Support enum variants in offset_of!

Author: GKFX

compiler/rustc_abi/src/lib.rs

@@ -1107,6 +1107,39 @@ impl Scalar {
 }
 
 // NOTE: This struct is generic over the FieldIdx for rust-analyzer usage.
+rustc_index::newtype_index! {
+    /// The *source-order* index of a field in a variant.
+    ///
+    /// This is how most code after type checking refers to fields, rather than
+    /// using names (as names have hygiene complications and more complex lookup).
+    ///
+    /// Particularly for `repr(Rust)` types, this may not be the same as *layout* order.
+    /// (It is for `repr(C)` `struct`s, however.)
+    ///
+    /// For example, in the following types,
+    /// ```rust
+    /// # enum Never {}
+    /// # #[repr(u16)]
+    /// enum Demo1 {
+    ///    Variant0 { a: Never, b: i32 } = 100,
+    ///    Variant1 { c: u8, d: u64 } = 10,
+    /// }
+    /// struct Demo2 { e: u8, f: u16, g: u8 }
+    /// ```
+    /// `b` is `FieldIdx(1)` in `VariantIdx(0)`,
+    /// `d` is `FieldIdx(1)` in `VariantIdx(1)`, and
+    /// `f` is `FieldIdx(1)` in `VariantIdx(0)`.
+    #[derive(HashStable_Generic)]
+    pub struct FieldIdx {}
+}
+
+/// `offset_of` can traverse fields and enum variants and should keep track of which is which.
+#[derive(Copy, Clone, Debug, Eq, Hash, HashStable_Generic, PartialEq, Encodable, Decodable)]
+pub enum OffsetOfIdx {
+    Field(FieldIdx),
+    Variant(VariantIdx),
+}
+
 /// Describes how the fields of a type are located in memory.
 #[derive(PartialEq, Eq, Hash, Clone, Debug)]
 #[cfg_attr(feature = "nightly", derive(HashStable_Generic))]

compiler/rustc_codegen_cranelift/src/base.rs

@@ -766,7 +766,7 @@ fn codegen_stmt<'tcx>(
                         NullOp::SizeOf => layout.size.bytes(),
                         NullOp::AlignOf => layout.align.abi.bytes(),
                         NullOp::OffsetOf(fields) => {
-                            layout.offset_of_subfield(fx, fields.iter().map(|f| f.index())).bytes()
+                            layout.offset_of_subfield(fx, fields.iter()).bytes()
                         }
                     };
                     let val = CValue::by_val(

compiler/rustc_codegen_ssa/src/mir/rvalue.rs

@@ -680,7 +680,7 @@ impl<'a, 'tcx, Bx: BuilderMethods<'a, 'tcx>> FunctionCx<'a, 'tcx, Bx> {
                         layout.align.abi.bytes()
                     }
                     mir::NullOp::OffsetOf(fields) => {
-                        layout.offset_of_subfield(bx.cx(), fields.iter().map(|f| f.index())).bytes()
+                        layout.offset_of_subfield(bx.cx(), fields.iter()).bytes()
                     }
                 };
                 let val = bx.cx().const_usize(val);

compiler/rustc_const_eval/src/interpret/step.rs

@@ -275,7 +275,7 @@ impl<'mir, 'tcx: 'mir, M: Machine<'mir, 'tcx>> InterpCx<'mir, 'tcx, M> {
                     mir::NullOp::SizeOf => layout.size.bytes(),
                     mir::NullOp::AlignOf => layout.align.abi.bytes(),
                     mir::NullOp::OffsetOf(fields) => {
-                        layout.offset_of_subfield(self, fields.iter().map(|f| f.index())).bytes()
+                        layout.offset_of_subfield(self, fields.iter()).bytes()
                     }
                 };
                 self.write_scalar(Scalar::from_target_usize(val, self), &dest)?;

compiler/rustc_const_eval/src/transform/validate.rs

@@ -1056,33 +1056,44 @@ impl<'a, 'tcx> Visitor<'tcx> for TypeChecker<'a, 'tcx> {
                     }
                 }
             }
-            Rvalue::NullaryOp(NullOp::OffsetOf(fields), container) => {
+            Rvalue::NullaryOp(NullOp::OffsetOf(indices), container) => {
                 let fail_out_of_bounds = |this: &mut Self, location, field, ty| {
                     this.fail(location, format!("Out of bounds field {field:?} for {ty:?}"));
                 };
 
                 let mut current_ty = *container;
+                let mut indices = indices.into_iter();
 
-                for field in fields.iter() {
-                    match current_ty.kind() {
-                        ty::Tuple(fields) => {
+                use rustc_target::abi::OffsetOfIdx::*;
+
+                while let Some(index) = indices.next() {
+                    match (current_ty.kind(), index) {
+                        (ty::Tuple(fields), Field(field)) => {
                             let Some(&f_ty) = fields.get(field.as_usize()) else {
                                 fail_out_of_bounds(self, location, field, current_ty);
                                 return;
                             };
 
                             current_ty = self.tcx.normalize_erasing_regions(self.param_env, f_ty);
                         }
-                        ty::Adt(adt_def, args) => {
-                            if adt_def.is_enum() {
+                        (ty::Adt(adt_def, args), Field(field)) if !adt_def.is_enum() => {
+                            let Some(field) = adt_def.non_enum_variant().fields.get(field) else {
+                                fail_out_of_bounds(self, location, field, current_ty);
+                                return;
+                            };
+
+                            let f_ty = field.ty(self.tcx, args);
+                            current_ty = self.tcx.normalize_erasing_regions(self.param_env, f_ty);
+                        }
+                        (ty::Adt(adt_def, args), Variant(variant)) if adt_def.is_enum() => {
+                            let Some(Field(field)) = indices.next() else {
                                 self.fail(
                                     location,
-                                    format!("Cannot get field offset from enum {current_ty:?}"),
+                                    format!("enum variant must be followed by field index in offset_of; in {current_ty:?}"),
                                 );
                                 return;
-                            }
-
-                            let Some(field) = adt_def.non_enum_variant().fields.get(field) else {
+                            };
+                            let Some(field) = adt_def.variant(variant).fields.get(field) else {
                                 fail_out_of_bounds(self, location, field, current_ty);
                                 return;
                             };
@@ -1093,7 +1104,7 @@ impl<'a, 'tcx> Visitor<'tcx> for TypeChecker<'a, 'tcx> {
                         _ => {
                             self.fail(
                                 location,
-                                format!("Cannot get field offset from non-adt type {current_ty:?}"),
+                                format!("Cannot get offset {index:?} from type {current_ty:?}"),
                             );
                             return;
                         }

compiler/rustc_error_codes/src/error_codes.rs

@@ -514,6 +514,7 @@ E0791: include_str!("./error_codes/E0791.md"),
 E0792: include_str!("./error_codes/E0792.md"),
 E0793: include_str!("./error_codes/E0793.md"),
 E0794: include_str!("./error_codes/E0794.md"),
+E0795: include_str!("./error_codes/E0795.md"),
 }
 
 // Undocumented removed error codes. Note that many removed error codes are kept in the list above

compiler/rustc_error_codes/src/error_codes/E0795.md

@@ -0,0 +1,28 @@
+Invalid argument for the `offset_of!` macro.
+
+Erroneous code example:
+
+```compile_fail,E0795
+#![feature(offset_of)]
+
+let x = std::mem::offset_of!(Option<u8>, Some);
+```
+
+The `offset_of!` macro gives the offset of a field within a type. It can
+navigate through enum variants, but the final component of its second argument
+must be a field and not a variant.
+
+The offset of the contained `u8` in the `Option<u8>` can be found by specifying
+the field name `0`:
+
+```
+#![feature(offset_of)]
+
+let x: usize = std::mem::offset_of!(Option<u8>, Some.0);
+```
+
+The discriminant of an enumeration may be read with `core::mem::discriminant`,
+but this is not always a value physically present within the enum.
+
+Further information about enum layout may be found at
+https://rust-lang.github.io/unsafe-code-guidelines/layout/enums.html.

compiler/rustc_hir_typeck/src/expr.rs

@@ -3103,16 +3103,68 @@ impl<'a, 'tcx> FnCtxt<'a, 'tcx> {
         fields: &[Ident],
         expr: &'tcx hir::Expr<'tcx>,
     ) -> Ty<'tcx> {
+        use rustc_target::abi::OffsetOfIdx::*;
+
         let container = self.to_ty(container).normalized;
 
         let mut field_indices = Vec::with_capacity(fields.len());
         let mut current_container = container;
+        let mut fields = fields.into_iter();
 
-        for &field in fields {
+        while let Some(&field) = fields.next() {
             let container = self.structurally_resolve_type(expr.span, current_container);
 
             match container.kind() {
-                ty::Adt(container_def, args) if !container_def.is_enum() => {
+                ty::Adt(container_def, args) if container_def.is_enum() => {
+                    let block = self.tcx.hir().local_def_id_to_hir_id(self.body_id);
+                    let (ident, _def_scope) =
+                        self.tcx.adjust_ident_and_get_scope(field, container_def.did(), block);
+
+                    if let Some((index, variant)) = container_def.variants()
+                        .iter_enumerated()
+                        .find(|(_, v)| v.ident(self.tcx).normalize_to_macros_2_0() == ident)
+                    {
+                        let Some(&subfield) = fields.next() else {
+                            let mut err = type_error_struct!(
+                                self.tcx().sess,
+                                ident.span,
+                                container,
+                                E0795,
+                                "`{ident}` is an enum variant; expected field at end of `offset_of`",
+                                );
+                            err.span_label(field.span, "enum variant");
+                            err.emit();
+                            break;
+                        };
+                        let (subident, sub_def_scope) =
+                            self.tcx.adjust_ident_and_get_scope(subfield, variant.def_id, block);
+
+                        if let Some((subindex, field)) = variant.fields
+                            .iter_enumerated()
+                            .find(|(_, f)| f.ident(self.tcx).normalize_to_macros_2_0() == subident)
+                        {
+                            let field_ty = self.field_ty(expr.span, field, args);
+
+                            // FIXME: DSTs with static alignment should be allowed
+                            self.require_type_is_sized(field_ty, expr.span, traits::MiscObligation);
+
+                            if field.vis.is_accessible_from(sub_def_scope, self.tcx) {
+                                self.tcx.check_stability(field.did, Some(expr.hir_id), expr.span, None);
+                            } else {
+                                self.private_field_err(ident, container_def.did()).emit();
+                            }
+
+                            // Save the index of all fields regardless of their visibility in case
+                            // of error recovery.
+                            field_indices.push(Variant(index));
+                            field_indices.push(Field(subindex));
+                            current_container = field_ty;
+
+                            continue;
+                        }
+                    }
+                }
+                ty::Adt(container_def, args) => {
                     let block = self.tcx.hir().local_def_id_to_hir_id(self.body_id);
                     let (ident, def_scope) =
                         self.tcx.adjust_ident_and_get_scope(field, container_def.did(), block);
@@ -3135,7 +3187,7 @@ impl<'a, 'tcx> FnCtxt<'a, 'tcx> {
 
                         // Save the index of all fields regardless of their visibility in case
                         // of error recovery.
-                        field_indices.push(index);
+                        field_indices.push(Field(index));
                         current_container = field_ty;
 
                         continue;
@@ -3149,7 +3201,7 @@ impl<'a, 'tcx> FnCtxt<'a, 'tcx> {
                             self.require_type_is_sized(ty, expr.span, traits::MiscObligation);
                         }
                         if let Some(&field_ty) = tys.get(index) {
-                            field_indices.push(index.into());
+                            field_indices.push(Field(index.into()));
                             current_container = field_ty;
 
                             continue;

compiler/rustc_middle/src/mir/syntax.rs

@@ -17,7 +17,7 @@ use rustc_hir::def_id::DefId;
 use rustc_hir::{self as hir};
 use rustc_hir::{self, CoroutineKind};
 use rustc_index::IndexVec;
-use rustc_target::abi::{FieldIdx, VariantIdx};
+use rustc_target::abi::{FieldIdx, OffsetOfIdx, VariantIdx};
 
 use rustc_ast::Mutability;
 use rustc_span::def_id::LocalDefId;
@@ -1354,7 +1354,7 @@ pub enum NullOp<'tcx> {
     /// Returns the minimum alignment of a type
     AlignOf,
     /// Returns the offset of a field
-    OffsetOf(&'tcx List<FieldIdx>),
+    OffsetOf(&'tcx List<OffsetOfIdx>),
 }
 
 #[derive(Copy, Clone, Debug, PartialEq, Eq, PartialOrd, Ord, Hash)]

compiler/rustc_middle/src/thir.rs

@@ -25,7 +25,8 @@ use rustc_middle::ty::{
 };
 use rustc_span::def_id::LocalDefId;
 use rustc_span::{sym, ErrorGuaranteed, Span, Symbol, DUMMY_SP};
-use rustc_target::abi::{FieldIdx, VariantIdx};
+use rustc_span::{sym, Span, Symbol, DUMMY_SP};
+use rustc_target::abi::{FieldIdx, OffsetOfIdx, VariantIdx};
 use rustc_target::asm::InlineAsmRegOrRegClass;
 use std::fmt;
 use std::ops::Index;
@@ -490,7 +491,7 @@ pub enum ExprKind<'tcx> {
     /// Field offset (`offset_of!`)
     OffsetOf {
         container: Ty<'tcx>,
-        fields: &'tcx List<FieldIdx>,
+        fields: &'tcx List<OffsetOfIdx>,
     },
     /// An expression taking a reference to a thread local.
     ThreadLocalRef(DefId),

compiler/rustc_middle/src/ty/codec.rs

@@ -19,7 +19,7 @@ use rustc_data_structures::fx::FxHashMap;
 use rustc_middle::ty::TyCtxt;
 use rustc_serialize::{Decodable, Encodable};
 use rustc_span::Span;
-use rustc_target::abi::FieldIdx;
+use rustc_target::abi::{FieldIdx, OffsetOfIdx};
 pub use rustc_type_ir::{TyDecoder, TyEncoder};
 use std::hash::Hash;
 use std::intrinsics;
@@ -414,6 +414,15 @@ impl<'tcx, D: TyDecoder<I = TyCtxt<'tcx>>> RefDecodable<'tcx, D> for ty::List<Fi
     }
 }
 
+impl<'tcx, D: TyDecoder<I = TyCtxt<'tcx>>> RefDecodable<'tcx, D> for ty::List<OffsetOfIdx> {
+    fn decode(decoder: &mut D) -> &'tcx Self {
+        let len = decoder.read_usize();
+        decoder
+            .interner()
+            .mk_offset_of_from_iter((0..len).map::<OffsetOfIdx, _>(|_| Decodable::decode(decoder)))
+    }
+}
+
 impl_decodable_via_ref! {
     &'tcx ty::TypeckResults<'tcx>,
     &'tcx ty::List<Ty<'tcx>>,
@@ -426,6 +435,7 @@ impl_decodable_via_ref! {
     &'tcx ty::List<ty::BoundVariableKind>,
     &'tcx ty::List<ty::Clause<'tcx>>,
     &'tcx ty::List<FieldIdx>,
+    &'tcx ty::List<OffsetOfIdx>,
 }
 
 #[macro_export]

compiler/rustc_middle/src/ty/context.rs

@@ -63,7 +63,7 @@ use rustc_session::{Limit, MetadataKind, Session};
 use rustc_span::def_id::{DefPathHash, StableCrateId};
 use rustc_span::symbol::{kw, sym, Ident, Symbol};
 use rustc_span::{Span, DUMMY_SP};
-use rustc_target::abi::{FieldIdx, Layout, LayoutS, TargetDataLayout, VariantIdx};
+use rustc_target::abi::{FieldIdx, Layout, LayoutS, OffsetOfIdx, TargetDataLayout, VariantIdx};
 use rustc_target::spec::abi;
 use rustc_type_ir::TyKind::*;
 use rustc_type_ir::WithCachedTypeInfo;
@@ -163,6 +163,7 @@ pub struct CtxtInterners<'tcx> {
     predefined_opaques_in_body: InternedSet<'tcx, PredefinedOpaquesData<'tcx>>,
     fields: InternedSet<'tcx, List<FieldIdx>>,
     local_def_ids: InternedSet<'tcx, List<LocalDefId>>,
+    offset_of: InternedSet<'tcx, List<OffsetOfIdx>>,
 }
 
 impl<'tcx> CtxtInterners<'tcx> {
@@ -189,6 +190,7 @@ impl<'tcx> CtxtInterners<'tcx> {
             predefined_opaques_in_body: Default::default(),
             fields: Default::default(),
             local_def_ids: Default::default(),
+            offset_of: Default::default(),
         }
     }
 
@@ -1587,6 +1589,7 @@ slice_interners!(
     bound_variable_kinds: pub mk_bound_variable_kinds(ty::BoundVariableKind),
     fields: pub mk_fields(FieldIdx),
     local_def_ids: intern_local_def_ids(LocalDefId),
+    offset_of: pub mk_offset_of(OffsetOfIdx),
 );
 
 impl<'tcx> TyCtxt<'tcx> {
@@ -1914,6 +1917,14 @@ impl<'tcx> TyCtxt<'tcx> {
         T::collect_and_apply(iter, |xs| self.mk_fields(xs))
     }
 
+    pub fn mk_offset_of_from_iter<I, T>(self, iter: I) -> T::Output
+    where
+        I: Iterator<Item = T>,
+        T: CollectAndApply<OffsetOfIdx, &'tcx List<OffsetOfIdx>>,
+    {
+        T::collect_and_apply(iter, |xs| self.mk_offset_of(xs))
+    }
+
     pub fn mk_args_trait(
         self,
         self_ty: Ty<'tcx>,