Ensure nested allocations in statics do not get deduplicated

Author: oli-obk

compiler/rustc_codegen_gcc/src/mono_item.rs

@@ -6,6 +6,7 @@ use rustc_middle::middle::codegen_fn_attrs::CodegenFnAttrFlags;
 use rustc_middle::mir::mono::{Linkage, Visibility};
 use rustc_middle::ty::{self, Instance, TypeVisitableExt};
 use rustc_middle::ty::layout::{FnAbiOf, LayoutOf};
+use rustc_hir::def::DefKind;
 
 use crate::attributes;
 use crate::base;
@@ -17,7 +18,10 @@ impl<'gcc, 'tcx> PreDefineMethods<'tcx> for CodegenCx<'gcc, 'tcx> {
     fn predefine_static(&self, def_id: DefId, _linkage: Linkage, visibility: Visibility, symbol_name: &str) {
         let attrs = self.tcx.codegen_fn_attrs(def_id);
         let instance = Instance::mono(self.tcx, def_id);
-        let ty = instance.ty(self.tcx, ty::ParamEnv::reveal_all());
+        let DefKind::Static { nested, .. } = self.tcx.def_kind(def_id) else { bug!() };
+        // Nested statics do not have a type, so pick a random type and let `define_static` figure out
+        // the llvm type from the actual evaluated initializer.
+        let ty = if nested { self.tcx.types.unit} else { instance.ty(self.tcx, ty::ParamEnv::reveal_all())  };
         let gcc_type = self.layout_of(ty).gcc_type(self);
 
         let is_tls = attrs.flags.contains(CodegenFnAttrFlags::THREAD_LOCAL);

compiler/rustc_codegen_llvm/src/consts.rs

@@ -9,6 +9,7 @@ use crate::type_::Type;
 use crate::type_of::LayoutLlvmExt;
 use crate::value::Value;
 use rustc_codegen_ssa::traits::*;
+use rustc_hir::def::DefKind;
 use rustc_hir::def_id::DefId;
 use rustc_middle::middle::codegen_fn_attrs::{CodegenFnAttrFlags, CodegenFnAttrs};
 use rustc_middle::mir::interpret::{
@@ -229,9 +230,17 @@ impl<'ll> CodegenCx<'ll, '_> {
     pub(crate) fn get_static(&self, def_id: DefId) -> &'ll Value {
         let instance = Instance::mono(self.tcx, def_id);
         trace!(?instance);
-        let ty = instance.ty(self.tcx, ty::ParamEnv::reveal_all());
-        trace!(?ty);
-        let llty = self.layout_of(ty).llvm_type(self);
+
+        let DefKind::Static { nested, .. } = self.tcx.def_kind(def_id) else { bug!() };
+        // Nested statics do not have a type, so pick a random type and let `define_static` figure out
+        // the llvm type from the actual evaluated initializer.
+        let llty = if nested {
+            self.type_i8()
+        } else {
+            let ty = instance.ty(self.tcx, ty::ParamEnv::reveal_all());
+            trace!(?ty);
+            self.layout_of(ty).llvm_type(self)
+        };
         self.get_static_inner(def_id, llty)
     }
 
@@ -346,6 +355,9 @@ impl<'ll> CodegenCx<'ll, '_> {
 
     fn codegen_static_item(&self, def_id: DefId) {
         unsafe {
+            assert!(
+                llvm::LLVMGetInitializer(self.statics.borrow().get(&def_id).unwrap()).is_none()
+            );
             let attrs = self.tcx.codegen_fn_attrs(def_id);
 
             let Ok((v, alloc)) = codegen_static_initializer(self, def_id) else {

compiler/rustc_codegen_llvm/src/mono_item.rs

@@ -5,7 +5,9 @@ use crate::errors::SymbolAlreadyDefined;
 use crate::llvm;
 use crate::type_of::LayoutLlvmExt;
 use rustc_codegen_ssa::traits::*;
+use rustc_hir::def::DefKind;
 use rustc_hir::def_id::{DefId, LOCAL_CRATE};
+use rustc_middle::bug;
 use rustc_middle::mir::mono::{Linkage, Visibility};
 use rustc_middle::ty::layout::{FnAbiOf, LayoutOf};
 use rustc_middle::ty::{self, Instance, TypeVisitableExt};
@@ -21,7 +23,14 @@ impl<'tcx> PreDefineMethods<'tcx> for CodegenCx<'_, 'tcx> {
         symbol_name: &str,
     ) {
         let instance = Instance::mono(self.tcx, def_id);
-        let ty = instance.ty(self.tcx, ty::ParamEnv::reveal_all());
+        let DefKind::Static { nested, .. } = self.tcx.def_kind(def_id) else { bug!() };
+        // Nested statics do not have a type, so pick a random type and let `define_static` figure out
+        // the llvm type from the actual evaluated initializer.
+        let ty = if nested {
+            self.tcx.types.unit
+        } else {
+            instance.ty(self.tcx, ty::ParamEnv::reveal_all())
+        };
         let llty = self.layout_of(ty).llvm_type(self);
 
         let g = self.define_global(symbol_name, llty).unwrap_or_else(|| {

compiler/rustc_const_eval/src/const_eval/eval_queries.rs

@@ -59,7 +59,7 @@ fn eval_body_using_ecx<'mir, 'tcx>(
     };
 
     let ret = if let InternKind::Static(_) = intern_kind {
-        create_static_alloc(ecx, cid.instance.def_id(), layout)?
+        create_static_alloc(ecx, cid.instance.def_id().expect_local(), layout)?
     } else {
         ecx.allocate(layout, MemoryKind::Stack)?
     };
@@ -381,10 +381,10 @@ pub fn eval_in_interpreter<'mir, 'tcx>(
         Ok(mplace) => {
             // Since evaluation had no errors, validate the resulting constant.
 
-            // Temporarily allow access to the static_root_alloc_id for the purpose of validation.
-            let static_root_alloc_id = ecx.machine.static_root_alloc_id.take();
+            // Temporarily allow access to the static_root_ids for the purpose of validation.
+            let static_root_ids = ecx.machine.static_root_ids.take();
             let validation = const_validate_mplace(&ecx, &mplace, cid);
-            ecx.machine.static_root_alloc_id = static_root_alloc_id;
+            ecx.machine.static_root_ids = static_root_ids;
 
             let alloc_id = mplace.ptr().provenance.unwrap().alloc_id();
 

compiler/rustc_const_eval/src/const_eval/machine.rs

@@ -8,6 +8,7 @@ use rustc_data_structures::fx::FxIndexMap;
 use rustc_data_structures::fx::IndexEntry;
 use rustc_hir::def::DefKind;
 use rustc_hir::def_id::DefId;
+use rustc_hir::def_id::LocalDefId;
 use rustc_hir::LangItem;
 use rustc_middle::mir;
 use rustc_middle::mir::AssertMessage;
@@ -60,7 +61,19 @@ pub struct CompileTimeInterpreter<'mir, 'tcx> {
     pub(super) check_alignment: CheckAlignment,
 
     /// Used to prevent reads from a static's base allocation, as that may allow for self-initialization.
-    pub(crate) static_root_alloc_id: Option<AllocId>,
+    pub(crate) static_root_ids: Option<(AllocId, LocalDefId)>,
+}
+
+pub trait HasStaticRootDefId {
+    /// Returns the `DefId` of the static item that is currently being evaluated.
+    /// Used for interning to be able to handle nested allocations.
+    fn static_def_id(&self) -> Option<LocalDefId>;
+}
+
+impl HasStaticRootDefId for CompileTimeInterpreter<'_, '_> {
+    fn static_def_id(&self) -> Option<LocalDefId> {
+        Some(self.static_root_ids?.1)
+    }
 }
 
 #[derive(Copy, Clone)]
@@ -94,7 +107,7 @@ impl<'mir, 'tcx> CompileTimeInterpreter<'mir, 'tcx> {
             stack: Vec::new(),
             can_access_mut_global,
             check_alignment,
-            static_root_alloc_id: None,
+            static_root_ids: None,
         }
     }
 }
@@ -751,7 +764,7 @@ impl<'mir, 'tcx> interpret::Machine<'mir, 'tcx> for CompileTimeInterpreter<'mir,
         ecx: &InterpCx<'mir, 'tcx, Self>,
         alloc_id: AllocId,
     ) -> InterpResult<'tcx> {
-        if Some(alloc_id) == ecx.machine.static_root_alloc_id {
+        if Some(alloc_id) == ecx.machine.static_root_ids.map(|(id, _)| id) {
             Err(ConstEvalErrKind::RecursiveStatic.into())
         } else {
             Ok(())

compiler/rustc_const_eval/src/interpret/intern.rs

@@ -13,12 +13,16 @@
 //! but that would require relying on type information, and given how many ways Rust has to lie
 //! about type information, we want to avoid doing that.
 
+use hir::def::DefKind;
 use rustc_ast::Mutability;
 use rustc_data_structures::fx::{FxHashSet, FxIndexMap};
 use rustc_errors::ErrorGuaranteed;
 use rustc_hir as hir;
-use rustc_middle::mir::interpret::{CtfeProvenance, InterpResult};
+use rustc_middle::mir::interpret::{ConstAllocation, CtfeProvenance, InterpResult};
+use rustc_middle::query::TyCtxtAt;
 use rustc_middle::ty::layout::TyAndLayout;
+use rustc_span::def_id::LocalDefId;
+use rustc_span::sym;
 
 use super::{AllocId, Allocation, InterpCx, MPlaceTy, Machine, MemoryKind, PlaceTy};
 use crate::const_eval;
@@ -33,7 +37,7 @@ pub trait CompileTimeMachine<'mir, 'tcx: 'mir, T> = Machine<
         FrameExtra = (),
         AllocExtra = (),
         MemoryMap = FxIndexMap<AllocId, (MemoryKind<T>, Allocation)>,
-    >;
+    > + const_eval::HasStaticRootDefId;
 
 /// Intern an allocation. Returns `Err` if the allocation does not exist in the local memory.
 ///
@@ -67,10 +71,34 @@ fn intern_shallow<'rt, 'mir, 'tcx, T, M: CompileTimeMachine<'mir, 'tcx, T>>(
     }
     // link the alloc id to the actual allocation
     let alloc = ecx.tcx.mk_const_alloc(alloc);
-    ecx.tcx.set_alloc_id_memory(alloc_id, alloc);
+    if let Some(static_id) = ecx.machine.static_def_id() {
+        intern_as_new_static(ecx.tcx, static_id, alloc_id, alloc);
+    } else {
+        ecx.tcx.set_alloc_id_memory(alloc_id, alloc);
+    }
     Ok(alloc.0.0.provenance().ptrs().iter().map(|&(_, prov)| prov))
 }
 
+/// Creates a new `DefId` and feeds all the right queries to make this `DefId`
+/// appear as if it were a user-written `static` (though it has no HIR).
+fn intern_as_new_static<'tcx>(
+    tcx: TyCtxtAt<'tcx>,
+    static_id: LocalDefId,
+    alloc_id: AllocId,
+    alloc: ConstAllocation<'tcx>,
+) {
+    let feed = tcx.create_def(
+        static_id,
+        sym::nested,
+        DefKind::Static { mt: alloc.0.mutability, nested: true },
+    );
+    tcx.set_nested_alloc_id_static(alloc_id, feed.def_id());
+    feed.codegen_fn_attrs(tcx.codegen_fn_attrs(static_id).clone());
+    feed.eval_static_initializer(Ok(alloc));
+    feed.generics_of(tcx.generics_of(static_id).clone());
+    feed.def_ident_span(tcx.def_ident_span(static_id));
+    feed.explicit_predicates_of(tcx.explicit_predicates_of(static_id));
+}
 /// How a constant value should be interned.
 #[derive(Copy, Clone, Debug, PartialEq, Hash, Eq)]
 pub enum InternKind {

compiler/rustc_const_eval/src/interpret/memory.rs

@@ -14,6 +14,7 @@ use std::ptr;
 
 use rustc_ast::Mutability;
 use rustc_data_structures::fx::{FxHashSet, FxIndexMap};
+use rustc_hir::def::DefKind;
 use rustc_middle::mir::display_allocation;
 use rustc_middle::ty::{self, Instance, ParamEnv, Ty, TyCtxt};
 use rustc_target::abi::{Align, HasDataLayout, Size};
@@ -744,19 +745,33 @@ impl<'mir, 'tcx: 'mir, M: Machine<'mir, 'tcx>> InterpCx<'mir, 'tcx, M> {
         // be held throughout the match.
         match self.tcx.try_get_global_alloc(id) {
             Some(GlobalAlloc::Static(def_id)) => {
-                assert!(self.tcx.is_static(def_id));
                 // Thread-local statics do not have a constant address. They *must* be accessed via
                 // `ThreadLocalRef`; we can never have a pointer to them as a regular constant value.
                 assert!(!self.tcx.is_thread_local_static(def_id));
-                // Use size and align of the type.
-                let ty = self
-                    .tcx
-                    .type_of(def_id)
-                    .no_bound_vars()
-                    .expect("statics should not have generic parameters");
-                let layout = self.tcx.layout_of(ParamEnv::empty().and(ty)).unwrap();
-                assert!(layout.is_sized());
-                (layout.size, layout.align.abi, AllocKind::LiveData)
+
+                let DefKind::Static { nested, .. } = self.tcx.def_kind(def_id) else {
+                    bug!("GlobalAlloc::Static is not a static")
+                };
+
+                let (size, align) = if nested {
+                    // Nested anonymous statics are untyped, so let's get their
+                    // size and alignment from the allocaiton itself. This always
+                    // succeeds, as the query is fed at DefId creation time, so no
+                    // evaluation actually occurs.
+                    let alloc = self.tcx.eval_static_initializer(def_id).unwrap();
+                    (alloc.0.size(), alloc.0.align)
+                } else {
+                    // Use size and align of the type.
+                    let ty = self
+                        .tcx
+                        .type_of(def_id)
+                        .no_bound_vars()
+                        .expect("statics should not have generic parameters");
+                    let layout = self.tcx.layout_of(ParamEnv::empty().and(ty)).unwrap();
+                    assert!(layout.is_sized());
+                    (layout.size, layout.align.abi)
+                };
+                (size, align, AllocKind::LiveData)
             }
             Some(GlobalAlloc::Memory(alloc)) => {
                 // Need to duplicate the logic here, because the global allocations have

compiler/rustc_const_eval/src/interpret/util.rs

@@ -1,11 +1,11 @@
 use crate::const_eval::CompileTimeEvalContext;
 use crate::interpret::{MemPlaceMeta, MemoryKind};
+use rustc_hir::def_id::LocalDefId;
 use rustc_middle::mir::interpret::{AllocId, Allocation, InterpResult, Pointer};
 use rustc_middle::ty::layout::TyAndLayout;
 use rustc_middle::ty::{
     self, Ty, TyCtxt, TypeSuperVisitable, TypeVisitable, TypeVisitableExt, TypeVisitor,
 };
-use rustc_span::def_id::DefId;
 use std::ops::ControlFlow;
 
 use super::MPlaceTy;
@@ -89,13 +89,13 @@ pub(crate) fn take_static_root_alloc<'mir, 'tcx: 'mir>(
 
 pub(crate) fn create_static_alloc<'mir, 'tcx: 'mir>(
     ecx: &mut CompileTimeEvalContext<'mir, 'tcx>,
-    static_def_id: DefId,
+    static_def_id: LocalDefId,
     layout: TyAndLayout<'tcx>,
 ) -> InterpResult<'tcx, MPlaceTy<'tcx>> {
     let alloc = Allocation::try_uninit(layout.size, layout.align.abi)?;
-    let alloc_id = ecx.tcx.reserve_and_set_static_alloc(static_def_id);
-    assert_eq!(ecx.machine.static_root_alloc_id, None);
-    ecx.machine.static_root_alloc_id = Some(alloc_id);
+    let alloc_id = ecx.tcx.reserve_and_set_static_alloc(static_def_id.into());
+    assert_eq!(ecx.machine.static_root_ids, None);
+    ecx.machine.static_root_ids = Some((alloc_id, static_def_id));
     assert!(ecx.memory.alloc_map.insert(alloc_id, (MemoryKind::Stack, alloc)).is_none());
     Ok(ecx.ptr_with_meta_to_mplace(Pointer::from(alloc_id).into(), MemPlaceMeta::None, layout))
 }

compiler/rustc_hir/src/def.rs

@@ -250,6 +250,9 @@ impl DefKind {
             | DefKind::AssocTy
             | DefKind::TyParam
             | DefKind::ExternCrate => DefPathData::TypeNs(name),
+            // It's not exactly an anon const, but wrt DefPathData, there
+            // is not difference.
+            DefKind::Static { nested: true, .. } => DefPathData::AnonConst,
             DefKind::Fn
             | DefKind::Const
             | DefKind::ConstParam

compiler/rustc_metadata/src/rmeta/encoder.rs

@@ -882,7 +882,7 @@ fn should_encode_attrs(def_kind: DefKind) -> bool {
         | DefKind::AssocTy
         | DefKind::Fn
         | DefKind::Const
-        | DefKind::Static { .. }
+        | DefKind::Static { nested: false, .. }
         | DefKind::AssocFn
         | DefKind::AssocConst
         | DefKind::Macro(_)
@@ -903,6 +903,7 @@ fn should_encode_attrs(def_kind: DefKind) -> bool {
         | DefKind::InlineConst
         | DefKind::OpaqueTy
         | DefKind::LifetimeParam
+        | DefKind::Static { nested: true, .. }
         | DefKind::GlobalAsm => false,
     }
 }
@@ -956,7 +957,7 @@ fn should_encode_visibility(def_kind: DefKind) -> bool {
         | DefKind::AssocTy
         | DefKind::Fn
         | DefKind::Const
-        | DefKind::Static { .. }
+        | DefKind::Static { nested: false, .. }
         | DefKind::Ctor(..)
         | DefKind::AssocFn
         | DefKind::AssocConst
@@ -969,6 +970,7 @@ fn should_encode_visibility(def_kind: DefKind) -> bool {
         | DefKind::LifetimeParam
         | DefKind::AnonConst
         | DefKind::InlineConst
+        | DefKind::Static { nested: true, .. }
         | DefKind::OpaqueTy
         | DefKind::GlobalAsm
         | DefKind::Impl { .. }
@@ -1148,7 +1150,7 @@ fn should_encode_type(tcx: TyCtxt<'_>, def_id: LocalDefId, def_kind: DefKind) ->
         | DefKind::Field
         | DefKind::Fn
         | DefKind::Const
-        | DefKind::Static { .. }
+        | DefKind::Static { nested: false, .. }
         | DefKind::TyAlias
         | DefKind::ForeignTy
         | DefKind::Impl { .. }
@@ -1190,6 +1192,7 @@ fn should_encode_type(tcx: TyCtxt<'_>, def_id: LocalDefId, def_kind: DefKind) ->
         | DefKind::Mod
         | DefKind::ForeignMod
         | DefKind::Macro(..)
+        | DefKind::Static { nested: true, .. }
         | DefKind::Use
         | DefKind::LifetimeParam
         | DefKind::GlobalAsm

compiler/rustc_middle/src/mir/interpret/mod.rs

@@ -130,7 +130,7 @@ use rustc_data_structures::fx::FxHashMap;
 use rustc_data_structures::sync::{HashMapExt, Lock};
 use rustc_data_structures::tiny_list::TinyList;
 use rustc_errors::ErrorGuaranteed;
-use rustc_hir::def_id::DefId;
+use rustc_hir::def_id::{DefId, LocalDefId};
 use rustc_macros::HashStable;
 use rustc_middle::ty::print::with_no_trimmed_paths;
 use rustc_serialize::{Decodable, Encodable};
@@ -627,6 +627,16 @@ impl<'tcx> TyCtxt<'tcx> {
         }
     }
 
+    /// Freezes an `AllocId` created with `reserve` by pointing it at a static item. Trying to
+    /// call this function twice, even with the same `DefId` will ICE the compiler.
+    pub fn set_nested_alloc_id_static(self, id: AllocId, def_id: LocalDefId) {
+        if let Some(old) =
+            self.alloc_map.lock().alloc_map.insert(id, GlobalAlloc::Static(def_id.to_def_id()))
+        {
+            bug!("tried to set allocation ID {id:?}, but it was already existing as {old:#?}");
+        }
+    }
+
     /// Freezes an `AllocId` created with `reserve` by pointing it at an `Allocation`. May be called
     /// twice for the same `(AllocId, Allocation)` pair.
     fn set_alloc_id_same_memory(self, id: AllocId, mem: ConstAllocation<'tcx>) {