Auto merge of #74019 - Manishearth:rollup-2st3jsk, r=Manishearth

Rollup of 12 pull requests

Successful merges:

 - #73140 (Fallback to xml.etree.ElementTree)
 - #73670 (Add `format_args_capture` feature)
 - #73693 (Use exhaustive match in const_prop.rs)
 - #73845 (Use &raw in A|Rc::as_ptr)
 - #73861 (Create E0768)
 - #73881 (Standardize bibliographic citations in rustc API docs)
 - #73925 (Improve comments from #72617, as suggested by RalfJung)
 - #73949 ([mir-opt] Fix mis-optimization and other issues with the SimplifyArmIdentity pass)
 - #73984 (Edit docs for rustc_data_structures::graph::scc)
 - #73985 (Fix "getting started" link)
 - #73997 (fix typo)
 - #73999 (Bump mingw-check CI image from Ubuntu 16.04 to 18.04.)

Failed merges:

 - #74000 (add `lazy_normalization_consts` feature gate)

r? @ghost

Author: bors

CONTRIBUTING.md

@@ -4,5 +4,5 @@ Thank you for your interest in contributing to Rust!
 
 To get started, read the [Getting Started] guide in the [rustc-dev-guide].
 
-[Getting Started]: https://rustc-dev-guide.rust-lang.org/getting-started.md
+[Getting Started]: https://rustc-dev-guide.rust-lang.org/getting-started.html
 [rustc-dev-guide]: https://rustc-dev-guide.rust-lang.org/

README.md

@@ -26,7 +26,7 @@ The Rust build system uses a Python script called `x.py` to build the compiler,
 which manages the bootstrapping process. More information about it can be found 
 by running `./x.py --help` or reading the [rustc dev guide][rustcguidebuild].
 
-[gettingstarted]: https://rustc-dev-guide.rust-lang.org/getting-started.md
+[gettingstarted]: https://rustc-dev-guide.rust-lang.org/getting-started.html
 [rustcguidebuild]: https://rustc-dev-guide.rust-lang.org/building/how-to-build-and-run.html
 
 ### Building on a Unix-like system

src/ci/docker/host-x86_64/mingw-check/Dockerfile

@@ -1,4 +1,4 @@
-FROM ubuntu:16.04
+FROM ubuntu:18.04
 
 RUN apt-get update && apt-get install -y --no-install-recommends \
   g++ \

src/doc/unstable-book/src/library-features/format-args-capture.md

@@ -0,0 +1,47 @@
+# `format_args_capture`
+
+The tracking issue for this feature is: [#67984]
+
+[#67984]: https://github.com/rust-lang/rust/issues/67984
+
+------------------------
+
+Enables `format_args!` (and macros which use `format_args!` in their implementation, such
+as `format!`, `print!` and `panic!`) to capture variables from the surrounding scope.
+This avoids the need to pass named parameters when the binding in question
+already exists in scope.
+
+```rust
+#![feature(format_args_capture)]
+
+let (person, species, name) = ("Charlie Brown", "dog", "Snoopy");
+
+// captures named argument `person`
+print!("Hello {person}");
+
+// captures named arguments `species` and `name`
+format!("The {species}'s name is {name}.");
+```
+
+This also works for formatting parameters such as width and precision:
+
+```rust
+#![feature(format_args_capture)]
+
+let precision = 2;
+let s = format!("{:.precision$}", 1.324223);
+
+assert_eq!(&s, "1.32");
+```
+
+A non-exhaustive list of macros which benefit from this functionality include:
+- `format!`
+- `print!` and `println!`
+- `eprint!` and `eprintln!`
+- `write!` and `writeln!`
+- `panic!`
+- `unreachable!`
+- `unimplemented!`
+- `todo!`
+- `assert!` and similar
+- macros in many thirdparty crates, such as `log`

src/etc/htmldocck.py

@@ -114,7 +114,10 @@
     from html.parser import HTMLParser
 except ImportError:
     from HTMLParser import HTMLParser
-from xml.etree import cElementTree as ET
+try:
+    from xml.etree import cElementTree as ET
+except ImportError:
+    from xml.etree import ElementTree as ET
 
 try:
     from html.entities import name2codepoint

src/liballoc/lib.rs

@@ -100,6 +100,7 @@
 #![feature(fundamental)]
 #![feature(internal_uninit_const)]
 #![feature(lang_items)]
+#![feature(layout_for_ptr)]
 #![feature(libc)]
 #![feature(negative_impls)]
 #![feature(new_uninit)]
@@ -109,6 +110,7 @@
 #![feature(pattern)]
 #![feature(ptr_internals)]
 #![feature(ptr_offset_from)]
+#![feature(raw_ref_op)]
 #![feature(rustc_attrs)]
 #![feature(receiver_trait)]
 #![feature(min_specialization)]

src/liballoc/rc.rs

@@ -245,7 +245,7 @@ use core::hash::{Hash, Hasher};
 use core::intrinsics::abort;
 use core::iter;
 use core::marker::{self, PhantomData, Unpin, Unsize};
-use core::mem::{self, align_of, align_of_val, forget, size_of_val};
+use core::mem::{self, align_of_val_raw, forget, size_of_val};
 use core::ops::{CoerceUnsized, Deref, DispatchFromDyn, Receiver};
 use core::pin::Pin;
 use core::ptr::{self, NonNull};
@@ -591,17 +591,11 @@ impl<T: ?Sized> Rc<T> {
     #[stable(feature = "weak_into_raw", since = "1.45.0")]
     pub fn as_ptr(this: &Self) -> *const T {
         let ptr: *mut RcBox<T> = NonNull::as_ptr(this.ptr);
-        let fake_ptr = ptr as *mut T;
 
-        // SAFETY: This cannot go through Deref::deref.
-        // Instead, we manually offset the pointer rather than manifesting a reference.
-        // This is so that the returned pointer retains the same provenance as our pointer.
-        // This is required so that e.g. `get_mut` can write through the pointer
-        // after the Rc is recovered through `from_raw`.
-        unsafe {
-            let offset = data_offset(&(*ptr).value);
-            set_data_ptr(fake_ptr, (ptr as *mut u8).offset(offset))
-        }
+        // SAFETY: This cannot go through Deref::deref or Rc::inner because
+        // this is required to retain raw/mut provenance such that e.g. `get_mut` can
+        // write through the pointer after the Rc is recovered through `from_raw`.
+        unsafe { &raw const (*ptr).value }
     }
 
     /// Constructs an `Rc<T>` from a raw pointer.
@@ -1647,6 +1641,7 @@ pub struct Weak<T: ?Sized> {
     // `Weak::new` sets this to `usize::MAX` so that it doesn’t need
     // to allocate space on the heap.  That's not a value a real pointer
     // will ever have because RcBox has alignment at least 2.
+    // This is only possible when `T: Sized`; unsized `T` never dangle.
     ptr: NonNull<RcBox<T>>,
 }
 
@@ -1708,9 +1703,18 @@ impl<T> Weak<T> {
     /// [`null`]: ../../std/ptr/fn.null.html
     #[stable(feature = "weak_into_raw", since = "1.45.0")]
     pub fn as_ptr(&self) -> *const T {
-        let offset = data_offset_sized::<T>();
-        let ptr = self.ptr.cast::<u8>().as_ptr().wrapping_offset(offset);
-        ptr as *const T
+        let ptr: *mut RcBox<T> = NonNull::as_ptr(self.ptr);
+
+        // SAFETY: we must offset the pointer manually, and said pointer may be
+        // a dangling weak (usize::MAX) if T is sized. data_offset is safe to call,
+        // because we know that a pointer to unsized T was derived from a real
+        // unsized T, as dangling weaks are only created for sized T. wrapping_offset
+        // is used so that we can use the same code path for the non-dangling
+        // unsized case and the potentially dangling sized case.
+        unsafe {
+            let offset = data_offset(ptr as *mut T);
+            set_data_ptr(ptr as *mut T, (ptr as *mut u8).wrapping_offset(offset))
+        }
     }
 
     /// Consumes the `Weak<T>` and turns it into a raw pointer.
@@ -2113,19 +2117,22 @@ impl<T: ?Sized> AsRef<T> for Rc<T> {
 #[stable(feature = "pin", since = "1.33.0")]
 impl<T: ?Sized> Unpin for Rc<T> {}
 
+/// Get the offset within an `ArcInner` for
+/// a payload of type described by a pointer.
+///
+/// # Safety
+///
+/// This has the same safety requirements as `align_of_val_raw`. In effect:
+///
+/// - This function is safe for any argument if `T` is sized, and
+/// - if `T` is unsized, the pointer must have appropriate pointer metadata
+///   aquired from the real instance that you are getting this offset for.
 unsafe fn data_offset<T: ?Sized>(ptr: *const T) -> isize {
     // Align the unsized value to the end of the `RcBox`.
     // Because it is ?Sized, it will always be the last field in memory.
     // Note: This is a detail of the current implementation of the compiler,
     // and is not a guaranteed language detail. Do not rely on it outside of std.
-    unsafe { data_offset_align(align_of_val(&*ptr)) }
-}
-
-/// Computes the offset of the data field within `RcBox`.
-///
-/// Unlike [`data_offset`], this doesn't need the pointer, but it works only on `T: Sized`.
-fn data_offset_sized<T>() -> isize {
-    data_offset_align(align_of::<T>())
+    unsafe { data_offset_align(align_of_val_raw(ptr)) }
 }
 
 #[inline]

src/liballoc/sync.rs

@@ -16,7 +16,7 @@ use core::hash::{Hash, Hasher};
 use core::intrinsics::abort;
 use core::iter;
 use core::marker::{PhantomData, Unpin, Unsize};
-use core::mem::{self, align_of, align_of_val, size_of_val};
+use core::mem::{self, align_of_val, size_of_val};
 use core::ops::{CoerceUnsized, Deref, DispatchFromDyn, Receiver};
 use core::pin::Pin;
 use core::ptr::{self, NonNull};
@@ -267,6 +267,7 @@ pub struct Weak<T: ?Sized> {
     // `Weak::new` sets this to `usize::MAX` so that it doesn’t need
     // to allocate space on the heap.  That's not a value a real pointer
     // will ever have because RcBox has alignment at least 2.
+    // This is only possible when `T: Sized`; unsized `T` never dangle.
     ptr: NonNull<ArcInner<T>>,
 }
 
@@ -590,17 +591,11 @@ impl<T: ?Sized> Arc<T> {
     #[stable(feature = "weak_into_raw", since = "1.45.0")]
     pub fn as_ptr(this: &Self) -> *const T {
         let ptr: *mut ArcInner<T> = NonNull::as_ptr(this.ptr);
-        let fake_ptr = ptr as *mut T;
 
-        // SAFETY: This cannot go through Deref::deref.
-        // Instead, we manually offset the pointer rather than manifesting a reference.
-        // This is so that the returned pointer retains the same provenance as our pointer.
-        // This is required so that e.g. `get_mut` can write through the pointer
-        // after the Arc is recovered through `from_raw`.
-        unsafe {
-            let offset = data_offset(&(*ptr).data);
-            set_data_ptr(fake_ptr, (ptr as *mut u8).offset(offset))
-        }
+        // SAFETY: This cannot go through Deref::deref or RcBoxPtr::inner because
+        // this is required to retain raw/mut provenance such that e.g. `get_mut` can
+        // write through the pointer after the Rc is recovered through `from_raw`.
+        unsafe { &raw const (*ptr).data }
     }
 
     /// Constructs an `Arc<T>` from a raw pointer.
@@ -1476,9 +1471,18 @@ impl<T> Weak<T> {
     /// [`null`]: ../../std/ptr/fn.null.html
     #[stable(feature = "weak_into_raw", since = "1.45.0")]
     pub fn as_ptr(&self) -> *const T {
-        let offset = data_offset_sized::<T>();
-        let ptr = self.ptr.cast::<u8>().as_ptr().wrapping_offset(offset);
-        ptr as *const T
+        let ptr: *mut ArcInner<T> = NonNull::as_ptr(self.ptr);
+
+        // SAFETY: we must offset the pointer manually, and said pointer may be
+        // a dangling weak (usize::MAX) if T is sized. data_offset is safe to call,
+        // because we know that a pointer to unsized T was derived from a real
+        // unsized T, as dangling weaks are only created for sized T. wrapping_offset
+        // is used so that we can use the same code path for the non-dangling
+        // unsized case and the potentially dangling sized case.
+        unsafe {
+            let offset = data_offset(ptr as *mut T);
+            set_data_ptr(ptr as *mut T, (ptr as *mut u8).wrapping_offset(offset))
+        }
     }
 
     /// Consumes the `Weak<T>` and turns it into a raw pointer.
@@ -2270,7 +2274,16 @@ impl<T: ?Sized> AsRef<T> for Arc<T> {
 #[stable(feature = "pin", since = "1.33.0")]
 impl<T: ?Sized> Unpin for Arc<T> {}
 
-/// Computes the offset of the data field within `ArcInner`.
+/// Get the offset within an `ArcInner` for
+/// a payload of type described by a pointer.
+///
+/// # Safety
+///
+/// This has the same safety requirements as `align_of_val_raw`. In effect:
+///
+/// - This function is safe for any argument if `T` is sized, and
+/// - if `T` is unsized, the pointer must have appropriate pointer metadata
+///   aquired from the real instance that you are getting this offset for.
 unsafe fn data_offset<T: ?Sized>(ptr: *const T) -> isize {
     // Align the unsized value to the end of the `ArcInner`.
     // Because it is `?Sized`, it will always be the last field in memory.
@@ -2279,13 +2292,6 @@ unsafe fn data_offset<T: ?Sized>(ptr: *const T) -> isize {
     unsafe { data_offset_align(align_of_val(&*ptr)) }
 }
 
-/// Computes the offset of the data field within `ArcInner`.
-///
-/// Unlike [`data_offset`], this doesn't need the pointer, but it works only on `T: Sized`.
-fn data_offset_sized<T>() -> isize {
-    data_offset_align(align_of::<T>())
-}
-
 #[inline]
 fn data_offset_align(align: usize) -> isize {
     let layout = Layout::new::<ArcInner<()>>();

src/libcore/marker.rs

@@ -694,7 +694,7 @@ mod impls {
 )]
 #[lang = "discriminant_kind"]
 pub trait DiscriminantKind {
-    /// The type of the dicriminant, which must satisfy the trait
+    /// The type of the discriminant, which must satisfy the trait
     /// bounds required by `mem::Discriminant`.
     type Discriminant: Clone + Copy + Debug + Eq + PartialEq + Hash + Send + Sync + Unpin;
 }

src/librustc_ast_pretty/pp.rs

@@ -1,10 +1,8 @@
 //! This pretty-printer is a direct reimplementation of Philip Karlton's
-//! Mesa pretty-printer, as described in appendix A of
-//!
-//! ```text
-//! STAN-CS-79-770: "Pretty Printing", by Derek C. Oppen.
-//! Stanford Department of Computer Science, 1979.
-//! ```
+//! Mesa pretty-printer, as described in the appendix to
+//! Derek C. Oppen, "Pretty Printing" (1979),
+//! Stanford Computer Science Department STAN-CS-79-770,
+//! <http://i.stanford.edu/pub/cstr/reports/cs/tr/79/770/CS-TR-79-770.pdf>.
 //!
 //! The algorithm's aim is to break a stream into as few lines as possible
 //! while respecting the indentation-consistency requirements of the enclosing

src/librustc_builtin_macros/format.rs

@@ -107,6 +107,9 @@ struct Context<'a, 'b> {
     arg_spans: Vec<Span>,
     /// All the formatting arguments that have formatting flags set, in order for diagnostics.
     arg_with_formatting: Vec<parse::FormatSpec<'a>>,
+
+    /// Whether this format string came from a string literal, as opposed to a macro.
+    is_literal: bool,
 }
 
 /// Parses the arguments from the given list of tokens, returning the diagnostic
@@ -498,10 +501,55 @@ impl<'a, 'b> Context<'a, 'b> {
                         self.verify_arg_type(Exact(idx), ty)
                     }
                     None => {
-                        let msg = format!("there is no argument named `{}`", name);
-                        let sp = *self.arg_spans.get(self.curpiece).unwrap_or(&self.fmtsp);
-                        let mut err = self.ecx.struct_span_err(sp, &msg[..]);
-                        err.emit();
+                        let capture_feature_enabled = self
+                            .ecx
+                            .ecfg
+                            .features
+                            .map_or(false, |features| features.format_args_capture);
+
+                        // For the moment capturing variables from format strings expanded from macros is
+                        // disabled (see RFC #2795)
+                        let can_capture = capture_feature_enabled && self.is_literal;
+
+                        if can_capture {
+                            // Treat this name as a variable to capture from the surrounding scope
+                            let idx = self.args.len();
+                            self.arg_types.push(Vec::new());
+                            self.arg_unique_types.push(Vec::new());
+                            self.args.push(
+                                self.ecx.expr_ident(self.fmtsp, Ident::new(name, self.fmtsp)),
+                            );
+                            self.names.insert(name, idx);
+                            self.verify_arg_type(Exact(idx), ty)
+                        } else {
+                            let msg = format!("there is no argument named `{}`", name);
+                            let sp = if self.is_literal {
+                                *self.arg_spans.get(self.curpiece).unwrap_or(&self.fmtsp)
+                            } else {
+                                self.fmtsp
+                            };
+                            let mut err = self.ecx.struct_span_err(sp, &msg[..]);
+
+                            if capture_feature_enabled && !self.is_literal {
+                                err.note(&format!(
+                                    "did you intend to capture a variable `{}` from \
+                                     the surrounding scope?",
+                                    name
+                                ));
+                                err.note(
+                                    "to avoid ambiguity, `format_args!` cannot capture variables \
+                                     when the format string is expanded from a macro",
+                                );
+                            } else if self.ecx.parse_sess().unstable_features.is_nightly_build() {
+                                err.help(&format!(
+                                    "if you intended to capture `{}` from the surrounding scope, add \
+                                     `#![feature(format_args_capture)]` to the crate attributes",
+                                    name
+                                ));
+                            }
+
+                            err.emit();
+                        }
                     }
                 }
             }
@@ -951,6 +999,7 @@ pub fn expand_preparsed_format_args(
         invalid_refs: Vec::new(),
         arg_spans,
         arg_with_formatting: Vec::new(),
+        is_literal: parser.is_literal,
     };
 
     // This needs to happen *after* the Parser has consumed all pieces to create all the spans

src/librustc_data_structures/graph/dominators/mod.rs

@@ -1,8 +1,9 @@
-//! Algorithm citation:
-//! A Simple, Fast Dominance Algorithm.
-//! Keith D. Cooper, Timothy J. Harvey, and Ken Kennedy
-//! Rice Computer Science TS-06-33870
-//! <https://www.cs.rice.edu/~keith/EMBED/dom.pdf>
+//! Finding the dominators in a control-flow graph.
+//!
+//! Algorithm based on Keith D. Cooper, Timothy J. Harvey, and Ken Kennedy,
+//! "A Simple, Fast Dominance Algorithm",
+//! Rice Computer Science TS-06-33870,
+//! <https://www.cs.rice.edu/~keith/EMBED/dom.pdf>.
 
 use super::iterate::reverse_post_order;
 use super::ControlFlowGraph;