Auto merge of rust-lang#127173 - bjorn3:mangle_rustc_std_internal_symbol, r=wesleywiser,jieyouxu

Mangle rustc_std_internal_symbols functions

This reduces the risk of issues when using a staticlib or rust dylib compiled with a different rustc version in a rust program. Currently this will either (in the case of staticlib) cause a linker error due to duplicate symbol definitions, or (in the case of rust dylibs) cause rustc_std_internal_symbols functions to be silently overridden. As rust gets more commonly used inside the implementation of libraries consumed with a C interface (like Spidermonkey, Ruby YJIT (curently has to do partial linking of all rust code to hide all symbols not part of the C api), the Rusticl OpenCL implementation in mesa) this is becoming much more of an issue. With this PR the only symbols remaining with an unmangled name are rust_eh_personality (LLVM doesn't allow renaming it) and `__rust_no_alloc_shim_is_unstable`.

Helps mitigate rust-lang#104707

try-job: aarch64-gnu-debug
try-job: aarch64-apple
try-job: x86_64-apple-1
try-job: x86_64-mingw-1
try-job: i686-mingw-1
try-job: x86_64-msvc-1
try-job: i686-msvc-1
try-job: test-various
try-job: armhf-gnu

Author: bors

compiler/rustc_codegen_cranelift/src/allocator.rs

@@ -7,13 +7,15 @@ use rustc_ast::expand::allocator::{
 };
 use rustc_codegen_ssa::base::allocator_kind_for_codegen;
 use rustc_session::config::OomStrategy;
+use rustc_symbol_mangling::mangle_internal_symbol;
 
 use crate::prelude::*;
 
 /// Returns whether an allocator shim was created
 pub(crate) fn codegen(tcx: TyCtxt<'_>, module: &mut dyn Module) -> bool {
     let Some(kind) = allocator_kind_for_codegen(tcx) else { return false };
     codegen_inner(
+        tcx,
         module,
         kind,
         tcx.alloc_error_handler_kind(()).unwrap(),
@@ -23,6 +25,7 @@ pub(crate) fn codegen(tcx: TyCtxt<'_>, module: &mut dyn Module) -> bool {
 }
 
 fn codegen_inner(
+    tcx: TyCtxt<'_>,
     module: &mut dyn Module,
     kind: AllocatorKind,
     alloc_error_handler_kind: AllocatorKind,
@@ -62,8 +65,8 @@ fn codegen_inner(
             crate::common::create_wrapper_function(
                 module,
                 sig,
-                &global_fn_name(method.name),
-                &default_fn_name(method.name),
+                &mangle_internal_symbol(tcx, &global_fn_name(method.name)),
+                &mangle_internal_symbol(tcx, &default_fn_name(method.name)),
             );
         }
     }
@@ -76,19 +79,32 @@ fn codegen_inner(
     crate::common::create_wrapper_function(
         module,
         sig,
-        "__rust_alloc_error_handler",
-        &alloc_error_handler_name(alloc_error_handler_kind),
+        &mangle_internal_symbol(tcx, "__rust_alloc_error_handler"),
+        &mangle_internal_symbol(tcx, alloc_error_handler_name(alloc_error_handler_kind)),
     );
 
-    let data_id = module.declare_data(OomStrategy::SYMBOL, Linkage::Export, false, false).unwrap();
+    let data_id = module
+        .declare_data(
+            &mangle_internal_symbol(tcx, OomStrategy::SYMBOL),
+            Linkage::Export,
+            false,
+            false,
+        )
+        .unwrap();
     let mut data = DataDescription::new();
     data.set_align(1);
     let val = oom_strategy.should_panic();
     data.define(Box::new([val]));
     module.define_data(data_id, &data).unwrap();
 
-    let data_id =
-        module.declare_data(NO_ALLOC_SHIM_IS_UNSTABLE, Linkage::Export, false, false).unwrap();
+    let data_id = module
+        .declare_data(
+            &mangle_internal_symbol(tcx, NO_ALLOC_SHIM_IS_UNSTABLE),
+            Linkage::Export,
+            false,
+            false,
+        )
+        .unwrap();
     let mut data = DataDescription::new();
     data.set_align(1);
     data.define(Box::new([0]));

compiler/rustc_codegen_cranelift/src/lib.rs

@@ -26,6 +26,7 @@ extern crate rustc_index;
 extern crate rustc_metadata;
 extern crate rustc_session;
 extern crate rustc_span;
+extern crate rustc_symbol_mangling;
 extern crate rustc_target;
 #[macro_use]
 extern crate tracing;

compiler/rustc_codegen_gcc/src/allocator.rs

@@ -8,6 +8,7 @@ use rustc_ast::expand::allocator::{
 use rustc_middle::bug;
 use rustc_middle::ty::TyCtxt;
 use rustc_session::config::OomStrategy;
+use rustc_symbol_mangling::mangle_internal_symbol;
 
 use crate::GccContext;
 #[cfg(feature = "master")]
@@ -53,8 +54,8 @@ pub(crate) unsafe fn codegen(
                     panic!("invalid allocator output")
                 }
             };
-            let from_name = global_fn_name(method.name);
-            let to_name = default_fn_name(method.name);
+            let from_name = mangle_internal_symbol(tcx, &global_fn_name(method.name));
+            let to_name = mangle_internal_symbol(tcx, &default_fn_name(method.name));
 
             create_wrapper_function(tcx, context, &from_name, &to_name, &types, output);
         }
@@ -64,13 +65,13 @@ pub(crate) unsafe fn codegen(
     create_wrapper_function(
         tcx,
         context,
-        "__rust_alloc_error_handler",
-        alloc_error_handler_name(alloc_error_handler_kind),
+        &mangle_internal_symbol(tcx, "__rust_alloc_error_handler"),
+        &mangle_internal_symbol(tcx, alloc_error_handler_name(alloc_error_handler_kind)),
         &[usize, usize],
         None,
     );
 
-    let name = OomStrategy::SYMBOL.to_string();
+    let name = mangle_internal_symbol(tcx, OomStrategy::SYMBOL);
     let global = context.new_global(None, GlobalKind::Exported, i8, name);
     #[cfg(feature = "master")]
     global.add_attribute(VarAttribute::Visibility(symbol_visibility_to_gcc(
@@ -80,7 +81,7 @@ pub(crate) unsafe fn codegen(
     let value = context.new_rvalue_from_int(i8, value as i32);
     global.global_set_initializer_rvalue(value);
 
-    let name = NO_ALLOC_SHIM_IS_UNSTABLE.to_string();
+    let name = mangle_internal_symbol(tcx, NO_ALLOC_SHIM_IS_UNSTABLE);
     let global = context.new_global(None, GlobalKind::Exported, i8, name);
     #[cfg(feature = "master")]
     global.add_attribute(VarAttribute::Visibility(symbol_visibility_to_gcc(

compiler/rustc_codegen_gcc/src/lib.rs

@@ -52,6 +52,7 @@ extern crate rustc_metadata;
 extern crate rustc_middle;
 extern crate rustc_session;
 extern crate rustc_span;
+extern crate rustc_symbol_mangling;
 extern crate rustc_target;
 
 // This prevents duplicating functions and statics that are already part of the host rustc process.

compiler/rustc_codegen_llvm/src/allocator.rs

@@ -7,6 +7,7 @@ use rustc_codegen_ssa::traits::BaseTypeCodegenMethods as _;
 use rustc_middle::bug;
 use rustc_middle::ty::TyCtxt;
 use rustc_session::config::{DebugInfo, OomStrategy};
+use rustc_symbol_mangling::mangle_internal_symbol;
 
 use crate::builder::SBuilder;
 use crate::declare::declare_simple_fn;
@@ -53,8 +54,8 @@ pub(crate) unsafe fn codegen(
                 }
             };
 
-            let from_name = global_fn_name(method.name);
-            let to_name = default_fn_name(method.name);
+            let from_name = mangle_internal_symbol(tcx, &global_fn_name(method.name));
+            let to_name = mangle_internal_symbol(tcx, &default_fn_name(method.name));
 
             create_wrapper_function(tcx, &cx, &from_name, &to_name, &args, output, false);
         }
@@ -64,24 +65,24 @@ pub(crate) unsafe fn codegen(
     create_wrapper_function(
         tcx,
         &cx,
-        "__rust_alloc_error_handler",
-        alloc_error_handler_name(alloc_error_handler_kind),
+        &mangle_internal_symbol(tcx, "__rust_alloc_error_handler"),
+        &mangle_internal_symbol(tcx, alloc_error_handler_name(alloc_error_handler_kind)),
         &[usize, usize], // size, align
         None,
         true,
     );
 
     unsafe {
         // __rust_alloc_error_handler_should_panic
-        let name = OomStrategy::SYMBOL;
-        let ll_g = cx.declare_global(name, i8);
+        let name = mangle_internal_symbol(tcx, OomStrategy::SYMBOL);
+        let ll_g = cx.declare_global(&name, i8);
         llvm::set_visibility(ll_g, llvm::Visibility::from_generic(tcx.sess.default_visibility()));
         let val = tcx.sess.opts.unstable_opts.oom.should_panic();
         let llval = llvm::LLVMConstInt(i8, val as u64, False);
         llvm::set_initializer(ll_g, llval);
 
-        let name = NO_ALLOC_SHIM_IS_UNSTABLE;
-        let ll_g = cx.declare_global(name, i8);
+        let name = mangle_internal_symbol(tcx, NO_ALLOC_SHIM_IS_UNSTABLE);
+        let ll_g = cx.declare_global(&name, i8);
         llvm::set_visibility(ll_g, llvm::Visibility::from_generic(tcx.sess.default_visibility()));
         let llval = llvm::LLVMConstInt(i8, 0, False);
         llvm::set_initializer(ll_g, llval);

compiler/rustc_codegen_llvm/src/context.rs

@@ -27,6 +27,7 @@ use rustc_session::config::{
 };
 use rustc_span::source_map::Spanned;
 use rustc_span::{DUMMY_SP, Span};
+use rustc_symbol_mangling::mangle_internal_symbol;
 use rustc_target::spec::{HasTargetSpec, RelocModel, SmallDataThresholdSupport, Target, TlsModel};
 use smallvec::SmallVec;
 
@@ -1199,7 +1200,7 @@ impl<'ll> CodegenCx<'ll, '_> {
             Some(def_id) => self.get_static(def_id),
             _ => {
                 let ty = self.type_struct(&[self.type_ptr(), self.type_ptr()], false);
-                self.declare_global("rust_eh_catch_typeinfo", ty)
+                self.declare_global(&mangle_internal_symbol(self.tcx, "rust_eh_catch_typeinfo"), ty)
             }
         };
         self.eh_catch_typeinfo.set(Some(eh_catch_typeinfo));

compiler/rustc_codegen_llvm/src/intrinsic.rs

@@ -14,6 +14,7 @@ use rustc_middle::ty::layout::{FnAbiOf, HasTyCtxt, HasTypingEnv, LayoutOf};
 use rustc_middle::ty::{self, GenericArgsRef, Ty};
 use rustc_middle::{bug, span_bug};
 use rustc_span::{Span, Symbol, sym};
+use rustc_symbol_mangling::mangle_internal_symbol;
 use rustc_target::callconv::{FnAbi, PassMode};
 use rustc_target::spec::{HasTargetSpec, PanicStrategy};
 use tracing::debug;
@@ -812,7 +813,10 @@ fn codegen_msvc_try<'ll>(
         let type_name = bx.const_bytes(b"rust_panic\0");
         let type_info =
             bx.const_struct(&[type_info_vtable, bx.const_null(bx.type_ptr()), type_name], false);
-        let tydesc = bx.declare_global("__rust_panic_type_info", bx.val_ty(type_info));
+        let tydesc = bx.declare_global(
+            &mangle_internal_symbol(bx.tcx, "__rust_panic_type_info"),
+            bx.val_ty(type_info),
+        );
 
         llvm::set_linkage(tydesc, llvm::Linkage::LinkOnceODRLinkage);
         if bx.cx.tcx.sess.target.supports_comdat() {

compiler/rustc_codegen_ssa/src/back/symbol_export.rs

@@ -1,6 +1,6 @@
 use std::collections::hash_map::Entry::*;
 
-use rustc_ast::expand::allocator::{ALLOCATOR_METHODS, NO_ALLOC_SHIM_IS_UNSTABLE};
+use rustc_ast::expand::allocator::{ALLOCATOR_METHODS, NO_ALLOC_SHIM_IS_UNSTABLE, global_fn_name};
 use rustc_data_structures::unord::UnordMap;
 use rustc_hir::def::DefKind;
 use rustc_hir::def_id::{CrateNum, DefId, DefIdMap, LOCAL_CRATE, LocalDefId};
@@ -13,6 +13,7 @@ use rustc_middle::query::LocalCrate;
 use rustc_middle::ty::{self, GenericArgKind, GenericArgsRef, Instance, SymbolName, Ty, TyCtxt};
 use rustc_middle::util::Providers;
 use rustc_session::config::{CrateType, OomStrategy};
+use rustc_symbol_mangling::mangle_internal_symbol;
 use rustc_target::callconv::Conv;
 use rustc_target::spec::{SanitizerSet, TlsModel};
 use tracing::debug;
@@ -219,8 +220,11 @@ fn exported_symbols_provider_local(
     if allocator_kind_for_codegen(tcx).is_some() {
         for symbol_name in ALLOCATOR_METHODS
             .iter()
-            .map(|method| format!("__rust_{}", method.name))
-            .chain(["__rust_alloc_error_handler".to_string(), OomStrategy::SYMBOL.to_string()])
+            .map(|method| mangle_internal_symbol(tcx, global_fn_name(method.name).as_str()))
+            .chain([
+                mangle_internal_symbol(tcx, "__rust_alloc_error_handler"),
+                mangle_internal_symbol(tcx, OomStrategy::SYMBOL),
+            ])
         {
             let exported_symbol = ExportedSymbol::NoDefId(SymbolName::new(tcx, &symbol_name));
 
@@ -234,8 +238,10 @@ fn exported_symbols_provider_local(
             ));
         }
 
-        let exported_symbol =
-            ExportedSymbol::NoDefId(SymbolName::new(tcx, NO_ALLOC_SHIM_IS_UNSTABLE));
+        let exported_symbol = ExportedSymbol::NoDefId(SymbolName::new(
+            tcx,
+            &mangle_internal_symbol(tcx, NO_ALLOC_SHIM_IS_UNSTABLE),
+        ));
         symbols.push((
             exported_symbol,
             SymbolExportInfo {

compiler/rustc_codegen_ssa/src/base.rs

@@ -28,6 +28,7 @@ use rustc_middle::ty::{self, Instance, Ty, TyCtxt};
 use rustc_session::Session;
 use rustc_session::config::{self, CrateType, EntryFnType, OutputType};
 use rustc_span::{DUMMY_SP, Symbol, sym};
+use rustc_symbol_mangling::mangle_internal_symbol;
 use rustc_trait_selection::infer::{BoundRegionConversionTime, TyCtxtInferExt};
 use rustc_trait_selection::traits::{ObligationCause, ObligationCtxt};
 use tracing::{debug, info};
@@ -989,7 +990,12 @@ impl CrateInfo {
                 .for_each(|(_, linked_symbols)| {
                     let mut symbols = missing_weak_lang_items
                         .iter()
-                        .map(|item| (format!("{prefix}{item}"), SymbolExportKind::Text))
+                        .map(|item| {
+                            (
+                                format!("{prefix}{}", mangle_internal_symbol(tcx, item.as_str())),
+                                SymbolExportKind::Text,
+                            )
+                        })
                         .collect::<Vec<_>>();
                     symbols.sort_unstable_by(|a, b| a.0.cmp(&b.0));
                     linked_symbols.extend(symbols);
@@ -1002,7 +1008,13 @@ impl CrateInfo {
                         // errors.
                         linked_symbols.extend(ALLOCATOR_METHODS.iter().map(|method| {
                             (
-                                format!("{prefix}{}", global_fn_name(method.name).as_str()),
+                                format!(
+                                    "{prefix}{}",
+                                    mangle_internal_symbol(
+                                        tcx,
+                                        global_fn_name(method.name).as_str()
+                                    )
+                                ),
                                 SymbolExportKind::Text,
                             )
                         }));

compiler/rustc_codegen_ssa/src/codegen_attrs.rs

@@ -601,25 +601,19 @@ fn codegen_fn_attrs(tcx: TyCtxt<'_>, did: LocalDefId) -> CodegenFnAttrs {
     // strippable by the linker.
     //
     // Additionally weak lang items have predetermined symbol names.
-    if WEAK_LANG_ITEMS.iter().any(|&l| tcx.lang_items().get(l) == Some(did.to_def_id())) {
-        codegen_fn_attrs.flags |= CodegenFnAttrFlags::RUSTC_STD_INTERNAL_SYMBOL;
-    }
     if let Some((name, _)) = lang_items::extract(attrs)
         && let Some(lang_item) = LangItem::from_name(name)
-        && let Some(link_name) = lang_item.link_name()
     {
-        codegen_fn_attrs.export_name = Some(link_name);
-        codegen_fn_attrs.link_name = Some(link_name);
+        if WEAK_LANG_ITEMS.iter().any(|&l| l == lang_item) {
+            codegen_fn_attrs.flags |= CodegenFnAttrFlags::RUSTC_STD_INTERNAL_SYMBOL;
+        }
+        if let Some(link_name) = lang_item.link_name() {
+            codegen_fn_attrs.export_name = Some(link_name);
+            codegen_fn_attrs.link_name = Some(link_name);
+        }
     }
     check_link_name_xor_ordinal(tcx, &codegen_fn_attrs, link_ordinal_span);
 
-    // Internal symbols to the standard library all have no_mangle semantics in
-    // that they have defined symbol names present in the function name. This
-    // also applies to weak symbols where they all have known symbol names.
-    if codegen_fn_attrs.flags.contains(CodegenFnAttrFlags::RUSTC_STD_INTERNAL_SYMBOL) {
-        codegen_fn_attrs.flags |= CodegenFnAttrFlags::NO_MANGLE;
-    }
-
     // Any linkage to LLVM intrinsics for now forcibly marks them all as never
     // unwinds since LLVM sometimes can't handle codegen which `invoke`s
     // intrinsic functions.