering, and forwards the query off to the correct item.
Fixes#132372
Rename `rustc_abi::Abi` to `BackendRepr`
Remove the confabulation of `rustc_abi::Abi` with what "ABI" actually means by renaming it to `BackendRepr`, and rename `Abi::Aggregate` to `BackendRepr::Memory`. The type never actually represented how things are passed, as that has to have `PassMode` considered, at minimum, but rather it just is how we represented some things to the backend. This conflation arose because LLVM, the primary backend at the time, would lower certain IR forms using certain ABIs. Even that only somewhat was true, as it broke down when one ventured significantly afield of what is described by the System V AMD64 ABI either by using different architectures, ABI-modifying IR annotations, the same architecture **with different ISA extensions enabled**, or other... unexpected delights.
Unfortunately both names are still somewhat of a misnomer right now, as people have written code for years based on this misunderstanding. Still, their original names are even moreso, and for better or worse, this backend code hasn't received as much maintenance as the rest of the compiler, lately. Actually arriving at a correct end-state will simply require us to disentangle a lot of code in order to fix, much of it pointlessly repeated in several places. Thus this is not an "actual fix", just a way to deflect further misunderstandings.
Remap impl-trait lifetimes on HIR instead of AST lowering
Current AST->HIR lowering goes out of its way to remap lifetimes for opaque types. This is complicated and leaks into upstream and downstream code.
This PR stops trying to be clever during lowering, and prefers to do this remapping during the HIR->ty lowering. The remapping computation easily fits into the bound var resolution code. Its result can be used in by `generics_of` and `hir_ty_lowering::new_opaque` to add the proper parameters and arguments.
See an example on the doc for query `opaque_captured_lifetimes`.
Based on https://github.com/rust-lang/rust/pull/129244/
Fixes https://github.com/rust-lang/rust/issues/125249
Fixes https://github.com/rust-lang/rust/issues/126850
cc `@compiler-errors` `@spastorino`
r? `@petrochenkov`
Fix AIX libc call char type from i8 to u8
There was an update to AIX `libc` default char type from `i8 -> u8`, we should reflect that on the call site to satisfy the type checker.
81f0cd3d97/src/unix/aix/mod.rs (L1)
The RFC for arbitrary self types v2 declares that we should reject
"generic" self types. This commit does so.
The definition of "generic" was unclear in the RFC, but has been
explored in
https://github.com/rust-lang/rust/issues/129147
and the conclusion is that "generic" means any `self` type which
is a type parameter defined on the method itself, or references
to such a type.
This approach was chosen because other definitions of "generic"
don't work. Specifically,
* we can't filter out generic type _arguments_, because that would
filter out Rc<Self> and all the other types of smart pointer
we want to support;
* we can't filter out all type params, because Self itself is a
type param, and because existing Rust code depends on other
type params declared on the type (as opposed to the method).
This PR decides to make a new error code for this case, instead of
reusing the existing E0307 error. This makes the code a
bit more complex, but it seems we have an opportunity to provide
specific diagnostics for this case so we should do so.
This PR filters out generic self types whether or not the
'arbitrary self types' feature is enabled. However, it's believed
that it can't have any effect on code which uses stable Rust, since
there are no stable traits which can be used to indicate a valid
generic receiver type, and thus it would have been impossible to
write code which could trigger this new error case.
It is however possible that this could break existing code which
uses either of the unstable `arbitrary_self_types` or
`receiver_trait` features. This breakage is intentional; as
we move arbitrary self types towards stabilization we don't want
to continue to support generic such types.
This PR adds lots of extra tests to arbitrary-self-from-method-substs.
Most of these are ways to trigger a "type mismatch" error which
9b82580c73/compiler/rustc_hir_typeck/src/method/confirm.rs (L519)
hopes can be minimized by filtering out generics in this way.
We remove a FIXME from confirm.rs suggesting that we make this change.
It's still possible to cause type mismatch errors, and a subsequent
PR may be able to improve diagnostics in this area, but it's harder
to cause these errors without contrived uses of the turbofish.
This is a part of the arbitrary self types v2 project,
https://github.com/rust-lang/rfcs/pull/3519https://github.com/rust-lang/rust/issues/44874
r? @wesleywiser
cg_llvm: Clean up FFI calls for operand bundles
All of these FFI functions have equivalents in the stable LLVM-C API, though `LLVMBuildCallBr` requires a temporary polyfill on LLVM 18.
This PR also creates a clear split between `OperandBundleOwned` and `OperandBundle`, and updates the internals of the owner to be a little less terrifying.
Use `token_descr` more in error messages
This is the first two commits from #124141, put into their own PR to get things rolling. Commit messages have the details.
r? ``@estebank``
cc ``@petrochenkov``
TypingMode: merge intercrate, reveal, and defining_opaque_types
This adds `TypingMode` and uses it in most places. We do not yet remove `Reveal` from `param_env`s. This and other future work as tracked in #132279 and via `FIXME`s.
Fetching the `TypingMode` of the `InferCtxt` asserts that the `TypingMode` agrees with `ParamEnv::reveal` to make sure we don't introduce any subtle bugs here. This will be unnecessary once `ParamEnv::reveal` no longer exists.
As the `TypingMode` is now a part of the query input, I've merged the coherence and non-coherence caches for the new solver. I've also enabled the local `infcx` cache during coherence by clearing the cache when forking it with a different `TypingMode`.
#### `TypingMode::from_param_env`
I am using this even in cases where I know that the `param_env` will always be `Reveal::UserFacing`. This is to make it easier to correctly refactor this code in the future, any time we use `Reveal::UserFacing` in a body while not defining its opaque types is incorrect and should use a `TypingMode` which only reveals opaques defined by that body instead, cc #124598
r? ``@compiler-errors``
Don't lint `irrefutable_let_patterns` on leading patterns if `else if` let-chains
fixes#128661
Is there any preference where the test goes? There looks to be several places it could fit.
`Formatter` currently has a `RefCell<Option<Results>>` field. This is so
the `Results` can be temporarily taken and put into a `ResultsCursor`
that is used by `BlockFormatter`, and then put back, which is messy.
This commit changes `Formatter` to have a `RefCell<ResultsCursor>` and
`BlockFormatter` to have a `&mut ResultsCursor`, which greatly
simplifies the code at the `Formatter`/`BlockFormatter` interaction
point in `Formatter::node_label`. It also means we construct a
`ResultsCursor` once per `Formatter`, instead of once per `node_label`
call.
The commit also:
- documents the reason for the `RefCell`;
- adds a `Formatter::body` method, replacing the `Formatter::body`
field.
It's no longer needed. `Engine::iterate_to_fixpoint` can be inlined into
`Analysis::iterate_to_fixpoint` and removed. The commit also renames
`engine.rs` as `results.rs`.
This is a standard pattern:
```
MyAnalysis.into_engine(tcx, body).iterate_to_fixpoint()
```
`into_engine` and `iterate_to_fixpoint` are always called in pairs, but
sometimes with a builder-style `pass_name` call between them. But a
builder-style interface is overkill here. This has been bugging me a for
a while.
This commit:
- Merges `Engine::new` and `Engine::iterate_to_fixpoint`. This removes
the need for `Engine` to have fields, leaving it as a trivial type
that the next commit will remove.
- Renames `Analysis::into_engine` as `Analysis::iterate_to_fixpoint`,
gives it an extra argument for the optional pass name, and makes it
call `Engine::iterate_to_fixpoint` instead of `Engine::new`.
This turns the pattern from above into this:
```
MyAnalysis.iterate_to_fixpoint(tcx, body, None)
```
which is shorter at every call site, and there's less plumbing required
to support it.
The initial naming of "Abi" was an awful mistake, conveying wrong ideas
about how psABIs worked and even more about what the enum meant.
It was only meant to represent the way the value would be described to
a codegen backend as it was lowered to that intermediate representation.
It was never meant to mean anything about the actual psABI handling!
The conflation is because LLVM typically will associate a certain form
with a certain ABI, but even that does not hold when the special cases
that actually exist arise, plus the IR annotations that modify the ABI.
Reframe `rustc_abi::Abi` as the `BackendRepr` of the type, and rename
`BackendRepr::Aggregate` as `BackendRepr::Memory`. Unfortunately, due to
the persistent misunderstandings, this too is now incorrect:
- Scattered ABI-relevant code is entangled with BackendRepr
- We do not always pre-compute a correct BackendRepr that reflects how
we "actually" want this value to be handled, so we leave the backend
interface to also inject various special-cases here
- In some cases `BackendRepr::Memory` is a "real" aggregate, but in
others it is in fact using memory, and in some cases it is a scalar!
Our rustc-to-backend lowering code handles this sort of thing right now.
That will eventually be addressed by lifting duplicated lowering code
to either rustc_codegen_ssa or rustc_target as appropriate.
cg_llvm: Clean up FFI calls for setting module flags
This is a combination of several inter-related changes to how module flags are set:
- Remove some unnecessary code for setting an `"LTOPostLink"` flag, which has been obsolete since LLVM 17.
- Define our own enum instead of relying on enum values defined by LLVM's unstable C++ API.
- Use safe wrapper functions to set module flags, instead of direct `unsafe` calls.
- Consistently pass pointer/length strings instead of C strings.
- Remove or shrink some `unsafe` blocks.
Remove detail from label/note that is already available in other note
Remove the "which is required by `{root_obligation}`" post-script in
"the trait `X` is not implemented for `Y`" explanation in E0277. This
information is already conveyed in the notes explaining requirements,
making it redundant while making the text (particularly in labels)
harder to read.
```
error[E0277]: the trait bound `NotCopy: Copy` is not satisfied
--> $DIR/wf-static-type.rs:10:13
|
LL | static FOO: IsCopy<Option<NotCopy>> = IsCopy { t: None };
| ^^^^^^^^^^^^^^^^^^^^^^^ the trait `Copy` is not implemented for `NotCopy`
|
= note: required for `Option<NotCopy>` to implement `Copy`
note: required by a bound in `IsCopy`
--> $DIR/wf-static-type.rs:7:17
|
LL | struct IsCopy<T:Copy> { t: T }
| ^^^^ required by this bound in `IsCopy`
```
vs the prior
```
error[E0277]: the trait bound `NotCopy: Copy` is not satisfied
--> $DIR/wf-static-type.rs:10:13
|
LL | static FOO: IsCopy<Option<NotCopy>> = IsCopy { t: None };
| ^^^^^^^^^^^^^^^^^^^^^^^ the trait `Copy` is not implemented for `NotCopy`, which is required by `Option<NotCopy>: Copy`
|
= note: required for `Option<NotCopy>` to implement `Copy`
note: required by a bound in `IsCopy`
--> $DIR/wf-static-type.rs:7:17
|
LL | struct IsCopy<T:Copy> { t: T }
| ^^^^ required by this bound in `IsCopy`
```
*Ignore first three commits from https://github.com/rust-lang/rust/pull/132086.*
Ensure that resume arg outlives region bound for coroutines
When proving that `{Coroutine}: 'region`, we must also prove that the coroutine's resume ty outlives that region as well. See the inline comment.
Fixes#132104
Remove the "which is required by `{root_obligation}`" post-script in
"the trait `X` is not implemented for `Y`" explanation in E0277. This
information is already conveyed in the notes explaining requirements,
making it redundant while making the text (particularly in labels)
harder to read.
```
error[E0277]: the trait bound `NotCopy: Copy` is not satisfied
--> $DIR/wf-static-type.rs:10:13
|
LL | static FOO: IsCopy<Option<NotCopy>> = IsCopy { t: None };
| ^^^^^^^^^^^^^^^^^^^^^^^ the trait `Copy` is not implemented for `NotCopy`
|
= note: required for `Option<NotCopy>` to implement `Copy`
note: required by a bound in `IsCopy`
--> $DIR/wf-static-type.rs:7:17
|
LL | struct IsCopy<T:Copy> { t: T }
| ^^^^ required by this bound in `IsCopy`
```
vs the prior
```
error[E0277]: the trait bound `NotCopy: Copy` is not satisfied
--> $DIR/wf-static-type.rs:10:13
|
LL | static FOO: IsCopy<Option<NotCopy>> = IsCopy { t: None };
| ^^^^^^^^^^^^^^^^^^^^^^^ the trait `Copy` is not implemented for `NotCopy`, which is required by `Option<NotCopy>: Copy`
|
= note: required for `Option<NotCopy>` to implement `Copy`
note: required by a bound in `IsCopy`
--> $DIR/wf-static-type.rs:7:17
|
LL | struct IsCopy<T:Copy> { t: T }
| ^^^^ required by this bound in `IsCopy`
```
- Don't rely on enum values defined by LLVM's C++ API
- Use safe wrapper functions instead of direct `unsafe` calls
- Consistently pass pointer/length strings instead of C strings
correct LLVMRustCreateThinLTOData arg types
`LLVMRustCreateThinLTOData` defined in rust as
```rust
pub fn LLVMRustCreateThinLTOData(
Modules: *const ThinLTOModule,
NumModules: c_uint,
PreservedSymbols: *const *const c_char,
PreservedSymbolsLen: c_uint,
) -> Option<&'static mut ThinLTOData>;
```
but in cpp as
```cpp
extern "C" LLVMRustThinLTOData *
LLVMRustCreateThinLTOData(LLVMRustThinLTOModule *modules, int num_modules,
const char **preserved_symbols, int num_symbols) {
```
(note `c_unit` vs `int` types). Let it be actually `size_t`.
Also fixes return type of `LLVMRustDIBuilderCreateOpLLVMFragment` to uint64_t as other similar functions around, which should be correct, i assume.
Collect item bounds for RPITITs from trait where clauses just like associated types
We collect item bounds from trait where clauses for *associated types*, i.e. this:
```rust
trait Foo
where
Self::Assoc: Send
{
type Assoc;
}
```
Becomes this:
```rust
trait Foo {
type Assoc: Send;
}
```
Today, with RPITITs/AFIT and return-type notation, we don't do that, i.e.:
```rust
trait Foo where Self::method(..): Send {
fn method() -> impl Sized;
}
fn is_send(_: impl Send) {}
fn test<T: Foo>() {
is_send(T::method());
}
```
...which fails on nightly today.
Turns out it's super easy to fix this, and we just need to use the `associated_type_bounds` lowering function in `explicit_item_bounds_with_filter`, which has that logic baked in.
Hack out effects support for old solver
Opening this for vibes ✨
Turns out that a basic, somewhat incomplete implementation of host effects is achievable in the old trait solver pretty easily. This should be sufficient for us to use in the standard library itself.
Regarding incompleteness, maybe we should always treat host predicates as ambiguous in intercrate mode (at least in the old solver) to avoid any worries about accidental impl overlap or something.
r? ```@lcnr``` cc ```@fee1-dead```
Lint against getting pointers from immediately dropped temporaries
Fixes#123613
## Changes:
1. New lint: `dangling_pointers_from_temporaries`. Is a generalization of `temporary_cstring_as_ptr` for more types and more ways to get a temporary.
2. `temporary_cstring_as_ptr` is removed and marked as renamed to `dangling_pointers_from_temporaries`.
3. `clippy::temporary_cstring_as_ptr` is marked as renamed to `dangling_pointers_from_temporaries`.
4. Fixed a false positive[^fp] for when the pointer is not actually dangling because of lifetime extension for function/method call arguments.
5. `core::cell::Cell` is now `rustc_diagnostic_item = "Cell"`
## Questions:
- [ ] Instead of manually checking for a list of known methods and diagnostic items, maybe add some sort of annotation to those methods in library and check for the presence of that annotation? https://github.com/rust-lang/rust/pull/128985#issuecomment-2318714312
## Known limitations:
### False negatives[^fn]:
See the comments in `compiler/rustc_lint/src/dangling.rs`
1. Method calls that are not checked for:
- `temporary_unsafe_cell.get()`
- `temporary_sync_unsafe_cell.get()`
2. Ways to get a temporary that are not recognized:
- `owning_temporary.field`
- `owning_temporary[index]`
3. No checks for ref-to-ptr conversions:
- `&raw [mut] temporary`
- `&temporary as *(const|mut) _`
- `ptr::from_ref(&temporary)` and friends
[^fn]: lint **should** be emitted, but **is not**
[^fp]: lint **should not** be emitted, but **is**
cg_llvm: Use a type-safe helper to cast `&str` and `&[u8]` to `*const c_char`
In `rustc_codegen_llvm` there are many uses of `.as_ptr().cast()` to convert a string or byte-slice to `*const c_char`, which then gets passed through FFI.
This works, but is fragile, because there's nothing constraining the pointer cast to actually be from `u8` to `c_char`. If the original value changes to something else that has an `as_ptr` method, or the context changes to expect something other than `c_char`, the cast will silently do the wrong thing.
By making the cast more explicit via a helper method, we can be sure that it will either perform the intended cast, or fail at compile time.
Add `LayoutS::is_uninhabited` and use it
Use accessors for the things that accessors are good at: reducing everyone's need to be nosy and peek at the internals of every data structure.
compiler: Add rustc_abi dependence to the compiler
Depend on rustc_abi in compiler crates that use it indirectly but have not yet taken on that dependency, and are not *significantly* entangled in my other PRs. This leaves an "excise rustc_target" step after the dust settles.
Lower AST node id only once
Fixes#96346.
I basically followed the given instructions except the inline part.
`lower_jump_destination` can't reuse local existing `HirId` due to unknown name resolution result so I created an additional mapping for labels.
r? ```@cjgillot```
Rollup of 4 pull requests
Successful merges:
- #131391 (Stabilize `isqrt` feature)
- #132248 (rustc_transmute: Directly use types from rustc_abi)
- #132252 (compiler: rename LayoutS to LayoutData)
- #132253 (Known-bug test for `keyword_idents` lint not propagating to other files)
r? `@ghost`
`@rustbot` modify labels: rollup
compiler: rename LayoutS to LayoutData
Bid `LayoutS` goodbye because it looks like a typo.
`LayoutS` is the last of the types that use the "`{TypeName}` is the interned type, `{TypeName}S` is the backing data that is interned" convention. This is pretty confusing to those not intimately familiar with the history of rustc's names for its types over time, and doubly so now that there are no other examples in the tree. Abolish this convention.
fix various linker warnings
separated out from https://github.com/rust-lang/rust/pull/119286; this doesn't have anything user-facing, i just want to land these changes so i can stop rebasing them.
r? `@bjorn3`
Remove `ObligationCause::span()` method
I think it's an incredibly confusing footgun to expose both `obligation_cause.span` and `obligation_cause.span()`. Especially because `ObligationCause::span()` (the method) seems to just be hacking around a single quirk in the way we set up obligation causes for match arms.
First commit removes the need for that hack, with only one diagnostic span changing (but IMO not really getting worse -- I'd argue that it was already confusing).
Depend on rustc_abi in compiler crates that use it indirectly but have
not yet taken on that dependency, and are not entangled in my other PRs.
This leaves an "excise rustc_target" step after the dust settles.
Much like the previous commit.
I think the removal of "the token" in each message is fine here. There
are many more error messages that mention tokens without saying "the
token" than those that do say it.
By using `token_descr`, as is done for many other errors, we can get
slightly better descriptions in error messages, e.g.
"macro expansion ignores token `let` and any following" becomes
"macro expansion ignores keyword `let` and any tokens following".
This will be more important once invisible delimiters start being
mentioned in error messages -- without this commit, that leads to error
messages such as "error at ``" because invisible delimiters are
pretty printed as an empty string.
Clean up some comments on lint implementation
This updates some doc comments that have gotten very out of date. Some of these macros were removed or renamed in #57726 and #104863 and others. Manual emitting of lints was significantly reworked when the `Diagnostic` infrastructure was added.
Rather than try to replicate the high-level documentation, I added pointers to the rustc-dev-guide.
I linkified some types so that if they are renamed/removed without updating the docs, it will break CI.
Cleanup: Move an impl-Trait check from AST validation to AST lowering
Namely the one that rejects `impl Trait` in qself types and non-final path segments.
There's no good reason to perform this during AST validation.
We have better infrastructure in place in the AST lowerer (`ImplTraitContext`).
This shaves off a lot of code.
We now lower `impl Trait` in bad positions to `{type error}` which allows us to
remove a special case from HIR ty lowering.
Coincidentally fixes#126725. Well, it only *masks* it by passing `{type error}` to HIR analysis instead of a "bad" opaque. I was able to find a new reproducer for it. See the issue.
Simplify param handling in `resolve_bound_vars`
I always found the flow of the `ResolvedArg` constructors to be a bit confusing; turns out they're also kinda redundantly passing around their data, too.
Also, deduplicate some code handling early-bound var to late-bound var conversion between return type notation's two styles: `where <T as Trait>::method(..): Bound` and `where T: Trait<method(..): Bound>`.
Rename macro `SmartPointer` to `CoercePointee`
As per resolution #129104 we will rename the macro to better reflect the technical specification of the feature and clarify the communication.
- `SmartPointer` is renamed to `CoerceReferent`
- `#[pointee]` attribute is renamed to `#[referent]`
- `#![feature(derive_smart_pointer)]` gate is renamed to `#![feature(derive_coerce_referent)]`.
- Any mention of `SmartPointer` in the file names are renamed accordingly.
r? `@compiler-errors`
cc `@nikomatsakis` `@Darksonn`
coverage: Don't rely on the custom traversal to find enclosing loops
This opens up the possibility of modifying or removing the custom graph traversal used in coverage counter creation, without losing access to the heuristics that care about a node's enclosing loops.
Actually changing the traversal is left for future work, because this PR on its own doesn't change the emitted coverage mappings at all.
Rollup of 4 pull requests
Successful merges:
- #132123 (allow type-based search on foreign functions)
- #132183 (Fix code HTML items making big blocks if too long)
- #132192 (expand: Stop using artificial `ast::Item` for macros loaded from metadata)
- #132205 (docs: Correctly link riscv32e from platform-support.md)
r? `@ghost`
`@rustbot` modify labels: rollup
expand: Stop using artificial `ast::Item` for macros loaded from metadata
You don't need a full `Item` for that, and not using a piece of AST helps with https://github.com/rust-lang/rust/pull/131808.
Replace some LLVMRust wrappers with calls to the LLVM C API
This PR removes the LLVMRust wrapper functions for getting/setting linkage and visibility, and replaces them with direct calls to the corresponding functions in LLVM's C API.
To make this convenient and sound, two pieces of supporting code have also been added:
- A simple proc-macro that derives `TryFrom<u32>` for fieldless enums
- A wrapper type for C enum values returned by LLVM functions, to ensure soundness if LLVM returns an enum value we don't know about
In a few places, the use of safe wrapper functions means that an `unsafe` block is no longer needed, so the affected code has changed its indentation level.
rustc_target: Add pauth-lr aarch64 target feature
Add the pauth-lr target feature, corresponding to aarch64 FEAT_PAuth_LR. This feature has been added in LLVM 19.
It is currently not supported by the Linux hwcap and so we cannot add runtime feature detection for it at this time.
r? `@Amanieu`
Downgrade `untranslatable_diagnostic` and `diagnostic_outside_of_impl` to `allow`
Current implementation of translatable diagnostics infrastructure unfortunately causes some friction for compiler contributors. While we don't have a redesign that causes less friction in place, let's downgrade the internal `untranslatable_diagnostic` and `diagnostic_outside_of_impl` lints so we don't indicate to contributors that they *have* to use the current translation infra.
I purposefully left `#[allow(untranslatable_diagnostic)]` and `#[allow(diagnostic_outside_of_impl)]` instances untouched because that seems like unnecessary additional churn.
See <https://github.com/rust-lang/rust/issues/132181> for context.
r? `@davidtwco` (or wg-diagnostics/compiler)
(Big performance change) Do not run lints that cannot emit
Before this change, adding a lint was a difficult matter because it always had some overhead involved. This was because all lints would run, no matter their default level, or if the user had `#![allow]`ed them. This PR changes that. This change would improve both the Rust lint infrastructure and Clippy, but Clippy will see the most benefit, as it has about 900 registered lints (and growing!)
So yeah, with this little patch we filter all lints pre-linting, and remove any lint that is either:
- Manually `#![allow]`ed in the whole crate,
- Allowed in the command line, or
- Not manually enabled with `#[warn]` or similar, and its default level is `Allow`
As some lints **need** to run, this PR also adds **loadbearing lints**. On a lint declaration, you can use the ``@eval_always` = true` marker to label it as loadbearing. A loadbearing lint will never be filtered (it will always run)
Fixes#106983
Deny calls to non-`#[const_trait]` methods in MIR constck
This is a (potentially temporary) fix that closes off the mismatch in assumptions between MIR constck and typeck which does the const traits checking. Before this PR, MIR constck assumed that typeck correctly handled all calls to trait methods in const contexts if effects is enabled. That is not true because typeck only correctly handles callees that are const. For non-const callees (such as methods in a non-const_trait), typeck had never created an error.
45089ec19e/compiler/rustc_hir_typeck/src/callee.rs (L876-L877)
I called this potentially temporary because the const checks could be moved to HIR entirely. Alongside the recent refactor in const stability checks where that component could be placed would need more discussion. (cc ```@compiler-errors``` ```@RalfJung)```
Tests are updated, mainly due to traits not being const in core, so tests that call them correctly error.
This fixes https://github.com/rust-lang/project-const-traits/issues/12.
coverage: Consolidate creation of covmap/covfun records
This code for creating covmap/covfun records during codegen was split across multiple functions and files for dubious historical reasons. Having it all in one place makes it easier to follow.
This PR also includes two semi-related cleanups:
- Getting the codegen context's `coverage_cx` state is made infallible, since it should always exist when running the code paths that need it.
- The value of `covfun_section_name` is saved in the codegen context, since it never changes at runtime, and the code that needs it has access to the context anyway.
---
Background: Coverage instrumentation generates two kinds of metadata that are embedded in the final binary. There is per-CGU information that goes in the `__llvm_covmap` linker section, and per-function information that goes in the `__llvm_covfun` section (except on Windows, where slightly different section names are used).
Effects cleanup
- removed extra bits from predicates queries that are no longer needed in the new system
- removed the need for `non_erasable_generics` to take in tcx and DefId, removed unused arguments in callers
r? compiler-errors
Use `Enabled{Lang,Lib}Feature` instead of n-tuples
Instead of passing around e.g. `(gate_name, attr_span, stable_since)` 3-tuples for enabled lang features or `(gate_name, attr_span)` 2-tuples for enabled lib features, use `Enabled{Lang,Lib}Feature` structs with named fields.
Also did some minor code-golfing of involved iterator chains to hopefully make them easier to follow.
Follow-up to https://github.com/rust-lang/rust/pull/132098#issuecomment-2434523431 cc `@RalfJung.`
- removed extra bits from predicates queries that are no longer needed in the new system
- removed the need for `non_erasable_generics` to take in tcx and DefId, removed unused arguments in callers
Const stability checks v2
The const stability system has served us well ever since `const fn` were first stabilized. It's main feature is that it enforces *recursive* validity -- a stable const fn cannot internally make use of unstable const features without an explicit marker in the form of `#[rustc_allow_const_fn_unstable]`. This is done to make sure that we don't accidentally expose unstable const features on stable in a way that would be hard to take back. As part of this, it is enforced that a `#[rustc_const_stable]` can only call `#[rustc_const_stable]` functions. However, some problems have been coming up with increased usage:
- It is baffling that we have to mark private or even unstable functions as `#[rustc_const_stable]` when they are used as helpers in regular stable `const fn`, and often people will rather add `#[rustc_allow_const_fn_unstable]` instead which was not our intention.
- The system has several gaping holes: a private `const fn` without stability attributes whose inherited stability (walking up parent modules) is `#[stable]` is allowed to call *arbitrary* unstable const operations, but can itself be called from stable `const fn`. Similarly, `#[allow_internal_unstable]` on a macro completely bypasses the recursive nature of the check.
Fundamentally, the problem is that we have *three* disjoint categories of functions, and not enough attributes to distinguish them:
1. const-stable functions
2. private/unstable functions that are meant to be callable from const-stable functions
3. functions that can make use of unstable const features
Functions in the first two categories cannot use unstable const features and they can only call functions from the first two categories.
This PR implements the following system:
- `#[rustc_const_stable]` puts functions in the first category. It may only be applied to `#[stable]` functions.
- `#[rustc_const_unstable]` by default puts functions in the third category. The new attribute `#[rustc_const_stable_indirect]` can be added to such a function to move it into the second category.
- `const fn` without a const stability marker are in the second category if they are still unstable. They automatically inherit the feature gate for regular calls, it can now also be used for const-calls.
Also, all the holes mentioned above have been closed. There's still one potential hole that is hard to avoid, which is when MIR building automatically inserts calls to a particular function in stable functions -- which happens in the panic machinery. Those need to be manually marked `#[rustc_const_stable_indirect]` to be sure they follow recursive const stability. But that's a fairly rare and special case so IMO it's fine.
The net effect of this is that a `#[unstable]` or unmarked function can be constified simply by marking it as `const fn`, and it will then be const-callable from stable `const fn` and subject to recursive const stability requirements. If it is publicly reachable (which implies it cannot be unmarked), it will be const-unstable under the same feature gate. Only if the function ever becomes `#[stable]` does it need a `#[rustc_const_unstable]` or `#[rustc_const_stable]` marker to decide if this should also imply const-stability.
Adding `#[rustc_const_unstable]` is only needed for (a) functions that need to use unstable const lang features (including intrinsics), or (b) `#[stable]` functions that are not yet intended to be const-stable. Adding `#[rustc_const_stable]` is only needed for functions that are actually meant to be directly callable from stable const code. `#[rustc_const_stable_indirect]` is used to mark intrinsics as const-callable and for `#[rustc_const_unstable]` functions that are actually called from other, exposed-on-stable `const fn`. No other attributes are required.
Also see the updated dev-guide at https://github.com/rust-lang/rustc-dev-guide/pull/2098.
I think in the future we may want to tweak this further, so that in the hopefully common case where a public function's const-stability just exactly mirrors its regular stability, we never have to add any attribute. But right now, once the function is stable this requires `#[rustc_const_stable]`.
### Open question
There is one point I could see we might want to do differently, and that is putting `#[rustc_const_unstable]` functions (but not intrinsics) in category 2 by default, and requiring an extra attribute for `#[rustc_const_not_exposed_on_stable]` or so. This would require a bunch of extra annotations, but would have the advantage that turning a `#[rustc_const_unstable]` into `#[rustc_const_stable]` will never change the way the function is const-checked. Currently, we often discover in the const stabilization PR that a function needs some other unstable const things, and then we rush to quickly deal with that. In this alternative universe, we'd work towards getting rid of the `rustc_const_not_exposed_on_stable` before stabilization, and once that is done stabilization becomes a trivial matter. `#[rustc_const_stable_indirect]` would then only be used for intrinsics.
I think I like this idea, but might want to do it in a follow-up PR, as it will need a whole bunch of annotations in the standard library. Also, we probably want to convert all const intrinsics to the "new" form (`#[rustc_intrinsic]` instead of an `extern` block) before doing this to avoid having to deal with two different ways of declaring intrinsics.
Cc `@rust-lang/wg-const-eval` `@rust-lang/libs-api`
Part of https://github.com/rust-lang/rust/issues/129815 (but not finished since this is not yet sufficient to safely let us expose `const fn` from hashbrown)
Fixes https://github.com/rust-lang/rust/issues/131073 by making it so that const-stable functions are always stable
try-job: test-various
Then we can rename the _raw functions to drop their suffix, and instead
explicitly use is_stable_const_fn for the few cases where that is really what
you want.
Pass Ident by reference in ast Visitor
`MutVisitor`'s version of `visit_ident` passes around `&Ident`, but `Visitor` copies `Ident`. This PR changes that
r? `@petrochenkov`
related to #128974
Fundamentally, we have *three* disjoint categories of functions:
1. const-stable functions
2. private/unstable functions that are meant to be callable from const-stable functions
3. functions that can make use of unstable const features
This PR implements the following system:
- `#[rustc_const_stable]` puts functions in the first category. It may only be applied to `#[stable]` functions.
- `#[rustc_const_unstable]` by default puts functions in the third category. The new attribute `#[rustc_const_stable_indirect]` can be added to such a function to move it into the second category.
- `const fn` without a const stability marker are in the second category if they are still unstable. They automatically inherit the feature gate for regular calls, it can now also be used for const-calls.
Also, several holes in recursive const stability checking are being closed.
There's still one potential hole that is hard to avoid, which is when MIR
building automatically inserts calls to a particular function in stable
functions -- which happens in the panic machinery. Those need to *not* be
`rustc_const_unstable` (or manually get a `rustc_const_stable_indirect`) to be
sure they follow recursive const stability. But that's a fairly rare and special
case so IMO it's fine.
The net effect of this is that a `#[unstable]` or unmarked function can be
constified simply by marking it as `const fn`, and it will then be
const-callable from stable `const fn` and subject to recursive const stability
requirements. If it is publicly reachable (which implies it cannot be unmarked),
it will be const-unstable under the same feature gate. Only if the function ever
becomes `#[stable]` does it need a `#[rustc_const_unstable]` or
`#[rustc_const_stable]` marker to decide if this should also imply
const-stability.
Adding `#[rustc_const_unstable]` is only needed for (a) functions that need to
use unstable const lang features (including intrinsics), or (b) `#[stable]`
functions that are not yet intended to be const-stable. Adding
`#[rustc_const_stable]` is only needed for functions that are actually meant to
be directly callable from stable const code. `#[rustc_const_stable_indirect]` is
used to mark intrinsics as const-callable and for `#[rustc_const_unstable]`
functions that are actually called from other, exposed-on-stable `const fn`. No
other attributes are required.
Because `note_obligation_cause_code` is recursive, if multiple types are too
long to print to the terminal, a `long_ty_file` will be created. Before, one was
created *per recursion*. Now, it is passed in so it gets printed only once.
Part of #132013.
When printing
```
= help: the trait `chumsky::private::ParserSealed<'_, &'a str, ((), ()), chumsky::extra::Full<EmptyErr, (), ()>>` is implemented for `Then<Ignored<chumsky::combinator::Filter<chumsky::primitive::Any<&str, chumsky::extra::Full<EmptyErr, (), ()>>, {closure@src/main.rs:9:17: 9:27}>, char>, chumsky::combinator::Map<impl CSTParser<'a, O>, O, {closure@src/main.rs:11:24: 11:27}>, (), (), chumsky::extra::Full<EmptyErr, (), ()>>`
= help: for that trait implementation, expected `((), ())`, found `()`
```
Highlight only the `expected` and `found` types, instead of the full type in the first `help`.
Emit future-incompatibility lint when calling/declaring functions with vectors that require missing target feature
On some architectures, vector types may have a different ABI depending on whether the relevant target features are enabled. (The ABI when the feature is disabled is often not specified, but LLVM implements some de-facto ABI.)
As discussed in https://github.com/rust-lang/lang-team/issues/235, this turns out to very easily lead to unsound code.
This commit makes it a post-monomorphization error to declare or call functions using those vector types in a context in which the corresponding target features are disabled, if using an ABI for which the difference is relevant. This ensures that these functions are always called with a consistent ABI.
See the [nomination comment](https://github.com/rust-lang/rust/pull/127731#issuecomment-2288558187) for more discussion.
r? RalfJung
Part of https://github.com/rust-lang/rust/issues/116558
On some architectures, vector types may have a different ABI when
relevant target features are enabled.
As discussed in https://github.com/rust-lang/lang-team/issues/235, this
turns out to very easily lead to unsound code.
This commit makes it an error to declare or call functions using those
vector types in a context in which the corresponding target features are
disabled, if using an ABI for which the difference is relevant.
coverage: Emit LLVM intrinsics using the normal helper method
Codegen already has convenient ways to declare and emit LLVM intrinsics, so there's no need for coverage instrumentation to jump through hoops to emit them manually.
Add support for `~const` item bounds
Supports the only missing capability of `~const` associated types that I can think of now (this is obviously excluding `~const` opaques, which I see as an extension to this; I'll probably do that next).
r? ``@lcnr`` mostly b/c it changes candidate assembly, or reassign
cc ``@fee1-dead``
Adding an extra `OnceCell` to `CrateCoverageContext` is much nicer than trying
to thread this string through multiple layers of function calls that already
have access to the context.
Rollup of 6 pull requests
Successful merges:
- #131851 ([musl] use posix_spawn if a directory change was requested)
- #132048 (AIX: use /dev/urandom for random implementation )
- #132093 (compiletest: suppress Windows Error Reporting (WER) for `run-make` tests)
- #132101 (Avoid using imports in thread_local_inner! in static)
- #132113 (Provide a default impl for Pattern::as_utf8_pattern)
- #132115 (rustdoc: Extend fake_variadic to "wrapped" tuples)
r? `@ghost`
`@rustbot` modify labels: rollup
Remove visit_expr_post from ast Visitor
`visit_expr_post` is only present in the immutable version of ast Visitors and its default implementation is a noop.
Given that its only implementer is on `rustc_lint/src/early.rs` and its name follows the same naming convention as some other lints (`_post`), it seems that `visit_expr_post` being in `Visitor` was a little mistake.
r? `@petrochenkov`
related to #128974
Stabilize shorter-tail-lifetimes
Close#131445
Tracked by #123739
We found a test case `tests/ui/drop/drop_order.rs` that had not been covered by the change. The test fixture is fixed now with the correct expectation.
Represent trait constness as a distinct predicate
cc `@rust-lang/project-const-traits`
r? `@ghost` for now
Also mirrored everything that is written below on this hackmd here: https://hackmd.io/`@compiler-errors/r12zoixg1l`
# Tl;dr:
* This PR removes the bulk of the old effect desugaring.
* This PR reimplements most of the effect desugaring as a new predicate and set of a couple queries. I believe it majorly simplifies the implementation and allows us to move forward more easily on its implementation.
I'm putting this up both as a request for comments and a vibe-check, but also as a legitimate implementation that I'd like to see land (though no rush of course on that last part).
## Background
### Early days
Once upon a time, we represented trait constness in the param-env and in `TraitPredicate`. This was very difficult to implement correctly; it had bugs and was also incomplete; I don't think this was anyone's fault though, it was just the limit of experimental knowledge we had at that point.
Dealing with `~const` within predicates themselves meant dealing with constness all throughout the trait solver. This was difficult to keep track of, and afaict was not handled well with all the corners of candidate assembly.
Specifically, we had to (in various places) remap constness according to the param-env constness:
574b64a97f/compiler/rustc_trait_selection/src/traits/select/mod.rs (L1498)
This was annoying and manual and also error prone.
### Beginning of the effects desugaring
Later on, #113210 reimplemented a new desugaring for const traits via a `<const HOST: bool>` predicate. This essentially "reified" the const checking and separated it from any of the remapping or separate tracking in param-envs. For example, if I was in a const-if-const environment, but I wanted to call a trait that was non-const, this reification would turn the constness mismatch into a simple *type* mismatch of the effect parameter.
While this was a monumental step towards straightening out const trait checking in the trait system, it had its own issues, since that meant that the constness of a trait (or any item within it, like an associated type) was *early-bound*. This essentially meant that `<T as Trait>::Assoc` was *distinct* from `<T as ~const Trait>::Assoc`, which was bad.
### Associated-type bound based effects desugaring
After this, #120639 implemented a new effects desugaring. This used an associated type to more clearly represent the fact that the constness is not an input parameter of a trait, but a property that could be computed of a impl. The write-up linked in that PR explains it better than I could.
However, I feel like it really reached the limits of what can comfortably be expressed in terms of associated type and trait calculus. Also, `<const HOST: bool>` remains a synthetic const parameter, which is observable in nested items like RPITs and closures, and comes with tons of its own hacks in the astconv and middle layer.
For example, there are pieces of unintuitive code that are needed to represent semantics like elaboration, and eventually will be needed to make error reporting intuitive, and hopefully in the future assist us in implementing built-in traits (eventually we'll want something like `~const Fn` trait bounds!).
elaboration hack: 8069f8d17a/compiler/rustc_type_ir/src/elaborate.rs (L133-L195)
trait bound remapping hack for diagnostics: 8069f8d17a/compiler/rustc_trait_selection/src/error_reporting/traits/fulfillment_errors.rs (L2370-L2413)
I want to be clear that I don't think this is a issue of implementation quality or anything like that; I think it's simply a very clear sign that we're using types and traits in a way that they're not fundamentally supposed to be used, especially given that constness deserves to be represented as a first-class concept.
### What now?
This PR implements a new desugaring for const traits. Specifically, it introduces a `HostEffect` predicate to represent the obligation an impl is const, rather than using associated type bounds and the compat trait that exists for effects today.
### `HostEffect` predicate
A `HostEffect` clause has two parts -- the `TraitRef` we're trying to prove, and a `HostPolarity::{Maybe, Const}`.
`HostPolarity::Const` corresponds to `T: const Trait` bounds, which must *always* be proven as const, and which can be written in any context. These are lowered directly into the predicates of an item, since they're not "context-specific".
On the other hand, `HostPolarity::Maybe` corresponds to `T: ~const Trait` bounds which must only exist in a conditionally-const context like a method in a `#[const_trait]`, or a `const fn` free function. We do not lower these immediately into the predicates of an item; instead, we collect them into a new query called the **`const_conditions`**. These are the set of trait refs that we need to prove have const implementations for an item to be const.
Notably, they're represented as bare (poly) trait refs because they are meant to be paired back together with a `HostPolarity` when they're being registered in typeck (see next section).
For example, given:
```rust
const fn foo<T: ~const A + const B>() {}
```
`foo`'s const conditions would contain `T: A`, but not `T: B`. On the flip side, foo's predicates (`predicates_of`) query would contain `HostEffect(T: B, HostPolarity::Const)` but not `HostEffect(T: A, HostPolarity::Maybe)` since we don't need to prove that predicate in a non-const environment (and it's not even the right predicate to prove in an unconditionally const environment).
### Type checking const bodies
When type checking bodies in HIR, when we encounter a call expression, we additionally register the callee item's const conditions with the `HostPolarity` from the body we're typechecking (`Const` for unconditionally const things like `const`/`static` items, and `Maybe` for conditionally const things like const fns; and we don't register `HostPolarity` predicates for non-const bodies).
When type-checking a conditionally const body, we augment its param-env with `HostEffect(..., Maybe)` predicates.
### Checking that const impls are WF
We extend the logic in `compare_method_predicate_entailment` to also check the const-conditions of the impl method, to make sure that we error for:
```rust
#[const_trait] Bar {}
#[const_trait] trait Foo {
fn method<T: Bar>();
}
impl Foo for () {
fn method<T: ~const Bar>() {} // stronger assumption!
}
```
We also extend the WF check for impls to register the const conditions of the trait that is being implemented. This is to make sure we error for:
```rust
#[const_trait] trait Bar {}
#[const_trait] trait Foo<T> where T: ~const Bar {}
impl<T> const Foo<T> for () {}
//~^ `T: ~const Bar` is missing!
```
### Proving a `HostEffect` predicate
We have several ways of proving a `HostEffect` predicate:
1. Matching a `HostEffect` predicate from the param-env
2. From an impl - we do impl selection very similar to confirming a trait goal, except we filter for only const impls, and we additionally register the impl's const conditions (i.e. the impl's `~const` where clauses).
Later I expect that we will add more built-in implementations for things like `Fn`.
## What next?
After this PR, I'd like to split out the work more so it can proceed in parallel and probably amongst others that are not me.
* Register `HostEffect` goal for places in HIR typeck that correspond to call terminators, like autoderef.
* Make traits in libstd const again.
* Probably need to impl host effect preds in old solver.
* Implement built-in `HostEffect` rules for traits like `Fn`.
* Rip out const checking from MIR altogether.
## So what?
This ends up being super convenient basically everywhere in the compiler. Due to the design of the new trait solver, we end up having an almost parallel structure to the existing trait and projection predicates for assembling `HostEffect` predicates; adding new candidates and especially new built-in implementations is now basically trivial, and it's quite straightforward to understand the confirmation logic for these predicates.
Same with diagnostics reporting; since we have predicates which represent the obligation to prove an impl is const, we can simplify and make these diagnostics richer without having to write a ton of logic to intercept and rewrite the existing `Compat` trait errors.
Finally, it gives us a much more straightforward path for supporting the const effect on the old trait solver. I'm personally quite passionate about getting const trait support into the hands of users without having to wait until the new solver lands[^1], so I think after this PR lands we can begin to gauge how difficult it would be to implement constness in the old trait solver too. This PR will not do this yet.
[^1]: Though this is not a prerequisite or by any means the only justification for this PR.
Replace an FTP link in comments with an equivalent HTTPS link
Modern browsers and editors often don't support following FTP links, so using an ordinary web link gives the same result in a more convenient way.
