1 //===-- SymbolFilePDB.cpp ---------------------------------------*- C++ -*-===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 
10 #include "SymbolFilePDB.h"
11 
12 #include "clang/Lex/Lexer.h"
13 
14 #include "lldb/Core/Module.h"
15 #include "lldb/Core/PluginManager.h"
16 #include "lldb/Symbol/ClangASTContext.h"
17 #include "lldb/Symbol/CompileUnit.h"
18 #include "lldb/Symbol/LineTable.h"
19 #include "lldb/Symbol/ObjectFile.h"
20 #include "lldb/Symbol/SymbolContext.h"
21 #include "lldb/Symbol/TypeMap.h"
22 
23 #include "llvm/DebugInfo/PDB/IPDBEnumChildren.h"
24 #include "llvm/DebugInfo/PDB/IPDBLineNumber.h"
25 #include "llvm/DebugInfo/PDB/IPDBSourceFile.h"
26 #include "llvm/DebugInfo/PDB/PDBSymbol.h"
27 #include "llvm/DebugInfo/PDB/PDBSymbolCompiland.h"
28 #include "llvm/DebugInfo/PDB/PDBSymbolCompilandDetails.h"
29 #include "llvm/DebugInfo/PDB/PDBSymbolExe.h"
30 #include "llvm/DebugInfo/PDB/PDBSymbolFunc.h"
31 #include "llvm/DebugInfo/PDB/PDBSymbolFuncDebugEnd.h"
32 #include "llvm/DebugInfo/PDB/PDBSymbolFuncDebugStart.h"
33 #include "llvm/DebugInfo/PDB/PDBSymbolTypeEnum.h"
34 #include "llvm/DebugInfo/PDB/PDBSymbolTypeTypedef.h"
35 #include "llvm/DebugInfo/PDB/PDBSymbolTypeUDT.h"
36 
37 #include "Plugins/SymbolFile/PDB/PDBASTParser.h"
38 
39 #include <regex>
40 
41 using namespace lldb_private;
42 
43 namespace
44 {
45     lldb::LanguageType TranslateLanguage(llvm::PDB_Lang lang)
46     {
47         switch (lang)
48         {
49         case llvm::PDB_Lang::Cpp:
50             return lldb::LanguageType::eLanguageTypeC_plus_plus;
51         case llvm::PDB_Lang::C:
52             return lldb::LanguageType::eLanguageTypeC;
53         default:
54             return lldb::LanguageType::eLanguageTypeUnknown;
55         }
56     }
57 
58     bool
59     ShouldAddLine(uint32_t requested_line, uint32_t actual_line, uint32_t addr_length)
60     {
61         return ((requested_line == 0 || actual_line == requested_line) && addr_length > 0);
62     }
63 }
64 
65 void
66 SymbolFilePDB::Initialize()
67 {
68     PluginManager::RegisterPlugin(GetPluginNameStatic(), GetPluginDescriptionStatic(), CreateInstance,
69                                   DebuggerInitialize);
70 }
71 
72 void
73 SymbolFilePDB::Terminate()
74 {
75     PluginManager::UnregisterPlugin(CreateInstance);
76 }
77 
78 void
79 SymbolFilePDB::DebuggerInitialize(lldb_private::Debugger &debugger)
80 {
81 }
82 
83 lldb_private::ConstString
84 SymbolFilePDB::GetPluginNameStatic()
85 {
86     static ConstString g_name("pdb");
87     return g_name;
88 }
89 
90 const char *
91 SymbolFilePDB::GetPluginDescriptionStatic()
92 {
93     return "Microsoft PDB debug symbol file reader.";
94 }
95 
96 lldb_private::SymbolFile *
97 SymbolFilePDB::CreateInstance(lldb_private::ObjectFile *obj_file)
98 {
99     return new SymbolFilePDB(obj_file);
100 }
101 
102 SymbolFilePDB::SymbolFilePDB(lldb_private::ObjectFile *object_file)
103     : SymbolFile(object_file), m_cached_compile_unit_count(0)
104 {
105 }
106 
107 SymbolFilePDB::~SymbolFilePDB()
108 {
109 }
110 
111 uint32_t
112 SymbolFilePDB::CalculateAbilities()
113 {
114     if (!m_session_up)
115     {
116         // Lazily load and match the PDB file, but only do this once.
117         std::string exePath = m_obj_file->GetFileSpec().GetPath();
118         auto error = llvm::loadDataForEXE(llvm::PDB_ReaderType::DIA, llvm::StringRef(exePath), m_session_up);
119         if (error != llvm::PDB_ErrorCode::Success)
120             return 0;
121     }
122     return CompileUnits | LineTables;
123 }
124 
125 void
126 SymbolFilePDB::InitializeObject()
127 {
128     lldb::addr_t obj_load_address = m_obj_file->GetFileOffset();
129     m_session_up->setLoadAddress(obj_load_address);
130 
131     TypeSystem *type_system = GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
132     ClangASTContext *clang_type_system = llvm::dyn_cast_or_null<ClangASTContext>(type_system);
133     m_tu_decl_ctx_up = llvm::make_unique<CompilerDeclContext>(type_system, clang_type_system->GetTranslationUnitDecl());
134 }
135 
136 uint32_t
137 SymbolFilePDB::GetNumCompileUnits()
138 {
139     if (m_cached_compile_unit_count == 0)
140     {
141         auto global = m_session_up->getGlobalScope();
142         auto compilands = global->findAllChildren<llvm::PDBSymbolCompiland>();
143         m_cached_compile_unit_count = compilands->getChildCount();
144 
145         // The linker can inject an additional "dummy" compilation unit into the PDB.
146         // Ignore this special compile unit for our purposes, if it is there.  It is
147         // always the last one.
148         auto last_cu = compilands->getChildAtIndex(m_cached_compile_unit_count - 1);
149         std::string name = last_cu->getName();
150         if (name == "* Linker *")
151             --m_cached_compile_unit_count;
152     }
153     return m_cached_compile_unit_count;
154 }
155 
156 lldb::CompUnitSP
157 SymbolFilePDB::ParseCompileUnitAtIndex(uint32_t index)
158 {
159     auto global = m_session_up->getGlobalScope();
160     auto compilands = global->findAllChildren<llvm::PDBSymbolCompiland>();
161     auto cu = compilands->getChildAtIndex(index);
162 
163     uint32_t id = cu->getSymIndexId();
164 
165     return ParseCompileUnitForSymIndex(id);
166 }
167 
168 lldb::LanguageType
169 SymbolFilePDB::ParseCompileUnitLanguage(const lldb_private::SymbolContext &sc)
170 {
171     // What fields should I expect to be filled out on the SymbolContext?  Is it
172     // safe to assume that `sc.comp_unit` is valid?
173     if (!sc.comp_unit)
174         return lldb::eLanguageTypeUnknown;
175 
176     auto cu = m_session_up->getConcreteSymbolById<llvm::PDBSymbolCompiland>(sc.comp_unit->GetID());
177     if (!cu)
178         return lldb::eLanguageTypeUnknown;
179     auto details = cu->findOneChild<llvm::PDBSymbolCompilandDetails>();
180     if (!details)
181         return lldb::eLanguageTypeUnknown;
182     return TranslateLanguage(details->getLanguage());
183 }
184 
185 size_t
186 SymbolFilePDB::ParseCompileUnitFunctions(const lldb_private::SymbolContext &sc)
187 {
188     // TODO: Implement this
189     return size_t();
190 }
191 
192 bool
193 SymbolFilePDB::ParseCompileUnitLineTable(const lldb_private::SymbolContext &sc)
194 {
195     return ParseCompileUnitLineTable(sc, 0);
196 }
197 
198 bool
199 SymbolFilePDB::ParseCompileUnitDebugMacros(const lldb_private::SymbolContext &sc)
200 {
201     // PDB doesn't contain information about macros
202     return false;
203 }
204 
205 bool
206 SymbolFilePDB::ParseCompileUnitSupportFiles(const lldb_private::SymbolContext &sc,
207                                             lldb_private::FileSpecList &support_files)
208 {
209     if (!sc.comp_unit)
210         return false;
211 
212     // In theory this is unnecessary work for us, because all of this information is easily
213     // (and quickly) accessible from DebugInfoPDB, so caching it a second time seems like a waste.
214     // Unfortunately, there's no good way around this short of a moderate refactor, since SymbolVendor
215     // depends on being able to cache this list.
216     auto cu = m_session_up->getConcreteSymbolById<llvm::PDBSymbolCompiland>(sc.comp_unit->GetID());
217     if (!cu)
218         return false;
219     auto files = m_session_up->getSourceFilesForCompiland(*cu);
220     if (!files || files->getChildCount() == 0)
221         return false;
222 
223     while (auto file = files->getNext())
224     {
225         FileSpec spec(file->getFileName(), false);
226         support_files.Append(spec);
227     }
228     return true;
229 }
230 
231 bool
232 SymbolFilePDB::ParseImportedModules(const lldb_private::SymbolContext &sc,
233                                     std::vector<lldb_private::ConstString> &imported_modules)
234 {
235     // PDB does not yet support module debug info
236     return false;
237 }
238 
239 size_t
240 SymbolFilePDB::ParseFunctionBlocks(const lldb_private::SymbolContext &sc)
241 {
242     // TODO: Implement this
243     return size_t();
244 }
245 
246 size_t
247 SymbolFilePDB::ParseTypes(const lldb_private::SymbolContext &sc)
248 {
249     // TODO: Implement this
250     return size_t();
251 }
252 
253 size_t
254 SymbolFilePDB::ParseVariablesForContext(const lldb_private::SymbolContext &sc)
255 {
256     // TODO: Implement this
257     return size_t();
258 }
259 
260 lldb_private::Type *
261 SymbolFilePDB::ResolveTypeUID(lldb::user_id_t type_uid)
262 {
263     auto find_result = m_types.find(type_uid);
264     if (find_result != m_types.end())
265         return find_result->second.get();
266 
267     TypeSystem *type_system = GetTypeSystemForLanguage(lldb::eLanguageTypeC_plus_plus);
268     ClangASTContext *clang_type_system = llvm::dyn_cast_or_null<ClangASTContext>(type_system);
269     if (!clang_type_system)
270         return nullptr;
271     PDBASTParser *pdb = llvm::dyn_cast<PDBASTParser>(clang_type_system->GetPDBParser());
272     if (!pdb)
273         return nullptr;
274 
275     auto pdb_type = m_session_up->getSymbolById(type_uid);
276     if (pdb_type == nullptr)
277         return nullptr;
278 
279     lldb::TypeSP result = pdb->CreateLLDBTypeFromPDBType(*pdb_type);
280     m_types.insert(std::make_pair(type_uid, result));
281     return result.get();
282 }
283 
284 bool
285 SymbolFilePDB::CompleteType(lldb_private::CompilerType &compiler_type)
286 {
287     // TODO: Implement this
288     return false;
289 }
290 
291 lldb_private::CompilerDecl
292 SymbolFilePDB::GetDeclForUID(lldb::user_id_t uid)
293 {
294     return lldb_private::CompilerDecl();
295 }
296 
297 lldb_private::CompilerDeclContext
298 SymbolFilePDB::GetDeclContextForUID(lldb::user_id_t uid)
299 {
300     // PDB always uses the translation unit decl context for everything.  We can improve this later
301     // but it's not easy because PDB doesn't provide a high enough level of type fidelity in this area.
302     return *m_tu_decl_ctx_up;
303 }
304 
305 lldb_private::CompilerDeclContext
306 SymbolFilePDB::GetDeclContextContainingUID(lldb::user_id_t uid)
307 {
308     return *m_tu_decl_ctx_up;
309 }
310 
311 void
312 SymbolFilePDB::ParseDeclsForContext(lldb_private::CompilerDeclContext decl_ctx)
313 {
314 }
315 
316 uint32_t
317 SymbolFilePDB::ResolveSymbolContext(const lldb_private::Address &so_addr, uint32_t resolve_scope,
318                                     lldb_private::SymbolContext &sc)
319 {
320     return uint32_t();
321 }
322 
323 uint32_t
324 SymbolFilePDB::ResolveSymbolContext(const lldb_private::FileSpec &file_spec, uint32_t line, bool check_inlines,
325                                     uint32_t resolve_scope, lldb_private::SymbolContextList &sc_list)
326 {
327     if (resolve_scope & lldb::eSymbolContextCompUnit)
328     {
329         // Locate all compilation units with line numbers referencing the specified file.  For example, if
330         // `file_spec` is <vector>, then this should return all source files and header files that reference
331         // <vector>, either directly or indirectly.
332         auto compilands =
333             m_session_up->findCompilandsForSourceFile(file_spec.GetPath(), llvm::PDB_NameSearchFlags::NS_CaseInsensitive);
334 
335         // For each one, either find get its previously parsed data, or parse it afresh and add it to
336         // the symbol context list.
337         while (auto compiland = compilands->getNext())
338         {
339             // If we're not checking inlines, then don't add line information for this file unless the FileSpec
340             // matches.
341             if (!check_inlines)
342             {
343                 // `getSourceFileName` returns the basename of the original source file used to generate this compiland.
344                 // It does not return the full path.  Currently the only way to get that is to do a basename lookup to
345                 // get the IPDBSourceFile, but this is ambiguous in the case of two source files with the same name
346                 // contributing to the same compiland.  This is a moderately extreme edge case, so we consider this ok
347                 // for now, although we need to find a long term solution.
348                 std::string source_file = compiland->getSourceFileName();
349                 auto pdb_file = m_session_up->findOneSourceFile(compiland.get(), source_file,
350                                                                 llvm::PDB_NameSearchFlags::NS_CaseInsensitive);
351                 source_file = pdb_file->getFileName();
352                 FileSpec this_spec(source_file, false, FileSpec::ePathSyntaxWindows);
353                 if (!file_spec.FileEquals(this_spec))
354                     continue;
355             }
356 
357             SymbolContext sc;
358             auto cu = ParseCompileUnitForSymIndex(compiland->getSymIndexId());
359             sc.comp_unit = cu.get();
360             sc.module_sp = cu->GetModule();
361             sc_list.Append(sc);
362 
363             // If we were asked to resolve line entries, add all entries to the line table that match the requested
364             // line (or all lines if `line` == 0)
365             if (resolve_scope & lldb::eSymbolContextLineEntry)
366                 ParseCompileUnitLineTable(sc, line);
367         }
368     }
369     return sc_list.GetSize();
370 }
371 
372 uint32_t
373 SymbolFilePDB::FindGlobalVariables(const lldb_private::ConstString &name,
374                                    const lldb_private::CompilerDeclContext *parent_decl_ctx, bool append,
375                                    uint32_t max_matches, lldb_private::VariableList &variables)
376 {
377     return uint32_t();
378 }
379 
380 uint32_t
381 SymbolFilePDB::FindGlobalVariables(const lldb_private::RegularExpression &regex, bool append, uint32_t max_matches,
382                                    lldb_private::VariableList &variables)
383 {
384     return uint32_t();
385 }
386 
387 uint32_t
388 SymbolFilePDB::FindFunctions(const lldb_private::ConstString &name,
389                              const lldb_private::CompilerDeclContext *parent_decl_ctx, uint32_t name_type_mask,
390                              bool include_inlines, bool append, lldb_private::SymbolContextList &sc_list)
391 {
392     return uint32_t();
393 }
394 
395 uint32_t
396 SymbolFilePDB::FindFunctions(const lldb_private::RegularExpression &regex, bool include_inlines, bool append,
397                              lldb_private::SymbolContextList &sc_list)
398 {
399     return uint32_t();
400 }
401 
402 void
403 SymbolFilePDB::GetMangledNamesForFunction(const std::string &scope_qualified_name,
404                                           std::vector<lldb_private::ConstString> &mangled_names)
405 {
406 }
407 
408 uint32_t
409 SymbolFilePDB::FindTypes(const lldb_private::SymbolContext &sc, const lldb_private::ConstString &name,
410                          const lldb_private::CompilerDeclContext *parent_decl_ctx, bool append, uint32_t max_matches,
411                          llvm::DenseSet<lldb_private::SymbolFile *> &searched_symbol_files,
412                          lldb_private::TypeMap &types)
413 {
414     if (!append)
415         types.Clear();
416     if (!name)
417         return 0;
418 
419     searched_symbol_files.clear();
420     searched_symbol_files.insert(this);
421 
422     std::string name_str = name.AsCString();
423 
424     // If this might be a regex, we have to return EVERY symbol and process them one by one, which is going
425     // to destroy performance on large PDB files.  So try really hard not to use a regex match.
426     if (name_str.find_first_of("[]?*.-+\\") != std::string::npos)
427         FindTypesByRegex(name_str, max_matches, types);
428     else
429         FindTypesByName(name_str, max_matches, types);
430     return types.GetSize();
431 }
432 
433 void
434 SymbolFilePDB::FindTypesByRegex(const std::string &regex, uint32_t max_matches, lldb_private::TypeMap &types)
435 {
436     // When searching by regex, we need to go out of our way to limit the search space as much as possible, since
437     // the way this is implemented is by searching EVERYTHING in the PDB and manually doing a regex compare.  PDB
438     // library isn't optimized for regex searches or searches across multiple symbol types at the same time, so the
439     // best we can do is to search enums, then typedefs, then classes one by one, and do a regex compare against all
440     // of them.
441     llvm::PDB_SymType tags_to_search[] = {llvm::PDB_SymType::Enum, llvm::PDB_SymType::Typedef, llvm::PDB_SymType::UDT};
442     auto global = m_session_up->getGlobalScope();
443     std::unique_ptr<llvm::IPDBEnumSymbols> results;
444 
445     std::regex re(regex);
446 
447     uint32_t matches = 0;
448 
449     for (auto tag : tags_to_search)
450     {
451         results = global->findAllChildren(tag);
452         while (auto result = results->getNext())
453         {
454             if (max_matches > 0 && matches >= max_matches)
455                 break;
456 
457             std::string type_name;
458             if (auto enum_type = llvm::dyn_cast<llvm::PDBSymbolTypeEnum>(result.get()))
459                 type_name = enum_type->getName();
460             else if (auto typedef_type = llvm::dyn_cast<llvm::PDBSymbolTypeTypedef>(result.get()))
461                 type_name = typedef_type->getName();
462             else if (auto class_type = llvm::dyn_cast<llvm::PDBSymbolTypeUDT>(result.get()))
463                 type_name = class_type->getName();
464             else
465             {
466                 // We're only looking for types that have names.  Skip symbols, as well as
467                 // unnamed types such as arrays, pointers, etc.
468                 continue;
469             }
470 
471             if (!std::regex_match(type_name, re))
472                 continue;
473 
474             // This should cause the type to get cached and stored in the `m_types` lookup.
475             if (!ResolveTypeUID(result->getSymIndexId()))
476                 continue;
477 
478             auto iter = m_types.find(result->getSymIndexId());
479             if (iter == m_types.end())
480                 continue;
481             types.Insert(iter->second);
482             ++matches;
483         }
484     }
485 }
486 
487 void
488 SymbolFilePDB::FindTypesByName(const std::string &name, uint32_t max_matches, lldb_private::TypeMap &types)
489 {
490     auto global = m_session_up->getGlobalScope();
491     std::unique_ptr<llvm::IPDBEnumSymbols> results;
492     results = global->findChildren(llvm::PDB_SymType::None, name.c_str(), llvm::PDB_NameSearchFlags::NS_Default);
493 
494     uint32_t matches = 0;
495 
496     while (auto result = results->getNext())
497     {
498         if (max_matches > 0 && matches >= max_matches)
499             break;
500         switch (result->getSymTag())
501         {
502             case llvm::PDB_SymType::Enum:
503             case llvm::PDB_SymType::UDT:
504             case llvm::PDB_SymType::Typedef:
505                 break;
506             default:
507                 // We're only looking for types that have names.  Skip symbols, as well as
508                 // unnamed types such as arrays, pointers, etc.
509                 continue;
510         }
511 
512         // This should cause the type to get cached and stored in the `m_types` lookup.
513         if (!ResolveTypeUID(result->getSymIndexId()))
514             continue;
515 
516         auto iter = m_types.find(result->getSymIndexId());
517         if (iter == m_types.end())
518             continue;
519         types.Insert(iter->second);
520         ++matches;
521     }
522 }
523 
524 size_t
525 SymbolFilePDB::FindTypes(const std::vector<lldb_private::CompilerContext> &contexts, bool append,
526                          lldb_private::TypeMap &types)
527 {
528     return 0;
529 }
530 
531 lldb_private::TypeList *
532 SymbolFilePDB::GetTypeList()
533 {
534     return nullptr;
535 }
536 
537 size_t
538 SymbolFilePDB::GetTypes(lldb_private::SymbolContextScope *sc_scope, uint32_t type_mask,
539                         lldb_private::TypeList &type_list)
540 {
541     return size_t();
542 }
543 
544 lldb_private::TypeSystem *
545 SymbolFilePDB::GetTypeSystemForLanguage(lldb::LanguageType language)
546 {
547     auto type_system = m_obj_file->GetModule()->GetTypeSystemForLanguage(language);
548     if (type_system)
549         type_system->SetSymbolFile(this);
550     return type_system;
551 }
552 
553 lldb_private::CompilerDeclContext
554 SymbolFilePDB::FindNamespace(const lldb_private::SymbolContext &sc, const lldb_private::ConstString &name,
555                              const lldb_private::CompilerDeclContext *parent_decl_ctx)
556 {
557     return lldb_private::CompilerDeclContext();
558 }
559 
560 lldb_private::ConstString
561 SymbolFilePDB::GetPluginName()
562 {
563     static ConstString g_name("pdb");
564     return g_name;
565 }
566 
567 uint32_t
568 SymbolFilePDB::GetPluginVersion()
569 {
570     return 1;
571 }
572 
573 llvm::IPDBSession &
574 SymbolFilePDB::GetPDBSession()
575 {
576     return *m_session_up;
577 }
578 
579 const llvm::IPDBSession &
580 SymbolFilePDB::GetPDBSession() const
581 {
582     return *m_session_up;
583 }
584 
585 lldb::CompUnitSP
586 SymbolFilePDB::ParseCompileUnitForSymIndex(uint32_t id)
587 {
588     auto found_cu = m_comp_units.find(id);
589     if (found_cu != m_comp_units.end())
590         return found_cu->second;
591 
592     auto cu = m_session_up->getConcreteSymbolById<llvm::PDBSymbolCompiland>(id);
593 
594     // `getSourceFileName` returns the basename of the original source file used to generate this compiland.  It does
595     // not return the full path.  Currently the only way to get that is to do a basename lookup to get the
596     // IPDBSourceFile, but this is ambiguous in the case of two source files with the same name contributing to the
597     // same compiland. This is a moderately extreme edge case, so we consider this ok for now, although we need to find
598     // a long term solution.
599     auto file = m_session_up->findOneSourceFile(cu.get(), cu->getSourceFileName(),
600                                                 llvm::PDB_NameSearchFlags::NS_CaseInsensitive);
601     std::string path = file->getFileName();
602 
603     lldb::LanguageType lang;
604     auto details = cu->findOneChild<llvm::PDBSymbolCompilandDetails>();
605     if (!details)
606         lang = lldb::eLanguageTypeC_plus_plus;
607     else
608         lang = TranslateLanguage(details->getLanguage());
609 
610     // Don't support optimized code for now, DebugInfoPDB does not return this information.
611     bool optimized = false;
612     auto result = std::make_shared<CompileUnit>(m_obj_file->GetModule(), nullptr, path.c_str(), id, lang, optimized);
613     m_comp_units.insert(std::make_pair(id, result));
614     return result;
615 }
616 
617 bool
618 SymbolFilePDB::ParseCompileUnitLineTable(const lldb_private::SymbolContext &sc, uint32_t match_line)
619 {
620     auto global = m_session_up->getGlobalScope();
621     auto cu = m_session_up->getConcreteSymbolById<llvm::PDBSymbolCompiland>(sc.comp_unit->GetID());
622 
623     // LineEntry needs the *index* of the file into the list of support files returned by
624     // ParseCompileUnitSupportFiles.  But the underlying SDK gives us a globally unique
625     // idenfitifier in the namespace of the PDB.  So, we have to do a mapping so that we
626     // can hand out indices.
627     llvm::DenseMap<uint32_t, uint32_t> index_map;
628     BuildSupportFileIdToSupportFileIndexMap(*cu, index_map);
629     auto line_table = llvm::make_unique<LineTable>(sc.comp_unit);
630 
631     // Find contributions to `cu` from all source and header files.
632     std::string path = sc.comp_unit->GetPath();
633     auto files = m_session_up->getSourceFilesForCompiland(*cu);
634 
635     // For each source and header file, create a LineSequence for contributions to the cu
636     // from that file, and add the sequence.
637     while (auto file = files->getNext())
638     {
639         std::unique_ptr<LineSequence> sequence(line_table->CreateLineSequenceContainer());
640         auto lines = m_session_up->findLineNumbers(*cu, *file);
641         int entry_count = lines->getChildCount();
642 
643         uint64_t prev_addr;
644         uint32_t prev_length;
645         uint32_t prev_line;
646         uint32_t prev_source_idx;
647 
648         for (int i = 0; i < entry_count; ++i)
649         {
650             auto line = lines->getChildAtIndex(i);
651 
652             uint64_t lno = line->getLineNumber();
653             uint64_t addr = line->getVirtualAddress();
654             uint32_t length = line->getLength();
655             uint32_t source_id = line->getSourceFileId();
656             uint32_t col = line->getColumnNumber();
657             uint32_t source_idx = index_map[source_id];
658 
659             // There was a gap between the current entry and the previous entry if the addresses don't perfectly line
660             // up.
661             bool is_gap = (i > 0) && (prev_addr + prev_length < addr);
662 
663             // Before inserting the current entry, insert a terminal entry at the end of the previous entry's address
664             // range if the current entry resulted in a gap from the previous entry.
665             if (is_gap && ShouldAddLine(match_line, prev_line, prev_length))
666             {
667                 line_table->AppendLineEntryToSequence(sequence.get(), prev_addr + prev_length, prev_line, 0,
668                                                       prev_source_idx, false, false, false, false, true);
669             }
670 
671             if (ShouldAddLine(match_line, lno, length))
672             {
673                 bool is_statement = line->isStatement();
674                 bool is_prologue = false;
675                 bool is_epilogue = false;
676                 auto func = m_session_up->findSymbolByAddress(addr, llvm::PDB_SymType::Function);
677                 if (func)
678                 {
679                     auto prologue = func->findOneChild<llvm::PDBSymbolFuncDebugStart>();
680                     is_prologue = (addr == prologue->getVirtualAddress());
681 
682                     auto epilogue = func->findOneChild<llvm::PDBSymbolFuncDebugEnd>();
683                     is_epilogue = (addr == epilogue->getVirtualAddress());
684                 }
685 
686                 line_table->AppendLineEntryToSequence(sequence.get(), addr, lno, col, source_idx, is_statement, false,
687                                                       is_prologue, is_epilogue, false);
688             }
689 
690             prev_addr = addr;
691             prev_length = length;
692             prev_line = lno;
693             prev_source_idx = source_idx;
694         }
695 
696         if (entry_count > 0 && ShouldAddLine(match_line, prev_line, prev_length))
697         {
698             // The end is always a terminal entry, so insert it regardless.
699             line_table->AppendLineEntryToSequence(sequence.get(), prev_addr + prev_length, prev_line, 0,
700                                                   prev_source_idx, false, false, false, false, true);
701         }
702 
703         line_table->InsertSequence(sequence.release());
704     }
705 
706     sc.comp_unit->SetLineTable(line_table.release());
707     return true;
708 }
709 
710 void
711 SymbolFilePDB::BuildSupportFileIdToSupportFileIndexMap(const llvm::PDBSymbolCompiland &cu,
712                                                        llvm::DenseMap<uint32_t, uint32_t> &index_map) const
713 {
714     // This is a hack, but we need to convert the source id into an index into the support
715     // files array.  We don't want to do path comparisons to avoid basename / full path
716     // issues that may or may not even be a problem, so we use the globally unique source
717     // file identifiers.  Ideally we could use the global identifiers everywhere, but LineEntry
718     // currently assumes indices.
719     auto source_files = m_session_up->getSourceFilesForCompiland(cu);
720     int index = 0;
721 
722     while (auto file = source_files->getNext())
723     {
724         uint32_t source_id = file->getUniqueId();
725         index_map[source_id] = index++;
726     }
727 }
728