Auto merge of #55674 - oli-obk:miri_engine_refactoring, r=RalfJung

Miri engine refactoring

r? @RalfJung

split out the "just moves stuff around" part of https://github.com/rust-lang/rust/pull/55293

src/librustc/mir/interpret/allocation.rs

@@ -0,0 +1,233 @@
+// Copyright 2018 The Rust Project Developers. See the COPYRIGHT
+// file at the top-level directory of this distribution and at
+// http://rust-lang.org/COPYRIGHT.
+//
+// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
+// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
+// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
+// option. This file may not be copied, modified, or distributed
+// except according to those terms.
+
+//! The virtual memory representation of the MIR interpreter
+
+use super::{Pointer, EvalResult, AllocId};
+
+use ty::layout::{Size, Align};
+use syntax::ast::Mutability;
+use std::iter;
+use mir;
+use std::ops::{Deref, DerefMut};
+use rustc_data_structures::sorted_map::SortedMap;
+
+#[derive(Clone, Debug, Eq, PartialEq, PartialOrd, Ord, Hash, RustcEncodable, RustcDecodable)]
+pub struct Allocation<Tag=(),Extra=()> {
+    /// The actual bytes of the allocation.
+    /// Note that the bytes of a pointer represent the offset of the pointer
+    pub bytes: Vec<u8>,
+    /// Maps from byte addresses to extra data for each pointer.
+    /// Only the first byte of a pointer is inserted into the map; i.e.,
+    /// every entry in this map applies to `pointer_size` consecutive bytes starting
+    /// at the given offset.
+    pub relocations: Relocations<Tag>,
+    /// Denotes undefined memory. Reading from undefined memory is forbidden in miri
+    pub undef_mask: UndefMask,
+    /// The alignment of the allocation to detect unaligned reads.
+    pub align: Align,
+    /// Whether the allocation is mutable.
+    /// Also used by codegen to determine if a static should be put into mutable memory,
+    /// which happens for `static mut` and `static` with interior mutability.
+    pub mutability: Mutability,
+    /// Extra state for the machine.
+    pub extra: Extra,
+}
+
+pub trait AllocationExtra<Tag>: ::std::fmt::Debug + Default + Clone {
+    /// Hook for performing extra checks on a memory read access.
+    ///
+    /// Takes read-only access to the allocation so we can keep all the memory read
+    /// operations take `&self`.  Use a `RefCell` in `AllocExtra` if you
+    /// need to mutate.
+    #[inline]
+    fn memory_read(
+        _alloc: &Allocation<Tag, Self>,
+        _ptr: Pointer<Tag>,
+        _size: Size,
+    ) -> EvalResult<'tcx> {
+        Ok(())
+    }
+
+    /// Hook for performing extra checks on a memory write access.
+    #[inline]
+    fn memory_written(
+        _alloc: &mut Allocation<Tag, Self>,
+        _ptr: Pointer<Tag>,
+        _size: Size,
+    ) -> EvalResult<'tcx> {
+        Ok(())
+    }
+}
+
+impl AllocationExtra<()> for () {}
+
+impl<Tag, Extra: Default> Allocation<Tag, Extra> {
+    /// Creates a read-only allocation initialized by the given bytes
+    pub fn from_bytes(slice: &[u8], align: Align) -> Self {
+        let mut undef_mask = UndefMask::new(Size::ZERO);
+        undef_mask.grow(Size::from_bytes(slice.len() as u64), true);
+        Self {
+            bytes: slice.to_owned(),
+            relocations: Relocations::new(),
+            undef_mask,
+            align,
+            mutability: Mutability::Immutable,
+            extra: Extra::default(),
+        }
+    }
+
+    pub fn from_byte_aligned_bytes(slice: &[u8]) -> Self {
+        Allocation::from_bytes(slice, Align::from_bytes(1, 1).unwrap())
+    }
+
+    pub fn undef(size: Size, align: Align) -> Self {
+        assert_eq!(size.bytes() as usize as u64, size.bytes());
+        Allocation {
+            bytes: vec![0; size.bytes() as usize],
+            relocations: Relocations::new(),
+            undef_mask: UndefMask::new(size),
+            align,
+            mutability: Mutability::Mutable,
+            extra: Extra::default(),
+        }
+    }
+}
+
+impl<'tcx> ::serialize::UseSpecializedDecodable for &'tcx Allocation {}
+
+#[derive(Clone, PartialEq, Eq, PartialOrd, Ord, Hash, Debug, RustcEncodable, RustcDecodable)]
+pub struct Relocations<Tag=(), Id=AllocId>(SortedMap<Size, (Tag, Id)>);
+
+impl<Tag, Id> Relocations<Tag, Id> {
+    pub fn new() -> Self {
+        Relocations(SortedMap::new())
+    }
+
+    // The caller must guarantee that the given relocations are already sorted
+    // by address and contain no duplicates.
+    pub fn from_presorted(r: Vec<(Size, (Tag, Id))>) -> Self {
+        Relocations(SortedMap::from_presorted_elements(r))
+    }
+}
+
+impl<Tag> Deref for Relocations<Tag> {
+    type Target = SortedMap<Size, (Tag, AllocId)>;
+
+    fn deref(&self) -> &Self::Target {
+        &self.0
+    }
+}
+
+impl<Tag> DerefMut for Relocations<Tag> {
+    fn deref_mut(&mut self) -> &mut Self::Target {
+        &mut self.0
+    }
+}
+
+////////////////////////////////////////////////////////////////////////////////
+// Undefined byte tracking
+////////////////////////////////////////////////////////////////////////////////
+
+type Block = u64;
+const BLOCK_SIZE: u64 = 64;
+
+#[derive(Clone, Debug, Eq, PartialEq, PartialOrd, Ord, Hash, RustcEncodable, RustcDecodable)]
+pub struct UndefMask {
+    blocks: Vec<Block>,
+    len: Size,
+}
+
+impl_stable_hash_for!(struct mir::interpret::UndefMask{blocks, len});
+
+impl UndefMask {
+    pub fn new(size: Size) -> Self {
+        let mut m = UndefMask {
+            blocks: vec![],
+            len: Size::ZERO,
+        };
+        m.grow(size, false);
+        m
+    }
+
+    /// Check whether the range `start..end` (end-exclusive) is entirely defined.
+    ///
+    /// Returns `Ok(())` if it's defined. Otherwise returns the index of the byte
+    /// at which the first undefined access begins.
+    #[inline]
+    pub fn is_range_defined(&self, start: Size, end: Size) -> Result<(), Size> {
+        if end > self.len {
+            return Err(self.len);
+        }
+
+        let idx = (start.bytes()..end.bytes())
+            .map(|i| Size::from_bytes(i))
+            .find(|&i| !self.get(i));
+
+        match idx {
+            Some(idx) => Err(idx),
+            None => Ok(())
+        }
+    }
+
+    pub fn set_range(&mut self, start: Size, end: Size, new_state: bool) {
+        let len = self.len;
+        if end > len {
+            self.grow(end - len, new_state);
+        }
+        self.set_range_inbounds(start, end, new_state);
+    }
+
+    pub fn set_range_inbounds(&mut self, start: Size, end: Size, new_state: bool) {
+        for i in start.bytes()..end.bytes() {
+            self.set(Size::from_bytes(i), new_state);
+        }
+    }
+
+    #[inline]
+    pub fn get(&self, i: Size) -> bool {
+        let (block, bit) = bit_index(i);
+        (self.blocks[block] & 1 << bit) != 0
+    }
+
+    #[inline]
+    pub fn set(&mut self, i: Size, new_state: bool) {
+        let (block, bit) = bit_index(i);
+        if new_state {
+            self.blocks[block] |= 1 << bit;
+        } else {
+            self.blocks[block] &= !(1 << bit);
+        }
+    }
+
+    pub fn grow(&mut self, amount: Size, new_state: bool) {
+        let unused_trailing_bits = self.blocks.len() as u64 * BLOCK_SIZE - self.len.bytes();
+        if amount.bytes() > unused_trailing_bits {
+            let additional_blocks = amount.bytes() / BLOCK_SIZE + 1;
+            assert_eq!(additional_blocks as usize as u64, additional_blocks);
+            self.blocks.extend(
+                iter::repeat(0).take(additional_blocks as usize),
+            );
+        }
+        let start = self.len;
+        self.len += amount;
+        self.set_range_inbounds(start, start + amount, new_state);
+    }
+}
+
+#[inline]
+fn bit_index(bits: Size) -> (usize, usize) {
+    let bits = bits.bytes();
+    let a = bits / BLOCK_SIZE;
+    let b = bits % BLOCK_SIZE;
+    assert_eq!(a as usize as u64, a);
+    assert_eq!(b as usize as u64, b);
+    (a as usize, b as usize)
+}

src/librustc/mir/interpret/mod.rs

@@ -17,27 +17,32 @@ macro_rules! err {
 
 mod error;
 mod value;
+mod allocation;
+mod pointer;
 
 pub use self::error::{
     EvalError, EvalResult, EvalErrorKind, AssertMessage, ConstEvalErr, struct_error,
     FrameInfo, ConstEvalResult, ErrorHandled,
 };
 
-pub use self::value::{Scalar, ConstValue};
+pub use self::value::{Scalar, ConstValue, ScalarMaybeUndef};
+
+pub use self::allocation::{
+    Allocation, AllocationExtra,
+    Relocations, UndefMask,
+};
+
+pub use self::pointer::{Pointer, PointerArithmetic};
 
 use std::fmt;
 use mir;
 use hir::def_id::DefId;
 use ty::{self, TyCtxt, Instance};
-use ty::layout::{self, Align, HasDataLayout, Size};
+use ty::layout::{self, Size};
 use middle::region;
-use std::iter;
 use std::io;
-use std::ops::{Deref, DerefMut};
 use std::hash::Hash;
-use syntax::ast::Mutability;
 use rustc_serialize::{Encoder, Decodable, Encodable};
-use rustc_data_structures::sorted_map::SortedMap;
 use rustc_data_structures::fx::FxHashMap;
 use rustc_data_structures::sync::{Lock as Mutex, HashMapExt};
 use rustc_data_structures::tiny_list::TinyList;
@@ -78,152 +83,6 @@ pub struct GlobalId<'tcx> {
     pub promoted: Option<mir::Promoted>,
 }
 
-////////////////////////////////////////////////////////////////////////////////
-// Pointer arithmetic
-////////////////////////////////////////////////////////////////////////////////
-
-pub trait PointerArithmetic: layout::HasDataLayout {
-    // These are not supposed to be overridden.
-
-    #[inline(always)]
-    fn pointer_size(&self) -> Size {
-        self.data_layout().pointer_size
-    }
-
-    //// Trunace the given value to the pointer size; also return whether there was an overflow
-    #[inline]
-    fn truncate_to_ptr(&self, val: u128) -> (u64, bool) {
-        let max_ptr_plus_1 = 1u128 << self.pointer_size().bits();
-        ((val % max_ptr_plus_1) as u64, val >= max_ptr_plus_1)
-    }
-
-    #[inline]
-    fn offset<'tcx>(&self, val: u64, i: u64) -> EvalResult<'tcx, u64> {
-        let (res, over) = self.overflowing_offset(val, i);
-        if over { err!(Overflow(mir::BinOp::Add)) } else { Ok(res) }
-    }
-
-    #[inline]
-    fn overflowing_offset(&self, val: u64, i: u64) -> (u64, bool) {
-        let (res, over1) = val.overflowing_add(i);
-        let (res, over2) = self.truncate_to_ptr(u128::from(res));
-        (res, over1 || over2)
-    }
-
-    #[inline]
-    fn signed_offset<'tcx>(&self, val: u64, i: i64) -> EvalResult<'tcx, u64> {
-        let (res, over) = self.overflowing_signed_offset(val, i128::from(i));
-        if over { err!(Overflow(mir::BinOp::Add)) } else { Ok(res) }
-    }
-
-    // Overflow checking only works properly on the range from -u64 to +u64.
-    #[inline]
-    fn overflowing_signed_offset(&self, val: u64, i: i128) -> (u64, bool) {
-        // FIXME: is it possible to over/underflow here?
-        if i < 0 {
-            // trickery to ensure that i64::min_value() works fine
-            // this formula only works for true negative values, it panics for zero!
-            let n = u64::max_value() - (i as u64) + 1;
-            val.overflowing_sub(n)
-        } else {
-            self.overflowing_offset(val, i as u64)
-        }
-    }
-}
-
-impl<T: layout::HasDataLayout> PointerArithmetic for T {}
-
-
-/// Pointer is generic over the type that represents a reference to Allocations,
-/// thus making it possible for the most convenient representation to be used in
-/// each context.
-///
-/// Defaults to the index based and loosely coupled AllocId.
-///
-/// Pointer is also generic over the `Tag` associated with each pointer,
-/// which is used to do provenance tracking during execution.
-#[derive(Copy, Clone, Debug, Eq, PartialEq, Ord, PartialOrd, RustcEncodable, RustcDecodable, Hash)]
-pub struct Pointer<Tag=(),Id=AllocId> {
-    pub alloc_id: Id,
-    pub offset: Size,
-    pub tag: Tag,
-}
-
-/// Produces a `Pointer` which points to the beginning of the Allocation
-impl From<AllocId> for Pointer {
-    #[inline(always)]
-    fn from(alloc_id: AllocId) -> Self {
-        Pointer::new(alloc_id, Size::ZERO)
-    }
-}
-
-impl<'tcx> Pointer<()> {
-    #[inline(always)]
-    pub fn new(alloc_id: AllocId, offset: Size) -> Self {
-        Pointer { alloc_id, offset, tag: () }
-    }
-
-    #[inline(always)]
-    pub fn with_default_tag<Tag>(self) -> Pointer<Tag>
-        where Tag: Default
-    {
-        Pointer::new_with_tag(self.alloc_id, self.offset, Default::default())
-    }
-}
-
-impl<'tcx, Tag> Pointer<Tag> {
-    #[inline(always)]
-    pub fn new_with_tag(alloc_id: AllocId, offset: Size, tag: Tag) -> Self {
-        Pointer { alloc_id, offset, tag }
-    }
-
-    #[inline]
-    pub fn offset(self, i: Size, cx: &impl HasDataLayout) -> EvalResult<'tcx, Self> {
-        Ok(Pointer::new_with_tag(
-            self.alloc_id,
-            Size::from_bytes(cx.data_layout().offset(self.offset.bytes(), i.bytes())?),
-            self.tag
-        ))
-    }
-
-    #[inline]
-    pub fn overflowing_offset(self, i: Size, cx: &impl HasDataLayout) -> (Self, bool) {
-        let (res, over) = cx.data_layout().overflowing_offset(self.offset.bytes(), i.bytes());
-        (Pointer::new_with_tag(self.alloc_id, Size::from_bytes(res), self.tag), over)
-    }
-
-    #[inline(always)]
-    pub fn wrapping_offset(self, i: Size, cx: &impl HasDataLayout) -> Self {
-        self.overflowing_offset(i, cx).0
-    }
-
-    #[inline]
-    pub fn signed_offset(self, i: i64, cx: &impl HasDataLayout) -> EvalResult<'tcx, Self> {
-        Ok(Pointer::new_with_tag(
-            self.alloc_id,
-            Size::from_bytes(cx.data_layout().signed_offset(self.offset.bytes(), i)?),
-            self.tag,
-        ))
-    }
-
-    #[inline]
-    pub fn overflowing_signed_offset(self, i: i128, cx: &impl HasDataLayout) -> (Self, bool) {
-        let (res, over) = cx.data_layout().overflowing_signed_offset(self.offset.bytes(), i);
-        (Pointer::new_with_tag(self.alloc_id, Size::from_bytes(res), self.tag), over)
-    }
-
-    #[inline(always)]
-    pub fn wrapping_signed_offset(self, i: i64, cx: &impl HasDataLayout) -> Self {
-        self.overflowing_signed_offset(i128::from(i), cx).0
-    }
-
-    #[inline(always)]
-    pub fn erase_tag(self) -> Pointer {
-        Pointer { alloc_id: self.alloc_id, offset: self.offset, tag: () }
-    }
-}
-
-
 #[derive(Copy, Clone, Eq, Hash, Ord, PartialEq, PartialOrd, Debug)]
 pub struct AllocId(pub u64);
 
@@ -528,91 +387,6 @@ impl<'tcx, M: fmt::Debug + Eq + Hash + Clone> AllocMap<'tcx, M> {
     }
 }
 
-#[derive(Clone, Debug, Eq, PartialEq, PartialOrd, Ord, Hash, RustcEncodable, RustcDecodable)]
-pub struct Allocation<Tag=(),Extra=()> {
-    /// The actual bytes of the allocation.
-    /// Note that the bytes of a pointer represent the offset of the pointer
-    pub bytes: Vec<u8>,
-    /// Maps from byte addresses to extra data for each pointer.
-    /// Only the first byte of a pointer is inserted into the map; i.e.,
-    /// every entry in this map applies to `pointer_size` consecutive bytes starting
-    /// at the given offset.
-    pub relocations: Relocations<Tag>,
-    /// Denotes undefined memory. Reading from undefined memory is forbidden in miri
-    pub undef_mask: UndefMask,
-    /// The alignment of the allocation to detect unaligned reads.
-    pub align: Align,
-    /// Whether the allocation is mutable.
-    /// Also used by codegen to determine if a static should be put into mutable memory,
-    /// which happens for `static mut` and `static` with interior mutability.
-    pub mutability: Mutability,
-    /// Extra state for the machine.
-    pub extra: Extra,
-}
-
-impl<Tag, Extra: Default> Allocation<Tag, Extra> {
-    /// Creates a read-only allocation initialized by the given bytes
-    pub fn from_bytes(slice: &[u8], align: Align) -> Self {
-        let mut undef_mask = UndefMask::new(Size::ZERO);
-        undef_mask.grow(Size::from_bytes(slice.len() as u64), true);
-        Self {
-            bytes: slice.to_owned(),
-            relocations: Relocations::new(),
-            undef_mask,
-            align,
-            mutability: Mutability::Immutable,
-            extra: Extra::default(),
-        }
-    }
-
-    pub fn from_byte_aligned_bytes(slice: &[u8]) -> Self {
-        Allocation::from_bytes(slice, Align::from_bytes(1, 1).unwrap())
-    }
-
-    pub fn undef(size: Size, align: Align) -> Self {
-        assert_eq!(size.bytes() as usize as u64, size.bytes());
-        Allocation {
-            bytes: vec![0; size.bytes() as usize],
-            relocations: Relocations::new(),
-            undef_mask: UndefMask::new(size),
-            align,
-            mutability: Mutability::Mutable,
-            extra: Extra::default(),
-        }
-    }
-}
-
-impl<'tcx> ::serialize::UseSpecializedDecodable for &'tcx Allocation {}
-
-#[derive(Clone, PartialEq, Eq, PartialOrd, Ord, Hash, Debug, RustcEncodable, RustcDecodable)]
-pub struct Relocations<Tag=(), Id=AllocId>(SortedMap<Size, (Tag, Id)>);
-
-impl<Tag, Id> Relocations<Tag, Id> {
-    pub fn new() -> Self {
-        Relocations(SortedMap::new())
-    }
-
-    // The caller must guarantee that the given relocations are already sorted
-    // by address and contain no duplicates.
-    pub fn from_presorted(r: Vec<(Size, (Tag, Id))>) -> Self {
-        Relocations(SortedMap::from_presorted_elements(r))
-    }
-}
-
-impl<Tag> Deref for Relocations<Tag> {
-    type Target = SortedMap<Size, (Tag, AllocId)>;
-
-    fn deref(&self) -> &Self::Target {
-        &self.0
-    }
-}
-
-impl<Tag> DerefMut for Relocations<Tag> {
-    fn deref_mut(&mut self) -> &mut Self::Target {
-        &mut self.0
-    }
-}
-
 ////////////////////////////////////////////////////////////////////////////////
 // Methods to access integers in the target endianness
 ////////////////////////////////////////////////////////////////////////////////
@@ -655,103 +429,3 @@ pub fn truncate(value: u128, size: Size) -> u128 {
     // truncate (shift left to drop out leftover values, shift right to fill with zeroes)
     (value << shift) >> shift
 }
-
-////////////////////////////////////////////////////////////////////////////////
-// Undefined byte tracking
-////////////////////////////////////////////////////////////////////////////////
-
-type Block = u64;
-const BLOCK_SIZE: u64 = 64;
-
-#[derive(Clone, Debug, Eq, PartialEq, PartialOrd, Ord, Hash, RustcEncodable, RustcDecodable)]
-pub struct UndefMask {
-    blocks: Vec<Block>,
-    len: Size,
-}
-
-impl_stable_hash_for!(struct mir::interpret::UndefMask{blocks, len});
-
-impl UndefMask {
-    pub fn new(size: Size) -> Self {
-        let mut m = UndefMask {
-            blocks: vec![],
-            len: Size::ZERO,
-        };
-        m.grow(size, false);
-        m
-    }
-
-    /// Check whether the range `start..end` (end-exclusive) is entirely defined.
-    ///
-    /// Returns `Ok(())` if it's defined. Otherwise returns the index of the byte
-    /// at which the first undefined access begins.
-    #[inline]
-    pub fn is_range_defined(&self, start: Size, end: Size) -> Result<(), Size> {
-        if end > self.len {
-            return Err(self.len);
-        }
-
-        let idx = (start.bytes()..end.bytes())
-            .map(|i| Size::from_bytes(i))
-            .find(|&i| !self.get(i));
-
-        match idx {
-            Some(idx) => Err(idx),
-            None => Ok(())
-        }
-    }
-
-    pub fn set_range(&mut self, start: Size, end: Size, new_state: bool) {
-        let len = self.len;
-        if end > len {
-            self.grow(end - len, new_state);
-        }
-        self.set_range_inbounds(start, end, new_state);
-    }
-
-    pub fn set_range_inbounds(&mut self, start: Size, end: Size, new_state: bool) {
-        for i in start.bytes()..end.bytes() {
-            self.set(Size::from_bytes(i), new_state);
-        }
-    }
-
-    #[inline]
-    pub fn get(&self, i: Size) -> bool {
-        let (block, bit) = bit_index(i);
-        (self.blocks[block] & 1 << bit) != 0
-    }
-
-    #[inline]
-    pub fn set(&mut self, i: Size, new_state: bool) {
-        let (block, bit) = bit_index(i);
-        if new_state {
-            self.blocks[block] |= 1 << bit;
-        } else {
-            self.blocks[block] &= !(1 << bit);
-        }
-    }
-
-    pub fn grow(&mut self, amount: Size, new_state: bool) {
-        let unused_trailing_bits = self.blocks.len() as u64 * BLOCK_SIZE - self.len.bytes();
-        if amount.bytes() > unused_trailing_bits {
-            let additional_blocks = amount.bytes() / BLOCK_SIZE + 1;
-            assert_eq!(additional_blocks as usize as u64, additional_blocks);
-            self.blocks.extend(
-                iter::repeat(0).take(additional_blocks as usize),
-            );
-        }
-        let start = self.len;
-        self.len += amount;
-        self.set_range_inbounds(start, start + amount, new_state);
-    }
-}
-
-#[inline]
-fn bit_index(bits: Size) -> (usize, usize) {
-    let bits = bits.bytes();
-    let a = bits / BLOCK_SIZE;
-    let b = bits % BLOCK_SIZE;
-    assert_eq!(a as usize as u64, a);
-    assert_eq!(b as usize as u64, b);
-    (a as usize, b as usize)
-}

src/librustc/mir/interpret/pointer.rs

@@ -0,0 +1,151 @@
+use mir;
+use ty::layout::{self, HasDataLayout, Size};
+
+use super::{
+    AllocId, EvalResult,
+};
+
+////////////////////////////////////////////////////////////////////////////////
+// Pointer arithmetic
+////////////////////////////////////////////////////////////////////////////////
+
+pub trait PointerArithmetic: layout::HasDataLayout {
+    // These are not supposed to be overridden.
+
+    #[inline(always)]
+    fn pointer_size(&self) -> Size {
+        self.data_layout().pointer_size
+    }
+
+    //// Trunace the given value to the pointer size; also return whether there was an overflow
+    #[inline]
+    fn truncate_to_ptr(&self, val: u128) -> (u64, bool) {
+        let max_ptr_plus_1 = 1u128 << self.pointer_size().bits();
+        ((val % max_ptr_plus_1) as u64, val >= max_ptr_plus_1)
+    }
+
+    #[inline]
+    fn offset<'tcx>(&self, val: u64, i: u64) -> EvalResult<'tcx, u64> {
+        let (res, over) = self.overflowing_offset(val, i);
+        if over { err!(Overflow(mir::BinOp::Add)) } else { Ok(res) }
+    }
+
+    #[inline]
+    fn overflowing_offset(&self, val: u64, i: u64) -> (u64, bool) {
+        let (res, over1) = val.overflowing_add(i);
+        let (res, over2) = self.truncate_to_ptr(u128::from(res));
+        (res, over1 || over2)
+    }
+
+    #[inline]
+    fn signed_offset<'tcx>(&self, val: u64, i: i64) -> EvalResult<'tcx, u64> {
+        let (res, over) = self.overflowing_signed_offset(val, i128::from(i));
+        if over { err!(Overflow(mir::BinOp::Add)) } else { Ok(res) }
+    }
+
+    // Overflow checking only works properly on the range from -u64 to +u64.
+    #[inline]
+    fn overflowing_signed_offset(&self, val: u64, i: i128) -> (u64, bool) {
+        // FIXME: is it possible to over/underflow here?
+        if i < 0 {
+            // trickery to ensure that i64::min_value() works fine
+            // this formula only works for true negative values, it panics for zero!
+            let n = u64::max_value() - (i as u64) + 1;
+            val.overflowing_sub(n)
+        } else {
+            self.overflowing_offset(val, i as u64)
+        }
+    }
+}
+
+impl<T: layout::HasDataLayout> PointerArithmetic for T {}
+
+
+/// Pointer is generic over the type that represents a reference to Allocations,
+/// thus making it possible for the most convenient representation to be used in
+/// each context.
+///
+/// Defaults to the index based and loosely coupled AllocId.
+///
+/// Pointer is also generic over the `Tag` associated with each pointer,
+/// which is used to do provenance tracking during execution.
+#[derive(Copy, Clone, Debug, Eq, PartialEq, Ord, PartialOrd, RustcEncodable, RustcDecodable, Hash)]
+pub struct Pointer<Tag=(),Id=AllocId> {
+    pub alloc_id: Id,
+    pub offset: Size,
+    pub tag: Tag,
+}
+
+/// Produces a `Pointer` which points to the beginning of the Allocation
+impl From<AllocId> for Pointer {
+    #[inline(always)]
+    fn from(alloc_id: AllocId) -> Self {
+        Pointer::new(alloc_id, Size::ZERO)
+    }
+}
+
+impl<'tcx> Pointer<()> {
+    #[inline(always)]
+    pub fn new(alloc_id: AllocId, offset: Size) -> Self {
+        Pointer { alloc_id, offset, tag: () }
+    }
+
+    #[inline(always)]
+    pub fn with_default_tag<Tag>(self) -> Pointer<Tag>
+        where Tag: Default
+    {
+        Pointer::new_with_tag(self.alloc_id, self.offset, Default::default())
+    }
+}
+
+impl<'tcx, Tag> Pointer<Tag> {
+    #[inline(always)]
+    pub fn new_with_tag(alloc_id: AllocId, offset: Size, tag: Tag) -> Self {
+        Pointer { alloc_id, offset, tag }
+    }
+
+    #[inline]
+    pub fn offset(self, i: Size, cx: &impl HasDataLayout) -> EvalResult<'tcx, Self> {
+        Ok(Pointer::new_with_tag(
+            self.alloc_id,
+            Size::from_bytes(cx.data_layout().offset(self.offset.bytes(), i.bytes())?),
+            self.tag
+        ))
+    }
+
+    #[inline]
+    pub fn overflowing_offset(self, i: Size, cx: &impl HasDataLayout) -> (Self, bool) {
+        let (res, over) = cx.data_layout().overflowing_offset(self.offset.bytes(), i.bytes());
+        (Pointer::new_with_tag(self.alloc_id, Size::from_bytes(res), self.tag), over)
+    }
+
+    #[inline(always)]
+    pub fn wrapping_offset(self, i: Size, cx: &impl HasDataLayout) -> Self {
+        self.overflowing_offset(i, cx).0
+    }
+
+    #[inline]
+    pub fn signed_offset(self, i: i64, cx: &impl HasDataLayout) -> EvalResult<'tcx, Self> {
+        Ok(Pointer::new_with_tag(
+            self.alloc_id,
+            Size::from_bytes(cx.data_layout().signed_offset(self.offset.bytes(), i)?),
+            self.tag,
+        ))
+    }
+
+    #[inline]
+    pub fn overflowing_signed_offset(self, i: i128, cx: &impl HasDataLayout) -> (Self, bool) {
+        let (res, over) = cx.data_layout().overflowing_signed_offset(self.offset.bytes(), i);
+        (Pointer::new_with_tag(self.alloc_id, Size::from_bytes(res), self.tag), over)
+    }
+
+    #[inline(always)]
+    pub fn wrapping_signed_offset(self, i: i64, cx: &impl HasDataLayout) -> Self {
+        self.overflowing_signed_offset(i128::from(i), cx).0
+    }
+
+    #[inline(always)]
+    pub fn erase_tag(self) -> Pointer {
+        Pointer { alloc_id: self.alloc_id, offset: self.offset, tag: () }
+    }
+}

src/librustc/mir/interpret/value.rs

@@ -392,3 +392,131 @@ impl<Tag> From<Pointer<Tag>> for Scalar<Tag> {
         Scalar::Ptr(ptr)
     }
 }
+
+#[derive(Clone, Copy, Debug, Eq, PartialEq, Ord, PartialOrd, RustcEncodable, RustcDecodable, Hash)]
+pub enum ScalarMaybeUndef<Tag=(), Id=AllocId> {
+    Scalar(Scalar<Tag, Id>),
+    Undef,
+}
+
+impl<Tag> From<Scalar<Tag>> for ScalarMaybeUndef<Tag> {
+    #[inline(always)]
+    fn from(s: Scalar<Tag>) -> Self {
+        ScalarMaybeUndef::Scalar(s)
+    }
+}
+
+impl<Tag> fmt::Display for ScalarMaybeUndef<Tag> {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            ScalarMaybeUndef::Undef => write!(f, "uninitialized bytes"),
+            ScalarMaybeUndef::Scalar(s) => write!(f, "{}", s),
+        }
+    }
+}
+
+impl<'tcx> ScalarMaybeUndef<()> {
+    #[inline]
+    pub fn with_default_tag<Tag>(self) -> ScalarMaybeUndef<Tag>
+        where Tag: Default
+    {
+        match self {
+            ScalarMaybeUndef::Scalar(s) => ScalarMaybeUndef::Scalar(s.with_default_tag()),
+            ScalarMaybeUndef::Undef => ScalarMaybeUndef::Undef,
+        }
+    }
+}
+
+impl<'tcx, Tag> ScalarMaybeUndef<Tag> {
+    #[inline]
+    pub fn erase_tag(self) -> ScalarMaybeUndef
+    {
+        match self {
+            ScalarMaybeUndef::Scalar(s) => ScalarMaybeUndef::Scalar(s.erase_tag()),
+            ScalarMaybeUndef::Undef => ScalarMaybeUndef::Undef,
+        }
+    }
+
+    #[inline]
+    pub fn not_undef(self) -> EvalResult<'static, Scalar<Tag>> {
+        match self {
+            ScalarMaybeUndef::Scalar(scalar) => Ok(scalar),
+            ScalarMaybeUndef::Undef => err!(ReadUndefBytes(Size::from_bytes(0))),
+        }
+    }
+
+    #[inline(always)]
+    pub fn to_ptr(self) -> EvalResult<'tcx, Pointer<Tag>> {
+        self.not_undef()?.to_ptr()
+    }
+
+    #[inline(always)]
+    pub fn to_bits(self, target_size: Size) -> EvalResult<'tcx, u128> {
+        self.not_undef()?.to_bits(target_size)
+    }
+
+    #[inline(always)]
+    pub fn to_bool(self) -> EvalResult<'tcx, bool> {
+        self.not_undef()?.to_bool()
+    }
+
+    #[inline(always)]
+    pub fn to_char(self) -> EvalResult<'tcx, char> {
+        self.not_undef()?.to_char()
+    }
+
+    #[inline(always)]
+    pub fn to_f32(self) -> EvalResult<'tcx, f32> {
+        self.not_undef()?.to_f32()
+    }
+
+    #[inline(always)]
+    pub fn to_f64(self) -> EvalResult<'tcx, f64> {
+        self.not_undef()?.to_f64()
+    }
+
+    #[inline(always)]
+    pub fn to_u8(self) -> EvalResult<'tcx, u8> {
+        self.not_undef()?.to_u8()
+    }
+
+    #[inline(always)]
+    pub fn to_u32(self) -> EvalResult<'tcx, u32> {
+        self.not_undef()?.to_u32()
+    }
+
+    #[inline(always)]
+    pub fn to_u64(self) -> EvalResult<'tcx, u64> {
+        self.not_undef()?.to_u64()
+    }
+
+    #[inline(always)]
+    pub fn to_usize(self, cx: &impl HasDataLayout) -> EvalResult<'tcx, u64> {
+        self.not_undef()?.to_usize(cx)
+    }
+
+    #[inline(always)]
+    pub fn to_i8(self) -> EvalResult<'tcx, i8> {
+        self.not_undef()?.to_i8()
+    }
+
+    #[inline(always)]
+    pub fn to_i32(self) -> EvalResult<'tcx, i32> {
+        self.not_undef()?.to_i32()
+    }
+
+    #[inline(always)]
+    pub fn to_i64(self) -> EvalResult<'tcx, i64> {
+        self.not_undef()?.to_i64()
+    }
+
+    #[inline(always)]
+    pub fn to_isize(self, cx: &impl HasDataLayout) -> EvalResult<'tcx, i64> {
+        self.not_undef()?.to_isize(cx)
+    }
+}
+
+impl_stable_hash_for!(enum ::mir::interpret::ScalarMaybeUndef {
+    Scalar(v),
+    Undef
+});

src/librustc_mir/interpret/machine.rs

@@ -20,7 +20,7 @@ use rustc::mir;
 use rustc::ty::{self, layout::{Size, TyLayout}, query::TyCtxtAt};
 
 use super::{
-    Allocation, AllocId, EvalResult, Scalar,
+    Allocation, AllocId, EvalResult, Scalar, AllocationExtra,
     EvalContext, PlaceTy, MPlaceTy, OpTy, Pointer, MemoryKind,
 };
 
@@ -78,7 +78,7 @@ pub trait Machine<'a, 'mir, 'tcx>: Sized {
     type PointerTag: ::std::fmt::Debug + Default + Copy + Eq + Hash + 'static;
 
     /// Extra data stored in every allocation.
-    type AllocExtra: ::std::fmt::Debug + Default + Clone;
+    type AllocExtra: AllocationExtra<Self::PointerTag>;
 
     /// Memory's allocation map
     type MemoryMap:
@@ -174,26 +174,6 @@ pub trait Machine<'a, 'mir, 'tcx>: Sized {
         dest: PlaceTy<'tcx, Self::PointerTag>,
     ) -> EvalResult<'tcx>;
 
-    /// Hook for performing extra checks on a memory read access.
-    #[inline]
-    fn memory_read(
-        _alloc: &Allocation<Self::PointerTag, Self::AllocExtra>,
-        _ptr: Pointer<Self::PointerTag>,
-        _size: Size,
-    ) -> EvalResult<'tcx> {
-        Ok(())
-    }
-
-    /// Hook for performing extra checks on a memory write access.
-    #[inline]
-    fn memory_written(
-        _alloc: &mut Allocation<Self::PointerTag, Self::AllocExtra>,
-        _ptr: Pointer<Self::PointerTag>,
-        _size: Size,
-    ) -> EvalResult<'tcx> {
-        Ok(())
-    }
-
     /// Hook for performing extra checks when memory gets deallocated.
     #[inline]
     fn memory_deallocated(

src/librustc_mir/interpret/memory.rs

@@ -28,7 +28,7 @@ use rustc_data_structures::fx::{FxHashSet, FxHashMap};
 use syntax::ast::Mutability;
 
 use super::{
-    Pointer, AllocId, Allocation, ConstValue, GlobalId,
+    Pointer, AllocId, Allocation, ConstValue, GlobalId, AllocationExtra,
     EvalResult, Scalar, EvalErrorKind, AllocType, PointerArithmetic,
     Machine, AllocMap, MayLeak, ScalarMaybeUndef, ErrorHandled,
 };
@@ -637,7 +637,7 @@ impl<'a, 'mir, 'tcx, M: Machine<'a, 'mir, 'tcx>> Memory<'a, 'mir, 'tcx, M> {
         }
 
         let alloc = self.get(ptr.alloc_id)?;
-        M::memory_read(alloc, ptr, size)?;
+        AllocationExtra::memory_read(alloc, ptr, size)?;
 
         assert_eq!(ptr.offset.bytes() as usize as u64, ptr.offset.bytes());
         assert_eq!(size.bytes() as usize as u64, size.bytes());
@@ -683,7 +683,7 @@ impl<'a, 'mir, 'tcx, M: Machine<'a, 'mir, 'tcx>> Memory<'a, 'mir, 'tcx, M> {
         self.clear_relocations(ptr, size)?;
 
         let alloc = self.get_mut(ptr.alloc_id)?;
-        M::memory_written(alloc, ptr, size)?;
+        AllocationExtra::memory_written(alloc, ptr, size)?;
 
         assert_eq!(ptr.offset.bytes() as usize as u64, ptr.offset.bytes());
         assert_eq!(size.bytes() as usize as u64, size.bytes());

src/librustc_mir/interpret/operand.rs

@@ -12,7 +12,6 @@
 //! All high-level functions to read from memory work on operands as sources.
 
 use std::convert::TryInto;
-use std::fmt;
 
 use rustc::{mir, ty};
 use rustc::ty::layout::{self, Size, LayoutOf, TyLayout, HasDataLayout, IntegerExt};
@@ -23,130 +22,7 @@ use rustc::mir::interpret::{
     EvalResult, EvalErrorKind
 };
 use super::{EvalContext, Machine, MemPlace, MPlaceTy, MemoryKind};
-
+pub use rustc::mir::interpret::ScalarMaybeUndef;
-#[derive(Clone, Copy, Debug, Eq, PartialEq, Ord, PartialOrd, RustcEncodable, RustcDecodable, Hash)]
-pub enum ScalarMaybeUndef<Tag=(), Id=AllocId> {
-    Scalar(Scalar<Tag, Id>),
-    Undef,
-}
-
-impl<Tag> From<Scalar<Tag>> for ScalarMaybeUndef<Tag> {
-    #[inline(always)]
-    fn from(s: Scalar<Tag>) -> Self {
-        ScalarMaybeUndef::Scalar(s)
-    }
-}
-
-impl<Tag> fmt::Display for ScalarMaybeUndef<Tag> {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
-        match self {
-            ScalarMaybeUndef::Undef => write!(f, "uninitialized bytes"),
-            ScalarMaybeUndef::Scalar(s) => write!(f, "{}", s),
-        }
-    }
-}
-
-impl<'tcx> ScalarMaybeUndef<()> {
-    #[inline]
-    pub fn with_default_tag<Tag>(self) -> ScalarMaybeUndef<Tag>
-        where Tag: Default
-    {
-        match self {
-            ScalarMaybeUndef::Scalar(s) => ScalarMaybeUndef::Scalar(s.with_default_tag()),
-            ScalarMaybeUndef::Undef => ScalarMaybeUndef::Undef,
-        }
-    }
-}
-
-impl<'tcx, Tag> ScalarMaybeUndef<Tag> {
-    #[inline]
-    pub fn erase_tag(self) -> ScalarMaybeUndef
-    {
-        match self {
-            ScalarMaybeUndef::Scalar(s) => ScalarMaybeUndef::Scalar(s.erase_tag()),
-            ScalarMaybeUndef::Undef => ScalarMaybeUndef::Undef,
-        }
-    }
-
-    #[inline]
-    pub fn not_undef(self) -> EvalResult<'static, Scalar<Tag>> {
-        match self {
-            ScalarMaybeUndef::Scalar(scalar) => Ok(scalar),
-            ScalarMaybeUndef::Undef => err!(ReadUndefBytes(Size::from_bytes(0))),
-        }
-    }
-
-    #[inline(always)]
-    pub fn to_ptr(self) -> EvalResult<'tcx, Pointer<Tag>> {
-        self.not_undef()?.to_ptr()
-    }
-
-    #[inline(always)]
-    pub fn to_bits(self, target_size: Size) -> EvalResult<'tcx, u128> {
-        self.not_undef()?.to_bits(target_size)
-    }
-
-    #[inline(always)]
-    pub fn to_bool(self) -> EvalResult<'tcx, bool> {
-        self.not_undef()?.to_bool()
-    }
-
-    #[inline(always)]
-    pub fn to_char(self) -> EvalResult<'tcx, char> {
-        self.not_undef()?.to_char()
-    }
-
-    #[inline(always)]
-    pub fn to_f32(self) -> EvalResult<'tcx, f32> {
-        self.not_undef()?.to_f32()
-    }
-
-    #[inline(always)]
-    pub fn to_f64(self) -> EvalResult<'tcx, f64> {
-        self.not_undef()?.to_f64()
-    }
-
-    #[inline(always)]
-    pub fn to_u8(self) -> EvalResult<'tcx, u8> {
-        self.not_undef()?.to_u8()
-    }
-
-    #[inline(always)]
-    pub fn to_u32(self) -> EvalResult<'tcx, u32> {
-        self.not_undef()?.to_u32()
-    }
-
-    #[inline(always)]
-    pub fn to_u64(self) -> EvalResult<'tcx, u64> {
-        self.not_undef()?.to_u64()
-    }
-
-    #[inline(always)]
-    pub fn to_usize(self, cx: &impl HasDataLayout) -> EvalResult<'tcx, u64> {
-        self.not_undef()?.to_usize(cx)
-    }
-
-    #[inline(always)]
-    pub fn to_i8(self) -> EvalResult<'tcx, i8> {
-        self.not_undef()?.to_i8()
-    }
-
-    #[inline(always)]
-    pub fn to_i32(self) -> EvalResult<'tcx, i32> {
-        self.not_undef()?.to_i32()
-    }
-
-    #[inline(always)]
-    pub fn to_i64(self) -> EvalResult<'tcx, i64> {
-        self.not_undef()?.to_i64()
-    }
-
-    #[inline(always)]
-    pub fn to_isize(self, cx: &impl HasDataLayout) -> EvalResult<'tcx, i64> {
-        self.not_undef()?.to_isize(cx)
-    }
-}
-
 
 /// A `Value` represents a single immediate self-contained Rust value.
 ///

src/librustc_mir/interpret/place.rs

@@ -24,7 +24,7 @@ use rustc::mir::interpret::{
     GlobalId, AllocId, Allocation, Scalar, EvalResult, Pointer, PointerArithmetic
 };
 use super::{
-    EvalContext, Machine, AllocMap,
+    EvalContext, Machine, AllocMap, AllocationExtra,
     Immediate, ImmTy, ScalarMaybeUndef, Operand, OpTy, MemoryKind
 };
 
@@ -264,6 +264,7 @@ where
     Tag: ::std::fmt::Debug+Default+Copy+Eq+Hash+'static,
     M: Machine<'a, 'mir, 'tcx, PointerTag=Tag>,
     M::MemoryMap: AllocMap<AllocId, (MemoryKind<M::MemoryKinds>, Allocation<Tag, M::AllocExtra>)>,
+    M::AllocExtra: AllocationExtra<Tag>,
 {
     /// Take a value, which represents a (thin or fat) reference, and make it a place.
     /// Alignment is just based on the type.  This is the inverse of `create_ref`.

src/librustc_mir/interpret/snapshot.rs

@@ -195,11 +195,6 @@ impl<'a, Ctx> Snapshot<'a, Ctx> for Scalar
     }
 }
 
-impl_stable_hash_for!(enum ::interpret::ScalarMaybeUndef {
-    Scalar(v),
-    Undef
-});
-
 impl_snapshot_for!(enum ScalarMaybeUndef {
     Scalar(s),
     Undef,

src/librustc_mir/interpret/validity.rs

@@ -17,7 +17,7 @@ use rustc::ty::layout::{self, Size, Align, TyLayout, LayoutOf};
 use rustc::ty;
 use rustc_data_structures::fx::FxHashSet;
 use rustc::mir::interpret::{
-    Scalar, AllocType, EvalResult, EvalErrorKind
+    Scalar, AllocType, EvalResult, EvalErrorKind,
 };
 
 use super::{