Google Git Google Git
Sign in
llvm/llvm-project/6b26457a0ef2994de54606eb5c356cb897cadda2/./lldb/source/Plugins/SymbolFile/NativePDB/SymbolFileNativePDB.cpp
blob: ca82d6ff36faa14524ab29366b5ca31733573180 [file]
//===-- SymbolFileNativePDB.cpp -------------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
#include "SymbolFileNativePDB.h"
#include "Plugins/ExpressionParser/Clang/ClangUtil.h"
#include "Plugins/Language/CPlusPlus/MSVCUndecoratedNameParser.h"
#include "Plugins/ObjectFile/PDB/ObjectFilePDB.h"
#include "Plugins/SymbolFile/PDB/SymbolFilePDB.h"
#include "Plugins/TypeSystem/Clang/TypeSystemClang.h"
#include "lldb/Core/Module.h"
#include "lldb/Core/PluginManager.h"
#include "lldb/Symbol/CompileUnit.h"
#include "lldb/Symbol/LineTable.h"
#include "lldb/Symbol/ObjectFile.h"
#include "lldb/Symbol/SymbolContext.h"
#include "lldb/Symbol/SymbolVendor.h"
#include "lldb/Symbol/Variable.h"
#include "lldb/Symbol/VariableList.h"
#include "lldb/Utility/LLDBLog.h"
#include "lldb/Utility/Log.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/DebugInfo/CodeView/CVRecord.h"
#include "llvm/DebugInfo/CodeView/CVTypeVisitor.h"
#include "llvm/DebugInfo/CodeView/DebugLinesSubsection.h"
#include "llvm/DebugInfo/CodeView/Formatters.h"
#include "llvm/DebugInfo/CodeView/LazyRandomTypeCollection.h"
#include "llvm/DebugInfo/CodeView/RecordName.h"
#include "llvm/DebugInfo/CodeView/SymbolDeserializer.h"
#include "llvm/DebugInfo/CodeView/SymbolRecordHelpers.h"
#include "llvm/DebugInfo/CodeView/TypeDeserializer.h"
#include "llvm/DebugInfo/PDB/Native/DbiStream.h"
#include "llvm/DebugInfo/PDB/Native/GlobalsStream.h"
#include "llvm/DebugInfo/PDB/Native/InfoStream.h"
#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"
#include "llvm/DebugInfo/PDB/PDBTypes.h"
#include "llvm/Demangle/MicrosoftDemangle.h"
#include "llvm/Object/COFF.h"
#include "llvm/Support/Allocator.h"
#include "llvm/Support/BinaryStreamReader.h"
#include "llvm/Support/Error.h"
#include "llvm/Support/ErrorOr.h"
#include "llvm/Support/MemoryBuffer.h"
#include "DWARFLocationExpression.h"
#include "PdbSymUid.h"
#include "PdbUtil.h"
#include "UdtRecordCompleter.h"
#include <optional>
#include <string_view>
using namespace lldb;
using namespace lldb_private;
using namespace npdb;
using namespace llvm::codeview;
using namespace llvm::pdb;
char SymbolFileNativePDB::ID;
static lldb::LanguageType TranslateLanguage(PDB_Lang lang) {
switch (lang) {
case PDB_Lang::Cpp:
return lldb::LanguageType::eLanguageTypeC_plus_plus;
case PDB_Lang::C:
return lldb::LanguageType::eLanguageTypeC;
case PDB_Lang::Swift:
return lldb::LanguageType::eLanguageTypeSwift;
case PDB_Lang::Rust:
return lldb::LanguageType::eLanguageTypeRust;
case PDB_Lang::ObjC:
return lldb::LanguageType::eLanguageTypeObjC;
case PDB_Lang::ObjCpp:
return lldb::LanguageType::eLanguageTypeObjC_plus_plus;
default:
return lldb::LanguageType::eLanguageTypeUnknown;
}
}
static std::optional<std::string>
findMatchingPDBFilePath(llvm::StringRef original_pdb_path,
llvm::StringRef exe_path) {
const FileSystem &fs = FileSystem::Instance();
if (fs.Exists(original_pdb_path))
return std::string(original_pdb_path);
const auto exe_dir = FileSpec(exe_path).CopyByRemovingLastPathComponent();
// While the exe_path uses the native style, the exe might be compiled on a
// different OS, so try to guess the style used.
const FileSpec original_pdb_spec(original_pdb_path,
FileSpec::GuessPathStyle(original_pdb_path)
.value_or(FileSpec::Style::native));
const llvm::StringRef pdb_filename = original_pdb_spec.GetFilename();
// If the file doesn't exist, perhaps the path specified at build time
// doesn't match the PDB's current location, so check the location of the
// executable.
const FileSpec local_pdb = exe_dir.CopyByAppendingPathComponent(pdb_filename);
if (fs.Exists(local_pdb))
return local_pdb.GetPath();
// Otherwise, search for one in target.debug-file-search-paths
FileSpecList search_paths = Target::GetDefaultDebugFileSearchPaths();
for (const FileSpec &search_dir : search_paths) {
FileSpec pdb_path = search_dir.CopyByAppendingPathComponent(pdb_filename);
if (fs.Exists(pdb_path))
return pdb_path.GetPath();
}
return std::nullopt;
}
static std::unique_ptr<PDBFile>
loadMatchingPDBFile(std::string exe_path, llvm::BumpPtrAllocator &allocator) {
// Try to find a matching PDB for an EXE.
using namespace llvm::object;
auto expected_binary = createBinary(exe_path);
// If the file isn't a PE/COFF executable, fail.
if (!expected_binary) {
llvm::consumeError(expected_binary.takeError());
return nullptr;
}
OwningBinary<Binary> binary = std::move(*expected_binary);
// TODO: Avoid opening the PE/COFF binary twice by reading this information
// directly from the lldb_private::ObjectFile.
auto *obj = llvm::dyn_cast<llvm::object::COFFObjectFile>(binary.getBinary());
if (!obj)
return nullptr;
const llvm::codeview::DebugInfo *pdb_info = nullptr;
// If it doesn't have a debug directory, fail.
llvm::StringRef pdb_file;
if (llvm::Error e = obj->getDebugPDBInfo(pdb_info, pdb_file)) {
consumeError(std::move(e));
return nullptr;
}
std::optional<std::string> resolved_pdb_path =
findMatchingPDBFilePath(pdb_file, exe_path);
if (!resolved_pdb_path)
return nullptr;
// If the file is not a PDB or if it doesn't have a matching GUID, fail.
auto pdb =
ObjectFilePDB::loadPDBFile(*std::move(resolved_pdb_path), allocator);
if (!pdb)
return nullptr;
auto expected_info = pdb->getPDBInfoStream();
if (!expected_info) {
llvm::consumeError(expected_info.takeError());
return nullptr;
}
llvm::codeview::GUID guid;
memcpy(&guid, pdb_info->PDB70.Signature, 16);
if (expected_info->getGuid() != guid)
return nullptr;
return pdb;
}
static bool IsFunctionPrologue(const CompilandIndexItem &cci,
lldb::addr_t addr) {
// FIXME: Implement this.
return false;
}
static bool IsFunctionEpilogue(const CompilandIndexItem &cci,
lldb::addr_t addr) {
// FIXME: Implement this.
return false;
}
// See llvm::codeview::TypeIndex::simpleTypeName as well as strForPrimitiveTi
// from the original pdbdump:
// https://github.com/microsoft/microsoft-pdb/blob/805655a28bd8198004be2ac27e6e0290121a5e89/pdbdump/pdbdump.cpp#L1896-L1974
//
// For 64bit integers we use "long long" like DIA instead of "__int64".
static llvm::StringRef GetSimpleTypeName(SimpleTypeKind kind) {
switch (kind) {
case SimpleTypeKind::Boolean128:
return "__bool128";
case SimpleTypeKind::Boolean64:
return "__bool64";
case SimpleTypeKind::Boolean32:
return "__bool32";
case SimpleTypeKind::Boolean16:
return "__bool16";
case SimpleTypeKind::Boolean8:
return "bool";
case SimpleTypeKind::Byte:
case SimpleTypeKind::UnsignedCharacter:
return "unsigned char";
case SimpleTypeKind::NarrowCharacter:
return "char";
case SimpleTypeKind::SignedCharacter:
case SimpleTypeKind::SByte:
return "signed char";
case SimpleTypeKind::Character32:
return "char32_t";
case SimpleTypeKind::Character16:
return "char16_t";
case SimpleTypeKind::Character8:
return "char8_t";
case SimpleTypeKind::Complex128:
return "_Complex __float128";
case SimpleTypeKind::Complex80:
return "_Complex long double";
case SimpleTypeKind::Complex64:
return "_Complex double";
case SimpleTypeKind::Complex48:
return "_Complex __float48";
case SimpleTypeKind::Complex32:
case SimpleTypeKind::Complex32PartialPrecision:
return "_Complex float";
case SimpleTypeKind::Complex16:
return "_Complex _Float16";
case SimpleTypeKind::Float128:
return "__float128";
case SimpleTypeKind::Float80:
return "long double";
case SimpleTypeKind::Float64:
return "double";
case SimpleTypeKind::Float48:
return "__float48";
case SimpleTypeKind::Float32:
case SimpleTypeKind::Float32PartialPrecision:
return "float";
case SimpleTypeKind::Float16:
return "_Float16";
case SimpleTypeKind::Int128Oct:
case SimpleTypeKind::Int128:
return "__int128";
case SimpleTypeKind::Int64:
case SimpleTypeKind::Int64Quad:
return "long long";
case SimpleTypeKind::Int32Long:
return "long";
case SimpleTypeKind::Int32:
return "int";
case SimpleTypeKind::Int16:
case SimpleTypeKind::Int16Short:
return "short";
case SimpleTypeKind::UInt128Oct:
case SimpleTypeKind::UInt128:
return "unsigned __int128";
case SimpleTypeKind::UInt64:
case SimpleTypeKind::UInt64Quad:
return "unsigned long long";
case SimpleTypeKind::UInt32:
return "unsigned";
case SimpleTypeKind::UInt16:
case SimpleTypeKind::UInt16Short:
return "unsigned short";
case SimpleTypeKind::UInt32Long:
return "unsigned long";
case SimpleTypeKind::HResult:
return "HRESULT";
case SimpleTypeKind::Void:
return "void";
case SimpleTypeKind::WideCharacter:
return "wchar_t";
case SimpleTypeKind::None:
case SimpleTypeKind::NotTranslated:
return "";
}
return "";
}
static bool IsClassRecord(TypeLeafKind kind) {
switch (kind) {
case LF_STRUCTURE:
case LF_CLASS:
case LF_INTERFACE:
return true;
default:
return false;
}
}
static std::optional<CVTagRecord>
GetNestedTagDefinition(const NestedTypeRecord &Record,
const CVTagRecord &parent, TpiStream &tpi) {
// An LF_NESTTYPE is essentially a nested typedef / using declaration, but it
// is also used to indicate the primary definition of a nested class. That is
// to say, if you have:
// struct A {
// struct B {};
// using C = B;
// };
// Then in the debug info, this will appear as:
// LF_STRUCTURE `A::B` [type index = N]
// LF_STRUCTURE `A`
// LF_NESTTYPE [name = `B`, index = N]
// LF_NESTTYPE [name = `C`, index = N]
// In order to accurately reconstruct the decl context hierarchy, we need to
// know which ones are actual definitions and which ones are just aliases.
// If it's a simple type, then this is something like `using foo = int`.
if (Record.Type.isSimple())
return std::nullopt;
CVType cvt = tpi.getType(Record.Type);
if (!IsTagRecord(cvt))
return std::nullopt;
// If it's an inner definition, then treat whatever name we have here as a
// single component of a mangled name. So we can inject it into the parent's
// mangled name to see if it matches.
CVTagRecord child = CVTagRecord::create(cvt);
std::string qname = std::string(parent.asTag().getUniqueName());
if (qname.size() < 4 || child.asTag().getUniqueName().size() < 4)
return std::nullopt;
// qname[3] is the tag type identifier (struct, class, union, etc). Since the
// inner tag type is not necessarily the same as the outer tag type, re-write
// it to match the inner tag type.
qname[3] = child.asTag().getUniqueName()[3];
std::string piece;
if (qname[3] == 'W')
piece = "4";
piece += Record.Name;
piece.push_back('@');
qname.insert(4, std::move(piece));
if (qname != child.asTag().UniqueName)
return std::nullopt;
return std::move(child);
}
void SymbolFileNativePDB::Initialize() {
PluginManager::RegisterPlugin(GetPluginNameStatic(),
GetPluginDescriptionStatic(), CreateInstance,
DebuggerInitialize);
}
void SymbolFileNativePDB::Terminate() {
PluginManager::UnregisterPlugin(CreateInstance);
}
void SymbolFileNativePDB::DebuggerInitialize(Debugger &debugger) {}
llvm::StringRef SymbolFileNativePDB::GetPluginDescriptionStatic() {
return "Microsoft PDB debug symbol cross-platform file reader.";
}
SymbolFile *SymbolFileNativePDB::CreateInstance(ObjectFileSP objfile_sp) {
if (!SymbolFilePDB::UseNativePDB())
return nullptr;
return new SymbolFileNativePDB(std::move(objfile_sp));
}
SymbolFileNativePDB::SymbolFileNativePDB(ObjectFileSP objfile_sp)
: SymbolFileCommon(std::move(objfile_sp)) {}
SymbolFileNativePDB::~SymbolFileNativePDB() = default;
uint32_t SymbolFileNativePDB::CalculateAbilities() {
uint32_t abilities = 0;
if (!m_objfile_sp)
return 0;
if (!m_index) {
// Lazily load and match the PDB file, but only do this once.
PDBFile *pdb_file;
if (auto *pdb = llvm::dyn_cast<ObjectFilePDB>(m_objfile_sp.get())) {
pdb_file = &pdb->GetPDBFile();
} else {
m_file_up = loadMatchingPDBFile(m_objfile_sp->GetFileSpec().GetPath(),
m_allocator);
pdb_file = m_file_up.get();
}
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());
return 0;
}
m_index = std::move(*expected_index);
}
if (!m_index)
return 0;
// We don't especially have to be precise here. We only distinguish between
// stripped and not stripped.
abilities = kAllAbilities;
if (m_index->dbi().isStripped())
abilities &= ~(Blocks | LocalVariables);
return abilities;
}
void SymbolFileNativePDB::InitializeObject() {
m_obj_load_address = m_objfile_sp->GetModule()
->GetObjectFile()
->GetBaseAddress()
.GetFileAddress();
m_index->SetLoadAddress(m_obj_load_address);
m_index->ParseSectionContribs();
auto ts_or_err = m_objfile_sp->GetModule()->GetTypeSystemForLanguage(
lldb::eLanguageTypeC_plus_plus);
if (auto err = ts_or_err.takeError()) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to initialize: {0}");
} else {
if (auto ts = *ts_or_err)
ts->SetSymbolFile(this);
}
}
uint32_t SymbolFileNativePDB::CalculateNumCompileUnits() {
const DbiModuleList &modules = m_index->dbi().modules();
uint32_t count = modules.getModuleCount();
if (count == 0)
return count;
// The linker can inject an additional "dummy" compilation unit into the
// PDB. Ignore this special compile unit for our purposes, if it is there.
// It is always the last one.
DbiModuleDescriptor last = modules.getModuleDescriptor(count - 1);
if (last.getModuleName() == "* Linker *")
--count;
return count;
}
Block *SymbolFileNativePDB::CreateBlock(PdbCompilandSymId block_id) {
CompilandIndexItem *cii = m_index->compilands().GetCompiland(block_id.modi);
CVSymbol sym = cii->m_debug_stream.readSymbolAtOffset(block_id.offset);
CompUnitSP comp_unit = GetOrCreateCompileUnit(*cii);
lldb::user_id_t opaque_block_uid = toOpaqueUid(block_id);
auto ts_or_err = GetTypeSystemForLanguage(comp_unit->GetLanguage());
if (auto err = ts_or_err.takeError())
return nullptr;
auto ts = *ts_or_err;
if (!ts)
return nullptr;
PdbAstBuilder* ast_builder = ts->GetNativePDBParser();
switch (sym.kind()) {
case S_GPROC32:
case S_LPROC32:
// This is a function. It must be global. Creating the Function entry
// for it automatically creates a block for it.
if (FunctionSP func = GetOrCreateFunction(block_id, *comp_unit))
return &func->GetBlock(false);
break;
case S_BLOCK32: {
// This is a block. Its parent is either a function or another block. In
// either case, its parent can be viewed as a block (e.g. a function
// contains 1 big block. So just get the parent block and add this block
// to it.
BlockSym block(static_cast<SymbolRecordKind>(sym.kind()));
if (auto err = SymbolDeserializer::deserializeAs<BlockSym>(sym, block)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize BlockSym record: {0}");
return nullptr;
}
if (block.Parent == 0) {
LLDB_LOG(GetLog(LLDBLog::Symbols), "BlockSym record ({0}) with parent=0",
block_id);
return nullptr;
}
PdbCompilandSymId parent_id(block_id.modi, block.Parent);
Block *parent_block = GetOrCreateBlock(parent_id);
if (!parent_block)
return nullptr;
Function *func = parent_block->CalculateSymbolContextFunction();
if (!func) {
LLDB_LOG(GetLog(LLDBLog::Symbols), "parent of {0} is not a function",
parent_id);
return nullptr;
}
lldb::addr_t block_base =
m_index->MakeVirtualAddress(block.Segment, block.CodeOffset);
lldb::addr_t func_base = func->GetAddress().GetFileAddress();
BlockSP child_block = parent_block->CreateChild(opaque_block_uid);
if (block_base >= func_base)
child_block->AddRange(Block::Range(block_base - func_base, block.CodeSize));
else {
GetObjectFile()->GetModule()->ReportError(
"S_BLOCK32 at modi: {0:d} offset: {1:d}: adding range "
"[{2:x16}-{3:x16}) which has a base that is less than the "
"function's "
"low PC 0x%" PRIx64 ". Please file a bug and attach the file at the "
"start of this error message",
block_id.modi, block_id.offset, block_base,
block_base + block.CodeSize, func_base);
}
if (ast_builder)
ast_builder->EnsureBlock(block_id);
m_blocks.insert({opaque_block_uid, child_block});
break;
}
case S_INLINESITE: {
// This ensures line table is parsed first so we have inline sites info.
comp_unit->GetLineTable();
std::shared_ptr<InlineSite> inline_site = m_inline_sites[opaque_block_uid];
Block *parent_block = GetOrCreateBlock(inline_site->parent_id);
if (!parent_block)
return nullptr;
BlockSP child_block = parent_block->CreateChild(opaque_block_uid);
if (ast_builder)
ast_builder->EnsureInlinedFunction(block_id);
// Copy ranges from InlineSite to Block.
for (size_t i = 0; i < inline_site->ranges.GetSize(); ++i) {
auto *entry = inline_site->ranges.GetEntryAtIndex(i);
child_block->AddRange(
Block::Range(entry->GetRangeBase(), entry->GetByteSize()));
}
child_block->FinalizeRanges();
// Get the inlined function callsite info.
Declaration &decl = inline_site->inline_function_info->GetDeclaration();
Declaration &callsite = inline_site->inline_function_info->GetCallSite();
child_block->SetInlinedFunctionInfo(
inline_site->inline_function_info->GetName().GetCString(), nullptr,
&decl, &callsite);
m_blocks.insert({opaque_block_uid, child_block});
break;
}
default:
LLDB_LOG(GetLog(LLDBLog::Symbols), "{0} is not a block", block_id);
return nullptr;
}
return nullptr;
}
lldb::FunctionSP SymbolFileNativePDB::CreateFunction(PdbCompilandSymId func_id,
CompileUnit &comp_unit) {
const CompilandIndexItem *cci =
m_index->compilands().GetCompiland(func_id.modi);
if (!cci) {
LLDB_LOG(GetLog(LLDBLog::Symbols), "missing compiland {0}", func_id.modi);
return nullptr;
}
CVSymbol sym_record = cci->m_debug_stream.readSymbolAtOffset(func_id.offset);
if (sym_record.kind() != S_LPROC32 && sym_record.kind() != S_GPROC32) {
LLDB_LOG(GetLog(LLDBLog::Symbols), "{0} is not a function", func_id);
return nullptr;
}
SegmentOffsetLength sol = GetSegmentOffsetAndLength(sym_record);
auto file_vm_addr =
m_index->MakeVirtualAddress(sol.so.segment, sol.so.offset);
if (file_vm_addr == LLDB_INVALID_ADDRESS || file_vm_addr == 0)
return nullptr;
Address func_addr(file_vm_addr, comp_unit.GetModule()->GetSectionList());
if (!func_addr.IsValid())
return nullptr;
ProcSym proc(static_cast<SymbolRecordKind>(sym_record.kind()));
if (auto err = SymbolDeserializer::deserializeAs<ProcSym>(sym_record, proc)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ProcSym record: {0}");
return nullptr;
}
if (proc.FunctionType == TypeIndex::None())
return nullptr;
TypeSP func_type = GetOrCreateType(proc.FunctionType);
if (!func_type)
return nullptr;
PdbTypeSymId sig_id(proc.FunctionType, false);
std::optional<llvm::StringRef> mangled_opt = FindMangledSymbol(
SegmentOffset(proc.Segment, proc.CodeOffset), proc.FunctionType);
Mangled mangled(mangled_opt.value_or(proc.Name));
FunctionSP func_sp = std::make_shared<Function>(
&comp_unit, toOpaqueUid(func_id), toOpaqueUid(sig_id), mangled,
func_type.get(), func_addr,
AddressRanges{AddressRange(func_addr, sol.length)});
comp_unit.AddFunction(func_sp);
auto ts_or_err = GetTypeSystemForLanguage(comp_unit.GetLanguage());
if (auto err = ts_or_err.takeError())
return func_sp;
auto ts = *ts_or_err;
if (ts) {
if (PdbAstBuilder *ast_builder = ts->GetNativePDBParser())
ast_builder->EnsureFunction(func_id);
}
return func_sp;
}
CompUnitSP
SymbolFileNativePDB::CreateCompileUnit(const CompilandIndexItem &cci) {
lldb::LanguageType lang =
cci.m_compile_opts ? TranslateLanguage(cci.m_compile_opts->getLanguage())
: lldb::eLanguageTypeUnknown;
LazyBool optimized = eLazyBoolNo;
if (cci.m_compile_opts && cci.m_compile_opts->hasOptimizations())
optimized = eLazyBoolYes;
llvm::SmallString<64> source_file_name;
if (auto main_file_or_err = m_index->compilands().GetMainSourceFile(cci)) {
source_file_name = std::move(*main_file_or_err);
} else {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), main_file_or_err.takeError(),
"Failed to determine main source file: {0}");
}
FileSpec fs(llvm::sys::path::convert_to_slash(
source_file_name, llvm::sys::path::Style::windows_backslash));
CompUnitSP cu_sp = std::make_shared<CompileUnit>(
m_objfile_sp->GetModule(), nullptr, std::make_shared<SupportFile>(fs),
toOpaqueUid(cci.m_id), lang, optimized);
SetCompileUnitAtIndex(cci.m_id.modi, cu_sp);
return cu_sp;
}
lldb::TypeSP SymbolFileNativePDB::CreateModifierType(PdbTypeSymId type_id,
const ModifierRecord &mr,
CompilerType ct) {
TpiStream &stream = m_index->tpi();
std::string name;
if ((mr.Modifiers & ModifierOptions::Const) != ModifierOptions::None)
name += "const ";
if ((mr.Modifiers & ModifierOptions::Volatile) != ModifierOptions::None)
name += "volatile ";
if ((mr.Modifiers & ModifierOptions::Unaligned) != ModifierOptions::None)
name += "__unaligned ";
if (mr.ModifiedType.isSimple())
name += GetSimpleTypeName(mr.ModifiedType.getSimpleKind());
else
name += computeTypeName(stream.typeCollection(), mr.ModifiedType);
Declaration decl;
lldb::TypeSP modified_type = GetOrCreateType(mr.ModifiedType);
return MakeType(toOpaqueUid(type_id), ConstString(name),
llvm::expectedToOptional(modified_type->GetByteSize(nullptr)),
nullptr, LLDB_INVALID_UID, Type::eEncodingIsUID, decl, ct,
Type::ResolveState::Full);
}
lldb::TypeSP
SymbolFileNativePDB::CreatePointerType(PdbTypeSymId type_id,
const llvm::codeview::PointerRecord &pr,
CompilerType ct) {
TypeSP pointee = GetOrCreateType(pr.ReferentType);
if (!pointee)
return nullptr;
if (pr.isPointerToMember()) {
MemberPointerInfo mpi = pr.getMemberInfo();
GetOrCreateType(mpi.ContainingType);
}
Declaration decl;
return MakeType(toOpaqueUid(type_id), ConstString(), pr.getSize(), nullptr,
LLDB_INVALID_UID, Type::eEncodingIsUID, decl, ct,
Type::ResolveState::Full);
}
lldb::TypeSP SymbolFileNativePDB::CreateSimpleType(TypeIndex ti,
CompilerType ct) {
uint64_t uid = toOpaqueUid(PdbTypeSymId(ti, false));
if (ti == TypeIndex::NullptrT()) {
Declaration decl;
return MakeType(uid, ConstString("decltype(nullptr)"), std::nullopt,
nullptr, LLDB_INVALID_UID, Type::eEncodingIsUID, decl, ct,
Type::ResolveState::Full);
}
if (ti.getSimpleMode() != SimpleTypeMode::Direct) {
TypeSP direct_sp = GetOrCreateType(ti.makeDirect());
uint32_t pointer_size = 0;
switch (ti.getSimpleMode()) {
case SimpleTypeMode::FarPointer32:
case SimpleTypeMode::NearPointer32:
pointer_size = 4;
break;
case SimpleTypeMode::NearPointer64:
pointer_size = 8;
break;
default:
// 128-bit and 16-bit pointers unsupported.
return nullptr;
}
Declaration decl;
return MakeType(uid, ConstString(), pointer_size, nullptr, LLDB_INVALID_UID,
Type::eEncodingIsUID, decl, ct, Type::ResolveState::Full);
}
if (ti.getSimpleKind() == SimpleTypeKind::NotTranslated)
return nullptr;
size_t size = GetTypeSizeForSimpleKind(ti.getSimpleKind());
llvm::StringRef type_name = GetSimpleTypeName(ti.getSimpleKind());
Declaration decl;
return MakeType(uid, ConstString(type_name), size, nullptr, LLDB_INVALID_UID,
Type::eEncodingIsUID, decl, ct, Type::ResolveState::Full);
}
static std::string GetUnqualifiedTypeName(const TagRecord &record) {
if (!record.hasUniqueName())
return std::string(MSVCUndecoratedNameParser::DropScope(record.Name));
llvm::ms_demangle::Demangler demangler;
std::string_view sv(record.UniqueName.begin(), record.UniqueName.size());
llvm::ms_demangle::TagTypeNode *ttn = demangler.parseTagUniqueName(sv);
if (demangler.Error)
return std::string(MSVCUndecoratedNameParser::DropScope(record.Name));
llvm::ms_demangle::IdentifierNode *idn =
ttn->QualifiedName->getUnqualifiedIdentifier();
return idn->toString();
}
lldb::TypeSP
SymbolFileNativePDB::CreateClassStructUnion(PdbTypeSymId type_id,
const TagRecord &record,
size_t size, CompilerType ct) {
std::string uname = GetUnqualifiedTypeName(record);
llvm::Expected<Declaration> maybeDecl = ResolveUdtDeclaration(type_id);
Declaration decl;
if (maybeDecl)
decl = std::move(*maybeDecl);
else
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), maybeDecl.takeError(),
"Failed to resolve declaration for '{1}': {0}", uname);
return MakeType(toOpaqueUid(type_id), ConstString(uname), size, nullptr,
LLDB_INVALID_UID, Type::eEncodingIsUID, decl, ct,
Type::ResolveState::Forward);
}
lldb::TypeSP SymbolFileNativePDB::CreateTagType(PdbTypeSymId type_id,
const ClassRecord &cr,
CompilerType ct) {
return CreateClassStructUnion(type_id, cr, cr.getSize(), ct);
}
lldb::TypeSP SymbolFileNativePDB::CreateTagType(PdbTypeSymId type_id,
const UnionRecord &ur,
CompilerType ct) {
return CreateClassStructUnion(type_id, ur, ur.getSize(), ct);
}
lldb::TypeSP SymbolFileNativePDB::CreateTagType(PdbTypeSymId type_id,
const EnumRecord &er,
CompilerType ct) {
std::string uname = GetUnqualifiedTypeName(er);
llvm::Expected<Declaration> maybeDecl = ResolveUdtDeclaration(type_id);
Declaration decl;
if (maybeDecl)
decl = std::move(*maybeDecl);
else
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), maybeDecl.takeError(),
"Failed to resolve declaration for '{1}': {0}", uname);
TypeSP underlying_type = GetOrCreateType(er.UnderlyingType);
return MakeType(
toOpaqueUid(type_id), ConstString(uname),
llvm::expectedToOptional(underlying_type->GetByteSize(nullptr)), nullptr,
LLDB_INVALID_UID, lldb_private::Type::eEncodingIsUID, decl, ct,
lldb_private::Type::ResolveState::Forward);
}
TypeSP SymbolFileNativePDB::CreateArrayType(PdbTypeSymId type_id,
const ArrayRecord &ar,
CompilerType ct) {
TypeSP element_type = GetOrCreateType(ar.ElementType);
Declaration decl;
TypeSP array_sp =
MakeType(toOpaqueUid(type_id), ConstString(), ar.Size, nullptr,
LLDB_INVALID_UID, lldb_private::Type::eEncodingIsUID, decl, ct,
lldb_private::Type::ResolveState::Full);
array_sp->SetEncodingType(element_type.get());
return array_sp;
}
TypeSP SymbolFileNativePDB::CreateFunctionType(PdbTypeSymId type_id,
const MemberFunctionRecord &mfr,
CompilerType ct) {
if (mfr.ReturnType.isSimple())
GetOrCreateType(mfr.ReturnType);
CreateSimpleArgumentListTypes(mfr.ArgumentList);
Declaration decl;
return MakeType(toOpaqueUid(type_id), ConstString(), 0, nullptr,
LLDB_INVALID_UID, lldb_private::Type::eEncodingIsUID, decl,
ct, lldb_private::Type::ResolveState::Full);
}
TypeSP SymbolFileNativePDB::CreateProcedureType(PdbTypeSymId type_id,
const ProcedureRecord &pr,
CompilerType ct) {
if (pr.ReturnType.isSimple())
GetOrCreateType(pr.ReturnType);
CreateSimpleArgumentListTypes(pr.ArgumentList);
Declaration decl;
return MakeType(toOpaqueUid(type_id), ConstString(), 0, nullptr,
LLDB_INVALID_UID, lldb_private::Type::eEncodingIsUID, decl,
ct, lldb_private::Type::ResolveState::Full);
}
void SymbolFileNativePDB::CreateSimpleArgumentListTypes(
llvm::codeview::TypeIndex arglist_ti) {
if (arglist_ti.isNoneType())
return;
CVType arglist_cvt = m_index->tpi().getType(arglist_ti);
if (arglist_cvt.kind() != LF_ARGLIST)
return; // invalid debug info
ArgListRecord alr;
if (auto err =
TypeDeserializer::deserializeAs<ArgListRecord>(arglist_cvt, alr)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ArgListRecord record ({1}): {0}",
arglist_ti);
return;
}
for (TypeIndex id : alr.getIndices())
if (!id.isNoneType() && id.isSimple())
GetOrCreateType(id);
}
TypeSP SymbolFileNativePDB::CreateType(PdbTypeSymId type_id, CompilerType ct) {
if (type_id.index.isSimple())
return CreateSimpleType(type_id.index, ct);
TpiStream &stream = type_id.is_ipi ? m_index->ipi() : m_index->tpi();
CVType cvt = stream.getType(type_id.index);
if (cvt.kind() == LF_MODIFIER) {
ModifierRecord modifier;
if (auto err =
TypeDeserializer::deserializeAs<ModifierRecord>(cvt, modifier)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ModifierRecord record ({1}): {0}",
type_id.index);
return nullptr;
}
return CreateModifierType(type_id, modifier, ct);
}
if (cvt.kind() == LF_POINTER) {
PointerRecord pointer;
if (auto err =
TypeDeserializer::deserializeAs<PointerRecord>(cvt, pointer)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize PointerRecord record ({1}): {0}",
type_id.index);
return nullptr;
}
return CreatePointerType(type_id, pointer, ct);
}
if (IsClassRecord(cvt.kind())) {
ClassRecord cr;
if (auto err = TypeDeserializer::deserializeAs<ClassRecord>(cvt, cr)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ClassRecord record ({1}): {0}",
type_id.index);
return nullptr;
}
return CreateTagType(type_id, cr, ct);
}
if (cvt.kind() == LF_ENUM) {
EnumRecord er;
if (auto err = TypeDeserializer::deserializeAs<EnumRecord>(cvt, er)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize EnumRecord record ({1}): {0}",
type_id.index);
return nullptr;
}
return CreateTagType(type_id, er, ct);
}
if (cvt.kind() == LF_UNION) {
UnionRecord ur;
if (auto err = TypeDeserializer::deserializeAs<UnionRecord>(cvt, ur)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize UnionRecord record ({1}): {0}",
type_id.index);
return nullptr;
}
return CreateTagType(type_id, ur, ct);
}
if (cvt.kind() == LF_ARRAY) {
ArrayRecord ar;
if (auto err = TypeDeserializer::deserializeAs<ArrayRecord>(cvt, ar)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ArrayRecord record ({1}): {0}",
type_id.index);
return nullptr;
}
return CreateArrayType(type_id, ar, ct);
}
if (cvt.kind() == LF_PROCEDURE) {
ProcedureRecord pr;
if (auto err = TypeDeserializer::deserializeAs<ProcedureRecord>(cvt, pr)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ProcedureRecord record ({1}): {0}",
type_id.index);
return nullptr;
}
return CreateProcedureType(type_id, pr, ct);
}
if (cvt.kind() == LF_MFUNCTION) {
MemberFunctionRecord mfr;
if (auto err =
TypeDeserializer::deserializeAs<MemberFunctionRecord>(cvt, mfr)) {
LLDB_LOG_ERROR(
GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize MemberFunctionRecord record ({1}): {0}",
type_id.index);
return nullptr;
}
return CreateFunctionType(type_id, mfr, ct);
}
return nullptr;
}
TypeSP SymbolFileNativePDB::CreateAndCacheType(PdbTypeSymId type_id) {
// If they search for a UDT which is a forward ref, try and resolve the full
// decl and just map the forward ref uid to the full decl record.
std::optional<PdbTypeSymId> full_decl_uid;
if (IsForwardRefUdt(type_id, m_index->tpi())) {
auto expected_full_ti =
m_index->tpi().findFullDeclForForwardRef(type_id.index);
if (!expected_full_ti)
llvm::consumeError(expected_full_ti.takeError());
else if (*expected_full_ti != type_id.index) {
full_decl_uid = PdbTypeSymId(*expected_full_ti, false);
// It's possible that a lookup would occur for the full decl causing it
// to be cached, then a second lookup would occur for the forward decl.
// We don't want to create a second full decl, so make sure the full
// decl hasn't already been cached.
auto full_iter = m_types.find(toOpaqueUid(*full_decl_uid));
if (full_iter != m_types.end()) {
TypeSP result = full_iter->second;
// Map the forward decl to the TypeSP for the full decl so we can take
// the fast path next time.
m_types[toOpaqueUid(type_id)] = result;
return result;
}
}
}
PdbTypeSymId best_decl_id = full_decl_uid ? *full_decl_uid : type_id;
auto ts_or_err = GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
if (auto err = ts_or_err.takeError())
return nullptr;
auto ts = *ts_or_err;
if (!ts)
return nullptr;
PdbAstBuilder *ast_builder = ts->GetNativePDBParser();
if (!ast_builder)
return nullptr;
CompilerType ct = ast_builder->GetOrCreateType(best_decl_id);
if (!ct)
return nullptr;
TypeSP result = CreateType(best_decl_id, ct);
if (!result)
return nullptr;
uint64_t best_uid = toOpaqueUid(best_decl_id);
m_types[best_uid] = result;
// If we had both a forward decl and a full decl, make both point to the new
// type.
if (full_decl_uid)
m_types[toOpaqueUid(type_id)] = result;
return result;
}
TypeSP SymbolFileNativePDB::GetOrCreateType(PdbTypeSymId type_id) {
// We can't use try_emplace / overwrite here because the process of creating
// a type could create nested types, which could invalidate iterators. So
// we have to do a 2-phase lookup / insert.
auto iter = m_types.find(toOpaqueUid(type_id));
if (iter != m_types.end())
return iter->second;
TypeSP type = CreateAndCacheType(type_id);
if (type)
GetTypeList().Insert(type);
return type;
}
VariableSP SymbolFileNativePDB::CreateGlobalVariable(PdbGlobalSymId var_id) {
CVSymbol sym = m_index->symrecords().readRecord(var_id.offset);
if (sym.kind() == S_CONSTANT)
return CreateConstantSymbol(var_id, sym);
lldb::ValueType scope = eValueTypeInvalid;
TypeIndex ti;
llvm::StringRef name;
lldb::addr_t addr = 0;
uint16_t section = 0;
uint32_t offset = 0;
bool is_external = false;
switch (sym.kind()) {
case S_GDATA32:
is_external = true;
[[fallthrough]];
case S_LDATA32: {
DataSym ds(sym.kind());
if (auto err = SymbolDeserializer::deserializeAs<DataSym>(sym, ds)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize DataSym record: {0}");
return nullptr;
}
ti = ds.Type;
scope = (sym.kind() == S_GDATA32) ? eValueTypeVariableGlobal
: eValueTypeVariableStatic;
name = ds.Name;
section = ds.Segment;
offset = ds.DataOffset;
addr = m_index->MakeVirtualAddress(ds.Segment, ds.DataOffset);
break;
}
case S_GTHREAD32:
is_external = true;
[[fallthrough]];
case S_LTHREAD32: {
ThreadLocalDataSym tlds(sym.kind());
if (auto err =
SymbolDeserializer::deserializeAs<ThreadLocalDataSym>(sym, tlds)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ThreadLocalDataSym record: {0}");
return nullptr;
}
ti = tlds.Type;
name = tlds.Name;
section = tlds.Segment;
offset = tlds.DataOffset;
addr = m_index->MakeVirtualAddress(tlds.Segment, tlds.DataOffset);
scope = eValueTypeVariableThreadLocal;
break;
}
default:
llvm_unreachable("unreachable!");
}
CompUnitSP comp_unit;
std::optional<uint16_t> modi = m_index->GetModuleIndexForVa(addr);
// Some globals has modi points to the linker module, ignore them.
if (!modi || modi >= GetNumCompileUnits())
return nullptr;
CompilandIndexItem &cci = m_index->compilands().GetOrCreateCompiland(*modi);
comp_unit = GetOrCreateCompileUnit(cci);
Declaration decl;
PdbTypeSymId tid(ti, false);
SymbolFileTypeSP type_sp =
std::make_shared<SymbolFileType>(*this, toOpaqueUid(tid));
Variable::RangeList ranges;
auto ts_or_err = GetTypeSystemForLanguage(comp_unit->GetLanguage());
if (auto err = ts_or_err.takeError())
return nullptr;
auto ts = *ts_or_err;
if (ts) {
if (PdbAstBuilder *ast_builder = ts->GetNativePDBParser())
ast_builder->EnsureVariable(var_id);
}
ModuleSP module_sp = GetObjectFile()->GetModule();
DWARFExpressionList location(
module_sp, MakeGlobalLocationExpression(section, offset, module_sp),
nullptr);
std::string global_name("::");
global_name += name;
bool artificial = false;
bool location_is_constant_data = false;
bool static_member = false;
VariableSP var_sp = std::make_shared<Variable>(
toOpaqueUid(var_id), name.str().c_str(), global_name.c_str(), type_sp,
scope, comp_unit.get(), ranges, &decl, location, is_external, artificial,
location_is_constant_data, static_member);
return var_sp;
}
lldb::VariableSP
SymbolFileNativePDB::CreateConstantSymbol(PdbGlobalSymId var_id,
const CVSymbol &cvs) {
TpiStream &tpi = m_index->tpi();
ConstantSym constant(cvs.kind());
if (cvs.kind() != S_CONSTANT)
return nullptr;
if (auto err =
SymbolDeserializer::deserializeAs<ConstantSym>(cvs, constant)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ConstantSym record: {0}");
return nullptr;
}
std::string global_name("::");
global_name += constant.Name;
PdbTypeSymId tid(constant.Type, false);
SymbolFileTypeSP type_sp =
std::make_shared<SymbolFileType>(*this, toOpaqueUid(tid));
Declaration decl;
Variable::RangeList ranges;
ModuleSP module = GetObjectFile()->GetModule();
auto location_or_err = MakeConstantLocationExpression(constant.Type, tpi,
constant.Value, module);
if (!location_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), location_or_err.takeError(),
"Failed to make constant location expression for {1}: {0}",
constant.Name);
return nullptr;
}
DWARFExpressionList location(module, std::move(*location_or_err), nullptr);
bool external = false;
bool artificial = false;
bool location_is_constant_data = true;
bool static_member = false;
VariableSP var_sp = std::make_shared<Variable>(
toOpaqueUid(var_id), constant.Name.str().c_str(), global_name.c_str(),
type_sp, eValueTypeVariableGlobal, module.get(), ranges, &decl, location,
external, artificial, location_is_constant_data, static_member);
return var_sp;
}
VariableSP
SymbolFileNativePDB::GetOrCreateGlobalVariable(PdbGlobalSymId var_id) {
auto emplace_result = m_global_vars.try_emplace(toOpaqueUid(var_id), nullptr);
if (emplace_result.second) {
if (VariableSP var_sp = CreateGlobalVariable(var_id))
emplace_result.first->second = var_sp;
else
return nullptr;
}
return emplace_result.first->second;
}
lldb::TypeSP SymbolFileNativePDB::GetOrCreateType(TypeIndex ti) {
return GetOrCreateType(PdbTypeSymId(ti, false));
}
FunctionSP SymbolFileNativePDB::GetOrCreateFunction(PdbCompilandSymId func_id,
CompileUnit &comp_unit) {
auto emplace_result = m_functions.try_emplace(toOpaqueUid(func_id), nullptr);
if (emplace_result.second)
emplace_result.first->second = CreateFunction(func_id, comp_unit);
return emplace_result.first->second;
}
CompUnitSP
SymbolFileNativePDB::GetOrCreateCompileUnit(const CompilandIndexItem &cci) {
auto emplace_result =
m_compilands.try_emplace(toOpaqueUid(cci.m_id), nullptr);
if (emplace_result.second) {
emplace_result.first->second = CreateCompileUnit(cci);
LLDB_LOG(GetLog(LLDBLog::Symbols), "failed to create compile unit for {0}",
cci.m_id.modi);
}
return emplace_result.first->second;
}
Block *SymbolFileNativePDB::GetOrCreateBlock(PdbCompilandSymId block_id) {
auto iter = m_blocks.find(toOpaqueUid(block_id));
if (iter != m_blocks.end())
return iter->second.get();
return CreateBlock(block_id);
}
void SymbolFileNativePDB::ParseDeclsForContext(
lldb_private::CompilerDeclContext decl_ctx) {
TypeSystem *ts = decl_ctx.GetTypeSystem();
if (!ts)
return;
PdbAstBuilder *ast_builder = ts->GetNativePDBParser();
if (!ast_builder)
return;
ast_builder->ParseDeclsForContext(decl_ctx);
}
lldb::CompUnitSP SymbolFileNativePDB::ParseCompileUnitAtIndex(uint32_t index) {
if (index >= GetNumCompileUnits())
return CompUnitSP();
assert(index < UINT16_MAX && "Invalid compile unit index");
if (index >= UINT16_MAX)
return nullptr;
CompilandIndexItem &item = m_index->compilands().GetOrCreateCompiland(index);
return GetOrCreateCompileUnit(item);
}
lldb::LanguageType SymbolFileNativePDB::ParseLanguage(CompileUnit &comp_unit) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
PdbSymUid uid(comp_unit.GetID());
if (uid.kind() != PdbSymUidKind::Compiland) {
assert(false && "uid of compile unit not a compiland");
return lldb::eLanguageTypeUnknown;
}
CompilandIndexItem *item =
m_index->compilands().GetCompiland(uid.asCompiland().modi);
assert(item);
if (!item || !item->m_compile_opts)
return lldb::eLanguageTypeUnknown;
return TranslateLanguage(item->m_compile_opts->getLanguage());
}
void SymbolFileNativePDB::AddSymbols(Symtab &symtab) {
auto *section_list =
m_objfile_sp->GetModule()->GetObjectFile()->GetSectionList();
if (!section_list)
return;
llvm::DenseMap<lldb::addr_t, llvm::SmallVector<uint32_t, 1>> symbols_by_addr;
for (uint32_t i = 0, n = symtab.GetNumSymbols(); i < n; ++i) {
Symbol *sym = symtab.SymbolAtIndex(i);
if (sym && sym->ValueIsAddress())
symbols_by_addr[sym->GetAddressRef().GetFileAddress()].push_back(i);
}
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) {
if (next->Offset < last_sym.Offset) {
LLDB_LOG(GetLog(LLDBLog::Symbols),
"Ignoring size estimate for '{0}': segment {1} offset {2} is "
"greater than the following offset {3}",
last_sym.Name, last_sym.Segment, last_sym.Offset,
next->Offset);
return;
}
last->SetByteSize(next->Offset - last_sym.Offset);
} else {
// the last symbol was the last in its section
if (section_sp->GetByteSize() < last_sym.Offset) {
LLDB_LOG(GetLog(LLDBLog::Symbols),
"Ignoring size estimate for '{0}': segment {1} offset {2} is "
"past the end of section '{3}' (size {4})",
last_sym.Name, last_sym.Segment, last_sym.Offset,
section_sp->GetName(), section_sp->GetByteSize());
return;
}
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;
auto pub_or_err = SymbolDeserializer::deserializeAs<PublicSym32>(sym);
if (!pub_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), pub_or_err.takeError(),
"Failed to deserialize PublicSym32 record: {0}");
continue;
}
PublicSym32 pub = std::move(*pub_or_err);
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;
Symbol pub_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);
auto it = symbols_by_addr.find(pub_symbol.GetAddressRef().GetFileAddress());
if (it != symbols_by_addr.end() &&
llvm::any_of(it->second, [&](uint32_t idx) {
Symbol *sym = symtab.SymbolAtIndex(idx);
return sym && sym->GetMangled() == pub_symbol.GetMangled();
}))
pub_symbol.SetType(lldb::eSymbolTypeAdditional);
last_sym_idx = symtab.AddSymbol(pub_symbol);
last_sym = pub;
}
finish_last_symbol(nullptr);
}
size_t SymbolFileNativePDB::ParseFunctions(CompileUnit &comp_unit) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
PdbSymUid uid{comp_unit.GetID()};
if (uid.kind() != PdbSymUidKind::Compiland) {
assert(false && "uid of compile unit not a compiland");
return 0;
}
uint16_t modi = uid.asCompiland().modi;
CompilandIndexItem &cii = m_index->compilands().GetOrCreateCompiland(modi);
size_t count = comp_unit.GetNumFunctions();
const CVSymbolArray &syms = cii.m_debug_stream.getSymbolArray();
for (auto iter = syms.begin(); iter != syms.end(); ++iter) {
if (iter->kind() != S_LPROC32 && iter->kind() != S_GPROC32)
continue;
PdbCompilandSymId sym_id{modi, iter.offset()};
FunctionSP func = GetOrCreateFunction(sym_id, comp_unit);
}
size_t new_count = comp_unit.GetNumFunctions();
if (new_count < count) {
assert(false && "less functions after parsing than before");
return 0;
}
return new_count - count;
}
static bool NeedsResolvedCompileUnit(uint32_t resolve_scope) {
// If any of these flags are set, we need to resolve the compile unit.
uint32_t flags = eSymbolContextCompUnit;
flags |= eSymbolContextVariable;
flags |= eSymbolContextFunction;
flags |= eSymbolContextBlock;
flags |= eSymbolContextLineEntry;
return (resolve_scope & flags) != 0;
}
uint32_t SymbolFileNativePDB::ResolveSymbolContext(
const Address &addr, SymbolContextItem resolve_scope, SymbolContext &sc) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
uint32_t resolved_flags = 0;
lldb::addr_t file_addr = addr.GetFileAddress();
if (NeedsResolvedCompileUnit(resolve_scope)) {
std::optional<uint16_t> modi = m_index->GetModuleIndexForVa(file_addr);
if (!modi)
return 0;
CompUnitSP cu_sp = GetCompileUnitAtIndex(*modi);
if (!cu_sp)
return 0;
sc.comp_unit = cu_sp.get();
resolved_flags |= eSymbolContextCompUnit;
}
if (resolve_scope & eSymbolContextFunction ||
resolve_scope & eSymbolContextBlock) {
if (!sc.comp_unit) {
LLDB_LOG(GetLog(LLDBLog::Symbols),
"missing compile unit for symbol at address {0:x}", file_addr);
return 0;
}
std::vector<SymbolAndUid> matches = m_index->FindSymbolsByVa(file_addr);
// Search the matches in reverse. This way if there are multiple matches
// (for example we are 3 levels deep in a nested scope) it will find the
// innermost one first.
for (const auto &match : llvm::reverse(matches)) {
if (match.uid.kind() != PdbSymUidKind::CompilandSym)
continue;
PdbCompilandSymId csid = match.uid.asCompilandSym();
CVSymbol cvs = m_index->ReadSymbolRecord(csid);
PDB_SymType type = CVSymToPDBSym(cvs.kind());
if (type != PDB_SymType::Function && type != PDB_SymType::Block)
continue;
if (type == PDB_SymType::Function) {
sc.function = GetOrCreateFunction(csid, *sc.comp_unit).get();
if (sc.function) {
Block &block = sc.function->GetBlock(true);
addr_t func_base = sc.function->GetAddress().GetFileAddress();
addr_t offset = file_addr - func_base;
sc.block = block.FindInnermostBlockByOffset(offset);
}
}
if (type == PDB_SymType::Block) {
Block *block = GetOrCreateBlock(csid);
if (!block)
continue;
sc.function = block->CalculateSymbolContextFunction();
if (sc.function) {
sc.function->GetBlock(true);
addr_t func_base = sc.function->GetAddress().GetFileAddress();
addr_t offset = file_addr - func_base;
sc.block = block->FindInnermostBlockByOffset(offset);
}
}
if (sc.function)
resolved_flags |= eSymbolContextFunction;
if (sc.block)
resolved_flags |= eSymbolContextBlock;
break;
}
}
if (resolve_scope & eSymbolContextLineEntry) {
if (!sc.comp_unit) {
LLDB_LOG(GetLog(LLDBLog::Symbols),
"missing compile unit for symbol at address {0:x}", file_addr);
return 0;
}
if (auto *line_table = sc.comp_unit->GetLineTable()) {
if (line_table->FindLineEntryByAddress(addr, sc.line_entry))
resolved_flags |= eSymbolContextLineEntry;
}
}
return resolved_flags;
}
uint32_t SymbolFileNativePDB::ResolveSymbolContext(
const SourceLocationSpec &src_location_spec,
lldb::SymbolContextItem resolve_scope, SymbolContextList &sc_list) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
const uint32_t prev_size = sc_list.GetSize();
if (resolve_scope & eSymbolContextCompUnit) {
for (uint32_t cu_idx = 0, num_cus = GetNumCompileUnits(); cu_idx < num_cus;
++cu_idx) {
CompileUnit *cu = ParseCompileUnitAtIndex(cu_idx).get();
if (!cu)
continue;
bool file_spec_matches_cu_file_spec = FileSpec::Match(
src_location_spec.GetFileSpec(), cu->GetPrimaryFile());
if (file_spec_matches_cu_file_spec) {
cu->ResolveSymbolContext(src_location_spec, resolve_scope, sc_list);
break;
}
}
}
return sc_list.GetSize() - prev_size;
}
bool SymbolFileNativePDB::ParseLineTable(CompileUnit &comp_unit) {
// Unfortunately LLDB is set up to parse the entire compile unit line table
// all at once, even if all it really needs is line info for a specific
// function. In the future it would be nice if it could set the sc.m_function
// member, and we could only get the line info for the function in question.
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
PdbSymUid cu_id(comp_unit.GetID());
if (cu_id.kind() != PdbSymUidKind::Compiland) {
assert(false && "uid of compile unit not a compiland");
return false;
}
uint16_t modi = cu_id.asCompiland().modi;
CompilandIndexItem *cii = m_index->compilands().GetCompiland(modi);
if (!cii) {
LLDB_LOG(GetLog(LLDBLog::Symbols), "missing compiland for modi={0}", modi);
return false;
}
// Parse DEBUG_S_LINES subsections first, then parse all S_INLINESITE records
// in this CU. Add line entries into the set first so that if there are line
// entries with same addres, the later is always more accurate than the
// former.
std::set<LineTable::Entry, LineTableEntryComparator> line_set;
// This is basically a copy of the .debug$S subsections from all original COFF
// object files merged together with address relocations applied. We are
// looking for all DEBUG_S_LINES subsections.
for (const DebugSubsectionRecord &dssr :
cii->m_debug_stream.getSubsectionsArray()) {
if (dssr.kind() != DebugSubsectionKind::Lines)
continue;
DebugLinesSubsectionRef lines;
llvm::BinaryStreamReader reader(dssr.getRecordData());
if (auto EC = lines.initialize(reader)) {
llvm::consumeError(std::move(EC));
return false;
}
const LineFragmentHeader *lfh = lines.header();
uint64_t virtual_addr =
m_index->MakeVirtualAddress(lfh->RelocSegment, lfh->RelocOffset);
if (virtual_addr == LLDB_INVALID_ADDRESS)
continue;
for (const LineColumnEntry &group : lines) {
llvm::Expected<uint32_t> file_index_or_err =
GetFileIndex(*cii, group.NameIndex);
if (!file_index_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), file_index_or_err.takeError(),
"failed to get file index for line entry: {0}");
continue;
}
uint32_t file_index = file_index_or_err.get();
if (group.LineNumbers.empty()) {
LLDB_LOG(GetLog(LLDBLog::Symbols),
"no line numbers for {0} in modi={1}", group.NameIndex, modi);
continue;
}
CompilandIndexItem::GlobalLineTable::Entry line_entry(
LLDB_INVALID_ADDRESS, 0);
for (const LineNumberEntry &entry : group.LineNumbers) {
LineInfo cur_info(entry.Flags);
if (cur_info.isAlwaysStepInto() || cur_info.isNeverStepInto())
continue;
uint64_t addr = virtual_addr + entry.Offset;
bool is_statement = cur_info.isStatement();
bool is_prologue = IsFunctionPrologue(*cii, addr);
bool is_epilogue = IsFunctionEpilogue(*cii, addr);
uint32_t lno = cur_info.getStartLine();
LineTable::Entry new_entry(addr, lno, 0, file_index, is_statement, false,
is_prologue, is_epilogue, false);
// Terminal entry has lower precedence than new entry.
auto iter = line_set.find(new_entry);
if (iter != line_set.end() && iter->is_terminal_entry)
line_set.erase(iter);
line_set.insert(new_entry);
if (line_entry.GetRangeBase() != LLDB_INVALID_ADDRESS) {
line_entry.SetRangeEnd(addr);
cii->m_global_line_table.Append(line_entry);
}
line_entry.SetRangeBase(addr);
line_entry.data = {file_index, lno};
}
LineInfo last_line(group.LineNumbers.back().Flags);
line_set.emplace(virtual_addr + lfh->CodeSize, last_line.getEndLine(), 0,
file_index, false, false, false, false, true);
if (line_entry.GetRangeBase() != LLDB_INVALID_ADDRESS) {
line_entry.SetRangeEnd(virtual_addr + lfh->CodeSize);
cii->m_global_line_table.Append(line_entry);
}
}
}
cii->m_global_line_table.Sort();
// Parse all S_INLINESITE in this CU.
const CVSymbolArray &syms = cii->m_debug_stream.getSymbolArray();
for (auto iter = syms.begin(); iter != syms.end();) {
if (iter->kind() != S_LPROC32 && iter->kind() != S_GPROC32) {
++iter;
continue;
}
uint32_t record_offset = iter.offset();
CVSymbol func_record =
cii->m_debug_stream.readSymbolAtOffset(record_offset);
SegmentOffsetLength sol = GetSegmentOffsetAndLength(func_record);
addr_t file_vm_addr =
m_index->MakeVirtualAddress(sol.so.segment, sol.so.offset);
if (file_vm_addr == LLDB_INVALID_ADDRESS)
continue;
Address func_base(file_vm_addr, comp_unit.GetModule()->GetSectionList());
PdbCompilandSymId func_id{modi, record_offset};
// Iterate all S_INLINESITEs in the function.
auto parse_inline_sites = [&](SymbolKind kind, PdbCompilandSymId id) {
if (kind != S_INLINESITE)
return false;
ParseInlineSite(id, func_base);
for (const auto &line_entry :
m_inline_sites[toOpaqueUid(id)]->line_entries) {
// If line_entry is not terminal entry, remove previous line entry at
// the same address and insert new one. Terminal entry inside an inline
// site might not be terminal entry for its parent.
if (!line_entry.is_terminal_entry)
line_set.erase(line_entry);
line_set.insert(line_entry);
}
// No longer useful after adding to line_set.
m_inline_sites[toOpaqueUid(id)]->line_entries.clear();
return true;
};
ParseSymbolArrayInScope(func_id, parse_inline_sites);
// Jump to the end of the function record.
iter = syms.at(getScopeEndOffset(func_record));
}
cii->m_global_line_table.Clear();
// Add line entries in line_set to line_table.
std::vector<LineTable::Sequence> sequence(1);
for (const auto &line_entry : line_set) {
LineTable::AppendLineEntryToSequence(
sequence.back(), line_entry.file_addr, line_entry.line,
line_entry.column, line_entry.file_idx,
line_entry.is_start_of_statement, line_entry.is_start_of_basic_block,
line_entry.is_prologue_end, line_entry.is_epilogue_begin,
line_entry.is_terminal_entry);
}
auto line_table =
std::make_unique<LineTable>(&comp_unit, std::move(sequence));
if (line_table->GetSize() == 0)
return false;
comp_unit.SetLineTable(line_table.release());
return true;
}
bool SymbolFileNativePDB::ParseDebugMacros(CompileUnit &comp_unit) {
// PDB doesn't contain information about macros
return false;
}
llvm::Expected<uint32_t>
SymbolFileNativePDB::GetFileIndex(const CompilandIndexItem &cii,
uint32_t file_id) {
if (!cii.m_strings.hasChecksums() || !cii.m_strings.hasStrings())
return llvm::make_error<RawError>(raw_error_code::no_entry);
const auto &checksums = cii.m_strings.checksums().getArray();
const auto &strings = cii.m_strings.strings();
// Indices in this structure are actually offsets of records in the
// DEBUG_S_FILECHECKSUMS subsection. Those entries then have an index
// into the global PDB string table.
auto iter = checksums.at(file_id);
if (iter == checksums.end())
return llvm::make_error<RawError>(raw_error_code::no_entry);
llvm::Expected<llvm::StringRef> efn = strings.getString(iter->FileNameOffset);
if (!efn) {
return efn.takeError();
}
// LLDB wants the index of the file in the list of support files.
auto fn_iter = llvm::find(cii.m_file_list, *efn);
if (fn_iter != cii.m_file_list.end())
return std::distance(cii.m_file_list.begin(), fn_iter);
return llvm::make_error<RawError>(raw_error_code::no_entry);
}
bool SymbolFileNativePDB::ParseSupportFiles(CompileUnit &comp_unit,
SupportFileList &support_files) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
PdbSymUid cu_id(comp_unit.GetID());
if (cu_id.kind() != PdbSymUidKind::Compiland) {
assert(false && "uid of compile unit not a compiland");
return false;
}
CompilandIndexItem *cci =
m_index->compilands().GetCompiland(cu_id.asCompiland().modi);
if (!cci) {
LLDB_LOG(GetLog(LLDBLog::Symbols), "missing compiland for modi={0}",
cu_id.asCompiland().modi);
return false;
}
for (llvm::StringRef f : cci->m_file_list) {
FileSpec::Style style =
f.starts_with("/") ? FileSpec::Style::posix : FileSpec::Style::windows;
FileSpec spec(f, style);
support_files.Append(spec);
}
return true;
}
bool SymbolFileNativePDB::ParseImportedModules(
const SymbolContext &sc, std::vector<SourceModule> &imported_modules) {
// PDB does not yet support module debug info
return false;
}
void SymbolFileNativePDB::ParseInlineSite(PdbCompilandSymId id,
Address func_addr) {
lldb::user_id_t opaque_uid = toOpaqueUid(id);
if (m_inline_sites.contains(opaque_uid))
return;
addr_t func_base = func_addr.GetFileAddress();
CompilandIndexItem *cii = m_index->compilands().GetCompiland(id.modi);
CVSymbol sym = cii->m_debug_stream.readSymbolAtOffset(id.offset);
CompUnitSP comp_unit = GetOrCreateCompileUnit(*cii);
if (sym.kind() != S_INLINESITE)
return;
InlineSiteSym inline_site(static_cast<SymbolRecordKind>(sym.kind()));
if (auto err =
SymbolDeserializer::deserializeAs<InlineSiteSym>(sym, inline_site)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize InlineSiteSym record: {0}");
return;
}
PdbCompilandSymId parent_id(id.modi, inline_site.Parent);
std::shared_ptr<InlineSite> inline_site_sp =
std::make_shared<InlineSite>(parent_id);
// Get the inlined function declaration info.
auto iter = cii->m_inline_map.find(inline_site.Inlinee);
if (iter == cii->m_inline_map.end())
return;
InlineeSourceLine inlinee_line = iter->second;
const SupportFileList &files = comp_unit->GetSupportFiles();
FileSpec decl_file;
llvm::Expected<uint32_t> file_index_or_err =
GetFileIndex(*cii, inlinee_line.Header->FileID);
if (!file_index_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), file_index_or_err.takeError(),
"failed to get file index for inline site: {0}");
return;
}
uint32_t file_offset = file_index_or_err.get();
decl_file = files.GetFileSpecAtIndex(file_offset);
uint32_t decl_line = inlinee_line.Header->SourceLineNum;
std::unique_ptr<Declaration> decl_up =
std::make_unique<Declaration>(decl_file, decl_line);
// Parse range and line info.
uint32_t code_offset = 0;
int32_t line_offset = 0;
std::optional<uint32_t> code_offset_base;
std::optional<uint32_t> code_offset_end;
std::optional<int32_t> cur_line_offset;
std::optional<int32_t> next_line_offset;
std::optional<uint32_t> next_file_offset;
bool is_terminal_entry = false;
bool is_start_of_statement = true;
// The first instruction is the prologue end.
bool is_prologue_end = true;
auto update_code_offset = [&](uint32_t code_delta) {
if (!code_offset_base)
code_offset_base = code_offset;
else if (!code_offset_end)
code_offset_end = *code_offset_base + code_delta;
};
auto update_line_offset = [&](int32_t line_delta) {
line_offset += line_delta;
if (!code_offset_base || !cur_line_offset)
cur_line_offset = line_offset;
else
next_line_offset = line_offset;
;
};
auto update_file_offset = [&](uint32_t offset) {
if (!code_offset_base)
file_offset = offset;
else
next_file_offset = offset;
};
for (auto &annot : inline_site.annotations()) {
switch (annot.OpCode) {
case BinaryAnnotationsOpCode::CodeOffset:
case BinaryAnnotationsOpCode::ChangeCodeOffset:
case BinaryAnnotationsOpCode::ChangeCodeOffsetBase:
code_offset += annot.U1;
update_code_offset(annot.U1);
break;
case BinaryAnnotationsOpCode::ChangeLineOffset:
update_line_offset(annot.S1);
break;
case BinaryAnnotationsOpCode::ChangeCodeLength:
update_code_offset(annot.U1);
code_offset += annot.U1;
is_terminal_entry = true;
break;
case BinaryAnnotationsOpCode::ChangeCodeOffsetAndLineOffset:
code_offset += annot.U1;
update_code_offset(annot.U1);
update_line_offset(annot.S1);
break;
case BinaryAnnotationsOpCode::ChangeCodeLengthAndCodeOffset:
code_offset += annot.U2;
update_code_offset(annot.U2);
update_code_offset(annot.U1);
code_offset += annot.U1;
is_terminal_entry = true;
break;
case BinaryAnnotationsOpCode::ChangeFile:
update_file_offset(annot.U1);
break;
default:
break;
}
// Add range if current range is finished.
if (code_offset_base && code_offset_end && cur_line_offset) {
inline_site_sp->ranges.Append(RangeSourceLineVector::Entry(
*code_offset_base, *code_offset_end - *code_offset_base,
decl_line + *cur_line_offset));
// Set base, end, file offset and line offset for next range.
if (next_file_offset)
file_offset = *next_file_offset;
if (next_line_offset) {
cur_line_offset = next_line_offset;
next_line_offset = std::nullopt;
}
code_offset_base = is_terminal_entry ? std::nullopt : code_offset_end;
code_offset_end = next_file_offset = std::nullopt;
}
if (code_offset_base && cur_line_offset) {
if (is_terminal_entry) {
LineTable::Entry line_entry(
func_base + *code_offset_base, decl_line + *cur_line_offset, 0,
file_offset, false, false, false, false, true);
inline_site_sp->line_entries.push_back(line_entry);
} else {
LineTable::Entry line_entry(func_base + *code_offset_base,
decl_line + *cur_line_offset, 0,
file_offset, is_start_of_statement, false,
is_prologue_end, false, false);
inline_site_sp->line_entries.push_back(line_entry);
is_prologue_end = false;
is_start_of_statement = false;
}
}
if (is_terminal_entry)
is_start_of_statement = true;
is_terminal_entry = false;
}
inline_site_sp->ranges.Sort();
// Get the inlined function callsite info.
std::unique_ptr<Declaration> callsite_up;
if (!inline_site_sp->ranges.IsEmpty()) {
auto *entry = inline_site_sp->ranges.GetEntryAtIndex(0);
addr_t base_offset = entry->GetRangeBase();
if (cii->m_debug_stream.readSymbolAtOffset(parent_id.offset).kind() ==
S_INLINESITE) {
// Its parent is another inline site, lookup parent site's range vector
// for callsite line.
ParseInlineSite(parent_id, Address(func_base));
std::shared_ptr<InlineSite> parent_site =
m_inline_sites[toOpaqueUid(parent_id)];
FileSpec &parent_decl_file =
parent_site->inline_function_info->GetDeclaration().GetFile();
if (auto *parent_entry =
parent_site->ranges.FindEntryThatContains(base_offset)) {
callsite_up =
std::make_unique<Declaration>(parent_decl_file, parent_entry->data);
}
} else {
// Its parent is a function, lookup global line table for callsite.
if (auto *entry = cii->m_global_line_table.FindEntryThatContains(
func_base + base_offset)) {
const FileSpec &callsite_file =
files.GetFileSpecAtIndex(entry->data.first);
callsite_up =
std::make_unique<Declaration>(callsite_file, entry->data.second);
}
}
}
// Get the inlined function name.
std::string inlinee_name;
llvm::Expected<CVType> inlinee_cvt =
m_index->ipi().typeCollection().getTypeOrError(inline_site.Inlinee);
if (!inlinee_cvt) {
inlinee_name = "[error reading function name: " +
llvm::toString(inlinee_cvt.takeError()) + "]";
} else if (inlinee_cvt->kind() == LF_MFUNC_ID) {
MemberFuncIdRecord mfr;
if (auto err = TypeDeserializer::deserializeAs<MemberFuncIdRecord>(
*inlinee_cvt, mfr)) {
inlinee_name =
"[error reading function name: " + llvm::toString(std::move(err)) +
"]";
} else {
LazyRandomTypeCollection &types = m_index->tpi().typeCollection();
inlinee_name.append(std::string(types.getTypeName(mfr.ClassType)));
inlinee_name.append("::");
inlinee_name.append(mfr.getName().str());
}
} else if (inlinee_cvt->kind() == LF_FUNC_ID) {
FuncIdRecord fir;
if (auto err =
TypeDeserializer::deserializeAs<FuncIdRecord>(*inlinee_cvt, fir)) {
inlinee_name =
"[error reading function name: " + llvm::toString(std::move(err)) +
"]";
} else {
TypeIndex parent_idx = fir.getParentScope();
if (!parent_idx.isNoneType()) {
LazyRandomTypeCollection &ids = m_index->ipi().typeCollection();
inlinee_name.append(std::string(ids.getTypeName(parent_idx)));
inlinee_name.append("::");
}
inlinee_name.append(fir.getName().str());
}
}
inline_site_sp->inline_function_info = std::make_shared<InlineFunctionInfo>(
inlinee_name.c_str(), llvm::StringRef(), decl_up.get(),
callsite_up.get());
m_inline_sites[opaque_uid] = inline_site_sp;
}
size_t SymbolFileNativePDB::ParseBlocksRecursive(Function &func) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
PdbCompilandSymId func_id = PdbSymUid(func.GetID()).asCompilandSym();
// After we iterate through inline sites inside the function, we already get
// all the info needed, removing from the map to save memory.
std::set<uint64_t> remove_uids;
auto parse_blocks = [&](SymbolKind kind, PdbCompilandSymId id) {
if (kind == S_GPROC32 || kind == S_LPROC32 || kind == S_BLOCK32 ||
kind == S_INLINESITE) {
GetOrCreateBlock(id);
if (kind == S_INLINESITE)
remove_uids.insert(toOpaqueUid(id));
return true;
}
return false;
};
size_t count = ParseSymbolArrayInScope(func_id, parse_blocks);
for (uint64_t uid : remove_uids) {
m_inline_sites.erase(uid);
}
func.GetBlock(false).SetBlockInfoHasBeenParsed(true, true);
return count;
}
size_t SymbolFileNativePDB::ParseSymbolArrayInScope(
PdbCompilandSymId parent_id,
llvm::function_ref<bool(SymbolKind, PdbCompilandSymId)> fn) {
CompilandIndexItem *cii = m_index->compilands().GetCompiland(parent_id.modi);
CVSymbolArray syms =
cii->m_debug_stream.getSymbolArrayForScope(parent_id.offset);
size_t count = 1;
for (auto iter = syms.begin(); iter != syms.end(); ++iter) {
PdbCompilandSymId child_id(parent_id.modi, iter.offset());
if (fn(iter->kind(), child_id))
++count;
}
return count;
}
void SymbolFileNativePDB::DumpClangAST(Stream &s, llvm::StringRef filter,
bool show_color) {
auto ts_or_err = GetTypeSystemForLanguage(eLanguageTypeC_plus_plus);
if (!ts_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), ts_or_err.takeError(),
"failed to get C++ type system: {0}");
return;
}
auto ts = *ts_or_err;
TypeSystemClang *clang = llvm::dyn_cast_or_null<TypeSystemClang>(ts.get());
if (!clang)
return;
PdbAstBuilder *ast_builder = clang->GetNativePDBParser();
if (!ast_builder)
return;
ast_builder->Dump(s, filter, show_color);
}
void SymbolFileNativePDB::CacheGlobalBaseNames() {
if (!m_func_full_names.IsEmpty() || !m_global_variable_base_names.IsEmpty())
return;
// (segment, code offset) -> gid
std::map<std::pair<uint16_t, uint32_t>, uint32_t> func_addr_ids;
// First, look through all items in the globals table.
for (const uint32_t gid : m_index->globals().getGlobalsTable()) {
CVSymbol sym = m_index->symrecords().readRecord(gid);
auto kind = sym.kind();
// If this is a global variable, we only need to look at the name
llvm::StringRef name;
switch (kind) {
case SymbolKind::S_GDATA32:
case SymbolKind::S_LDATA32: {
auto data_or_err = SymbolDeserializer::deserializeAs<DataSym>(sym);
if (!data_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), data_or_err.takeError(),
"Failed to deserialize DataSym record: {0}");
continue;
}
name = data_or_err->Name;
break;
}
case SymbolKind::S_GTHREAD32:
case SymbolKind::S_LTHREAD32: {
auto data_or_err =
SymbolDeserializer::deserializeAs<ThreadLocalDataSym>(sym);
if (!data_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), data_or_err.takeError(),
"Failed to deserialize ThreadLocalDataSym record: {0}");
continue;
}
name = data_or_err->Name;
break;
}
case SymbolKind::S_CONSTANT: {
auto data_or_err = SymbolDeserializer::deserializeAs<ConstantSym>(sym);
if (!data_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), data_or_err.takeError(),
"Failed to deserialize ConstantSym record: {0}");
continue;
}
name = data_or_err->Name;
break;
}
default:
break;
}
if (!name.empty()) {
llvm::StringRef base = MSVCUndecoratedNameParser::DropScope(name);
if (base.empty())
base = name;
m_global_variable_base_names.Append(ConstString(base), gid);
continue;
}
if (kind != S_PROCREF && kind != S_LPROCREF)
continue;
// For functions, we need to follow the reference to the procedure and look
// at the type
auto ref_or_err = SymbolDeserializer::deserializeAs<ProcRefSym>(sym);
if (!ref_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), ref_or_err.takeError(),
"Failed to deserialize ProcRefSym record: {0}");
continue;
}
ProcRefSym ref = std::move(*ref_or_err);
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;
auto proc_or_err = SymbolDeserializer::deserializeAs<ProcSym>(*iter);
if (!proc_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), proc_or_err.takeError(),
"Failed to deserialize ProcSym record: {0}");
continue;
}
ProcSym proc = std::move(*proc_or_err);
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.
func_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;
if (auto err = TypeDeserializer::deserializeAs<MemberFunctionRecord>(
type, mfr)) {
LLDB_LOG_ERROR(
GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize MemberFunctionRecord record ({1}): {0}",
proc.FunctionType);
} else 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;
auto pub_or_err = SymbolDeserializer::deserializeAs<PublicSym32>(sym);
if (!pub_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), pub_or_err.takeError(),
"Failed to deserialize PublicSym32 record: {0}");
continue;
}
PublicSym32 pub = std::move(*pub_or_err);
// 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 = func_addr_ids.find({pub.Segment, pub.Offset});
if (it != func_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(std::less<uint32_t>());
m_func_full_names.SizeToFit();
m_func_method_names.Sort(std::less<uint32_t>());
m_func_method_names.SizeToFit();
m_func_base_names.Sort(std::less<uint32_t>());
m_func_base_names.SizeToFit();
m_global_variable_base_names.Sort(std::less<uint32_t>());
m_global_variable_base_names.SizeToFit();
}
void SymbolFileNativePDB::FindGlobalVariables(
ConstString name, const CompilerDeclContext &parent_decl_ctx,
uint32_t max_matches, VariableList &variables) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
CacheGlobalBaseNames();
std::vector<uint32_t> results;
m_global_variable_base_names.GetValues(name, results);
size_t n_matches = 0;
for (uint32_t gid : results) {
PdbGlobalSymId global(gid, false);
if (parent_decl_ctx.IsValid() &&
GetDeclContextContainingUID(toOpaqueUid(global)) != parent_decl_ctx)
continue;
VariableSP var = GetOrCreateGlobalVariable(global);
if (!var)
continue;
variables.AddVariable(var);
if (++n_matches >= max_matches)
break;
}
}
void SymbolFileNativePDB::FindFunctions(
const Module::LookupInfo &lookup_info,
const CompilerDeclContext &parent_decl_ctx, bool include_inlines,
SymbolContextList &sc_list) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
ConstString name = lookup_info.GetLookupName();
FunctionNameType name_type_mask = lookup_info.GetNameTypeMask();
if (name_type_mask & eFunctionNameTypeFull)
name = lookup_info.GetName();
if (!(name_type_mask & eFunctionNameTypeFull ||
name_type_mask & eFunctionNameTypeBase ||
name_type_mask & eFunctionNameTypeMethod))
return;
CacheGlobalBaseNames();
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;
for (uint32_t id : ids) {
if (!resolved_ids.insert(id).second)
continue;
PdbGlobalSymId global{id, false};
if (parent_decl_ctx.IsValid() &&
GetDeclContextContainingUID(toOpaqueUid(global)) != parent_decl_ctx)
continue;
CVSymbol sym = m_index->ReadSymbolRecord(global);
auto kind = sym.kind();
if (kind != S_PROCREF && kind != S_LPROCREF) {
LLDB_LOG(GetLog(LLDBLog::Symbols), "{0} is not a proc reference",
global);
continue;
}
auto proc_or_err = SymbolDeserializer::deserializeAs<ProcRefSym>(sym);
if (!proc_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), proc_or_err.takeError(),
"Failed to deserialize ProcRefSym record: {0}");
continue;
}
ProcRefSym proc = std::move(*proc_or_err);
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,
bool include_inlines,
SymbolContextList &sc_list) {}
void SymbolFileNativePDB::FindTypes(const lldb_private::TypeQuery &query,
lldb_private::TypeResults &results) {
// Make sure we haven't already searched this SymbolFile before.
if (results.AlreadySearched(this))
return;
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
BuildParentMap();
// We can't query for the full name because the type might reside
// in an anonymous namespace. Search for the basename in our map and check the
// matching types afterwards.
std::vector<uint32_t> matches;
m_type_base_names.GetValues(query.GetTypeBasename(), matches);
for (uint32_t match_idx : matches) {
std::vector context = GetContextForType(TypeIndex(match_idx));
if (context.empty())
continue;
if (query.ContextMatches(context)) {
TypeSP type_sp = GetOrCreateType(TypeIndex(match_idx));
if (!type_sp)
continue;
results.InsertUnique(type_sp);
if (results.Done(query))
return;
}
}
}
void SymbolFileNativePDB::FindTypesByName(llvm::StringRef name,
uint32_t max_matches,
TypeMap &types) {
std::vector<TypeIndex> matches = m_index->tpi().findRecordsByName(name);
if (max_matches > 0 && max_matches < matches.size())
matches.resize(max_matches);
for (TypeIndex ti : matches) {
TypeSP type = GetOrCreateType(ti);
if (!type)
continue;
types.Insert(type);
}
}
size_t SymbolFileNativePDB::ParseTypes(CompileUnit &comp_unit) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
// Only do the full type scan the first time.
if (m_done_full_type_scan)
return 0;
const size_t old_count = GetTypeList().GetSize();
LazyRandomTypeCollection &types = m_index->tpi().typeCollection();
// First process the entire TPI stream.
for (auto ti = types.getFirst(); ti; ti = types.getNext(*ti)) {
TypeSP type = GetOrCreateType(*ti);
if (type)
(void)type->GetFullCompilerType();
}
// Next look for S_UDT records in the globals stream.
for (const uint32_t gid : m_index->globals().getGlobalsTable()) {
PdbGlobalSymId global{gid, false};
CVSymbol sym = m_index->ReadSymbolRecord(global);
if (sym.kind() != S_UDT)
continue;
auto udt_or_err = SymbolDeserializer::deserializeAs<UDTSym>(sym);
if (!udt_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), udt_or_err.takeError(),
"Failed to deserialize UDTSym record: {0}");
continue;
}
UDTSym udt = std::move(*udt_or_err);
bool is_typedef = true;
if (IsTagRecord(PdbTypeSymId{udt.Type, false}, m_index->tpi())) {
CVType cvt = m_index->tpi().getType(udt.Type);
llvm::StringRef name = CVTagRecord::create(cvt).name();
if (name == udt.Name)
is_typedef = false;
}
if (is_typedef)
GetOrCreateTypedef(global);
}
const size_t new_count = GetTypeList().GetSize();
m_done_full_type_scan = true;
return new_count - old_count;
}
size_t
SymbolFileNativePDB::ParseVariablesForCompileUnit(CompileUnit &comp_unit,
VariableList &variables) {
for (const uint32_t gid : m_index->globals().getGlobalsTable()) {
PdbGlobalSymId global{gid, false};
CVSymbol sym = m_index->ReadSymbolRecord(global);
// TODO: S_CONSTANT is not handled here to prevent a possible crash in
// lldb_private::npdb::MakeConstantLocationExpression when it's a record
// type (e.g. std::strong_ordering::equal). That function needs to be
// updated to handle this case when we add S_CONSTANT case here.
switch (sym.kind()) {
case SymbolKind::S_GDATA32:
case SymbolKind::S_LDATA32:
case SymbolKind::S_GTHREAD32:
case SymbolKind::S_LTHREAD32: {
if (VariableSP var = GetOrCreateGlobalVariable(global))
variables.AddVariable(var);
break;
}
default:
break;
}
}
return variables.GetSize();
}
VariableSP SymbolFileNativePDB::CreateLocalVariable(PdbCompilandSymId scope_id,
PdbCompilandSymId var_id,
bool is_param,
bool is_constant) {
ModuleSP module = GetObjectFile()->GetModule();
Block *block = GetOrCreateBlock(scope_id);
if (!block)
return nullptr;
CompilandIndexItem *cii = m_index->compilands().GetCompiland(var_id.modi);
if (!cii)
return nullptr;
CompUnitSP comp_unit_sp = GetOrCreateCompileUnit(*cii);
VariableInfo var_info;
bool location_is_constant_data = is_constant;
if (is_constant) {
CVSymbol sym = cii->m_debug_stream.readSymbolAtOffset(var_id.offset);
if (sym.kind() != S_CONSTANT)
return nullptr;
ConstantSym constant(sym.kind());
if (auto err =
SymbolDeserializer::deserializeAs<ConstantSym>(sym, constant)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ConstantSym record: {0}");
return nullptr;
}
var_info.name = constant.Name;
var_info.type = constant.Type;
auto location_or_err = MakeConstantLocationExpression(
constant.Type, m_index->tpi(), constant.Value, module);
if (!location_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), location_or_err.takeError(),
"Failed to make constant location expression for {1}: {0}",
constant.Name);
return nullptr;
}
var_info.location =
DWARFExpressionList(module, std::move(*location_or_err), nullptr);
} else {
// Get function block.
Block *func_block = block;
while (func_block->GetParent())
func_block = func_block->GetParent();
Address addr;
func_block->GetStartAddress(addr);
var_info = GetVariableLocationInfo(*m_index, var_id, *func_block, module);
Function *func = func_block->CalculateSymbolContextFunction();
if (!func)
return nullptr;
// Use empty dwarf expr if optimized away so that it won't be filtered out
// when lookuping local variables in this scope.
if (!var_info.location.IsValid())
var_info.location =
DWARFExpressionList(module, DWARFExpression(), nullptr);
var_info.location.SetFuncFileAddress(func->GetAddress().GetFileAddress());
}
TypeSP type_sp = GetOrCreateType(var_info.type);
if (!type_sp)
return nullptr;
std::string name = var_info.name.str();
Declaration decl;
SymbolFileTypeSP sftype =
std::make_shared<SymbolFileType>(*this, type_sp->GetID());
is_param |= var_info.is_param;
ValueType var_scope =
is_param ? eValueTypeVariableArgument : eValueTypeVariableLocal;
bool external = false;
bool artificial = false;
bool static_member = false;
Variable::RangeList scope_ranges;
VariableSP var_sp = std::make_shared<Variable>(
toOpaqueUid(var_id), name.c_str(), name.c_str(), sftype, var_scope, block,
scope_ranges, &decl, var_info.location, external, artificial,
location_is_constant_data, static_member);
if (!is_param) {
auto ts_or_err = GetTypeSystemForLanguage(comp_unit_sp->GetLanguage());
if (auto err = ts_or_err.takeError())
return nullptr;
auto ts = *ts_or_err;
if (ts) {
if (PdbAstBuilder *ast_builder = ts->GetNativePDBParser())
ast_builder->EnsureVariable(scope_id, var_id);
}
}
m_local_variables[toOpaqueUid(var_id)] = var_sp;
return var_sp;
}
VariableSP
SymbolFileNativePDB::GetOrCreateLocalVariable(PdbCompilandSymId scope_id,
PdbCompilandSymId var_id,
bool is_param, bool is_constant) {
auto iter = m_local_variables.find(toOpaqueUid(var_id));
if (iter != m_local_variables.end())
return iter->second;
return CreateLocalVariable(scope_id, var_id, is_param, is_constant);
}
TypeSP SymbolFileNativePDB::CreateTypedef(PdbGlobalSymId id) {
CVSymbol sym = m_index->ReadSymbolRecord(id);
if (sym.kind() != S_UDT) {
LLDB_LOG(GetLog(LLDBLog::Symbols), "{0} is not an S_UDT", id);
return nullptr;
}
auto udt_or_err = SymbolDeserializer::deserializeAs<UDTSym>(sym);
if (!udt_or_err) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), udt_or_err.takeError(),
"Failed to deserialize UDTSym record: {0}");
return nullptr;
}
UDTSym udt = std::move(*udt_or_err);
TypeSP target_type = GetOrCreateType(udt.Type);
auto ts_or_err = GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
if (auto err = ts_or_err.takeError())
return nullptr;
auto ts = *ts_or_err;
if (!ts)
return nullptr;
PdbAstBuilder *ast_builder = ts->GetNativePDBParser();
if (!ast_builder)
return nullptr;
CompilerType ct = ast_builder->GetOrCreateTypedefType(id);
if (!ct)
ct = target_type->GetForwardCompilerType();
Declaration decl;
return MakeType(toOpaqueUid(id), ConstString(udt.Name),
llvm::expectedToOptional(target_type->GetByteSize(nullptr)),
nullptr, target_type->GetID(),
lldb_private::Type::eEncodingIsTypedefUID, decl, ct,
lldb_private::Type::ResolveState::Forward);
}
TypeSP SymbolFileNativePDB::GetOrCreateTypedef(PdbGlobalSymId id) {
auto iter = m_types.find(toOpaqueUid(id));
if (iter != m_types.end())
return iter->second;
return CreateTypedef(id);
}
size_t SymbolFileNativePDB::ParseVariablesForBlock(PdbCompilandSymId block_id) {
Block *block = GetOrCreateBlock(block_id);
if (!block)
return 0;
size_t count = 0;
CompilandIndexItem *cii = m_index->compilands().GetCompiland(block_id.modi);
CVSymbol sym = cii->m_debug_stream.readSymbolAtOffset(block_id.offset);
uint32_t params_remaining = 0;
switch (sym.kind()) {
case S_GPROC32:
case S_LPROC32: {
ProcSym proc(static_cast<SymbolRecordKind>(sym.kind()));
if (auto err = SymbolDeserializer::deserializeAs<ProcSym>(sym, proc)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ProcSym record: {0}");
return 0;
}
CVType signature = m_index->tpi().getType(proc.FunctionType);
if (signature.kind() == LF_PROCEDURE) {
ProcedureRecord sig;
if (llvm::Error e = TypeDeserializer::deserializeAs<ProcedureRecord>(
signature, sig)) {
llvm::consumeError(std::move(e));
return 0;
}
params_remaining = sig.getParameterCount();
} else if (signature.kind() == LF_MFUNCTION) {
MemberFunctionRecord sig;
if (llvm::Error e = TypeDeserializer::deserializeAs<MemberFunctionRecord>(
signature, sig)) {
llvm::consumeError(std::move(e));
return 0;
}
params_remaining = sig.getParameterCount();
} else
return 0;
break;
}
case S_BLOCK32:
break;
case S_INLINESITE:
break;
default:
LLDB_LOG(GetLog(LLDBLog::Symbols), "{0} is not a block", block_id);
return 0;
}
VariableListSP variables = block->GetBlockVariableList(false);
if (!variables) {
variables = std::make_shared<VariableList>();
block->SetVariableList(variables);
}
CVSymbolArray syms = limitSymbolArrayToScope(
cii->m_debug_stream.getSymbolArray(), block_id.offset);
// Skip the first record since it's a PROC32 or BLOCK32, and there's
// no point examining it since we know it's not a local variable.
syms.drop_front();
auto iter = syms.begin();
auto end = syms.end();
while (iter != end) {
uint32_t record_offset = iter.offset();
CVSymbol variable_cvs = *iter;
PdbCompilandSymId child_sym_id(block_id.modi, record_offset);
++iter;
// If this is a block or inline site, recurse into its children and then
// skip it.
if (variable_cvs.kind() == S_BLOCK32 ||
variable_cvs.kind() == S_INLINESITE) {
uint32_t block_end = getScopeEndOffset(variable_cvs);
count += ParseVariablesForBlock(child_sym_id);
iter = syms.at(block_end);
continue;
}
bool is_param = params_remaining > 0;
VariableSP variable;
switch (variable_cvs.kind()) {
case S_REGREL32:
case S_REGREL32_INDIR:
case S_REGISTER:
case S_LOCAL:
variable = GetOrCreateLocalVariable(block_id, child_sym_id, is_param);
if (is_param)
--params_remaining;
if (variable)
variables->AddVariableIfUnique(variable);
break;
case S_CONSTANT:
variable = GetOrCreateLocalVariable(block_id, child_sym_id,
/*is_param=*/false,
/*is_constant=*/true);
if (variable)
variables->AddVariableIfUnique(variable);
break;
default:
break;
}
}
// Pass false for set_children, since we call this recursively so that the
// children will call this for themselves.
block->SetDidParseVariables(true, false);
return count;
}
size_t SymbolFileNativePDB::ParseVariablesForContext(const SymbolContext &sc) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
if (sc.block) {
PdbSymUid block_id(sc.block->GetID());
size_t count = ParseVariablesForBlock(block_id.asCompilandSym());
return count;
}
if (sc.function) {
PdbSymUid block_id(sc.function->GetID());
size_t count = ParseVariablesForBlock(block_id.asCompilandSym());
return count;
}
if (sc.comp_unit) {
VariableListSP variables = sc.comp_unit->GetVariableList(false);
if (!variables) {
variables = std::make_shared<VariableList>();
sc.comp_unit->SetVariableList(variables);
}
return ParseVariablesForCompileUnit(*sc.comp_unit, *variables);
}
LLDB_LOG(GetLog(LLDBLog::Symbols),
"missing missing block, function, or module for symbol context");
return 0;
}
CompilerDecl SymbolFileNativePDB::GetDeclForUID(lldb::user_id_t uid) {
auto ts_or_err = GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
if (auto err = ts_or_err.takeError())
return CompilerDecl();
auto ts = *ts_or_err;
if (!ts)
return {};
PdbAstBuilder *ast_builder = ts->GetNativePDBParser();
if (!ast_builder)
return {};
return ast_builder->GetOrCreateDeclForUid(uid);
}
CompilerDeclContext
SymbolFileNativePDB::GetDeclContextForUID(lldb::user_id_t uid) {
auto ts_or_err = GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
if (auto err = ts_or_err.takeError())
return {};
auto ts = *ts_or_err;
if (!ts)
return {};
PdbAstBuilder *ast_builder = ts->GetNativePDBParser();
if (!ast_builder)
return {};
return ast_builder->GetOrCreateDeclContextForUid(PdbSymUid(uid));
}
CompilerDeclContext
SymbolFileNativePDB::GetDeclContextContainingUID(lldb::user_id_t uid) {
auto ts_or_err = GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
if (auto err = ts_or_err.takeError())
return CompilerDeclContext();
auto ts = *ts_or_err;
if (!ts)
return {};
PdbAstBuilder *ast_builder = ts->GetNativePDBParser();
if (!ast_builder)
return {};
return ast_builder->GetParentDeclContext(PdbSymUid(uid));
}
Type *SymbolFileNativePDB::ResolveTypeUID(lldb::user_id_t type_uid) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
auto iter = m_types.find(type_uid);
// lldb should not be passing us non-sensical type uids. the only way it
// could have a type uid in the first place is if we handed it out, in which
// case we should know about the type. However, that doesn't mean we've
// instantiated it yet. We can vend out a UID for a future type. So if the
// type doesn't exist, let's instantiate it now.
if (iter != m_types.end())
return &*iter->second;
PdbSymUid uid(type_uid);
if (uid.kind() != PdbSymUidKind::Type) {
assert(false && "uid is not a type index");
return nullptr;
}
PdbTypeSymId type_id = uid.asTypeSym();
if (type_id.index.isNoneType())
return nullptr;
TypeSP type_sp = CreateAndCacheType(type_id);
if (!type_sp)
return nullptr;
return &*type_sp;
}
std::optional<SymbolFile::ArrayInfo>
SymbolFileNativePDB::GetDynamicArrayInfoForUID(
lldb::user_id_t type_uid, const lldb_private::ExecutionContext *exe_ctx) {
return std::nullopt;
}
bool SymbolFileNativePDB::CompleteType(CompilerType &compiler_type) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
auto ts = compiler_type.GetTypeSystem();
if (!ts)
return false;
PdbAstBuilder *ast_builder = ts->GetNativePDBParser();
if (!ast_builder)
return false;
return ast_builder->CompleteType(compiler_type);
}
void SymbolFileNativePDB::GetTypes(lldb_private::SymbolContextScope *sc_scope,
TypeClass type_mask,
lldb_private::TypeList &type_list) {}
CompilerDeclContext
SymbolFileNativePDB::FindNamespace(ConstString name,
const CompilerDeclContext &parent_decl_ctx,
bool /* only_root_namespaces */) {
std::lock_guard<std::recursive_mutex> guard(GetModuleMutex());
auto ts_or_err = GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
if (auto err = ts_or_err.takeError())
return {};
auto ts = *ts_or_err;
if (!ts)
return {};
auto *clang = llvm::dyn_cast_or_null<TypeSystemClang>(ts.get());
if (!clang)
return {};
PdbAstBuilder *ast_builder = clang->GetNativePDBParser();
if (!ast_builder)
return {};
return ast_builder->FindNamespaceDecl(parent_decl_ctx, name.GetStringRef());
}
llvm::Expected<lldb::TypeSystemSP>
SymbolFileNativePDB::GetTypeSystemForLanguage(lldb::LanguageType language) {
auto type_system_or_err =
m_objfile_sp->GetModule()->GetTypeSystemForLanguage(language);
if (type_system_or_err)
if (auto ts = *type_system_or_err)
ts->SetSymbolFile(this);
return type_system_or_err;
}
uint64_t SymbolFileNativePDB::GetDebugInfoSize(bool load_all_debug_info) {
// PDB files are a separate file that contains all debug info.
return m_index->pdb().getFileSize();
}
void SymbolFileNativePDB::BuildParentMap() {
if (m_parent_map_built)
return;
m_parent_map_built = true;
LazyRandomTypeCollection &types = m_index->tpi().typeCollection();
llvm::DenseMap<TypeIndex, TypeIndex> forward_to_full;
llvm::DenseMap<TypeIndex, TypeIndex> full_to_forward;
struct RecordIndices {
TypeIndex forward;
TypeIndex full;
};
llvm::StringMap<RecordIndices> record_indices;
for (auto ti = types.getFirst(); ti; ti = types.getNext(*ti)) {
CVType type = types.getType(*ti);
if (!IsTagRecord(type))
continue;
CVTagRecord tag = CVTagRecord::create(type);
RecordIndices &indices = record_indices[tag.asTag().getUniqueName()];
if (tag.asTag().isForwardRef()) {
indices.forward = *ti;
} else {
indices.full = *ti;
auto base_name = MSVCUndecoratedNameParser::DropScope(tag.name());
m_type_base_names.Append(ConstString(base_name), ti->getIndex());
}
if (indices.full != TypeIndex::None() &&
indices.forward != TypeIndex::None()) {
forward_to_full[indices.forward] = indices.full;
full_to_forward[indices.full] = indices.forward;
}
// We're looking for LF_NESTTYPE records in the field list, so ignore
// forward references (no field list), and anything without a nested class
// (since there won't be any LF_NESTTYPE records).
if (tag.asTag().isForwardRef() || !tag.asTag().containsNestedClass())
continue;
struct ProcessTpiStream : public TypeVisitorCallbacks {
ProcessTpiStream(PdbIndex &index, TypeIndex parent,
const CVTagRecord &parent_cvt,
llvm::DenseMap<TypeIndex, TypeIndex> &parents)
: index(index), parents(parents), parent(parent),
parent_cvt(parent_cvt) {}
PdbIndex &index;
llvm::DenseMap<TypeIndex, TypeIndex> &parents;
unsigned unnamed_type_index = 1;
TypeIndex parent;
const CVTagRecord &parent_cvt;
llvm::Error visitKnownMember(CVMemberRecord &CVR,
NestedTypeRecord &Record) override {
std::string unnamed_type_name;
if (Record.Name.empty()) {
unnamed_type_name =
llvm::formatv("<unnamed-type-$S{0}>", unnamed_type_index).str();
Record.Name = unnamed_type_name;
++unnamed_type_index;
}
std::optional<CVTagRecord> tag =
GetNestedTagDefinition(Record, parent_cvt, index.tpi());
if (!tag)
return llvm::ErrorSuccess();
parents[Record.Type] = parent;
return llvm::ErrorSuccess();
}
};
CVType field_list_cvt = m_index->tpi().getType(tag.asTag().FieldList);
if (field_list_cvt.kind() != LF_FIELDLIST)
continue; // Invalid reference to a field list.
ProcessTpiStream process(*m_index, *ti, tag, m_parent_types);
FieldListRecord field_list;
if (llvm::Error error = TypeDeserializer::deserializeAs<FieldListRecord>(
field_list_cvt, field_list))
llvm::consumeError(std::move(error));
if (llvm::Error error = visitMemberRecordStream(field_list.Data, process))
llvm::consumeError(std::move(error));
}
// After calling Append(), the type-name map needs to be sorted again to be
// able to look up a type by its name.
m_type_base_names.Sort(std::less<uint32_t>());
// Now that we know the forward -> full mapping of all type indices, we can
// re-write all the indices. At the end of this process, we want a mapping
// consisting of fwd -> full and full -> full for all child -> parent indices.
// We can re-write the values in place, but for the keys, we must save them
// off so that we don't modify the map in place while also iterating it.
std::vector<TypeIndex> full_keys;
std::vector<TypeIndex> fwd_keys;
for (auto &entry : m_parent_types) {
TypeIndex key = entry.first;
TypeIndex value = entry.second;
auto iter = forward_to_full.find(value);
if (iter != forward_to_full.end())
entry.second = iter->second;
iter = forward_to_full.find(key);
if (iter != forward_to_full.end())
fwd_keys.push_back(key);
else
full_keys.push_back(key);
}
for (TypeIndex fwd : fwd_keys) {
TypeIndex full = forward_to_full[fwd];
TypeIndex parent_idx = m_parent_types[fwd];
m_parent_types[full] = parent_idx;
}
for (TypeIndex full : full_keys) {
TypeIndex fwd = full_to_forward[full];
m_parent_types[fwd] = m_parent_types[full];
}
}
std::optional<PdbCompilandSymId>
SymbolFileNativePDB::FindSymbolScope(PdbCompilandSymId id) {
CVSymbol sym = m_index->ReadSymbolRecord(id);
if (symbolOpensScope(sym.kind())) {
// If this exact symbol opens a scope, we can just directly access its
// parent.
id.offset = getScopeParentOffset(sym);
// Global symbols have parent offset of 0. Return std::nullopt to indicate
// this.
if (id.offset == 0)
return std::nullopt;
return id;
}
// Otherwise we need to start at the beginning and iterate forward until we
// reach (or pass) this particular symbol
CompilandIndexItem &cii = m_index->compilands().GetOrCreateCompiland(id.modi);
const CVSymbolArray &syms = cii.m_debug_stream.getSymbolArray();
auto begin = syms.begin();
auto end = syms.at(id.offset);
std::vector<PdbCompilandSymId> scope_stack;
while (begin != end) {
if (begin.offset() > id.offset) {
// We passed it. We couldn't even find this symbol record.
LLDB_LOG(GetLog(LLDBLog::Symbols), "invalid compiland symbol id: {0}",
id);
return std::nullopt;
}
// We haven't found the symbol yet. Check if we need to open or close the
// scope stack.
if (symbolOpensScope(begin->kind())) {
// We can use the end offset of the scope to determine whether or not
// we can just outright skip this entire scope.
uint32_t scope_end = getScopeEndOffset(*begin);
if (scope_end < id.offset) {
begin = syms.at(scope_end);
} else {
// The symbol we're looking for is somewhere in this scope.
scope_stack.emplace_back(id.modi, begin.offset());
}
} else if (symbolEndsScope(begin->kind())) {
scope_stack.pop_back();
}
++begin;
}
if (scope_stack.empty())
return std::nullopt;
// We have a match! Return the top of the stack
return scope_stack.back();
}
std::optional<llvm::codeview::TypeIndex>
SymbolFileNativePDB::GetParentType(llvm::codeview::TypeIndex ti) {
BuildParentMap();
auto parent_iter = m_parent_types.find(ti);
if (parent_iter == m_parent_types.end())
return std::nullopt;
return parent_iter->second;
}
std::vector<CompilerContext>
SymbolFileNativePDB::GetContextForType(TypeIndex ti) {
CVType type = m_index->tpi().getType(ti);
if (!IsTagRecord(type))
return {};
CVTagRecord tag = CVTagRecord::create(type);
std::optional<Type::ParsedName> parsed_name =
Type::GetTypeScopeAndBasename(tag.name());
if (!parsed_name)
return {{tag.contextKind(), ConstString(tag.name())}};
std::vector<CompilerContext> ctx;
// assume everything is a namespace at first
for (llvm::StringRef scope : parsed_name->scope) {
ctx.emplace_back(CompilerContextKind::Namespace, ConstString(scope));
}
// we know the kind of our own type
ctx.emplace_back(tag.contextKind(), ConstString(parsed_name->basename));
// try to find the kind of parents
for (auto &el : llvm::reverse(llvm::drop_end(ctx))) {
std::optional<TypeIndex> parent = GetParentType(ti);
if (!parent)
break;
ti = *parent;
type = m_index->tpi().getType(ti);
switch (type.kind()) {
case LF_CLASS:
case LF_STRUCTURE:
case LF_INTERFACE:
el.kind = CompilerContextKind::ClassOrStruct;
continue;
case LF_UNION:
el.kind = CompilerContextKind::Union;
continue;
case LF_ENUM:
el.kind = CompilerContextKind::Enum;
continue;
default:
break;
}
break;
}
return ctx;
}
std::optional<llvm::StringRef>
SymbolFileNativePDB::FindMangledFunctionName(PdbCompilandSymId func_id) {
const CompilandIndexItem *cci =
m_index->compilands().GetCompiland(func_id.modi);
if (!cci)
return std::nullopt;
CVSymbol sym_record = cci->m_debug_stream.readSymbolAtOffset(func_id.offset);
if (sym_record.kind() != S_LPROC32 && sym_record.kind() != S_GPROC32)
return std::nullopt;
ProcSym proc(static_cast<SymbolRecordKind>(sym_record.kind()));
if (auto err = SymbolDeserializer::deserializeAs<ProcSym>(sym_record, proc)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize ProcSym record: {0}");
return std::nullopt;
}
return FindMangledSymbol(SegmentOffset(proc.Segment, proc.CodeOffset),
proc.FunctionType);
}
std::optional<llvm::StringRef>
SymbolFileNativePDB::FindMangledSymbol(SegmentOffset so,
TypeIndex function_type) {
auto symbol = m_index->publics().findByAddress(m_index->symrecords(),
so.segment, so.offset);
if (!symbol)
return std::nullopt;
llvm::StringRef name = symbol->first.Name;
// For functions, we might need to strip the mangled name. See
// StripMangledFunctionName for more info.
if (!function_type.isNoneType() &&
(symbol->first.Flags & PublicSymFlags::Function) != PublicSymFlags::None)
name = StripMangledFunctionName(name, function_type);
return name;
}
llvm::StringRef
SymbolFileNativePDB::StripMangledFunctionName(const llvm::StringRef mangled,
PdbTypeSymId func_ty) {
// "In non-64 bit environments" (on x86 in pactice), __cdecl functions get
// prefixed with an underscore. For compilers using LLVM, this happens in LLVM
// (as opposed to the compiler frontend). Because of this, DWARF doesn't
// contain the "full" mangled name in DW_AT_linkage_name for these functions.
// We strip the mangling here for compatibility with DWARF. See
// llvm.org/pr161676 and
// https://learn.microsoft.com/en-us/cpp/build/reference/decorated-names#FormatC
if (!mangled.starts_with('_') ||
m_index->dbi().getMachineType() != PDB_Machine::x86)
return mangled;
CVType cvt = m_index->tpi().getType(func_ty.index);
PDB_CallingConv cc = PDB_CallingConv::NearC;
if (cvt.kind() == LF_PROCEDURE) {
ProcedureRecord proc;
if (llvm::Error error =
TypeDeserializer::deserializeAs<ProcedureRecord>(cvt, proc))
llvm::consumeError(std::move(error));
cc = proc.CallConv;
} else if (cvt.kind() == LF_MFUNCTION) {
MemberFunctionRecord mfunc;
if (llvm::Error error =
TypeDeserializer::deserializeAs<MemberFunctionRecord>(cvt, mfunc))
llvm::consumeError(std::move(error));
cc = mfunc.CallConv;
} else {
LLDB_LOG(GetLog(LLDBLog::Symbols), "Unexpected function type, got {0}",
cvt.kind());
return mangled;
}
if (cc == PDB_CallingConv::NearC || cc == PDB_CallingConv::FarC)
return mangled.drop_front();
return mangled;
}
void SymbolFileNativePDB::CacheUdtDeclarations() {
for (CVType cvt : m_index->ipi().typeArray()) {
switch (cvt.kind()) {
case LF_UDT_SRC_LINE: {
UdtSourceLineRecord udt_src;
if (auto err = TypeDeserializer::deserializeAs(cvt, udt_src)) {
LLDB_LOG_ERROR(GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize UdtSourceLineRecord record: {0}");
continue;
}
m_udt_declarations.try_emplace(
udt_src.UDT, UdtDeclaration{/*FileNameIndex=*/udt_src.SourceFile,
/*IsIpiIndex=*/true,
/*Line=*/udt_src.LineNumber});
} break;
case LF_UDT_MOD_SRC_LINE: {
UdtModSourceLineRecord udt_mod_src;
if (auto err = TypeDeserializer::deserializeAs(cvt, udt_mod_src)) {
LLDB_LOG_ERROR(
GetLog(LLDBLog::Symbols), std::move(err),
"Failed to deserialize UdtModSourceLineRecord record: {0}");
continue;
}
// Some types might be contributed by multiple modules. We assume that
// they all point to the same file and line because we can only provide
// one location.
m_udt_declarations.try_emplace(
udt_mod_src.UDT,
UdtDeclaration{/*FileNameIndex=*/udt_mod_src.SourceFile,
/*IsIpiIndex=*/false,
/*Line=*/udt_mod_src.LineNumber});
} break;
default:
break;
}
}
}
llvm::Expected<Declaration>
SymbolFileNativePDB::ResolveUdtDeclaration(PdbTypeSymId type_id) {
std::call_once(m_cached_udt_declarations, [this] { CacheUdtDeclarations(); });
auto it = m_udt_declarations.find(type_id.index);
if (it == m_udt_declarations.end())
return llvm::createStringError("no UDT declaration found");
llvm::StringRef file_name;
if (it->second.IsIpiIndex) {
CVType cvt = m_index->ipi().getType(it->second.FileNameIndex);
if (cvt.kind() != LF_STRING_ID)
return llvm::createStringError("file name was not a LF_STRING_ID");
StringIdRecord sid;
if (auto err = TypeDeserializer::deserializeAs(cvt, sid))
return std::move(err);
file_name = sid.String;
} else {
// The file name index is an index into the string table
auto string_table = m_index->pdb().getStringTable();
if (!string_table)
return string_table.takeError();
llvm::Expected<llvm::StringRef> string =
string_table->getStringTable().getString(
it->second.FileNameIndex.getIndex());
if (!string)
return string.takeError();
file_name = *string;
}
// rustc sets the filename to "<unknown>" for some files
if (file_name == "\\<unknown>")
return Declaration();
return Declaration(FileSpec(file_name), it->second.Line);
}