Add instance evaluation and methods to read alloc

The instance evaluation is needed to handle intrinsics such as
`type_id` and `type_name`.

Since we now use Allocation to represent all evaluated constants,
provide a few methods to help process the data inside an allocation.

compiler/stable_mir/src/compiler_interface.rs

@@ -8,6 +8,7 @@ use std::cell::Cell;
 use crate::mir::alloc::{AllocId, GlobalAlloc};
 use crate::mir::mono::{Instance, InstanceDef, StaticDef};
 use crate::mir::Body;
+use crate::target::MachineInfo;
 use crate::ty::{
     AdtDef, AdtKind, Allocation, ClosureDef, ClosureKind, Const, FieldDef, FnDef, GenericArgs,
     GenericPredicates, Generics, ImplDef, ImplTrait, LineInfo, PolyFnSig, RigidTy, Span, TraitDecl,
@@ -150,6 +151,9 @@ pub trait Context {
     /// Evaluate a static's initializer.
     fn eval_static_initializer(&self, def: StaticDef) -> Result<Allocation, Error>;
 
+    /// Try to evaluate an instance into a constant.
+    fn eval_instance(&self, def: InstanceDef, const_ty: Ty) -> Result<Allocation, Error>;
+
     /// Retrieve global allocation for the given allocation ID.
     fn global_alloc(&self, id: AllocId) -> GlobalAlloc;
 
@@ -157,6 +161,9 @@ pub trait Context {
     fn vtable_allocation(&self, global_alloc: &GlobalAlloc) -> Option<AllocId>;
     fn krate(&self, def_id: DefId) -> Crate;
     fn instance_name(&self, def: InstanceDef, trimmed: bool) -> Symbol;
+
+    /// Return the number of bytes for a pointer size.
+    fn target_info(&self) -> MachineInfo;
 }
 
 // A thread local variable that stores a pointer to the tables mapping between TyCtxt

compiler/stable_mir/src/error.rs

@@ -6,11 +6,11 @@
 
 use std::convert::From;
 use std::fmt::{Debug, Display, Formatter};
-use std::{error, fmt};
+use std::{error, fmt, io};
 
 macro_rules! error {
      ($fmt: literal $(,)?) => { Error(format!($fmt)) };
-     ($fmt: literal, $($arg:tt)*) => { Error(format!($fmt, $($arg:tt)*)) };
+     ($fmt: literal, $($arg:tt)*) => { Error(format!($fmt, $($arg)*)) };
  }
 
 /// An error type used to represent an error that has already been reported by the compiler.
@@ -79,3 +79,9 @@ where
 
 impl error::Error for Error {}
 impl<T> error::Error for CompilerError<T> where T: Display + Debug {}
+
+impl From<io::Error> for Error {
+    fn from(value: io::Error) -> Self {
+        Error(value.to_string())
+    }
+}

compiler/stable_mir/src/lib.rs

@@ -39,6 +39,7 @@ pub mod compiler_interface;
 #[macro_use]
 pub mod error;
 pub mod mir;
+pub mod target;
 pub mod ty;
 pub mod visitor;
 

compiler/stable_mir/src/mir/alloc.rs

@@ -1,7 +1,9 @@
 //! This module provides methods to retrieve allocation information, such as static variables.
 use crate::mir::mono::{Instance, StaticDef};
+use crate::target::{Endian, MachineInfo};
 use crate::ty::{Allocation, Binder, ExistentialTraitRef, IndexedVal, Ty};
-use crate::with;
+use crate::{with, Error};
+use std::io::Read;
 
 /// An allocation in the SMIR global memory can be either a function pointer,
 /// a static, or a "real" allocation with some data in it.
@@ -38,7 +40,7 @@ impl GlobalAlloc {
 }
 
 /// A unique identification number for each provenance
-#[derive(Clone, Copy, PartialEq, Eq, Debug)]
+#[derive(Clone, Copy, PartialEq, Eq, Debug, Hash)]
 pub struct AllocId(usize);
 
 impl IndexedVal for AllocId {
@@ -49,3 +51,33 @@ impl IndexedVal for AllocId {
         self.0
     }
 }
+
+/// Utility function used to read an allocation data into a unassigned integer.
+pub(crate) fn read_target_uint(mut bytes: &[u8]) -> Result<u128, Error> {
+    let mut buf = [0u8; std::mem::size_of::<u128>()];
+    match MachineInfo::target_endianess() {
+        Endian::Little => {
+            bytes.read(&mut buf)?;
+            Ok(u128::from_le_bytes(buf))
+        }
+        Endian::Big => {
+            bytes.read(&mut buf[16 - bytes.len()..])?;
+            Ok(u128::from_be_bytes(buf))
+        }
+    }
+}
+
+/// Utility function used to read an allocation data into an assigned integer.
+pub(crate) fn read_target_int(mut bytes: &[u8]) -> Result<i128, Error> {
+    let mut buf = [0u8; std::mem::size_of::<i128>()];
+    match MachineInfo::target_endianess() {
+        Endian::Little => {
+            bytes.read(&mut buf)?;
+            Ok(i128::from_le_bytes(buf))
+        }
+        Endian::Big => {
+            bytes.read(&mut buf[16 - bytes.len()..])?;
+            Ok(i128::from_be_bytes(buf))
+        }
+    }
+}

compiler/stable_mir/src/mir/body.rs

@@ -832,7 +832,7 @@ pub enum MutBorrowKind {
     ClosureCapture,
 }
 
-#[derive(Copy, Clone, Debug, PartialEq, Eq)]
+#[derive(Copy, Clone, Debug, PartialEq, Eq, Hash)]
 pub enum Mutability {
     Not,
     Mut,

compiler/stable_mir/src/mir/mono.rs

@@ -132,6 +132,14 @@ impl Instance {
     pub fn is_empty_shim(&self) -> bool {
         self.kind == InstanceKind::Shim && with(|cx| cx.is_empty_drop_shim(self.def))
     }
+
+    /// Try to constant evaluate the instance into a constant with the given type.
+    ///
+    /// This can be used to retrieve a constant that represents an intrinsic return such as
+    /// `type_id`.
+    pub fn try_const_eval(&self, const_ty: Ty) -> Result<Allocation, Error> {
+        with(|cx| cx.eval_instance(self.def, const_ty))
+    }
 }
 
 impl Debug for Instance {

compiler/stable_mir/src/target.rs

@@ -0,0 +1,50 @@
+//! Provide information about the machine that this is being compiled into.
+
+use crate::compiler_interface::with;
+
+/// The properties of the target machine being compiled into.
+#[derive(Clone, PartialEq, Eq)]
+pub struct MachineInfo {
+    pub endian: Endian,
+    pub pointer_width: MachineSize,
+}
+
+impl MachineInfo {
+    pub fn target() -> MachineInfo {
+        with(|cx| cx.target_info().clone())
+    }
+
+    pub fn target_endianess() -> Endian {
+        with(|cx| cx.target_info().endian)
+    }
+
+    pub fn target_pointer_width() -> MachineSize {
+        with(|cx| cx.target_info().pointer_width)
+    }
+}
+
+#[derive(Copy, Clone, PartialEq, Eq)]
+pub enum Endian {
+    Little,
+    Big,
+}
+
+/// Represent the size of a component.
+#[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord)]
+pub struct MachineSize {
+    num_bits: usize,
+}
+
+impl MachineSize {
+    pub fn bytes(self) -> usize {
+        self.num_bits / 8
+    }
+
+    pub fn bits(self) -> usize {
+        self.num_bits
+    }
+
+    pub fn from_bits(num_bits: usize) -> MachineSize {
+        MachineSize { num_bits }
+    }
+}

compiler/stable_mir/src/ty.rs

@@ -4,9 +4,11 @@ use super::{
     with, DefId, Error, Symbol,
 };
 use crate::crate_def::CrateDef;
-use crate::mir::alloc::AllocId;
+use crate::mir::alloc::{read_target_int, read_target_uint, AllocId};
+use crate::target::MachineInfo;
 use crate::{Filename, Opaque};
 use std::fmt::{self, Debug, Display, Formatter};
+use std::ops::Range;
 
 #[derive(Copy, Clone, Eq, PartialEq, Hash)]
 pub struct Ty(pub usize);
@@ -366,6 +368,19 @@ pub enum IntTy {
     I128,
 }
 
+impl IntTy {
+    pub fn num_bytes(self) -> usize {
+        match self {
+            IntTy::Isize => crate::target::MachineInfo::target_pointer_width().bytes().into(),
+            IntTy::I8 => 1,
+            IntTy::I16 => 2,
+            IntTy::I32 => 4,
+            IntTy::I64 => 8,
+            IntTy::I128 => 16,
+        }
+    }
+}
+
 #[derive(Clone, Copy, Debug, PartialEq, Eq)]
 pub enum UintTy {
     Usize,
@@ -376,6 +391,19 @@ pub enum UintTy {
     U128,
 }
 
+impl UintTy {
+    pub fn num_bytes(self) -> usize {
+        match self {
+            UintTy::Usize => crate::target::MachineInfo::target_pointer_width().bytes().into(),
+            UintTy::U8 => 1,
+            UintTy::U16 => 2,
+            UintTy::U32 => 4,
+            UintTy::U64 => 8,
+            UintTy::U128 => 16,
+        }
+    }
+}
+
 #[derive(Clone, Copy, Debug, PartialEq, Eq)]
 pub enum FloatTy {
     F32,
@@ -821,21 +849,21 @@ pub struct BoundTy {
 pub type Bytes = Vec<Option<u8>>;
 pub type Size = usize;
 
-#[derive(Clone, Copy, PartialEq, Eq, Debug)]
+#[derive(Clone, Copy, PartialEq, Eq, Debug, Hash)]
 pub struct Prov(pub AllocId);
 pub type Align = u64;
 pub type Promoted = u32;
 pub type InitMaskMaterialized = Vec<u64>;
 
 /// Stores the provenance information of pointers stored in memory.
-#[derive(Clone, Debug, Eq, PartialEq)]
+#[derive(Clone, Debug, Eq, PartialEq, Hash)]
 pub struct ProvenanceMap {
     /// Provenance in this map applies from the given offset for an entire pointer-size worth of
     /// bytes. Two entries in this map are always at least a pointer size apart.
     pub ptrs: Vec<(Size, Prov)>,
 }
 
-#[derive(Clone, Debug, Eq, PartialEq)]
+#[derive(Clone, Debug, Eq, PartialEq, Hash)]
 pub struct Allocation {
     pub bytes: Bytes,
     pub provenance: ProvenanceMap,
@@ -843,6 +871,70 @@ pub struct Allocation {
     pub mutability: Mutability,
 }
 
+impl Allocation {
+    /// Get a vector of bytes for an Allocation that has been fully initialized
+    pub fn raw_bytes(&self) -> Result<Vec<u8>, Error> {
+        self.bytes
+            .iter()
+            .copied()
+            .collect::<Option<Vec<_>>>()
+            .ok_or_else(|| error!("Found uninitialized bytes: `{:?}`", self.bytes))
+    }
+
+    /// Read a uint value from the specified range.
+    pub fn read_partial_uint(&self, range: Range<usize>) -> Result<u128, Error> {
+        if range.end - range.start > 16 {
+            return Err(error!("Allocation is bigger than largest integer"));
+        }
+        if range.end > self.bytes.len() {
+            return Err(error!(
+                "Range is out of bounds. Allocation length is `{}`, but requested range `{:?}`",
+                self.bytes.len(),
+                range
+            ));
+        }
+        let raw = self.bytes[range]
+            .iter()
+            .copied()
+            .collect::<Option<Vec<_>>>()
+            .ok_or_else(|| error!("Found uninitialized bytes: `{:?}`", self.bytes))?;
+        read_target_uint(&raw)
+    }
+
+    pub fn read_uint(&self) -> Result<u128, Error> {
+        if self.bytes.len() > 16 {
+            return Err(error!("Allocation is bigger than largest integer"));
+        }
+        let raw = self.raw_bytes()?;
+        read_target_uint(&raw)
+    }
+
+    pub fn read_int(&self) -> Result<i128, Error> {
+        if self.bytes.len() > 16 {
+            return Err(error!("Allocation is bigger than largest integer"));
+        }
+        let raw = self.raw_bytes()?;
+        read_target_int(&raw)
+    }
+
+    pub fn read_bool(&self) -> Result<bool, Error> {
+        match self.read_int()? {
+            0 => Ok(false),
+            1 => Ok(true),
+            val @ _ => Err(error!("Unexpected value for bool: `{val}`")),
+        }
+    }
+
+    pub fn is_null(&self) -> Result<bool, Error> {
+        let len = self.bytes.len();
+        let ptr_len = MachineInfo::target_pointer_width().bytes();
+        if len != ptr_len {
+            return Err(error!("Expected width of pointer (`{ptr_len}`), but found: `{len}`"));
+        }
+        Ok(self.read_uint()? == 0)
+    }
+}
+
 #[derive(Clone, Debug, Eq, PartialEq)]
 pub enum ConstantKind {
     Allocated(Allocation),