From 7e2cc725dbe00a9c99b0d58df020ab970e53dd51 Mon Sep 17 00:00:00 2001 From: nerix Date: Tue, 12 Aug 2025 03:51:37 +0200 Subject: [PATCH] [LLDB][NativePDB] Find functions by basename (#152295) This adds the ability for functions to be matched by their basename. Before, the globals were searched for the name. This works if the full name is available but fails for basenames. PDB only includes the full names of functions, so we need to cache all basenames. This is (again) very similar to [SymbolFilePDB](https://github.com/llvm/llvm-project/blob/b242150b075a8a720b00821682a9469258bbcd30/lldb/source/Plugins/SymbolFile/PDB/SymbolFilePDB.cpp#L1291-L1383). There are two main differences: - We can't just access the parent of a function to determine that it's a member function - we need to check the type of the function, and its "this type". - SymbolFilePDB saved the full method name in the map. However, when searching for methods, only the basename is passed, so the function never found any methods. Fixes #51933. --------- Co-authored-by: Jonas Devlieghere --- .../NativePDB/SymbolFileNativePDB.cpp | 153 +++++++++++++++--- .../NativePDB/SymbolFileNativePDB.h | 9 ++ .../SymbolFile/NativePDB/find-functions.cpp | 87 +++++++++- 3 files changed, 224 insertions(+), 25 deletions(-) diff --git a/lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.cpp b/lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.cpp index dcea33dd9f85..986d647b4de2 100644 --- a/lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.cpp +++ b/lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.cpp @@ -39,6 +39,7 @@ #include "llvm/DebugInfo/PDB/Native/ModuleDebugStream.h" #include "llvm/DebugInfo/PDB/Native/NativeSession.h" #include "llvm/DebugInfo/PDB/Native/PDBFile.h" +#include "llvm/DebugInfo/PDB/Native/PublicsStream.h" #include "llvm/DebugInfo/PDB/Native/SymbolStream.h" #include "llvm/DebugInfo/PDB/Native/TpiStream.h" #include "llvm/DebugInfo/PDB/PDB.h" @@ -1641,6 +1642,94 @@ void SymbolFileNativePDB::DumpClangAST(Stream &s, llvm::StringRef filter) { clang->GetNativePDBParser()->Dump(s, filter); } +void SymbolFileNativePDB::CacheFunctionNames() { + if (!m_func_full_names.IsEmpty()) + return; + + // (segment, code offset) -> gid + std::map, uint32_t> addr_ids; + + // First, find all function references in the globals table. + for (const uint32_t gid : m_index->globals().getGlobalsTable()) { + CVSymbol ref_sym = m_index->symrecords().readRecord(gid); + auto kind = ref_sym.kind(); + if (kind != S_PROCREF && kind != S_LPROCREF) + continue; + + ProcRefSym ref = + llvm::cantFail(SymbolDeserializer::deserializeAs(ref_sym)); + if (ref.Name.empty()) + continue; + + // Find the function this is referencing. + CompilandIndexItem &cci = + m_index->compilands().GetOrCreateCompiland(ref.modi()); + auto iter = cci.m_debug_stream.getSymbolArray().at(ref.SymOffset); + if (iter == cci.m_debug_stream.getSymbolArray().end()) + continue; + kind = iter->kind(); + if (kind != S_GPROC32 && kind != S_LPROC32) + continue; + + ProcSym proc = + llvm::cantFail(SymbolDeserializer::deserializeAs(*iter)); + if ((proc.Flags & ProcSymFlags::IsUnreachable) != ProcSymFlags::None) + continue; + if (proc.Name.empty()) + continue; + + // The function/procedure symbol only contains the demangled name. + // The mangled names are in the publics table. Save the address of this + // function to lookup the mangled name later. + addr_ids.emplace(std::make_pair(proc.Segment, proc.CodeOffset), gid); + + llvm::StringRef basename = MSVCUndecoratedNameParser::DropScope(proc.Name); + if (basename.empty()) + basename = proc.Name; + + m_func_base_names.Append(ConstString(basename), gid); + m_func_full_names.Append(ConstString(proc.Name), gid); + + // To see if this is a member function, check the type. + auto type = m_index->tpi().getType(proc.FunctionType); + if (type.kind() == LF_MFUNCTION) { + MemberFunctionRecord mfr; + llvm::cantFail( + TypeDeserializer::deserializeAs(type, mfr)); + if (!mfr.getThisType().isNoneType()) + m_func_method_names.Append(ConstString(basename), gid); + } + } + + // The publics stream contains all mangled function names and their address. + for (auto pid : m_index->publics().getPublicsTable()) { + PdbGlobalSymId global{pid, true}; + CVSymbol sym = m_index->ReadSymbolRecord(global); + auto kind = sym.kind(); + if (kind != S_PUB32) + continue; + PublicSym32 pub = + llvm::cantFail(SymbolDeserializer::deserializeAs(sym)); + // We only care about mangled names - if the name isn't mangled, it's + // already in the full name map. + if (!Mangled::IsMangledName(pub.Name)) + continue; + + // Check if this symbol is for one of our functions. + auto it = addr_ids.find({pub.Segment, pub.Offset}); + if (it != addr_ids.end()) + m_func_full_names.Append(ConstString(pub.Name), it->second); + } + + // Sort them before value searching is working properly. + m_func_full_names.Sort(); + m_func_full_names.SizeToFit(); + m_func_method_names.Sort(); + m_func_method_names.SizeToFit(); + m_func_base_names.Sort(); + m_func_base_names.SizeToFit(); +} + void SymbolFileNativePDB::FindGlobalVariables( ConstString name, const CompilerDeclContext &parent_decl_ctx, uint32_t max_matches, VariableList &variables) { @@ -1677,34 +1766,60 @@ void SymbolFileNativePDB::FindFunctions( if (name_type_mask & eFunctionNameTypeFull) name = lookup_info.GetName(); - // For now we only support lookup by method name or full name. if (!(name_type_mask & eFunctionNameTypeFull || + name_type_mask & eFunctionNameTypeBase || name_type_mask & eFunctionNameTypeMethod)) return; + CacheFunctionNames(); - using SymbolAndOffset = std::pair; + std::set resolved_ids; // avoid duplicate lookups + auto resolve_from = [&](UniqueCStringMap &Names) { + std::vector ids; + if (!Names.GetValues(name, ids)) + return; - std::vector matches = m_index->globals().findRecordsByName( - name.GetStringRef(), m_index->symrecords()); - for (const SymbolAndOffset &match : matches) { - if (match.second.kind() != S_PROCREF && match.second.kind() != S_LPROCREF) - continue; - ProcRefSym proc(match.second.kind()); - cantFail(SymbolDeserializer::deserializeAs(match.second, proc)); + for (uint32_t id : ids) { + if (!resolved_ids.insert(id).second) + continue; - if (!IsValidRecord(proc)) - continue; + PdbGlobalSymId global{id, false}; + if (parent_decl_ctx.IsValid() && + GetDeclContextContainingUID(toOpaqueUid(global)) != parent_decl_ctx) + continue; - CompilandIndexItem &cci = - m_index->compilands().GetOrCreateCompiland(proc.modi()); - SymbolContext sc; + CVSymbol sym = m_index->ReadSymbolRecord(global); + auto kind = sym.kind(); + lldbassert(kind == S_PROCREF || kind == S_LPROCREF); - sc.comp_unit = GetOrCreateCompileUnit(cci).get(); - PdbCompilandSymId func_id(proc.modi(), proc.SymOffset); - sc.function = GetOrCreateFunction(func_id, *sc.comp_unit).get(); + ProcRefSym proc = + cantFail(SymbolDeserializer::deserializeAs(sym)); - sc_list.Append(sc); - } + if (!IsValidRecord(proc)) + continue; + + CompilandIndexItem &cci = + m_index->compilands().GetOrCreateCompiland(proc.modi()); + SymbolContext sc; + + sc.comp_unit = GetOrCreateCompileUnit(cci).get(); + if (!sc.comp_unit) + continue; + + PdbCompilandSymId func_id(proc.modi(), proc.SymOffset); + sc.function = GetOrCreateFunction(func_id, *sc.comp_unit).get(); + if (!sc.function) + continue; + + sc_list.Append(sc); + } + }; + + if (name_type_mask & eFunctionNameTypeFull) + resolve_from(m_func_full_names); + if (name_type_mask & eFunctionNameTypeBase) + resolve_from(m_func_base_names); + if (name_type_mask & eFunctionNameTypeMethod) + resolve_from(m_func_method_names); } void SymbolFileNativePDB::FindFunctions(const RegularExpression ®ex, diff --git a/lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.h b/lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.h index eda375d4cebe..6bbeb8bb1442 100644 --- a/lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.h +++ b/lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.h @@ -260,6 +260,8 @@ private: std::vector GetContextForType(llvm::codeview::TypeIndex ti); + void CacheFunctionNames(); + llvm::BumpPtrAllocator m_allocator; lldb::addr_t m_obj_load_address = 0; @@ -282,6 +284,13 @@ private: m_parent_types; lldb_private::UniqueCStringMap m_type_base_names; + + /// mangled name/full function name -> Global ID(s) + lldb_private::UniqueCStringMap m_func_full_names; + /// basename -> Global ID(s) + lldb_private::UniqueCStringMap m_func_base_names; + /// method basename -> Global ID(s) + lldb_private::UniqueCStringMap m_func_method_names; }; } // namespace npdb diff --git a/lldb/test/Shell/SymbolFile/NativePDB/find-functions.cpp b/lldb/test/Shell/SymbolFile/NativePDB/find-functions.cpp index 5ebef61bdbfe..37a0b10d445f 100644 --- a/lldb/test/Shell/SymbolFile/NativePDB/find-functions.cpp +++ b/lldb/test/Shell/SymbolFile/NativePDB/find-functions.cpp @@ -6,24 +6,52 @@ // RUN: lldb-test symbols --find=function --name=main --function-flags=full %t.exe \ // RUN: | FileCheck %s --check-prefix=FIND-MAIN +// RUN: lldb-test symbols --find=function --name=main --function-flags=method %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-NO-FUNCTION +// RUN: lldb-test symbols --find=function --name=main --function-flags=base %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-MAIN // RUN: lldb-test symbols --find=function --name=static_fn --function-flags=full %t.exe \ // RUN: | FileCheck %s --check-prefix=FIND-STATIC +// RUN: lldb-test symbols --find=function --name=static_fn --function-flags=method %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-NO-FUNCTION +// RUN: lldb-test symbols --find=function --name=static_fn --function-flags=base %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-STATIC // RUN: lldb-test symbols --find=function --name=varargs_fn --function-flags=full %t.exe \ // RUN: | FileCheck %s --check-prefix=FIND-VAR +// RUN: lldb-test symbols --find=function --name=varargs_fn --function-flags=method %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-NO-FUNCTION +// RUN: lldb-test symbols --find=function --name=varargs_fn --function-flags=base %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-VAR // RUN: lldb-test symbols --find=function --name=Struct::simple_method --function-flags=full %t.exe \ // RUN: | FileCheck %s --check-prefix=FIND-SIMPLE +// RUN: lldb-test symbols --find=function --name=Struct::simple_method --function-flags=method %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-SIMPLE +// RUN: lldb-test symbols --find=function --name=Struct::simple_method --function-flags=base %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-SIMPLE // RUN: lldb-test symbols --find=function --name=Struct::virtual_method --function-flags=full %t.exe \ // RUN: | FileCheck %s --check-prefix=FIND-VIRTUAL +// RUN: lldb-test symbols --find=function --name=Struct::virtual_method --function-flags=method %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-VIRTUAL +// RUN: lldb-test symbols --find=function --name=Struct::virtual_method --function-flags=base %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-VIRTUAL // RUN: lldb-test symbols --find=function --name=Struct::static_method --function-flags=full %t.exe \ // RUN: | FileCheck %s --check-prefix=FIND-STATIC-METHOD +// RUN: lldb-test symbols --find=function --name=Struct::static_method --function-flags=method %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-NO-FUNCTION +// RUN: lldb-test symbols --find=function --name=Struct::static_method --function-flags=base %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-STATIC-METHOD // RUN: lldb-test symbols --find=function --name=Struct::overloaded_method --function-flags=full %t.exe \ -// RUN: | FileCheck %s --check-prefix=FIND-OVERLOAD +// RUN: | FileCheck %s --check-prefix=FIND-OVERLOAD-FULL +// RUN: lldb-test symbols --find=function --name=Struct::overloaded_method --function-flags=method %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-OVERLOAD-METHOD +// RUN: lldb-test symbols --find=function --name=Struct::overloaded_method --function-flags=base %t.exe \ +// RUN: | FileCheck %s --check-prefix=FIND-OVERLOAD-BASE struct Struct { int simple_method() { @@ -51,7 +79,28 @@ private: } }; +class Class { +public: + bool overloaded_method() { + return false; + } + bool overloaded_method(int i) { + return i > 0; + } + static int overloaded_method(bool b) { + return b ? 1 : 2; + } +}; + +char overloaded_method() { + return 0; +} +char overloaded_method(int i) { + return 0; +} + Struct s; +Class c; static int static_fn() { return 42; @@ -63,12 +112,16 @@ int varargs_fn(int x, int y, ...) { int main(int argc, char **argv) { return static_fn() + varargs_fn(argc, argc) + s.simple_method() + - Struct::static_method() + s.virtual_method() + s.overloaded_method(); + Struct::static_method() + s.virtual_method() + s.overloaded_method() + + Class::overloaded_method(false) + c.overloaded_method(1) + c.overloaded_method() + + overloaded_method() + overloaded_method(1); } // FIND-MAIN: Function: id = {{.*}}, name = "main" // FIND-MAIN-NEXT: FuncType: id = {{.*}}, compiler_type = "int (int, char **)" +// FIND-NO-FUNCTION: Found 0 functions + // FIND-STATIC: Function: id = {{.*}}, name = "{{.*}}static_fn{{.*}}" // FIND-STATIC-NEXT: FuncType: id = {{.*}}, compiler_type = "int (void)" @@ -84,7 +137,29 @@ int main(int argc, char **argv) { // FIND-STATIC-METHOD: Function: id = {{.*}}, name = "{{.*}}Struct::static_method{{.*}}" // FIND-STATIC-METHOD-NEXT: FuncType: id = {{.*}}, compiler_type = "int (void)" -// FIND-OVERLOAD: Function: id = {{.*}}, name = "{{.*}}Struct::overloaded_method{{.*}}" -// FIND-OVERLOAD: FuncType: id = {{.*}}, compiler_type = "int (void)" -// FIND-OVERLOAD: FuncType: id = {{.*}}, compiler_type = "int (char)" -// FIND-OVERLOAD: FuncType: id = {{.*}}, compiler_type = "int (char, int, ...)" +// FIND-OVERLOAD-FULL-NOT: "Class::overloaded_method" +// FIND-OVERLOAD-FULL-NOT: "overloaded_method" +// FIND-OVERLOAD-FULL: Function: id = {{.*}}, name = "{{.*}}Struct::overloaded_method{{.*}}" +// FIND-OVERLOAD-FULL: FuncType: id = {{.*}}, compiler_type = "int (void)" +// FIND-OVERLOAD-FULL: FuncType: id = {{.*}}, compiler_type = "int (char)" +// FIND-OVERLOAD-FULL: FuncType: id = {{.*}}, compiler_type = "int (char, int, ...)" + +// FIND-OVERLOAD-BASE-DAG: Function: id = {{.*}}, name = "{{.*}}Struct::overloaded_method{{.*}}" +// FIND-OVERLOAD-BASE-DAG: FuncType: id = {{.*}}, compiler_type = "int (void)" +// FIND-OVERLOAD-BASE-DAG: FuncType: id = {{.*}}, compiler_type = "int (char)" +// FIND-OVERLOAD-BASE-DAG: FuncType: id = {{.*}}, compiler_type = "int (char, int, ...)" +// FIND-OVERLOAD-BASE-DAG: Function: id = {{.*}}, name = "Class::overloaded_method" +// FIND-OVERLOAD-BASE-DAG: FuncType: id = {{.*}}, compiler_type = "_Bool (void)" +// FIND-OVERLOAD-BASE-DAG: FuncType: id = {{.*}}, compiler_type = "_Bool (int)" +// FIND-OVERLOAD-BASE-DAG: FuncType: id = {{.*}}, compiler_type = "int (_Bool)" +// FIND-OVERLOAD-BASE-DAG: Function: id = {{.*}}, name = "overloaded_method" +// FIND-OVERLOAD-BASE-DAG: FuncType: id = {{.*}}, compiler_type = "char (void)" +// FIND-OVERLOAD-BASE-DAG: FuncType: id = {{.*}}, compiler_type = "char (int)" + +// FIND-OVERLOAD-METHOD-NOT: "overloaded_method" +// FIND-OVERLOAD-METHOD-DAG: Function: id = {{.*}}, name = "{{.*}}Struct::overloaded_method{{.*}}" +// FIND-OVERLOAD-METHOD-DAG: FuncType: id = {{.*}}, compiler_type = "int (void)" +// FIND-OVERLOAD-METHOD-DAG: FuncType: id = {{.*}}, compiler_type = "int (char)" +// FIND-OVERLOAD-METHOD-DAG: Function: id = {{.*}}, name = "Class::overloaded_method" +// FIND-OVERLOAD-METHOD-DAG: FuncType: id = {{.*}}, compiler_type = "_Bool (void)" +// FIND-OVERLOAD-METHOD-DAG: FuncType: id = {{.*}}, compiler_type = "_Bool (int)"