interp/validity: enforce Scalar::Initialized
This is a follow-up to https://github.com/rust-lang/rust/pull/94527, to also account for the new kind of `Scalar` layout inside the validity checker.
r? `@oli-obk`
enhance `ConstGoto` mir-opt by moving up `StorageDead` statements
From the `FIXME` in the implementation of `ConstGoto` miropt. We can move `StorageDead` statements up to the predecessor. This can expand the scope of application of this opt.
interp: pass TyCtxt to Machine methods that do not take InterpCx
This just seems like something you might need, so let's consistently have it.
One day we might have to add `ParamEnv` as well, though that seems less likely (and in Miri you can always use `reveal_all` anyway). It might make sense to have a type that packages `TyCtxt` and `ParamEnv`, this pairing occurs quite frequently in rustc...
r? `@oli-obk`
Currently it's called in `parse_tt` every time a match rule is invoked.
This commit moves it so it's called instead once per match rule, in
`compile_declarative_macro. This is a performance win.
The commit also moves `compute_locs` out of `TtParser`, because there's
no longer any reason for it to be in there.
Let CTFE to handle partially uninitialized unions without marking the entire value as uninitialized.
follow up to #94411
To fix https://github.com/rust-lang/rust/issues/69488 and by extension fix https://github.com/rust-lang/rust/issues/94371, we should stop treating types like `MaybeUninit<usize>` as something that the `Scalar` type in the interpreter engine can represent. So we add a new field to `abi::Primitive` that records whether the primitive is nested in a union
cc `@RalfJung`
r? `@ghost`
Use the proc-macro descr to track their individual expansions with
self-profiling events. This will help diagnose performance issues
with slow proc-macros.
Note invariance reason for FnDef types
Fixes#95272. Is it worthwhile even printing a variance explanation here? Or should I try to track down which function parameter is responsible for the invariance?
r? ``@Aaron1011`` since you wrote #89336
diagnostics: translation infrastructure
An implementation of the infrastructure required to have translatable diagnostic messages.
- Introduces a `DiagnosticMessage` type which can represent both the current non-translatable messages and identifiers for [Fluent](https://projectfluent.org/).
- Modifies current diagnostic API so that existing calls still work but `DiagnosticMessage`s can be provided too.
- Adds support for always loading a "fallback bundle" containing the English diagnostic messages, which are used when a `DiagnosticMessage::FluentIdentifier` is used in a diagnostic being emitted.
- Adds support for loading a "primary bundle" which contains the user's preferred language translation, and is used preferentially when it contains a diagnostic message being emitted. Primary bundles are loaded either from the path provided to `-Ztranslate-alternate-ftl` (for testing), or from the sysroot at `$sysroot/locale/$locale/*.ftl` given a locale with `-Ztranslate-lang` (which is parsed as a language identifier).
- Adds "diagnostic args" which enable normally-interpolated variables to be made available as variables for Fluent messages to use.
- Updates `#[derive(SessionDiagnostic)]` so that it can only be used for translatable diagnostics and update the handful of diagnostics which used the derive to be translatable.
For example, the following diagnostic...
```rust
#[derive(SessionDiagnostic)]
#[error = "E0195"]
pub struct LifetimesOrBoundsMismatchOnTrait {
#[message = "lifetime parameters or bounds on {item_kind} `{ident}` do not match the trait declaration"]
#[label = "lifetimes do not match {item_kind} in trait"]
pub span: Span,
#[label = "lifetimes in impl do not match this {item_kind} in trait"]
pub generics_span: Option<Span>,
pub item_kind: &'static str,
pub ident: Ident,
}
```
...becomes...
```rust
#[derive(SessionDiagnostic)]
#[error(code = "E0195", slug = "typeck-lifetimes-or-bounds-mismatch-on-trait")]
pub struct LifetimesOrBoundsMismatchOnTrait {
#[primary_span]
#[label]
pub span: Span,
#[label = "generics-label"]
pub generics_span: Option<Span>,
pub item_kind: &'static str,
pub ident: Ident,
}
```
```fluent
typeck-lifetimes-or-bounds-mismatch-on-trait =
lifetime parameters or bounds on {$item_kind} `{$ident}` do not match the trait declaration
.label = lifetimes do not match {$item_kind} in trait
.generics-label = lifetimes in impl do not match this {$item_kind} in trait
```
r? `@estebank`
cc `@oli-obk` `@Manishearth`
In #95555 this was moved out of `parse_tt_inner` and `nameize` into
`compute_locs`. But the next commit will be moving `compute_locs`
outwards to a place that isn't suitable for the missing fragment
identifier checking. So this reinstates the old checking.
Conditional on the parallel compiler being enabled, use a different
`IntlLangMemoizer` which supports being sent between threads in
`FluentBundle`.
Signed-off-by: David Wood <david.wood@huawei.com>
`FluentId` is the type alias that is used everywhere else so it should
be used here too so that this doesn't need updated if the alias changes.
Signed-off-by: David Wood <david.wood@huawei.com>
Add an option for enabling and disabling Fluent's directionality
isolation markers in output. Disabled by default as these can render in
some terminals and applications.
Signed-off-by: David Wood <david.wood@huawei.com>
Add some links to the Fluent documentation to
`DiagnosticMessage::FluentIdentifier` which explain what a Fluent
message and attribute are.
Signed-off-by: David Wood <david.wood@huawei.com>
Removes `expected_pluralize` parameter from diagnostic struct which is
no longer necessary as the Fluent message can determine the correct
pluralization.
Signed-off-by: David Wood <david.wood@huawei.com>
If the user requests a diagnostic locale of "en-US" then it doesn't make
sense to try and load that from the `$sysroot` because it is just the
default built-in locale.
Signed-off-by: David Wood <david.wood@huawei.com>
Extends support for generating `DiagnosticMessage::FluentIdentifier`
messages from `SessionDiagnostic` derive to `#[suggestion]`.
Signed-off-by: David Wood <david.wood@huawei.com>
Extends support for generating `DiagnosticMessage::FluentIdentifier`
messages from `SessionDiagnostic` derive to `#[label]`.
Signed-off-by: David Wood <david.wood@huawei.com>
In an effort to make it easier to port diagnostics to
`SessionDiagnostic` (for translation) and since translation slugs could
replace error codes, make error codes optional in the
`SessionDiagnostic` derive.
Signed-off-by: David Wood <david.wood@huawei.com>
A call to `set_arg` is generated for every field of a
`SessionDiagnostic` struct without attributes, but not all types support
being an argument, so `#[no_arg]` is introduced to skip these fields.
Signed-off-by: David Wood <david.wood@huawei.com>
Small commit renaming `#[message]` to `#[primary_span]` as this more
accurately reflects what it does now.
Signed-off-by: David Wood <david.wood@huawei.com>
Small commit adding backticks around types and annotations in the error
messages from the session diagnostic derive.
Signed-off-by: David Wood <david.wood@huawei.com>
Fluent diagnostics can insert directionality isolation markers around
interpolated variables indicating that there may be a shift from
right-to-left to left-to-right text (or vice-versa). These are disabled
because they are sometimes visible in the error output, but may be worth
investigating in future (for example: if type names are left-to-right
and the surrounding diagnostic messages are right-to-left, then these
might be helpful).
Signed-off-by: David Wood <david.wood@huawei.com>
Non-subdiagnostic fields (i.e. those that don't have `#[label]`
attributes or similar and are just additional context) have to be added
as arguments for Fluent messages to refer them. This commit extends the
`SessionDiagnostic` derive to do this for all fields that do not have
attributes and introduces an `IntoDiagnosticArg` trait that is
implemented on all types that can be converted to a argument for Fluent.
Signed-off-by: David Wood <david.wood@huawei.com>
Move the handling of `Span` or `(Span, Applicability)` types in
`#[suggestion]` attributes to its own function.
Signed-off-by: David Wood <david.wood@huawei.com>
Various small changes to comments, like wrapping code in backticks,
changing comments to doc comments and adding newlines.
Signed-off-by: David Wood <david.wood@huawei.com>
Extend loading of Fluent bundles so that bundles can be loaded from the
sysroot based on the language requested by the user, or using a nightly
flag.
Sysroot bundles are loaded from `$sysroot/share/locale/$locale/*.ftl`.
Signed-off-by: David Wood <david.wood@huawei.com>
This commit updates the signatures of all diagnostic functions to accept
types that can be converted into a `DiagnosticMessage`. This enables
existing diagnostic calls to continue to work as before and Fluent
identifiers to be provided. The `SessionDiagnostic` derive just
generates normal diagnostic calls, so these APIs had to be modified to
accept Fluent identifiers.
In addition, loading of the "fallback" Fluent bundle, which contains the
built-in English messages, has been implemented.
Each diagnostic now has "arguments" which correspond to variables in the
Fluent messages (necessary to render a Fluent message) but no API for
adding arguments has been added yet. Therefore, diagnostics (that do not
require interpolation) can be converted to use Fluent identifiers and
will be output as before.
`MultiSpan` contains labels, which are more complicated with the
introduction of diagnostic translation and will use types from
`rustc_errors` - however, `rustc_errors` depends on `rustc_span` so
`rustc_span` cannot use types like `DiagnosticMessage` without
dependency cycles. Introduce a new `rustc_error_messages` crate that can
contain `DiagnosticMessage` and `MultiSpan`.
Signed-off-by: David Wood <david.wood@huawei.com>
Introduce a `DiagnosticMessage` type that will enable diagnostic
messages to be simple strings or Fluent identifiers.
`DiagnosticMessage` is now used in the implementation of the standard
`DiagnosticBuilder` APIs.
Signed-off-by: David Wood <david.wood@huawei.com>
rustdoc: Fix resolution of `crate`-relative paths in doc links
Resolve `crate::foo` paths transparently to rustdoc, so their resolution no longer affects diagnostics and modules used for determining traits in scope.
The proper solution is to account for the current `module_id`/`parent_scope` in `fn resolve_crate_root`, but it's a slightly larger compiler changes. This PR moves the code closer to it, but keeps it rustdoc-specific.
Fixes https://github.com/rust-lang/rust/issues/78696
Fixes https://github.com/rust-lang/rust/issues/94924
Fix late-bound ICE in `dyn` return type suggestion
This fixes the root-cause of the attached issues -- the root problem is that we're using the return type from a signature with late-bound instead of early-bound regions. The change on line 1087 (`let Some(liberated_sig) = typeck_results.liberated_fn_sigs().get(fn_hir_id) else { return false; };`) makes sure we're grabbing the _right_ return type for this suggestion to check the `dyn` predicates with.
Fixes#91801Fixes#91803
This fix also includes some drive-by changes, specifically:
1. Don't suggest boxing when we have `-> dyn Trait` and are already returning `Box<T>` where `T: Trait` (before we always boxed the value).
2. Suggestion applies even when the return type is a type alias (e.g. `type Foo = dyn Trait`). This does cause the suggestion to expand to the aliased type, but I think it's still beneficial.
3. Split up the multipart suggestion because there's a 6-line max in the printed output...
I am open to splitting out the above changes, if we just want to fix the ICE first.
cc: ```@terrarier2111``` and #92289
Mention implementers of unsatisfied trait
When encountering an unsatisfied trait bound, if there are no other
suggestions, mention all the types that *do* implement that trait:
```
error[E0277]: the trait bound `f32: Foo` is not satisfied
--> $DIR/impl_wf.rs:22:6
|
LL | impl Baz<f32> for f32 { }
| ^^^^^^^^ the trait `Foo` is not implemented for `f32`
|
= help: the trait `Foo` is implemented for `i32`
note: required by a bound in `Baz`
--> $DIR/impl_wf.rs:18:31
|
LL | trait Baz<U: ?Sized> where U: Foo { }
| ^^^ required by this bound in `Baz`
```
```
error[E0277]: the trait bound `u32: Foo` is not satisfied
--> $DIR/associated-types-path-2.rs:29:5
|
LL | f1(2u32, 4u32);
| ^^ the trait `Foo` is not implemented for `u32`
|
= help: the trait `Foo` is implemented for `i32`
note: required by a bound in `f1`
--> $DIR/associated-types-path-2.rs:13:14
|
LL | pub fn f1<T: Foo>(a: T, x: T::A) {}
| ^^^ required by this bound in `f1`
```
Suggest dereferencing in more cases.
Fix#87437, fix#90970.
When encountering an unsatisfied trait bound, if there are no other
suggestions, mention all the types that *do* implement that trait:
```
error[E0277]: the trait bound `f32: Foo` is not satisfied
--> $DIR/impl_wf.rs:22:6
|
LL | impl Baz<f32> for f32 { }
| ^^^^^^^^ the trait `Foo` is not implemented for `f32`
|
= help: the following other types implement trait `Foo`:
Option<T>
i32
str
note: required by a bound in `Baz`
--> $DIR/impl_wf.rs:18:31
|
LL | trait Baz<U: ?Sized> where U: Foo { }
| ^^^ required by this bound in `Baz`
```
Mention implementers of traits in `ImplObligation`s.
Do not mention other `impl`s for closures, ranges and `?`.
Suggest borrowing when trying to coerce unsized type into `dyn Trait`
A helpful error in response to #95598, since we can't coerce e.g. `&str` into `&dyn Display`, but we can coerce `&&str` into `&dyn Display` :)
Not sure if the suggestion message needs some help. Let me know, and I can refine this PR.
Reduce unnecessary escaping in proc_macro::Literal::character/string
I noticed that https://doc.rust-lang.org/proc_macro/struct.Literal.html#method.character is producing unreadable literals that make macro-expanded code unnecessarily hard to read. Since the proc macro server was using `escape_unicode()`, every char is escaped using `\u{…}` regardless of whether there is any need to do so. For example `Literal::character('=')` would previously produce `'\u{3d}'` which unnecessarily obscures the meaning when reading the macro-expanded code.
I've changed Literal::string also in this PR because `str`'s `Debug` impl is also smarter than just calling `escape_debug` on every char. For example `Literal::string("ferris's")` would previously produce `"ferris\'s"` but will now produce `"ferris's"`.
A new matcher representation for use in `parse_tt`
By transforming the matcher into a different form, `parse_tt` can run faster and be easier to understand.
r? `@petrochenkov`
Improve method name suggestions
Attempts to improve method name suggestions when a matching method name
is not found. The approach taken is use the Levenshtein distance and
account for substrings having a high distance but can sometimes be very
close to the intended method (eg. empty vs is_empty).
resolves#94747
`parse_tt` currently traverses a `&[TokenTree]` to do matching. But this
is a bad representation for the traversal.
- `TokenTree` is nested, and there's a bunch of expensive and fiddly
state required to handle entering and exiting nested submatchers.
- There are three positions (sequence separators, sequence Kleene ops,
and end of the matcher) that are represented by an index that exceeds
the end of the `&[TokenTree]`, which is clumsy and error-prone.
This commit introduces a new representation called `MatcherLoc` that is
designed specifically for matching. It fixes all the above problems,
making the code much easier to read. A `&[TokenTree]` is converted to a
`&[MatcherLoc]` before matching begins. Despite the cost of the
conversion, it's still a net performance win, because various pieces of
traversal state are computed once up-front, rather than having to be
recomputed repeatedly during the macro matching.
Some improvements worth noting.
- `parse_tt_inner` is *much* easier to read. No more having to compare
`idx` against `len` and read comments to understand what the result
means.
- The handling of `Delimited` in `parse_tt_inner` is now trivial.
- The three end-of-sequence cases in `parse_tt_inner` are now handled in
three separate match arms, and the control flow is much simpler.
- `nameize` is no longer recursive.
- There were two places that issued "missing fragment specifier" errors:
one in `parse_tt_inner()`, and one in `nameize()`. Presumably the
latter was never executed. There's now a single place issuing these
errors, in `compute_locs()`.
- The number of heap allocations done for a `check full` build of
`async-std-1.10.0` (an extreme example of heavy macro use) drops from
11.8M to 2.6M, and most of these occur outside of macro matching.
- The size of `MatcherPos` drops from 64 bytes to 16 bytes. Small enough
that it no longer needs boxing, which partly accounts for the
reduction in allocations.
- The rest of the drop in allocations is due to the removal of
`MatcherKind`, because we no longer need to record anything for the
parent matcher when entering a submatcher.
- Overall it reduces code size by 45 lines.
Do not use `ParamEnv::and` when building a cache key from a param-env and trait eval candidate
Do not use `ParamEnv::and` to cache a param-env with a selection/evaluation candidate.
This is because if the param-env is `RevealAll` mode, and the candidate looks global (i.e. it has erased regions, which can show up when we normalize a projection type under a binder<sup>1</sup>), then when we use `ParamEnv::and` to pair the candidate and the param-env for use as a cache key, we will throw away the param-env's caller bounds, and we'll end up caching a candidate that we inferred from the param-env with a empty param-env, which may cause cache-hit later when we have an empty param-env, and possibly mess with normalization like we see in the referenced issue during codegen.
Not sure how to trigger this with a more structured test, but changing `check-pass` to `build-pass` triggers the case that https://github.com/rust-lang/rust/issues/94903 detected.
<sup>1.</sup> That is, we will replace the late-bound region with a placeholder, which gets canonicalized and turned into an infererence variable, which gets erased during region freshening right before we cache the result. Sorry, it's quite a few steps.
Fixes#94903
r? `@Aaron1011` (or reassign as you see fit)
Don't emit non-asm contents error for naked function composed of errors
## Motivation
For naked functions an error is emitted when they are composed of anything other than a single asm!() block. However, this error triggers in a couple situations in which it adds no additional information or is actively misleading.
One example is if you do have an asm!() block but simply one with a syntax error:
```rust
#[naked]
unsafe extern "C" fn compiler_errors() {
asm!(invalid_syntax)
}
```
This results in two errors, one for the syntax error itself and another telling you that you need an asm block in your function:
```rust
error[E0787]: naked functions must contain a single asm block
--> src/main.rs:6:1
|
6 | / unsafe extern "C" fn naked_compile_error() {
7 | | asm!(blah)
8 | | }
| |_^
```
This issue also comes up when [utilizing `compile_error!()` for improving your diagnostics](https://twitter.com/steveklabnik/status/1509538243020218372), such as raising a compiler error when compiling for an unsupported target.
## Implementation
The rules this PR implements are as follows:
1. If any non-erroneous non-asm statement is included, an error will still occur
2. If multiple asm statements are included, an error will still occur
3. If 0 or 1 asm statements are present, as well as any non-zero number of erroneous statements, then this error will *not* be raised as it is likely either redundant or incorrect
The rule of thumb is effectively "if an error is present and its correction could change things, don't raise an error".
Reduce the cost of loading all built-ins targets
This PR started by measuring the exact slowdown of checking of well known conditional values.
Than this PR implemented some technics to reduce the cost of loading all built-ins targets.
cf. https://github.com/rust-lang/rust/issues/82450#issuecomment-1073992323
Add debug assertions to some unsafe functions
As suggested by https://github.com/rust-lang/rust/issues/51713
~~Some similar code calls `abort()` instead of `panic!()` but aborting doesn't work in a `const fn`, and the intrinsic for doing dispatch based on whether execution is in a const is unstable.~~
This picked up some invalid uses of `get_unchecked` in the compiler, and fixes them.
I can confirm that they do in fact pick up invalid uses of `get_unchecked` in the wild, though the user experience is less-than-awesome:
```
Running unittests (target/x86_64-unknown-linux-gnu/debug/deps/rle_decode_fast-04b7918da2001b50)
running 6 tests
error: test failed, to rerun pass '--lib'
Caused by:
process didn't exit successfully: `/home/ben/rle-decode-helper/target/x86_64-unknown-linux-gnu/debug/deps/rle_decode_fast-04b7918da2001b50` (signal: 4, SIGILL: illegal instruction)
```
~~As best I can tell these changes produce a 6% regression in the runtime of `./x.py test` when `[rust] debug = true` is set.~~
Latest commit (6894d559bd) brings the additional overhead from this PR down to 0.5%, while also adding a few more assertions. I think this actually covers all the places in `core` that it is reasonable to check for safety requirements at runtime.
Thoughts?
