Auto merge of #13490 - HKalbasi:layout, r=jonas-schievink

Compute data layout of types

cc #4091

Things that aren't working:
* Closures
* Generators (so no support for `Future` I think)
* Opaque types
* Type alias and associated types which may need normalization

Things that show wrong result:
* ~Enums with explicit discriminant~
* SIMD types
* ~`NonZero*` and similar standard library items which control layout with special attributes~

At the user level, I didn't put much work, since I wasn't confident about what is the best way to present this information. Currently it shows size and align for ADTs, and size, align, offset for struct fields, in the hover, similar to clangd. I used it some days and I feel I liked it, but we may consider it too noisy and move it to an assist or command.

Author: bors

Cargo.lock

@@ -33,6 +33,8 @@ base-db = { path = "../base-db", version = "0.0.0" }
 syntax = { path = "../syntax", version = "0.0.0" }
 profile = { path = "../profile", version = "0.0.0" }
 hir-expand = { path = "../hir-expand", version = "0.0.0" }
+rustc_abi = { version = "0.0.20221125", package = "hkalbasi-rustc-ap-rustc_abi", default-features = false }
+rustc_index = { version = "0.0.20221125", package = "hkalbasi-rustc-ap-rustc_index", default-features = false }
 mbe = { path = "../mbe", version = "0.0.0" }
 cfg = { path = "../cfg", version = "0.0.0" }
 tt = { path = "../tt", version = "0.0.0" }

crates/hir-def/Cargo.toml

@@ -1,6 +1,6 @@
 //! Defines hir-level representation of structs, enums and unions
 
-use std::{num::NonZeroU32, sync::Arc};
+use std::sync::Arc;
 
 use base_db::CrateId;
 use either::Either;
@@ -9,6 +9,7 @@ use hir_expand::{
     HirFileId, InFile,
 };
 use la_arena::{Arena, ArenaMap};
+use rustc_abi::{Integer, IntegerType};
 use syntax::ast::{self, HasName, HasVisibility};
 use tt::{Delimiter, DelimiterKind, Leaf, Subtree, TokenTree};
 
@@ -18,6 +19,7 @@ use crate::{
     db::DefDatabase,
     intern::Interned,
     item_tree::{AttrOwner, Field, FieldAstId, Fields, ItemTree, ModItem, RawVisibilityId},
+    layout::{Align, ReprFlags, ReprOptions},
     nameres::diagnostics::DefDiagnostic,
     src::HasChildSource,
     src::HasSource,
@@ -34,7 +36,7 @@ use cfg::CfgOptions;
 pub struct StructData {
     pub name: Name,
     pub variant_data: Arc<VariantData>,
-    pub repr: Option<ReprData>,
+    pub repr: Option<ReprOptions>,
     pub visibility: RawVisibility,
     pub rustc_has_incoherent_inherent_impls: bool,
 }
@@ -43,7 +45,7 @@ pub struct StructData {
 pub struct EnumData {
     pub name: Name,
     pub variants: Arena<EnumVariantData>,
-    pub repr: Option<ReprData>,
+    pub repr: Option<ReprOptions>,
     pub visibility: RawVisibility,
     pub rustc_has_incoherent_inherent_impls: bool,
 }
@@ -69,80 +71,91 @@ pub struct FieldData {
     pub visibility: RawVisibility,
 }
 
-#[derive(Copy, Debug, Clone, PartialEq, Eq)]
-pub enum ReprKind {
-    C,
-    BuiltinInt { builtin: Either<BuiltinInt, BuiltinUint>, is_c: bool },
-    Transparent,
-    Default,
-}
-
-#[derive(Copy, Debug, Clone, PartialEq, Eq)]
-pub struct ReprData {
-    pub kind: ReprKind,
-    pub packed: bool,
-    pub align: Option<NonZeroU32>,
-}
-
 fn repr_from_value(
     db: &dyn DefDatabase,
     krate: CrateId,
     item_tree: &ItemTree,
     of: AttrOwner,
-) -> Option<ReprData> {
+) -> Option<ReprOptions> {
     item_tree.attrs(db, krate, of).by_key("repr").tt_values().find_map(parse_repr_tt)
 }
 
-fn parse_repr_tt(tt: &Subtree) -> Option<ReprData> {
+fn parse_repr_tt(tt: &Subtree) -> Option<ReprOptions> {
     match tt.delimiter {
         Some(Delimiter { kind: DelimiterKind::Parenthesis, .. }) => {}
         _ => return None,
     }
 
-    let mut data = ReprData { kind: ReprKind::Default, packed: false, align: None };
+    let mut flags = ReprFlags::empty();
+    let mut int = None;
+    let mut max_align: Option<Align> = None;
+    let mut min_pack: Option<Align> = None;
 
     let mut tts = tt.token_trees.iter().peekable();
     while let Some(tt) = tts.next() {
         if let TokenTree::Leaf(Leaf::Ident(ident)) = tt {
-            match &*ident.text {
+            flags.insert(match &*ident.text {
                 "packed" => {
-                    data.packed = true;
-                    if let Some(TokenTree::Subtree(_)) = tts.peek() {
+                    let pack = if let Some(TokenTree::Subtree(tt)) = tts.peek() {
                         tts.next();
-                    }
+                        if let Some(TokenTree::Leaf(Leaf::Literal(lit))) = tt.token_trees.first() {
+                            lit.text.parse().unwrap_or_default()
+                        } else {
+                            0
+                        }
+                    } else {
+                        0
+                    };
+                    let pack = Align::from_bytes(pack).unwrap();
+                    min_pack =
+                        Some(if let Some(min_pack) = min_pack { min_pack.min(pack) } else { pack });
+                    ReprFlags::empty()
                 }
                 "align" => {
                     if let Some(TokenTree::Subtree(tt)) = tts.peek() {
                         tts.next();
                         if let Some(TokenTree::Leaf(Leaf::Literal(lit))) = tt.token_trees.first() {
                             if let Ok(align) = lit.text.parse() {
-                                data.align = Some(align);
+                                let align = Align::from_bytes(align).ok();
+                                max_align = max_align.max(align);
                             }
                         }
                     }
+                    ReprFlags::empty()
                 }
-                "C" => {
-                    if let ReprKind::BuiltinInt { is_c, .. } = &mut data.kind {
-                        *is_c = true;
-                    } else {
-                        data.kind = ReprKind::C;
-                    }
-                }
-                "transparent" => data.kind = ReprKind::Transparent,
+                "C" => ReprFlags::IS_C,
+                "transparent" => ReprFlags::IS_TRANSPARENT,
                 repr => {
-                    let is_c = matches!(data.kind, ReprKind::C);
                     if let Some(builtin) = BuiltinInt::from_suffix(repr)
                         .map(Either::Left)
                         .or_else(|| BuiltinUint::from_suffix(repr).map(Either::Right))
                     {
-                        data.kind = ReprKind::BuiltinInt { builtin, is_c };
+                        int = Some(match builtin {
+                            Either::Left(bi) => match bi {
+                                BuiltinInt::Isize => IntegerType::Pointer(true),
+                                BuiltinInt::I8 => IntegerType::Fixed(Integer::I8, true),
+                                BuiltinInt::I16 => IntegerType::Fixed(Integer::I16, true),
+                                BuiltinInt::I32 => IntegerType::Fixed(Integer::I32, true),
+                                BuiltinInt::I64 => IntegerType::Fixed(Integer::I64, true),
+                                BuiltinInt::I128 => IntegerType::Fixed(Integer::I128, true),
+                            },
+                            Either::Right(bu) => match bu {
+                                BuiltinUint::Usize => IntegerType::Pointer(false),
+                                BuiltinUint::U8 => IntegerType::Fixed(Integer::I8, false),
+                                BuiltinUint::U16 => IntegerType::Fixed(Integer::I16, false),
+                                BuiltinUint::U32 => IntegerType::Fixed(Integer::I32, false),
+                                BuiltinUint::U64 => IntegerType::Fixed(Integer::I64, false),
+                                BuiltinUint::U128 => IntegerType::Fixed(Integer::I128, false),
+                            },
+                        });
                     }
+                    ReprFlags::empty()
                 }
-            }
+            })
         }
     }
 
-    Some(data)
+    Some(ReprOptions { int, align: max_align, pack: min_pack, flags, field_shuffle_seed: 0 })
 }
 
 impl StructData {
@@ -299,10 +312,10 @@ impl EnumData {
         Some(id)
     }
 
-    pub fn variant_body_type(&self) -> Either<BuiltinInt, BuiltinUint> {
+    pub fn variant_body_type(&self) -> IntegerType {
         match self.repr {
-            Some(ReprData { kind: ReprKind::BuiltinInt { builtin, .. }, .. }) => builtin,
-            _ => Either::Left(BuiltinInt::Isize),
+            Some(ReprOptions { int: Some(builtin), .. }) => builtin,
+            _ => IntegerType::Pointer(true),
         }
     }
 }

crates/hir-def/src/adt.rs

@@ -0,0 +1,96 @@
+//! Definitions needed for computing data layout of types.
+
+use std::cmp;
+
+use la_arena::{Idx, RawIdx};
+pub use rustc_abi::{
+    Abi, AbiAndPrefAlign, AddressSpace, Align, Endian, FieldsShape, Integer, IntegerType,
+    LayoutCalculator, Niche, Primitive, ReprFlags, ReprOptions, Scalar, Size, StructKind,
+    TargetDataLayout, WrappingRange,
+};
+
+use crate::LocalEnumVariantId;
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
+pub struct RustcEnumVariantIdx(pub LocalEnumVariantId);
+
+impl rustc_index::vec::Idx for RustcEnumVariantIdx {
+    fn new(idx: usize) -> Self {
+        RustcEnumVariantIdx(Idx::from_raw(RawIdx::from(idx as u32)))
+    }
+
+    fn index(self) -> usize {
+        u32::from(self.0.into_raw()) as usize
+    }
+}
+
+pub type Layout = rustc_abi::LayoutS<RustcEnumVariantIdx>;
+pub type TagEncoding = rustc_abi::TagEncoding<RustcEnumVariantIdx>;
+pub type Variants = rustc_abi::Variants<RustcEnumVariantIdx>;
+
+pub trait IntegerExt {
+    fn repr_discr(
+        dl: &TargetDataLayout,
+        repr: &ReprOptions,
+        min: i128,
+        max: i128,
+    ) -> Result<(Integer, bool), LayoutError>;
+}
+
+impl IntegerExt for Integer {
+    /// Finds the appropriate Integer type and signedness for the given
+    /// signed discriminant range and `#[repr]` attribute.
+    /// N.B.: `u128` values above `i128::MAX` will be treated as signed, but
+    /// that shouldn't affect anything, other than maybe debuginfo.
+    fn repr_discr(
+        dl: &TargetDataLayout,
+        repr: &ReprOptions,
+        min: i128,
+        max: i128,
+    ) -> Result<(Integer, bool), LayoutError> {
+        // Theoretically, negative values could be larger in unsigned representation
+        // than the unsigned representation of the signed minimum. However, if there
+        // are any negative values, the only valid unsigned representation is u128
+        // which can fit all i128 values, so the result remains unaffected.
+        let unsigned_fit = Integer::fit_unsigned(cmp::max(min as u128, max as u128));
+        let signed_fit = cmp::max(Integer::fit_signed(min), Integer::fit_signed(max));
+
+        if let Some(ity) = repr.int {
+            let discr = Integer::from_attr(dl, ity);
+            let fit = if ity.is_signed() { signed_fit } else { unsigned_fit };
+            if discr < fit {
+                return Err(LayoutError::UserError(
+                    "Integer::repr_discr: `#[repr]` hint too small for \
+                      discriminant range of enum "
+                        .to_string(),
+                ));
+            }
+            return Ok((discr, ity.is_signed()));
+        }
+
+        let at_least = if repr.c() {
+            // This is usually I32, however it can be different on some platforms,
+            // notably hexagon and arm-none/thumb-none
+            dl.c_enum_min_size
+        } else {
+            // repr(Rust) enums try to be as small as possible
+            Integer::I8
+        };
+
+        // If there are no negative values, we can use the unsigned fit.
+        Ok(if min >= 0 {
+            (cmp::max(unsigned_fit, at_least), false)
+        } else {
+            (cmp::max(signed_fit, at_least), true)
+        })
+    }
+}
+
+#[derive(Debug, PartialEq, Eq, Clone)]
+pub enum LayoutError {
+    UserError(String),
+    SizeOverflow,
+    HasPlaceholder,
+    NotImplemented,
+    Unknown,
+}

crates/hir-def/src/layout.rs

@@ -34,6 +34,7 @@ pub mod adt;
 pub mod data;
 pub mod generics;
 pub mod lang_item;
+pub mod layout;
 
 pub mod expr;
 pub mod body;

crates/hir-def/src/lib.rs

@@ -419,6 +419,7 @@ pub mod known {
         shr,
         sub_assign,
         sub,
+        unsafe_cell,
         va_list
     );
 

crates/hir-expand/src/name.rs

@@ -25,6 +25,7 @@ chalk-derive = "0.88.0"
 la-arena = { version = "0.3.0", path = "../../lib/la-arena" }
 once_cell = "1.15.0"
 typed-arena = "2.0.1"
+rustc_index = { version = "0.0.20221125", package = "hkalbasi-rustc-ap-rustc_index", default-features = false }
 
 stdx = { path = "../stdx", version = "0.0.0" }
 hir-def = { path = "../hir-def", version = "0.0.0" }

crates/hir-ty/Cargo.toml

@@ -5,8 +5,11 @@ use std::sync::Arc;
 
 use base_db::{impl_intern_key, salsa, CrateId, Upcast};
 use hir_def::{
-    db::DefDatabase, expr::ExprId, BlockId, ConstId, ConstParamId, DefWithBodyId, EnumVariantId,
-    FunctionId, GenericDefId, ImplId, LifetimeParamId, LocalFieldId, TypeOrConstParamId, VariantId,
+    db::DefDatabase,
+    expr::ExprId,
+    layout::{Layout, LayoutError, TargetDataLayout},
+    AdtId, BlockId, ConstId, ConstParamId, DefWithBodyId, EnumVariantId, FunctionId, GenericDefId,
+    ImplId, LifetimeParamId, LocalFieldId, TypeOrConstParamId, VariantId,
 };
 use la_arena::ArenaMap;
 use smallvec::SmallVec;
@@ -16,7 +19,7 @@ use crate::{
     consteval::{ComputedExpr, ConstEvalError},
     method_resolution::{InherentImpls, TraitImpls, TyFingerprint},
     Binders, CallableDefId, FnDefId, GenericArg, ImplTraitId, InferenceResult, Interner, PolyFnSig,
-    QuantifiedWhereClause, ReturnTypeImplTraits, TraitRef, Ty, TyDefId, ValueTyDefId,
+    QuantifiedWhereClause, ReturnTypeImplTraits, Substitution, TraitRef, Ty, TyDefId, ValueTyDefId,
 };
 use hir_expand::name::Name;
 
@@ -57,6 +60,13 @@ pub trait HirDatabase: DefDatabase + Upcast<dyn DefDatabase> {
     #[salsa::invoke(crate::lower::field_types_query)]
     fn field_types(&self, var: VariantId) -> Arc<ArenaMap<LocalFieldId, Binders<Ty>>>;
 
+    #[salsa::invoke(crate::layout::layout_of_adt_query)]
+    #[salsa::cycle(crate::layout::layout_of_adt_recover)]
+    fn layout_of_adt(&self, def: AdtId, subst: Substitution) -> Result<Layout, LayoutError>;
+
+    #[salsa::invoke(crate::layout::current_target_data_layout_query)]
+    fn current_target_data_layout(&self) -> Arc<TargetDataLayout>;
+
     #[salsa::invoke(crate::lower::callable_item_sig)]
     fn callable_item_signature(&self, def: CallableDefId) -> PolyFnSig;