7 files changed, 132 insertions, 32 deletions
diff --git a/include/llvm/DebugInfo/PDB/Native/DbiModuleDescriptor.h b/include/llvm/DebugInfo/PDB/Native/DbiModuleDescriptor.h
index d1f791b9daede..7e77f5a3eef92 100644
--- a/include/llvm/DebugInfo/PDB/Native/DbiModuleDescriptor.h
+++ b/include/llvm/DebugInfo/PDB/Native/DbiModuleDescriptor.h
@@ -53,14 +53,6 @@ private:
   const ModuleInfoHeader *Layout = nullptr;
 };
 
-struct ModuleInfoEx {
-  ModuleInfoEx(const DbiModuleDescriptor &Info) : Info(Info) {}
-  ModuleInfoEx(const ModuleInfoEx &Ex) = default;
-
-  DbiModuleDescriptor Info;
-  std::vector<StringRef> SourceFiles;
-};
-
 } // end namespace pdb
 
 template <> struct VarStreamArrayExtractor<pdb::DbiModuleDescriptor> {
diff --git a/include/llvm/DebugInfo/PDB/Native/DbiModuleList.h b/include/llvm/DebugInfo/PDB/Native/DbiModuleList.h
new file mode 100644
index 0000000000000..bcf1cff8f6e5d
--- /dev/null
+++ b/include/llvm/DebugInfo/PDB/Native/DbiModuleList.h
@@ -0,0 +1,114 @@
+//===- DbiModuleList.h - PDB module information list ------------*- C++ -*-===//
+//
+//                     The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_DEBUGINFO_PDB_RAW_DBIMODULELIST_H
+#define LLVM_DEBUGINFO_PDB_RAW_DBIMODULELIST_H
+
+#include "llvm/ADT/StringRef.h"
+#include "llvm/ADT/iterator.h"
+#include "llvm/DebugInfo/PDB/Native/DbiModuleDescriptor.h"
+#include "llvm/Support/BinaryStreamArray.h"
+#include "llvm/Support/BinaryStreamRef.h"
+#include "llvm/Support/Endian.h"
+#include "llvm/Support/Error.h"
+#include <cstdint>
+#include <vector>
+
+namespace llvm {
+namespace pdb {
+
+class DbiModuleList;
+struct FileInfoSubstreamHeader;
+
+class DbiModuleSourceFilesIterator
+    : public iterator_facade_base<DbiModuleSourceFilesIterator,
+                                  std::random_access_iterator_tag, StringRef> {
+  typedef iterator_facade_base<DbiModuleSourceFilesIterator,
+                               std::random_access_iterator_tag, StringRef>
+      BaseType;
+
+public:
+  DbiModuleSourceFilesIterator(const DbiModuleList &Modules, uint32_t Modi,
+                               uint16_t Filei);
+  DbiModuleSourceFilesIterator() = default;
+  DbiModuleSourceFilesIterator &
+  operator=(const DbiModuleSourceFilesIterator &R) = default;
+
+  bool operator==(const DbiModuleSourceFilesIterator &R) const;
+
+  const StringRef &operator*() const { return ThisValue; }
+  StringRef &operator*() { return ThisValue; }
+
+  bool operator<(const DbiModuleSourceFilesIterator &RHS) const;
+  std::ptrdiff_t operator-(const DbiModuleSourceFilesIterator &R) const;
+  DbiModuleSourceFilesIterator &operator+=(std::ptrdiff_t N);
+  DbiModuleSourceFilesIterator &operator-=(std::ptrdiff_t N);
+
+private:
+  void setValue();
+
+  bool isEnd() const;
+  bool isCompatible(const DbiModuleSourceFilesIterator &R) const;
+  bool isUniversalEnd() const;
+
+  StringRef ThisValue;
+  const DbiModuleList *Modules{nullptr};
+  uint32_t Modi{0};
+  uint16_t Filei{0};
+};
+
+class DbiModuleList {
+  friend DbiModuleSourceFilesIterator;
+
+public:
+  Error initialize(BinaryStreamRef ModInfo, BinaryStreamRef FileInfo);
+
+  Expected<StringRef> getFileName(uint32_t Index) const;
+  uint32_t getModuleCount() const;
+  uint32_t getSourceFileCount() const;
+  uint16_t getSourceFileCount(uint32_t Modi) const;
+
+  iterator_range<DbiModuleSourceFilesIterator>
+  source_files(uint32_t Modi) const;
+
+  DbiModuleDescriptor getModuleDescriptor(uint32_t Modi) const;
+
+private:
+  Error initializeModInfo(BinaryStreamRef ModInfo);
+  Error initializeFileInfo(BinaryStreamRef FileInfo);
+
+  VarStreamArray<DbiModuleDescriptor> Descriptors;
+
+  FixedStreamArray<support::little32_t> FileNameOffsets;
+  FixedStreamArray<support::ulittle16_t> ModFileCountArray;
+
+  // For each module, there are multiple filenames, which can be obtained by
+  // knowing the index of the file.  Given the index of the file, one can use
+  // that as an offset into the FileNameOffsets array, which contains the
+  // absolute offset of the file name in NamesBuffer.  Thus, for each module
+  // we store the first index in the FileNameOffsets array for this module.
+  // The number of files for the corresponding module is stored in
+  // ModFileCountArray.
+  std::vector<uint32_t> ModuleInitialFileIndex;
+
+  // In order to provide random access into the Descriptors array, we iterate it
+  // once up front to find the offsets of the individual items and store them in
+  // this array.
+  std::vector<uint32_t> ModuleDescriptorOffsets;
+
+  const FileInfoSubstreamHeader *FileInfoHeader = nullptr;
+
+  BinaryStreamRef ModInfoSubstream;
+  BinaryStreamRef FileInfoSubstream;
+  BinaryStreamRef NamesBuffer;
+};
+}
+}
+
+#endif // LLVM_DEBUGINFO_PDB_RAW_DBIMODULELIST_H
+\ No newline at end of file
diff --git a/include/llvm/DebugInfo/PDB/Native/DbiStream.h b/include/llvm/DebugInfo/PDB/Native/DbiStream.h
index 08262e47f77f5..8f95481f41521 100644
--- a/include/llvm/DebugInfo/PDB/Native/DbiStream.h
+++ b/include/llvm/DebugInfo/PDB/Native/DbiStream.h
@@ -13,6 +13,7 @@
 #include "llvm/DebugInfo/CodeView/ModuleDebugFragment.h"
 #include "llvm/DebugInfo/MSF/MappedBlockStream.h"
 #include "llvm/DebugInfo/PDB/Native/DbiModuleDescriptor.h"
+#include "llvm/DebugInfo/PDB/Native/DbiModuleList.h"
 #include "llvm/DebugInfo/PDB/Native/PDBStringTable.h"
 #include "llvm/DebugInfo/PDB/Native/RawConstants.h"
 #include "llvm/DebugInfo/PDB/Native/RawTypes.h"
@@ -68,9 +69,7 @@ public:
   /// not present, returns InvalidStreamIndex.
   uint32_t getDebugStreamIndex(DbgHeaderType Type) const;
 
-  ArrayRef<ModuleInfoEx> modules() const;
-
-  Expected<StringRef> getFileNameForIndex(uint32_t Index) const;
+  const DbiModuleList &modules() const;
 
   FixedStreamArray<object::coff_section> getSectionHeaders();
 
@@ -80,27 +79,22 @@ public:
   void visitSectionContributions(ISectionContribVisitor &Visitor) const;
 
 private:
-  Error initializeModInfoArray();
   Error initializeSectionContributionData();
   Error initializeSectionHeadersData();
   Error initializeSectionMapData();
-  Error initializeFileInfo();
   Error initializeFpoRecords();
 
   PDBFile &Pdb;
   std::unique_ptr<msf::MappedBlockStream> Stream;
 
-  std::vector<ModuleInfoEx> ModuleInfos;
   PDBStringTable ECNames;
 
-  BinaryStreamRef ModInfoSubstream;
   BinaryStreamRef SecContrSubstream;
   BinaryStreamRef SecMapSubstream;
-  BinaryStreamRef FileInfoSubstream;
   BinaryStreamRef TypeServerMapSubstream;
   BinaryStreamRef ECSubstream;
 
-  BinaryStreamRef NamesBuffer;
+  DbiModuleList Modules;
 
   FixedStreamArray<support::ulittle16_t> DbgStreams;
 
@@ -108,7 +102,6 @@ private:
   FixedStreamArray<SectionContrib> SectionContribs;
   FixedStreamArray<SectionContrib2> SectionContribs2;
   FixedStreamArray<SecMapEntry> SectionMap;
-  FixedStreamArray<support::little32_t> FileNameOffsets;
 
   std::unique_ptr<msf::MappedBlockStream> SectionHeaderStream;
   FixedStreamArray<object::coff_section> SectionHeaders;
diff --git a/include/llvm/DebugInfo/PDB/Native/NativeCompilandSymbol.h b/include/llvm/DebugInfo/PDB/Native/NativeCompilandSymbol.h
index b1d980679a455..22ed61910d94a 100644
--- a/include/llvm/DebugInfo/PDB/Native/NativeCompilandSymbol.h
+++ b/include/llvm/DebugInfo/PDB/Native/NativeCompilandSymbol.h
@@ -18,7 +18,7 @@ namespace pdb {
 
 class NativeCompilandSymbol : public NativeRawSymbol {
 public:
-  NativeCompilandSymbol(NativeSession &Session, const ModuleInfoEx &MI);
+  NativeCompilandSymbol(NativeSession &Session, DbiModuleDescriptor MI);
   PDB_SymType getSymTag() const override;
   bool isEditAndContinueEnabled() const override;
   uint32_t getLexicalParentId() const override;
@@ -26,7 +26,7 @@ public:
   std::string getName() const override;
 
 private:
-  ModuleInfoEx Module;
+  DbiModuleDescriptor Module;
 };
 
 } // namespace pdb
diff --git a/include/llvm/DebugInfo/PDB/Native/NativeEnumModules.h b/include/llvm/DebugInfo/PDB/Native/NativeEnumModules.h
index 18022f599bbaf..6aa1460dbb4e6 100644
--- a/include/llvm/DebugInfo/PDB/Native/NativeEnumModules.h
+++ b/include/llvm/DebugInfo/PDB/Native/NativeEnumModules.h
@@ -16,13 +16,13 @@
 namespace llvm {
 namespace pdb {
 
+class DbiModuleList;
 class NativeSession;
 
 class NativeEnumModules : public IPDBEnumChildren<PDBSymbol> {
 public:
-  explicit NativeEnumModules(NativeSession &Session,
-                             ArrayRef<ModuleInfoEx> Modules,
-                             uint32_t Index = 0);
+  NativeEnumModules(NativeSession &Session, const DbiModuleList &Modules,
+                    uint32_t Index = 0);
 
   uint32_t getChildCount() const override;
   std::unique_ptr<PDBSymbol> getChildAtIndex(uint32_t Index) const override;
@@ -32,7 +32,7 @@ public:
 
 private:
   NativeSession &Session;
-  ArrayRef<ModuleInfoEx> Modules;
+  const DbiModuleList &Modules;
   uint32_t Index;
 };
 }
diff --git a/include/llvm/DebugInfo/PDB/Native/RawTypes.h b/include/llvm/DebugInfo/PDB/Native/RawTypes.h
index 93622d0a43949..979b8454dd5ec 100644
--- a/include/llvm/DebugInfo/PDB/Native/RawTypes.h
+++ b/include/llvm/DebugInfo/PDB/Native/RawTypes.h
@@ -211,7 +211,7 @@ struct ModInfoFlags {
 };
 
 /// The header preceeding each entry in the Module Info substream of the DBI
-/// stream.
+/// stream.  Corresponds to the type MODI in the reference implementation.
 struct ModuleInfoHeader {
   /// Currently opened module. This field is a pointer in the reference
   /// implementation, but that won't work on 64-bit systems, and anyway it
@@ -243,9 +243,12 @@ struct ModuleInfoHeader {
   /// Padding so the next field is 4-byte aligned.
   char Padding1[2];
 
-  /// Array of [0..NumFiles) DBI name buffer offsets.  This field is a pointer
-  /// in the reference implementation, but as with `Mod`, we ignore it for now
-  /// since it is unused.
+  /// Array of [0..NumFiles) DBI name buffer offsets.  In the reference
+  /// implementation this field is a pointer.  But since you can't portably
+  /// serialize a pointer, on 64-bit platforms they copy all the values except
+  /// this one into the 32-bit version of the struct and use that for
+  /// serialization.  Regardless, this field is unused, it is only there to
+  /// store a pointer that can be accessed at runtime.
   support::ulittle32_t FileNameOffs;
 
   /// Name Index for src file name
diff --git a/include/llvm/DebugInfo/PDB/Native/TpiStream.h b/include/llvm/DebugInfo/PDB/Native/TpiStream.h
index 62dde0ef08b76..9fef9bee5e1a4 100644
--- a/include/llvm/DebugInfo/PDB/Native/TpiStream.h
+++ b/include/llvm/DebugInfo/PDB/Native/TpiStream.h
@@ -40,12 +40,12 @@ public:
 
   uint32_t TypeIndexBegin() const;
   uint32_t TypeIndexEnd() const;
-  uint32_t NumTypeRecords() const;
+  uint32_t getNumTypeRecords() const;
   uint16_t getTypeHashStreamIndex() const;
   uint16_t getTypeHashStreamAuxIndex() const;
 
   uint32_t getHashKeySize() const;
-  uint32_t NumHashBuckets() const;
+  uint32_t getNumHashBuckets() const;
   FixedStreamArray<support::ulittle32_t> getHashValues() const;
   FixedStreamArray<TypeIndexOffset> getTypeIndexOffsets() const;
   HashTable &getHashAdjusters();
@@ -55,8 +55,6 @@ public:
   Error commit();
 
 private:
-  Error verifyHashValues();
-
   const PDBFile &Pdb;
   std::unique_ptr<msf::MappedBlockStream> Stream;