expose drop_in_place as ptr::drop_in_place

Author: Gankra

src/liballoc/lib.rs

@@ -103,6 +103,8 @@
 #![feature(unsize)]
 #![feature(core_slice_ext)]
 #![feature(core_str_ext)]
+#![feature(drop_in_place)]
+
 #![cfg_attr(stage0, feature(alloc_system))]
 #![cfg_attr(not(stage0), feature(needs_allocator))]
 

src/libcore/intrinsics.rs

@@ -195,7 +195,26 @@ extern "rust-intrinsic" {
 
     pub fn size_of_val<T: ?Sized>(_: &T) -> usize;
     pub fn min_align_of_val<T: ?Sized>(_: &T) -> usize;
-    pub fn drop_in_place<T: ?Sized>(_: *mut T);
+
+    /// Executes the destructor (if any) of the pointed-to value.
+    ///
+    /// This has two usecases:
+    ///
+    /// * It is *required* to use `drop_in_place` to drop unsized types like
+    ///   trait objects, because they can't be read out onto the stack and
+    ///   dropped normally.
+    ///
+    /// * It is friendlier to the optimizer to do this over `ptr::read` when
+    ///   dropping manually allocated memory (e.g. when writing Box/Rc/Vec),
+    ///   as the compiler doesn't need to prove that it's sound to elide the
+    ///   copy.
+    ///
+    /// # Undefined Behaviour
+    ///
+    /// This has all the same safety problems as `ptr::read` with respect to
+    /// invalid pointers, types, and double drops.
+    #[unstable(feature = "drop_in_place", reason = "just exposed, needs FCP", issue = "27908")]
+    pub fn drop_in_place<T: ?Sized>(to_drop: *mut T);
 
     /// Gets a static string slice containing the name of a type.
     pub fn type_name<T: ?Sized>() -> &'static str;

src/libcore/ptr.rs

@@ -40,6 +40,8 @@ pub use intrinsics::copy;
 #[stable(feature = "rust1", since = "1.0.0")]
 pub use intrinsics::write_bytes;
 
+pub use intrinsics::drop_in_place;
+
 /// Creates a null raw pointer.
 ///
 /// # Examples