Auto merge of rust-lang#127014 - jhpratt:rollup-45ic8f5, r=jhpratt

Rollup of 6 pull requests

Successful merges:

 - rust-lang#126571 (Less `maybe_whole_expr`, take 2)
 - rust-lang#126721 (coverage: Make `#[coverage(..)]` apply recursively to nested functions)
 - rust-lang#126928 (Some `Nonterminal` removal precursors)
 - rust-lang#126929 (Remove `__rust_force_expr`.)
 - rust-lang#126980 (set self.is_known_utf8 to false in extend_from_slice)
 - rust-lang#126983 (Remove `f16` and `f128` ICE paths from smir)

r? `@ghost`
`@rustbot` modify labels: rollup

Author: bors

compiler/rustc_ast/src/attr/mod.rs

@@ -327,7 +327,8 @@ impl MetaItem {
         I: Iterator<Item = &'a TokenTree>,
     {
         // FIXME: Share code with `parse_path`.
-        let path = match tokens.next().map(|tt| TokenTree::uninterpolate(tt)).as_deref() {
+        let tt = tokens.next().map(|tt| TokenTree::uninterpolate(tt));
+        let path = match tt.as_deref() {
             Some(&TokenTree::Token(
                 Token { kind: ref kind @ (token::Ident(..) | token::PathSep), span },
                 _,
@@ -368,6 +369,12 @@ impl MetaItem {
                 token::Nonterminal::NtPath(path) => (**path).clone(),
                 _ => return None,
             },
+            Some(TokenTree::Token(
+                Token { kind: token::OpenDelim(_) | token::CloseDelim(_), .. },
+                _,
+            )) => {
+                panic!("Should be `AttrTokenTree::Delimited`, not delim tokens: {:?}", tt);
+            }
             _ => return None,
         };
         let list_closing_paren_pos = tokens.peek().map(|tt| tt.span().hi());

compiler/rustc_ast/src/tokenstream.rs

@@ -224,7 +224,7 @@ impl AttrTokenStream {
                                 // Inner attributes are only supported on extern blocks, functions,
                                 // impls, and modules. All of these have their inner attributes
                                 // placed at the beginning of the rightmost outermost braced group:
-                                // e.g. fn foo() { #![my_attr} }
+                                // e.g. fn foo() { #![my_attr] }
                                 //
                                 // Therefore, we can insert them back into the right location
                                 // without needing to do any extra position tracking.

compiler/rustc_codegen_ssa/src/codegen_attrs.rs

@@ -124,22 +124,6 @@ fn codegen_fn_attrs(tcx: TyCtxt<'_>, did: LocalDefId) -> CodegenFnAttrs {
                         .emit();
                 }
             }
-            sym::coverage => {
-                let inner = attr.meta_item_list();
-                match inner.as_deref() {
-                    Some([item]) if item.has_name(sym::off) => {
-                        codegen_fn_attrs.flags |= CodegenFnAttrFlags::NO_COVERAGE;
-                    }
-                    Some([item]) if item.has_name(sym::on) => {
-                        // Allow #[coverage(on)] for being explicit, maybe also in future to enable
-                        // coverage on a smaller scope within an excluded larger scope.
-                    }
-                    Some(_) | None => {
-                        tcx.dcx()
-                            .span_delayed_bug(attr.span, "unexpected value of coverage attribute");
-                    }
-                }
-            }
             sym::rustc_std_internal_symbol => {
                 codegen_fn_attrs.flags |= CodegenFnAttrFlags::RUSTC_STD_INTERNAL_SYMBOL
             }
@@ -584,7 +568,6 @@ fn codegen_fn_attrs(tcx: TyCtxt<'_>, did: LocalDefId) -> CodegenFnAttrs {
     }
 
     if codegen_fn_attrs.flags.contains(CodegenFnAttrFlags::NAKED) {
-        codegen_fn_attrs.flags |= CodegenFnAttrFlags::NO_COVERAGE;
         codegen_fn_attrs.inline = InlineAttr::Never;
     }
 

compiler/rustc_expand/src/config.rs

@@ -214,6 +214,12 @@ impl<'a> StripUnconfigured<'a> {
                 ) => {
                     panic!("Nonterminal should have been flattened: {:?}", tree);
                 }
+                AttrTokenTree::Token(
+                    Token { kind: TokenKind::OpenDelim(_) | TokenKind::CloseDelim(_), .. },
+                    _,
+                ) => {
+                    panic!("Should be `AttrTokenTree::Delimited`, not delim tokens: {:?}", tree);
+                }
                 AttrTokenTree::Token(token, spacing) => {
                     Some(AttrTokenTree::Token(token, spacing)).into_iter()
                 }

compiler/rustc_middle/src/middle/codegen_fn_attrs.rs

@@ -87,10 +87,7 @@ bitflags::bitflags! {
         /// #[cmse_nonsecure_entry]: with a TrustZone-M extension, declare a
         /// function as an entry function from Non-Secure code.
         const CMSE_NONSECURE_ENTRY      = 1 << 13;
-        /// `#[coverage(off)]`: indicates that the function should be ignored by
-        /// the MIR `InstrumentCoverage` pass and not added to the coverage map
-        /// during codegen.
-        const NO_COVERAGE               = 1 << 14;
+        // (Bit 14 was used for `#[coverage(off)]`, but is now unused.)
         /// `#[used(linker)]`:
         /// indicates that neither LLVM nor the linker will eliminate this function.
         const USED_LINKER               = 1 << 15;

compiler/rustc_middle/src/query/mod.rs

@@ -572,6 +572,15 @@ rustc_queries! {
         separate_provide_extern
     }
 
+    /// Checks for the nearest `#[coverage(off)]` or `#[coverage(on)]` on
+    /// this def and any enclosing defs, up to the crate root.
+    ///
+    /// Returns `false` if `#[coverage(off)]` was found, or `true` if
+    /// either `#[coverage(on)]` or no coverage attribute was found.
+    query coverage_attr_on(key: LocalDefId) -> bool {
+        desc { |tcx| "checking for `#[coverage(..)]` on `{}`", tcx.def_path_str(key) }
+    }
+
     /// Summarizes coverage IDs inserted by the `InstrumentCoverage` MIR pass
     /// (for compiler option `-Cinstrument-coverage`), after MIR optimizations
     /// have had a chance to potentially remove some of them.

compiler/rustc_mir_transform/src/coverage/query.rs

@@ -6,11 +6,13 @@ use rustc_middle::query::TyCtxtAt;
 use rustc_middle::ty::{self, TyCtxt};
 use rustc_middle::util::Providers;
 use rustc_span::def_id::LocalDefId;
+use rustc_span::sym;
 
 /// Registers query/hook implementations related to coverage.
 pub(crate) fn provide(providers: &mut Providers) {
     providers.hooks.is_eligible_for_coverage =
         |TyCtxtAt { tcx, .. }, def_id| is_eligible_for_coverage(tcx, def_id);
+    providers.queries.coverage_attr_on = coverage_attr_on;
     providers.queries.coverage_ids_info = coverage_ids_info;
 }
 
@@ -38,14 +40,43 @@ fn is_eligible_for_coverage(tcx: TyCtxt<'_>, def_id: LocalDefId) -> bool {
         return false;
     }
 
-    if tcx.codegen_fn_attrs(def_id).flags.contains(CodegenFnAttrFlags::NO_COVERAGE) {
+    if tcx.codegen_fn_attrs(def_id).flags.contains(CodegenFnAttrFlags::NAKED) {
+        trace!("InstrumentCoverage skipped for {def_id:?} (`#[naked]`)");
+        return false;
+    }
+
+    if !tcx.coverage_attr_on(def_id) {
         trace!("InstrumentCoverage skipped for {def_id:?} (`#[coverage(off)]`)");
         return false;
     }
 
     true
 }
 
+/// Query implementation for `coverage_attr_on`.
+fn coverage_attr_on(tcx: TyCtxt<'_>, def_id: LocalDefId) -> bool {
+    // Check for annotations directly on this def.
+    if let Some(attr) = tcx.get_attr(def_id, sym::coverage) {
+        match attr.meta_item_list().as_deref() {
+            Some([item]) if item.has_name(sym::off) => return false,
+            Some([item]) if item.has_name(sym::on) => return true,
+            Some(_) | None => {
+                // Other possibilities should have been rejected by `rustc_parse::validate_attr`.
+                tcx.dcx().span_bug(attr.span, "unexpected value of coverage attribute");
+            }
+        }
+    }
+
+    match tcx.opt_local_parent(def_id) {
+        // Check the parent def (and so on recursively) until we find an
+        // enclosing attribute or reach the crate root.
+        Some(parent) => tcx.coverage_attr_on(parent),
+        // We reached the crate root without seeing a coverage attribute, so
+        // allow coverage instrumentation by default.
+        None => true,
+    }
+}
+
 /// Query implementation for `coverage_ids_info`.
 fn coverage_ids_info<'tcx>(
     tcx: TyCtxt<'tcx>,

compiler/rustc_parse/src/parser/expr.rs

@@ -39,36 +39,6 @@ use rustc_span::{BytePos, ErrorGuaranteed, Pos, Span};
 use thin_vec::{thin_vec, ThinVec};
 use tracing::instrument;
 
-/// Possibly accepts an `token::Interpolated` expression (a pre-parsed expression
-/// dropped into the token stream, which happens while parsing the result of
-/// macro expansion). Placement of these is not as complex as I feared it would
-/// be. The important thing is to make sure that lookahead doesn't balk at
-/// `token::Interpolated` tokens.
-macro_rules! maybe_whole_expr {
-    ($p:expr) => {
-        if let token::Interpolated(nt) = &$p.token.kind {
-            match &**nt {
-                token::NtExpr(e) | token::NtLiteral(e) => {
-                    let e = e.clone();
-                    $p.bump();
-                    return Ok(e);
-                }
-                token::NtPath(path) => {
-                    let path = (**path).clone();
-                    $p.bump();
-                    return Ok($p.mk_expr($p.prev_token.span, ExprKind::Path(None, path)));
-                }
-                token::NtBlock(block) => {
-                    let block = block.clone();
-                    $p.bump();
-                    return Ok($p.mk_expr($p.prev_token.span, ExprKind::Block(block, None)));
-                }
-                _ => {}
-            };
-        }
-    };
-}
-
 #[derive(Debug)]
 pub(super) enum LhsExpr {
     // Already parsed just the outer attributes.
@@ -1421,7 +1391,27 @@ impl<'a> Parser<'a> {
     /// correctly if called from `parse_dot_or_call_expr()`.
     fn parse_expr_bottom(&mut self) -> PResult<'a, P<Expr>> {
         maybe_recover_from_interpolated_ty_qpath!(self, true);
-        maybe_whole_expr!(self);
+
+        if let token::Interpolated(nt) = &self.token.kind {
+            match &**nt {
+                token::NtExpr(e) | token::NtLiteral(e) => {
+                    let e = e.clone();
+                    self.bump();
+                    return Ok(e);
+                }
+                token::NtPath(path) => {
+                    let path = (**path).clone();
+                    self.bump();
+                    return Ok(self.mk_expr(self.prev_token.span, ExprKind::Path(None, path)));
+                }
+                token::NtBlock(block) => {
+                    let block = block.clone();
+                    self.bump();
+                    return Ok(self.mk_expr(self.prev_token.span, ExprKind::Block(block, None)));
+                }
+                _ => {}
+            };
+        }
 
         // Outer attributes are already parsed and will be
         // added to the return value after the fact.
@@ -2190,7 +2180,26 @@ impl<'a> Parser<'a> {
     /// Matches `'-' lit | lit` (cf. `ast_validation::AstValidator::check_expr_within_pat`).
     /// Keep this in sync with `Token::can_begin_literal_maybe_minus`.
     pub fn parse_literal_maybe_minus(&mut self) -> PResult<'a, P<Expr>> {
-        maybe_whole_expr!(self);
+        if let token::Interpolated(nt) = &self.token.kind {
+            match &**nt {
+                // FIXME(nnethercote) The `NtExpr` case should only match if
+                // `e` is an `ExprKind::Lit` or an `ExprKind::Unary` containing
+                // an `UnOp::Neg` and an `ExprKind::Lit`, like how
+                // `can_begin_literal_maybe_minus` works. But this method has
+                // been over-accepting for a long time, and to make that change
+                // here requires also changing some `parse_literal_maybe_minus`
+                // call sites to accept additional expression kinds. E.g.
+                // `ExprKind::Path` must be accepted when parsing range
+                // patterns. That requires some care. So for now, we continue
+                // being less strict here than we should be.
+                token::NtExpr(e) | token::NtLiteral(e) => {
+                    let e = e.clone();
+                    self.bump();
+                    return Ok(e);
+                }
+                _ => {}
+            };
+        }
 
         let lo = self.token.span;
         let minus_present = self.eat(&token::BinOp(token::Minus));

compiler/rustc_passes/src/check_attr.rs

@@ -369,13 +369,16 @@ impl<'tcx> CheckAttrVisitor<'tcx> {
         }
     }
 
-    /// Checks that `#[coverage(..)]` is applied to a function or closure.
+    /// Checks that `#[coverage(..)]` is applied to a function/closure/method,
+    /// or to an impl block or module.
     fn check_coverage(&self, attr: &Attribute, span: Span, target: Target) -> bool {
         match target {
-            // #[coverage(..)] on function is fine
             Target::Fn
             | Target::Closure
-            | Target::Method(MethodKind::Trait { body: true } | MethodKind::Inherent) => true,
+            | Target::Method(MethodKind::Trait { body: true } | MethodKind::Inherent)
+            | Target::Impl
+            | Target::Mod => true,
+
             _ => {
                 self.dcx().emit_err(errors::CoverageNotFnOrClosure {
                     attr_span: attr.span,

compiler/rustc_smir/src/rustc_internal/internal.rs

@@ -188,8 +188,10 @@ impl RustcInternal for FloatTy {
 
     fn internal<'tcx>(&self, _tables: &mut Tables<'_>, _tcx: TyCtxt<'tcx>) -> Self::T<'tcx> {
         match self {
+            FloatTy::F16 => rustc_ty::FloatTy::F16,
             FloatTy::F32 => rustc_ty::FloatTy::F32,
             FloatTy::F64 => rustc_ty::FloatTy::F64,
+            FloatTy::F128 => rustc_ty::FloatTy::F128,
         }
     }
 }

compiler/rustc_smir/src/rustc_smir/convert/ty.rs

@@ -304,10 +304,10 @@ impl<'tcx> Stable<'tcx> for ty::FloatTy {
 
     fn stable(&self, _: &mut Tables<'_>) -> Self::T {
         match self {
-            ty::FloatTy::F16 => unimplemented!("f16_f128"),
+            ty::FloatTy::F16 => FloatTy::F16,
             ty::FloatTy::F32 => FloatTy::F32,
             ty::FloatTy::F64 => FloatTy::F64,
-            ty::FloatTy::F128 => unimplemented!("f16_f128"),
+            ty::FloatTy::F128 => FloatTy::F128,
         }
     }
 }

compiler/stable_mir/src/ty.rs

@@ -608,8 +608,10 @@ impl UintTy {
 
 #[derive(Clone, Copy, Debug, PartialEq, Eq)]
 pub enum FloatTy {
+    F16,
     F32,
     F64,
+    F128,
 }
 
 #[derive(Clone, Copy, Debug, PartialEq, Eq)]

library/alloc/src/macros.rs

@@ -41,18 +41,18 @@
 #[allow_internal_unstable(rustc_attrs, liballoc_internals)]
 macro_rules! vec {
     () => (
-        $crate::__rust_force_expr!($crate::vec::Vec::new())
+        $crate::vec::Vec::new()
     );
     ($elem:expr; $n:expr) => (
-        $crate::__rust_force_expr!($crate::vec::from_elem($elem, $n))
+        $crate::vec::from_elem($elem, $n)
     );
     ($($x:expr),+ $(,)?) => (
-        $crate::__rust_force_expr!(<[_]>::into_vec(
+        <[_]>::into_vec(
             // This rustc_box is not required, but it produces a dramatic improvement in compile
             // time when constructing arrays with many elements.
             #[rustc_box]
             $crate::boxed::Box::new([$($x),+])
-        ))
+        )
     );
 }
 
@@ -126,13 +126,3 @@ macro_rules! format {
         res
     }}
 }
-
-/// Force AST node to an expression to improve diagnostics in pattern position.
-#[doc(hidden)]
-#[macro_export]
-#[unstable(feature = "liballoc_internals", issue = "none", reason = "implementation detail")]
-macro_rules! __rust_force_expr {
-    ($e:expr) => {
-        $e
-    };
-}

library/std/src/sys_common/wtf8.rs

@@ -480,7 +480,7 @@ impl Wtf8Buf {
     #[inline]
     pub(crate) fn extend_from_slice(&mut self, other: &[u8]) {
         self.bytes.extend_from_slice(other);
-        self.is_known_utf8 = self.is_known_utf8 || self.next_surrogate(0).is_none();
+        self.is_known_utf8 = false;
     }
 }
 

library/std/src/sys_common/wtf8/tests.rs

@@ -725,3 +725,27 @@ fn wtf8_utf8_boundary_between_surrogates() {
     string.push(CodePoint::from_u32(0xD800).unwrap());
     check_utf8_boundary(&string, 3);
 }
+
+#[test]
+fn wobbled_wtf8_plus_bytes_isnt_utf8() {
+    let mut string: Wtf8Buf = unsafe { Wtf8::from_bytes_unchecked(b"\xED\xA0\x80").to_owned() };
+    assert!(!string.is_known_utf8);
+    string.extend_from_slice(b"some utf-8");
+    assert!(!string.is_known_utf8);
+}
+
+#[test]
+fn wobbled_wtf8_plus_str_isnt_utf8() {
+    let mut string: Wtf8Buf = unsafe { Wtf8::from_bytes_unchecked(b"\xED\xA0\x80").to_owned() };
+    assert!(!string.is_known_utf8);
+    string.push_str("some utf-8");
+    assert!(!string.is_known_utf8);
+}
+
+#[test]
+fn unwobbly_wtf8_plus_utf8_is_utf8() {
+    let mut string: Wtf8Buf = Wtf8Buf::from_str("hello world");
+    assert!(string.is_known_utf8);
+    string.push_str("some utf-8");
+    assert!(string.is_known_utf8);
+}