Fix e_flags for 32-bit MIPS targets in generated object file
In #95604 the compiler started generating a temporary symbols.o which is added to the linker invocation. This object file has an `e_flags` which is invalid for 32-bit MIPS targets. Even though symbols.o doesn't contain code, linking these targets with [lld fails](https://github.com/llvm/llvm-project/blob/main/lld/ELF/Arch/MipsArchTree.cpp#L76-L79) with
```
rust-lld: error: foo-cgu.0.rcgu.o: ABI 'o32' is incompatible with target ABI 'n64'
```
because it omits the ABI bits (`EF_MIPS_ABI_O32`) so lld assumes it's using the N64 ABI. This breaks linking on nightly for the out-of-tree [mipsel-sony-psx target](https://github.com/ayrtonm/psx-sdk-rs/issues/9), the builtin mipsel-sony-psp target (cc `@overdrivenpotato)` and probably any other 32-bit MIPS target using lld.
This PR sets the ABI in `e_flags` to O32 since that's the only ABI for 32-bit MIPS that LLVM supports. It also sets other `e_flags` bits based on the target to avoid similar issues with the object file arch and PIC. I had to bump the object crate version since some of these constants were [added recently](https://github.com/gimli-rs/object/pull/433). I'm not sure if this PR needs a test, but I can confirm that it fixes the linking issue on both targets I mentioned.
Gracefully handle normalization failures in the prospective inliner cycle detector
Preliminary work for adding the regression test in #96950 to our test suite (it was flaky on glacier).
If this PR solves the flakiness on glacier, we can then merge #96950
don't encode only locally used attrs
Part of https://github.com/rust-lang/compiler-team/issues/505.
We now filter builtin attributes before encoding them in the crate metadata in case they should only be used in the local crate. To prevent accidental misuse `get_attrs` now requires the caller to state which attribute they are interested in. For places where that isn't trivially possible, I've added a method `fn get_attrs_unchecked` which I intend to remove in a followup PR.
After this pull request landed, we can then slowly move all attributes to only be used in the local crate while being certain that we don't accidentally try to access them from extern crates.
cc https://github.com/rust-lang/rust/pull/94963#issuecomment-1082924289
diagnostics: port more diagnostics to derive + support for `()` fields
- Extend diagnostic derive so that spanless subdiagnostics (e.g. some uses of `help`/`note`) can be applied via attributes to fields of type `()` (currently spanless subdiagnostics are applied via attributes on the diagnostic struct itself). A consequence of this is that `Option<()>` fields can be used to represent optional spanless subdiagnostics, which are sometimes useful (e.g. for a `help` that should only show on nightly builds).
- Simplify the "explicit generic args with impl trait" diagnostic struct (from #96760) using support for `Option<()>` spanless subdiagnostics.
- Change `DiagnosticBuilder::set_arg`, used to provide context for Fluent messages, so that it takes anything that implements `IntoDiagnosticArg`, rather than `DiagnosticArgValue` - this improves the ergonomics of manual implementations of `SessionDiagnostic` which are translatable.
- Port "the type parameter `T` must be explicitly specified", "manual implementations of `X` are experimental", "could not resolve substs on overridden impl" diagnostics to diagnostic structs.
- When testing macros from `rustc_macros` in `ui-fulldeps` tests, sometimes paths from the compiler source tree can be shown in error messages - these need to be normalized in `compiletest`.
r? `@oli-obk`
cc `@pvdrz`
Port the "the type parameter `T` must be explicitly specified"
diagnostic to using a diagnostic struct.
Signed-off-by: David Wood <david.wood@huawei.com>
Manual implementors of translatable diagnostics will need to call
`set_arg`, not just the derive, so make this function a bit more
ergonomic by taking `IntoDiagnosticArg` rather than
`DiagnosticArgValue`.
Signed-off-by: David Wood <david.wood@huawei.com>
Using new support for spanless subdiagnostics from `()` fields in the
diagnostic derive, simplify the "explicit generic args with impl trait"
diagnostic's struct.
Signed-off-by: David Wood <david.wood@huawei.com>
Type attributes could previously be used to support spanless
subdiagnostics but these couldn't easily be made optional in the same
way that spanned subdiagnostics could by using a field attribute on a
field with an `Option<Span>` type. Spanless subdiagnostics can now be
specified on fields with `()` type or `Option<()>` type.
Signed-off-by: David Wood <david.wood@huawei.com>
Remove `PartialOrd`/`Ord` impl for `PlaceRef`
This is a new attempt at #93315. It removes one usage
of the `Ord` impl for `DefId`, which should make it easier
to eventually remove that impl.
Add `sub_ptr` on pointers (the `usize` version of `offset_from`)
We have `add`/`sub` which are the `usize` versions of `offset`, this adds the `usize` equivalent of `offset_from`. Like how `.add(d)` replaced a whole bunch of `.offset(d as isize)`, you can see from the changes here that it's fairly common that code actually knows the order between the pointers and *wants* a `usize`, not an `isize`.
As a bonus, this can do `sub nuw`+`udiv exact`, rather than `sub`+`sdiv exact`, which can be optimized slightly better because it doesn't have to worry about negatives. That's why the slice iterators weren't using `offset_from`, though I haven't updated that code in this PR because slices are so perf-critical that I'll do it as its own change.
This is an intrinsic, like `offset_from`, so that it can eventually be allowed in CTFE. It also allows checking the extra safety condition -- see the test confirming that CTFE catches it if you pass the pointers in the wrong order.
Like we have `add`/`sub` which are the `usize` version of `offset`, this adds the `usize` equivalent of `offset_from`. Like how `.add(d)` replaced a whole bunch of `.offset(d as isize)`, you can see from the changes here that it's fairly common that code actually knows the order between the pointers and *wants* a `usize`, not an `isize`.
As a bonus, this can do `sub nuw`+`udiv exact`, rather than `sub`+`sdiv exact`, which can be optimized slightly better because it doesn't have to worry about negatives. That's why the slice iterators weren't using `offset_from`, though I haven't updated that code in this PR because slices are so perf-critical that I'll do it as its own change.
This is an intrinsic, like `offset_from`, so that it can eventually be allowed in CTFE. It also allows checking the extra safety condition -- see the test confirming that CTFE catches it if you pass the pointers in the wrong order.
Implement a lint to warn about unused macro rules
This implements a new lint to warn about unused macro rules (arms/matchers), similar to the `unused_macros` lint added by #41907 that warns about entire macros.
```rust
macro_rules! unused_empty {
(hello) => { println!("Hello, world!") };
() => { println!("empty") }; //~ ERROR: 1st rule of macro `unused_empty` is never used
}
fn main() {
unused_empty!(hello);
}
```
Builds upon #96149 and #96156.
Fixes#73576
Gracefully fail to resolve associated items instead of `delay_span_bug`.
`codegen_fulfill_obligation` is used during instance resolution for trait items.
In case of insufficient normalization issues during MIR inlining, it caused ICEs.
It's better to gracefully refuse to resolve the associated item, and let the caller decide what to do with this.
Split from https://github.com/rust-lang/rust/pull/91743Closes#69121Closes#73021Closes#88599Closes#93008Closes#93248Closes#94680Closes#96170
r? `@oli-obk`
tighten sanity checks around Scalar and ScalarPair
While investigating https://github.com/rust-lang/rust/issues/96185 I noticed codegen has tighter sanity checks here than Miri does, so I added some more assertions. Strangely, some of them fail, so I also needed to add a HACK... that is probably worth looking into.
This does not fix that issue, but it changes the ICE messages, making it quite clear that we have a scalar whose size is not the same as that of the surrounding layout.
r? `@oli-obk`
logging: add env var to control verbose scope entry/exit logging
~~This got removed in https://github.com/rust-lang/rust/pull/75143, and I find this makes long traces a lot harder to read, so I propose we add this back.~~
Example trace:
```
│ │ ├─0ms INFO rustc_const_eval::interpret::step return
│ │ ├─0ms INFO rustc_const_eval::interpret::eval_context popping stack frame (returning from function)
│ │┌┘rustc_const_eval::interpret::eval_context::frame std::ptr::mut_ptr::<impl *mut u8>::guaranteed_eq
│ ├┘rustc_const_eval::interpret::eval_context::frame std::ptr::mut_ptr::<impl *mut u8>::is_null
│ ├─1ms INFO rustc_const_eval::interpret::step // executing bb2
│ ├─1ms INFO rustc_const_eval::interpret::step StorageDead(_4)
│ ├─1ms INFO rustc_const_eval::interpret::step StorageDead(_2)
│ ├─1ms INFO rustc_const_eval::interpret::step return
│ ├─1ms INFO rustc_const_eval::interpret::eval_context popping stack frame (returning from function)
│┌┘rustc_const_eval::interpret::eval_context::frame std::ptr::mut_ptr::<impl *mut u8>::is_null
├┘rustc_const_eval::interpret::eval_context::frame std::sys_common::thread_local_dtor::register_dtor_fallback::run_dtors
├─178ms INFO rustc_const_eval::interpret::step // executing bb2
├─178ms INFO rustc_const_eval::interpret::step StorageDead(_5)
```
r? `@oli-obk`
Fix inaccurate function name in `rustc_const_eval` docs
Looks to me like this fixes#85513. I had trouble making a intra-docs link to `eval_place_to_op` work, though...
Use lifetimes on type-alias-impl-trait used in function signatures to infer output type lifetimes
fixes https://github.com/rust-lang/rust/issues/96564
TLDR:
```rust
fn execute(ty: Ty<'_>) -> &str { todo!() }
```
(`Ty` being a type alias impl trait) used to produce the following error before this PR
```
error[E0581]: return type references an anonymous lifetime, which is not constrained by the fn input types
--> src/lib.rs:4:27
|
4 | fn execute(ty: Ty<'_>) -> &str { todo!() }
| ^^^^
|
= note: lifetimes appearing in an associated type are not considered constrained
```
Remove hacks in `make_token_stream`.
`make_tokenstream` has three commented hacks, and a comment at the top
referring to #67062. These hacks have no observable effect, at least as judged
by running the test suite. The hacks were added in #82608, with an explanation
[here](https://github.com/rust-lang/rust/pull/82608#issuecomment-812877329). It
appears that one of the following is true: (a) they never did anything useful,
(b) they do something useful but we have no test coverage for them, or (c)
something has changed in the meantime that means they are no longer necessary.
This commit removes the hacks and the comments, in the hope that (b) is not
true.
r? `@Aaron1011`
Use `FxIndexSet` to avoid sorting fake borrows
This fixes#96449, but I haven't yet been able to
make the reproducer work using `#[cfg]` attributes,
so we can't use the 'revision' infra to write a test
The previous implementation relied on sorting by `PlaceRef`.
This requires sorting by a `DefId`, which uses untracked state
(see #93315)
In #95604 the compiler started generating a temporary symbols.o which is added
to the linker invocation. This object file has an `e_flags` which may be invalid
for 32-bit MIPS targets. Even though symbols.o doesn't contain code, linking
with [lld fails](https://github.com/llvm/llvm-project/blob/main/lld/ELF/Arch/MipsArchTree.cpp#L79) with
```
rust-lld: error: foo-cgu.0.rcgu.o: ABI 'o32' is incompatible with target ABI 'n64'
```
because it omits the ABI bits (EF_MIPS_ABI_O32) so lld assumes it's using the
N64 ABI. This breaks linking on nightly for the out-of-tree [psx
target](https://github.com/ayrtonm/psx-sdk-rs/issues/9), the builtin
mipsel-sony-psp target (cc @overdrivenpotato) and any other 32-bit MIPS
target using lld.
This PR sets the ABI in `e_flags` to O32 since that's the only ABI for 32-bit
MIPS that LLVM supports. It also sets other `e_flags` bits based on the target.
I had to bump the object crate version since some of these constants were [added
recently](https://github.com/gimli-rs/object/pull/433). I'm not sure if this
PR needs a test, but I can confirm that it fixes the linking issue on both
targets I mentioned.
Handle mismatched generic param kinds in trait impls betterly
- Check that generic params on a generic associated type are the same as in the trait definition
- Check that const generics are not used in place of type generics (and the other way round too)
r? `@lcnr`
Check hidden types for well formedness at the definition site instead of only at the opaque type itself
work towards #90409 . We'll need to look into closure and generator bodies of closures and generators nested inside the hidden type in order to fix that. In hindsight this PR is not necessary for that, but it may be a bit easier with it and we'll get better diagnostics from it on its own.
make sure ScalarPair enums have ScalarPair variants; add some layout sanity checks
`@eddyb` suggested that it might be reasonable for `ScalarPair` enums to simply adjust the ABI of their variants accordingly, such that the layout invariant Miri expects actually holds. This PR implements that. I should note though that I don't know much about this layout computation code and what non-Miri consumers expect from it, so tread with caution!
I also added a function to sanity-check that computed layouts are internally consistent. This helped a lot in figuring out the final shape of this PR, though I am also not 100% sure that these sanity checks are the right ones.
Cc `@oli-obk`
Fixes https://github.com/rust-lang/rust/issues/96221
Properly fix#96638Closes#96638
The main part of this change is `Error::Invalid` now returns both the input and arg indices. However, I realized the code here was kind of confusing and not internally consistent (and thus I was having trouble getting the right behavior). So I've also switched `input_indices` and `arg_indices` to more closely match some naming in `checks` (although I think a more thorough cleanup there could be beneficial). I've added comments, but essentially `input_indices` refers to *user provided* inputs and `arg_indices` refers to *expected* args.
Do not lint on explicit outlives requirements from external macros.
The current implementation of the list rightfully skipped where predicates from external macros.
However, if the where predicate came from the current macro but the bounds were from an external macro, the lint still fired.
Closes https://github.com/rust-lang/rust/issues/96640
This fixes#96449, but I haven't yet been able to
make the reproducer work using `#[cfg]` attributes,
so we can't use the 'revision' infra to write a test
The previous implementation relied on sorting by `PlaceRef`.
This requires sorting by a `DefId`, which uses untracked state
(see #93315)
Fortify handing of where bounds on trait & trait alias definitions
Closes https://github.com/rust-lang/rust/issues/96664
Closes https://github.com/rust-lang/rust/issues/96665
Since https://github.com/rust-lang/rust/pull/93803, when listing all bounds and predicates we now need to account for the possible presence of predicates on any of the generic parameters. Both bugs were hidden by the special handling of bounds at the generic parameter declaration position.
Trait alias expansion used to confuse predicates on `Self` and where predicates.
Exiting too late when listing all the bounds caused a cycle error.
Optimize switch sources representation and usage
* Avoid constructing switch sources unless necessary - switch sources are used by backward analysis with a custom switch int edge effects, but are otherwise unnecessarily computed.
* Use sparse representation of switch sources to avoid quadratic space overhead.
Some subst cleanup
Two separate things here. Both changes are useful for some refactoring I'm doing to add an "EarlyBinder" newtype. (Part of chalkification).
1) Remove `subst_spanned` and just use `subst`. It wasn't used much anyways. In practice, I think we can probably get most of the info just from the actual error message. If not, outputting logs should do the trick. (The specific line probably wouldn't help much anyways).
2) Call `.subst()` before `replace_bound_vars_with_fresh_vars` and `erase_late_bound_regions` in three places that do the opposite. I think there might have been some time in the past that the order here matter for something, but this shouldn't be the case anymore. Conceptually, it makes more sense to the of the *early bound* vars on `fn`s as "outside" the late bound vars.
Actually fix ICE from #96583
PR #96746 fixed a very similar bug, so the same logic is used in a different place.
I originally concluded that the two issues (#96583 and #96738) were identical by comparing the backtrace, but I didn't look close enough.
Warn on unused `#[doc(hidden)]` attributes on trait impl items
[Zulip conversation](https://rust-lang.zulipchat.com/#narrow/stream/266220-rustdoc/topic/.E2.9C.94.20Validy.20checks.20for.20.60.23.5Bdoc.28hidden.29.5D.60).
Whether an associated item in a trait impl is shown or hidden in the documentation entirely depends on the corresponding item in the trait declaration. Rustdoc completely ignores `#[doc(hidden)]` attributes on impl items. No error or warning is emitted:
```rust
pub trait Tr { fn f(); }
pub struct Ty;
impl Tr for Ty { #[doc(hidden)] fn f() {} }
// ^^^^^^^^^^^^^^ ignored by rustdoc and currently
// no error or warning issued
```
This may lead users to the wrong belief that the attribute has an effect. In fact, several such cases are found in the standard library (I've removed all of them in this PR).
There does not seem to exist any incentive to allow this in the future either: Impl'ing a trait for a type means the type *fully* conforms to its API. Users can add `#[doc(hidden)]` to the whole impl if they want to hide the implementation or add the attribute to the corresponding associated item in the trait declaration to hide the specific item. Hiding an implementation of an associated item does not make much sense: The associated item can still be found on the trait page.
This PR emits the warn-by-default lint `unused_attribute` for this case with a future-incompat warning.
`@rustbot` label T-compiler T-rustdoc A-lint
optimize `promote_consts` by caching the results of `validate_local`
From the FIXME in the impl of `promote_consts`. Early return the `validate_local` should save some compile time.
`qualif_local` is similar to this, but requires futher changing because there are different types of qualif checks. If this PR is effective, I will do it as well.
Remove `#[rustc_deprecated]`
This removes `#[rustc_deprecated]` and introduces diagnostics to help users to the right direction (that being `#[deprecated]`). All uses of `#[rustc_deprecated]` have been converted. CI is expected to fail initially; this requires #95958, which includes converting `stdarch`.
I plan on following up in a short while (maybe a bootstrap cycle?) removing the diagnostics, as they're only intended to be short-term.
Support tool lints with the `#[expect]` attribute (RFC 2383)
This PR fixes the ICE https://github.com/rust-lang/rust/issues/94953 by making the assert for converted expectation IDs conditional.
Additionally, it moves the lint expectation check into a separate query to support rustdoc and other tools. On the way, I've also added some tests to ensure that the attribute works for Clippy and rustdoc lints.
The number of changes comes from the long test file. This may look like a monster PR, this may smell like a monster PR and this may be a monster PR, but it's a harmless monster. 🦕
---
Closes: https://github.com/rust-lang/rust/issues/94953
cc: https://github.com/rust-lang/rust/issues/85549
r? `@wesleywiser`
cc: `@rust-lang/rustdoc`
Switch sources are used by backward analysis with a custom switch int
edge effects, but are otherwise unnecessarily computed.
Delay the computation until we know that switch sources are indeed
required and avoid the computation otherwise.
Track if a where bound comes from a impl Trait desugar
With https://github.com/rust-lang/rust/pull/93803 `impl Trait` function arguments get desugared to hidden where bounds. However, Clippy needs to know if a bound was originally a `impl Trait` or an actual bound. This adds a field to the `WhereBoundPredicate` struct to keep track of this information during AST->HIR lowering.
r? `@cjgillot`
cc `@estebank` (as the reviewer of #93803)
