Struct InterpCx

pub struct InterpCx<'tcx, M: Machine<'tcx>> {
    pub machine: M,
    pub tcx: TyCtxtAt<'tcx>,
    pub(crate) param_env: ParamEnv<'tcx>,
    pub memory: Memory<'tcx, M>,
    pub recursion_limit: Limit,
}

Fields

machine: M

Stores the Machine instance.

Note: the stack is provided by the machine.

tcx: TyCtxtAt<'tcx>

The results of the type checker, from rustc. The span in this is the “root” of the evaluation, i.e., the const we are evaluating (if this is CTFE).

param_env: ParamEnv<'tcx>

Bounds in scope for polymorphic evaluations.

memory: Memory<'tcx, M>

The virtual memory system.

recursion_limit: Limit

The recursion limit (cached from tcx.recursion_limit(()))

Implementations

impl<'tcx> InterpCx<'tcx, CompileTimeMachine<'tcx>>

fn location_triple_for_span(&self, span: Span) -> (Symbol, u32, u32)

fn hook_special_const_fn( &mut self, instance: Instance<'tcx>, args: &[FnArg<'tcx>], dest: &MPlaceTy<'tcx>, ret: Option<BasicBlock>, ) -> InterpResult<'tcx, Option<Instance<'tcx>>>

“Intercept” a function call, because we have something special to do for it. All #[rustc_do_not_const_check] functions should be hooked here. If this returns Some function, which may be instance or a different function with compatible arguments, then evaluation should continue with that function. If this returns None, the function call has been handled and the function has returned.

fn align_offset( &mut self, instance: Instance<'tcx>, args: &[OpTy<'tcx>], dest: &MPlaceTy<'tcx>, ret: Option<BasicBlock>, ) -> InterpResult<'tcx, ControlFlow<()>>

align_offset(ptr, target_align) needs special handling in const eval, because the pointer may not have an address.

If ptr does have a known address, then we return Continue(()) and the function call should proceed as normal.

If ptr doesn’t have an address, but its underlying allocation’s alignment is at most target_align, then we call the function again with an dummy address relative to the allocation.

If ptr doesn’t have an address and target_align is stricter than the underlying allocation’s alignment, then we return usize::MAX immediately.

fn guaranteed_cmp(&mut self, a: Scalar, b: Scalar) -> InterpResult<'tcx, u8>

See documentation on the ptr_guaranteed_cmp intrinsic.

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn copy_fn_arg( &self, arg: &FnArg<'tcx, M::Provenance>, ) -> OpTy<'tcx, M::Provenance>

Make a copy of the given fn_arg. Any InPlace are degenerated to copies, no protection of the original memory occurs.

pub fn copy_fn_args( &self, args: &[FnArg<'tcx, M::Provenance>], ) -> Vec<OpTy<'tcx, M::Provenance>>

Make a copy of the given fn_args. Any InPlace are degenerated to copies, no protection of the original memory occurs.

pub(super) fn fn_arg_field( &self, arg: &FnArg<'tcx, M::Provenance>, field: usize, ) -> InterpResult<'tcx, FnArg<'tcx, M::Provenance>>

Helper function for argument untupling.

fn unfold_transparent( &self, layout: TyAndLayout<'tcx>, may_unfold: impl Fn(AdtDef<'tcx>) -> bool, ) -> TyAndLayout<'tcx>

Find the wrapped inner type of a transparent wrapper. Must not be called on 1-ZST (as they don’t have a uniquely defined “wrapped field”).

We work with TyAndLayout here since that makes it much easier to iterate over all fields.

fn unfold_npo( &self, layout: TyAndLayout<'tcx>, ) -> InterpResult<'tcx, TyAndLayout<'tcx>>

Unwrap types that are guaranteed a null-pointer-optimization

fn layout_compat( &self, caller: TyAndLayout<'tcx>, callee: TyAndLayout<'tcx>, ) -> InterpResult<'tcx, bool>

Check if these two layouts look like they are fn-ABI-compatible. (We also compare the PassMode, so this doesn’t have to check everything. But it turns out that only checking the PassMode is insufficient.)

fn check_argument_compat( &self, caller_abi: &ArgAbi<'tcx, Ty<'tcx>>, callee_abi: &ArgAbi<'tcx, Ty<'tcx>>, ) -> InterpResult<'tcx, bool>

fn pass_argument<'x, 'y>( &mut self, caller_args: &mut impl Iterator<Item = (&'x FnArg<'tcx, M::Provenance>, &'y ArgAbi<'tcx, Ty<'tcx>>)>, callee_abi: &ArgAbi<'tcx, Ty<'tcx>>, callee_arg: &Place<'tcx>, callee_ty: Ty<'tcx>, already_live: bool, ) -> InterpResult<'tcx>

where 'tcx: 'x + 'y,

Initialize a single callee argument, checking the types for compatibility.

pub fn init_stack_frame( &mut self, instance: Instance<'tcx>, body: &'tcx Body<'tcx>, caller_fn_abi: &FnAbi<'tcx, Ty<'tcx>>, args: &[FnArg<'tcx, M::Provenance>], with_caller_location: bool, destination: &MPlaceTy<'tcx, M::Provenance>, stack_pop: StackPopCleanup, ) -> InterpResult<'tcx>

The main entry point for creating a new stack frame: performs ABI checks and initializes arguments.

pub(super) fn init_fn_call( &mut self, fn_val: FnVal<'tcx, M::ExtraFnVal>, (caller_abi, caller_fn_abi): (Abi, &FnAbi<'tcx, Ty<'tcx>>), args: &[FnArg<'tcx, M::Provenance>], with_caller_location: bool, destination: &MPlaceTy<'tcx, M::Provenance>, target: Option<BasicBlock>, unwind: UnwindAction, ) -> InterpResult<'tcx>

Initiate a call to this function – pushing the stack frame and initializing the arguments.

caller_fn_abi is used to determine if all the arguments are passed the proper way. However, we also need caller_abi to determine if we need to do untupling of arguments.

with_caller_location indicates whether the caller passed a caller location. Miri implements caller locations without argument passing, but to match FnAbi we need to know when those arguments are present.

pub(super) fn init_fn_tail_call( &mut self, fn_val: FnVal<'tcx, M::ExtraFnVal>, (caller_abi, caller_fn_abi): (Abi, &FnAbi<'tcx, Ty<'tcx>>), args: &[FnArg<'tcx, M::Provenance>], with_caller_location: bool, ) -> InterpResult<'tcx>

Initiate a tail call to this function – popping the current stack frame, pushing the new stack frame and initializing the arguments.

pub(super) fn init_drop_in_place_call( &mut self, place: &PlaceTy<'tcx, M::Provenance>, instance: Instance<'tcx>, target: BasicBlock, unwind: UnwindAction, ) -> InterpResult<'tcx>

pub(super) fn return_from_current_stack_frame( &mut self, unwinding: bool, ) -> InterpResult<'tcx>

Pops the current frame from the stack, copies the return value to the caller, deallocates the memory for allocated locals, and jumps to an appropriate place.

If unwinding is false, then we are performing a normal return from a function. In this case, we jump back into the frame of the caller, and continue execution as normal.

If unwinding is true, then we are in the middle of a panic, and need to unwind this frame. In this case, we jump to the cleanup block for the function, which is responsible for running Drop impls for any locals that have been initialized at this point. The cleanup block ends with a special Resume terminator, which will cause us to continue unwinding.

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn cast( &mut self, src: &OpTy<'tcx, M::Provenance>, cast_kind: CastKind, cast_ty: Ty<'tcx>, dest: &PlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

pub fn int_to_int_or_float( &self, src: &ImmTy<'tcx, M::Provenance>, cast_to: TyAndLayout<'tcx>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

Handles ‘IntToInt’ and ‘IntToFloat’ casts.

pub fn float_to_float_or_int( &self, src: &ImmTy<'tcx, M::Provenance>, cast_to: TyAndLayout<'tcx>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

Handles ‘FloatToFloat’ and ‘FloatToInt’ casts.

pub fn ptr_to_ptr( &self, src: &ImmTy<'tcx, M::Provenance>, cast_to: TyAndLayout<'tcx>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

Handles ‘FnPtrToPtr’ and ‘PtrToPtr’ casts.

pub fn pointer_expose_provenance_cast( &mut self, src: &ImmTy<'tcx, M::Provenance>, cast_to: TyAndLayout<'tcx>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

pub fn pointer_with_exposed_provenance_cast( &self, src: &ImmTy<'tcx, M::Provenance>, cast_to: TyAndLayout<'tcx>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

fn cast_from_int_like( &self, scalar: Scalar<M::Provenance>, src_layout: TyAndLayout<'tcx>, cast_ty: Ty<'tcx>, ) -> InterpResult<'tcx, Scalar<M::Provenance>>

Low-level cast helper function. This works directly on scalars and can take ‘int-like’ input type (basically everything with a scalar layout) to int/float/char types.

fn cast_from_float<F>(&self, f: F, dest_ty: Ty<'tcx>) -> Scalar<M::Provenance>

where F: Float + Into<Scalar<M::Provenance>> + FloatConvert<Half> + FloatConvert<Single> + FloatConvert<Double> + FloatConvert<Quad>,

Low-level cast helper function. Converts an apfloat f into int or float types.

fn unsize_into_ptr( &mut self, src: &OpTy<'tcx, M::Provenance>, dest: &PlaceTy<'tcx, M::Provenance>, source_ty: Ty<'tcx>, cast_ty: Ty<'tcx>, ) -> InterpResult<'tcx>

src is a pointer to a source_ty, and in dest we should store a pointer to th same data at type cast_ty.

pub fn unsize_into( &mut self, src: &OpTy<'tcx, M::Provenance>, cast_ty: TyAndLayout<'tcx>, dest: &PlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn write_discriminant( &mut self, variant_index: VariantIdx, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Writes the discriminant of the given variant.

If the variant is uninhabited, this is UB.

pub fn read_discriminant( &self, op: &impl Projectable<'tcx, M::Provenance>, ) -> InterpResult<'tcx, VariantIdx>

Read discriminant, return the runtime value as well as the variant index. Can also legally be called on non-enums (e.g. through the discriminant_value intrinsic)!

Will never return an uninhabited variant.

pub fn discriminant_for_variant( &self, ty: Ty<'tcx>, variant: VariantIdx, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

pub(crate) fn tag_for_variant( &self, ty: Ty<'tcx>, variant_index: VariantIdx, ) -> InterpResult<'tcx, Option<(ScalarInt, usize)>>

Computes how to write the tag of a given variant of enum ty:

• None means that nothing needs to be done as the variant is encoded implicitly
• Some((val, field_idx)) means that the given integer value needs to be stored at the given field index.

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn new( tcx: TyCtxt<'tcx>, root_span: Span, param_env: ParamEnv<'tcx>, machine: M, ) -> Self

pub fn cur_span(&self) -> Span

Returns the span of the currently executed statement/terminator. This is the span typically used for error reporting.

pub(crate) fn stack(&self) -> &[Frame<'tcx, M::Provenance, M::FrameExtra>]

pub(crate) fn stack_mut( &mut self, ) -> &mut Vec<Frame<'tcx, M::Provenance, M::FrameExtra>>

pub fn frame_idx(&self) -> usize

pub fn frame(&self) -> &Frame<'tcx, M::Provenance, M::FrameExtra>

pub fn frame_mut(&mut self) -> &mut Frame<'tcx, M::Provenance, M::FrameExtra>

pub fn body(&self) -> &'tcx Body<'tcx>

pub fn type_is_freeze(&self, ty: Ty<'tcx>) -> bool

pub fn load_mir( &self, instance: InstanceKind<'tcx>, promoted: Option<Promoted>, ) -> InterpResult<'tcx, &'tcx Body<'tcx>>

pub(super) fn instantiate_from_current_frame_and_normalize_erasing_regions<T: TypeFoldable<TyCtxt<'tcx>>>( &self, value: T, ) -> Result<T, ErrorHandled>

Call this on things you got out of the MIR (so it is as generic as the current stack frame), to bring it into the proper environment for this interpreter.

pub(super) fn instantiate_from_frame_and_normalize_erasing_regions<T: TypeFoldable<TyCtxt<'tcx>>>( &self, frame: &Frame<'tcx, M::Provenance, M::FrameExtra>, value: T, ) -> Result<T, ErrorHandled>

Call this on things you got out of the MIR (so it is as generic as the provided stack frame), to bring it into the proper environment for this interpreter.

pub(super) fn resolve( &self, def: DefId, args: GenericArgsRef<'tcx>, ) -> InterpResult<'tcx, Instance<'tcx>>

The args are assumed to already be in our interpreter “universe” (param_env).

pub(super) fn eq_in_param_env<T>(&self, a: T, b: T) -> bool

where T: PartialEq + TypeFoldable<TyCtxt<'tcx>> + ToTrace<'tcx>,

Check if the two things are equal in the current param_env, using an infctx to get proper equality checks.

pub(crate) fn find_closest_untracked_caller_location(&self) -> Span

Walks up the callstack from the intrinsic’s callsite, searching for the first callsite in a frame which is not #[track_caller]. This matches the caller_location intrinsic, and is primarily intended for the panic machinery.

pub(super) fn size_and_align_of( &self, metadata: &MemPlaceMeta<M::Provenance>, layout: &TyAndLayout<'tcx>, ) -> InterpResult<'tcx, Option<(Size, Align)>>

Returns the actual dynamic size and alignment of the place at the given type. Only the “meta” (metadata) part of the place matters. This can fail to provide an answer for extern types.

pub fn size_and_align_of_mplace( &self, mplace: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Option<(Size, Align)>>

pub fn go_to_block(&mut self, target: BasicBlock)

Jump to the given block.

pub fn return_to_block( &mut self, target: Option<BasicBlock>, ) -> InterpResult<'tcx>

Return to the given target basic block. Do not use for unwinding! Use unwind_to_block instead.

If target is None, that indicates the function cannot return, so we raise UB.

pub fn unwind_to_block(&mut self, target: UnwindAction) -> InterpResult<'tcx>

Unwind to the given target basic block. Do not use for returning! Use return_to_block instead.

If target is UnwindAction::Continue, that indicates the function does not need cleanup during unwinding, and we will just keep propagating that upwards.

If target is UnwindAction::Unreachable, that indicates the function does not allow unwinding, and doing so is UB.

pub fn ctfe_query<T>( &self, query: impl FnOnce(TyCtxtAt<'tcx>) -> Result<T, ErrorHandled>, ) -> Result<T, ErrorHandled>

Call a query that can return ErrorHandled. Should be used for statics and other globals. (mir::Const/ty::Const have eval methods that can be used directly instead.)

pub fn eval_global( &self, instance: Instance<'tcx>, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

pub fn eval_mir_constant( &self, val: &Const<'tcx>, span: Span, layout: Option<TyAndLayout<'tcx>>, ) -> InterpResult<'tcx, OpTy<'tcx, M::Provenance>>

pub fn dump_place( &self, place: &PlaceTy<'tcx, M::Provenance>, ) -> PlacePrinter<'_, 'tcx, M>

pub fn generate_stacktrace(&self) -> Vec<FrameInfo<'tcx>>

pub fn adjust_nan<F1, F2>(&self, f: F2, inputs: &[F1]) -> F2

where F1: Float + FloatConvert<F2>, F2: Float,

impl<'tcx, M: CompileTimeMachine<'tcx, !>> InterpCx<'tcx, M>

pub fn intern_with_temp_alloc( &mut self, layout: TyAndLayout<'tcx>, f: impl FnOnce(&mut InterpCx<'tcx, M>, &PlaceTy<'tcx, M::Provenance>) -> InterpResult<'tcx, ()>, ) -> InterpResult<'tcx, AllocId>

A helper function that allocates memory for the layout given and gives you access to mutate it. Once your own mutation code is done, the backing Allocation is removed from the current Memory and interned as read-only into the global memory.

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn eval_intrinsic( &mut self, instance: Instance<'tcx>, args: &[OpTy<'tcx, M::Provenance>], dest: &MPlaceTy<'tcx, M::Provenance>, ret: Option<BasicBlock>, ) -> InterpResult<'tcx, bool>

Returns true if emulation happened. Here we implement the intrinsics that are common to all Miri instances; individual machines can add their own intrinsic handling.

pub(super) fn eval_nondiverging_intrinsic( &mut self, intrinsic: &NonDivergingIntrinsic<'tcx>, ) -> InterpResult<'tcx>

pub fn numeric_intrinsic( &self, name: Symbol, val: Scalar<M::Provenance>, layout: TyAndLayout<'tcx>, ret_layout: TyAndLayout<'tcx>, ) -> InterpResult<'tcx, Scalar<M::Provenance>>

pub fn exact_div( &mut self, a: &ImmTy<'tcx, M::Provenance>, b: &ImmTy<'tcx, M::Provenance>, dest: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

pub fn saturating_arith( &self, mir_op: BinOp, l: &ImmTy<'tcx, M::Provenance>, r: &ImmTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Scalar<M::Provenance>>

pub fn ptr_offset_inbounds( &self, ptr: Pointer<Option<M::Provenance>>, offset_bytes: i64, ) -> InterpResult<'tcx, Pointer<Option<M::Provenance>>>

Offsets a pointer by some multiple of its type, returning an error if the pointer leaves its allocation.

pub(crate) fn copy_intrinsic( &mut self, src: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, dst: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, count: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, nonoverlapping: bool, ) -> InterpResult<'tcx>

Copy count*size_of::<T>() many bytes from *src to *dst.

fn typed_swap_intrinsic( &mut self, left: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, right: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, ) -> InterpResult<'tcx>

Does a typed swap of *left and *right.

pub fn write_bytes_intrinsic( &mut self, dst: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, byte: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, count: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, name: &'static str, ) -> InterpResult<'tcx>

pub(crate) fn compare_bytes_intrinsic( &mut self, left: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, right: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, byte_count: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, ) -> InterpResult<'tcx, Scalar<M::Provenance>>

pub(crate) fn raw_eq_intrinsic( &mut self, lhs: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, rhs: &OpTy<'tcx, <M as Machine<'tcx>>::Provenance>, ) -> InterpResult<'tcx, Scalar<M::Provenance>>

fn float_min_intrinsic<F>( &mut self, args: &[OpTy<'tcx, M::Provenance>], dest: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ()>

where F: Float + FloatConvert<F> + Into<Scalar<M::Provenance>>,

fn float_max_intrinsic<F>( &mut self, args: &[OpTy<'tcx, M::Provenance>], dest: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ()>

where F: Float + FloatConvert<F> + Into<Scalar<M::Provenance>>,

fn float_copysign_intrinsic<F>( &mut self, args: &[OpTy<'tcx, M::Provenance>], dest: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ()>

where F: Float + FloatConvert<F> + Into<Scalar<M::Provenance>>,

fn float_abs_intrinsic<F>( &mut self, args: &[OpTy<'tcx, M::Provenance>], dest: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ()>

where F: Float + FloatConvert<F> + Into<Scalar<M::Provenance>>,

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn global_root_pointer( &self, ptr: Pointer<CtfeProvenance>, ) -> InterpResult<'tcx, Pointer<M::Provenance>>

Call this to turn untagged “global” pointers (obtained via tcx) into the machine pointer to the allocation. Must never be used for any other pointers, nor for TLS statics.

Using the resulting pointer represents a direct access to that memory (e.g. by directly using a static), as opposed to access through a pointer that was created by the program.

This function can fail only if ptr points to an extern static.

pub fn fn_ptr( &mut self, fn_val: FnVal<'tcx, M::ExtraFnVal>, ) -> Pointer<M::Provenance>

pub fn allocate_ptr( &mut self, size: Size, align: Align, kind: MemoryKind<M::MemoryKind>, ) -> InterpResult<'tcx, Pointer<M::Provenance>>

pub fn allocate_bytes_ptr( &mut self, bytes: &[u8], align: Align, kind: MemoryKind<M::MemoryKind>, mutability: Mutability, ) -> InterpResult<'tcx, Pointer<M::Provenance>>

pub fn insert_allocation( &mut self, alloc: Allocation<M::Provenance, (), M::Bytes>, kind: MemoryKind<M::MemoryKind>, ) -> InterpResult<'tcx, Pointer<M::Provenance>>

pub fn reallocate_ptr( &mut self, ptr: Pointer<Option<M::Provenance>>, old_size_and_align: Option<(Size, Align)>, new_size: Size, new_align: Align, kind: MemoryKind<M::MemoryKind>, ) -> InterpResult<'tcx, Pointer<M::Provenance>>

pub fn deallocate_ptr( &mut self, ptr: Pointer<Option<M::Provenance>>, old_size_and_align: Option<(Size, Align)>, kind: MemoryKind<M::MemoryKind>, ) -> InterpResult<'tcx>

fn get_ptr_access( &self, ptr: Pointer<Option<M::Provenance>>, size: Size, ) -> InterpResult<'tcx, Option<(AllocId, Size, M::ProvenanceExtra)>>

Internal helper function to determine the allocation and offset of a pointer (if any).

pub fn check_ptr_access( &self, ptr: Pointer<Option<M::Provenance>>, size: Size, msg: CheckInAllocMsg, ) -> InterpResult<'tcx>

Check if the given pointer points to live memory of the given size. The caller can control the error message for the out-of-bounds case.

pub fn check_ptr_access_signed( &self, ptr: Pointer<Option<M::Provenance>>, size: i64, msg: CheckInAllocMsg, ) -> InterpResult<'tcx>

Check whether the given pointer points to live memory for a signed amount of bytes. A negative amounts means that the given range of memory to the left of the pointer needs to be dereferenceable.

fn check_and_deref_ptr<T, R: Borrow<Self>>( this: R, ptr: Pointer<Option<M::Provenance>>, size: i64, msg: CheckInAllocMsg, alloc_size: impl FnOnce(R, AllocId, Size, M::ProvenanceExtra) -> InterpResult<'tcx, (Size, Align, T)>, ) -> InterpResult<'tcx, Option<T>>

Low-level helper function to check if a ptr is in-bounds and potentially return a reference to the allocation it points to. Supports both shared and mutable references, as the actual checking is offloaded to a helper closure. Supports signed sizes for checks “to the left” of a pointer.

alloc_size will only get called for non-zero-sized accesses.

Returns None if and only if the size is 0.

pub(super) fn check_misalign( &self, misaligned: Option<Misalignment>, msg: CheckAlignMsg, ) -> InterpResult<'tcx>

pub(super) fn is_ptr_misaligned( &self, ptr: Pointer<Option<M::Provenance>>, align: Align, ) -> Option<Misalignment>

pub fn check_ptr_align( &self, ptr: Pointer<Option<M::Provenance>>, align: Align, ) -> InterpResult<'tcx>

Checks a pointer for misalignment.

The error assumes this is checking the pointer used directly for an access.

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn remove_unreachable_allocs( &mut self, reachable_allocs: &FxHashSet<AllocId>, )

This function is used by Miri’s provenance GC to remove unreachable entries from the dead_alloc_map.

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

Allocation accessors

fn get_global_alloc( &self, id: AllocId, is_write: bool, ) -> InterpResult<'tcx, Cow<'tcx, Allocation<M::Provenance, M::AllocExtra, M::Bytes>>>

Helper function to obtain a global (tcx) allocation. This attempts to return a reference to an existing allocation if one can be found in tcx. That, however, is only possible if tcx and this machine use the same pointer provenance, so it is indirected through M::adjust_allocation.

fn get_alloc_raw( &self, id: AllocId, ) -> InterpResult<'tcx, &Allocation<M::Provenance, M::AllocExtra, M::Bytes>>

Gives raw access to the Allocation, without bounds or alignment checks. The caller is responsible for calling the access hooks!

You almost certainly want to use get_ptr_alloc/get_ptr_alloc_mut instead.

pub fn get_alloc_bytes_unchecked_raw( &self, id: AllocId, ) -> InterpResult<'tcx, *const u8>

Gives raw, immutable access to the Allocation address, without bounds or alignment checks. The caller is responsible for calling the access hooks!

pub fn get_ptr_alloc<'a>( &'a self, ptr: Pointer<Option<M::Provenance>>, size: Size, ) -> InterpResult<'tcx, Option<AllocRef<'a, 'tcx, M::Provenance, M::AllocExtra, M::Bytes>>>

Bounds-checked but not align-checked allocation access.

pub fn get_alloc_extra<'a>( &'a self, id: AllocId, ) -> InterpResult<'tcx, &'a M::AllocExtra>

Return the extra field of the given allocation.

pub fn get_alloc_mutability<'a>( &'a self, id: AllocId, ) -> InterpResult<'tcx, Mutability>

Return the mutability field of the given allocation.

fn get_alloc_raw_mut( &mut self, id: AllocId, ) -> InterpResult<'tcx, (&mut Allocation<M::Provenance, M::AllocExtra, M::Bytes>, &mut M)>

Gives raw mutable access to the Allocation, without bounds or alignment checks. The caller is responsible for calling the access hooks!

Also returns a ptr to self.extra so that the caller can use it in parallel with the allocation.

pub fn get_alloc_bytes_unchecked_raw_mut( &mut self, id: AllocId, ) -> InterpResult<'tcx, *mut u8>

Gives raw, mutable access to the Allocation address, without bounds or alignment checks. The caller is responsible for calling the access hooks!

pub fn get_ptr_alloc_mut<'a>( &'a mut self, ptr: Pointer<Option<M::Provenance>>, size: Size, ) -> InterpResult<'tcx, Option<AllocRefMut<'a, 'tcx, M::Provenance, M::AllocExtra, M::Bytes>>>

Bounds-checked but not align-checked allocation access.

pub fn get_alloc_extra_mut<'a>( &'a mut self, id: AllocId, ) -> InterpResult<'tcx, (&'a mut M::AllocExtra, &'a mut M)>

Return the extra field of the given allocation.

pub fn is_alloc_live(&self, id: AllocId) -> bool

Check whether an allocation is live. This is faster than calling InterpCx::get_alloc_info if all you need to check is whether the kind is AllocKind::Dead because it doesn’t have to look up the type and layout of statics.

pub fn get_alloc_info(&self, id: AllocId) -> (Size, Align, AllocKind)

Obtain the size and alignment of an allocation, even if that allocation has been deallocated.

fn get_live_alloc_size_and_align( &self, id: AllocId, msg: CheckInAllocMsg, ) -> InterpResult<'tcx, (Size, Align)>

Obtain the size and alignment of a live allocation.

fn get_fn_alloc(&self, id: AllocId) -> Option<FnVal<'tcx, M::ExtraFnVal>>

pub fn get_ptr_fn( &self, ptr: Pointer<Option<M::Provenance>>, ) -> InterpResult<'tcx, FnVal<'tcx, M::ExtraFnVal>>

pub fn get_ptr_vtable_ty( &self, ptr: Pointer<Option<M::Provenance>>, expected_trait: Option<&'tcx List<PolyExistentialPredicate<'tcx>>>, ) -> InterpResult<'tcx, Ty<'tcx>>

Get the dynamic type of the given vtable pointer. If expected_trait is Some, it must be a vtable for the given trait.

pub fn alloc_mark_immutable(&mut self, id: AllocId) -> InterpResult<'tcx>

pub fn dump_alloc<'a>(&'a self, id: AllocId) -> DumpAllocs<'a, 'tcx, M>

Create a lazy debug printer that prints the given allocation and all allocations it points to, recursively.

pub fn dump_allocs<'a>( &'a self, allocs: Vec<AllocId>, ) -> DumpAllocs<'a, 'tcx, M>

Create a lazy debug printer for a list of allocations and all allocations they point to, recursively.

pub fn print_alloc_bytes_for_diagnostics(&self, id: AllocId) -> String

Print the allocation’s bytes, without any nested allocations.

pub fn take_leaked_allocations( &mut self, static_roots: impl FnOnce(&Self) -> &[AllocId], ) -> Vec<(AllocId, MemoryKind<M::MemoryKind>, Allocation<M::Provenance, M::AllocExtra, M::Bytes>)>

Find leaked allocations, remove them from memory and return them. Allocations reachable from static_roots or a Global allocation are not considered leaked, as well as leaks whose kind’s may_leak() returns true.

This is highly destructive, no more execution can happen after this!

pub fn run_for_validation<R>(&mut self, f: impl FnOnce(&mut Self) -> R) -> R

Runs the closure in “validation” mode, which means the machine’s memory read hooks will be suppressed. Needless to say, this must only be set with great care! Cannot be nested.

We do this so Miri’s allocation access tracking does not show the validation reads as spurious accesses.

pub(super) fn validation_in_progress(&self) -> bool

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn read_bytes_ptr_strip_provenance( &self, ptr: Pointer<Option<M::Provenance>>, size: Size, ) -> InterpResult<'tcx, &[u8]>

Reads the given number of bytes from memory, and strips their provenance if possible. Returns them as a slice.

Performs appropriate bounds checks.

pub fn write_bytes_ptr( &mut self, ptr: Pointer<Option<M::Provenance>>, src: impl IntoIterator<Item = u8>, ) -> InterpResult<'tcx>

Writes the given stream of bytes into memory.

Performs appropriate bounds checks.

pub fn mem_copy( &mut self, src: Pointer<Option<M::Provenance>>, dest: Pointer<Option<M::Provenance>>, size: Size, nonoverlapping: bool, ) -> InterpResult<'tcx>

pub fn mem_copy_repeatedly( &mut self, src: Pointer<Option<M::Provenance>>, dest: Pointer<Option<M::Provenance>>, size: Size, num_copies: u64, nonoverlapping: bool, ) -> InterpResult<'tcx>

Performs num_copies many copies of size many bytes from src to dest + i*size (where i is the index of the copy).

Either nonoverlapping must be true or num_copies must be 1; doing repeated copies that may overlap is not supported.

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

Machine pointer introspection.

pub fn scalar_may_be_null( &self, scalar: Scalar<M::Provenance>, ) -> InterpResult<'tcx, bool>

Test if this value might be null. If the machine does not support ptr-to-int casts, this is conservative.

pub fn ptr_try_get_alloc_id( &self, ptr: Pointer<Option<M::Provenance>>, size: i64, ) -> Result<(AllocId, Size, M::ProvenanceExtra), u64>

Turning a “maybe pointer” into a proper pointer (and some information about where it points), or an absolute address.

size says how many bytes of memory are expected at that pointer. This is largely only used for error messages; however, the sign of size can be used to disambiguate situations where a wildcard pointer sits right in between two allocations. It is almost always okay to just set the size to 0; this will be treated like a positive size for handling wildcard pointers.

The result must be used immediately; it is not allowed to convert the returned data back into a Pointer and store that in machine state. (In fact that’s not even possible since M::ProvenanceExtra is generic and we don’t have an operation to turn it back into M::Provenance.)

pub fn ptr_get_alloc_id( &self, ptr: Pointer<Option<M::Provenance>>, size: i64, ) -> InterpResult<'tcx, (AllocId, Size, M::ProvenanceExtra)>

Turning a “maybe pointer” into a proper pointer (and some information about where it points).

size says how many bytes of memory are expected at that pointer. This is largely only used for error messages; however, the sign of size can be used to disambiguate situations where a wildcard pointer sits right in between two allocations. It is almost always okay to just set the size to 0; this will be treated like a positive size for handling wildcard pointers.

The result must be used immediately; it is not allowed to convert the returned data back into a Pointer and store that in machine state. (In fact that’s not even possible since M::ProvenanceExtra is generic and we don’t have an operation to turn it back into M::Provenance.)

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

fn read_immediate_from_mplace_raw( &self, mplace: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Option<ImmTy<'tcx, M::Provenance>>>

Try reading an immediate in memory; this is interesting particularly for ScalarPair. Returns None if the layout does not permit loading this as a value.

This is an internal function; call read_immediate instead.

pub fn read_immediate_raw( &self, src: &impl Projectable<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Either<MPlaceTy<'tcx, M::Provenance>, ImmTy<'tcx, M::Provenance>>>

Try returning an immediate for the operand. If the layout does not permit loading this as an immediate, return where in memory we can find the data. Note that for a given layout, this operation will either always return Left or Right! succeed! Whether it returns Left depends on whether the layout can be represented in an Immediate, not on which data is stored there currently.

This is an internal function that should not usually be used; call read_immediate instead. ConstProp needs it, though.

pub fn read_immediate( &self, op: &impl Projectable<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

Read an immediate from a place, asserting that that is possible with the given layout.

If this succeeds, the ImmTy is never Uninit.

pub fn read_scalar( &self, op: &impl Projectable<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Scalar<M::Provenance>>

Read a scalar from a place

pub fn read_pointer( &self, op: &impl Projectable<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Pointer<Option<M::Provenance>>>

Read a pointer from a place.

pub fn read_target_usize( &self, op: &impl Projectable<'tcx, M::Provenance>, ) -> InterpResult<'tcx, u64>

Read a pointer-sized unsigned integer from a place.

pub fn read_target_isize( &self, op: &impl Projectable<'tcx, M::Provenance>, ) -> InterpResult<'tcx, i64>

Read a pointer-sized signed integer from a place.

pub fn read_str( &self, mplace: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, &str>

Turn the wide MPlace into a string (must already be dereferenced!)

pub fn local_to_op( &self, local: Local, layout: Option<TyAndLayout<'tcx>>, ) -> InterpResult<'tcx, OpTy<'tcx, M::Provenance>>

Read from a local of the current frame. Will not access memory, instead an indirect Operand is returned.

This is public because it is used by priroda to get an OpTy from a local.

pub fn place_to_op( &self, place: &PlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, OpTy<'tcx, M::Provenance>>

Every place can be read from, so we can turn them into an operand. This will definitely return Indirect if the place is a Ptr, i.e., this will never actually read from memory.

pub fn eval_place_to_op( &self, mir_place: Place<'tcx>, layout: Option<TyAndLayout<'tcx>>, ) -> InterpResult<'tcx, OpTy<'tcx, M::Provenance>>

Evaluate a place with the goal of reading from it. This lets us sometimes avoid allocations.

pub fn eval_operand( &self, mir_op: &Operand<'tcx>, layout: Option<TyAndLayout<'tcx>>, ) -> InterpResult<'tcx, OpTy<'tcx, M::Provenance>>

Evaluate the operand, returning a place where you can then find the data. If you already know the layout, you can save two table lookups by passing it in here.

pub(crate) fn const_val_to_op( &self, val_val: ConstValue<'tcx>, ty: Ty<'tcx>, layout: Option<TyAndLayout<'tcx>>, ) -> InterpResult<'tcx, OpTy<'tcx, M::Provenance>>

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

fn three_way_compare<T: Ord>( &self, lhs: T, rhs: T, ) -> ImmTy<'tcx, M::Provenance>

fn binary_char_op( &self, bin_op: BinOp, l: char, r: char, ) -> ImmTy<'tcx, M::Provenance>

fn binary_bool_op( &self, bin_op: BinOp, l: bool, r: bool, ) -> ImmTy<'tcx, M::Provenance>

fn binary_float_op<F: Float + FloatConvert<F> + Into<Scalar<M::Provenance>>>( &self, bin_op: BinOp, layout: TyAndLayout<'tcx>, l: F, r: F, ) -> ImmTy<'tcx, M::Provenance>

fn binary_int_op( &self, bin_op: BinOp, left: &ImmTy<'tcx, M::Provenance>, right: &ImmTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

pub fn compute_size_in_bytes(&self, elem_size: Size, count: u64) -> Option<Size>

Computes the total size of this access, count * elem_size, checking for overflow beyond isize::MAX.

fn binary_ptr_op( &self, bin_op: BinOp, left: &ImmTy<'tcx, M::Provenance>, right: &ImmTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

pub fn binary_op( &self, bin_op: BinOp, left: &ImmTy<'tcx, M::Provenance>, right: &ImmTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

Returns the result of the specified operation.

Whether this produces a scalar or a pair depends on the specific bin_op.

pub fn unary_op( &self, un_op: UnOp, val: &ImmTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

Returns the result of the specified operation, whether it overflowed, and the result type.

pub fn nullary_op( &self, null_op: NullOp<'tcx>, arg_ty: Ty<'tcx>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

impl<'tcx, Prov, M> InterpCx<'tcx, M>

where Prov: Provenance, M: Machine<'tcx, Provenance = Prov>,

fn ptr_with_meta_to_mplace( &self, ptr: Pointer<Option<M::Provenance>>, meta: MemPlaceMeta<M::Provenance>, layout: TyAndLayout<'tcx>, unaligned: bool, ) -> MPlaceTy<'tcx, M::Provenance>

pub fn ptr_to_mplace( &self, ptr: Pointer<Option<M::Provenance>>, layout: TyAndLayout<'tcx>, ) -> MPlaceTy<'tcx, M::Provenance>

pub fn ptr_to_mplace_unaligned( &self, ptr: Pointer<Option<M::Provenance>>, layout: TyAndLayout<'tcx>, ) -> MPlaceTy<'tcx, M::Provenance>

pub fn ref_to_mplace( &self, val: &ImmTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

Take a value, which represents a (thin or wide) reference, and make it a place. Alignment is just based on the type. This is the inverse of mplace_to_ref().

Only call this if you are sure the place is “valid” (aligned and inbounds), or do not want to ever use the place for memory access! Generally prefer deref_pointer.

pub fn mplace_to_ref( &self, mplace: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, ImmTy<'tcx, M::Provenance>>

Turn a mplace into a (thin or wide) mutable raw pointer, pointing to the same space. align information is lost! This is the inverse of ref_to_mplace.

pub fn deref_pointer( &self, src: &impl Projectable<'tcx, M::Provenance>, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

Take an operand, representing a pointer, and dereference it to a place. Corresponds to the * operator in Rust.

pub(super) fn get_place_alloc( &self, mplace: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Option<AllocRef<'_, 'tcx, M::Provenance, M::AllocExtra, M::Bytes>>>

pub(super) fn get_place_alloc_mut( &mut self, mplace: &MPlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Option<AllocRefMut<'_, 'tcx, M::Provenance, M::AllocExtra, M::Bytes>>>

pub fn local_to_place( &self, local: Local, ) -> InterpResult<'tcx, PlaceTy<'tcx, M::Provenance>>

Turn a local in the current frame into a place.

pub fn eval_place( &self, mir_place: Place<'tcx>, ) -> InterpResult<'tcx, PlaceTy<'tcx, M::Provenance>>

Computes a place. You should only use this if you intend to write into this place; for reading, a more efficient alternative is eval_place_to_op.

fn as_mplace_or_mutable_local( &mut self, place: &PlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, Either<MPlaceTy<'tcx, M::Provenance>, (&mut Immediate<M::Provenance>, TyAndLayout<'tcx>, Local)>>

Given a place, returns either the underlying mplace or a reference to where the value of this place is stored.

pub fn write_immediate( &mut self, src: Immediate<M::Provenance>, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Write an immediate to a place

pub fn write_scalar( &mut self, val: impl Into<Scalar<M::Provenance>>, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Write a scalar to a place

pub fn write_pointer( &mut self, ptr: impl Into<Pointer<Option<M::Provenance>>>, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Write a pointer to a place

pub(super) fn write_immediate_no_validate( &mut self, src: Immediate<M::Provenance>, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Write an immediate to a place. If you use this you are responsible for validating that things got copied at the right type.

fn write_immediate_to_mplace_no_validate( &mut self, value: Immediate<M::Provenance>, layout: TyAndLayout<'tcx>, dest: MemPlace<M::Provenance>, ) -> InterpResult<'tcx>

Write an immediate to memory. If you use this you are responsible for validating that things got copied at the right layout.

pub fn write_uninit( &mut self, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

pub fn clear_provenance( &mut self, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Remove all provenance in the given place.

pub(super) fn copy_op_no_dest_validation( &mut self, src: &impl Projectable<'tcx, M::Provenance>, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Copies the data from an operand to a place. The layouts of the src and dest may disagree. Does not perform validation of the destination. The only known use case for this function is checking the return value of a static during stack frame popping.

pub fn copy_op_allow_transmute( &mut self, src: &impl Projectable<'tcx, M::Provenance>, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Copies the data from an operand to a place. The layouts of the src and dest may disagree.

pub fn copy_op( &mut self, src: &impl Projectable<'tcx, M::Provenance>, dest: &impl Writeable<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Copies the data from an operand to a place. src and dest must have the same layout and the copied value will be validated.

fn copy_op_inner( &mut self, src: &impl Projectable<'tcx, M::Provenance>, dest: &impl Writeable<'tcx, M::Provenance>, allow_transmute: bool, validate_dest: bool, ) -> InterpResult<'tcx>

Copies the data from an operand to a place. allow_transmute indicates whether the layouts may disagree.

fn copy_op_no_validate( &mut self, src: &impl Projectable<'tcx, M::Provenance>, dest: &impl Writeable<'tcx, M::Provenance>, allow_transmute: bool, ) -> InterpResult<'tcx>

Copies the data from an operand to a place. allow_transmute indicates whether the layouts may disagree. Also, if you use this you are responsible for validating that things get copied at the right type.

pub fn force_allocation( &mut self, place: &PlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

Ensures that a place is in memory, and returns where it is. If the place currently refers to a local that doesn’t yet have a matching allocation, create such an allocation. This is essentially force_to_memplace.

pub fn allocate_dyn( &mut self, layout: TyAndLayout<'tcx>, kind: MemoryKind<M::MemoryKind>, meta: MemPlaceMeta<M::Provenance>, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

pub fn allocate( &mut self, layout: TyAndLayout<'tcx>, kind: MemoryKind<M::MemoryKind>, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

pub fn allocate_str( &mut self, str: &str, kind: MemoryKind<M::MemoryKind>, mutbl: Mutability, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

Returns a wide MPlace of type str to a new 1-aligned allocation. Immutable strings are deduplicated and stored in global memory.

pub fn raw_const_to_mplace( &self, raw: ConstAlloc<'tcx>, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

impl<'tcx, Prov, M> InterpCx<'tcx, M>

where Prov: Provenance, M: Machine<'tcx, Provenance = Prov>,

pub fn project_field<P: Projectable<'tcx, M::Provenance>>( &self, base: &P, field: usize, ) -> InterpResult<'tcx, P>

Offset a pointer to project to a field of a struct/union. Unlike place_field, this is always possible without allocating, so it can take &self. Also return the field’s layout. This supports both struct and array fields, but not slices!

This also works for arrays, but then the usize index type is restricting. For indexing into arrays, use mplace_index.

pub fn project_downcast<P: Projectable<'tcx, M::Provenance>>( &self, base: &P, variant: VariantIdx, ) -> InterpResult<'tcx, P>

Downcasting to an enum variant.

pub fn project_index<P: Projectable<'tcx, M::Provenance>>( &self, base: &P, index: u64, ) -> InterpResult<'tcx, P>

Compute the offset and field layout for accessing the given index.

pub fn project_to_simd<P: Projectable<'tcx, M::Provenance>>( &self, base: &P, ) -> InterpResult<'tcx, (P, u64)>

Converts a repr(simd) value into an array of the right size, such that project_index accesses the SIMD elements. Also returns the number of elements.

fn project_constant_index<P: Projectable<'tcx, M::Provenance>>( &self, base: &P, offset: u64, min_length: u64, from_end: bool, ) -> InterpResult<'tcx, P>

pub fn project_array_fields<'a, P: Projectable<'tcx, M::Provenance>>( &self, base: &'a P, ) -> InterpResult<'tcx, ArrayIterator<'a, 'tcx, M::Provenance, P>>

Iterates over all fields of an array. Much more efficient than doing the same by repeatedly calling project_index.

fn project_subslice<P: Projectable<'tcx, M::Provenance>>( &self, base: &P, from: u64, to: u64, from_end: bool, ) -> InterpResult<'tcx, P>

Subslicing

pub fn project<P>( &self, base: &P, proj_elem: PlaceElem<'tcx>, ) -> InterpResult<'tcx, P>

where P: Projectable<'tcx, M::Provenance> + From<MPlaceTy<'tcx, M::Provenance>> + Debug,

Applying a general projection

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub(crate) fn push_stack_frame_raw( &mut self, instance: Instance<'tcx>, body: &'tcx Body<'tcx>, return_place: &MPlaceTy<'tcx, M::Provenance>, return_to_block: StackPopCleanup, ) -> InterpResult<'tcx>

Very low-level helper that pushes a stack frame without initializing the arguments or local variables.

The high-level version of this is init_stack_frame.

pub(super) fn pop_stack_frame_raw( &mut self, unwinding: bool, ) -> InterpResult<'tcx, StackPopInfo<'tcx, M::Provenance>>

Low-level helper that pops a stack frame from the stack and returns some information about it.

This also deallocates locals, if necessary.

M::before_stack_pop should be called before calling this function. M::after_stack_pop is called by this function automatically.

The high-level version of this is return_from_current_stack_frame.

fn cleanup_current_frame_locals(&mut self) -> InterpResult<'tcx, bool>

A private helper for pop_stack_frame_raw. Returns true if cleanup has been done, false otherwise.

pub(crate) fn storage_live_for_always_live_locals( &mut self, ) -> InterpResult<'tcx>

In the current stack frame, mark all locals as live that are not arguments and don’t have Storage* annotations (this includes the return place).

pub fn storage_live_dyn( &mut self, local: Local, meta: MemPlaceMeta<M::Provenance>, ) -> InterpResult<'tcx>

pub fn storage_live(&mut self, local: Local) -> InterpResult<'tcx>

Mark a storage as live, killing the previous content.

pub fn storage_dead(&mut self, local: Local) -> InterpResult<'tcx>

fn deallocate_local( &mut self, local: LocalValue<M::Provenance>, ) -> InterpResult<'tcx>

pub(super) fn layout_of_local( &self, frame: &Frame<'tcx, M::Provenance, M::FrameExtra>, local: Local, layout: Option<TyAndLayout<'tcx>>, ) -> InterpResult<'tcx, TyAndLayout<'tcx>>

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn step(&mut self) -> InterpResult<'tcx, bool>

Returns true as long as there are more things to do.

This is used by priroda

This is marked #inline(always) to work around adversarial codegen when opt-level = 3

pub fn eval_statement(&mut self, stmt: &Statement<'tcx>) -> InterpResult<'tcx>

Runs the interpretation logic for the given mir::Statement at the current frame and statement counter.

This does NOT move the statement counter forward, the caller has to do that!

pub fn eval_rvalue_into_place( &mut self, rvalue: &Rvalue<'tcx>, place: Place<'tcx>, ) -> InterpResult<'tcx>

Evaluate an assignment statement.

There is no separate eval_rvalue function. Instead, the code for handling each rvalue type writes its results directly into the memory specified by the place.

fn write_aggregate( &mut self, kind: &AggregateKind<'tcx>, operands: &IndexSlice<FieldIdx, Operand<'tcx>>, dest: &PlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Writes the aggregate to the destination.

fn write_repeat( &mut self, operand: &Operand<'tcx>, dest: &PlaceTy<'tcx, M::Provenance>, ) -> InterpResult<'tcx>

Repeats operand into the destination. dest must have array type, and that type determines how often operand is repeated.

fn eval_fn_call_argument( &self, op: &Operand<'tcx>, ) -> InterpResult<'tcx, FnArg<'tcx, M::Provenance>>

Evaluate the arguments of a function call

fn eval_callee_and_args( &self, terminator: &Terminator<'tcx>, func: &Operand<'tcx>, args: &[Spanned<Operand<'tcx>>], ) -> InterpResult<'tcx, EvaluatedCalleeAndArgs<'tcx, M>>

Shared part of Call and TailCall implementation — finding and evaluating all the necessary information about callee and arguments to make a call.

fn eval_terminator( &mut self, terminator: &Terminator<'tcx>, ) -> InterpResult<'tcx>

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

pub fn get_vtable_ptr( &self, ty: Ty<'tcx>, dyn_ty: &'tcx List<PolyExistentialPredicate<'tcx>>, ) -> InterpResult<'tcx, Pointer<Option<M::Provenance>>>

Creates a dynamic vtable for the given type and vtable origin. This is used only for objects.

The dyn_ty encodes the erased self type. Hence, if we are making an object Foo<dyn Trait<Assoc = A> + Send> from a value of type Foo<T>, then dyn_ty would be Trait<Assoc = A> + Send. If this list doesn’t have a principal trait ref, we only need the basic vtable prefix (drop, size, align).

pub fn get_vtable_size_and_align( &self, vtable: Pointer<Option<M::Provenance>>, expected_trait: Option<&'tcx List<PolyExistentialPredicate<'tcx>>>, ) -> InterpResult<'tcx, (Size, Align)>

pub(super) fn vtable_entries( &self, trait_: Option<PolyExistentialTraitRef<'tcx>>, dyn_ty: Ty<'tcx>, ) -> &'tcx [VtblEntry<'tcx>]

pub(super) fn check_vtable_for_type( &self, vtable_dyn_type: &'tcx List<PolyExistentialPredicate<'tcx>>, expected_dyn_type: &'tcx List<PolyExistentialPredicate<'tcx>>, ) -> InterpResult<'tcx>

Check that the given vtable trait is valid for a pointer/reference/place with the given expected trait type.

pub(super) fn unpack_dyn_trait( &self, mplace: &MPlaceTy<'tcx, M::Provenance>, expected_trait: &'tcx List<PolyExistentialPredicate<'tcx>>, ) -> InterpResult<'tcx, MPlaceTy<'tcx, M::Provenance>>

Turn a place with a dyn Trait type into a place with the actual dynamic type.

pub(super) fn unpack_dyn_star<P: Projectable<'tcx, M::Provenance>>( &self, val: &P, expected_trait: &'tcx List<PolyExistentialPredicate<'tcx>>, ) -> InterpResult<'tcx, P>

Turn a dyn* Trait type into an value with the actual dynamic type.

impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M>

fn validate_operand_internal( &mut self, val: &PlaceTy<'tcx, M::Provenance>, path: Vec<PathElem>, ref_tracking: Option<&mut RefTracking<MPlaceTy<'tcx, M::Provenance>, Vec<PathElem>>>, ctfe_mode: Option<CtfeValidationMode>, reset_provenance_and_padding: bool, ) -> InterpResult<'tcx>

pub(crate) fn const_validate_operand( &mut self, val: &PlaceTy<'tcx, M::Provenance>, path: Vec<PathElem>, ref_tracking: &mut RefTracking<MPlaceTy<'tcx, M::Provenance>, Vec<PathElem>>, ctfe_mode: CtfeValidationMode, ) -> InterpResult<'tcx>

This function checks the data at op to be const-valid. op is assumed to cover valid memory if it is an indirect operand. It will error if the bits at the destination do not match the ones described by the layout.

ref_tracking is used to record references that we encounter so that they can be checked recursively by an outside driving loop.

constant controls whether this must satisfy the rules for constants:

• no pointers to statics.
• no UnsafeCell or non-ZST &mut.

pub fn validate_operand( &mut self, val: &PlaceTy<'tcx, M::Provenance>, recursive: bool, reset_provenance_and_padding: bool, ) -> InterpResult<'tcx>

This function checks the data at op to be runtime-valid. op is assumed to cover valid memory if it is an indirect operand. It will error if the bits at the destination do not match the ones described by the layout.

Trait Implementations

impl<'tcx, M: Machine<'tcx>> FnAbiOfHelpers<'tcx> for InterpCx<'tcx, M>

type FnAbiOfResult = Result<&'tcx FnAbi<'tcx, Ty<'tcx>>, InterpErrorKind<'tcx>>

The &FnAbi-wrapping type (or &FnAbi itself), which will be returned from fn_abi_of_* (see also handle_fn_abi_err).

fn handle_fn_abi_err( &self, err: FnAbiError<'tcx>, _span: Span, _fn_abi_request: FnAbiRequest<'tcx>, ) -> InterpErrorKind<'tcx>

Helper used for fn_abi_of_*, to adapt tcx.fn_abi_of_*(...) into a Self::FnAbiOfResult (which does not need to be a Result<...>).

impl<'tcx, M: Machine<'tcx>> HasDataLayout for InterpCx<'tcx, M>

fn data_layout(&self) -> &TargetDataLayout

impl<'tcx, M> HasParamEnv<'tcx> for InterpCx<'tcx, M>

where M: Machine<'tcx>,

fn param_env(&self) -> ParamEnv<'tcx>

impl<'tcx, M> HasTyCtxt<'tcx> for InterpCx<'tcx, M>

where M: Machine<'tcx>,

fn tcx(&self) -> TyCtxt<'tcx>

impl<'tcx, M: Machine<'tcx>> LayoutOfHelpers<'tcx> for InterpCx<'tcx, M>

type LayoutOfResult = Result<TyAndLayout<'tcx, Ty<'tcx>>, InterpErrorKind<'tcx>>

The TyAndLayout-wrapping type (or TyAndLayout itself), which will be returned from layout_of (see also handle_layout_err).

fn layout_tcx_at_span(&self) -> Span

Span to use for tcx.at(span), from layout_of.

fn handle_layout_err( &self, err: LayoutError<'tcx>, _: Span, _: Ty<'tcx>, ) -> InterpErrorKind<'tcx>

Helper used for layout_of, to adapt tcx.layout_of(...) into a Self::LayoutOfResult (which does not need to be a Result<...>).

Layout

Note: Unable to compute type layout, possibly due to this type having generic parameters. Layout can only be computed for concrete, fully-instantiated types.