Rollup merge of rust-lang#111543 - Urgau:uplift_invalid_utf8_in_unchecked, r=WaffleLapkin

Uplift `clippy::invalid_utf8_in_unchecked` lint

This PR aims at uplifting the `clippy::invalid_utf8_in_unchecked` lint into two lints.

## `invalid_from_utf8_unchecked`

(deny-by-default)

The `invalid_from_utf8_unchecked` lint checks for calls to `std::str::from_utf8_unchecked` and `std::str::from_utf8_unchecked_mut` with an invalid UTF-8 literal.

### Example

```rust
unsafe {
    std::str::from_utf8_unchecked(b"cl\x82ippy");
}
```

### Explanation

Creating such a `str` would result in undefined behavior as per documentation for `std::str::from_utf8_unchecked` and `std::str::from_utf8_unchecked_mut`.

## `invalid_from_utf8`

(warn-by-default)

The `invalid_from_utf8` lint checks for calls to `std::str::from_utf8` and `std::str::from_utf8_mut` with an invalid UTF-8 literal.

### Example

```rust
std::str::from_utf8(b"ru\x82st");
```

### Explanation

Trying to create such a `str` would always return an error as per documentation for `std::str::from_utf8` and `std::str::from_utf8_mut`.

-----

Mostly followed the instructions for uplifting a clippy lint described here: rust-lang#99696 (review)

````@rustbot```` label: +I-lang-nominated
r? compiler

-----

For Clippy:

changelog: Moves: Uplifted `clippy::invalid_utf8_in_unchecked` into rustc

Author: Noratrieb

compiler/rustc_lint/messages.ftl

@@ -304,6 +304,14 @@ lint_improper_ctypes_union_layout_help = consider adding a `#[repr(C)]` or `#[re
 lint_improper_ctypes_union_layout_reason = this union has unspecified layout
 lint_improper_ctypes_union_non_exhaustive = this union is non-exhaustive
 
+# FIXME: we should ordinalize $valid_up_to when we add support for doing so
+lint_invalid_from_utf8_checked = calls to `{$method}` with a invalid literal always return an error
+    .label = the literal was valid UTF-8 up to the {$valid_up_to} bytes
+
+# FIXME: we should ordinalize $valid_up_to when we add support for doing so
+lint_invalid_from_utf8_unchecked = calls to `{$method}` with a invalid literal are undefined behavior
+    .label = the literal was valid UTF-8 up to the {$valid_up_to} bytes
+
 lint_lintpass_by_hand = implementing `LintPass` by hand
     .help = try using `declare_lint_pass!` or `impl_lint_pass!` instead
 

compiler/rustc_lint/src/invalid_from_utf8.rs

@@ -0,0 +1,118 @@
+use std::str::Utf8Error;
+
+use rustc_ast::{BorrowKind, LitKind};
+use rustc_hir::{Expr, ExprKind};
+use rustc_span::source_map::Spanned;
+use rustc_span::sym;
+
+use crate::lints::InvalidFromUtf8Diag;
+use crate::{LateContext, LateLintPass, LintContext};
+
+declare_lint! {
+    /// The `invalid_from_utf8_unchecked` lint checks for calls to
+    /// `std::str::from_utf8_unchecked` and `std::str::from_utf8_unchecked_mut`
+    /// with an invalid UTF-8 literal.
+    ///
+    /// ### Example
+    ///
+    /// ```rust,compile_fail
+    /// # #[allow(unused)]
+    /// unsafe {
+    ///     std::str::from_utf8_unchecked(b"Ru\x82st");
+    /// }
+    /// ```
+    ///
+    /// {{produces}}
+    ///
+    /// ### Explanation
+    ///
+    /// Creating such a `str` would result in undefined behavior as per documentation
+    /// for `std::str::from_utf8_unchecked` and `std::str::from_utf8_unchecked_mut`.
+    pub INVALID_FROM_UTF8_UNCHECKED,
+    Deny,
+    "using a non UTF-8 literal in `std::str::from_utf8_unchecked`"
+}
+
+declare_lint! {
+    /// The `invalid_from_utf8` lint checks for calls to
+    /// `std::str::from_utf8` and `std::str::from_utf8_mut`
+    /// with an invalid UTF-8 literal.
+    ///
+    /// ### Example
+    ///
+    /// ```rust
+    /// # #[allow(unused)]
+    /// std::str::from_utf8(b"Ru\x82st");
+    /// ```
+    ///
+    /// {{produces}}
+    ///
+    /// ### Explanation
+    ///
+    /// Trying to create such a `str` would always return an error as per documentation
+    /// for `std::str::from_utf8` and `std::str::from_utf8_mut`.
+    pub INVALID_FROM_UTF8,
+    Warn,
+    "using a non UTF-8 literal in `std::str::from_utf8`"
+}
+
+declare_lint_pass!(InvalidFromUtf8 => [INVALID_FROM_UTF8_UNCHECKED, INVALID_FROM_UTF8]);
+
+impl<'tcx> LateLintPass<'tcx> for InvalidFromUtf8 {
+    fn check_expr(&mut self, cx: &LateContext<'tcx>, expr: &'tcx Expr<'tcx>) {
+        if let ExprKind::Call(path, [arg]) = expr.kind
+            && let ExprKind::Path(ref qpath) = path.kind
+            && let Some(def_id) = cx.qpath_res(qpath, path.hir_id).opt_def_id()
+            && let Some(diag_item) = cx.tcx.get_diagnostic_name(def_id)
+            && [sym::str_from_utf8, sym::str_from_utf8_mut,
+                sym::str_from_utf8_unchecked, sym::str_from_utf8_unchecked_mut].contains(&diag_item)
+        {
+            let lint = |utf8_error: Utf8Error| {
+                let label = arg.span;
+                let method = diag_item.as_str().strip_prefix("str_").unwrap();
+                let method = format!("std::str::{method}");
+                let valid_up_to = utf8_error.valid_up_to();
+                let is_unchecked_variant = diag_item.as_str().contains("unchecked");
+
+                cx.emit_spanned_lint(
+                    if is_unchecked_variant { INVALID_FROM_UTF8_UNCHECKED } else { INVALID_FROM_UTF8 },
+                    expr.span,
+                    if is_unchecked_variant {
+                        InvalidFromUtf8Diag::Unchecked { method,  valid_up_to, label }
+                    } else {
+                        InvalidFromUtf8Diag::Checked { method,  valid_up_to, label }
+                    }
+                )
+            };
+
+            match &arg.kind {
+                ExprKind::Lit(Spanned { node: lit, .. }) => {
+                    if let LitKind::ByteStr(bytes, _) = &lit
+                        && let Err(utf8_error) = std::str::from_utf8(bytes)
+                    {
+                        lint(utf8_error);
+                    }
+                },
+                ExprKind::AddrOf(BorrowKind::Ref, _, Expr { kind: ExprKind::Array(args), .. }) => {
+                    let elements = args.iter().map(|e|{
+                        match &e.kind {
+                            ExprKind::Lit(Spanned { node: lit, .. }) => match lit {
+                                LitKind::Byte(b) => Some(*b),
+                                LitKind::Int(b, _) => Some(*b as u8),
+                                _ => None
+                            }
+                            _ => None
+                        }
+                    }).collect::<Option<Vec<_>>>();
+
+                    if let Some(elements) = elements
+                        && let Err(utf8_error) = std::str::from_utf8(&elements)
+                    {
+                        lint(utf8_error);
+                    }
+                }
+                _ => {}
+            }
+        }
+    }
+}

compiler/rustc_lint/src/lib.rs

@@ -60,6 +60,7 @@ mod expect;
 mod for_loops_over_fallibles;
 pub mod hidden_unicode_codepoints;
 mod internal;
+mod invalid_from_utf8;
 mod late;
 mod let_underscore;
 mod levels;
@@ -102,6 +103,7 @@ use enum_intrinsics_non_enums::EnumIntrinsicsNonEnums;
 use for_loops_over_fallibles::*;
 use hidden_unicode_codepoints::*;
 use internal::*;
+use invalid_from_utf8::*;
 use let_underscore::*;
 use map_unit_fn::*;
 use methods::*;
@@ -207,6 +209,7 @@ late_lint_methods!(
             HardwiredLints: HardwiredLints,
             ImproperCTypesDeclarations: ImproperCTypesDeclarations,
             ImproperCTypesDefinitions: ImproperCTypesDefinitions,
+            InvalidFromUtf8: InvalidFromUtf8,
             VariantSizeDifferences: VariantSizeDifferences,
             BoxPointers: BoxPointers,
             PathStatements: PathStatements,

compiler/rustc_lint/src/lints.rs

@@ -699,6 +699,25 @@ pub struct ForgetCopyDiag<'a> {
     pub label: Span,
 }
 
+// invalid_from_utf8.rs
+#[derive(LintDiagnostic)]
+pub enum InvalidFromUtf8Diag {
+    #[diag(lint_invalid_from_utf8_unchecked)]
+    Unchecked {
+        method: String,
+        valid_up_to: usize,
+        #[label]
+        label: Span,
+    },
+    #[diag(lint_invalid_from_utf8_checked)]
+    Checked {
+        method: String,
+        valid_up_to: usize,
+        #[label]
+        label: Span,
+    },
+}
+
 // hidden_unicode_codepoints.rs
 #[derive(LintDiagnostic)]
 #[diag(lint_hidden_unicode_codepoints)]

compiler/rustc_span/src/symbol.rs

@@ -1454,6 +1454,10 @@ symbols! {
         stop_after_dataflow,
         store,
         str,
+        str_from_utf8,
+        str_from_utf8_mut,
+        str_from_utf8_unchecked,
+        str_from_utf8_unchecked_mut,
         str_split_whitespace,
         str_trim,
         str_trim_end,

library/alloc/tests/str.rs

@@ -1,3 +1,5 @@
+#![cfg_attr(not(bootstrap), allow(invalid_from_utf8))]
+
 use std::assert_matches::assert_matches;
 use std::borrow::Cow;
 use std::cmp::Ordering::{Equal, Greater, Less};

library/core/src/str/converts.rs

@@ -84,6 +84,7 @@ use super::Utf8Error;
 #[stable(feature = "rust1", since = "1.0.0")]
 #[rustc_const_stable(feature = "const_str_from_utf8_shared", since = "1.63.0")]
 #[rustc_allow_const_fn_unstable(str_internals)]
+#[rustc_diagnostic_item = "str_from_utf8"]
 pub const fn from_utf8(v: &[u8]) -> Result<&str, Utf8Error> {
     // FIXME: This should use `?` again, once it's `const`
     match run_utf8_validation(v) {
@@ -127,6 +128,7 @@ pub const fn from_utf8(v: &[u8]) -> Result<&str, Utf8Error> {
 /// errors that can be returned.
 #[stable(feature = "str_mut_extras", since = "1.20.0")]
 #[rustc_const_unstable(feature = "const_str_from_utf8", issue = "91006")]
+#[rustc_diagnostic_item = "str_from_utf8_mut"]
 pub const fn from_utf8_mut(v: &mut [u8]) -> Result<&mut str, Utf8Error> {
     // This should use `?` again, once it's `const`
     match run_utf8_validation(v) {
@@ -167,6 +169,7 @@ pub const fn from_utf8_mut(v: &mut [u8]) -> Result<&mut str, Utf8Error> {
 #[must_use]
 #[stable(feature = "rust1", since = "1.0.0")]
 #[rustc_const_stable(feature = "const_str_from_utf8_unchecked", since = "1.55.0")]
+#[rustc_diagnostic_item = "str_from_utf8_unchecked"]
 pub const unsafe fn from_utf8_unchecked(v: &[u8]) -> &str {
     // SAFETY: the caller must guarantee that the bytes `v` are valid UTF-8.
     // Also relies on `&str` and `&[u8]` having the same layout.
@@ -194,6 +197,7 @@ pub const unsafe fn from_utf8_unchecked(v: &[u8]) -> &str {
 #[must_use]
 #[stable(feature = "str_mut_extras", since = "1.20.0")]
 #[rustc_const_unstable(feature = "const_str_from_utf8_unchecked_mut", issue = "91005")]
+#[rustc_diagnostic_item = "str_from_utf8_unchecked_mut"]
 pub const unsafe fn from_utf8_unchecked_mut(v: &mut [u8]) -> &mut str {
     // SAFETY: the caller must guarantee that the bytes `v`
     // are valid UTF-8, thus the cast to `*mut str` is safe.

src/tools/clippy/clippy_lints/src/declared_lints.rs

@@ -212,7 +212,6 @@ pub(crate) static LINTS: &[&crate::LintInfo] = &[
     crate::instant_subtraction::UNCHECKED_DURATION_SUBTRACTION_INFO,
     crate::int_plus_one::INT_PLUS_ONE_INFO,
     crate::invalid_upcast_comparisons::INVALID_UPCAST_COMPARISONS_INFO,
-    crate::invalid_utf8_in_unchecked::INVALID_UTF8_IN_UNCHECKED_INFO,
     crate::items_after_statements::ITEMS_AFTER_STATEMENTS_INFO,
     crate::items_after_test_module::ITEMS_AFTER_TEST_MODULE_INFO,
     crate::iter_not_returning_iterator::ITER_NOT_RETURNING_ITERATOR_INFO,

src/tools/clippy/clippy_lints/src/invalid_utf8_in_unchecked.rs

@@ -157,7 +157,6 @@ mod inline_fn_without_body;
 mod instant_subtraction;
 mod int_plus_one;
 mod invalid_upcast_comparisons;
-mod invalid_utf8_in_unchecked;
 mod items_after_statements;
 mod items_after_test_module;
 mod iter_not_returning_iterator;
@@ -937,7 +936,6 @@ pub fn register_plugins(store: &mut rustc_lint::LintStore, sess: &Session, conf:
     store.register_late_pass(move |_| Box::new(manual_retain::ManualRetain::new(msrv())));
     let verbose_bit_mask_threshold = conf.verbose_bit_mask_threshold;
     store.register_late_pass(move |_| Box::new(operators::Operators::new(verbose_bit_mask_threshold)));
-    store.register_late_pass(|_| Box::new(invalid_utf8_in_unchecked::InvalidUtf8InUnchecked));
     store.register_late_pass(|_| Box::<std_instead_of_core::StdReexports>::default());
     store.register_late_pass(move |_| Box::new(instant_subtraction::InstantSubtraction::new(msrv())));
     store.register_late_pass(|_| Box::new(partialeq_to_none::PartialeqToNone));