Rollup merge of #66564 - foeb:66219-document-unsafe-sync-cell-str, r=Amanieu

Document unsafe blocks in core::{cell, str, sync}

Split from #66506 (issue #66219). Hopefully doing a chunk at a time is more manageable!

r? @RalfJung

3 files changed, 59 insertions, 30 deletions

diff --git a/src/libcore/str/lossy.rs b/src/libcore/str/lossy.rs
index 8a1fb9de546..88b2bc551b7 100644
--- a/src/libcore/str/lossy.rs
+++ b/src/libcore/str/lossy.rs
@@ -3,8 +3,6 @@ use crate::fmt::{self, Write};
 use crate::mem;
 use crate::str as core_str;
 
-// ignore-tidy-undocumented-unsafe
-
 /// Lossy UTF-8 string.
 #[unstable(feature = "str_internals", issue = "none")]
 pub struct Utf8Lossy {
@@ -17,6 +15,7 @@ impl Utf8Lossy {
     }
 
     pub fn from_bytes(bytes: &[u8]) -> &Utf8Lossy {
+        // SAFETY: Both use the same memory layout, and UTF-8 correctness isn't required.
         unsafe { mem::transmute(bytes) }
     }
 
@@ -60,6 +59,8 @@ impl<'a> Iterator for Utf8LossyChunksIter<'a> {
         while i < self.source.len() {
             let i_ = i;
 
+            // SAFETY: `i` starts at `0`, is less than `self.source.len()`, and
+            // only increases, so `0 <= i < self.source.len()`.
             let byte = unsafe { *self.source.get_unchecked(i) };
             i += 1;
 
@@ -69,6 +70,7 @@ impl<'a> Iterator for Utf8LossyChunksIter<'a> {
 
                 macro_rules! error {
                     () => {{
+                        // SAFETY: We have checked up to `i` that source is valid UTF-8.
                         unsafe {
                             let r = Utf8LossyChunk {
                                 valid: core_str::from_utf8_unchecked(&self.source[0..i_]),
@@ -130,6 +132,7 @@ impl<'a> Iterator for Utf8LossyChunksIter<'a> {
         }
 
         let r = Utf8LossyChunk {
+            // SAFETY: We have checked that the entire source is valid UTF-8.
             valid: unsafe { core_str::from_utf8_unchecked(self.source) },
             broken: &[],
         };
diff --git a/src/libcore/str/mod.rs b/src/libcore/str/mod.rs
index ab771b1164b..5a7cddd4041 100644
--- a/src/libcore/str/mod.rs
+++ b/src/libcore/str/mod.rs
@@ -1,5 +1,4 @@
 // ignore-tidy-filelength
-// ignore-tidy-undocumented-unsafe
 
 //! String manipulation.
 //!
@@ -341,6 +340,7 @@ impl Utf8Error {
 #[stable(feature = "rust1", since = "1.0.0")]
 pub fn from_utf8(v: &[u8]) -> Result<&str, Utf8Error> {
     run_utf8_validation(v)?;
+    // SAFETY: Just ran validation.
     Ok(unsafe { from_utf8_unchecked(v) })
 }
 
@@ -379,6 +379,7 @@ pub fn from_utf8(v: &[u8]) -> Result<&str, Utf8Error> {
 #[stable(feature = "str_mut_extras", since = "1.20.0")]
 pub fn from_utf8_mut(v: &mut [u8]) -> Result<&mut str, Utf8Error> {
     run_utf8_validation(v)?;
+    // SAFETY: Just ran validation.
     Ok(unsafe { from_utf8_unchecked_mut(v) })
 }
 
@@ -581,7 +582,7 @@ impl<'a> Iterator for Chars<'a> {
     #[inline]
     fn next(&mut self) -> Option<char> {
         next_code_point(&mut self.iter).map(|ch| {
-            // str invariant says `ch` is a valid Unicode Scalar Value
+            // SAFETY: `str` invariant says `ch` is a valid Unicode Scalar Value.
             unsafe { char::from_u32_unchecked(ch) }
         })
     }
@@ -628,7 +629,7 @@ impl<'a> DoubleEndedIterator for Chars<'a> {
     #[inline]
     fn next_back(&mut self) -> Option<char> {
         next_code_point_reverse(&mut self.iter).map(|ch| {
-            // str invariant says `ch` is a valid Unicode Scalar Value
+            // SAFETY: `str` invariant says `ch` is a valid Unicode Scalar Value.
             unsafe { char::from_u32_unchecked(ch) }
         })
     }
@@ -658,6 +659,7 @@ impl<'a> Chars<'a> {
     #[stable(feature = "iter_to_slice", since = "1.4.0")]
     #[inline]
     pub fn as_str(&self) -> &'a str {
+        // SAFETY: `Chars` is only made from a str, which guarantees the iter is valid UTF-8.
         unsafe { from_utf8_unchecked(self.iter.as_slice()) }
     }
 }
@@ -1102,6 +1104,7 @@ impl<'a, P: Pattern<'a>> SplitInternal<'a, P> {
     fn get_end(&mut self) -> Option<&'a str> {
         if !self.finished && (self.allow_trailing_empty || self.end - self.start > 0) {
             self.finished = true;
+            // SAFETY: `self.start` and `self.end` always lie on unicode boundaries.
             unsafe {
                 let string = self.matcher.haystack().get_unchecked(self.start..self.end);
                 Some(string)
@@ -1119,6 +1122,7 @@ impl<'a, P: Pattern<'a>> SplitInternal<'a, P> {
 
         let haystack = self.matcher.haystack();
         match self.matcher.next_match() {
+            // SAFETY: `Searcher` guarantees that `a` and `b` lie on unicode boundaries.
             Some((a, b)) => unsafe {
                 let elt = haystack.get_unchecked(self.start..a);
                 self.start = b;
@@ -1151,11 +1155,13 @@ impl<'a, P: Pattern<'a>> SplitInternal<'a, P> {
 
         let haystack = self.matcher.haystack();
         match self.matcher.next_match_back() {
+            // SAFETY: `Searcher` guarantees that `a` and `b` lie on unicode boundaries.
             Some((a, b)) => unsafe {
                 let elt = haystack.get_unchecked(b..self.end);
                 self.end = a;
                 Some(elt)
             },
+            // SAFETY: `self.start` and `self.end` always lie on unicode boundaries.
             None => unsafe {
                 self.finished = true;
                 Some(haystack.get_unchecked(self.start..self.end))
@@ -1297,6 +1303,7 @@ impl<'a, P: Pattern<'a>> MatchIndicesInternal<'a, P> {
     fn next(&mut self) -> Option<(usize, &'a str)> {
         self.0
             .next_match()
+            // SAFETY: `Searcher` guarantees that `start` and `end` lie on unicode boundaries.
             .map(|(start, end)| unsafe { (start, self.0.haystack().get_unchecked(start..end)) })
     }
 
@@ -1307,6 +1314,7 @@ impl<'a, P: Pattern<'a>> MatchIndicesInternal<'a, P> {
     {
         self.0
             .next_match_back()
+            // SAFETY: `Searcher` guarantees that `start` and `end` lie on unicode boundaries.
             .map(|(start, end)| unsafe { (start, self.0.haystack().get_unchecked(start..end)) })
     }
 }
@@ -1348,6 +1356,7 @@ where
 impl<'a, P: Pattern<'a>> MatchesInternal<'a, P> {
     #[inline]
     fn next(&mut self) -> Option<&'a str> {
+        // SAFETY: `Searcher` guarantees that `start` and `end` lie on unicode boundaries.
         self.0.next_match().map(|(a, b)| unsafe {
             // Indices are known to be on utf8 boundaries
             self.0.haystack().get_unchecked(a..b)
@@ -1359,6 +1368,7 @@ impl<'a, P: Pattern<'a>> MatchesInternal<'a, P> {
     where
         P::Searcher: ReverseSearcher<'a>,
     {
+        // SAFETY: `Searcher` guarantees that `start` and `end` lie on unicode boundaries.
         self.0.next_match_back().map(|(a, b)| unsafe {
             // Indices are known to be on utf8 boundaries
             self.0.haystack().get_unchecked(a..b)
@@ -1579,6 +1589,10 @@ fn run_utf8_validation(v: &[u8]) -> Result<(), Utf8Error> {
             if align != usize::max_value() && align.wrapping_sub(index) % usize_bytes == 0 {
                 let ptr = v.as_ptr();
                 while index < blocks_end {
+                    // SAFETY: since `align - index` and `ascii_block_size` are
+                    // multiples of `usize_bytes`, `block = ptr.add(index)` is
+                    // always aligned with a `usize` so it's safe to dereference
+                    // both `block` and `block.offset(1)`.
                     unsafe {
                         let block = ptr.add(index) as *const usize;
                         // break if there is a nonascii byte
@@ -1804,6 +1818,7 @@ mod traits {
                 && slice.is_char_boundary(self.start)
                 && slice.is_char_boundary(self.end)
             {
+                // SAFETY: just checked that `start` and `end` are on a char boundary.
                 Some(unsafe { self.get_unchecked(slice) })
             } else {
                 None
@@ -1815,6 +1830,7 @@ mod traits {
                 && slice.is_char_boundary(self.start)
                 && slice.is_char_boundary(self.end)
             {
+                // SAFETY: just checked that `start` and `end` are on a char boundary.
                 Some(unsafe { self.get_unchecked_mut(slice) })
             } else {
                 None
@@ -1845,6 +1861,7 @@ mod traits {
                 && slice.is_char_boundary(self.start)
                 && slice.is_char_boundary(self.end)
             {
+                // SAFETY: just checked that `start` and `end` are on a char boundary.
                 unsafe { self.get_unchecked_mut(slice) }
             } else {
                 super::slice_error_fail(slice, self.start, self.end)
@@ -1873,6 +1890,7 @@ mod traits {
         #[inline]
         fn get(self, slice: &str) -> Option<&Self::Output> {
             if slice.is_char_boundary(self.end) {
+                // SAFETY: just checked that `end` is on a char boundary.
                 Some(unsafe { self.get_unchecked(slice) })
             } else {
                 None
@@ -1881,6 +1899,7 @@ mod traits {
         #[inline]
         fn get_mut(self, slice: &mut str) -> Option<&mut Self::Output> {
             if slice.is_char_boundary(self.end) {
+                // SAFETY: just checked that `end` is on a char boundary.
                 Some(unsafe { self.get_unchecked_mut(slice) })
             } else {
                 None
@@ -1903,8 +1922,8 @@ mod traits {
         }
         #[inline]
         fn index_mut(self, slice: &mut str) -> &mut Self::Output {
-            // is_char_boundary checks that the index is in [0, .len()]
             if slice.is_char_boundary(self.end) {
+                // SAFETY: just checked that `end` is on a char boundary.
                 unsafe { self.get_unchecked_mut(slice) }
             } else {
                 super::slice_error_fail(slice, 0, self.end)
@@ -1934,6 +1953,7 @@ mod traits {
         #[inline]
         fn get(self, slice: &str) -> Option<&Self::Output> {
             if slice.is_char_boundary(self.start) {
+                // SAFETY: just checked that `start` is on a char boundary.
                 Some(unsafe { self.get_unchecked(slice) })
             } else {
                 None
@@ -1942,6 +1962,7 @@ mod traits {
         #[inline]
         fn get_mut(self, slice: &mut str) -> Option<&mut Self::Output> {
             if slice.is_char_boundary(self.start) {
+                // SAFETY: just checked that `start` is on a char boundary.
                 Some(unsafe { self.get_unchecked_mut(slice) })
             } else {
                 None
@@ -1966,8 +1987,8 @@ mod traits {
         }
         #[inline]
         fn index_mut(self, slice: &mut str) -> &mut Self::Output {
-            // is_char_boundary checks that the index is in [0, .len()]
             if slice.is_char_boundary(self.start) {
+                // SAFETY: just checked that `start` is on a char boundary.
                 unsafe { self.get_unchecked_mut(slice) }
             } else {
                 super::slice_error_fail(slice, self.start, slice.len())
@@ -2238,7 +2259,6 @@ impl str {
     #[stable(feature = "rust1", since = "1.0.0")]
     #[rustc_const_stable(feature = "str_as_bytes", since = "1.32.0")]
     #[inline(always)]
-    // SAFETY: const sound because we transmute two types with the same layout
     #[allow(unused_attributes)]
     #[allow_internal_unstable(const_fn_union)]
     pub const fn as_bytes(&self) -> &[u8] {
@@ -2247,6 +2267,7 @@ impl str {
             str: &'a str,
             slice: &'a [u8],
         }
+        // SAFETY: const sound because we transmute two types with the same layout
         unsafe { Slices { str: self }.slice }
     }
 
@@ -2573,6 +2594,7 @@ impl str {
     pub fn split_at(&self, mid: usize) -> (&str, &str) {
         // is_char_boundary checks that the index is in [0, .len()]
         if self.is_char_boundary(mid) {
+            // SAFETY: just checked that `mid` is on a char boundary.
             unsafe { (self.get_unchecked(0..mid), self.get_unchecked(mid..self.len())) }
         } else {
             slice_error_fail(self, 0, mid)
@@ -2617,6 +2639,7 @@ impl str {
         if self.is_char_boundary(mid) {
             let len = self.len();
             let ptr = self.as_mut_ptr();
+            // SAFETY: just checked that `mid` is on a char boundary.
             unsafe {
                 (
                     from_utf8_unchecked_mut(slice::from_raw_parts_mut(ptr, mid)),
@@ -3805,10 +3828,8 @@ impl str {
         if let Some((_, b)) = matcher.next_reject_back() {
             j = b;
         }
-        unsafe {
-            // Searcher is known to return valid indices
-            self.get_unchecked(i..j)
-        }
+        // SAFETY: `Searcher` is known to return valid indices.
+        unsafe { self.get_unchecked(i..j) }
     }
 
     /// Returns a string slice with all prefixes that match a pattern
@@ -3844,10 +3865,8 @@ impl str {
         if let Some((a, _)) = matcher.next_reject() {
             i = a;
         }
-        unsafe {
-            // Searcher is known to return valid indices
-            self.get_unchecked(i..self.len())
-        }
+        // SAFETY: `Searcher` is known to return valid indices.
+        unsafe { self.get_unchecked(i..self.len()) }
     }
 
     /// Returns a string slice with the prefix removed.
@@ -3878,10 +3897,8 @@ impl str {
                 "The first search step from Searcher \
                 must include the first character"
             );
-            unsafe {
-                // Searcher is known to return valid indices.
-                Some(self.get_unchecked(len..))
-            }
+            // SAFETY: `Searcher` is known to return valid indices.
+            unsafe { Some(self.get_unchecked(len..)) }
         } else {
             None
         }
@@ -3919,10 +3936,8 @@ impl str {
                 "The first search step from ReverseSearcher \
                 must include the last character"
             );
-            unsafe {
-                // Searcher is known to return valid indices.
-                Some(self.get_unchecked(..start))
-            }
+            // SAFETY: `Searcher` is known to return valid indices.
+            unsafe { Some(self.get_unchecked(..start)) }
         } else {
             None
         }
@@ -3970,10 +3985,8 @@ impl str {
         if let Some((_, b)) = matcher.next_reject_back() {
             j = b;
         }
-        unsafe {
-            // Searcher is known to return valid indices
-            self.get_unchecked(0..j)
-        }
+        // SAFETY: `Searcher` is known to return valid indices.
+        unsafe { self.get_unchecked(0..j) }
     }
 
     /// Returns a string slice with all prefixes that match a pattern
@@ -4166,6 +4179,7 @@ impl str {
     /// ```
     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
     pub fn make_ascii_uppercase(&mut self) {
+        // SAFETY: safe because we transmute two types with the same layout.
         let me = unsafe { self.as_bytes_mut() };
         me.make_ascii_uppercase()
     }
@@ -4191,6 +4205,7 @@ impl str {
     /// ```
     #[stable(feature = "ascii_methods_on_intrinsics", since = "1.23.0")]
     pub fn make_ascii_lowercase(&mut self) {
+        // SAFETY: safe because we transmute two types with the same layout.
         let me = unsafe { self.as_bytes_mut() };
         me.make_ascii_lowercase()
     }
@@ -4357,6 +4372,7 @@ impl Default for &str {
 impl Default for &mut str {
     /// Creates an empty mutable str
     fn default() -> Self {
+        // SAFETY: The empty string is valid UTF-8.
         unsafe { from_utf8_unchecked_mut(&mut []) }
     }
 }
@@ -4412,6 +4428,7 @@ impl_fn_for_zst! {
 
     #[derive(Clone)]
     struct UnsafeBytesToStr impl<'a> Fn = |bytes: &'a [u8]| -> &'a str {
+        // SAFETY: not safe
         unsafe { from_utf8_unchecked(bytes) }
     };
 }
diff --git a/src/libcore/str/pattern.rs b/src/libcore/str/pattern.rs
index 46d9499394a..ef64d8b0fdf 100644
--- a/src/libcore/str/pattern.rs
+++ b/src/libcore/str/pattern.rs
@@ -3,8 +3,6 @@
 //! For more details, see the traits [`Pattern`], [`Searcher`],
 //! [`ReverseSearcher`], and [`DoubleEndedSearcher`].
 
-// ignore-tidy-undocumented-unsafe
-
 #![unstable(
     feature = "pattern",
     reason = "API not fully fleshed out and ready to be stabilized",
@@ -271,6 +269,14 @@ unsafe impl<'a> Searcher<'a> for CharSearcher<'a> {
     #[inline]
     fn next(&mut self) -> SearchStep {
         let old_finger = self.finger;
+        // SAFETY: 1-4 guarantee safety of `get_unchecked`
+        // 1. `self.finger` and `self.finger_back` are kept on unicode boundaries
+        //    (this is invariant)
+        // 2. `self.finger >= 0` since it starts at 0 and only increases
+        // 3. `self.finger < self.finger_back` because otherwise the char `iter`
+        //    would return `SearchStep::Done`
+        // 4. `self.finger` comes before the end of the haystack because `self.finger_back`
+        //    starts at the end and only decreases
         let slice = unsafe { self.haystack.get_unchecked(old_finger..self.finger_back) };
         let mut iter = slice.chars();
         let old_len = iter.iter.len();
@@ -293,6 +299,7 @@ unsafe impl<'a> Searcher<'a> for CharSearcher<'a> {
             // get the haystack after the last character found
             let bytes = self.haystack.as_bytes().get(self.finger..self.finger_back)?;
             // the last byte of the utf8 encoded needle
+            // SAFETY: we have an invariant that `utf8_size < 5`
             let last_byte = unsafe { *self.utf8_encoded.get_unchecked(self.utf8_size - 1) };
             if let Some(index) = memchr::memchr(last_byte, bytes) {
                 // The new finger is the index of the byte we found,
@@ -336,6 +343,7 @@ unsafe impl<'a> ReverseSearcher<'a> for CharSearcher<'a> {
     #[inline]
     fn next_back(&mut self) -> SearchStep {
         let old_finger = self.finger_back;
+        // SAFETY: see the comment for next() above
         let slice = unsafe { self.haystack.get_unchecked(self.finger..old_finger) };
         let mut iter = slice.chars();
         let old_len = iter.iter.len();
@@ -363,6 +371,7 @@ unsafe impl<'a> ReverseSearcher<'a> for CharSearcher<'a> {
                 return None;
             };
             // the last byte of the utf8 encoded needle
+            // SAFETY: we have an invariant that `utf8_size < 5`
             let last_byte = unsafe { *self.utf8_encoded.get_unchecked(self.utf8_size - 1) };
             if let Some(index) = memchr::memrchr(last_byte, bytes) {
                 // we searched a slice that was offset by self.finger,