[PDB] Output first section contribution for each module

[zturner]

Summary

One piece of the DBI stream is a list of module descriptors that include information about object file name, Stream index of the module's symbols, etc.

One piece of this descriptor that we have been leaving blank is the section contribution structure. Although I'm not entirely sure what part of the debugger uses this (if any) it's important that we try to emit something correct to minimize the risk of strange hard-to-diagnose failures.

Comments in the refernece source indicate that this is supposed be the *first* section contribution for the module. So, we simply emit a copy of the same structure that we emit for the section contribution list, but only for the first chunk for a given module.

Comparing LLD and link PDBs side by side, the results are now very similar.

[smeenai]

Typo in the summary: refernece -> reference

[rnk]

lgtm

lld/COFF/PDB.cpp

public:
 
   Expected<const CVIndexMap&> maybeMergeTypeServerPDB(ObjFile *File,
                                                       TypeServer2Record &TS);
 
   /// Add the section map and section contributions to the PDB.
   void addSections(ArrayRef<OutputSection *> OutputSections,
                    ArrayRef<uint8_t> SectionTable);
 
-  void addSectionContrib(pdb::DbiModuleDescriptorBuilder &LinkerModule,
-                         OutputSection *OS, Chunk *C);
-
   /// Write the PDB to disk.
   void commit();
 
 private:
   BumpPtrAllocator Alloc;
 
   SymbolTable *Symtab;
 
static ArrayRef<uint8_t> relocateDebugChunk(BumpPtrAllocator &Alloc,
   uint8_t *Buffer = Alloc.Allocate<uint8_t>(DebugChunk->getSize());
   assert(DebugChunk->OutputSectionOff == 0 &&
          "debug sections should not be in output sections");
   DebugChunk->writeTo(Buffer);
   return consumeDebugMagic(makeArrayRef(Buffer, DebugChunk->getSize()),
                            ".debug$S");
 }
 
+static pdb::SectionContrib createSectionContrib(const Chunk *C, uint32_t Modi) {
+  OutputSection *OS = C->getOutputSection();
+  pdb::SectionContrib SC;
+  memset(&SC, 0, sizeof(SC));
+  SC.ISect = OS->SectionIndex;
+  SC.Off = C->getRVA() - OS->getRVA();
+  SC.Size = C->getSize();
+  if (auto *SecChunk = dyn_cast<SectionChunk>(C)) {
+    SC.Characteristics = SecChunk->Header->Characteristics;
+    SC.Imod = SecChunk->File->ModuleDBI->getModuleIndex();
+    ArrayRef<uint8_t> Contents = SecChunk->getContents();
+    JamCRC CRC(0);
+    ArrayRef<char> CharContents = makeArrayRef(
+        reinterpret_cast<const char *>(Contents.data()), Contents.size());
+    CRC.update(CharContents);
+    SC.DataCrc = CRC.getCRC();
+  } else {
+    SC.Characteristics = OS->getCharacteristics();
+    // FIXME: When we start creating DBI for import libraries, use those here.
+    SC.Imod = Modi;
+  }
+  SC.RelocCrc = 0; // FIXME
+
+  return SC;
+}
+
 void PDBLinker::addObjFile(ObjFile *File) {
   // Add a module descriptor for every object file. We need to put an absolute
   // path to the object into the PDB. If this is a plain object, we make its
   // path absolute. If it's an object in an archive, we make the archive path
   // absolute.
   bool InArchive = !File->ParentName.empty();
   SmallString<128> Path = InArchive ? File->ParentName : File->getName();
   sys::fs::make_absolute(Path);
   sys::path::native(Path, sys::path::Style::windows);
   StringRef Name = InArchive ? File->getName() : StringRef(Path);
 
   File->ModuleDBI = &ExitOnErr(Builder.getDbiBuilder().addModuleInfo(Name));
   File->ModuleDBI->setObjFileName(Path);
 
+  auto Chunks = File->getChunks();
+  assert(!Chunks.empty());
+  uint32_t Modi = File->ModuleDBI->getModuleIndex();
+  for (Chunk *C : Chunks) {
+    auto *SecChunk = dyn_cast<SectionChunk>(C);
+    if (!SecChunk || !SecChunk->isLive())
+      continue;
+    pdb::SectionContrib SC = createSectionContrib(SecChunk, Modi);
+    File->ModuleDBI->setFirstSectionContrib(SC);
+    break;
+  }
+
   // Before we can process symbol substreams from .debug$S, we need to process
   // type information, file checksums, and the string table.  Add type info to
   // the PDB first, so that we can get the map from object file type and item
   // indices to PDB type and item indices.
   CVIndexMap ObjectIndexMap;
   auto IndexMapResult = mergeDebugT(File, ObjectIndexMap);
 
   // If the .debug$T sections fail to merge, assume there is no debug info.
void PDBLinker::initialize(const llvm::codeview::DebugInfo &BuildId) {
   DbiBuilder.setMachineType(Config->Machine);
   // Technically we are not link.exe 14.11, but there are known cases where
   // debugging tools on Windows expect Microsoft-specific version numbers or
   // they fail to work at all.  Since we know we produce PDBs that are
   // compatible with LINK 14.11, we set that version number here.
   DbiBuilder.setBuildNumber(14, 11);
 }
 
-void PDBLinker::addSectionContrib(pdb::DbiModuleDescriptorBuilder &LinkerModule,
-                                  OutputSection *OS, Chunk *C) {
-  pdb::SectionContrib SC;
-  memset(&SC, 0, sizeof(SC));
-  SC.ISect = OS->SectionIndex;
-  SC.Off = C->getRVA() - OS->getRVA();
-  SC.Size = C->getSize();
-  if (auto *SecChunk = dyn_cast<SectionChunk>(C)) {
-    SC.Characteristics = SecChunk->Header->Characteristics;
-    SC.Imod = SecChunk->File->ModuleDBI->getModuleIndex();
-    ArrayRef<uint8_t> Contents = SecChunk->getContents();
-    JamCRC CRC(0);
-    ArrayRef<char> CharContents = makeArrayRef(
-        reinterpret_cast<const char *>(Contents.data()), Contents.size());
-    CRC.update(CharContents);
-    SC.DataCrc = CRC.getCRC();
-  } else {
-    SC.Characteristics = OS->getCharacteristics();
-    // FIXME: When we start creating DBI for import libraries, use those here.
-    SC.Imod = LinkerModule.getModuleIndex();
-  }
-  SC.RelocCrc = 0; // FIXME
-  Builder.getDbiBuilder().addSectionContrib(SC);
-}
-
 void PDBLinker::addSections(ArrayRef<OutputSection *> OutputSections,
                             ArrayRef<uint8_t> SectionTable) {
   // It's not entirely clear what this is, but the * Linker * module uses it.
   pdb::DbiStreamBuilder &DbiBuilder = Builder.getDbiBuilder();
   NativePath = Config->PDBPath;
   sys::fs::make_absolute(NativePath);
   sys::path::native(NativePath, sys::path::Style::windows);
   uint32_t PdbFilePathNI = DbiBuilder.addECName(NativePath);
   auto &LinkerModule = ExitOnErr(DbiBuilder.addModuleInfo("* Linker *"));
   LinkerModule.setPdbFilePathNI(PdbFilePathNI);
   addCommonLinkerModuleSymbols(NativePath, LinkerModule, Alloc);
 
   // Add section contributions. They must be ordered by ascending RVA.
   for (OutputSection *OS : OutputSections) {
     addLinkerModuleSectionSymbol(LinkerModule, *OS, Alloc);
-    for (Chunk *C : OS->getChunks())
-      addSectionContrib(LinkerModule, OS, C);
+    for (Chunk *C : OS->getChunks()) {
+      pdb::SectionContrib SC =
+          createSectionContrib(C, LinkerModule.getModuleIndex());
+      Builder.getDbiBuilder().addSectionContrib(SC);
+    }
   }
 
   // Add Section Map stream.
   ArrayRef<object::coff_section> Sections = {
       (const object::coff_section *)SectionTable.data(),
       SectionTable.size() / sizeof(object::coff_section)};
   SectionMap = pdb::DbiStreamBuilder::createSectionMap(Sections);
   DbiBuilder.setSectionMap(SectionMap);

lld/test/COFF/pdb.test

 # CHECK-NEXT:         String:          ret42-sub.c
 # CHECK-NEXT:     - Kind:            LF_BUILDINFO
 # CHECK-NEXT:       BuildInfo:
 # CHECK-NEXT:         ArgIndices:      [ 4098, 4099, 4106, 4104, 4102 ]
 
 RAW:                               Modules
 RAW-NEXT: ============================================================
 RAW-NEXT:   Mod 0000 | `{{.*}}pdb.test.tmp1.obj`:
-RAW-NEXT:              SC[???]  | mod = 0, 0000:0000, size = 0, data crc = 0, reloc crc = 0
-RAW-NEXT:                         none
+RAW-NEXT:              SC[.text]  | mod = 0, 0001:0000, size = 14, data crc = 1682752513, reloc crc = 0
+RAW-NEXT:                           IMAGE_SCN_CNT_CODE | IMAGE_SCN_ALIGN_16BYTES | IMAGE_SCN_MEM_EXECUTE |
+RAW-NEXT:                           IMAGE_SCN_MEM_READ
 RAW-NEXT:              Obj: `{{.*}}pdb.test.tmp1.obj`:
 RAW-NEXT:              debug stream: 11, # files: 1, has ec info: false
 RAW-NEXT:              pdb file ni: 0 ``, src file ni: 0 ``
 RAW-NEXT:   Mod 0001 | `{{.*}}pdb.test.tmp2.obj`:
-RAW-NEXT:              SC[???]  | mod = 1, 0000:0000, size = 0, data crc = 0, reloc crc = 0
-RAW-NEXT:                         none
+RAW-NEXT:              SC[.text]  | mod = 1, 0001:0016, size = 6, data crc = 2139436471, reloc crc = 0
+RAW-NEXT:                           IMAGE_SCN_CNT_CODE | IMAGE_SCN_ALIGN_16BYTES | IMAGE_SCN_MEM_EXECUTE |
+RAW-NEXT:                           IMAGE_SCN_MEM_READ
 RAW-NEXT:              Obj: `{{.*}}pdb.test.tmp2.obj`:
 RAW-NEXT:              debug stream: 12, # files: 1, has ec info: false
 RAW-NEXT:              pdb file ni: 0 ``, src file ni: 0 ``
 RAW-NEXT:   Mod 0002 | `* Linker *`:
 RAW-NEXT:              SC[???]  | mod = 2, 0000:0000, size = 0, data crc = 0, reloc crc = 0
 RAW-NEXT:                         none
 RAW-NEXT:              Obj: ``:
 RAW-NEXT:              debug stream: 13, # files: 0, has ec info: false

llvm/include/llvm/DebugInfo/PDB/Native/DbiModuleDescriptorBuilder.h

public:
   ~DbiModuleDescriptorBuilder();
 
   DbiModuleDescriptorBuilder(const DbiModuleDescriptorBuilder &) = delete;
   DbiModuleDescriptorBuilder &
   operator=(const DbiModuleDescriptorBuilder &) = delete;
 
   void setPdbFilePathNI(uint32_t NI);
   void setObjFileName(StringRef Name);
+  void setFirstSectionContrib(const SectionContrib &SC);
   void addSymbol(codeview::CVSymbol Symbol);
 
   void
   addDebugSubsection(std::shared_ptr<codeview::DebugSubsection> Subsection);
 
   void
   addDebugSubsection(const codeview::DebugSubsectionRecord &SubsectionContents);
 

llvm/lib/DebugInfo/PDB/Native/DbiModuleDescriptorBuilder.cpp

 void DbiModuleDescriptorBuilder::setObjFileName(StringRef Name) {
   ObjFileName = Name;
 }
 
 void DbiModuleDescriptorBuilder::setPdbFilePathNI(uint32_t NI) {
   PdbFilePathNI = NI;
 }
 
+void DbiModuleDescriptorBuilder::setFirstSectionContrib(
+    const SectionContrib &SC) {
+  Layout.SC = SC;
+}
+
 void DbiModuleDescriptorBuilder::addSymbol(CVSymbol Symbol) {
   Symbols.push_back(Symbol);
   // Symbols written to a PDB file are required to be 4 byte aligned.  The same
   // is not true of object files.
   assert(Symbol.length() % alignOf(CodeViewContainer::Pdb) == 0 &&
          "Invalid Symbol alignment!");
   SymbolByteSize += Symbol.length();
 }