Handle macros with multiple kinds, and improve errors
(I recommend reviewing this commit-by-commit.)
Switch to a bitflags `MacroKinds` to support macros with more than one kind
Review everything that uses `MacroKind`, and switch anything that could refer to more than one kind to use `MacroKinds`.
Add a new `SyntaxExtensionKind::MacroRules` for `macro_rules!` macros, using the concrete `MacroRulesMacroExpander` type, and have it track which kinds it can handle. Eliminate the separate optional `attr_ext`, now that a `SyntaxExtension` can handle multiple macro kinds.
This also avoids the need to downcast when calling methods on `MacroRulesMacroExpander`, such as `get_unused_rule`.
Integrate macro kind checking into name resolution's `sub_namespace_match`, so that we only find a macro if it's the right type, and eliminate the special-case hack for attributes.
This allows detecting and report macro kind mismatches early, and more precisely, improving various error messages. In particular, this eliminates the case in `failed_to_match_macro` to check for a function-like invocation of a macro with no function-like rules.
Instead, macro kind mismatches now result in an unresolved macro, and we detect this case in `unresolved_macro_suggestions`, which now carefully distinguishes between a kind mismatch and other errors.
This also handles cases of forward-referenced attributes and cyclic attributes.
----
In this PR, I've minimally fixed up `rustdoc` so that it compiles and passes tests. This is just the minimal necessary fixes to handle the switch to `MacroKinds`, and it only works for macros that don't actually have multiple kinds. This will panic (with a `todo!`) if it encounters a macro with multiple kinds.
rustdoc needs further fixes to handle macros with multiple kinds, and to handle attributes and derive macros that aren't proc macros. I'd appreciate some help from a rustdoc expert on that.
----
r? ````````@petrochenkov````````
Review everything that uses `MacroKind`, and switch anything that could
refer to more than one kind to use `MacroKinds`.
Add a new `SyntaxExtensionKind::MacroRules` for `macro_rules!` macros,
using the concrete `MacroRulesMacroExpander` type, and have it track
which kinds it can handle. Eliminate the separate optional `attr_ext`,
now that a `SyntaxExtension` can handle multiple macro kinds.
This also avoids the need to downcast when calling methods on
`MacroRulesMacroExpander`, such as `get_unused_rule`.
Integrate macro kind checking into name resolution's
`sub_namespace_match`, so that we only find a macro if it's the right
type, and eliminate the special-case hack for attributes.
Reject relaxed bounds inside associated type bounds (ATB)
**Reject** relaxed bounds — most notably `?Sized` — inside associated type bounds `TraitRef<AssocTy: …>`.
This was previously accepted without warning despite being incorrect: ATBs are *not* a place where we perform *sized elaboration*, meaning `TraitRef<AssocTy: …>` does *not* elaborate to `TraitRef<AssocTy: Sized + …>` if `…` doesn't contain `?Sized`. Therefore `?Sized` is meaningless. In no other (stable) place do we (intentionally) allow relaxed bounds where we don't also perform sized elab, this is highly inconsistent and confusing! Another point of comparison: For the desugared `$SelfTy: TraitRef, $SelfTy::AssocTy: …` we don't do sized elab either (and thus also don't allow relaxed bounds).
Moreover — as I've alluded to back in https://github.com/rust-lang/rust/pull/135841#pullrequestreview-2619462717 — some later validation steps only happen during sized elaboration during HIR ty lowering[^1]. Namely, rejecting duplicates (e.g., `?Trait + ?Trait`) and ensuring that `Trait` in `?Trait` is equal to `Sized`[^2]. As you can probably guess, on stable/master we don't run these checks for ATBs (so we allow even more nonsensical bounds like `Iterator<Item: ?Copy>` despite T-types's ruling established in the FCP'ed rust-lang/rust#135841).
This PR rectifies all of this. I cratered this back in 2025-01-10 with (allegedly) no regressions found ([report](https://github.com/rust-lang/rust/pull/135331#issuecomment-2585330783), [its analysis](https://github.com/rust-lang/rust/pull/135331#issuecomment-2585356422)). [However a contributor manually found two occurrences](https://github.com/rust-lang/rust/issues/135229#issuecomment-2581832852) of `TraitRef<AssocTy: ?Sized>` in small hobby projects (presumably via GH code search). I immediately sent downstream PRs: https://github.com/Gui-Yom/turbo-metrics/pull/14, https://github.com/ireina7/summon/pull/1 (however, the owners have showed no reaction so far).
I'm leaning towards banning these forms **without a FCW** because a FCW isn't worth the maintenance cost[^3]. Note that associated type bounds were stabilized in 1.79.0 (released 2024-06-13 which is 13 months ago), so the proliferation of ATBs shouldn't be that high yet. If you think we should do another crater run since the last one was 6 months ago, I'm fine with that.
Fixesrust-lang/rust#135229.
[^1]: I consider this a flaw in the implementation and [I've already added a huge FIXME](82a02aefe0/compiler/rustc_hir_analysis/src/hir_ty_lowering/bounds.rs (L195-L207)).
[^2]: To be more precise, if the internal flag `-Zexperimental-default-bounds` is provided other "default traits" (needs internal feature `lang_items`) are permitted as well (cc closely related internal feature: `more_maybe_bounds`).
[^3]: Having to track this and adding an entire lint whose remnants would remain in the code base forever (we never *fully* remove lints).
Return a struct with named fields from `hash_owner_nodes`
While looking through this code for other reasons, I noticed a nice opportunity to return a struct with named fields instead of a tuple. The first patch also introduces an early-return to flatten the rest of `hash_owner_nodes`.
There are further changes that could potentially be made here (renaming things, `Option<Hashes>` instead of optional fields), but I'm not deeply familiar with this code so I didn't want to disturb the calling code too much.
de-duplicate condition scoping logic between AST→HIR lowering and `ScopeTree` construction
There was some overlap between `rustc_ast_lowering::LoweringContext::lower_cond` and `rustc_hir_analysis::check::region::resolve_expr`, so I've removed the former and migrated its logic to the latter, with some simplifications.
Consequences:
- For `while` and `if` expressions' `let`-chains, this changes the `HirId`s for the `&&`s to properly correspond to their AST nodes. This is how guards were handled already.
- This makes match guards share previously-duplicated logic with `if`/`while` expressions. This will also be used by guard pattern[^1] guards.
- Aside from legacy syntax extensions (e.g. some builtin macros) that directly feed AST to the compiler, it's currently impossible to put attributes directly on `&&` operators in `let` chains[^2]. Nonetheless, attributes on `&&` operators in `let` chains in `if`/`while` expression conditions are no longer silently ignored and will be lowered.
- This no longer wraps conditions in `DropTemps`, so the HIR and THIR will be slightly smaller.
- `DesugaringKind::CondTemporary` is now gone. It's no longer applied to any spans, and all uses of it were dead since they were made to account for `if` and `while` being desugared to `match` on a boolean scrutinee.
- Should be a marginal perf improvement beyond that due to leveraging [`ScopeTree` construction](5e749eb66f/compiler/rustc_hir_analysis/src/check/region.rs (L312-L355))'s clever handling of `&&` and `||`:
- This removes some unnecessary terminating scopes that were placed around top-level `&&` and `||` operators in conditions. When lowered to MIR, logical operator chains don't create intermediate boolean temporaries, so there's no temporary to drop. The linked snippet handles wrapping the operands in terminating scopes as necessary, in case they create temporaries.
- The linked snippet takes care of letting `let` temporaries live and terminating other operands, so we don't need separate traversals of `&&` chains for that.
[^1]: rust-lang/rust#129967
[^2]: Case-by-case, here's my justification: `#[attr] e1 && e2` applies the attribute to `e1`. In `#[attr] (e1 && e2)` , the attribute is on the parentheses in the AST, plus it'd fail to parse if `e1` or `e2` contains a `let`. In `#[attr] expands_to_let_chain!()`, the attribute would already be ignored (rust-lang/rust#63221) and it'd fail to parse anyway; even if the expansion site is a condition, the expansion wouldn't be parsed with `Restrictions::ALLOW_LET`. If it *was* allowed, the notion of a "reparse context" from https://github.com/rust-lang/rust/issues/61733#issuecomment-509626449 would be necessary in order to make `let`-chains left-associative; multiple places in the compiler assume they are.
Allow custom default address spaces and parse `p-` specifications in the datalayout string
Some targets, such as CHERI, use as default an address space different from the "normal" default address space `0` (in the case of CHERI, [200 is used](https://www.cl.cam.ac.uk/techreports/UCAM-CL-TR-877.pdf)). Currently, `rustc` does not allow to specify custom address spaces and does not take into consideration [`p-` specifications in the datalayout string](https://llvm.org/docs/LangRef.html#langref-datalayout).
This patch tries to mitigate these problems by allowing targets to define a custom default address space (while keeping the default value to address space `0`) and adding the code to parse the `p-` specifications in `rustc_abi`. The main changes are that `TargetDataLayout` now uses functions to refer to pointer-related informations, instead of having specific fields for the size and alignment of pointers in the default address space; furthermore, the two `pointer_size` and `pointer_align` fields in `TargetDataLayout` are replaced with an `FxHashMap` that holds info for all the possible address spaces, as parsed by the `p-` specifications.
The potential performance drawbacks of not having ad-hoc fields for the default address space will be tested in this PR's CI run.
r? workingjubilee
Fix some comments and related types and locals where it is obvious, e.g.
- bare_fn -> fn_ptr
- LifetimeBinderKind::BareFnType -> LifetimeBinderKind::FnPtrType
Co-authored-by: León Orell Valerian Liehr <me@fmease.dev>
It's like `Symbol` but for byte strings. The interner is now used for
both `Symbol` and `ByteSymbol`. E.g. if you intern `"dog"` and `b"dog"`
you'll get a `Symbol` and a `ByteSymbol` with the same index and the
characters will only be stored once.
The motivation for this is to eliminate the `Arc`s in `ast::LitKind`, to
make `ast::LitKind` impl `Copy`, and to avoid the need to arena-allocate
`ast::LitKind` in HIR. The latter change reduces peak memory by a
non-trivial amount on literal-heavy benchmarks such as `deep-vector` and
`tuple-stress`.
`Encoder`, `Decoder`, `SpanEncoder`, and `SpanDecoder` all get some
changes so that they can handle normal strings and byte strings.
This change does slow down compilation of programs that use
`include_bytes!` on large files, because the contents of those files are
now interned (hashed). This makes `include_bytes!` more similar to
`include_str!`, though `include_bytes!` contents still aren't escaped,
and hashing is still much cheaper than escaping.
Taint body on invalid call ABI
Fixes https://github.com/rust-lang/rust/issues/142969
I'm not certain if there are any other paths that should be tainted, but they would operate similarly. Perhaps pointer coercion.
Introduces `extern "rust-invalid"` for testing purposes.
r? ```@workingjubilee``` or ```@oli-obk``` (or anyone)
We modify rustc_ast_lowering to prevent all unsupported ABIs
from leaking through the HIR without being checked for target support.
Previously ad-hoc checking on various HIR items required making sure
we check every HIR item which could contain an `extern "{abi}"` string.
This is a losing proposition compared to gating the lowering itself.
As a consequence, unsupported ABI strings will now hard-error instead of
triggering the FCW `unsupported_fn_ptr_calling_conventions`.
This FCW was upgraded to warn in dependencies in Rust 1.87 which was
released on 2025 May 17, and it is now 2025 June, so it has become
active within a stable Rust version.
As we already had errored on these ABIs in most other positions, and
have warned for fn ptrs, this breakage has had reasonable foreshadowing.
However, this does cause errors for usages of `extern "{abi}"` that were
theoretically writeable within source but could not actually be applied
in any useful way by Rust programmers without either warning or error.
For instance, trait declarations without impls were never checked.
These are the exact kinds of leakages that this new approach prevents.
A deprecation cycle is not useful for these marginal cases as upon impl,
even default impls within traits, different HIR objects would be used.
Details of our HIR analysis meant that those objects did get checked.
We choose to error twice if an ABI is also barred by a feature gate
on the presumption that usage of a target-incorrect ABI is intentional.
Co-authored-by: Ralf Jung <post@ralfj.de>
