Verify the symbol table contents before and after all

GCs (not just mark-compacts) and make the mark-compact
shortcutting of cons-strings identical to the scavenge
version.
Review URL: http://codereview.chromium.org/67125

git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@1699 ce2b1a6d-e550-0410-aec6-3dcde31c8c00

src/heap.cc

@@ -374,9 +374,34 @@ void Heap::PerformScavenge() {
 }
 
 
+static void VerifySymbolTable() {
+#ifdef DEBUG
+  // Helper class for verifying the symbol table.
+  class SymbolTableVerifier : public ObjectVisitor {
+   public:
+    SymbolTableVerifier() { }
+    void VisitPointers(Object** start, Object** end) {
+      // Visit all HeapObject pointers in [start, end).
+      for (Object** p = start; p < end; p++) {
+        if ((*p)->IsHeapObject()) {
+          // Check that the symbol is actually a symbol.
+          ASSERT((*p)->IsNull() || (*p)->IsUndefined() || (*p)->IsSymbol());
+        }
+      }
+    }
+  };
+
+  SymbolTableVerifier verifier;
+  SymbolTable* symbol_table = SymbolTable::cast(Heap::symbol_table());
+  symbol_table->IterateElements(&verifier);
+#endif  // DEBUG
+}
+
+
 void Heap::PerformGarbageCollection(AllocationSpace space,
                                     GarbageCollector collector,
                                     GCTracer* tracer) {
+  VerifySymbolTable();
   if (collector == MARK_COMPACTOR && global_gc_prologue_callback_) {
     ASSERT(!allocation_allowed_);
     global_gc_prologue_callback_();
@@ -421,6 +446,7 @@ void Heap::PerformGarbageCollection(AllocationSpace space,
     ASSERT(!allocation_allowed_);
     global_gc_epilogue_callback_();
   }
+  VerifySymbolTable();
 }
 
 
@@ -813,18 +839,9 @@ void Heap::ScavengeObject(HeapObject** p, HeapObject* object) {
 
 
 static inline bool IsShortcutCandidate(HeapObject* object, Map* map) {
-  // A ConsString with an empty string as the right side is a
-  // candidate for being shortcut by the scavenger unless it is a
-  // symbol. It's not common to have non-flat symbols, so we do not
-  // shortcut them thereby avoiding turning symbols into strings.
-  ASSERT(kNotStringTag != 0 && kSymbolTag != 0);
-  static const uint32_t kShortcutTypeMask =
-      kIsNotStringMask |
-      kIsSymbolMask |
-      kStringRepresentationMask;
   ASSERT(object->map() == map);
   InstanceType type = map->instance_type();
-  if ((type & kShortcutTypeMask) != kConsStringTag) return false;
+  if ((type & kShortcutTypeMask) != kShortcutTypeTag) return false;
   ASSERT(object->IsString() && !object->IsSymbol());
   return ConsString::cast(object)->unchecked_second() == Heap::empty_string();
 }

src/mark-compact.cc

@@ -96,24 +96,6 @@ void MarkCompactCollector::CollectGarbage() {
 }
 
 
-#ifdef DEBUG
-// Helper class for verifying the symbol table.
-class SymbolTableVerifier : public ObjectVisitor {
- public:
-  SymbolTableVerifier() { }
-  void VisitPointers(Object** start, Object** end) {
-    // Visit all HeapObject pointers in [start, end).
-    for (Object** p = start; p < end; p++) {
-      if ((*p)->IsHeapObject()) {
-        // Check that the symbol is actually a symbol.
-        ASSERT((*p)->IsNull() || (*p)->IsUndefined() || (*p)->IsSymbol());
-      }
-    }
-  }
-};
-#endif  // DEBUG
-
-
 void MarkCompactCollector::Prepare(GCTracer* tracer) {
   // Rather than passing the tracer around we stash it in a static member
   // variable.
@@ -166,10 +148,6 @@ void MarkCompactCollector::Prepare(GCTracer* tracer) {
   }
 
 #ifdef DEBUG
-  SymbolTable* symbol_table = SymbolTable::cast(Heap::symbol_table());
-  SymbolTableVerifier v;
-  symbol_table->IterateElements(&v);
-
   live_bytes_ = 0;
   live_young_objects_ = 0;
   live_old_pointer_objects_ = 0;
@@ -183,10 +161,6 @@ void MarkCompactCollector::Prepare(GCTracer* tracer) {
 
 void MarkCompactCollector::Finish() {
 #ifdef DEBUG
-  SymbolTable* symbol_table = SymbolTable::cast(Heap::symbol_table());
-  SymbolTableVerifier v;
-  symbol_table->IterateElements(&v);
-
   ASSERT(state_ == SWEEP_SPACES || state_ == REBUILD_RSETS);
   state_ = IDLE;
 #endif
@@ -246,13 +220,7 @@ static inline HeapObject* ShortCircuitConsString(Object** p) {
   MapWord map_word = object->map_word();
   map_word.ClearMark();
   InstanceType type = map_word.ToMap()->instance_type();
-  if (type >= FIRST_NONSTRING_TYPE || (type & kIsSymbolMask) != 0) {
-    return object;
-  }
-
-  StringRepresentationTag rep =
-      static_cast<StringRepresentationTag>(type & kStringRepresentationMask);
-  if (rep != kConsStringTag) return object;
+  if ((type & kShortcutTypeMask) != kShortcutTypeTag) return object;
 
   Object* second = reinterpret_cast<ConsString*>(object)->unchecked_second();
   if (reinterpret_cast<String*>(second) != Heap::empty_string()) return object;

src/objects.cc

@@ -679,7 +679,7 @@ bool String::MakeExternal(v8::String::ExternalStringResource* resource) {
     SmartPointer<uc16> smart_chars = this->ToWideCString();
     ASSERT(memcmp(*smart_chars,
                   resource->data(),
-                  resource->length()*sizeof(**smart_chars)) == 0);
+                  resource->length() * sizeof(**smart_chars)) == 0);
   }
 #endif  // DEBUG
 
@@ -6304,7 +6304,7 @@ class SymbolKey : public HashTableKey {
     if (StringShape(string_).IsCons()) {
       ConsString* cons_string = ConsString::cast(string_);
       cons_string->TryFlatten();
-      if (cons_string->second() == Heap::empty_string()) {
+      if (cons_string->second()->length() == 0) {
         string_ = cons_string->first();
       }
     }
@@ -6312,6 +6312,7 @@ class SymbolKey : public HashTableKey {
     Map* map = Heap::SymbolMapForString(string_);
     if (map != NULL) {
       string_->set_map(map);
+      ASSERT(string_->IsSymbol());
       return string_;
     }
     // Otherwise allocate a new symbol.

src/objects.h

@@ -441,6 +441,20 @@ enum StringRepresentationTag {
   kExternalStringTag = 0x3
 };
 
+
+// A ConsString with an empty string as the right side is a candidate
+// for being shortcut by the garbage collector unless it is a
+// symbol. It's not common to have non-flat symbols, so we do not
+// shortcut them thereby avoiding turning symbols into strings. See
+// heap.cc and mark-compact.cc.
+STATIC_ASSERT(kNotStringTag != 0 && kSymbolTag != 0);
+const uint32_t kShortcutTypeMask =
+    kIsNotStringMask |
+    kIsSymbolMask |
+    kStringRepresentationMask;
+const uint32_t kShortcutTypeTag = kConsStringTag;
+
+
 enum InstanceType {
   SHORT_SYMBOL_TYPE = kShortStringTag | kSymbolTag | kSeqStringTag,
   MEDIUM_SYMBOL_TYPE = kMediumStringTag | kSymbolTag | kSeqStringTag,