Rollup merge of #95785 - RalfJung:interpret-size-mismatch, r=oli-obk

interpret: err instead of ICE on size mismatches in to_bits_or_ptr_internal

We did this a while ago already for `to_i32()` and friends, but missed this one. That became quite annoying when I was debugging an ICE caused by `read_pointer` in a Miri shim where the code was passing an argument at the wrong type.

Having `scalar_to_ptr` be fallible is consistent with all the other `Scalar::to_*` methods being fallible. I added `unwrap` only in code outside the interpreter, which is no worse off than before now in terms of panics.

r? ````@oli-obk````

compiler/rustc_middle/src/mir/interpret/allocation.rs

@@ -15,8 +15,8 @@ use rustc_target::abi::{Align, HasDataLayout, Size};
 
 use super::{
     read_target_uint, write_target_uint, AllocId, InterpError, InterpResult, Pointer, Provenance,
-    ResourceExhaustionInfo, Scalar, ScalarMaybeUninit, UndefinedBehaviorInfo, UninitBytesAccess,
-    UnsupportedOpInfo,
+    ResourceExhaustionInfo, Scalar, ScalarMaybeUninit, ScalarSizeMismatch, UndefinedBehaviorInfo,
+    UninitBytesAccess, UnsupportedOpInfo,
 };
 use crate::ty;
 
@@ -81,6 +81,8 @@ impl<'tcx, Tag, Extra> ConstAllocation<'tcx, Tag, Extra> {
 /// is added when converting to `InterpError`.
 #[derive(Debug)]
 pub enum AllocError {
+    /// A scalar had the wrong size.
+    ScalarSizeMismatch(ScalarSizeMismatch),
     /// Encountered a pointer where we needed raw bytes.
     ReadPointerAsBytes,
     /// Partially overwriting a pointer.
@@ -90,10 +92,19 @@ pub enum AllocError {
 }
 pub type AllocResult<T = ()> = Result<T, AllocError>;
 
+impl From<ScalarSizeMismatch> for AllocError {
+    fn from(s: ScalarSizeMismatch) -> Self {
+        AllocError::ScalarSizeMismatch(s)
+    }
+}
+
 impl AllocError {
     pub fn to_interp_error<'tcx>(self, alloc_id: AllocId) -> InterpError<'tcx> {
         use AllocError::*;
         match self {
+            ScalarSizeMismatch(s) => {
+                InterpError::UndefinedBehavior(UndefinedBehaviorInfo::ScalarSizeMismatch(s))
+            }
             ReadPointerAsBytes => InterpError::Unsupported(UnsupportedOpInfo::ReadPointerAsBytes),
             PartialPointerOverwrite(offset) => InterpError::Unsupported(
                 UnsupportedOpInfo::PartialPointerOverwrite(Pointer::new(alloc_id, offset)),
@@ -425,7 +436,7 @@ impl<Tag: Provenance, Extra> Allocation<Tag, Extra> {
 
         // `to_bits_or_ptr_internal` is the right method because we just want to store this data
         // as-is into memory.
-        let (bytes, provenance) = match val.to_bits_or_ptr_internal(range.size) {
+        let (bytes, provenance) = match val.to_bits_or_ptr_internal(range.size)? {
             Err(val) => {
                 let (provenance, offset) = val.into_parts();
                 (u128::from(offset.bytes()), Some(provenance))

compiler/rustc_middle/src/mir/interpret/error.rs

@@ -221,6 +221,13 @@ pub struct UninitBytesAccess {
     pub uninit_size: Size,
 }
 
+/// Information about a size mismatch.
+#[derive(Debug)]
+pub struct ScalarSizeMismatch {
+    pub target_size: u64,
+    pub data_size: u64,
+}
+
 /// Error information for when the program caused Undefined Behavior.
 pub enum UndefinedBehaviorInfo<'tcx> {
     /// Free-form case. Only for errors that are never caught!
@@ -298,10 +305,7 @@ pub enum UndefinedBehaviorInfo<'tcx> {
     /// Working with a local that is not currently live.
     DeadLocal,
     /// Data size is not equal to target size.
-    ScalarSizeMismatch {
-        target_size: u64,
-        data_size: u64,
-    },
+    ScalarSizeMismatch(ScalarSizeMismatch),
     /// A discriminant of an uninhabited enum variant is written.
     UninhabitedEnumVariantWritten,
 }
@@ -408,7 +412,7 @@ impl fmt::Display for UndefinedBehaviorInfo<'_> {
                 "using uninitialized data, but this operation requires initialized memory"
             ),
             DeadLocal => write!(f, "accessing a dead local variable"),
-            ScalarSizeMismatch { target_size, data_size } => write!(
+            ScalarSizeMismatch(self::ScalarSizeMismatch { target_size, data_size }) => write!(
                 f,
                 "scalar size mismatch: expected {} bytes but got {} bytes instead",
                 target_size, data_size

compiler/rustc_middle/src/mir/interpret/mod.rs

@@ -120,7 +120,8 @@ use crate::ty::{self, Instance, Ty, TyCtxt};
 pub use self::error::{
     struct_error, CheckInAllocMsg, ErrorHandled, EvalToAllocationRawResult, EvalToConstValueResult,
     InterpError, InterpErrorInfo, InterpResult, InvalidProgramInfo, MachineStopType,
-    ResourceExhaustionInfo, UndefinedBehaviorInfo, UninitBytesAccess, UnsupportedOpInfo,
+    ResourceExhaustionInfo, ScalarSizeMismatch, UndefinedBehaviorInfo, UninitBytesAccess,
+    UnsupportedOpInfo,
 };
 
 pub use self::value::{get_slice_bytes, ConstAlloc, ConstValue, Scalar, ScalarMaybeUninit};

compiler/rustc_middle/src/mir/interpret/value.rs

@@ -12,6 +12,7 @@ use crate::ty::{Lift, ParamEnv, ScalarInt, Ty, TyCtxt};
 
 use super::{
     AllocId, AllocRange, ConstAllocation, InterpResult, Pointer, PointerArithmetic, Provenance,
+    ScalarSizeMismatch,
 };
 
 /// Represents the result of const evaluation via the `eval_to_allocation` query.
@@ -300,16 +301,29 @@ impl<Tag> Scalar<Tag> {
     ///
     /// This method only exists for the benefit of low-level operations that truly need to treat the
     /// scalar in whatever form it is.
+    ///
+    /// This throws UB (instead of ICEing) on a size mismatch since size mismatches can arise in
+    /// Miri when someone declares a function that we shim (such as `malloc`) with a wrong type.
     #[inline]
-    pub fn to_bits_or_ptr_internal(self, target_size: Size) -> Result<u128, Pointer<Tag>> {
+    pub fn to_bits_or_ptr_internal(
+        self,
+        target_size: Size,
+    ) -> Result<Result<u128, Pointer<Tag>>, ScalarSizeMismatch> {
         assert_ne!(target_size.bytes(), 0, "you should never look at the bits of a ZST");
-        match self {
-            Scalar::Int(int) => Ok(int.assert_bits(target_size)),
+        Ok(match self {
+            Scalar::Int(int) => Ok(int.to_bits(target_size).map_err(|size| {
+                ScalarSizeMismatch { target_size: target_size.bytes(), data_size: size.bytes() }
+            })?),
             Scalar::Ptr(ptr, sz) => {
-                assert_eq!(target_size.bytes(), u64::from(sz));
+                if target_size.bytes() != sz.into() {
+                    return Err(ScalarSizeMismatch {
+                        target_size: target_size.bytes(),
+                        data_size: sz.into(),
+                    });
+                }
                 Err(ptr)
             }
-        }
+        })
     }
 }
 
@@ -348,10 +362,10 @@ impl<'tcx, Tag: Provenance> Scalar<Tag> {
         assert_ne!(target_size.bytes(), 0, "you should never look at the bits of a ZST");
         self.try_to_int().map_err(|_| err_unsup!(ReadPointerAsBytes))?.to_bits(target_size).map_err(
             |size| {
-                err_ub!(ScalarSizeMismatch {
+                err_ub!(ScalarSizeMismatch(ScalarSizeMismatch {
                     target_size: target_size.bytes(),
                     data_size: size.bytes(),
-                })
+                }))
                 .into()
             },
         )