Auto merge of #94963 - lcnr:inherent-impls-std, r=oli-obk,m-ou-se

allow arbitrary inherent impls for builtin types in core

Part of rust-lang/compiler-team#487. Slightly adjusted after some talks with `@m-ou-se` about the requirements of `t-libs-api`.

This adds a crate attribute `#![rustc_coherence_is_core]` which allows arbitrary impls for builtin types in core.

For other library crates impls for builtin types should be avoided if possible. We do have to allow the existing stable impls however. To prevent us from accidentally adding more of these in the future, there is a second attribute `#[rustc_allow_incoherent_impl]` which has to be added to **all impl items**. This only supports impls for builtin types but can easily be extended to additional types in a future PR.

This implementation does not check for overlaps in these impls. Perfectly checking that requires us to check the coherence of these incoherent impls in every crate, as two distinct dependencies may add overlapping methods. It should be easy enough to detect if it goes wrong and the attribute is only intended for use inside of std.

The first two commits are mostly unrelated cleanups.

Author: bors

alloc/src/slice.rs

@@ -237,7 +237,7 @@ mod hack {
     }
 }
 
-#[lang = "slice_alloc"]
+#[cfg_attr(bootstrap, lang = "slice_alloc")]
 #[cfg(not(test))]
 impl<T> [T] {
     /// Sorts the slice.
@@ -267,6 +267,7 @@ impl<T> [T] {
     /// assert!(v == [-5, -3, 1, 2, 4]);
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "rust1", since = "1.0.0")]
     #[inline]
     pub fn sort(&mut self)
@@ -322,6 +323,7 @@ impl<T> [T] {
     /// assert!(v == [5, 4, 3, 2, 1]);
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "rust1", since = "1.0.0")]
     #[inline]
     pub fn sort_by<F>(&mut self, mut compare: F)
@@ -363,6 +365,7 @@ impl<T> [T] {
     /// assert!(v == [1, 2, -3, 4, -5]);
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "slice_sort_by_key", since = "1.7.0")]
     #[inline]
     pub fn sort_by_key<K, F>(&mut self, mut f: F)
@@ -409,6 +412,7 @@ impl<T> [T] {
     ///
     /// [pdqsort]: https://github.com/orlp/pdqsort
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "slice_sort_by_cached_key", since = "1.34.0")]
     #[inline]
     pub fn sort_by_cached_key<K, F>(&mut self, f: F)
@@ -467,6 +471,7 @@ impl<T> [T] {
     /// // Here, `s` and `x` can be modified independently.
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[rustc_conversion_suggestion]
     #[stable(feature = "rust1", since = "1.0.0")]
     #[inline]
@@ -491,6 +496,7 @@ impl<T> [T] {
     /// // Here, `s` and `x` can be modified independently.
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[inline]
     #[unstable(feature = "allocator_api", issue = "32838")]
     pub fn to_vec_in<A: Allocator>(&self, alloc: A) -> Vec<T, A>
@@ -515,6 +521,7 @@ impl<T> [T] {
     ///
     /// assert_eq!(x, vec![10, 40, 30]);
     /// ```
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "rust1", since = "1.0.0")]
     #[inline]
     pub fn into_vec<A: Allocator>(self: Box<Self, A>) -> Vec<T, A> {
@@ -542,6 +549,7 @@ impl<T> [T] {
     /// // this will panic at runtime
     /// b"0123456789abcdef".repeat(usize::MAX);
     /// ```
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[cfg(not(no_global_oom_handling))]
     #[stable(feature = "repeat_generic_slice", since = "1.40.0")]
     pub fn repeat(&self, n: usize) -> Vec<T>
@@ -610,6 +618,7 @@ impl<T> [T] {
     /// assert_eq!(["hello", "world"].concat(), "helloworld");
     /// assert_eq!([[1, 2], [3, 4]].concat(), [1, 2, 3, 4]);
     /// ```
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "rust1", since = "1.0.0")]
     pub fn concat<Item: ?Sized>(&self) -> <Self as Concat<Item>>::Output
     where
@@ -628,6 +637,7 @@ impl<T> [T] {
     /// assert_eq!([[1, 2], [3, 4]].join(&0), [1, 2, 0, 3, 4]);
     /// assert_eq!([[1, 2], [3, 4]].join(&[0, 0][..]), [1, 2, 0, 0, 3, 4]);
     /// ```
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "rename_connect_to_join", since = "1.3.0")]
     pub fn join<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output
     where
@@ -646,6 +656,7 @@ impl<T> [T] {
     /// assert_eq!(["hello", "world"].connect(" "), "hello world");
     /// assert_eq!([[1, 2], [3, 4]].connect(&0), [1, 2, 0, 3, 4]);
     /// ```
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "rust1", since = "1.0.0")]
     #[rustc_deprecated(since = "1.3.0", reason = "renamed to join")]
     pub fn connect<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output
@@ -656,7 +667,7 @@ impl<T> [T] {
     }
 }
 
-#[lang = "slice_u8_alloc"]
+#[cfg_attr(bootstrap, lang = "slice_u8_alloc")]
 #[cfg(not(test))]
 impl [u8] {
     /// Returns a vector containing a copy of this slice where each byte
@@ -669,6 +680,7 @@ impl [u8] {
     ///
     /// [`make_ascii_uppercase`]: slice::make_ascii_uppercase
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "this returns the uppercase bytes as a new Vec, \
                   without modifying the original"]
     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
@@ -689,6 +701,7 @@ impl [u8] {
     ///
     /// [`make_ascii_lowercase`]: slice::make_ascii_lowercase
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "this returns the lowercase bytes as a new Vec, \
                   without modifying the original"]
     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]

alloc/src/str.rs

@@ -235,7 +235,7 @@ impl ToOwned for str {
 }
 
 /// Methods for string slices.
-#[lang = "str_alloc"]
+#[cfg_attr(bootstrap, lang = "str_alloc")]
 #[cfg(not(test))]
 impl str {
     /// Converts a `Box<str>` into a `Box<[u8]>` without copying or allocating.
@@ -250,6 +250,7 @@ impl str {
     /// let boxed_bytes = boxed_str.into_boxed_bytes();
     /// assert_eq!(*boxed_bytes, *s.as_bytes());
     /// ```
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[stable(feature = "str_box_extras", since = "1.20.0")]
     #[must_use = "`self` will be dropped if the result is not used"]
     #[inline]
@@ -280,6 +281,7 @@ impl str {
     /// assert_eq!(s, s.replace("cookie monster", "little lamb"));
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "this returns the replaced string as a new allocation, \
                   without modifying the original"]
     #[stable(feature = "rust1", since = "1.0.0")]
@@ -320,6 +322,7 @@ impl str {
     /// assert_eq!(s, s.replacen("cookie monster", "little lamb", 10));
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "this returns the replaced string as a new allocation, \
                   without modifying the original"]
     #[stable(feature = "str_replacen", since = "1.16.0")]
@@ -376,6 +379,7 @@ impl str {
     /// assert_eq!(new_year, new_year.to_lowercase());
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "this returns the lowercase string as a new String, \
                   without modifying the original"]
     #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
@@ -458,6 +462,7 @@ impl str {
     /// assert_eq!("TSCHÜSS", s.to_uppercase());
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "this returns the uppercase string as a new String, \
                   without modifying the original"]
     #[stable(feature = "unicode_case_mapping", since = "1.2.0")]
@@ -493,6 +498,7 @@ impl str {
     /// assert_eq!(boxed_str.into_string(), string);
     /// ```
     #[stable(feature = "box_str", since = "1.4.0")]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "`self` will be dropped if the result is not used"]
     #[inline]
     pub fn into_string(self: Box<str>) -> String {
@@ -521,6 +527,7 @@ impl str {
     /// let huge = "0123456789abcdef".repeat(usize::MAX);
     /// ```
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use]
     #[stable(feature = "repeat_str", since = "1.16.0")]
     pub fn repeat(&self, n: usize) -> String {
@@ -549,6 +556,7 @@ impl str {
     /// [`make_ascii_uppercase`]: str::make_ascii_uppercase
     /// [`to_uppercase`]: #method.to_uppercase
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "to uppercase the value in-place, use `make_ascii_uppercase()`"]
     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
     #[inline]
@@ -581,6 +589,7 @@ impl str {
     /// [`make_ascii_lowercase`]: str::make_ascii_lowercase
     /// [`to_lowercase`]: #method.to_lowercase
     #[cfg(not(no_global_oom_handling))]
+    #[cfg_attr(not(bootstrap), rustc_allow_incoherent_impl)]
     #[must_use = "to lowercase the value in-place, use `make_ascii_lowercase()`"]
     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
     #[inline]

core/src/array/mod.rs

@@ -395,7 +395,7 @@ macro_rules! array_impl_default {
 
 array_impl_default! {32, T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T T}
 
-#[lang = "array"]
+#[cfg_attr(bootstrap, lang = "array")]
 impl<T, const N: usize> [T; N] {
     /// Returns an array of the same size as `self`, with function `f` applied to each element
     /// in order.

core/src/bool.rs

@@ -2,7 +2,7 @@
 
 use crate::marker::Destruct;
 
-#[lang = "bool"]
+#[cfg_attr(bootstrap, lang = "bool")]
 impl bool {
     /// Returns `Some(t)` if the `bool` is [`true`](../std/keyword.true.html),
     /// or `None` otherwise.

core/src/char/methods.rs

@@ -7,7 +7,7 @@ use crate::unicode::{self, conversions};
 
 use super::*;
 
-#[lang = "char"]
+#[cfg_attr(bootstrap, lang = "char")]
 impl char {
     /// The highest valid code point a `char` can have, `'\u{10FFFF}'`.
     ///

core/src/lib.rs

@@ -84,6 +84,7 @@
     target_has_atomic_load_store = "ptr",
 ))]
 #![no_core]
+#![cfg_attr(not(bootstrap), rustc_coherence_is_core)]
 //
 // Lints:
 #![deny(rust_2021_incompatible_or_patterns)]

core/src/num/f32.rs

@@ -370,7 +370,7 @@ pub mod consts {
     pub const LN_10: f32 = 2.30258509299404568401799145468436421_f32;
 }
 
-#[lang = "f32"]
+#[cfg_attr(bootstrap, lang = "f32")]
 #[cfg(not(test))]
 impl f32 {
     /// The radix or base of the internal representation of `f32`.

core/src/num/f64.rs

@@ -370,7 +370,7 @@ pub mod consts {
     pub const LN_10: f64 = 2.30258509299404568401799145468436421_f64;
 }
 
-#[lang = "f64"]
+#[cfg_attr(bootstrap, lang = "f64")]
 #[cfg(not(test))]
 impl f64 {
     /// The radix or base of the internal representation of `f64`.

core/src/num/mod.rs

@@ -193,34 +193,34 @@ macro_rules! widening_impl {
     };
 }
 
-#[lang = "i8"]
+#[cfg_attr(bootstrap, lang = "i8")]
 impl i8 {
     int_impl! { i8, i8, u8, 8, 7, -128, 127, 2, "-0x7e", "0xa", "0x12", "0x12", "0x48",
     "[0x12]", "[0x12]", "", "" }
 }
 
-#[lang = "i16"]
+#[cfg_attr(bootstrap, lang = "i16")]
 impl i16 {
     int_impl! { i16, i16, u16, 16, 15, -32768, 32767, 4, "-0x5ffd", "0x3a", "0x1234", "0x3412",
     "0x2c48", "[0x34, 0x12]", "[0x12, 0x34]", "", "" }
 }
 
-#[lang = "i32"]
+#[cfg_attr(bootstrap, lang = "i32")]
 impl i32 {
     int_impl! { i32, i32, u32, 32, 31, -2147483648, 2147483647, 8, "0x10000b3", "0xb301",
     "0x12345678", "0x78563412", "0x1e6a2c48", "[0x78, 0x56, 0x34, 0x12]",
     "[0x12, 0x34, 0x56, 0x78]", "", "" }
 }
 
-#[lang = "i64"]
+#[cfg_attr(bootstrap, lang = "i64")]
 impl i64 {
     int_impl! { i64, i64, u64, 64, 63, -9223372036854775808, 9223372036854775807, 12,
     "0xaa00000000006e1", "0x6e10aa", "0x1234567890123456", "0x5634129078563412",
     "0x6a2c48091e6a2c48", "[0x56, 0x34, 0x12, 0x90, 0x78, 0x56, 0x34, 0x12]",
     "[0x12, 0x34, 0x56, 0x78, 0x90, 0x12, 0x34, 0x56]", "", "" }
 }
 
-#[lang = "i128"]
+#[cfg_attr(bootstrap, lang = "i128")]
 impl i128 {
     int_impl! { i128, i128, u128, 128, 127, -170141183460469231731687303715884105728,
     170141183460469231731687303715884105727, 16,
@@ -233,15 +233,15 @@ impl i128 {
 }
 
 #[cfg(target_pointer_width = "16")]
-#[lang = "isize"]
+#[cfg_attr(bootstrap, lang = "isize")]
 impl isize {
     int_impl! { isize, i16, usize, 16, 15, -32768, 32767, 4, "-0x5ffd", "0x3a", "0x1234",
     "0x3412", "0x2c48", "[0x34, 0x12]", "[0x12, 0x34]",
     usize_isize_to_xe_bytes_doc!(), usize_isize_from_xe_bytes_doc!() }
 }
 
 #[cfg(target_pointer_width = "32")]
-#[lang = "isize"]
+#[cfg_attr(bootstrap, lang = "isize")]
 impl isize {
     int_impl! { isize, i32, usize, 32, 31, -2147483648, 2147483647, 8, "0x10000b3", "0xb301",
     "0x12345678", "0x78563412", "0x1e6a2c48", "[0x78, 0x56, 0x34, 0x12]",
@@ -250,7 +250,7 @@ impl isize {
 }
 
 #[cfg(target_pointer_width = "64")]
-#[lang = "isize"]
+#[cfg_attr(bootstrap, lang = "isize")]
 impl isize {
     int_impl! { isize, i64, usize, 64, 63, -9223372036854775808, 9223372036854775807,
     12, "0xaa00000000006e1", "0x6e10aa",  "0x1234567890123456", "0x5634129078563412",
@@ -262,7 +262,7 @@ impl isize {
 /// If 6th bit set ascii is upper case.
 const ASCII_CASE_MASK: u8 = 0b0010_0000;
 
-#[lang = "u8"]
+#[cfg_attr(bootstrap, lang = "u8")]
 impl u8 {
     uint_impl! { u8, u8, i8, NonZeroU8, 8, 255, 2, "0x82", "0xa", "0x12", "0x12", "0x48", "[0x12]",
     "[0x12]", "", "" }
@@ -816,7 +816,7 @@ impl u8 {
     }
 }
 
-#[lang = "u16"]
+#[cfg_attr(bootstrap, lang = "u16")]
 impl u16 {
     uint_impl! { u16, u16, i16, NonZeroU16, 16, 65535, 4, "0xa003", "0x3a", "0x1234", "0x3412", "0x2c48",
     "[0x34, 0x12]", "[0x12, 0x34]", "", "" }
@@ -848,14 +848,14 @@ impl u16 {
     }
 }
 
-#[lang = "u32"]
+#[cfg_attr(bootstrap, lang = "u32")]
 impl u32 {
     uint_impl! { u32, u32, i32, NonZeroU32, 32, 4294967295, 8, "0x10000b3", "0xb301", "0x12345678",
     "0x78563412", "0x1e6a2c48", "[0x78, 0x56, 0x34, 0x12]", "[0x12, 0x34, 0x56, 0x78]", "", "" }
     widening_impl! { u32, u64, 32, unsigned }
 }
 
-#[lang = "u64"]
+#[cfg_attr(bootstrap, lang = "u64")]
 impl u64 {
     uint_impl! { u64, u64, i64, NonZeroU64, 64, 18446744073709551615, 12, "0xaa00000000006e1", "0x6e10aa",
     "0x1234567890123456", "0x5634129078563412", "0x6a2c48091e6a2c48",
@@ -865,7 +865,7 @@ impl u64 {
     widening_impl! { u64, u128, 64, unsigned }
 }
 
-#[lang = "u128"]
+#[cfg_attr(bootstrap, lang = "u128")]
 impl u128 {
     uint_impl! { u128, u128, i128, NonZeroU128, 128, 340282366920938463463374607431768211455, 16,
     "0x13f40000000000000000000000004f76", "0x4f7613f4", "0x12345678901234567890123456789012",
@@ -878,15 +878,15 @@ impl u128 {
 }
 
 #[cfg(target_pointer_width = "16")]
-#[lang = "usize"]
+#[cfg_attr(bootstrap, lang = "usize")]
 impl usize {
     uint_impl! { usize, u16, isize, NonZeroUsize, 16, 65535, 4, "0xa003", "0x3a", "0x1234", "0x3412", "0x2c48",
     "[0x34, 0x12]", "[0x12, 0x34]",
     usize_isize_to_xe_bytes_doc!(), usize_isize_from_xe_bytes_doc!() }
     widening_impl! { usize, u32, 16, unsigned }
 }
 #[cfg(target_pointer_width = "32")]
-#[lang = "usize"]
+#[cfg_attr(bootstrap, lang = "usize")]
 impl usize {
     uint_impl! { usize, u32, isize, NonZeroUsize, 32, 4294967295, 8, "0x10000b3", "0xb301", "0x12345678",
     "0x78563412", "0x1e6a2c48", "[0x78, 0x56, 0x34, 0x12]", "[0x12, 0x34, 0x56, 0x78]",
@@ -895,7 +895,7 @@ impl usize {
 }
 
 #[cfg(target_pointer_width = "64")]
-#[lang = "usize"]
+#[cfg_attr(bootstrap, lang = "usize")]
 impl usize {
     uint_impl! { usize, u64, isize, NonZeroUsize, 64, 18446744073709551615, 12, "0xaa00000000006e1", "0x6e10aa",
     "0x1234567890123456", "0x5634129078563412", "0x6a2c48091e6a2c48",