From 85a0f87636c7107e5ebb34f57fb3444dc27ff52d Mon Sep 17 00:00:00 2001
From: Chris Lattner <sabre@nondot.org>
Date: Wed, 14 Jan 2004 23:35:21 +0000
Subject: [PATCH] Version 1.2 now supports encoding strings as a special case,
 to avoid having to emit all of those sbyte constants.

llvm-svn: 10863
---
 lib/Bytecode/Reader/ConstantReader.cpp | 39 ++++++++++++++++++++++++++
 lib/Bytecode/Reader/Reader.cpp         |  7 +++--
 lib/Bytecode/Reader/ReaderInternals.h  |  3 ++
 3 files changed, 47 insertions(+), 2 deletions(-)

diff --git a/lib/Bytecode/Reader/ConstantReader.cpp b/lib/Bytecode/Reader/ConstantReader.cpp
index 660566b2664..5fb62dae20c 100644
--- a/lib/Bytecode/Reader/ConstantReader.cpp
+++ b/lib/Bytecode/Reader/ConstantReader.cpp
@@ -298,6 +298,42 @@ void BytecodeParser::ParseGlobalTypes(const unsigned char *&Buf,
   ParseConstantPool(Buf, EndBuf, T, ModuleTypeValues);
 }
 
+void BytecodeParser::parseStringConstants(const unsigned char *&Buf,
+                                          const unsigned char *EndBuf,
+                                          unsigned NumEntries, ValueTable &Tab){
+  unsigned Typ;
+  for (; NumEntries; --NumEntries) {
+    if (read_vbr(Buf, EndBuf, Typ)) throw Error_readvbr;
+    const Type *Ty = getType(Typ);
+    if (!isa<ArrayType>(Ty))
+      throw std::string("String constant data invalid!");
+    
+    const ArrayType *ATy = cast<ArrayType>(Ty);
+    if (ATy->getElementType() != Type::SByteTy &&
+        ATy->getElementType() != Type::UByteTy)
+      throw std::string("String constant data invalid!");
+    
+    // Read character data.  The type tells us how long the string is.
+    char Data[ATy->getNumElements()];
+    if (input_data(Buf, EndBuf, Data, Data+ATy->getNumElements()))
+      throw Error_inputdata;
+
+    std::vector<Constant*> Elements(ATy->getNumElements());
+    if (ATy->getElementType() == Type::SByteTy)
+      for (unsigned i = 0, e = ATy->getNumElements(); i != e; ++i)
+        Elements[i] = ConstantSInt::get(Type::SByteTy, Data[i]);
+    else
+      for (unsigned i = 0, e = ATy->getNumElements(); i != e; ++i)
+        Elements[i] = ConstantSInt::get(Type::UByteTy, Data[i]);
+
+    // Create the constant, inserting it as needed.
+    Constant *C = ConstantArray::get(ATy, Elements);
+    unsigned Slot = insertValue(C, Typ, Tab);
+    ResolveReferencesToConstant(C, Slot);
+  }
+}
+
+
 void BytecodeParser::ParseConstantPool(const unsigned char *&Buf,
                                        const unsigned char *EndBuf,
                                        ValueTable &Tab, 
@@ -310,6 +346,9 @@ void BytecodeParser::ParseConstantPool(const unsigned char *&Buf,
     if (Typ == Type::TypeTyID) {
       BCR_TRACE(3, "Type: 'type'  NumEntries: " << NumEntries << "\n");
       parseTypeConstants(Buf, EndBuf, TypeTab, NumEntries);
+    } else if (Typ == Type::VoidTyID) {
+      assert(&Tab == &ModuleValues && "Cannot read strings in functions!");
+      parseStringConstants(Buf, EndBuf, NumEntries, Tab);
     } else {
       BCR_TRACE(3, "Type: '" << *getType(Typ) << "'  NumEntries: "
                 << NumEntries << "\n");
diff --git a/lib/Bytecode/Reader/Reader.cpp b/lib/Bytecode/Reader/Reader.cpp
index a0b40be6e39..7a6fa875114 100644
--- a/lib/Bytecode/Reader/Reader.cpp
+++ b/lib/Bytecode/Reader/Reader.cpp
@@ -552,8 +552,8 @@ void BytecodeParser::ParseVersionInfo(const unsigned char *&Buf,
 
   switch (RevisionNum) {
   case 2:               // LLVM pre-1.0 release: will be deleted on the next rev
-    // Version #2 added information about all 4 linkage types instead of just
-    // having internal and external.
+    // Version #2 only supported 4 linkage types.  It didn't support weak
+    // linkage.
     hasExtendedLinkageSpecs = false;
     hasOldStyleVarargs = true;
     hasVarArgCallPadding = true;
@@ -561,8 +561,11 @@ void BytecodeParser::ParseVersionInfo(const unsigned char *&Buf,
   case 0:               //  LLVM 1.0, 1.1 release version
     // Compared to rev #2, we added support for weak linkage, a more dense
     // encoding, and better varargs support.
+
+    // Base LLVM 1.0 bytecode format.
     break;
   case 1:               // LLVM 1.2 release version
+    // LLVM 1.2 added explicit support for emitting strings efficiently.
     break;
 
   default:
diff --git a/lib/Bytecode/Reader/ReaderInternals.h b/lib/Bytecode/Reader/ReaderInternals.h
index b1d9c605f78..54a38e457b4 100644
--- a/lib/Bytecode/Reader/ReaderInternals.h
+++ b/lib/Bytecode/Reader/ReaderInternals.h
@@ -177,6 +177,9 @@ private:
                           TypeValuesListTy &Tab, unsigned NumEntries);
   const Type *parseTypeConstant(const unsigned char *&Buf,
                                 const unsigned char *EndBuf);
+  void parseStringConstants(const unsigned char *&Buf,
+                            const unsigned char *EndBuf,
+                            unsigned NumEntries, ValueTable &Tab);
 
   Value      *getValue(unsigned TypeID, unsigned num, bool Create = true);
   const Type *getType(unsigned ID);