Auto merge of #144325 - matthiaskrgr:rollup-0jbo4dl, r=matthiaskrgr

Rollup of 8 pull requests

Successful merges:

 - #144094 (Ensure we codegen the main fn)
 - #144173 (Remove tidy checks for `tests/ui/issues/`)
 - #144218 (Use serde for target spec json deserialize)
 - #144221 (generate elf symbol version in raw-dylib)
 - #144234 (Fix broken TLS destructors on 32-bit win7)
 - #144256 (Don't ICE on non-TypeId metadata within TypeId)
 - #144272 (resolve: Make disambiguators for underscore bindings module-local (take 2))
 - #144276 (Use less HIR in check_private_in_public.)

r? `@ghost`
`@rustbot` modify labels: rollup

Author: bors

Cargo.lock

@@ -4563,7 +4563,10 @@ dependencies = [
  "rustc_macros",
  "rustc_serialize",
  "rustc_span",
+ "serde",
+ "serde_derive",
  "serde_json",
+ "serde_path_to_error",
  "tracing",
 ]
 
@@ -4955,6 +4958,16 @@ dependencies = [
  "serde",
 ]
 
+[[package]]
+name = "serde_path_to_error"
+version = "0.1.17"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "59fab13f937fa393d08645bf3a84bdfe86e296747b506ada67bb15f10f218b2a"
+dependencies = [
+ "itoa",
+ "serde",
+]
+
 [[package]]
 name = "serde_spanned"
 version = "0.6.9"

compiler/rustc_codegen_ssa/src/back/link/raw_dylib.rs

@@ -4,7 +4,7 @@ use std::path::{Path, PathBuf};
 
 use rustc_abi::Endian;
 use rustc_data_structures::base_n::{CASE_INSENSITIVE, ToBaseN};
-use rustc_data_structures::fx::FxIndexMap;
+use rustc_data_structures::fx::{FxHashMap, FxIndexMap};
 use rustc_data_structures::stable_hasher::StableHasher;
 use rustc_hashes::Hash128;
 use rustc_session::Session;
@@ -230,40 +230,66 @@ fn create_elf_raw_dylib_stub(sess: &Session, soname: &str, symbols: &[DllImport]
         Endian::Little => object::Endianness::Little,
         Endian::Big => object::Endianness::Big,
     };
+
     let mut stub = write::Writer::new(endianness, true, &mut stub_buf);
 
+    let mut vers = Vec::new();
+    let mut vers_map = FxHashMap::default();
+    let mut syms = Vec::new();
+
+    for symbol in symbols {
+        let symbol_name = symbol.name.as_str();
+        if let Some((name, version_name)) = symbol_name.split_once('@') {
+            assert!(!version_name.contains('@'));
+            let dynstr = stub.add_dynamic_string(name.as_bytes());
+            let ver = if let Some(&ver_id) = vers_map.get(version_name) {
+                ver_id
+            } else {
+                let id = vers.len();
+                vers_map.insert(version_name, id);
+                let dynstr = stub.add_dynamic_string(version_name.as_bytes());
+                vers.push((version_name, dynstr));
+                id
+            };
+            syms.push((name, dynstr, Some(ver)));
+        } else {
+            let dynstr = stub.add_dynamic_string(symbol_name.as_bytes());
+            syms.push((symbol_name, dynstr, None));
+        }
+    }
+
+    let soname = stub.add_dynamic_string(soname.as_bytes());
+
     // These initial reservations don't reserve any bytes in the binary yet,
     // they just allocate in the internal data structures.
 
-    // First, we crate the dynamic symbol table. It starts with a null symbol
+    // First, we create the dynamic symbol table. It starts with a null symbol
     // and then all the symbols and their dynamic strings.
     stub.reserve_null_dynamic_symbol_index();
 
-    let dynstrs = symbols
-        .iter()
-        .map(|sym| {
-            stub.reserve_dynamic_symbol_index();
-            (sym, stub.add_dynamic_string(sym.name.as_str().as_bytes()))
-        })
-        .collect::<Vec<_>>();
-
-    let soname = stub.add_dynamic_string(soname.as_bytes());
+    for _ in syms.iter() {
+        stub.reserve_dynamic_symbol_index();
+    }
 
     // Reserve the sections.
     // We have the minimal sections for a dynamic SO and .text where we point our dummy symbols to.
     stub.reserve_shstrtab_section_index();
     let text_section_name = stub.add_section_name(".text".as_bytes());
     let text_section = stub.reserve_section_index();
-    stub.reserve_dynstr_section_index();
     stub.reserve_dynsym_section_index();
+    stub.reserve_dynstr_section_index();
+    stub.reserve_gnu_versym_section_index();
+    stub.reserve_gnu_verdef_section_index();
     stub.reserve_dynamic_section_index();
 
     // These reservations now determine the actual layout order of the object file.
     stub.reserve_file_header();
     stub.reserve_shstrtab();
     stub.reserve_section_headers();
-    stub.reserve_dynstr();
     stub.reserve_dynsym();
+    stub.reserve_dynstr();
+    stub.reserve_gnu_versym();
+    stub.reserve_gnu_verdef(1 + vers.len(), 1 + vers.len());
     stub.reserve_dynamic(2); // DT_SONAME, DT_NULL
 
     // First write the ELF header with the arch information.
@@ -342,18 +368,17 @@ fn create_elf_raw_dylib_stub(sess: &Session, soname: &str, symbols: &[DllImport]
         sh_addralign: 1,
         sh_entsize: 0,
     });
-    stub.write_dynstr_section_header(0);
     stub.write_dynsym_section_header(0, 1);
+    stub.write_dynstr_section_header(0);
+    stub.write_gnu_versym_section_header(0);
+    stub.write_gnu_verdef_section_header(0);
     stub.write_dynamic_section_header(0);
 
-    // .dynstr
-    stub.write_dynstr();
-
     // .dynsym
     stub.write_null_dynamic_symbol();
-    for (_, name) in dynstrs {
+    for (_name, dynstr, _ver) in syms.iter().copied() {
         stub.write_dynamic_symbol(&write::Sym {
-            name: Some(name),
+            name: Some(dynstr),
             st_info: (elf::STB_GLOBAL << 4) | elf::STT_NOTYPE,
             st_other: elf::STV_DEFAULT,
             section: Some(text_section),
@@ -363,10 +388,44 @@ fn create_elf_raw_dylib_stub(sess: &Session, soname: &str, symbols: &[DllImport]
         });
     }
 
+    // .dynstr
+    stub.write_dynstr();
+
+    // .gnu_version
+    stub.write_null_gnu_versym();
+    for (_name, _dynstr, ver) in syms.iter().copied() {
+        stub.write_gnu_versym(if let Some(ver) = ver {
+            assert!((2 + ver as u16) < elf::VERSYM_HIDDEN);
+            elf::VERSYM_HIDDEN | (2 + ver as u16)
+        } else {
+            1
+        });
+    }
+
+    // .gnu_version_d
+    stub.write_align_gnu_verdef();
+    stub.write_gnu_verdef(&write::Verdef {
+        version: elf::VER_DEF_CURRENT,
+        flags: elf::VER_FLG_BASE,
+        index: 1,
+        aux_count: 1,
+        name: soname,
+    });
+    for (ver, (_name, dynstr)) in vers.into_iter().enumerate() {
+        stub.write_gnu_verdef(&write::Verdef {
+            version: elf::VER_DEF_CURRENT,
+            flags: 0,
+            index: 2 + ver as u16,
+            aux_count: 1,
+            name: dynstr,
+        });
+    }
+
     // .dynamic
     // the DT_SONAME will be used by the linker to populate DT_NEEDED
     // which the loader uses to find the library.
     // DT_NULL terminates the .dynamic table.
+    stub.write_align_dynamic();
     stub.write_dynamic_string(elf::DT_SONAME, soname);
     stub.write_dynamic(elf::DT_NULL, 0);
 

compiler/rustc_const_eval/src/interpret/memory.rs

@@ -1000,7 +1000,7 @@ impl<'tcx, M: Machine<'tcx>> InterpCx<'tcx, M> {
         ptr: Pointer<Option<M::Provenance>>,
     ) -> InterpResult<'tcx, (Ty<'tcx>, u64)> {
         let (alloc_id, offset, _meta) = self.ptr_get_alloc_id(ptr, 0)?;
-        let GlobalAlloc::TypeId { ty } = self.tcx.global_alloc(alloc_id) else {
+        let Some(GlobalAlloc::TypeId { ty }) = self.tcx.try_get_global_alloc(alloc_id) else {
             throw_ub_format!("invalid `TypeId` value: not all bytes carry type id metadata")
         };
         interp_ok((ty, offset.bytes()))

compiler/rustc_metadata/messages.ftl

@@ -330,3 +330,6 @@ metadata_wasm_import_form =
 
 metadata_whole_archive_needs_static =
     linking modifier `whole-archive` is only compatible with `static` linking kind
+
+metadata_raw_dylib_malformed =
+    link name must be well-formed if link kind is `raw-dylib`

compiler/rustc_metadata/src/errors.rs

@@ -815,3 +815,10 @@ pub struct AsyncDropTypesInDependency {
     pub extern_crate: Symbol,
     pub local_crate: Symbol,
 }
+
+#[derive(Diagnostic)]
+#[diag(metadata_raw_dylib_malformed)]
+pub struct RawDylibMalformed {
+    #[primary_span]
+    pub span: Span,
+}

compiler/rustc_metadata/src/native_libs.rs

@@ -700,8 +700,21 @@ impl<'tcx> Collector<'tcx> {
             .link_ordinal
             .map_or(import_name_type, |ord| Some(PeImportNameType::Ordinal(ord)));
 
+        let name = codegen_fn_attrs.link_name.unwrap_or_else(|| self.tcx.item_name(item));
+
+        if self.tcx.sess.target.binary_format == BinaryFormat::Elf {
+            let name = name.as_str();
+            if name.contains('\0') {
+                self.tcx.dcx().emit_err(errors::RawDylibMalformed { span });
+            } else if let Some((left, right)) = name.split_once('@')
+                && (left.is_empty() || right.is_empty() || right.contains('@'))
+            {
+                self.tcx.dcx().emit_err(errors::RawDylibMalformed { span });
+            }
+        }
+
         DllImport {
-            name: codegen_fn_attrs.link_name.unwrap_or_else(|| self.tcx.item_name(item)),
+            name,
             import_name_type,
             calling_convention,
             span,

compiler/rustc_middle/src/mir/mono.rs

@@ -143,10 +143,8 @@ impl<'tcx> MonoItem<'tcx> {
         };
 
         // Similarly, the executable entrypoint must be instantiated exactly once.
-        if let Some((entry_def_id, _)) = tcx.entry_fn(()) {
-            if instance.def_id() == entry_def_id {
-                return InstantiationMode::GloballyShared { may_conflict: false };
-            }
+        if tcx.is_entrypoint(instance.def_id()) {
+            return InstantiationMode::GloballyShared { may_conflict: false };
         }
 
         // If the function is #[naked] or contains any other attribute that requires exactly-once

compiler/rustc_middle/src/ty/context.rs

@@ -3410,6 +3410,20 @@ impl<'tcx> TyCtxt<'tcx> {
     pub fn do_not_recommend_impl(self, def_id: DefId) -> bool {
         self.get_diagnostic_attr(def_id, sym::do_not_recommend).is_some()
     }
+
+    /// Whether this def is one of the special bin crate entrypoint functions that must have a
+    /// monomorphization and also not be internalized in the bin crate.
+    pub fn is_entrypoint(self, def_id: DefId) -> bool {
+        if self.is_lang_item(def_id, LangItem::Start) {
+            return true;
+        }
+        if let Some((entry_def_id, _)) = self.entry_fn(())
+            && entry_def_id == def_id
+        {
+            return true;
+        }
+        false
+    }
 }
 
 /// Parameter attributes that can only be determined by examining the body of a function instead

compiler/rustc_monomorphize/src/collector.rs

@@ -1582,6 +1582,15 @@ impl<'v> RootCollector<'_, 'v> {
             return;
         };
 
+        let main_instance = Instance::mono(self.tcx, main_def_id);
+        if self.tcx.should_codegen_locally(main_instance) {
+            self.output.push(create_fn_mono_item(
+                self.tcx,
+                main_instance,
+                self.tcx.def_span(main_def_id),
+            ));
+        }
+
         let Some(start_def_id) = self.tcx.lang_items().start_fn() else {
             self.tcx.dcx().emit_fatal(errors::StartNotFound);
         };

compiler/rustc_monomorphize/src/partitioning.rs

@@ -223,11 +223,7 @@ where
         // So even if its mode is LocalCopy, we need to treat it like a root.
         match mono_item.instantiation_mode(cx.tcx) {
             InstantiationMode::GloballyShared { .. } => {}
-            InstantiationMode::LocalCopy => {
-                if !cx.tcx.is_lang_item(mono_item.def_id(), LangItem::Start) {
-                    continue;
-                }
-            }
+            InstantiationMode::LocalCopy => continue,
         }
 
         let characteristic_def_id = characteristic_def_id_of_mono_item(cx.tcx, mono_item);
@@ -821,10 +817,9 @@ fn mono_item_visibility<'tcx>(
         | InstanceKind::FnPtrAddrShim(..) => return Visibility::Hidden,
     };
 
-    // The `start_fn` lang item is actually a monomorphized instance of a
-    // function in the standard library, used for the `main` function. We don't
-    // want to export it so we tag it with `Hidden` visibility but this symbol
-    // is only referenced from the actual `main` symbol which we unfortunately
+    // Both the `start_fn` lang item and `main` itself should not be exported,
+    // so we give them with `Hidden` visibility but these symbols are
+    // only referenced from the actual `main` symbol which we unfortunately
     // don't know anything about during partitioning/collection. As a result we
     // forcibly keep this symbol out of the `internalization_candidates` set.
     //
@@ -834,7 +829,7 @@ fn mono_item_visibility<'tcx>(
     //        from the `main` symbol we'll generate later.
     //
     //        This may be fixable with a new `InstanceKind` perhaps? Unsure!
-    if tcx.is_lang_item(def_id, LangItem::Start) {
+    if tcx.is_entrypoint(def_id) {
         *can_be_internalized = false;
         return Visibility::Hidden;
     }