🍒 [lldb] SymbolLocatorSymStore 1/2 (local lookup)

lldb/source/Plugins/ObjectFile/PECOFF/ObjectFilePECOFF.cpp

@@ -1108,6 +1108,27 @@ std::optional<FileSpec> ObjectFilePECOFF::GetDebugLink() {
   return std::nullopt;
 }
 
+std::optional<FileSpec> ObjectFilePECOFF::GetPDBPath() {
+  llvm::StringRef pdb_file;
+  const llvm::codeview::DebugInfo *pdb_info = nullptr;
+  if (llvm::Error Err = m_binary->getDebugPDBInfo(pdb_info, pdb_file)) {
+    // DebugInfo section is corrupt.
+    Log *log = GetLog(LLDBLog::Object);
+    llvm::StringRef file = m_binary->getFileName();
+    LLDB_LOG_ERROR(
+        log, std::move(Err),
+        "Failed to read Codeview record for PDB debug info file ({1}): {0}",
+        file);
+    return std::nullopt;
+  }
+  if (pdb_file.empty()) {
+    // No DebugInfo section present.
+    return std::nullopt;
+  }
+  return FileSpec(pdb_file, FileSpec::GuessPathStyle(pdb_file).value_or(
+                                FileSpec::Style::native));
+}
+
 uint32_t ObjectFilePECOFF::ParseDependentModules() {
   ModuleSP module_sp(GetModule());
   if (!module_sp)

lldb/source/Plugins/ObjectFile/PECOFF/ObjectFilePECOFF.h

@@ -130,6 +130,8 @@ class ObjectFilePECOFF : public lldb_private::ObjectFile {
   /// contains it.
   std::optional<lldb_private::FileSpec> GetDebugLink();
 
+  std::optional<lldb_private::FileSpec> GetPDBPath();
+
   uint32_t GetDependentModules(lldb_private::FileSpecList &files) override;
 
   lldb_private::Address GetEntryPointAddress() override;

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"
@@ -367,6 +368,11 @@ uint32_t SymbolFileNativePDB::CalculateAbilities() {
     if (!pdb_file)
       return 0;
 
+    LLDB_LOG(
+        GetLog(LLDBLog::Symbols), "Loading {0} for {1}",
+        pdb_file->getFilePath(),
+        m_objfile_sp->GetModule()->GetObjectFile()->GetFileSpec().GetPath());
+
     auto expected_index = PdbIndex::create(pdb_file);
     if (!expected_index) {
       llvm::consumeError(expected_index.takeError());
@@ -1163,7 +1169,80 @@ lldb::LanguageType SymbolFileNativePDB::ParseLanguage(CompileUnit &comp_unit) {
   return TranslateLanguage(item->m_compile_opts->getLanguage());
 }
 
-void SymbolFileNativePDB::AddSymbols(Symtab &symtab) {}
+void SymbolFileNativePDB::AddSymbols(Symtab &symtab) {
+  auto *section_list =
+      m_objfile_sp->GetModule()->GetObjectFile()->GetSectionList();
+  if (!section_list)
+    return;
+
+  PublicSym32 last_sym;
+  size_t last_sym_idx = 0;
+  lldb::SectionSP section_sp;
+
+  // To estimate the size of a symbol, we use the difference to the next symbol.
+  // If there's no next symbol or the section/segment changed, the symbol will
+  // take the remaining space. The estimate can be too high in case there's
+  // padding between symbols. This similar to the algorithm used by the DIA
+  // SDK.
+  auto finish_last_symbol = [&](const PublicSym32 *next) {
+    if (!section_sp)
+      return;
+    Symbol *last = symtab.SymbolAtIndex(last_sym_idx);
+    if (!last)
+      return;
+
+    if (next && last_sym.Segment == next->Segment) {
+      assert(last_sym.Offset <= next->Offset);
+      last->SetByteSize(next->Offset - last_sym.Offset);
+    } else {
+      // the last symbol was the last in its section
+      assert(section_sp->GetByteSize() >= last_sym.Offset);
+      assert(!next || next->Segment > last_sym.Segment);
+      last->SetByteSize(section_sp->GetByteSize() - last_sym.Offset);
+    }
+  };
+
+  // The address map is sorted by the address of a symbol.
+  for (auto pid : m_index->publics().getAddressMap()) {
+    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<PublicSym32>(sym));
+    finish_last_symbol(&pub);
+
+    if (!section_sp || last_sym.Segment != pub.Segment)
+      section_sp = section_list->FindSectionByID(pub.Segment);
+
+    if (!section_sp)
+      continue;
+
+    lldb::SymbolType type = eSymbolTypeData;
+    if ((pub.Flags & PublicSymFlags::Function) != PublicSymFlags::None ||
+        (pub.Flags & PublicSymFlags::Code) != PublicSymFlags::None)
+      type = eSymbolTypeCode;
+
+    last_sym_idx =
+        symtab.AddSymbol(Symbol(/*symID=*/pid,
+                                /*name=*/pub.Name,
+                                /*type=*/type,
+                                /*external=*/true,
+                                /*is_debug=*/true,
+                                /*is_trampoline=*/false,
+                                /*is_artificial=*/false,
+                                /*section_sp=*/section_sp,
+                                /*value=*/pub.Offset,
+                                /*size=*/0,
+                                /*size_is_valid=*/false,
+                                /*contains_linker_annotations=*/false,
+                                /*flags=*/0));
+    last_sym = pub;
+  }
+
+  finish_last_symbol(nullptr);
+}
 
 size_t SymbolFileNativePDB::ParseFunctions(CompileUnit &comp_unit) {
   std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
@@ -1783,6 +1862,94 @@ void SymbolFileNativePDB::DumpClangAST(Stream &s, llvm::StringRef filter) {
   ast_builder->Dump(s, filter);
 }
 
+void SymbolFileNativePDB::CacheFunctionNames() {
+  if (!m_func_full_names.IsEmpty())
+    return;
+
+  // (segment, code offset) -> gid
+  std::map<std::pair<uint16_t, uint32_t>, 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<ProcRefSym>(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<ProcSym>(*iter));
+    if ((proc.Flags & ProcSymFlags::IsUnreachable) != ProcSymFlags::None)
+      continue;
+    if (proc.Name.empty() || proc.FunctionType.isSimple())
+      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<MemberFunctionRecord>(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<PublicSym32>(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) {
@@ -1819,34 +1986,60 @@ 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<uint32_t, llvm::codeview::CVSymbol>;
+  std::set<uint32_t> resolved_ids; // avoid duplicate lookups
+  auto resolve_from = [&](UniqueCStringMap<uint32_t> &Names) {
+    std::vector<uint32_t> ids;
+    if (!Names.GetValues(name, ids))
+      return;
 
-  std::vector<SymbolAndOffset> 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<ProcRefSym>(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<ProcRefSym>(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 &regex,

lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.h

@@ -259,6 +259,8 @@ class SymbolFileNativePDB : public SymbolFileCommon {
           fn);
 
   void ParseInlineSite(PdbCompilandSymId inline_site_id, Address func_addr);
+
+  void CacheFunctionNames();
 
   llvm::BumpPtrAllocator m_allocator;
 
@@ -280,6 +282,13 @@ class SymbolFileNativePDB : public SymbolFileCommon {
   llvm::DenseMap<lldb::user_id_t, std::shared_ptr<InlineSite>> m_inline_sites;
   llvm::DenseMap<llvm::codeview::TypeIndex, llvm::codeview::TypeIndex>
       m_parent_types;
+
+  /// mangled name/full function name -> Global ID(s)
+  lldb_private::UniqueCStringMap<uint32_t> m_func_full_names;
+  /// basename -> Global ID(s)
+  lldb_private::UniqueCStringMap<uint32_t> m_func_base_names;
+  /// method basename -> Global ID(s)
+  lldb_private::UniqueCStringMap<uint32_t> m_func_method_names;
 };
 
 } // namespace npdb

lldb/source/Plugins/SymbolFile/PDB/SymbolFilePDB.cpp

@@ -153,7 +153,8 @@ uint32_t SymbolFilePDB::CalculateAbilities() {
 
   if (!m_session_up) {
     // Lazily load and match the PDB file, but only do this once.
-    std::string exePath = m_objfile_sp->GetFileSpec().GetPath();
+    std::string exePath =
+        m_objfile_sp->GetModule()->GetObjectFile()->GetFileSpec().GetPath();
     auto error = loadDataForEXE(PDB_ReaderType::DIA, llvm::StringRef(exePath),
                                 m_session_up);
     if (error) {
@@ -163,6 +164,10 @@ uint32_t SymbolFilePDB::CalculateAbilities() {
         return 0;
       // See if any symbol file is specified through `--symfile` option.
       FileSpec symfile = module_sp->GetSymbolFileFileSpec();
+      // If m_objfile_sp is a supplemental ObjectFilePDB (e.g. located via
+      // SymStore by SymbolVendorPECOFF), its file spec is the PDB path.
+      if (!symfile && m_objfile_sp.get() != module_sp->GetObjectFile())
+        symfile = m_objfile_sp->GetFileSpec();
       if (!symfile)
         return 0;
       error = loadDataForPDB(PDB_ReaderType::DIA,
@@ -201,8 +206,10 @@ uint32_t SymbolFilePDB::CalculateAbilities() {
 }
 
 void SymbolFilePDB::InitializeObject() {
-  lldb::addr_t obj_load_address =
-      m_objfile_sp->GetBaseAddress().GetFileAddress();
+  lldb::addr_t obj_load_address = m_objfile_sp->GetModule()
+                                      ->GetObjectFile()
+                                      ->GetBaseAddress()
+                                      .GetFileAddress();
   lldbassert(obj_load_address && obj_load_address != LLDB_INVALID_ADDRESS);
   m_session_up->setLoadAddress(obj_load_address);
   if (!m_global_scope_up)
@@ -1392,7 +1399,8 @@ void SymbolFilePDB::AddSymbols(lldb_private::Symtab &symtab) {
   if (!results)
     return;
 
-  auto section_list = m_objfile_sp->GetSectionList();
+  auto section_list =
+      m_objfile_sp->GetModule()->GetObjectFile()->GetSectionList();
   if (!section_list)
     return;
 

lldb/source/Plugins/SymbolLocator/CMakeLists.txt

@@ -6,6 +6,7 @@ set_property(DIRECTORY PROPERTY LLDB_PLUGIN_KIND SymbolLocator)
 # prevents an unstripped binary from being requested from the Debuginfod
 # provider.
 add_subdirectory(Debuginfod)
+add_subdirectory(SymStore)
 add_subdirectory(Default)
 if (CMAKE_SYSTEM_NAME MATCHES "Darwin")
   add_subdirectory(DebugSymbols)