Struct rustc_const_eval::interpret::intern::InternVisitor [−][src]
struct InternVisitor<'rt, 'mir, 'tcx, M: CompileTimeMachine<'mir, 'tcx, MemoryKind>> {
ecx: &'rt mut InterpCx<'mir, 'tcx, M>,
ref_tracking: &'rt mut RefTracking<(MPlaceTy<'tcx>, InternMode)>,
leftover_allocations: &'rt mut FxHashSet<AllocId>,
mode: InternMode,
inside_unsafe_cell: bool,
}
Fields
ecx: &'rt mut InterpCx<'mir, 'tcx, M>
The ectx from which we intern.
ref_tracking: &'rt mut RefTracking<(MPlaceTy<'tcx>, InternMode)>
Previously encountered safe references.
leftover_allocations: &'rt mut FxHashSet<AllocId>
A list of all encountered allocations. After type-based interning, we traverse this list to also intern allocations that are only referenced by a raw pointer or inside a union.
mode: InternMode
The root kind of the value that we’re looking at. This field is never mutated for a particular allocation. It is primarily used to make as many allocations as possible read-only so LLVM can place them in const memory.
inside_unsafe_cell: bool
This field stores whether we are currently inside an UnsafeCell
. This can affect
the intern mode of references we encounter.
Implementations
impl<'rt, 'mir, 'tcx, M: CompileTimeMachine<'mir, 'tcx, MemoryKind>> InternVisitor<'rt, 'mir, 'tcx, M>
impl<'rt, 'mir, 'tcx, M: CompileTimeMachine<'mir, 'tcx, MemoryKind>> InternVisitor<'rt, 'mir, 'tcx, M>
fn intern_shallow(
&mut self,
alloc_id: AllocId,
mode: InternMode,
ty: Option<Ty<'tcx>>
) -> Option<IsStaticOrFn>
Trait Implementations
impl<'rt, 'mir, 'tcx: 'mir, M: CompileTimeMachine<'mir, 'tcx, MemoryKind>> ValueVisitor<'mir, 'tcx, M> for InternVisitor<'rt, 'mir, 'tcx, M>
impl<'rt, 'mir, 'tcx: 'mir, M: CompileTimeMachine<'mir, 'tcx, MemoryKind>> ValueVisitor<'mir, 'tcx, M> for InternVisitor<'rt, 'mir, 'tcx, M>
fn visit_aggregate(
&mut self,
mplace: &MPlaceTy<'tcx>,
fields: impl Iterator<Item = InterpResult<'tcx, Self::V>>
) -> InterpResult<'tcx>
fn visit_aggregate(
&mut self,
mplace: &MPlaceTy<'tcx>,
fields: impl Iterator<Item = InterpResult<'tcx, Self::V>>
) -> InterpResult<'tcx>
Visits this value as an aggregate, you are getting an iterator yielding
all the fields (still in an InterpResult
, you have to do error handling yourself).
Recurses into the fields. Read more
Visits the given value, dispatching as appropriate to more specialized visitors.
fn read_discriminant(
&mut self,
op: &OpTy<'tcx, M::PointerTag>
) -> InterpResult<'tcx, VariantIdx>
fn read_discriminant(
&mut self,
op: &OpTy<'tcx, M::PointerTag>
) -> InterpResult<'tcx, VariantIdx>
read_discriminant
can be hooked for better error messages.
Visits the given value as a union. No automatic recursion can happen here.
fn visit_field(
&mut self,
_old_val: &Self::V,
_field: usize,
new_val: &Self::V
) -> InterpResult<'tcx>
fn visit_field(
&mut self,
_old_val: &Self::V,
_field: usize,
new_val: &Self::V
) -> InterpResult<'tcx>
Called each time we recurse down to a field of a “product-like” aggregate (structs, tuples, arrays and the like, but not enums), passing in old (outer) and new (inner) value. This gives the visitor the chance to track the stack of nested fields that we are descending through. Read more
fn visit_variant(
&mut self,
_old_val: &Self::V,
_variant: VariantIdx,
new_val: &Self::V
) -> InterpResult<'tcx>
fn visit_variant(
&mut self,
_old_val: &Self::V,
_variant: VariantIdx,
new_val: &Self::V
) -> InterpResult<'tcx>
Called when recursing into an enum variant. This gives the visitor the chance to track the stack of nested fields that we are descending through. Read more
fn walk_aggregate(
&mut self,
v: &Self::V,
fields: impl Iterator<Item = InterpResult<'tcx, Self::V>>
) -> InterpResult<'tcx>
Auto Trait Implementations
impl<'rt, 'mir, 'tcx, M> !RefUnwindSafe for InternVisitor<'rt, 'mir, 'tcx, M>
impl<'rt, 'mir, 'tcx, M> !Send for InternVisitor<'rt, 'mir, 'tcx, M>
impl<'rt, 'mir, 'tcx, M> !Sync for InternVisitor<'rt, 'mir, 'tcx, M>
impl<'rt, 'mir, 'tcx, M> Unpin for InternVisitor<'rt, 'mir, 'tcx, M> where
'mir: 'rt,
'tcx: 'rt,
impl<'rt, 'mir, 'tcx, M> !UnwindSafe for InternVisitor<'rt, 'mir, 'tcx, M>
Blanket Implementations
Layout
Note: Most layout information is completely unstable and may even differ between compilations. The only exception is types with certain repr(...)
attributes. Please see the Rust Reference’s “Type Layout” chapter for details on type layout guarantees.
Size: 32 bytes