X-Git-Url: http://demsky.eecs.uci.edu/git/?a=blobdiff_plain;f=lib%2FBitcode%2FWriter%2FBitcodeWriter.cpp;h=279e447873cab76fc90b4097f95d852e64366816;hb=e2e9d8e6aa914b81d79ff496736165f3358c6f3a;hp=7a875a06769dc903b349b891ca711a34d0b96ad1;hpb=f6f9cd1bf2e50244273cfdaf10573f10a545cdd5;p=oota-llvm.git diff --git a/lib/Bitcode/Writer/BitcodeWriter.cpp b/lib/Bitcode/Writer/BitcodeWriter.cpp index 7a875a06769..279e447873c 100644 --- a/lib/Bitcode/Writer/BitcodeWriter.cpp +++ b/lib/Bitcode/Writer/BitcodeWriter.cpp @@ -2,8 +2,8 @@ // // The LLVM Compiler Infrastructure // -// This file was developed by Chris Lattner and is distributed under -// the University of Illinois Open Source License. See LICENSE.TXT for details. +// This file is distributed under the University of Illinois Open Source +// License. See LICENSE.TXT for details. // //===----------------------------------------------------------------------===// // @@ -17,15 +17,43 @@ #include "ValueEnumerator.h" #include "llvm/Constants.h" #include "llvm/DerivedTypes.h" +#include "llvm/InlineAsm.h" #include "llvm/Instructions.h" #include "llvm/Module.h" -#include "llvm/ParameterAttributes.h" #include "llvm/TypeSymbolTable.h" #include "llvm/ValueSymbolTable.h" #include "llvm/Support/MathExtras.h" +#include "llvm/Support/Streams.h" +#include "llvm/Support/raw_ostream.h" +#include "llvm/System/Program.h" using namespace llvm; -static const unsigned CurVersion = 0; +/// These are manifest constants used by the bitcode writer. They do not need to +/// be kept in sync with the reader, but need to be consistent within this file. +enum { + CurVersion = 0, + + // VALUE_SYMTAB_BLOCK abbrev id's. + VST_ENTRY_8_ABBREV = bitc::FIRST_APPLICATION_ABBREV, + VST_ENTRY_7_ABBREV, + VST_ENTRY_6_ABBREV, + VST_BBENTRY_6_ABBREV, + + // CONSTANTS_BLOCK abbrev id's. + CONSTANTS_SETTYPE_ABBREV = bitc::FIRST_APPLICATION_ABBREV, + CONSTANTS_INTEGER_ABBREV, + CONSTANTS_CE_CAST_Abbrev, + CONSTANTS_NULL_Abbrev, + + // FUNCTION_BLOCK abbrev id's. + FUNCTION_INST_LOAD_ABBREV = bitc::FIRST_APPLICATION_ABBREV, + FUNCTION_INST_BINOP_ABBREV, + FUNCTION_INST_CAST_ABBREV, + FUNCTION_INST_RET_VOID_ABBREV, + FUNCTION_INST_RET_VAL_ABBREV, + FUNCTION_INST_UNREACHABLE_ABBREV +}; + static unsigned GetEncodedCastOpcode(unsigned Opcode) { switch (Opcode) { @@ -72,8 +100,7 @@ static void WriteStringRecord(unsigned Code, const std::string &Str, unsigned AbbrevToUse, BitstreamWriter &Stream) { SmallVector Vals; - // Code: [strlen, strchar x N] - Vals.push_back(Str.size()); + // Code: [strchar x N] for (unsigned i = 0, e = Str.size(); i != e; ++i) Vals.push_back(Str[i]); @@ -82,19 +109,20 @@ static void WriteStringRecord(unsigned Code, const std::string &Str, } // Emit information about parameter attributes. -static void WriteParamAttrTable(const ValueEnumerator &VE, +static void WriteAttributeTable(const ValueEnumerator &VE, BitstreamWriter &Stream) { - const std::vector &Attrs = VE.getParamAttrs(); + const std::vector &Attrs = VE.getAttributes(); if (Attrs.empty()) return; Stream.EnterSubblock(bitc::PARAMATTR_BLOCK_ID, 3); SmallVector Record; for (unsigned i = 0, e = Attrs.size(); i != e; ++i) { - const ParamAttrsList *A = Attrs[i]; - for (unsigned op = 0, e = A->size(); op != e; ++op) { - Record.push_back(A->getParamIndex(op)); - Record.push_back(A->getParamAttrsAtIndex(op)); + const AttrListPtr &A = Attrs[i]; + for (unsigned i = 0, e = A.getNumSlots(); i != e; ++i) { + const AttributeWithIndex &PAWI = A.getSlot(i); + Record.push_back(PAWI.Index); + Record.push_back(PAWI.Attrs); } Stream.EmitRecord(bitc::PARAMATTR_CODE_ENTRY, Record); @@ -111,7 +139,40 @@ static void WriteTypeTable(const ValueEnumerator &VE, BitstreamWriter &Stream) { Stream.EnterSubblock(bitc::TYPE_BLOCK_ID, 4 /*count from # abbrevs */); SmallVector TypeVals; - // FIXME: Set up abbrevs now that we know the width of the type fields, etc. + // Abbrev for TYPE_CODE_POINTER. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::TYPE_CODE_POINTER)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, + Log2_32_Ceil(VE.getTypes().size()+1))); + Abbv->Add(BitCodeAbbrevOp(0)); // Addrspace = 0 + unsigned PtrAbbrev = Stream.EmitAbbrev(Abbv); + + // Abbrev for TYPE_CODE_FUNCTION. + Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::TYPE_CODE_FUNCTION)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); // isvararg + Abbv->Add(BitCodeAbbrevOp(0)); // FIXME: DEAD value, remove in LLVM 3.0 + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, + Log2_32_Ceil(VE.getTypes().size()+1))); + unsigned FunctionAbbrev = Stream.EmitAbbrev(Abbv); + + // Abbrev for TYPE_CODE_STRUCT. + Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::TYPE_CODE_STRUCT)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); // ispacked + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, + Log2_32_Ceil(VE.getTypes().size()+1))); + unsigned StructAbbrev = Stream.EmitAbbrev(Abbv); + + // Abbrev for TYPE_CODE_ARRAY. + Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::TYPE_CODE_ARRAY)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); // size + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, + Log2_32_Ceil(VE.getTypes().size()+1))); + unsigned ArrayAbbrev = Stream.EmitAbbrev(Abbv); // Emit an entry count so the reader can reserve space. TypeVals.push_back(TypeList.size()); @@ -125,11 +186,13 @@ static void WriteTypeTable(const ValueEnumerator &VE, BitstreamWriter &Stream) { unsigned Code = 0; switch (T->getTypeID()) { - case Type::PackedStructTyID: // FIXME: Delete Type::PackedStructTyID. default: assert(0 && "Unknown type!"); case Type::VoidTyID: Code = bitc::TYPE_CODE_VOID; break; case Type::FloatTyID: Code = bitc::TYPE_CODE_FLOAT; break; case Type::DoubleTyID: Code = bitc::TYPE_CODE_DOUBLE; break; + case Type::X86_FP80TyID: Code = bitc::TYPE_CODE_X86_FP80; break; + case Type::FP128TyID: Code = bitc::TYPE_CODE_FP128; break; + case Type::PPC_FP128TyID: Code = bitc::TYPE_CODE_PPC_FP128; break; case Type::LabelTyID: Code = bitc::TYPE_CODE_LABEL; break; case Type::OpaqueTyID: Code = bitc::TYPE_CODE_OPAQUE; break; case Type::IntegerTyID: @@ -137,34 +200,38 @@ static void WriteTypeTable(const ValueEnumerator &VE, BitstreamWriter &Stream) { Code = bitc::TYPE_CODE_INTEGER; TypeVals.push_back(cast(T)->getBitWidth()); break; - case Type::PointerTyID: - // POINTER: [pointee type] + case Type::PointerTyID: { + const PointerType *PTy = cast(T); + // POINTER: [pointee type, address space] Code = bitc::TYPE_CODE_POINTER; - TypeVals.push_back(VE.getTypeID(cast(T)->getElementType())); + TypeVals.push_back(VE.getTypeID(PTy->getElementType())); + unsigned AddressSpace = PTy->getAddressSpace(); + TypeVals.push_back(AddressSpace); + if (AddressSpace == 0) AbbrevToUse = PtrAbbrev; break; - + } case Type::FunctionTyID: { const FunctionType *FT = cast(T); - // FUNCTION: [isvararg, #pararms, paramty x N] + // FUNCTION: [isvararg, attrid, retty, paramty x N] Code = bitc::TYPE_CODE_FUNCTION; TypeVals.push_back(FT->isVarArg()); + TypeVals.push_back(0); // FIXME: DEAD: remove in llvm 3.0 TypeVals.push_back(VE.getTypeID(FT->getReturnType())); - // FIXME: PARAM ATTR ID! - TypeVals.push_back(FT->getNumParams()); for (unsigned i = 0, e = FT->getNumParams(); i != e; ++i) TypeVals.push_back(VE.getTypeID(FT->getParamType(i))); + AbbrevToUse = FunctionAbbrev; break; } case Type::StructTyID: { const StructType *ST = cast(T); - // STRUCT: [ispacked, #elts, eltty x N] + // STRUCT: [ispacked, eltty x N] Code = bitc::TYPE_CODE_STRUCT; TypeVals.push_back(ST->isPacked()); - TypeVals.push_back(ST->getNumElements()); - // Output all of the element types... + // Output all of the element types. for (StructType::element_iterator I = ST->element_begin(), E = ST->element_end(); I != E; ++I) TypeVals.push_back(VE.getTypeID(*I)); + AbbrevToUse = StructAbbrev; break; } case Type::ArrayTyID: { @@ -173,6 +240,7 @@ static void WriteTypeTable(const ValueEnumerator &VE, BitstreamWriter &Stream) { Code = bitc::TYPE_CODE_ARRAY; TypeVals.push_back(AT->getNumElements()); TypeVals.push_back(VE.getTypeID(AT->getElementType())); + AbbrevToUse = ArrayAbbrev; break; } case Type::VectorTyID: { @@ -196,6 +264,7 @@ static void WriteTypeTable(const ValueEnumerator &VE, BitstreamWriter &Stream) { static unsigned getEncodedLinkage(const GlobalValue *GV) { switch (GV->getLinkage()) { default: assert(0 && "Invalid linkage!"); + case GlobalValue::GhostLinkage: // Map ghost linkage onto external. case GlobalValue::ExternalLinkage: return 0; case GlobalValue::WeakLinkage: return 1; case GlobalValue::AppendingLinkage: return 2; @@ -204,6 +273,7 @@ static unsigned getEncodedLinkage(const GlobalValue *GV) { case GlobalValue::DLLImportLinkage: return 5; case GlobalValue::DLLExportLinkage: return 6; case GlobalValue::ExternalWeakLinkage: return 7; + case GlobalValue::CommonLinkage: return 8; } } @@ -235,9 +305,10 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, WriteStringRecord(bitc::MODULE_CODE_ASM, M->getModuleInlineAsm(), 0/*TODO*/, Stream); - // Emit information about sections, computing how many there are. Also + // Emit information about sections and GC, computing how many there are. Also // compute the maximum alignment value. std::map SectionMap; + std::map GCMap; unsigned MaxAlignment = 0; unsigned MaxGlobalType = 0; for (Module::const_global_iterator GV = M->global_begin(),E = M->global_end(); @@ -255,13 +326,24 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, } for (Module::const_iterator F = M->begin(), E = M->end(); F != E; ++F) { MaxAlignment = std::max(MaxAlignment, F->getAlignment()); - if (!F->hasSection()) continue; - // Give section names unique ID's. - unsigned &Entry = SectionMap[F->getSection()]; - if (Entry != 0) continue; - WriteStringRecord(bitc::MODULE_CODE_SECTIONNAME, F->getSection(), - 0/*TODO*/, Stream); - Entry = SectionMap.size(); + if (F->hasSection()) { + // Give section names unique ID's. + unsigned &Entry = SectionMap[F->getSection()]; + if (!Entry) { + WriteStringRecord(bitc::MODULE_CODE_SECTIONNAME, F->getSection(), + 0/*TODO*/, Stream); + Entry = SectionMap.size(); + } + } + if (F->hasGC()) { + // Same for GC names. + unsigned &Entry = GCMap[F->getGC()]; + if (!Entry) { + WriteStringRecord(bitc::MODULE_CODE_GCNAME, F->getGC(), + 0/*TODO*/, Stream); + Entry = GCMap.size(); + } + } } // Emit abbrev for globals, now that we know # sections and max alignment. @@ -270,22 +352,22 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, // Add an abbrev for common globals with no visibility or thread localness. BitCodeAbbrev *Abbv = new BitCodeAbbrev(); Abbv->Add(BitCodeAbbrevOp(bitc::MODULE_CODE_GLOBALVAR)); - Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::FixedWidth, + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, Log2_32_Ceil(MaxGlobalType+1))); - Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::FixedWidth, 1)); // Constant. + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); // Constant. Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // Initializer. - Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::FixedWidth, 3)); // Linkage. - if (MaxAlignment == 0) // Alignment. + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 4)); // Linkage. + if (MaxAlignment == 0) // Alignment. Abbv->Add(BitCodeAbbrevOp(0)); else { unsigned MaxEncAlignment = Log2_32(MaxAlignment)+1; - Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::FixedWidth, + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, Log2_32_Ceil(MaxEncAlignment+1))); } if (SectionMap.empty()) // Section. Abbv->Add(BitCodeAbbrevOp(0)); else - Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::FixedWidth, + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, Log2_32_Ceil(SectionMap.size()+1))); // Don't bother emitting vis + thread local. SimpleGVarAbbrev = Stream.EmitAbbrev(Abbv); @@ -320,15 +402,17 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, // Emit the function proto information. for (Module::const_iterator F = M->begin(), E = M->end(); F != E; ++F) { - // FUNCTION: [type, callingconv, isproto, linkage, alignment, section, - // visibility] + // FUNCTION: [type, callingconv, isproto, paramattr, + // linkage, alignment, section, visibility, gc] Vals.push_back(VE.getTypeID(F->getType())); Vals.push_back(F->getCallingConv()); Vals.push_back(F->isDeclaration()); Vals.push_back(getEncodedLinkage(F)); + Vals.push_back(VE.getAttributeID(F->getAttributes())); Vals.push_back(Log2_32(F->getAlignment())+1); Vals.push_back(F->hasSection() ? SectionMap[F->getSection()] : 0); Vals.push_back(getEncodedVisibility(F)); + Vals.push_back(F->hasGC() ? GCMap[F->getGC()] : 0); unsigned AbbrevToUse = 0; Stream.EmitRecord(bitc::MODULE_CODE_FUNCTION, Vals, AbbrevToUse); @@ -342,6 +426,7 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, Vals.push_back(VE.getTypeID(AI->getType())); Vals.push_back(VE.getValueID(AI->getAliasee())); Vals.push_back(getEncodedLinkage(AI)); + Vals.push_back(getEncodedVisibility(AI)); unsigned AbbrevToUse = 0; Stream.EmitRecord(bitc::MODULE_CODE_ALIAS, Vals, AbbrevToUse); Vals.clear(); @@ -351,13 +436,43 @@ static void WriteModuleInfo(const Module *M, const ValueEnumerator &VE, static void WriteConstants(unsigned FirstVal, unsigned LastVal, const ValueEnumerator &VE, - BitstreamWriter &Stream) { + BitstreamWriter &Stream, bool isGlobal) { if (FirstVal == LastVal) return; - Stream.EnterSubblock(bitc::CONSTANTS_BLOCK_ID, 2); + Stream.EnterSubblock(bitc::CONSTANTS_BLOCK_ID, 4); + + unsigned AggregateAbbrev = 0; + unsigned String8Abbrev = 0; + unsigned CString7Abbrev = 0; + unsigned CString6Abbrev = 0; + // If this is a constant pool for the module, emit module-specific abbrevs. + if (isGlobal) { + // Abbrev for CST_CODE_AGGREGATE. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_AGGREGATE)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, Log2_32_Ceil(LastVal+1))); + AggregateAbbrev = Stream.EmitAbbrev(Abbv); - // FIXME: Install and use abbrevs to reduce size. Install them globally so - // they don't need to be reemitted for each function body. + // Abbrev for CST_CODE_STRING. + Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_STRING)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 8)); + String8Abbrev = Stream.EmitAbbrev(Abbv); + // Abbrev for CST_CODE_CSTRING. + Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_CSTRING)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 7)); + CString7Abbrev = Stream.EmitAbbrev(Abbv); + // Abbrev for CST_CODE_CSTRING. + Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_CSTRING)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Char6)); + CString6Abbrev = Stream.EmitAbbrev(Abbv); + } SmallVector Record; @@ -369,12 +484,27 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, if (V->getType() != LastTy) { LastTy = V->getType(); Record.push_back(VE.getTypeID(LastTy)); - Stream.EmitRecord(bitc::CST_CODE_SETTYPE, Record); + Stream.EmitRecord(bitc::CST_CODE_SETTYPE, Record, + CONSTANTS_SETTYPE_ABBREV); Record.clear(); } if (const InlineAsm *IA = dyn_cast(V)) { - assert(0 && IA && "FIXME: Inline asm writing unimp!"); + Record.push_back(unsigned(IA->hasSideEffects())); + + // Add the asm string. + const std::string &AsmStr = IA->getAsmString(); + Record.push_back(AsmStr.size()); + for (unsigned i = 0, e = AsmStr.size(); i != e; ++i) + Record.push_back(AsmStr[i]); + + // Add the constraint string. + const std::string &ConstraintStr = IA->getConstraintString(); + Record.push_back(ConstraintStr.size()); + for (unsigned i = 0, e = ConstraintStr.size(); i != e; ++i) + Record.push_back(ConstraintStr[i]); + Stream.EmitRecord(bitc::CST_CODE_INLINEASM, Record); + Record.clear(); continue; } const Constant *C = cast(V); @@ -392,6 +522,7 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, else Record.push_back((-V << 1) | 1); Code = bitc::CST_CODE_INTEGER; + AbbrevToUse = CONSTANTS_INTEGER_ABBREV; } else { // Wide integers, > 64 bits in size. // We have an arbitrary precision integer value to write whose // bit width is > 64. However, in canonical unsigned integer @@ -399,7 +530,6 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, // So, we only write the number of active words. unsigned NWords = IV->getValue().getActiveWords(); const uint64_t *RawWords = IV->getValue().getRawData(); - Record.push_back(NWords); for (unsigned i = 0; i != NWords; ++i) { int64_t V = RawWords[i]; if (V >= 0) @@ -411,18 +541,54 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, } } else if (const ConstantFP *CFP = dyn_cast(C)) { Code = bitc::CST_CODE_FLOAT; - if (CFP->getType() == Type::FloatTy) { - Record.push_back(FloatToBits((float)CFP->getValue())); + const Type *Ty = CFP->getType(); + if (Ty == Type::FloatTy || Ty == Type::DoubleTy) { + Record.push_back(CFP->getValueAPF().bitcastToAPInt().getZExtValue()); + } else if (Ty == Type::X86_FP80Ty) { + // api needed to prevent premature destruction + APInt api = CFP->getValueAPF().bitcastToAPInt(); + const uint64_t *p = api.getRawData(); + Record.push_back(p[0]); + Record.push_back((uint16_t)p[1]); + } else if (Ty == Type::FP128Ty || Ty == Type::PPC_FP128Ty) { + APInt api = CFP->getValueAPF().bitcastToAPInt(); + const uint64_t *p = api.getRawData(); + Record.push_back(p[0]); + Record.push_back(p[1]); + } else { + assert (0 && "Unknown FP type!"); + } + } else if (isa(C) && cast(C)->isString()) { + // Emit constant strings specially. + unsigned NumOps = C->getNumOperands(); + // If this is a null-terminated string, use the denser CSTRING encoding. + if (C->getOperand(NumOps-1)->isNullValue()) { + Code = bitc::CST_CODE_CSTRING; + --NumOps; // Don't encode the null, which isn't allowed by char6. } else { - assert (CFP->getType() == Type::DoubleTy && "Unknown FP type!"); - Record.push_back(DoubleToBits((double)CFP->getValue())); + Code = bitc::CST_CODE_STRING; + AbbrevToUse = String8Abbrev; } + bool isCStr7 = Code == bitc::CST_CODE_CSTRING; + bool isCStrChar6 = Code == bitc::CST_CODE_CSTRING; + for (unsigned i = 0; i != NumOps; ++i) { + unsigned char V = cast(C->getOperand(i))->getZExtValue(); + Record.push_back(V); + isCStr7 &= (V & 128) == 0; + if (isCStrChar6) + isCStrChar6 = BitCodeAbbrevOp::isChar6(V); + } + + if (isCStrChar6) + AbbrevToUse = CString6Abbrev; + else if (isCStr7) + AbbrevToUse = CString7Abbrev; } else if (isa(C) || isa(V) || isa(V)) { Code = bitc::CST_CODE_AGGREGATE; - Record.push_back(C->getNumOperands()); for (unsigned i = 0, e = C->getNumOperands(); i != e; ++i) Record.push_back(VE.getValueID(C->getOperand(i))); + AbbrevToUse = AggregateAbbrev; } else if (const ConstantExpr *CE = dyn_cast(C)) { switch (CE->getOpcode()) { default: @@ -431,6 +597,7 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, Record.push_back(GetEncodedCastOpcode(CE->getOpcode())); Record.push_back(VE.getTypeID(C->getOperand(0)->getType())); Record.push_back(VE.getValueID(C->getOperand(0))); + AbbrevToUse = CONSTANTS_CE_CAST_Abbrev; } else { assert(CE->getNumOperands() == 2 && "Unknown constant expr!"); Code = bitc::CST_CODE_CE_BINOP; @@ -441,7 +608,6 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, break; case Instruction::GetElementPtr: Code = bitc::CST_CODE_CE_GEP; - Record.push_back(CE->getNumOperands()); for (unsigned i = 0, e = CE->getNumOperands(); i != e; ++i) { Record.push_back(VE.getTypeID(C->getOperand(i)->getType())); Record.push_back(VE.getValueID(C->getOperand(i))); @@ -473,7 +639,16 @@ static void WriteConstants(unsigned FirstVal, unsigned LastVal, break; case Instruction::ICmp: case Instruction::FCmp: - Code = bitc::CST_CODE_CE_CMP; + case Instruction::VICmp: + case Instruction::VFCmp: + if (isa(C->getOperand(0)->getType()) + && (CE->getOpcode() == Instruction::ICmp + || CE->getOpcode() == Instruction::FCmp)) { + // compare returning vector of Int1Ty + assert(0 && "Unsupported constant!"); + } else { + Code = bitc::CST_CODE_CE_CMP; + } Record.push_back(VE.getTypeID(C->getOperand(0)->getType())); Record.push_back(VE.getValueID(C->getOperand(0))); Record.push_back(VE.getValueID(C->getOperand(1))); @@ -498,15 +673,35 @@ static void WriteModuleConstants(const ValueEnumerator &VE, // We know globalvalues have been emitted by WriteModuleInfo. for (unsigned i = 0, e = Vals.size(); i != e; ++i) { if (!isa(Vals[i].first)) { - WriteConstants(i, Vals.size(), VE, Stream); + WriteConstants(i, Vals.size(), VE, Stream, true); return; } } } +/// PushValueAndType - The file has to encode both the value and type id for +/// many values, because we need to know what type to create for forward +/// references. However, most operands are not forward references, so this type +/// field is not needed. +/// +/// This function adds V's value ID to Vals. If the value ID is higher than the +/// instruction ID, then it is a forward reference, and it also includes the +/// type ID. +static bool PushValueAndType(Value *V, unsigned InstID, + SmallVector &Vals, + ValueEnumerator &VE) { + unsigned ValID = VE.getValueID(V); + Vals.push_back(ValID); + if (ValID >= InstID) { + Vals.push_back(VE.getTypeID(V->getType())); + return true; + } + return false; +} + /// WriteInstruction - Emit an instruction to the specified stream. -static void WriteInstruction(const Instruction &I, ValueEnumerator &VE, - BitstreamWriter &Stream, +static void WriteInstruction(const Instruction &I, unsigned InstID, + ValueEnumerator &VE, BitstreamWriter &Stream, SmallVector &Vals) { unsigned Code = 0; unsigned AbbrevToUse = 0; @@ -514,68 +709,94 @@ static void WriteInstruction(const Instruction &I, ValueEnumerator &VE, default: if (Instruction::isCast(I.getOpcode())) { Code = bitc::FUNC_CODE_INST_CAST; - Vals.push_back(GetEncodedCastOpcode(I.getOpcode())); + if (!PushValueAndType(I.getOperand(0), InstID, Vals, VE)) + AbbrevToUse = FUNCTION_INST_CAST_ABBREV; Vals.push_back(VE.getTypeID(I.getType())); - Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); + Vals.push_back(GetEncodedCastOpcode(I.getOpcode())); } else { assert(isa(I) && "Unknown instruction!"); Code = bitc::FUNC_CODE_INST_BINOP; - Vals.push_back(GetEncodedBinaryOpcode(I.getOpcode())); - Vals.push_back(VE.getTypeID(I.getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); + if (!PushValueAndType(I.getOperand(0), InstID, Vals, VE)) + AbbrevToUse = FUNCTION_INST_BINOP_ABBREV; Vals.push_back(VE.getValueID(I.getOperand(1))); + Vals.push_back(GetEncodedBinaryOpcode(I.getOpcode())); } break; case Instruction::GetElementPtr: Code = bitc::FUNC_CODE_INST_GEP; - for (unsigned i = 0, e = I.getNumOperands(); i != e; ++i) { - Vals.push_back(VE.getTypeID(I.getOperand(i)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(i))); - } + for (unsigned i = 0, e = I.getNumOperands(); i != e; ++i) + PushValueAndType(I.getOperand(i), InstID, Vals, VE); + break; + case Instruction::ExtractValue: { + Code = bitc::FUNC_CODE_INST_EXTRACTVAL; + PushValueAndType(I.getOperand(0), InstID, Vals, VE); + const ExtractValueInst *EVI = cast(&I); + for (const unsigned *i = EVI->idx_begin(), *e = EVI->idx_end(); i != e; ++i) + Vals.push_back(*i); break; + } + case Instruction::InsertValue: { + Code = bitc::FUNC_CODE_INST_INSERTVAL; + PushValueAndType(I.getOperand(0), InstID, Vals, VE); + PushValueAndType(I.getOperand(1), InstID, Vals, VE); + const InsertValueInst *IVI = cast(&I); + for (const unsigned *i = IVI->idx_begin(), *e = IVI->idx_end(); i != e; ++i) + Vals.push_back(*i); + break; + } case Instruction::Select: - Code = bitc::FUNC_CODE_INST_SELECT; - Vals.push_back(VE.getTypeID(I.getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); - Vals.push_back(VE.getValueID(I.getOperand(1))); + Code = bitc::FUNC_CODE_INST_VSELECT; + PushValueAndType(I.getOperand(1), InstID, Vals, VE); Vals.push_back(VE.getValueID(I.getOperand(2))); + PushValueAndType(I.getOperand(0), InstID, Vals, VE); break; case Instruction::ExtractElement: Code = bitc::FUNC_CODE_INST_EXTRACTELT; - Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); + PushValueAndType(I.getOperand(0), InstID, Vals, VE); Vals.push_back(VE.getValueID(I.getOperand(1))); break; case Instruction::InsertElement: Code = bitc::FUNC_CODE_INST_INSERTELT; - Vals.push_back(VE.getTypeID(I.getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); + PushValueAndType(I.getOperand(0), InstID, Vals, VE); Vals.push_back(VE.getValueID(I.getOperand(1))); Vals.push_back(VE.getValueID(I.getOperand(2))); break; case Instruction::ShuffleVector: Code = bitc::FUNC_CODE_INST_SHUFFLEVEC; - Vals.push_back(VE.getTypeID(I.getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); + PushValueAndType(I.getOperand(0), InstID, Vals, VE); Vals.push_back(VE.getValueID(I.getOperand(1))); Vals.push_back(VE.getValueID(I.getOperand(2))); break; case Instruction::ICmp: case Instruction::FCmp: - Code = bitc::FUNC_CODE_INST_CMP; - Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); + case Instruction::VICmp: + case Instruction::VFCmp: + if (I.getOpcode() == Instruction::ICmp + || I.getOpcode() == Instruction::FCmp) { + // compare returning Int1Ty or vector of Int1Ty + Code = bitc::FUNC_CODE_INST_CMP2; + } else { + Code = bitc::FUNC_CODE_INST_CMP; + } + PushValueAndType(I.getOperand(0), InstID, Vals, VE); Vals.push_back(VE.getValueID(I.getOperand(1))); Vals.push_back(cast(I).getPredicate()); break; - case Instruction::Ret: - Code = bitc::FUNC_CODE_INST_RET; - if (I.getNumOperands()) { - Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); + case Instruction::Ret: + { + Code = bitc::FUNC_CODE_INST_RET; + unsigned NumOperands = I.getNumOperands(); + if (NumOperands == 0) + AbbrevToUse = FUNCTION_INST_RET_VOID_ABBREV; + else if (NumOperands == 1) { + if (!PushValueAndType(I.getOperand(0), InstID, Vals, VE)) + AbbrevToUse = FUNCTION_INST_RET_VAL_ABBREV; + } else { + for (unsigned i = 0, e = NumOperands; i != e; ++i) + PushValueAndType(I.getOperand(i), InstID, Vals, VE); + } } break; case Instruction::Br: @@ -593,28 +814,26 @@ static void WriteInstruction(const Instruction &I, ValueEnumerator &VE, Vals.push_back(VE.getValueID(I.getOperand(i))); break; case Instruction::Invoke: { + const PointerType *PTy = cast(I.getOperand(0)->getType()); + const FunctionType *FTy = cast(PTy->getElementType()); Code = bitc::FUNC_CODE_INST_INVOKE; - Vals.push_back(cast(I).getCallingConv()); - Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); // callee - Vals.push_back(VE.getValueID(I.getOperand(1))); // normal - Vals.push_back(VE.getValueID(I.getOperand(2))); // unwind + + const InvokeInst *II = cast(&I); + Vals.push_back(VE.getAttributeID(II->getAttributes())); + Vals.push_back(II->getCallingConv()); + Vals.push_back(VE.getValueID(I.getOperand(1))); // normal dest + Vals.push_back(VE.getValueID(I.getOperand(2))); // unwind dest + PushValueAndType(I.getOperand(0), InstID, Vals, VE); // callee // Emit value #'s for the fixed parameters. - const PointerType *PTy = cast(I.getOperand(0)->getType()); - const FunctionType *FTy = cast(PTy->getElementType()); for (unsigned i = 0, e = FTy->getNumParams(); i != e; ++i) Vals.push_back(VE.getValueID(I.getOperand(i+3))); // fixed param. // Emit type/value pairs for varargs params. if (FTy->isVarArg()) { - unsigned NumVarargs = I.getNumOperands()-3-FTy->getNumParams(); - Vals.push_back(NumVarargs); - for (unsigned i = I.getNumOperands()-NumVarargs, e = I.getNumOperands(); - i != e; ++i) { - Vals.push_back(VE.getTypeID(I.getOperand(i)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(i))); - } + for (unsigned i = 3+FTy->getNumParams(), e = I.getNumOperands(); + i != e; ++i) + PushValueAndType(I.getOperand(i), InstID, Vals, VE); // vararg } break; } @@ -623,12 +842,12 @@ static void WriteInstruction(const Instruction &I, ValueEnumerator &VE, break; case Instruction::Unreachable: Code = bitc::FUNC_CODE_INST_UNREACHABLE; + AbbrevToUse = FUNCTION_INST_UNREACHABLE_ABBREV; break; case Instruction::PHI: Code = bitc::FUNC_CODE_INST_PHI; Vals.push_back(VE.getTypeID(I.getType())); - Vals.push_back(I.getNumOperands()); for (unsigned i = 0, e = I.getNumOperands(); i != e; ++i) Vals.push_back(VE.getValueID(I.getOperand(i))); break; @@ -642,8 +861,7 @@ static void WriteInstruction(const Instruction &I, ValueEnumerator &VE, case Instruction::Free: Code = bitc::FUNC_CODE_INST_FREE; - Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); + PushValueAndType(I.getOperand(0), InstID, Vals, VE); break; case Instruction::Alloca: @@ -655,29 +873,31 @@ static void WriteInstruction(const Instruction &I, ValueEnumerator &VE, case Instruction::Load: Code = bitc::FUNC_CODE_INST_LOAD; - Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); // ptr. + if (!PushValueAndType(I.getOperand(0), InstID, Vals, VE)) // ptr + AbbrevToUse = FUNCTION_INST_LOAD_ABBREV; + Vals.push_back(Log2_32(cast(I).getAlignment())+1); Vals.push_back(cast(I).isVolatile()); break; case Instruction::Store: - Code = bitc::FUNC_CODE_INST_STORE; - Vals.push_back(VE.getTypeID(I.getOperand(1)->getType())); // Pointer - Vals.push_back(VE.getValueID(I.getOperand(0))); // val. - Vals.push_back(VE.getValueID(I.getOperand(1))); // ptr. + Code = bitc::FUNC_CODE_INST_STORE2; + PushValueAndType(I.getOperand(1), InstID, Vals, VE); // ptrty + ptr + Vals.push_back(VE.getValueID(I.getOperand(0))); // val. Vals.push_back(Log2_32(cast(I).getAlignment())+1); Vals.push_back(cast(I).isVolatile()); break; case Instruction::Call: { + const PointerType *PTy = cast(I.getOperand(0)->getType()); + const FunctionType *FTy = cast(PTy->getElementType()); + Code = bitc::FUNC_CODE_INST_CALL; - Vals.push_back((cast(I).getCallingConv() << 1) | - cast(I).isTailCall()); - Vals.push_back(VE.getTypeID(I.getOperand(0)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(0))); // callee + + const CallInst *CI = cast(&I); + Vals.push_back(VE.getAttributeID(CI->getAttributes())); + Vals.push_back((CI->getCallingConv() << 1) | unsigned(CI->isTailCall())); + PushValueAndType(CI->getOperand(0), InstID, Vals, VE); // Callee // Emit value #'s for the fixed parameters. - const PointerType *PTy = cast(I.getOperand(0)->getType()); - const FunctionType *FTy = cast(PTy->getElementType()); for (unsigned i = 0, e = FTy->getNumParams(); i != e; ++i) Vals.push_back(VE.getValueID(I.getOperand(i+1))); // fixed param. @@ -685,10 +905,8 @@ static void WriteInstruction(const Instruction &I, ValueEnumerator &VE, if (FTy->isVarArg()) { unsigned NumVarargs = I.getNumOperands()-1-FTy->getNumParams(); for (unsigned i = I.getNumOperands()-NumVarargs, e = I.getNumOperands(); - i != e; ++i) { - Vals.push_back(VE.getTypeID(I.getOperand(i)->getType())); - Vals.push_back(VE.getValueID(I.getOperand(i))); - } + i != e; ++i) + PushValueAndType(I.getOperand(i), InstID, Vals, VE); // varargs } break; } @@ -709,29 +927,50 @@ static void WriteValueSymbolTable(const ValueSymbolTable &VST, const ValueEnumerator &VE, BitstreamWriter &Stream) { if (VST.empty()) return; - Stream.EnterSubblock(bitc::VALUE_SYMTAB_BLOCK_ID, 3); - + Stream.EnterSubblock(bitc::VALUE_SYMTAB_BLOCK_ID, 4); + // FIXME: Set up the abbrev, we know how many values there are! // FIXME: We know if the type names can use 7-bit ascii. SmallVector NameVals; for (ValueSymbolTable::const_iterator SI = VST.begin(), SE = VST.end(); SI != SE; ++SI) { - unsigned AbbrevToUse = 0; - // VST_ENTRY: [valueid, namelen, namechar x N] - // VST_BBENTRY: [bbid, namelen, namechar x N] + const ValueName &Name = *SI; + + // Figure out the encoding to use for the name. + bool is7Bit = true; + bool isChar6 = true; + for (const char *C = Name.getKeyData(), *E = C+Name.getKeyLength(); + C != E; ++C) { + if (isChar6) + isChar6 = BitCodeAbbrevOp::isChar6(*C); + if ((unsigned char)*C & 128) { + is7Bit = false; + break; // don't bother scanning the rest. + } + } + + unsigned AbbrevToUse = VST_ENTRY_8_ABBREV; + + // VST_ENTRY: [valueid, namechar x N] + // VST_BBENTRY: [bbid, namechar x N] unsigned Code; if (isa(SI->getValue())) { Code = bitc::VST_CODE_BBENTRY; + if (isChar6) + AbbrevToUse = VST_BBENTRY_6_ABBREV; } else { Code = bitc::VST_CODE_ENTRY; + if (isChar6) + AbbrevToUse = VST_ENTRY_6_ABBREV; + else if (is7Bit) + AbbrevToUse = VST_ENTRY_7_ABBREV; } NameVals.push_back(VE.getValueID(SI->getValue())); - NameVals.push_back(SI->getKeyLength()); - for (const char *P = SI->getKeyData(), - *E = SI->getKeyData()+SI->getKeyLength(); P != E; ++P) + for (const char *P = Name.getKeyData(), + *E = Name.getKeyData()+Name.getKeyLength(); P != E; ++P) NameVals.push_back((unsigned char)*P); // Emit the finished record. @@ -744,7 +983,7 @@ static void WriteValueSymbolTable(const ValueSymbolTable &VST, /// WriteFunction - Emit a function body to the module stream. static void WriteFunction(const Function &F, ValueEnumerator &VE, BitstreamWriter &Stream) { - Stream.EnterSubblock(bitc::FUNCTION_BLOCK_ID, 3); + Stream.EnterSubblock(bitc::FUNCTION_BLOCK_ID, 4); VE.incorporateFunction(F); SmallVector Vals; @@ -755,17 +994,22 @@ static void WriteFunction(const Function &F, ValueEnumerator &VE, Stream.EmitRecord(bitc::FUNC_CODE_DECLAREBLOCKS, Vals); Vals.clear(); - // FIXME: Function attributes? - // If there are function-local constants, emit them now. unsigned CstStart, CstEnd; VE.getFunctionConstantRange(CstStart, CstEnd); - WriteConstants(CstStart, CstEnd, VE, Stream); + WriteConstants(CstStart, CstEnd, VE, Stream, false); + + // Keep a running idea of what the instruction ID is. + unsigned InstID = CstEnd; // Finally, emit all the instructions, in order. for (Function::const_iterator BB = F.begin(), E = F.end(); BB != E; ++BB) - for (BasicBlock::const_iterator I = BB->begin(), E = BB->end(); I != E; ++I) - WriteInstruction(*I, VE, Stream, Vals); + for (BasicBlock::const_iterator I = BB->begin(), E = BB->end(); + I != E; ++I) { + WriteInstruction(*I, InstID, VE, Stream, Vals); + if (I->getType() != Type::VoidTy) + ++InstID; + } // Emit names for all the instructions etc. WriteValueSymbolTable(F.getValueSymbolTable(), VE, Stream); @@ -782,31 +1026,188 @@ static void WriteTypeSymbolTable(const TypeSymbolTable &TST, Stream.EnterSubblock(bitc::TYPE_SYMTAB_BLOCK_ID, 3); - // FIXME: Set up the abbrev, we know how many types there are! - // FIXME: We know if the type names can use 7-bit ascii. + // 7-bit fixed width VST_CODE_ENTRY strings. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::VST_CODE_ENTRY)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, + Log2_32_Ceil(VE.getTypes().size()+1))); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 7)); + unsigned V7Abbrev = Stream.EmitAbbrev(Abbv); SmallVector NameVals; for (TypeSymbolTable::const_iterator TI = TST.begin(), TE = TST.end(); TI != TE; ++TI) { - unsigned AbbrevToUse = 0; - - // TST_ENTRY: [typeid, namelen, namechar x N] + // TST_ENTRY: [typeid, namechar x N] NameVals.push_back(VE.getTypeID(TI->second)); const std::string &Str = TI->first; - NameVals.push_back(Str.size()); - for (unsigned i = 0, e = Str.size(); i != e; ++i) - NameVals.push_back(Str[i]); + bool is7Bit = true; + for (unsigned i = 0, e = Str.size(); i != e; ++i) { + NameVals.push_back((unsigned char)Str[i]); + if (Str[i] & 128) + is7Bit = false; + } // Emit the finished record. - Stream.EmitRecord(bitc::VST_CODE_ENTRY, NameVals, AbbrevToUse); + Stream.EmitRecord(bitc::VST_CODE_ENTRY, NameVals, is7Bit ? V7Abbrev : 0); NameVals.clear(); } Stream.ExitBlock(); } +// Emit blockinfo, which defines the standard abbreviations etc. +static void WriteBlockInfo(const ValueEnumerator &VE, BitstreamWriter &Stream) { + // We only want to emit block info records for blocks that have multiple + // instances: CONSTANTS_BLOCK, FUNCTION_BLOCK and VALUE_SYMTAB_BLOCK. Other + // blocks can defined their abbrevs inline. + Stream.EnterBlockInfoBlock(2); + + { // 8-bit fixed-width VST_ENTRY/VST_BBENTRY strings. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 3)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 8)); + if (Stream.EmitBlockInfoAbbrev(bitc::VALUE_SYMTAB_BLOCK_ID, + Abbv) != VST_ENTRY_8_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + + { // 7-bit fixed width VST_ENTRY strings. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::VST_CODE_ENTRY)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 7)); + if (Stream.EmitBlockInfoAbbrev(bitc::VALUE_SYMTAB_BLOCK_ID, + Abbv) != VST_ENTRY_7_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + { // 6-bit char6 VST_ENTRY strings. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::VST_CODE_ENTRY)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Char6)); + if (Stream.EmitBlockInfoAbbrev(bitc::VALUE_SYMTAB_BLOCK_ID, + Abbv) != VST_ENTRY_6_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + { // 6-bit char6 VST_BBENTRY strings. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::VST_CODE_BBENTRY)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Array)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Char6)); + if (Stream.EmitBlockInfoAbbrev(bitc::VALUE_SYMTAB_BLOCK_ID, + Abbv) != VST_BBENTRY_6_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + + + + { // SETTYPE abbrev for CONSTANTS_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_SETTYPE)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, + Log2_32_Ceil(VE.getTypes().size()+1))); + if (Stream.EmitBlockInfoAbbrev(bitc::CONSTANTS_BLOCK_ID, + Abbv) != CONSTANTS_SETTYPE_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + + { // INTEGER abbrev for CONSTANTS_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_INTEGER)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); + if (Stream.EmitBlockInfoAbbrev(bitc::CONSTANTS_BLOCK_ID, + Abbv) != CONSTANTS_INTEGER_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + + { // CE_CAST abbrev for CONSTANTS_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_CE_CAST)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 4)); // cast opc + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, // typeid + Log2_32_Ceil(VE.getTypes().size()+1))); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 8)); // value id + + if (Stream.EmitBlockInfoAbbrev(bitc::CONSTANTS_BLOCK_ID, + Abbv) != CONSTANTS_CE_CAST_Abbrev) + assert(0 && "Unexpected abbrev ordering!"); + } + { // NULL abbrev for CONSTANTS_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::CST_CODE_NULL)); + if (Stream.EmitBlockInfoAbbrev(bitc::CONSTANTS_BLOCK_ID, + Abbv) != CONSTANTS_NULL_Abbrev) + assert(0 && "Unexpected abbrev ordering!"); + } + + // FIXME: This should only use space for first class types! + + { // INST_LOAD abbrev for FUNCTION_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::FUNC_CODE_INST_LOAD)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // Ptr + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 4)); // Align + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 1)); // volatile + if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, + Abbv) != FUNCTION_INST_LOAD_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + { // INST_BINOP abbrev for FUNCTION_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::FUNC_CODE_INST_BINOP)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // LHS + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // RHS + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 4)); // opc + if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, + Abbv) != FUNCTION_INST_BINOP_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + { // INST_CAST abbrev for FUNCTION_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::FUNC_CODE_INST_CAST)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // OpVal + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, // dest ty + Log2_32_Ceil(VE.getTypes().size()+1))); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::Fixed, 4)); // opc + if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, + Abbv) != FUNCTION_INST_CAST_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + + { // INST_RET abbrev for FUNCTION_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::FUNC_CODE_INST_RET)); + if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, + Abbv) != FUNCTION_INST_RET_VOID_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + { // INST_RET abbrev for FUNCTION_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::FUNC_CODE_INST_RET)); + Abbv->Add(BitCodeAbbrevOp(BitCodeAbbrevOp::VBR, 6)); // ValID + if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, + Abbv) != FUNCTION_INST_RET_VAL_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + { // INST_UNREACHABLE abbrev for FUNCTION_BLOCK. + BitCodeAbbrev *Abbv = new BitCodeAbbrev(); + Abbv->Add(BitCodeAbbrevOp(bitc::FUNC_CODE_INST_UNREACHABLE)); + if (Stream.EmitBlockInfoAbbrev(bitc::FUNCTION_BLOCK_ID, + Abbv) != FUNCTION_INST_UNREACHABLE_ABBREV) + assert(0 && "Unexpected abbrev ordering!"); + } + + Stream.ExitBlock(); +} + /// WriteModule - Emit the specified module to the bitstream. static void WriteModule(const Module *M, BitstreamWriter &Stream) { @@ -821,9 +1222,12 @@ static void WriteModule(const Module *M, BitstreamWriter &Stream) { // Analyze the module, enumerating globals, functions, etc. ValueEnumerator VE(M); + + // Emit blockinfo, which defines the standard abbreviations etc. + WriteBlockInfo(VE, Stream); // Emit information about parameter attributes. - WriteParamAttrTable(VE, Stream); + WriteAttributeTable(VE, Stream); // Emit information describing all of the types in the module. WriteTypeTable(VE, Stream); @@ -859,14 +1263,94 @@ static void WriteModule(const Module *M, BitstreamWriter &Stream) { Stream.ExitBlock(); } +/// EmitDarwinBCHeader - If generating a bc file on darwin, we have to emit a +/// header and trailer to make it compatible with the system archiver. To do +/// this we emit the following header, and then emit a trailer that pads the +/// file out to be a multiple of 16 bytes. +/// +/// struct bc_header { +/// uint32_t Magic; // 0x0B17C0DE +/// uint32_t Version; // Version, currently always 0. +/// uint32_t BitcodeOffset; // Offset to traditional bitcode file. +/// uint32_t BitcodeSize; // Size of traditional bitcode file. +/// uint32_t CPUType; // CPU specifier. +/// ... potentially more later ... +/// }; +enum { + DarwinBCSizeFieldOffset = 3*4, // Offset to bitcode_size. + DarwinBCHeaderSize = 5*4 +}; + +static void EmitDarwinBCHeader(BitstreamWriter &Stream, + const std::string &TT) { + unsigned CPUType = ~0U; + + // Match x86_64-*, i[3-9]86-*, powerpc-*, powerpc64-*. The CPUType is a + // magic number from /usr/include/mach/machine.h. It is ok to reproduce the + // specific constants here because they are implicitly part of the Darwin ABI. + enum { + DARWIN_CPU_ARCH_ABI64 = 0x01000000, + DARWIN_CPU_TYPE_X86 = 7, + DARWIN_CPU_TYPE_POWERPC = 18 + }; + + if (TT.find("x86_64-") == 0) + CPUType = DARWIN_CPU_TYPE_X86 | DARWIN_CPU_ARCH_ABI64; + else if (TT.size() >= 5 && TT[0] == 'i' && TT[2] == '8' && TT[3] == '6' && + TT[4] == '-' && TT[1] - '3' < 6) + CPUType = DARWIN_CPU_TYPE_X86; + else if (TT.find("powerpc-") == 0) + CPUType = DARWIN_CPU_TYPE_POWERPC; + else if (TT.find("powerpc64-") == 0) + CPUType = DARWIN_CPU_TYPE_POWERPC | DARWIN_CPU_ARCH_ABI64; + + // Traditional Bitcode starts after header. + unsigned BCOffset = DarwinBCHeaderSize; + + Stream.Emit(0x0B17C0DE, 32); + Stream.Emit(0 , 32); // Version. + Stream.Emit(BCOffset , 32); + Stream.Emit(0 , 32); // Filled in later. + Stream.Emit(CPUType , 32); +} + +/// EmitDarwinBCTrailer - Emit the darwin epilog after the bitcode file and +/// finalize the header. +static void EmitDarwinBCTrailer(BitstreamWriter &Stream, unsigned BufferSize) { + // Update the size field in the header. + Stream.BackpatchWord(DarwinBCSizeFieldOffset, BufferSize-DarwinBCHeaderSize); + + // If the file is not a multiple of 16 bytes, insert dummy padding. + while (BufferSize & 15) { + Stream.Emit(0, 8); + ++BufferSize; + } +} + + /// WriteBitcodeToFile - Write the specified module to the specified output /// stream. void llvm::WriteBitcodeToFile(const Module *M, std::ostream &Out) { + raw_os_ostream RawOut(Out); + // If writing to stdout, set binary mode. + if (llvm::cout == Out) + sys::Program::ChangeStdoutToBinary(); + WriteBitcodeToFile(M, RawOut); +} + +/// WriteBitcodeToFile - Write the specified module to the specified output +/// stream. +void llvm::WriteBitcodeToFile(const Module *M, raw_ostream &Out) { std::vector Buffer; BitstreamWriter Stream(Buffer); Buffer.reserve(256*1024); + // If this is darwin, emit a file header and trailer if needed. + bool isDarwin = M->getTargetTriple().find("-darwin") != std::string::npos; + if (isDarwin) + EmitDarwinBCHeader(Stream, M->getTargetTriple()); + // Emit the file header. Stream.Emit((unsigned)'B', 8); Stream.Emit((unsigned)'C', 8); @@ -877,7 +1361,18 @@ void llvm::WriteBitcodeToFile(const Module *M, std::ostream &Out) { // Emit the module. WriteModule(M, Stream); + + if (isDarwin) + EmitDarwinBCTrailer(Stream, Buffer.size()); + + // If writing to stdout, set binary mode. + if (&llvm::outs() == &Out) + sys::Program::ChangeStdoutToBinary(); + // Write the generated bitstream to "Out". Out.write((char*)&Buffer.front(), Buffer.size()); + + // Make sure it hits disk now. + Out.flush(); }