// Copyright 2013 The Rust Project Developers. See the COPYRIGHT // file at the top-level directory of this distribution and at // http://rust-lang.org/COPYRIGHT. // // Licensed under the Apache License, Version 2.0 or the MIT license // , at your // option. This file may not be copied, modified, or distributed // except according to those terms. //! # Representation of Algebraic Data Types //! //! This module determines how to represent enums, structs, and tuples //! based on their monomorphized types; it is responsible both for //! choosing a representation and translating basic operations on //! values of those types. (Note: exporting the representations for //! debuggers is handled in debuginfo.rs, not here.) //! //! Note that the interface treats everything as a general case of an //! enum, so structs/tuples/etc. have one pseudo-variant with //! discriminant 0; i.e., as if they were a univariant enum. //! //! Having everything in one place will enable improvements to data //! structure representation; possibilities include: //! //! - User-specified alignment (e.g., cacheline-aligning parts of //! concurrently accessed data structures); LLVM can't represent this //! directly, so we'd have to insert padding fields in any structure //! that might contain one and adjust GEP indices accordingly. See //! issue #4578. //! //! - Store nested enums' discriminants in the same word. Rather, if //! some variants start with enums, and those enums representations //! have unused alignment padding between discriminant and body, the //! outer enum's discriminant can be stored there and those variants //! can start at offset 0. Kind of fancy, and might need work to //! make copies of the inner enum type cooperate, but it could help //! with `Option` or `Result` wrapped around another enum. //! //! - Tagged pointers would be neat, but given that any type can be //! used unboxed and any field can have pointers (including mutable) //! taken to it, implementing them for Rust seems difficult. pub use self::Repr::*; use super::Disr; use std; use std::rc::Rc; use llvm::{ValueRef, True, IntEQ, IntNE}; use rustc::ty::subst::Substs; use rustc::ty::{self, AdtKind, Ty, TyCtxt}; use syntax::ast; use syntax::attr; use syntax::attr::IntType; use abi::FAT_PTR_ADDR; use build::*; use common::*; use debuginfo::DebugLoc; use glue; use machine; use monomorphize; use type_::Type; use type_of; use value::Value; #[derive(Copy, Clone, PartialEq)] pub enum BranchKind { Switch, Single } type Hint = attr::ReprAttr; /// Representations. #[derive(Eq, PartialEq, Debug)] pub enum Repr<'tcx> { /// C-like enums; basically an int. CEnum(IntType, Disr, Disr), // discriminant range (signedness based on the IntType) /// Single-case variants, and structs/tuples/records. Univariant(Struct<'tcx>), /// Untagged unions. UntaggedUnion(Union<'tcx>), /// General-case enums: for each case there is a struct, and they /// all start with a field for the discriminant. General(IntType, Vec>), /// Two cases distinguished by a nullable pointer: the case with discriminant /// `nndiscr` must have single field which is known to be nonnull due to its type. /// The other case is known to be zero sized. Hence we represent the enum /// as simply a nullable pointer: if not null it indicates the `nndiscr` variant, /// otherwise it indicates the other case. RawNullablePointer { nndiscr: Disr, nnty: Ty<'tcx>, nullfields: Vec> }, /// Two cases distinguished by a nullable pointer: the case with discriminant /// `nndiscr` is represented by the struct `nonnull`, where the `discrfield`th /// field is known to be nonnull due to its type; if that field is null, then /// it represents the other case, which is inhabited by at most one value /// (and all other fields are undefined/unused). /// /// For example, `std::option::Option` instantiated at a safe pointer type /// is represented such that `None` is a null pointer and `Some` is the /// identity function. StructWrappedNullablePointer { nonnull: Struct<'tcx>, nndiscr: Disr, discrfield: DiscrField, nullfields: Vec>, } } /// For structs, and struct-like parts of anything fancier. #[derive(Eq, PartialEq, Debug)] pub struct Struct<'tcx> { // If the struct is DST, then the size and alignment do not take into // account the unsized fields of the struct. pub size: u64, pub align: u32, pub sized: bool, pub packed: bool, pub fields: Vec>, } /// For untagged unions. #[derive(Eq, PartialEq, Debug)] pub struct Union<'tcx> { pub min_size: u64, pub align: u32, pub packed: bool, pub fields: Vec>, } #[derive(Copy, Clone)] pub struct MaybeSizedValue { pub value: ValueRef, pub meta: ValueRef, } impl MaybeSizedValue { pub fn sized(value: ValueRef) -> MaybeSizedValue { MaybeSizedValue { value: value, meta: std::ptr::null_mut() } } pub fn unsized_(value: ValueRef, meta: ValueRef) -> MaybeSizedValue { MaybeSizedValue { value: value, meta: meta } } pub fn has_meta(&self) -> bool { !self.meta.is_null() } } /// Decides how to represent a given type. pub fn represent_type<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, t: Ty<'tcx>) -> Rc> { debug!("Representing: {}", t); if let Some(repr) = cx.adt_reprs().borrow().get(&t) { return repr.clone(); } let repr = Rc::new(represent_type_uncached(cx, t)); debug!("Represented as: {:?}", repr); cx.adt_reprs().borrow_mut().insert(t, repr.clone()); repr } fn represent_type_uncached<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, t: Ty<'tcx>) -> Repr<'tcx> { match t.sty { ty::TyTuple(ref elems) => { Univariant(mk_struct(cx, &elems[..], false, t)) } ty::TyClosure(_, ref substs) => { Univariant(mk_struct(cx, &substs.upvar_tys, false, t)) } ty::TyAdt(def, substs) => match def.adt_kind() { AdtKind::Struct => { let ftys = def.struct_variant().fields.iter().map(|field| { monomorphize::field_ty(cx.tcx(), substs, field) }).collect::>(); let packed = cx.tcx().lookup_packed(def.did); Univariant(mk_struct(cx, &ftys[..], packed, t)) } AdtKind::Union => { let ftys = def.struct_variant().fields.iter().map(|field| { monomorphize::field_ty(cx.tcx(), substs, field) }).collect::>(); let packed = cx.tcx().lookup_packed(def.did); UntaggedUnion(mk_union(cx, &ftys[..], packed, t)) } AdtKind::Enum => { let cases = get_cases(cx.tcx(), def, substs); let hint = *cx.tcx().lookup_repr_hints(def.did).get(0) .unwrap_or(&attr::ReprAny); if cases.is_empty() { // Uninhabitable; represent as unit // (Typechecking will reject discriminant-sizing attrs.) assert_eq!(hint, attr::ReprAny); return Univariant(mk_struct(cx, &[], false, t)); } if cases.iter().all(|c| c.tys.is_empty()) { // All bodies empty -> intlike let discrs: Vec<_> = cases.iter().map(|c| Disr::from(c.discr)).collect(); let bounds = IntBounds { ulo: discrs.iter().min().unwrap().0, uhi: discrs.iter().max().unwrap().0, slo: discrs.iter().map(|n| n.0 as i64).min().unwrap(), shi: discrs.iter().map(|n| n.0 as i64).max().unwrap() }; return mk_cenum(cx, hint, &bounds); } // Since there's at least one // non-empty body, explicit discriminants should have // been rejected by a checker before this point. if !cases.iter().enumerate().all(|(i,c)| c.discr == Disr::from(i)) { bug!("non-C-like enum {} with specified discriminants", cx.tcx().item_path_str(def.did)); } if cases.len() == 1 && hint == attr::ReprAny { // Equivalent to a struct or tuple. return Univariant(mk_struct(cx, &cases[0].tys, false, t)); } if cases.len() == 2 && hint == attr::ReprAny { // Nullable pointer optimization let mut discr = 0; while discr < 2 { if cases[1 - discr].is_zerolen(cx, t) { let st = mk_struct(cx, &cases[discr].tys, false, t); match cases[discr].find_ptr(cx) { Some(ref df) if df.len() == 1 && st.fields.len() == 1 => { return RawNullablePointer { nndiscr: Disr::from(discr), nnty: st.fields[0], nullfields: cases[1 - discr].tys.clone() }; } Some(mut discrfield) => { discrfield.push(0); discrfield.reverse(); return StructWrappedNullablePointer { nndiscr: Disr::from(discr), nonnull: st, discrfield: discrfield, nullfields: cases[1 - discr].tys.clone() }; } None => {} } } discr += 1; } } // The general case. assert!((cases.len() - 1) as i64 >= 0); let bounds = IntBounds { ulo: 0, uhi: (cases.len() - 1) as u64, slo: 0, shi: (cases.len() - 1) as i64 }; let min_ity = range_to_inttype(cx, hint, &bounds); // Create the set of structs that represent each variant // Use the minimum integer type we figured out above let fields : Vec<_> = cases.iter().map(|c| { let mut ftys = vec!(ty_of_inttype(cx.tcx(), min_ity)); ftys.extend_from_slice(&c.tys); mk_struct(cx, &ftys, false, t) }).collect(); // Check to see if we should use a different type for the // discriminant. If the overall alignment of the type is // the same as the first field in each variant, we can safely use // an alignment-sized type. // We increase the size of the discriminant to avoid LLVM copying // padding when it doesn't need to. This normally causes unaligned // load/stores and excessive memcpy/memset operations. By using a // bigger integer size, LLVM can be sure about it's contents and // won't be so conservative. // This check is needed to avoid increasing the size of types when // the alignment of the first field is smaller than the overall // alignment of the type. let (_, align) = union_size_and_align(&fields); let mut use_align = true; for st in &fields { // Get the first non-zero-sized field let field = st.fields.iter().skip(1).filter(|ty| { let t = type_of::sizing_type_of(cx, **ty); machine::llsize_of_real(cx, t) != 0 || // This case is only relevant for zero-sized types with large alignment machine::llalign_of_min(cx, t) != 1 }).next(); if let Some(field) = field { let field_align = type_of::align_of(cx, *field); if field_align != align { use_align = false; break; } } } // If the alignment is smaller than the chosen discriminant size, don't use the // alignment as the final size. let min_ty = ll_inttype(&cx, min_ity); let min_size = machine::llsize_of_real(cx, min_ty); if (align as u64) < min_size { use_align = false; } let ity = if use_align { // Use the overall alignment match align { 1 => attr::UnsignedInt(ast::UintTy::U8), 2 => attr::UnsignedInt(ast::UintTy::U16), 4 => attr::UnsignedInt(ast::UintTy::U32), 8 if machine::llalign_of_min(cx, Type::i64(cx)) == 8 => attr::UnsignedInt(ast::UintTy::U64), _ => min_ity // use min_ity as a fallback } } else { min_ity }; let fields : Vec<_> = cases.iter().map(|c| { let mut ftys = vec!(ty_of_inttype(cx.tcx(), ity)); ftys.extend_from_slice(&c.tys); mk_struct(cx, &ftys[..], false, t) }).collect(); ensure_enum_fits_in_address_space(cx, &fields[..], t); General(ity, fields) } }, _ => bug!("adt::represent_type called on non-ADT type: {}", t) } } // this should probably all be in ty struct Case<'tcx> { discr: Disr, tys: Vec> } /// This represents the (GEP) indices to follow to get to the discriminant field pub type DiscrField = Vec; fn find_discr_field_candidate<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>, ty: Ty<'tcx>, mut path: DiscrField) -> Option { match ty.sty { // Fat &T/&mut T/Box i.e. T is [T], str, or Trait ty::TyRef(_, ty::TypeAndMut { ty, .. }) | ty::TyBox(ty) if !type_is_sized(tcx, ty) => { path.push(FAT_PTR_ADDR); Some(path) }, // Regular thin pointer: &T/&mut T/Box ty::TyRef(..) | ty::TyBox(..) => Some(path), // Function pointer: `fn() -> i32` ty::TyFnPtr(_) => Some(path), // Is this the NonZero lang item wrapping a pointer or integer type? ty::TyAdt(def, substs) if Some(def.did) == tcx.lang_items.non_zero() => { let nonzero_fields = &def.struct_variant().fields; assert_eq!(nonzero_fields.len(), 1); let field_ty = monomorphize::field_ty(tcx, substs, &nonzero_fields[0]); match field_ty.sty { ty::TyRawPtr(ty::TypeAndMut { ty, .. }) if !type_is_sized(tcx, ty) => { path.extend_from_slice(&[0, FAT_PTR_ADDR]); Some(path) }, ty::TyRawPtr(..) | ty::TyInt(..) | ty::TyUint(..) => { path.push(0); Some(path) }, _ => None } }, // Perhaps one of the fields of this struct is non-zero // let's recurse and find out ty::TyAdt(def, substs) if def.is_struct() => { for (j, field) in def.struct_variant().fields.iter().enumerate() { let field_ty = monomorphize::field_ty(tcx, substs, field); if let Some(mut fpath) = find_discr_field_candidate(tcx, field_ty, path.clone()) { fpath.push(j); return Some(fpath); } } None }, // Perhaps one of the upvars of this struct is non-zero // Let's recurse and find out! ty::TyClosure(_, ref substs) => { for (j, &ty) in substs.upvar_tys.iter().enumerate() { if let Some(mut fpath) = find_discr_field_candidate(tcx, ty, path.clone()) { fpath.push(j); return Some(fpath); } } None }, // Can we use one of the fields in this tuple? ty::TyTuple(ref tys) => { for (j, &ty) in tys.iter().enumerate() { if let Some(mut fpath) = find_discr_field_candidate(tcx, ty, path.clone()) { fpath.push(j); return Some(fpath); } } None }, // Is this a fixed-size array of something non-zero // with at least one element? ty::TyArray(ety, d) if d > 0 => { if let Some(mut vpath) = find_discr_field_candidate(tcx, ety, path) { vpath.push(0); Some(vpath) } else { None } }, // Anything else is not a pointer _ => None } } impl<'tcx> Case<'tcx> { fn is_zerolen<'a>(&self, cx: &CrateContext<'a, 'tcx>, scapegoat: Ty<'tcx>) -> bool { mk_struct(cx, &self.tys, false, scapegoat).size == 0 } fn find_ptr<'a>(&self, cx: &CrateContext<'a, 'tcx>) -> Option { for (i, &ty) in self.tys.iter().enumerate() { if let Some(mut path) = find_discr_field_candidate(cx.tcx(), ty, vec![]) { path.push(i); return Some(path); } } None } } fn get_cases<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>, adt: ty::AdtDef<'tcx>, substs: &Substs<'tcx>) -> Vec> { adt.variants.iter().map(|vi| { let field_tys = vi.fields.iter().map(|field| { monomorphize::field_ty(tcx, substs, field) }).collect(); Case { discr: Disr::from(vi.disr_val), tys: field_tys } }).collect() } fn mk_struct<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, tys: &[Ty<'tcx>], packed: bool, scapegoat: Ty<'tcx>) -> Struct<'tcx> { let sized = tys.iter().all(|&ty| type_is_sized(cx.tcx(), ty)); let lltys : Vec = if sized { tys.iter().map(|&ty| type_of::sizing_type_of(cx, ty)).collect() } else { tys.iter().filter(|&ty| type_is_sized(cx.tcx(), *ty)) .map(|&ty| type_of::sizing_type_of(cx, ty)).collect() }; ensure_struct_fits_in_address_space(cx, &lltys[..], packed, scapegoat); let llty_rec = Type::struct_(cx, &lltys[..], packed); Struct { size: machine::llsize_of_alloc(cx, llty_rec), align: machine::llalign_of_min(cx, llty_rec), sized: sized, packed: packed, fields: tys.to_vec(), } } fn mk_union<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, tys: &[Ty<'tcx>], packed: bool, _scapegoat: Ty<'tcx>) -> Union<'tcx> { let mut min_size = 0; let mut align = 0; for llty in tys.iter().map(|&ty| type_of::sizing_type_of(cx, ty)) { let field_size = machine::llsize_of_alloc(cx, llty); if min_size < field_size { min_size = field_size; } let field_align = machine::llalign_of_min(cx, llty); if align < field_align { align = field_align; } } Union { min_size: min_size, align: if packed { 1 } else { align }, packed: packed, fields: tys.to_vec(), } } #[derive(Debug)] struct IntBounds { slo: i64, shi: i64, ulo: u64, uhi: u64 } fn mk_cenum<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, hint: Hint, bounds: &IntBounds) -> Repr<'tcx> { let it = range_to_inttype(cx, hint, bounds); match it { attr::SignedInt(_) => CEnum(it, Disr(bounds.slo as u64), Disr(bounds.shi as u64)), attr::UnsignedInt(_) => CEnum(it, Disr(bounds.ulo), Disr(bounds.uhi)) } } fn range_to_inttype(cx: &CrateContext, hint: Hint, bounds: &IntBounds) -> IntType { debug!("range_to_inttype: {:?} {:?}", hint, bounds); // Lists of sizes to try. u64 is always allowed as a fallback. #[allow(non_upper_case_globals)] const choose_shortest: &'static [IntType] = &[ attr::UnsignedInt(ast::UintTy::U8), attr::SignedInt(ast::IntTy::I8), attr::UnsignedInt(ast::UintTy::U16), attr::SignedInt(ast::IntTy::I16), attr::UnsignedInt(ast::UintTy::U32), attr::SignedInt(ast::IntTy::I32)]; #[allow(non_upper_case_globals)] const at_least_32: &'static [IntType] = &[ attr::UnsignedInt(ast::UintTy::U32), attr::SignedInt(ast::IntTy::I32)]; let attempts; match hint { attr::ReprInt(span, ity) => { if !bounds_usable(cx, ity, bounds) { span_bug!(span, "representation hint insufficient for discriminant range") } return ity; } attr::ReprExtern => { attempts = match &cx.sess().target.target.arch[..] { // WARNING: the ARM EABI has two variants; the one corresponding to `at_least_32` // appears to be used on Linux and NetBSD, but some systems may use the variant // corresponding to `choose_shortest`. However, we don't run on those yet...? "arm" => at_least_32, _ => at_least_32, } } attr::ReprAny => { attempts = choose_shortest; }, attr::ReprPacked => { bug!("range_to_inttype: found ReprPacked on an enum"); } attr::ReprSimd => { bug!("range_to_inttype: found ReprSimd on an enum"); } } for &ity in attempts { if bounds_usable(cx, ity, bounds) { return ity; } } return attr::UnsignedInt(ast::UintTy::U64); } pub fn ll_inttype(cx: &CrateContext, ity: IntType) -> Type { match ity { attr::SignedInt(t) => Type::int_from_ty(cx, t), attr::UnsignedInt(t) => Type::uint_from_ty(cx, t) } } fn bounds_usable(cx: &CrateContext, ity: IntType, bounds: &IntBounds) -> bool { debug!("bounds_usable: {:?} {:?}", ity, bounds); match ity { attr::SignedInt(_) => { let lllo = C_integral(ll_inttype(cx, ity), bounds.slo as u64, true); let llhi = C_integral(ll_inttype(cx, ity), bounds.shi as u64, true); bounds.slo == const_to_int(lllo) as i64 && bounds.shi == const_to_int(llhi) as i64 } attr::UnsignedInt(_) => { let lllo = C_integral(ll_inttype(cx, ity), bounds.ulo, false); let llhi = C_integral(ll_inttype(cx, ity), bounds.uhi, false); bounds.ulo == const_to_uint(lllo) as u64 && bounds.uhi == const_to_uint(llhi) as u64 } } } pub fn ty_of_inttype<'a, 'tcx>(tcx: TyCtxt<'a, 'tcx, 'tcx>, ity: IntType) -> Ty<'tcx> { match ity { attr::SignedInt(t) => tcx.mk_mach_int(t), attr::UnsignedInt(t) => tcx.mk_mach_uint(t) } } // LLVM doesn't like types that don't fit in the address space fn ensure_struct_fits_in_address_space<'a, 'tcx>(ccx: &CrateContext<'a, 'tcx>, fields: &[Type], packed: bool, scapegoat: Ty<'tcx>) { let mut offset = 0; for &llty in fields { // Invariant: offset < ccx.obj_size_bound() <= 1<<61 if !packed { let type_align = machine::llalign_of_min(ccx, llty); offset = roundup(offset, type_align); } // type_align is a power-of-2, so still offset < ccx.obj_size_bound() // llsize_of_alloc(ccx, llty) is also less than ccx.obj_size_bound() // so the sum is less than 1<<62 (and therefore can't overflow). offset += machine::llsize_of_alloc(ccx, llty); if offset >= ccx.obj_size_bound() { ccx.report_overbig_object(scapegoat); } } } fn union_size_and_align(sts: &[Struct]) -> (machine::llsize, machine::llalign) { let size = sts.iter().map(|st| st.size).max().unwrap(); let align = sts.iter().map(|st| st.align).max().unwrap(); (roundup(size, align), align) } fn ensure_enum_fits_in_address_space<'a, 'tcx>(ccx: &CrateContext<'a, 'tcx>, fields: &[Struct], scapegoat: Ty<'tcx>) { let (total_size, _) = union_size_and_align(fields); if total_size >= ccx.obj_size_bound() { ccx.report_overbig_object(scapegoat); } } /// LLVM-level types are a little complicated. /// /// C-like enums need to be actual ints, not wrapped in a struct, /// because that changes the ABI on some platforms (see issue #10308). /// /// For nominal types, in some cases, we need to use LLVM named structs /// and fill in the actual contents in a second pass to prevent /// unbounded recursion; see also the comments in `trans::type_of`. pub fn type_of<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, r: &Repr<'tcx>) -> Type { generic_type_of(cx, r, None, false, false) } // Pass dst=true if the type you are passing is a DST. Yes, we could figure // this out, but if you call this on an unsized type without realising it, you // are going to get the wrong type (it will not include the unsized parts of it). pub fn sizing_type_of<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, r: &Repr<'tcx>, dst: bool) -> Type { generic_type_of(cx, r, None, true, dst) } pub fn incomplete_type_of<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, r: &Repr<'tcx>, name: &str) -> Type { generic_type_of(cx, r, Some(name), false, false) } pub fn finish_type_of<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, r: &Repr<'tcx>, llty: &mut Type) { match *r { CEnum(..) | General(..) | UntaggedUnion(..) | RawNullablePointer { .. } => { } Univariant(ref st) | StructWrappedNullablePointer { nonnull: ref st, .. } => llty.set_struct_body(&struct_llfields(cx, st, false, false), st.packed) } } fn generic_type_of<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, r: &Repr<'tcx>, name: Option<&str>, sizing: bool, dst: bool) -> Type { debug!("adt::generic_type_of r: {:?} name: {:?} sizing: {} dst: {}", r, name, sizing, dst); match *r { CEnum(ity, ..) => ll_inttype(cx, ity), RawNullablePointer { nnty, .. } => type_of::sizing_type_of(cx, nnty), StructWrappedNullablePointer { nonnull: ref st, .. } => { match name { None => { Type::struct_(cx, &struct_llfields(cx, st, sizing, dst), st.packed) } Some(name) => { assert_eq!(sizing, false); Type::named_struct(cx, name) } } } Univariant(ref st) => { match name { None => { let fields = struct_llfields(cx, st, sizing, dst); Type::struct_(cx, &fields, st.packed) } Some(name) => { // Hypothesis: named_struct's can never need a // drop flag. (... needs validation.) assert_eq!(sizing, false); Type::named_struct(cx, name) } } } UntaggedUnion(ref un) => { // Use alignment-sized ints to fill all the union storage. let (size, align) = (roundup(un.min_size, un.align), un.align); let align_s = align as u64; assert_eq!(size % align_s, 0); // Ensure division in align_units comes out evenly let align_units = size / align_s; let fill_ty = match align_s { 1 => Type::array(&Type::i8(cx), align_units), 2 => Type::array(&Type::i16(cx), align_units), 4 => Type::array(&Type::i32(cx), align_units), 8 if machine::llalign_of_min(cx, Type::i64(cx)) == 8 => Type::array(&Type::i64(cx), align_units), a if a.count_ones() == 1 => Type::array(&Type::vector(&Type::i32(cx), a / 4), align_units), _ => bug!("unsupported union alignment: {}", align) }; match name { None => { Type::struct_(cx, &[fill_ty], un.packed) } Some(name) => { let mut llty = Type::named_struct(cx, name); llty.set_struct_body(&[fill_ty], un.packed); llty } } } General(ity, ref sts) => { // We need a representation that has: // * The alignment of the most-aligned field // * The size of the largest variant (rounded up to that alignment) // * No alignment padding anywhere any variant has actual data // (currently matters only for enums small enough to be immediate) // * The discriminant in an obvious place. // // So we start with the discriminant, pad it up to the alignment with // more of its own type, then use alignment-sized ints to get the rest // of the size. // // FIXME #10604: this breaks when vector types are present. let (size, align) = union_size_and_align(&sts[..]); let align_s = align as u64; let discr_ty = ll_inttype(cx, ity); let discr_size = machine::llsize_of_alloc(cx, discr_ty); let padded_discr_size = roundup(discr_size, align); assert_eq!(size % align_s, 0); // Ensure division in align_units comes out evenly let align_units = (size - padded_discr_size) / align_s; let fill_ty = match align_s { 1 => Type::array(&Type::i8(cx), align_units), 2 => Type::array(&Type::i16(cx), align_units), 4 => Type::array(&Type::i32(cx), align_units), 8 if machine::llalign_of_min(cx, Type::i64(cx)) == 8 => Type::array(&Type::i64(cx), align_units), a if a.count_ones() == 1 => Type::array(&Type::vector(&Type::i32(cx), a / 4), align_units), _ => bug!("unsupported enum alignment: {}", align) }; assert_eq!(machine::llalign_of_min(cx, fill_ty), align); assert_eq!(padded_discr_size % discr_size, 0); // Ensure discr_ty can fill pad evenly let fields: Vec = [discr_ty, Type::array(&discr_ty, (padded_discr_size - discr_size)/discr_size), fill_ty].iter().cloned().collect(); match name { None => { Type::struct_(cx, &fields[..], false) } Some(name) => { let mut llty = Type::named_struct(cx, name); llty.set_struct_body(&fields[..], false); llty } } } } } fn struct_llfields<'a, 'tcx>(cx: &CrateContext<'a, 'tcx>, st: &Struct<'tcx>, sizing: bool, dst: bool) -> Vec { if sizing { st.fields.iter().filter(|&ty| !dst || type_is_sized(cx.tcx(), *ty)) .map(|&ty| type_of::sizing_type_of(cx, ty)).collect() } else { st.fields.iter().map(|&ty| type_of::in_memory_type_of(cx, ty)).collect() } } /// Obtain a representation of the discriminant sufficient to translate /// destructuring; this may or may not involve the actual discriminant. pub fn trans_switch<'blk, 'tcx>(bcx: Block<'blk, 'tcx>, r: &Repr<'tcx>, scrutinee: ValueRef, range_assert: bool) -> (BranchKind, Option) { match *r { CEnum(..) | General(..) | RawNullablePointer { .. } | StructWrappedNullablePointer { .. } => { (BranchKind::Switch, Some(trans_get_discr(bcx, r, scrutinee, None, range_assert))) } Univariant(..) | UntaggedUnion(..) => { // N.B.: Univariant means <= 1 enum variants (*not* == 1 variants). (BranchKind::Single, None) } } } pub fn is_discr_signed<'tcx>(r: &Repr<'tcx>) -> bool { match *r { CEnum(ity, ..) => ity.is_signed(), General(ity, _) => ity.is_signed(), Univariant(..) | UntaggedUnion(..) => false, RawNullablePointer { .. } => false, StructWrappedNullablePointer { .. } => false, } } /// Obtain the actual discriminant of a value. pub fn trans_get_discr<'blk, 'tcx>(bcx: Block<'blk, 'tcx>, r: &Repr<'tcx>, scrutinee: ValueRef, cast_to: Option, range_assert: bool) -> ValueRef { debug!("trans_get_discr r: {:?}", r); let val = match *r { CEnum(ity, min, max) => { load_discr(bcx, ity, scrutinee, min, max, range_assert) } General(ity, ref cases) => { let ptr = StructGEP(bcx, scrutinee, 0); load_discr(bcx, ity, ptr, Disr(0), Disr(cases.len() as u64 - 1), range_assert) } Univariant(..) | UntaggedUnion(..) => C_u8(bcx.ccx(), 0), RawNullablePointer { nndiscr, nnty, .. } => { let cmp = if nndiscr == Disr(0) { IntEQ } else { IntNE }; let llptrty = type_of::sizing_type_of(bcx.ccx(), nnty); ICmp(bcx, cmp, Load(bcx, scrutinee), C_null(llptrty), DebugLoc::None) } StructWrappedNullablePointer { nndiscr, ref discrfield, .. } => { struct_wrapped_nullable_bitdiscr(bcx, nndiscr, discrfield, scrutinee) } }; match cast_to { None => val, Some(llty) => if is_discr_signed(r) { SExt(bcx, val, llty) } else { ZExt(bcx, val, llty) } } } fn struct_wrapped_nullable_bitdiscr(bcx: Block, nndiscr: Disr, discrfield: &DiscrField, scrutinee: ValueRef) -> ValueRef { let llptrptr = GEPi(bcx, scrutinee, &discrfield[..]); let llptr = Load(bcx, llptrptr); let cmp = if nndiscr == Disr(0) { IntEQ } else { IntNE }; ICmp(bcx, cmp, llptr, C_null(val_ty(llptr)), DebugLoc::None) } /// Helper for cases where the discriminant is simply loaded. fn load_discr(bcx: Block, ity: IntType, ptr: ValueRef, min: Disr, max: Disr, range_assert: bool) -> ValueRef { let llty = ll_inttype(bcx.ccx(), ity); assert_eq!(val_ty(ptr), llty.ptr_to()); let bits = machine::llbitsize_of_real(bcx.ccx(), llty); assert!(bits <= 64); let bits = bits as usize; let mask = Disr(!0u64 >> (64 - bits)); // For a (max) discr of -1, max will be `-1 as usize`, which overflows. // However, that is fine here (it would still represent the full range), if max.wrapping_add(Disr(1)) & mask == min & mask || !range_assert { // i.e., if the range is everything. The lo==hi case would be // rejected by the LLVM verifier (it would mean either an // empty set, which is impossible, or the entire range of the // type, which is pointless). Load(bcx, ptr) } else { // llvm::ConstantRange can deal with ranges that wrap around, // so an overflow on (max + 1) is fine. LoadRangeAssert(bcx, ptr, min.0, max.0.wrapping_add(1), /* signed: */ True) } } /// Yield information about how to dispatch a case of the /// discriminant-like value returned by `trans_switch`. /// /// This should ideally be less tightly tied to `_match`. pub fn trans_case<'blk, 'tcx>(bcx: Block<'blk, 'tcx>, r: &Repr, discr: Disr) -> ValueRef { match *r { CEnum(ity, ..) => { C_integral(ll_inttype(bcx.ccx(), ity), discr.0, true) } General(ity, _) => { C_integral(ll_inttype(bcx.ccx(), ity), discr.0, true) } Univariant(..) | UntaggedUnion(..) => { bug!("no cases for univariants, structs or unions") } RawNullablePointer { .. } | StructWrappedNullablePointer { .. } => { assert!(discr == Disr(0) || discr == Disr(1)); C_bool(bcx.ccx(), discr != Disr(0)) } } } /// Set the discriminant for a new value of the given case of the given /// representation. pub fn trans_set_discr<'blk, 'tcx>(bcx: Block<'blk, 'tcx>, r: &Repr<'tcx>, val: ValueRef, discr: Disr) { match *r { CEnum(ity, min, max) => { assert_discr_in_range(ity, min, max, discr); Store(bcx, C_integral(ll_inttype(bcx.ccx(), ity), discr.0, true), val); } General(ity, _) => { Store(bcx, C_integral(ll_inttype(bcx.ccx(), ity), discr.0, true), StructGEP(bcx, val, 0)); } Univariant(_) => { assert_eq!(discr, Disr(0)); } UntaggedUnion(..) => { assert_eq!(discr, Disr(0)); } RawNullablePointer { nndiscr, nnty, ..} => { if discr != nndiscr { let llptrty = type_of::sizing_type_of(bcx.ccx(), nnty); Store(bcx, C_null(llptrty), val); } } StructWrappedNullablePointer { nndiscr, ref discrfield, .. } => { if discr != nndiscr { let llptrptr = GEPi(bcx, val, &discrfield[..]); let llptrty = val_ty(llptrptr).element_type(); Store(bcx, C_null(llptrty), llptrptr); } } } } fn assert_discr_in_range(ity: IntType, min: Disr, max: Disr, discr: Disr) { match ity { attr::UnsignedInt(_) => { assert!(min <= discr); assert!(discr <= max); }, attr::SignedInt(_) => { assert!(min.0 as i64 <= discr.0 as i64); assert!(discr.0 as i64 <= max.0 as i64); }, } } /// Access a field, at a point when the value's case is known. pub fn trans_field_ptr<'blk, 'tcx>(bcx: Block<'blk, 'tcx>, r: &Repr<'tcx>, val: MaybeSizedValue, discr: Disr, ix: usize) -> ValueRef { trans_field_ptr_builder(&bcx.build(), r, val, discr, ix) } /// Access a field, at a point when the value's case is known. pub fn trans_field_ptr_builder<'blk, 'tcx>(bcx: &BlockAndBuilder<'blk, 'tcx>, r: &Repr<'tcx>, val: MaybeSizedValue, discr: Disr, ix: usize) -> ValueRef { // Note: if this ever needs to generate conditionals (e.g., if we // decide to do some kind of cdr-coding-like non-unique repr // someday), it will need to return a possibly-new bcx as well. match *r { CEnum(..) => { bug!("element access in C-like enum") } Univariant(ref st) => { assert_eq!(discr, Disr(0)); struct_field_ptr(bcx, st, val, ix, false) } General(_, ref cases) => { struct_field_ptr(bcx, &cases[discr.0 as usize], val, ix + 1, true) } UntaggedUnion(ref un) => { let ty = type_of::in_memory_type_of(bcx.ccx(), un.fields[ix]); if bcx.is_unreachable() { return C_undef(ty.ptr_to()); } bcx.pointercast(val.value, ty.ptr_to()) } RawNullablePointer { nndiscr, ref nullfields, .. } | StructWrappedNullablePointer { nndiscr, ref nullfields, .. } if discr != nndiscr => { // The unit-like case might have a nonzero number of unit-like fields. // (e.d., Result of Either with (), as one side.) let ty = type_of::type_of(bcx.ccx(), nullfields[ix]); assert_eq!(machine::llsize_of_alloc(bcx.ccx(), ty), 0); // The contents of memory at this pointer can't matter, but use // the value that's "reasonable" in case of pointer comparison. if bcx.is_unreachable() { return C_undef(ty.ptr_to()); } bcx.pointercast(val.value, ty.ptr_to()) } RawNullablePointer { nndiscr, nnty, .. } => { assert_eq!(ix, 0); assert_eq!(discr, nndiscr); let ty = type_of::type_of(bcx.ccx(), nnty); if bcx.is_unreachable() { return C_undef(ty.ptr_to()); } bcx.pointercast(val.value, ty.ptr_to()) } StructWrappedNullablePointer { ref nonnull, nndiscr, .. } => { assert_eq!(discr, nndiscr); struct_field_ptr(bcx, nonnull, val, ix, false) } } } fn struct_field_ptr<'blk, 'tcx>(bcx: &BlockAndBuilder<'blk, 'tcx>, st: &Struct<'tcx>, val: MaybeSizedValue, ix: usize, needs_cast: bool) -> ValueRef { let ccx = bcx.ccx(); let fty = st.fields[ix]; let ll_fty = type_of::in_memory_type_of(bcx.ccx(), fty); if bcx.is_unreachable() { return C_undef(ll_fty.ptr_to()); } let ptr_val = if needs_cast { let fields = st.fields.iter().map(|&ty| { type_of::in_memory_type_of(ccx, ty) }).collect::>(); let real_ty = Type::struct_(ccx, &fields[..], st.packed); bcx.pointercast(val.value, real_ty.ptr_to()) } else { val.value }; // Simple case - we can just GEP the field // * First field - Always aligned properly // * Packed struct - There is no alignment padding // * Field is sized - pointer is properly aligned already if ix == 0 || st.packed || type_is_sized(bcx.tcx(), fty) { return bcx.struct_gep(ptr_val, ix); } // If the type of the last field is [T] or str, then we don't need to do // any adjusments match fty.sty { ty::TySlice(..) | ty::TyStr => { return bcx.struct_gep(ptr_val, ix); } _ => () } // There's no metadata available, log the case and just do the GEP. if !val.has_meta() { debug!("Unsized field `{}`, of `{:?}` has no metadata for adjustment", ix, Value(ptr_val)); return bcx.struct_gep(ptr_val, ix); } let dbloc = DebugLoc::None; // We need to get the pointer manually now. // We do this by casting to a *i8, then offsetting it by the appropriate amount. // We do this instead of, say, simply adjusting the pointer from the result of a GEP // because the field may have an arbitrary alignment in the LLVM representation // anyway. // // To demonstrate: // struct Foo { // x: u16, // y: T // } // // The type Foo> is represented in LLVM as { u16, { u16, u8 }}, meaning that // the `y` field has 16-bit alignment. let meta = val.meta; // Calculate the unaligned offset of the unsized field. let mut offset = 0; for &ty in &st.fields[0..ix] { let llty = type_of::sizing_type_of(ccx, ty); let type_align = type_of::align_of(ccx, ty); offset = roundup(offset, type_align); offset += machine::llsize_of_alloc(ccx, llty); } let unaligned_offset = C_uint(bcx.ccx(), offset); // Get the alignment of the field let (_, align) = glue::size_and_align_of_dst(bcx, fty, meta); // Bump the unaligned offset up to the appropriate alignment using the // following expression: // // (unaligned offset + (align - 1)) & -align // Calculate offset dbloc.apply(bcx.fcx()); let align_sub_1 = bcx.sub(align, C_uint(bcx.ccx(), 1u64)); let offset = bcx.and(bcx.add(unaligned_offset, align_sub_1), bcx.neg(align)); debug!("struct_field_ptr: DST field offset: {:?}", Value(offset)); // Cast and adjust pointer let byte_ptr = bcx.pointercast(ptr_val, Type::i8p(bcx.ccx())); let byte_ptr = bcx.gep(byte_ptr, &[offset]); // Finally, cast back to the type expected let ll_fty = type_of::in_memory_type_of(bcx.ccx(), fty); debug!("struct_field_ptr: Field type is {:?}", ll_fty); bcx.pointercast(byte_ptr, ll_fty.ptr_to()) } /// Construct a constant value, suitable for initializing a /// GlobalVariable, given a case and constant values for its fields. /// Note that this may have a different LLVM type (and different /// alignment!) from the representation's `type_of`, so it needs a /// pointer cast before use. /// /// The LLVM type system does not directly support unions, and only /// pointers can be bitcast, so a constant (and, by extension, the /// GlobalVariable initialized by it) will have a type that can vary /// depending on which case of an enum it is. /// /// To understand the alignment situation, consider `enum E { V64(u64), /// V32(u32, u32) }` on Windows. The type has 8-byte alignment to /// accommodate the u64, but `V32(x, y)` would have LLVM type `{i32, /// i32, i32}`, which is 4-byte aligned. /// /// Currently the returned value has the same size as the type, but /// this could be changed in the future to avoid allocating unnecessary /// space after values of shorter-than-maximum cases. pub fn trans_const<'a, 'tcx>(ccx: &CrateContext<'a, 'tcx>, r: &Repr<'tcx>, discr: Disr, vals: &[ValueRef]) -> ValueRef { match *r { CEnum(ity, min, max) => { assert_eq!(vals.len(), 0); assert_discr_in_range(ity, min, max, discr); C_integral(ll_inttype(ccx, ity), discr.0, true) } General(ity, ref cases) => { let case = &cases[discr.0 as usize]; let (max_sz, _) = union_size_and_align(&cases[..]); let lldiscr = C_integral(ll_inttype(ccx, ity), discr.0 as u64, true); let mut f = vec![lldiscr]; f.extend_from_slice(vals); let mut contents = build_const_struct(ccx, case, &f[..]); contents.extend_from_slice(&[padding(ccx, max_sz - case.size)]); C_struct(ccx, &contents[..], false) } UntaggedUnion(ref un) => { assert_eq!(discr, Disr(0)); let contents = build_const_union(ccx, un, vals[0]); C_struct(ccx, &contents, un.packed) } Univariant(ref st) => { assert_eq!(discr, Disr(0)); let contents = build_const_struct(ccx, st, vals); C_struct(ccx, &contents[..], st.packed) } RawNullablePointer { nndiscr, nnty, .. } => { if discr == nndiscr { assert_eq!(vals.len(), 1); vals[0] } else { C_null(type_of::sizing_type_of(ccx, nnty)) } } StructWrappedNullablePointer { ref nonnull, nndiscr, .. } => { if discr == nndiscr { C_struct(ccx, &build_const_struct(ccx, nonnull, vals), false) } else { let vals = nonnull.fields.iter().map(|&ty| { // Always use null even if it's not the `discrfield`th // field; see #8506. C_null(type_of::sizing_type_of(ccx, ty)) }).collect::>(); C_struct(ccx, &build_const_struct(ccx, nonnull, &vals[..]), false) } } } } /// Compute struct field offsets relative to struct begin. fn compute_struct_field_offsets<'a, 'tcx>(ccx: &CrateContext<'a, 'tcx>, st: &Struct<'tcx>) -> Vec { let mut offsets = vec!(); let mut offset = 0; for &ty in &st.fields { let llty = type_of::sizing_type_of(ccx, ty); if !st.packed { let type_align = type_of::align_of(ccx, ty); offset = roundup(offset, type_align); } offsets.push(offset); offset += machine::llsize_of_alloc(ccx, llty); } assert_eq!(st.fields.len(), offsets.len()); offsets } /// Building structs is a little complicated, because we might need to /// insert padding if a field's value is less aligned than its type. /// /// Continuing the example from `trans_const`, a value of type `(u32, /// E)` should have the `E` at offset 8, but if that field's /// initializer is 4-byte aligned then simply translating the tuple as /// a two-element struct will locate it at offset 4, and accesses to it /// will read the wrong memory. fn build_const_struct<'a, 'tcx>(ccx: &CrateContext<'a, 'tcx>, st: &Struct<'tcx>, vals: &[ValueRef]) -> Vec { assert_eq!(vals.len(), st.fields.len()); let target_offsets = compute_struct_field_offsets(ccx, st); // offset of current value let mut offset = 0; let mut cfields = Vec::new(); for (&val, target_offset) in vals.iter().zip(target_offsets) { if !st.packed { let val_align = machine::llalign_of_min(ccx, val_ty(val)); offset = roundup(offset, val_align); } if offset != target_offset { cfields.push(padding(ccx, target_offset - offset)); offset = target_offset; } assert!(!is_undef(val)); cfields.push(val); offset += machine::llsize_of_alloc(ccx, val_ty(val)); } assert!(st.sized && offset <= st.size); if offset != st.size { cfields.push(padding(ccx, st.size - offset)); } cfields } fn build_const_union<'a, 'tcx>(ccx: &CrateContext<'a, 'tcx>, un: &Union<'tcx>, field_val: ValueRef) -> Vec { let mut cfields = vec![field_val]; let offset = machine::llsize_of_alloc(ccx, val_ty(field_val)); let size = roundup(un.min_size, un.align); if offset != size { cfields.push(padding(ccx, size - offset)); } cfields } fn padding(ccx: &CrateContext, size: u64) -> ValueRef { C_undef(Type::array(&Type::i8(ccx), size)) } // FIXME this utility routine should be somewhere more general #[inline] fn roundup(x: u64, a: u32) -> u64 { let a = a as u64; ((x + (a - 1)) / a) * a } /// Extract a field of a constant value, as appropriate for its /// representation. /// /// (Not to be confused with `common::const_get_elt`, which operates on /// raw LLVM-level structs and arrays.) pub fn const_get_field(r: &Repr, val: ValueRef, _discr: Disr, ix: usize) -> ValueRef { match *r { CEnum(..) => bug!("element access in C-like enum const"), Univariant(..) => const_struct_field(val, ix), UntaggedUnion(..) => const_struct_field(val, 0), General(..) => const_struct_field(val, ix + 1), RawNullablePointer { .. } => { assert_eq!(ix, 0); val }, StructWrappedNullablePointer{ .. } => const_struct_field(val, ix) } } /// Extract field of struct-like const, skipping our alignment padding. fn const_struct_field(val: ValueRef, ix: usize) -> ValueRef { // Get the ix-th non-undef element of the struct. let mut real_ix = 0; // actual position in the struct let mut ix = ix; // logical index relative to real_ix let mut field; loop { loop { field = const_get_elt(val, &[real_ix]); if !is_undef(field) { break; } real_ix = real_ix + 1; } if ix == 0 { return field; } ix = ix - 1; real_ix = real_ix + 1; } }