1 //===-- LLVMSymbolize.cpp -------------------------------------------------===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 // Implementation for LLVM symbolization library.
10 //
11 //===----------------------------------------------------------------------===//
12 
13 #include "llvm/DebugInfo/Symbolize/Symbolize.h"
14 
15 #include "SymbolizableObjectFile.h"
16 
17 #include "llvm/ADT/STLExtras.h"
18 #include "llvm/BinaryFormat/COFF.h"
19 #include "llvm/Config/config.h"
20 #include "llvm/DebugInfo/DWARF/DWARFContext.h"
21 #include "llvm/DebugInfo/PDB/PDB.h"
22 #include "llvm/DebugInfo/PDB/PDBContext.h"
23 #include "llvm/Demangle/Demangle.h"
24 #include "llvm/Object/COFF.h"
25 #include "llvm/Object/MachO.h"
26 #include "llvm/Object/MachOUniversal.h"
27 #include "llvm/Support/CRC.h"
28 #include "llvm/Support/Casting.h"
29 #include "llvm/Support/Compression.h"
30 #include "llvm/Support/DataExtractor.h"
31 #include "llvm/Support/Errc.h"
32 #include "llvm/Support/FileSystem.h"
33 #include "llvm/Support/MemoryBuffer.h"
34 #include "llvm/Support/Path.h"
35 #include <algorithm>
36 #include <cassert>
37 #include <cstring>
38 
39 namespace llvm {
40 namespace symbolize {
41 
42 Expected<DILineInfo>
symbolizeCodeCommon(SymbolizableModule * Info,object::SectionedAddress ModuleOffset)43 LLVMSymbolizer::symbolizeCodeCommon(SymbolizableModule *Info,
44                                     object::SectionedAddress ModuleOffset) {
45   // A null module means an error has already been reported. Return an empty
46   // result.
47   if (!Info)
48     return DILineInfo();
49 
50   // If the user is giving us relative addresses, add the preferred base of the
51   // object to the offset before we do the query. It's what DIContext expects.
52   if (Opts.RelativeAddresses)
53     ModuleOffset.Address += Info->getModulePreferredBase();
54 
55   DILineInfo LineInfo = Info->symbolizeCode(
56       ModuleOffset, DILineInfoSpecifier(Opts.PathStyle, Opts.PrintFunctions),
57       Opts.UseSymbolTable);
58   if (Opts.Demangle)
59     LineInfo.FunctionName = DemangleName(LineInfo.FunctionName, Info);
60   return LineInfo;
61 }
62 
63 Expected<DILineInfo>
symbolizeCode(const ObjectFile & Obj,object::SectionedAddress ModuleOffset)64 LLVMSymbolizer::symbolizeCode(const ObjectFile &Obj,
65                               object::SectionedAddress ModuleOffset) {
66   StringRef ModuleName = Obj.getFileName();
67   auto I = Modules.find(ModuleName);
68   if (I != Modules.end())
69     return symbolizeCodeCommon(I->second.get(), ModuleOffset);
70 
71   std::unique_ptr<DIContext> Context = DWARFContext::create(Obj);
72   Expected<SymbolizableModule *> InfoOrErr =
73                      createModuleInfo(&Obj, std::move(Context), ModuleName);
74   if (!InfoOrErr)
75     return InfoOrErr.takeError();
76   return symbolizeCodeCommon(*InfoOrErr, ModuleOffset);
77 }
78 
79 Expected<DILineInfo>
symbolizeCode(const std::string & ModuleName,object::SectionedAddress ModuleOffset)80 LLVMSymbolizer::symbolizeCode(const std::string &ModuleName,
81                               object::SectionedAddress ModuleOffset) {
82   Expected<SymbolizableModule *> InfoOrErr = getOrCreateModuleInfo(ModuleName);
83   if (!InfoOrErr)
84     return InfoOrErr.takeError();
85   return symbolizeCodeCommon(*InfoOrErr, ModuleOffset);
86 }
87 
88 Expected<DIInliningInfo>
symbolizeInlinedCode(const std::string & ModuleName,object::SectionedAddress ModuleOffset)89 LLVMSymbolizer::symbolizeInlinedCode(const std::string &ModuleName,
90                                      object::SectionedAddress ModuleOffset) {
91   SymbolizableModule *Info;
92   if (auto InfoOrErr = getOrCreateModuleInfo(ModuleName))
93     Info = InfoOrErr.get();
94   else
95     return InfoOrErr.takeError();
96 
97   // A null module means an error has already been reported. Return an empty
98   // result.
99   if (!Info)
100     return DIInliningInfo();
101 
102   // If the user is giving us relative addresses, add the preferred base of the
103   // object to the offset before we do the query. It's what DIContext expects.
104   if (Opts.RelativeAddresses)
105     ModuleOffset.Address += Info->getModulePreferredBase();
106 
107   DIInliningInfo InlinedContext = Info->symbolizeInlinedCode(
108       ModuleOffset, DILineInfoSpecifier(Opts.PathStyle, Opts.PrintFunctions),
109       Opts.UseSymbolTable);
110   if (Opts.Demangle) {
111     for (int i = 0, n = InlinedContext.getNumberOfFrames(); i < n; i++) {
112       auto *Frame = InlinedContext.getMutableFrame(i);
113       Frame->FunctionName = DemangleName(Frame->FunctionName, Info);
114     }
115   }
116   return InlinedContext;
117 }
118 
119 Expected<DIGlobal>
symbolizeData(const std::string & ModuleName,object::SectionedAddress ModuleOffset)120 LLVMSymbolizer::symbolizeData(const std::string &ModuleName,
121                               object::SectionedAddress ModuleOffset) {
122   SymbolizableModule *Info;
123   if (auto InfoOrErr = getOrCreateModuleInfo(ModuleName))
124     Info = InfoOrErr.get();
125   else
126     return InfoOrErr.takeError();
127 
128   // A null module means an error has already been reported. Return an empty
129   // result.
130   if (!Info)
131     return DIGlobal();
132 
133   // If the user is giving us relative addresses, add the preferred base of
134   // the object to the offset before we do the query. It's what DIContext
135   // expects.
136   if (Opts.RelativeAddresses)
137     ModuleOffset.Address += Info->getModulePreferredBase();
138 
139   DIGlobal Global = Info->symbolizeData(ModuleOffset);
140   if (Opts.Demangle)
141     Global.Name = DemangleName(Global.Name, Info);
142   return Global;
143 }
144 
145 Expected<std::vector<DILocal>>
symbolizeFrame(const std::string & ModuleName,object::SectionedAddress ModuleOffset)146 LLVMSymbolizer::symbolizeFrame(const std::string &ModuleName,
147                                object::SectionedAddress ModuleOffset) {
148   SymbolizableModule *Info;
149   if (auto InfoOrErr = getOrCreateModuleInfo(ModuleName))
150     Info = InfoOrErr.get();
151   else
152     return InfoOrErr.takeError();
153 
154   // A null module means an error has already been reported. Return an empty
155   // result.
156   if (!Info)
157     return std::vector<DILocal>();
158 
159   // If the user is giving us relative addresses, add the preferred base of
160   // the object to the offset before we do the query. It's what DIContext
161   // expects.
162   if (Opts.RelativeAddresses)
163     ModuleOffset.Address += Info->getModulePreferredBase();
164 
165   return Info->symbolizeFrame(ModuleOffset);
166 }
167 
flush()168 void LLVMSymbolizer::flush() {
169   ObjectForUBPathAndArch.clear();
170   BinaryForPath.clear();
171   ObjectPairForPathArch.clear();
172   Modules.clear();
173 }
174 
175 namespace {
176 
177 // For Path="/path/to/foo" and Basename="foo" assume that debug info is in
178 // /path/to/foo.dSYM/Contents/Resources/DWARF/foo.
179 // For Path="/path/to/bar.dSYM" and Basename="foo" assume that debug info is in
180 // /path/to/bar.dSYM/Contents/Resources/DWARF/foo.
getDarwinDWARFResourceForPath(const std::string & Path,const std::string & Basename)181 std::string getDarwinDWARFResourceForPath(
182     const std::string &Path, const std::string &Basename) {
183   SmallString<16> ResourceName = StringRef(Path);
184   if (sys::path::extension(Path) != ".dSYM") {
185     ResourceName += ".dSYM";
186   }
187   sys::path::append(ResourceName, "Contents", "Resources", "DWARF");
188   sys::path::append(ResourceName, Basename);
189   return std::string(ResourceName.str());
190 }
191 
checkFileCRC(StringRef Path,uint32_t CRCHash)192 bool checkFileCRC(StringRef Path, uint32_t CRCHash) {
193   ErrorOr<std::unique_ptr<MemoryBuffer>> MB =
194       MemoryBuffer::getFileOrSTDIN(Path);
195   if (!MB)
196     return false;
197   return CRCHash == llvm::crc32(arrayRefFromStringRef(MB.get()->getBuffer()));
198 }
199 
findDebugBinary(const std::string & OrigPath,const std::string & DebuglinkName,uint32_t CRCHash,const std::string & FallbackDebugPath,std::string & Result)200 bool findDebugBinary(const std::string &OrigPath,
201                      const std::string &DebuglinkName, uint32_t CRCHash,
202                      const std::string &FallbackDebugPath,
203                      std::string &Result) {
204   SmallString<16> OrigDir(OrigPath);
205   llvm::sys::path::remove_filename(OrigDir);
206   SmallString<16> DebugPath = OrigDir;
207   // Try relative/path/to/original_binary/debuglink_name
208   llvm::sys::path::append(DebugPath, DebuglinkName);
209   if (checkFileCRC(DebugPath, CRCHash)) {
210     Result = std::string(DebugPath.str());
211     return true;
212   }
213   // Try relative/path/to/original_binary/.debug/debuglink_name
214   DebugPath = OrigDir;
215   llvm::sys::path::append(DebugPath, ".debug", DebuglinkName);
216   if (checkFileCRC(DebugPath, CRCHash)) {
217     Result = std::string(DebugPath.str());
218     return true;
219   }
220   // Make the path absolute so that lookups will go to
221   // "/usr/lib/debug/full/path/to/debug", not
222   // "/usr/lib/debug/to/debug"
223   llvm::sys::fs::make_absolute(OrigDir);
224   if (!FallbackDebugPath.empty()) {
225     // Try <FallbackDebugPath>/absolute/path/to/original_binary/debuglink_name
226     DebugPath = FallbackDebugPath;
227   } else {
228 #if defined(__NetBSD__)
229     // Try /usr/libdata/debug/absolute/path/to/original_binary/debuglink_name
230     DebugPath = "/usr/libdata/debug";
231 #else
232     // Try /usr/lib/debug/absolute/path/to/original_binary/debuglink_name
233     DebugPath = "/usr/lib/debug";
234 #endif
235   }
236   llvm::sys::path::append(DebugPath, llvm::sys::path::relative_path(OrigDir),
237                           DebuglinkName);
238   if (checkFileCRC(DebugPath, CRCHash)) {
239     Result = std::string(DebugPath.str());
240     return true;
241   }
242   return false;
243 }
244 
getGNUDebuglinkContents(const ObjectFile * Obj,std::string & DebugName,uint32_t & CRCHash)245 bool getGNUDebuglinkContents(const ObjectFile *Obj, std::string &DebugName,
246                              uint32_t &CRCHash) {
247   if (!Obj)
248     return false;
249   for (const SectionRef &Section : Obj->sections()) {
250     StringRef Name;
251     if (Expected<StringRef> NameOrErr = Section.getName())
252       Name = *NameOrErr;
253     else
254       consumeError(NameOrErr.takeError());
255 
256     Name = Name.substr(Name.find_first_not_of("._"));
257     if (Name == "gnu_debuglink") {
258       Expected<StringRef> ContentsOrErr = Section.getContents();
259       if (!ContentsOrErr) {
260         consumeError(ContentsOrErr.takeError());
261         return false;
262       }
263       DataExtractor DE(*ContentsOrErr, Obj->isLittleEndian(), 0);
264       uint64_t Offset = 0;
265       if (const char *DebugNameStr = DE.getCStr(&Offset)) {
266         // 4-byte align the offset.
267         Offset = (Offset + 3) & ~0x3;
268         if (DE.isValidOffsetForDataOfSize(Offset, 4)) {
269           DebugName = DebugNameStr;
270           CRCHash = DE.getU32(&Offset);
271           return true;
272         }
273       }
274       break;
275     }
276   }
277   return false;
278 }
279 
darwinDsymMatchesBinary(const MachOObjectFile * DbgObj,const MachOObjectFile * Obj)280 bool darwinDsymMatchesBinary(const MachOObjectFile *DbgObj,
281                              const MachOObjectFile *Obj) {
282   ArrayRef<uint8_t> dbg_uuid = DbgObj->getUuid();
283   ArrayRef<uint8_t> bin_uuid = Obj->getUuid();
284   if (dbg_uuid.empty() || bin_uuid.empty())
285     return false;
286   return !memcmp(dbg_uuid.data(), bin_uuid.data(), dbg_uuid.size());
287 }
288 
289 template <typename ELFT>
getBuildID(const ELFFile<ELFT> * Obj)290 Optional<ArrayRef<uint8_t>> getBuildID(const ELFFile<ELFT> *Obj) {
291   if (!Obj)
292     return {};
293   auto PhdrsOrErr = Obj->program_headers();
294   if (!PhdrsOrErr) {
295     consumeError(PhdrsOrErr.takeError());
296     return {};
297   }
298   for (const auto &P : *PhdrsOrErr) {
299     if (P.p_type != ELF::PT_NOTE)
300       continue;
301     Error Err = Error::success();
302     for (auto N : Obj->notes(P, Err))
303       if (N.getType() == ELF::NT_GNU_BUILD_ID && N.getName() == ELF::ELF_NOTE_GNU)
304         return N.getDesc();
305     consumeError(std::move(Err));
306   }
307   return {};
308 }
309 
getBuildID(const ELFObjectFileBase * Obj)310 Optional<ArrayRef<uint8_t>> getBuildID(const ELFObjectFileBase *Obj) {
311   Optional<ArrayRef<uint8_t>> BuildID;
312   if (auto *O = dyn_cast<ELFObjectFile<ELF32LE>>(Obj))
313     BuildID = getBuildID(O->getELFFile());
314   else if (auto *O = dyn_cast<ELFObjectFile<ELF32BE>>(Obj))
315     BuildID = getBuildID(O->getELFFile());
316   else if (auto *O = dyn_cast<ELFObjectFile<ELF64LE>>(Obj))
317     BuildID = getBuildID(O->getELFFile());
318   else if (auto *O = dyn_cast<ELFObjectFile<ELF64BE>>(Obj))
319     BuildID = getBuildID(O->getELFFile());
320   else
321     llvm_unreachable("unsupported file format");
322   return BuildID;
323 }
324 
findDebugBinary(const std::vector<std::string> & DebugFileDirectory,const ArrayRef<uint8_t> BuildID,std::string & Result)325 bool findDebugBinary(const std::vector<std::string> &DebugFileDirectory,
326                      const ArrayRef<uint8_t> BuildID,
327                      std::string &Result) {
328   auto getDebugPath = [&](StringRef Directory) {
329     SmallString<128> Path{Directory};
330     sys::path::append(Path, ".build-id",
331                       llvm::toHex(BuildID[0], /*LowerCase=*/true),
332                       llvm::toHex(BuildID.slice(1), /*LowerCase=*/true));
333     Path += ".debug";
334     return Path;
335   };
336   if (DebugFileDirectory.empty()) {
337     SmallString<128> Path = getDebugPath(
338 #if defined(__NetBSD__)
339       // Try /usr/libdata/debug/.build-id/../...
340       "/usr/libdata/debug"
341 #else
342       // Try /usr/lib/debug/.build-id/../...
343       "/usr/lib/debug"
344 #endif
345     );
346     if (llvm::sys::fs::exists(Path)) {
347       Result = std::string(Path.str());
348       return true;
349     }
350   } else {
351     for (const auto &Directory : DebugFileDirectory) {
352       // Try <debug-file-directory>/.build-id/../...
353       SmallString<128> Path = getDebugPath(Directory);
354       if (llvm::sys::fs::exists(Path)) {
355         Result = std::string(Path.str());
356         return true;
357       }
358     }
359   }
360   return false;
361 }
362 
363 } // end anonymous namespace
364 
lookUpDsymFile(const std::string & ExePath,const MachOObjectFile * MachExeObj,const std::string & ArchName)365 ObjectFile *LLVMSymbolizer::lookUpDsymFile(const std::string &ExePath,
366     const MachOObjectFile *MachExeObj, const std::string &ArchName) {
367   // On Darwin we may find DWARF in separate object file in
368   // resource directory.
369   std::vector<std::string> DsymPaths;
370   StringRef Filename = sys::path::filename(ExePath);
371   DsymPaths.push_back(
372       getDarwinDWARFResourceForPath(ExePath, std::string(Filename)));
373   for (const auto &Path : Opts.DsymHints) {
374     DsymPaths.push_back(
375         getDarwinDWARFResourceForPath(Path, std::string(Filename)));
376   }
377   for (const auto &Path : DsymPaths) {
378     auto DbgObjOrErr = getOrCreateObject(Path, ArchName);
379     if (!DbgObjOrErr) {
380       // Ignore errors, the file might not exist.
381       consumeError(DbgObjOrErr.takeError());
382       continue;
383     }
384     ObjectFile *DbgObj = DbgObjOrErr.get();
385     if (!DbgObj)
386       continue;
387     const MachOObjectFile *MachDbgObj = dyn_cast<const MachOObjectFile>(DbgObj);
388     if (!MachDbgObj)
389       continue;
390     if (darwinDsymMatchesBinary(MachDbgObj, MachExeObj))
391       return DbgObj;
392   }
393   return nullptr;
394 }
395 
lookUpDebuglinkObject(const std::string & Path,const ObjectFile * Obj,const std::string & ArchName)396 ObjectFile *LLVMSymbolizer::lookUpDebuglinkObject(const std::string &Path,
397                                                   const ObjectFile *Obj,
398                                                   const std::string &ArchName) {
399   std::string DebuglinkName;
400   uint32_t CRCHash;
401   std::string DebugBinaryPath;
402   if (!getGNUDebuglinkContents(Obj, DebuglinkName, CRCHash))
403     return nullptr;
404   if (!findDebugBinary(Path, DebuglinkName, CRCHash, Opts.FallbackDebugPath,
405                        DebugBinaryPath))
406     return nullptr;
407   auto DbgObjOrErr = getOrCreateObject(DebugBinaryPath, ArchName);
408   if (!DbgObjOrErr) {
409     // Ignore errors, the file might not exist.
410     consumeError(DbgObjOrErr.takeError());
411     return nullptr;
412   }
413   return DbgObjOrErr.get();
414 }
415 
lookUpBuildIDObject(const std::string & Path,const ELFObjectFileBase * Obj,const std::string & ArchName)416 ObjectFile *LLVMSymbolizer::lookUpBuildIDObject(const std::string &Path,
417                                                 const ELFObjectFileBase *Obj,
418                                                 const std::string &ArchName) {
419   auto BuildID = getBuildID(Obj);
420   if (!BuildID)
421     return nullptr;
422   if (BuildID->size() < 2)
423     return nullptr;
424   std::string DebugBinaryPath;
425   if (!findDebugBinary(Opts.DebugFileDirectory, *BuildID, DebugBinaryPath))
426     return nullptr;
427   auto DbgObjOrErr = getOrCreateObject(DebugBinaryPath, ArchName);
428   if (!DbgObjOrErr) {
429     consumeError(DbgObjOrErr.takeError());
430     return nullptr;
431   }
432   return DbgObjOrErr.get();
433 }
434 
435 Expected<LLVMSymbolizer::ObjectPair>
getOrCreateObjectPair(const std::string & Path,const std::string & ArchName)436 LLVMSymbolizer::getOrCreateObjectPair(const std::string &Path,
437                                       const std::string &ArchName) {
438   auto I = ObjectPairForPathArch.find(std::make_pair(Path, ArchName));
439   if (I != ObjectPairForPathArch.end())
440     return I->second;
441 
442   auto ObjOrErr = getOrCreateObject(Path, ArchName);
443   if (!ObjOrErr) {
444     ObjectPairForPathArch.emplace(std::make_pair(Path, ArchName),
445                                   ObjectPair(nullptr, nullptr));
446     return ObjOrErr.takeError();
447   }
448 
449   ObjectFile *Obj = ObjOrErr.get();
450   assert(Obj != nullptr);
451   ObjectFile *DbgObj = nullptr;
452 
453   if (auto MachObj = dyn_cast<const MachOObjectFile>(Obj))
454     DbgObj = lookUpDsymFile(Path, MachObj, ArchName);
455   else if (auto ELFObj = dyn_cast<const ELFObjectFileBase>(Obj))
456     DbgObj = lookUpBuildIDObject(Path, ELFObj, ArchName);
457   if (!DbgObj)
458     DbgObj = lookUpDebuglinkObject(Path, Obj, ArchName);
459   if (!DbgObj)
460     DbgObj = Obj;
461   ObjectPair Res = std::make_pair(Obj, DbgObj);
462   ObjectPairForPathArch.emplace(std::make_pair(Path, ArchName), Res);
463   return Res;
464 }
465 
466 Expected<ObjectFile *>
getOrCreateObject(const std::string & Path,const std::string & ArchName)467 LLVMSymbolizer::getOrCreateObject(const std::string &Path,
468                                   const std::string &ArchName) {
469   Binary *Bin;
470   auto Pair = BinaryForPath.emplace(Path, OwningBinary<Binary>());
471   if (!Pair.second) {
472     Bin = Pair.first->second.getBinary();
473   } else {
474     Expected<OwningBinary<Binary>> BinOrErr = createBinary(Path);
475     if (!BinOrErr)
476       return BinOrErr.takeError();
477     Pair.first->second = std::move(BinOrErr.get());
478     Bin = Pair.first->second.getBinary();
479   }
480 
481   if (!Bin)
482     return static_cast<ObjectFile *>(nullptr);
483 
484   if (MachOUniversalBinary *UB = dyn_cast_or_null<MachOUniversalBinary>(Bin)) {
485     auto I = ObjectForUBPathAndArch.find(std::make_pair(Path, ArchName));
486     if (I != ObjectForUBPathAndArch.end())
487       return I->second.get();
488 
489     Expected<std::unique_ptr<ObjectFile>> ObjOrErr =
490         UB->getMachOObjectForArch(ArchName);
491     if (!ObjOrErr) {
492       ObjectForUBPathAndArch.emplace(std::make_pair(Path, ArchName),
493                                      std::unique_ptr<ObjectFile>());
494       return ObjOrErr.takeError();
495     }
496     ObjectFile *Res = ObjOrErr->get();
497     ObjectForUBPathAndArch.emplace(std::make_pair(Path, ArchName),
498                                    std::move(ObjOrErr.get()));
499     return Res;
500   }
501   if (Bin->isObject()) {
502     return cast<ObjectFile>(Bin);
503   }
504   return errorCodeToError(object_error::arch_not_found);
505 }
506 
507 Expected<SymbolizableModule *>
createModuleInfo(const ObjectFile * Obj,std::unique_ptr<DIContext> Context,StringRef ModuleName)508 LLVMSymbolizer::createModuleInfo(const ObjectFile *Obj,
509                                  std::unique_ptr<DIContext> Context,
510                                  StringRef ModuleName) {
511   auto InfoOrErr = SymbolizableObjectFile::create(Obj, std::move(Context),
512                                                   Opts.UntagAddresses);
513   std::unique_ptr<SymbolizableModule> SymMod;
514   if (InfoOrErr)
515     SymMod = std::move(*InfoOrErr);
516   auto InsertResult = Modules.insert(
517       std::make_pair(std::string(ModuleName), std::move(SymMod)));
518   assert(InsertResult.second);
519   if (!InfoOrErr)
520     return InfoOrErr.takeError();
521   return InsertResult.first->second.get();
522 }
523 
524 Expected<SymbolizableModule *>
getOrCreateModuleInfo(const std::string & ModuleName)525 LLVMSymbolizer::getOrCreateModuleInfo(const std::string &ModuleName) {
526   auto I = Modules.find(ModuleName);
527   if (I != Modules.end())
528     return I->second.get();
529 
530   std::string BinaryName = ModuleName;
531   std::string ArchName = Opts.DefaultArch;
532   size_t ColonPos = ModuleName.find_last_of(':');
533   // Verify that substring after colon form a valid arch name.
534   if (ColonPos != std::string::npos) {
535     std::string ArchStr = ModuleName.substr(ColonPos + 1);
536     if (Triple(ArchStr).getArch() != Triple::UnknownArch) {
537       BinaryName = ModuleName.substr(0, ColonPos);
538       ArchName = ArchStr;
539     }
540   }
541   auto ObjectsOrErr = getOrCreateObjectPair(BinaryName, ArchName);
542   if (!ObjectsOrErr) {
543     // Failed to find valid object file.
544     Modules.emplace(ModuleName, std::unique_ptr<SymbolizableModule>());
545     return ObjectsOrErr.takeError();
546   }
547   ObjectPair Objects = ObjectsOrErr.get();
548 
549   std::unique_ptr<DIContext> Context;
550   // If this is a COFF object containing PDB info, use a PDBContext to
551   // symbolize. Otherwise, use DWARF.
552   if (auto CoffObject = dyn_cast<COFFObjectFile>(Objects.first)) {
553     const codeview::DebugInfo *DebugInfo;
554     StringRef PDBFileName;
555     auto EC = CoffObject->getDebugPDBInfo(DebugInfo, PDBFileName);
556     if (!EC && DebugInfo != nullptr && !PDBFileName.empty()) {
557       using namespace pdb;
558       std::unique_ptr<IPDBSession> Session;
559 
560       PDB_ReaderType ReaderType = PDB_ReaderType::Native;
561 #if LLVM_ENABLE_DIA_SDK
562       if (!Opts.UseNativePDBReader)
563         ReaderType = PDB_ReaderType::DIA;
564 #endif
565       if (auto Err = loadDataForEXE(ReaderType, Objects.first->getFileName(),
566                                     Session)) {
567         Modules.emplace(ModuleName, std::unique_ptr<SymbolizableModule>());
568         // Return along the PDB filename to provide more context
569         return createFileError(PDBFileName, std::move(Err));
570       }
571       Context.reset(new PDBContext(*CoffObject, std::move(Session)));
572     }
573   }
574   if (!Context)
575     Context = DWARFContext::create(*Objects.second, nullptr, Opts.DWPName);
576   return createModuleInfo(Objects.first, std::move(Context), ModuleName);
577 }
578 
579 namespace {
580 
581 // Undo these various manglings for Win32 extern "C" functions:
582 // cdecl       - _foo
583 // stdcall     - _foo@12
584 // fastcall    - @foo@12
585 // vectorcall  - foo@@12
586 // These are all different linkage names for 'foo'.
demanglePE32ExternCFunc(StringRef SymbolName)587 StringRef demanglePE32ExternCFunc(StringRef SymbolName) {
588   // Remove any '_' or '@' prefix.
589   char Front = SymbolName.empty() ? '\0' : SymbolName[0];
590   if (Front == '_' || Front == '@')
591     SymbolName = SymbolName.drop_front();
592 
593   // Remove any '@[0-9]+' suffix.
594   if (Front != '?') {
595     size_t AtPos = SymbolName.rfind('@');
596     if (AtPos != StringRef::npos &&
597         std::all_of(SymbolName.begin() + AtPos + 1, SymbolName.end(),
598                     [](char C) { return C >= '0' && C <= '9'; })) {
599       SymbolName = SymbolName.substr(0, AtPos);
600     }
601   }
602 
603   // Remove any ending '@' for vectorcall.
604   if (SymbolName.endswith("@"))
605     SymbolName = SymbolName.drop_back();
606 
607   return SymbolName;
608 }
609 
610 } // end anonymous namespace
611 
612 std::string
DemangleName(const std::string & Name,const SymbolizableModule * DbiModuleDescriptor)613 LLVMSymbolizer::DemangleName(const std::string &Name,
614                              const SymbolizableModule *DbiModuleDescriptor) {
615   // We can spoil names of symbols with C linkage, so use an heuristic
616   // approach to check if the name should be demangled.
617   if (Name.substr(0, 2) == "_Z") {
618     int status = 0;
619     char *DemangledName = itaniumDemangle(Name.c_str(), nullptr, nullptr, &status);
620     if (status != 0)
621       return Name;
622     std::string Result = DemangledName;
623     free(DemangledName);
624     return Result;
625   }
626 
627   if (!Name.empty() && Name.front() == '?') {
628     // Only do MSVC C++ demangling on symbols starting with '?'.
629     int status = 0;
630     char *DemangledName = microsoftDemangle(
631         Name.c_str(), nullptr, nullptr, nullptr, &status,
632         MSDemangleFlags(MSDF_NoAccessSpecifier | MSDF_NoCallingConvention |
633                         MSDF_NoMemberType | MSDF_NoReturnType));
634     if (status != 0)
635       return Name;
636     std::string Result = DemangledName;
637     free(DemangledName);
638     return Result;
639   }
640 
641   if (DbiModuleDescriptor && DbiModuleDescriptor->isWin32Module())
642     return std::string(demanglePE32ExternCFunc(Name));
643   return Name;
644 }
645 
646 } // namespace symbolize
647 } // namespace llvm
648