Optimize unqualified/global name lookup in modules by introducing a

generational scheme for identifiers that avoids searching the hash
tables of a given module more than once for a given
identifier. Previously, loading any new module invalidated all of the
previous lookup results for all identifiers, causing us to perform the
lookups repeatedly.

llvm-svn: 148412

clang/include/clang/Serialization/ASTReader.h

@@ -579,6 +579,10 @@ private:
   /// \brief Whether to disable the use of stat caches in AST files.
   bool DisableStatCache;
 
+  /// \brief The current "generation" of the module file import stack, which 
+  /// indicates how many separate module file load operations have occurred.
+  unsigned CurrentGeneration;
+
   /// \brief Mapping from switch-case IDs in the chain to switch-case statements
   ///
   /// Statements usually don't have IDs, but switch cases need them, so that the
@@ -652,6 +656,9 @@ private:
   /// loaded once the recursive loading has completed.
   std::deque<PendingIdentifierInfo> PendingIdentifierInfos;
 
+  /// \brief The generation number of 
+  llvm::DenseMap<IdentifierInfo *, unsigned> IdentifierGeneration;
+  
   /// \brief Contains declarations and definitions that will be
   /// "interesting" to the ASTConsumer, when we get that AST consumer.
   ///
@@ -1432,6 +1439,9 @@ public:
   /// \brief Update an out-of-date identifier.
   virtual void updateOutOfDateIdentifier(IdentifierInfo &II);
 
+  /// \brief Note that this identifier is up-to-date.
+  void markIdentifierUpToDate(IdentifierInfo *II);
+  
   /// \brief Read the macro definition corresponding to this iterator
   /// into the unread macro record offsets table.
   void LoadMacroDefinition(

clang/include/clang/Serialization/Module.h

@@ -57,7 +57,7 @@ struct DeclContextInfo {
 /// other modules.
 class ModuleFile {
 public:
-  ModuleFile(ModuleKind Kind);
+  ModuleFile(ModuleKind Kind, unsigned Generation);
   ~ModuleFile();
 
   // === General information ===
@@ -72,6 +72,9 @@ public:
   /// user.
   bool DirectlyImported;
 
+  /// \brief The generation of which this module file is a part.
+  unsigned Generation;
+  
   /// \brief The memory buffer that stores the data associated with
   /// this AST file.
   llvm::OwningPtr<llvm::MemoryBuffer> Buffer;

clang/include/clang/Serialization/ModuleManager.h

@@ -95,6 +95,8 @@ public:
   /// \param ImportedBy The module that is importing this module, or NULL if
   /// this module is imported directly by the user.
   ///
+  /// \param Generation The generation in which this module was loaded.
+  ///
   /// \param ErrorStr Will be set to a non-empty string if any errors occurred
   /// while trying to load the module.
   ///
@@ -102,7 +104,7 @@ public:
   /// and a boolean indicating whether the module was newly added.
   std::pair<ModuleFile *, bool> 
   addModule(StringRef FileName, ModuleKind Type, ModuleFile *ImportedBy,
-            std::string &ErrorStr);
+            unsigned Generation, std::string &ErrorStr);
   
   /// \brief Add an in-memory buffer the list of known buffers
   void addInMemoryBuffer(StringRef FileName, llvm::MemoryBuffer *Buffer);

clang/lib/Serialization/ASTReader.cpp

@@ -514,7 +514,7 @@ IdentifierInfo *ASTIdentifierLookupTrait::ReadData(const internal_key_type& k,
       II = &Reader.getIdentifierTable().getOwn(StringRef(k.first, k.second));
     Reader.SetIdentifierInfo(ID, II);
     II->setIsFromAST();
-    II->setOutOfDate(false);
+    Reader.markIdentifierUpToDate(II);    
     return II;
   }
 
@@ -540,7 +540,7 @@ IdentifierInfo *ASTIdentifierLookupTrait::ReadData(const internal_key_type& k,
   IdentifierInfo *II = KnownII;
   if (!II)
     II = &Reader.getIdentifierTable().getOwn(StringRef(k.first, k.second));
-  II->setOutOfDate(false);
+  Reader.markIdentifierUpToDate(II);
   II->setIsFromAST();
 
   // Set or check the various bits in the IdentifierInfo structure.
@@ -1560,14 +1560,20 @@ namespace {
   /// \brief Visitor class used to look up identifirs in an AST file.
   class IdentifierLookupVisitor {
     StringRef Name;
+    unsigned PriorGeneration;
     IdentifierInfo *Found;
   public:
-    explicit IdentifierLookupVisitor(StringRef Name) : Name(Name), Found() { }
+    IdentifierLookupVisitor(StringRef Name, unsigned PriorGeneration) 
+      : Name(Name), PriorGeneration(PriorGeneration), Found() { }
     
     static bool visit(ModuleFile &M, void *UserData) {
       IdentifierLookupVisitor *This
         = static_cast<IdentifierLookupVisitor *>(UserData);
       
+      // If we've already searched this module file, skip it now.
+      if (M.Generation <= This->PriorGeneration)
+        return true;
+      
       ASTIdentifierLookupTable *IdTable
         = (ASTIdentifierLookupTable *)M.IdentifierLookupTable;
       if (!IdTable)
@@ -1593,7 +1599,24 @@ namespace {
 }
 
 void ASTReader::updateOutOfDateIdentifier(IdentifierInfo &II) {
-  get(II.getName());
+  unsigned PriorGeneration = 0;
+  if (getContext().getLangOptions().Modules)
+    PriorGeneration = IdentifierGeneration[&II];
+  
+  IdentifierLookupVisitor Visitor(II.getName(), PriorGeneration);
+  ModuleMgr.visit(IdentifierLookupVisitor::visit, &Visitor);
+  markIdentifierUpToDate(&II);
+}
+
+void ASTReader::markIdentifierUpToDate(IdentifierInfo *II) {
+  if (!II)
+    return;
+  
+  II->setOutOfDate(false);
+
+  // Update the generation for this identifier.
+  if (getContext().getLangOptions().Modules)
+    IdentifierGeneration[II] = CurrentGeneration;
 }
 
 const FileEntry *ASTReader::getFileEntry(StringRef filenameStrRef) {
@@ -2612,6 +2635,9 @@ void ASTReader::makeModuleVisible(Module *Mod,
 
 ASTReader::ASTReadResult ASTReader::ReadAST(const std::string &FileName,
                                             ModuleKind Type) {
+  // Bump the generation number.
+  ++CurrentGeneration;
+  
   switch(ReadASTCore(FileName, Type, /*ImportedBy=*/0)) {
   case Failure: return Failure;
   case IgnorePCH: return IgnorePCH;
@@ -2619,7 +2645,7 @@ ASTReader::ASTReadResult ASTReader::ReadAST(const std::string &FileName,
   }
 
   // Here comes stuff that we only do once the entire chain is loaded.
-  
+
   // Check the predefines buffers.
   if (!DisableValidation && Type == MK_PCH &&
       // FIXME: CheckPredefinesBuffers also sets the SuggestedPredefines;
@@ -2635,7 +2661,7 @@ ASTReader::ASTReadResult ASTReader::ReadAST(const std::string &FileName,
                               IdEnd = PP.getIdentifierTable().end();
        Id != IdEnd; ++Id)
     Id->second->setOutOfDate(true);
-
+  
   // Resolve any unresolved module exports.
   for (unsigned I = 0, N = UnresolvedModuleImportExports.size(); I != N; ++I) {
     UnresolvedModuleImportExport &Unresolved = UnresolvedModuleImportExports[I];
@@ -2683,7 +2709,7 @@ ASTReader::ASTReadResult ASTReader::ReadASTCore(StringRef FileName,
   bool NewModule;
   std::string ErrorStr;
   llvm::tie(M, NewModule) = ModuleMgr.addModule(FileName, Type, ImportedBy,
-                                                ErrorStr);
+                                                CurrentGeneration, ErrorStr);
 
   if (!M) {
     // We couldn't load the module.
@@ -5109,11 +5135,11 @@ void ASTReader::InitializeSema(Sema &S) {
 }
 
 IdentifierInfo* ASTReader::get(const char *NameStart, const char *NameEnd) {
-  IdentifierLookupVisitor Visitor(StringRef(NameStart, NameEnd - NameStart));
+  IdentifierLookupVisitor Visitor(StringRef(NameStart, NameEnd - NameStart),
+                                  /*PriorGeneration=*/0);
   ModuleMgr.visit(IdentifierLookupVisitor::visit, &Visitor);
   IdentifierInfo *II = Visitor.getIdentifierInfo();
-  if (II)
-    II->setOutOfDate(false);
+  markIdentifierUpToDate(II);
   return II;
 }
 
@@ -6236,7 +6262,8 @@ ASTReader::ASTReader(Preprocessor &PP, ASTContext &Context,
     Consumer(0), ModuleMgr(FileMgr.getFileSystemOptions()),
     RelocatablePCH(false), isysroot(isysroot),
     DisableValidation(DisableValidation),
-    DisableStatCache(DisableStatCache), NumStatHits(0), NumStatMisses(0), 
+    DisableStatCache(DisableStatCache), 
+    CurrentGeneration(0), NumStatHits(0), NumStatMisses(0), 
     NumSLocEntriesRead(0), TotalNumSLocEntries(0), 
     NumStatementsRead(0), TotalNumStatements(0), NumMacrosRead(0), 
     TotalNumMacros(0), NumSelectorsRead(0), NumMethodPoolEntriesRead(0), 

clang/lib/Serialization/Module.cpp

@@ -20,8 +20,8 @@ using namespace clang;
 using namespace serialization;
 using namespace reader;
 
-ModuleFile::ModuleFile(ModuleKind Kind)
-  : Kind(Kind), DirectlyImported(false), SizeInBits(0), 
+ModuleFile::ModuleFile(ModuleKind Kind, unsigned Generation)
+  : Kind(Kind), DirectlyImported(false), Generation(Generation), SizeInBits(0), 
     LocalNumSLocEntries(0), SLocEntryBaseID(0),
     SLocEntryBaseOffset(0), SLocEntryOffsets(0),
     SLocFileOffsets(0), LocalNumIdentifiers(0), 

clang/lib/Serialization/ModuleManager.cpp

@@ -35,7 +35,8 @@ llvm::MemoryBuffer *ModuleManager::lookupBuffer(StringRef Name) {
 
 std::pair<ModuleFile *, bool>
 ModuleManager::addModule(StringRef FileName, ModuleKind Type, 
-                         ModuleFile *ImportedBy, std::string &ErrorStr) {
+                         ModuleFile *ImportedBy, unsigned Generation,
+                         std::string &ErrorStr) {
   const FileEntry *Entry = FileMgr.getFile(FileName);
   if (!Entry && FileName != "-") {
     ErrorStr = "file not found";
@@ -47,7 +48,7 @@ ModuleManager::addModule(StringRef FileName, ModuleKind Type,
   bool NewModule = false;
   if (!ModuleEntry) {
     // Allocate a new module.
-    ModuleFile *New = new ModuleFile(Type);
+    ModuleFile *New = new ModuleFile(Type, Generation);
     New->FileName = FileName.str();
     Chain.push_back(New);
     NewModule = true;